Full text of "Advances In Agronomy Vol II"

See other formats

IS SOU 168104

uj <*-"J_

OSMANIA UraVERSITY LIBRARY

OOlNo. 2>t>/ fltFlf /4 Acomioalo,

Author

Thi> book should be returned on or before the date last marked below.

ADVANCES IN AGRONOMY

VOLUME II

ADVANCES IN

AGRONOMY

Prepared under the Auspices of the

AMERICAN SOCIETY OF AGRONOMY

VOLUME II

Edited by A. G. NORMAN

Camp Detrick, Frederick, Maryland

ADVISORY BOARD

R. BRADFIELD K. S. QUISENBERRY

H. H. LAUDE L. A. RICHARDS

C. E. MARSHALL V. G. SPRAGUE

N. P. NEAL E. WINTERS

1950

ACADEMIC PRESS INC., PUBLISHERS
NEW YORK

125 EAST 23RD STREET
NEW YORK 10, N. Y.

No part of this book may be reproduced in any
form, by photostat, microfilm, or any other means t
without written permission from the publishers.

PRINTED IN THE UNITED STATES OP AMERICA

CONTRIBUTORS TO VOLUME II

J. E. ADAMS, Head, Department of Agronomy, Texas Agricultural and
Mechanical College System, College Station, Texas.

GILBERT H. AHLGREN, Professor of Farm Crops, Rutgers University, New
Brunswick, New Jersey.

W. H. ALLAWAY, Research Professor of Soils, Iowa Agricultural Experi-
ment Station, Ames, Iowa.

HENRY D. BARKER, Principal Pathologist, U. S. Department of Agricul-
ture, Beltsville, Maryland.

CHARLES A. BENNETT, Principal Agricultural Engineer, U. S. Department
of Agriculture, Cotton Ginning Laboratory, Stoneville, Mississippi.

E. C. CHILDS, Assistant Director of Research in Soil Physics, School of

Agriculture, University of Cambridge, England.

F. M. EATON, Principal Physiologist, U. S. Department of Agriculture,

College Station, Texas.

L. E. ENSMINGER, Associate Soil Chemist, Alabama Agricultural Experi-
ment Station, Auburn, Alabama.

R. F. FUELLEMAN, Professor of Agronomy, University of Illinois, Urbana,
Illinois.

J. C. GAINES, Professor of Entomology, Texas Agricultural and Mechani-
cal College System, College Station, Texas.

N. COLLIS-GEORGE, Demonstrator, School of Agriculture, University of
Cambridge, England.

M. K. HORNE, JR., Dean, School of Commerce and Business Administra-
tion, University of Mississippi, University, Mississippi.

WESLEY KELLER, Geneticist, U. S. Department of Agriculture, Logan,
Utah.

W. K. KENNEDY, Professor of Agronomy, Cornell University, Ithaca,
New York.

J. E. KNOTT, Professor of Truck Crops, University of California, Davis,
California.

HELMUT KOHNKE, Soil Scientist, Department of Agronomy, Purdue Uni-
versity Agricultural Experiment Station, Lafayette, Indiana.

VI CONTRIBUTORS TO VOLUME II

0. A. LORENZ, Assistant Professor of Truck Crops, University of Cali-
fornia, Davis, California.

WILLIAM E. MEEK, Senior Agricultural Engineer, U. S. Department of
Agriculture, Stoneville, Mississippi.

R. B. MUSGRAVE, Associate Professor of Agronomy, Cornell University,
Ithaca, New York.

R. W. PEARSON, Senior Soil Scientist, U. S. Department of Agriculture,
Auburn, Alabama.

MAURICE L. PETERSON, Assistant Agrononvist, California Agricultural Ex-
periment Station, Davis, California.

JOHN T. PRESLEY, Head, Department of Plant Pathology and Physiology,
Mississippi Agricultural Experiment Station, State College, Missis-
sippi.

T. R. RICHMOND, Professor of Agronomy, Texas Agricultural Experiment
Station and Senior Agronomist, U. S. Department of Agriculture,
College Station, Texas.

F. F. RIECKEN, Research Professor of Soil*, Iowa Agricultural Experi-
ment Station, Ames, Iowa.

GUY D. SMITH, Senior Soil Correlator, U. S. Department of Agriculture,
Ames, Iowa.

HARRIS P. SMITH, Professor of Agricultural Engineering, Texas Agricul-
tural Experiment Station, College Station, Texas.

Preface

In the preface to Volume I it was pointed out that the pressure of
progress in the many sub-fields that collectively constitute agronomy
tends to produce specialists who find it difficult to keep abreast of newer
developments somewhat removed from their immediate interests, yet of
professional importance to them. It was further explained that the edi-
tors were not inclined to quibble about the precise definition of the word
"agronomy." In selecting topics for treatment they would be guided
more by the consideration of what might be useful to agronomists than
what constitutes agronomy. The authors arc urged to present, as far
as possible, unified, complete and authoritative accounts of the recent
developments in their particular fields. Topics will reappear from time
to time as new material and new viewpoints develop.

This is the mid-century year. It would be presumptious on the part of
a publication so recently established as Advances in Agronomy to
prepare a mid-century number delineating and weighing the achieve-
ments and accomplishments of the first half of this century. The editor
had no difficulty in resisting the urge to follow the lead set in this matter
by long-established and more popular publications. However, this
thought did set in train some speculations as to what topics might have
been selected for a similar volume had one been prepared fifty years ago.
Largely, this amounted to a realization of the topics which would not
have been included because their development has taken place almost
entirely since 1900. Crop improvement through genetics, soil physics,
and soil genesis are examples. A cursory glance at the contents of this
volume will show r that about half of them would not have appeared in
any form in a 1900 edition. Thus fast has agronomy grown.

Speculation in another direction is possible. One might consider the
extent of the agronomic achievements of the past half century in terms
of the changes in U. S. agriculture and agricultural practice. These are
dramatic enough. Here is a nation which in fifty years has doubled in
population but has no more farms now than then. Twenty-five per cent
more acres are harvested, meat production has been doubled, fertilizer
consumption has increased eight-fold, but there are fewer sheep, and
far fewer horses and men on farms now than at the turn of the century.
Perhaps in the last lies a clue to much that has been accomplished by the
mechanization of many operations through the availability of power
equipment.

What can be anticipated in the remainder of this troubled century?

vii

Vlii PREFACE

Have the easy things been done? Will the tempo of progress be slowed?
Are the land use systems that have developed stable or exploitive? Will
the crop surpluses of domestic production be absorbed in meeting the
deficits elsewhere, or are they merely a temporary feature soon to be
dissipated by population increase? Should production be curtailed in the
interests of conservation? Will there be changes in food habits on the
part of the consumer that will call for great shifts in types of farming?
Will the era of surpluses even give way to a period when the demands
for food will be such that exploitive land use is forced upon us?

The resolution of many vital questions such as these will not pri-
marily lie in the hands of those practicing the profession of agronomy,
yet they will be required to use all their skills and ingenuity and re-
sourcefulness in providing the solution to the innumerable practical prob-
lems that taken together will determine the answers to such major
questions. Subsequent volumes of the Advances will record and sum-
marize their methods, recount their achievements and measure their
accomplishments.

A. G. NORMAN
Frederick, Md.
October, 1950.

CONTENTS

Page

Contributors to Volume II v

Preface vii

Cotton

COORDINATED BY J E. ADAMS, Texas Agricultural and Mechanical College System,

College Station, Texas

I. Introduction, BY J. E. ADAMS 2

II. Competitive Position of Cotton Among Fibers, BY M. K. HORNE, JR. . 5

III. Physiology of the Cotton Plant, BY F. M. EATON 11

IV. Diseases of Cotton, BY JOHN T. PRESLEY 26

V. Insect Pests, BY J. C. GAINES 32

VI. Improvements in Production Practices,

BY WILLIAM E. MEEK and HARRIS P. SMITH 40

VII. Improvements in Ginning Practices, BY CHARLES A. BENNETT ... 50

VIII. Fiber Properties and Their Significance, BY HENRY D. BARKER ... 56

IX. Breeding and Improvement, BY T. R. RICHMOND 63

References 74

Soil Nitrogen

BY L. E. ENSMINGER AND R. W. PEARSON, Alabama Agricultural Experiment Station,
and U. S. Department of Agriculture, Auburn, Alabama

I. Introduction 81

II. Factors Affecting Nitrogen Content of Soils 83

III. Nature of Organic Nitrogen in the Soil 87

IV. Nitrogen Transformations 89

V. Effect of Cropping Practices on Nitrogen Level 94

VI. Nitrogen Economy of Eroded Soils 98

VII. Commercial Nitrogen vs. Barnyard Manure and Green Manures . . . 100

VIII. Nitrogen Trends in Various Parts of the U. S 104

References 109

Vegetable Production
BY J. E. KNOTT AND 0. A. LORENZ, University of California, Davis, California

I. Introduction 114

II. Fertilization 116

III. Trace Elements 120

IV. Development of New Vegetable Varieties 121

V. Utilization of Heterosis 128

X CONTENTS

Page

VI. Growth Control Techniques 135

VII. Labor Saving Devices 141

VIII. Possible Future Developments 151

References 152

Prairie Soils of the Upper Mississippi Valley

BY GUY D. SMITH, W. H. ALLAWAY AND F. F. RIECKEN, U. S. Department of Agri-
culture, Ames, Iowa, and the Iowa Agricultural Experiment Station, Ames, Iowa

I. Introduction 157

II. Characteristics of a Modal Prairie Soil 159

III. Variability of Prairie Soils as Functions of Soil-Forming Factors . . . 166

IV. Classification of Prairie Soils 192

V. Distribution of the Prairie Soils 196

VI. Crop Yields from Prairie Soils 197

References 203

Ladino Clover

BY GILBERT H. AHLUREN AND R. F. FUELLEMAN, Rutgers University, New Brunswick,
New Jersey, and University of Illinois, Urbana, Illinois

I. History and Distribution 208

II. Characteristics and Adaptation 209

III. Establishment and Management 213

IV. Utilization 226

V. Summary 230

References 230

The Control of Soil Water

BY E. C. CHILDS AND N. COLLIS-GEORGE, School of Agriculture, University of

Cambridge, England

I. The Scope of the Review ^ 234

II. Research Methods , 234

III. The Basic Approach 235

IV. Drainage and Irrigation 258

References 269

Preservation and Storage of Forage Crops

BY R. B. MUSGRAVE AND W. K. KENNEDY, Department of Agronomy, Cornell
University, Ithaca, New York

I. Introduction 274

II. Measurements of Changes in Quality During Preservation and Storage . 275
III. Silage 279

CONTENTS XI

Page

IV. Field-Cured Hay 2M

V. Barn Hay Drying 299

VI. Artificial Drying 304

VII. Experiments Comparing Silage, Barn-Cured and Field-Cured Hay . . 306

VIII. Conclusions 309

References 311

The Reclamation of Coal Mine Spoils

BY HELMUT KOHNKE, Purdue University Agricultural Experiment Station,

Lafayette, Indiana

I. Introduction 318

II. The Condition of Spoil Banks 319

III. Mrthods of Testing Spoil Bank Materials 330

IV. Land Use Capabilities 332

V. Methods of Revegelation 335

VI. Grading 338

VII. Economic Aspects 341

VIII Legislation 344

IX. Discussion 344

X. Summary 347

XI. Acknowledgments 348

Reformers ... 349

Irrigated Pastures

BY WESLEY KELLER AND MAURICE L. PETERSON, U. ft. Department of Agriculture,
Logan, Utah, and California Agricultural Experiment Station, Davis, California

I Introduction 351

II. Pasture Soils 352

III. Choosing Productive Mixtures 356

IV. Establishing Pastures 360

V. Management of Pastures 365

VI. Economy of Pastures 375

VII. Pastures in Relation to Other Sources of Feed 380

References 382

Author Index 385

Subject Index 402

Cotton

Coordinated by J. E. ADAMS
Texas Agricultural and Mechanical College System, College Station, Texas

CONTENTS

Page

I. Introduction by J. E. ADAMS , 2

II. Competitive Position of Cotton Among Fibers by M. K. HORNE, JR. . 5

1. Cotton Loses Markets 5

2. End-Uses of Cotton and Other Fibers 6

3. A Static vs. a Dynamic Position for Cotton 8

4. Need for Expanded Research 9

III. Physiology of the Cotton Plant by F. M. EATON 11

1. Floral Initiation and Plant Development 11

2. Mineral Nutrition 14

3. Nitrogen 20

4. Carbohydrates, Nitrogen and Fruitfulness 22

5. Effect of Drought on Plant Composition and Fruiting 24

6. Drought and Other Factors Affecting Boll Development and Lint

Properties 24

7. Oxygen Requirements for Root Growth 25

IV. Diseases of Cotton by J. T. PRESLEY 26

1. Seed Treatment 26

2. Phymatotrichum Root Rot 26

3. Fusarium Wilt 28

4. Verticillium Wilt 29

5. Bacterial Blight 30

6. Root-Knot 31

7. Summary 32

V. Insect Pests by J. C. GAINES 32

1. Thrips 32

2. Cotton Aphid 33

3. Cotton Fleahopper 33

4. Boll Weevil 34

5. Bollworm 36

6. Pink Bollworm 37

7. Hemipterous Insects 38

8. Cotton Leafworm 38

9. Spider Mites 39

10. General Recommendations for Chemical Control 40

VI. Improvements in Production Practices 40

1. In Humid Areas by W. E. MEEK 40

2. In Low-Rainfall and Subhumid Areas by H. P. SMITH 46

2 J. E. ADAMS

Page
VII. Improvements in Ginning Practices by C. A. BENNETT 50

1. Regulation of Moisture 50

2. Cleaning 52

3. Extraction and Interrelated Processes 53

4. Summary 55

VIII. Fiber Properties and Their Significance by H. D. BARKER 56

1. Fiber Structure and Development 56

2. Fiber Length 58

3. Fiber Strength 59

4. Fiber Fineness 60

5. Significance of Fiber Properties 62

IX. Breeding and Improvement by T. R. RICHMOND 63

1. General 63

2. The Breeding Problem 64

3. Breeding Systems 66

4. Hybrid Vigor in Fi and Advanced Generations 70

5. Special Phases 71

References 74

I. INTRODUCTION

J. E. ADAMS
Texas Agricultural and Mechanical College System, College Station, Texas

Cotton is the most important cash crop grown in the United States,
and is the only major crop which produces the 3 products, fiber," food and
feed. When all phases of the industry are considered, some 20 to 25
million people are wholly or partially dependent on cotton as a source of
income. Of these, approximately 1,500,000 are engaged in production
and 3,000,000 in ginning, marketing and processing.

The fact that the entire fruit of the cotton plant is used makes it an
unusual crop. In addition^to lint, the embryos or "meats" of the seed
furnish both a protein concentrate and a high-grade oil. The "hulls"
or seed coats are used as feed, as well as the concentrate. The fuzz left
on the seed after normal ginning is removed mechanically and these short
fibers known as "linters" are used in upholstering, low-grade mattresses,
and as a source of cellulose for synthetics.

The present status of cotton is both artificial and in a state of flux.
Although cotton lint is the most versatile fiber known when all end-uses
are considered, economic conditions since the early thirties, with inter-
mittent acreage control and price subsidies, have resulted in increased
competition of old and new synthetic fibers. Unrestricted production
during and following the war, with impending acreage controls, is re-
flected in Table I, released as of December 8, 1949, by the Bureau of

COTTON 3

Agricultural Economics, Austin, Texas. Every cotton-producing state
shows an increase in acreage in 1949 over 1948. In spite of lower produc-
tion per acre in some areas, there was a net increase of 1,157,000 500-lb.
gross-weight bales in 1949, as compared with 1948. More striking is
the increase of 4,728,000 bales in 1949 over the average of 11,306,000
bales in the 1938-1947 period. The increase in production in the irri-
gated areas, particularly California and also New Mexico and Arizona,
is phenomenal.

Although the sections on production treat mechanization of the cotton
crop in many of its details, it is to be emphasized that mechanization
lor all areas really dates to the World War II period, during which there
was an intensification in research and more ready acceptance by the
farmer due to the dearth of labor. Development of flame cultivation,
rotary hoes, efficient fenders for cultivation equipment, along with the
culmination of years of research on harvesting machinery, resulting in
usable harvesters, has made full mechanization of the crop a reality in
some sections. Remaining difficulties for other areas and particular
seasons are yielding to intensified research.

Estimates furnished by the National Cotton Council indicate that
approximately 2900 spindle-type pickers were used in 1949. Mississippi
led with 990. California used approximately 850, Arkansas, 350, and
Louisiana, 200. The acreage harvested would be difficult to estimate,
but probable harvest per machine varied from 100 to 250 acres. A total
of better than 7000 strippers were used, with the Texas High-Plains area
accounting for at least 6000. Oklahoma was the only other large user,
the estimated number being 950. The capacity per machine ranges from
125 to 400 acres, depending on local conditions.

Considering that practically all of the cotton harvesters have become
available since the war, the increase in use of machinery is outstanding.
The general feeling that costs of production for cotton must eventually
be radically lowered, leads to the conclusion that production per acre
must be increased. Better fertility practices, control of insects and dis-
eases, better varieties of cotton for mechanical harvesting, along with
more compl^e^and assured defoliation are of prime importance.

All of the cottons of the world, whether cultivated or wild, belong
to the genus, Gossypium. They may be divided into 3 main groups: (1)
Old World or Asiatic cultivated (n = 13), (2) New World or American
cultivated (n = 26), and (3) wild, (n = 13, with one anomalous excep-
tion). Though the reported number of species of Gossypium varies
widely, depending on the classification system employed and the inclina-
tion of the taxonomist, a recent work by Hutchinson et al. (1947) (see
References in IX) recognizes twenty. These writers place the cultivated

J. E. ADAMS

cottons under four species: G. arboreum L., G. hirsutum L., and G. bar-
badense L. The first two are designated as Asiatic and the last two as
American cottons, and all bear spinnable seed hairs, called lint, which
distinguishes them from the wild cottons which do not have spinnable
lint. American cultivated cottons (n = 26), according to the theory
advanced by Skovsted (1937) (see References in IX) and confirmed
independently by Beasley (1940) and Harland (1940) (see References

TABLE I
Cotton Production in United States

Acreage Harvested

Lint Yield per

Production

Harvested Acre

(Ginnings)*

500 Ib. gross

wt. bales

State

Aver-

1948

1949

age

1948

1949

age

1948

1949

age

Crop

1938-

(re-

(Dec. 1

1938-

(re-

(Dec. 1

1938-

(re-

(Dec. 1

1947,

vised) ,

est.),

1947,

vised)

est.),

1947,

vised)

, est.),

1000

acres

Ibs.

bales

Missouri

375

555

583

451

436

377

356

506

460

Virginia

348

447

300

North Carolina

743

725

815

355

447

270

549

678

460

South Carolina

1,118

1,120

1,270

309

372

211

716

871

560

Georgia

1,608

1,289

1,550

235

279

189

779

751

610

Florida

164

249

196

Tennessee

684

770

830

368

417

375

523

669

650

Alabama

1,691

1,630

1,810

262

353

229

901

1,197

865

Mississippi

2,397

2,560

2,770

318

441

258

1,588

2,353

1,490

Arkansas

1,916

2,220

2,450

334

428

325

1,329

1,982

1,660

Louisiana

968

950

1,060

261

382

294

528

756

650

Oklahoma

1,491

1,025

1,300

163

175

229

521

374

620

Texas

7,642

8,610

10,725

170

176

264

2,722

3,153

5,900

New Mexico

115

209

310

497

542

394

119

236

255

Arizona

200

281

373

423

558

641

174

328

500

California

352

804

957

602

576

651

447

968

1,300

Other States 1 *

413

435

383

UNITED STATES

21,396

22,821

26,898

254.0

312.6

285.8

11,306

14,877

16,034

Amer. Egypt*

635

4.0

6.3

279

434

327

29.5

3.6

4.3

* Allowances made for interstate movement of seed cotton for ginning.
b Illinois, Kansas, and Kentucky for all years and Nevada for 1948 and 1949.
c Included in State and United States totals. Grown principally in Arizona, New
Mexico, and Texas.

COTTON 5

in IX) , are tetraploids which have arisen by amphidiploidy from hybrids
of Asiatic (n = 13) and American Wild (n = 13) parentage.

This paper will present important aspects of the production of cotton
each of which, because of space limitations, can be treated only briefly.

II. COMPETITIVE POSITION OF COTTON AMONG FIBERS

M. K. HORNE, JR.

University of Mississippi, University, Mississippi

1. Cotton Loses Markets

Except for the abnormal experience of the war and early postwar
years, it can be said that over the past 4 decades the per capita con-
sumption of cotton in this country has shown no tendency to rise. Over
this long period, it has gravitated around a central figure of 25 or 26
Ibs. a year, displaying no trend either up or down. In net effect, the
entire new market created by rising standards of living has been cap-
tured by rayon, paper, and to a smaller extent, other materials. In this
fact we have an indication of what competition has done to the cotton
market down to the present time, and why the students of the demand
for cotton are engrossed in its competitive position.

There are at least 35 materials which give cotton substantial compe-
tition. The more interesting ones include rayon in its various forms,
paper, glass fiber, nylon and the other fibers of synthesized polymers,
the synthetic protein fibers, plastic film, and jute, ramie and the other
bast fibers.

In seeking the competitive meaning of these numerous materials, it
seems helpful, and in some degree defensible, to think primarily in terms
of rayon. This fiber is not only cotton's biggest, but by a wide margin
its most serious, most threatening competitor. In Project IV of the
cotton fact-finding program, "A study of the agricultural and economic
problems of the Cotton Belt," presented in 1947 before the Cotton Sub-
committee of the Committee on Agriculture, U. S. House of Represen-
tatives, it was found that some form of rayon was cotton's closest
competitor in 65 out of a group of 106 end-uses analyzed. In 25 years,
rayon has advanced from a trivial position among all fibers to second
place in the volume consumed in the United States. In 1948, 1,124,000,-
000 Ibs. of rayon were produced in this country. Factory capacity has
now reached an estimated 1,235,000,000 Ibs., or the equivalent in usable
fiber of about 2,900,000 bales of cotton. This amounts to 48 per cent
of the average annual consumption of cotton in this country during the

6 M. K. HORNE, JR.

decade of the 1930's, and to 35 per cent of the cotton consumption in the
best year of that decade, which was 1937. Two-thirds of this rayon
capacity has been built since the end of that decade. With the return
to a buyer's market for textiles in the United States, cotton is inevitably
feeling a terrific impact from a competitor which has grown so rapidly
ajid become so large. In foreign lands, the tendency is for rayon to
assume an even stronger competitive position than in the United States.

2. End-Uses of Cotton and Other Fibers

Every end-use market for cotton has a separate pattern of require-
ments in price and in the scores of distinct qualities which characterize
a fiber product. Likewise, there is a separate price and quality pattern
for every competing material. Most materials are quite different from
cotton in price and quality, and they must compete in a more limited
number of end-uses where their special characteristics give them an ad-
vantage. As competitors of cotton they are specialty materials, con-
centrating upon and limited to some segment of cotton's end-use markets.

Paper, for example, is cotton's second most aggressive competitor
today. It has a big advantage in price and a few small advantages in
quality. Within a limited range of uses, these factors can sometimes
overcome the important advantages of cotton in other properties. Paper
is a formidable competitor in the great bag market, and in large sectors
of the towel, cordage, napkin, and handkerchief markets. Its quality
is being steadily improved through research. But paper's differences in
texture, strength, and absorbency quite obviously restrict it to a fraction
of the cotton market.

For a second example, the leading synthetics other than rayon are
considered. They are all far above the price range of cotton. One may
be tempted to reason that since rayon came down to the price of cotton,
these other materials can eventually do likewise. Any such reasoning
would seem to be premature, ai least for the better fibers. In the strong,
resilient synthesized polymers, including nylon, the textile scientists point
out that basic differences in the chemical approach seem to invalidate
the idea of ever bringing costs down to the level of rayon. At the same
time, these various synthetics are quite superior to cotton in some quali-
ties, and quite inferior to it in others. There are certain uses for cotton
in which their patterns of quality can overcome the price handicap, but
again these uses are limited.

Rayon ,is distinguished by the fact that to an ever-increasing degree
it competes for markets, not because of its differences from cotton, but
because of its similarities to cotton. Its price and quality pattern has
relentlessly shaped itself toward the price and quality pattern of cotton.

COTTON 7

Very generally, it can be said that today rayon is in the same price
range with cotton. In reference to quality, there still are sharp differences
between the two fibers, but rayon has made marvelous progress in over-
coming its quality handicaps through research. That progress can be
seen in the development of staple fiber, in delustering, in crimping, in
higher wet and dry strength, in softer yarns, and in better finishes for di-
mensional stability. Several of rayon's biggest quality handicaps remain,
but we cannot overlook the fact that the extensive research program
of the rayon industry is going vigorously forward. As a competitor of
cotton, rayon looks less and less like a specialty fiber and more and more
like a fiber whksh eventually may contend for virtually all of cotton's
markets.

The two chief weapons with which rayon might be expected to im-
prove its present competitive position are: (1) the lowering of price;
and (2) the improvement of quality. Which of these two possibilities
is the more threatening to cotton? From the excellent research in recent
years, it now seems fairly clear that the rayon industry is likely to rely
primarily on improved quality to buttress its competitive position. The
era of continual price reductions in rayon, year by year, seems to have
ended in 1938. The average production cost of viscose staple at a recent
time was 29 cents a pound. This of course was before any allowance
for income tax. The selling price of viscose staple was 35 cents in
March, 1950. Obviously it is unlikely that any drastic price reduc-
tion will be achieved by reducing profits unless market conditions take
a serious turn for the worse. On the side of cost reductions, it must be
recognized that real technical progress is still being made, but the nature
of the cost is such that the further reductions are likely to be much more
gradual than in the past. Therefore, it seems that any declines in the
selling price of viscose staple are likely to be modest in amount unless
a serious business recession occurs. For types of rayon other than
viscose staple, the possibilities of cost reductions, through technical ad-
vances, may be somewhat greater, but nowhere can any reductions of
major proportions be foreseen.

On the other hand, the rayon industry is spending large sums of money
on research, a major part of it apparently aimed at the improvement of
quality. It is perhaps a reasonable guess that the present rayon research
and technical-service programs of the 5 leading companies are costing
ten million dollars a year. We can never predict what research labora-
tories may bring forth, but we must be impressed with the success of the
rayon research program down to the present time. A continued, and
perhaps an accelerated, improvement in the quality of the various rayons
appears to be the greatest threat to the market for cotton. A continuing

8 M. K. HOENE, JE.

trend toward general equality in the price and quality patterns of the
two fibers is therefore to be expected.

3. A Static vs. a Dynamic Position for Cotton

In the face of this trend, how shall we appraise the competitive
position of cotton?

Although the situation is essentially dynamic, let us examine it first
under assumed conditions which would make it static. The assumptions
include the following: (1) no change from the situation of 1946 in the
relative qualities of cotton and rayon products; (2) no change in the
relative merchandising efficiency of the two fibers; (3) no war-created
shortages or deferred demands; (4) a level of economic activity repre-
sented by 7 million unemployed; and (5) the price levels prevailing in
January, 1946.

The two most interesting price assumptions for cotton are, first, the
price actually prevailing in January 1946, or about 25 cents, and second,
a price about half as high, or 12 cents.

With these assumptions a group of textile economists made a sys-
tematic effort to estimate the amount of cotton that would be consumed
in 127 end-uses, representing about 83 per cent of the domestic market.
For the other 17 per cent, made up of countless small uses, it was assumed
that changes would follow the pattern of the 83 per cent. The sources
of information already available were supplemented by field trips, in
which numbers of the best informed business executives were questioned
with regard to- each end-use. The estimates which came out of this work
were as follows: first, that at 25 cents per Ib. cotton would find a domestic
market for about 7,700,000 bales; second, that at 12 cents per Ib. cotton
would find a domestic market for about 9,600,000 bales.

Since prices in general, and rayon prices in particular, are now about
40 per cent higher than in January, 1946, the assumed cotton price may
logically be increased to this -extent. With this revision, and under the
other assumptions stated, the study indicated that, at today's general
price level, the domestic market for cotton would tend to be 7,700,000
bales at 35 cents, compared with 9,600,000 bales at 17 cents.
Two lessons from these figures are outstanding:
First, by cutting the higher price in half, the quantity consumed
would be increased by less than two million bales. The demand, con-
sidered in this sense, is inelastic, even when allowance is made (as it was
in Project IV) for enough time lapse to permit a given price to exert
a real influence on an end-use market. For certain end-uses (notably
tire cord, bags, insulation, and plastic laminates) the demand is elastic,
but in the great bulk of the domestic market the differences in consump-

COTTON 9

tion at the two price levels would be quite small, so small as to make the
overall domestic demand rather inelastic. The inelasticity results chiefly
from two facts: (a) the price of raw cotton is a small factor in the aver-
age retail price of cotton products, and (b) under the assumption of no
change in quality, the substitutability of other materials is quite limited.

Second, at either price, cotton still seems to have the competitive
strength to hold a very substantial domestic market. Even at the higher
price, the quantity consumed would be no smaller than in 1939. The
explanations are: (1) the increase in population; (2) some assumed im-
provement in business conditions; and (3) a decisive margin of quality
advantages for cotton in many important uses. Despite rayon's gains,
cotton still holds firmly to many markets by the strength of some very
real quality advantages, most of which can be summarized under the
headings of launderability, durability, and versatility. Cotton is endowed
with a remarkable combination of qualities, all present in the same fiber
at the same time: they include wet and dry strength, abrasion resistance,
good absorbency and dyeing properties, vapor permeability, chemical
stability, softness, pliancy, and ease of preshrinking.

In view of this finding and of what we have already said about rayon,
it seems that two facts should be made equally emphatic on the vital
point of quality: (1) (a static concept) at a very recent time, cotton's
quality advantages were strong enough to protect most of its domestic
markets from rayon. Since that time there has been no drastic change
in the quality picture. If cotton could hold its present position in quality,
the great part of its domestic market would appear to be reasonably
safe from competition, even at a high price; (2) (a dynamic concept)
there is little reason to hope that under the present circumstances cotton
is holding its position in quality. Over the past 25 years, rayon has
made enormous gains on cotton in quality, and the research and devel-
opment program which produced those gains is now being pushed forward
on a record scale. Cotton does not have a research and development
program of equal size and scope. Some excellent advances are being made
in cotton research, but the program is simply too small to give it a rea-
sonable chance of equalling rayon's achievements. It is essentially a
mere token of what is needed.

4. Need for Expanded Research

On the dynamic side of the quality problem, however, one further
point needs emphasis: cotton's lag in quality improvement results from
the lack of an adequate research program and not from the lack of op-
portunities in the fiber. The Project IV report outlined 41 broad fields
of quality in which cotton's markets might be strengthened through

10 M. K. HORNE, JR.

research. Cotton is a promising subject for quality improvement at
every stage from plant breeding through production, ginning, spinning
and weaving, finishing and fabricating. Cotton has the opportunity to
become a dynamic fiber like rayon, matching rayon's progress with
progress of its own, and thereby postponing indefinitely the day when
rayon will overtake it in quality. The opportunities exist, but in spite
of recent expansion, an adequate program does not.

Let us now attempt to summarize the significance of two competi-
tive factors, price and quality, and their interrelationship with one an-
other, in the domestic market for cotton. There seem to be 3 points
which deserve attention: (1) as long as cotton holds its present quality
advantages, the demand on the domestic market will be rather inelastic
in response to price change. There will be a very substantial market
for cotton at what has usually been regarded as a high price; (2) if
rayon continues to improve in quality more rapidly than cotton, in the
course of time the amount of cotton that can be sold at any price will
decline; (3) if rayon continues to improve in quality more rapidly than
cotton, in the course of time the significance of price as a competitive
factor will increase. As rayon becomes more substitutable for cotton,
the demand for cotton will become more elastic in its response to price
change. If rayon ultimately becomes as launderable and as durable as
cotton, and cotton makes no offsetting gains in quality, it will then be
out of the question for cotton to sell on the domestic market at a figure
which would not give it a marked price advantage over rayon.

Thus, from the standpoint of the cotton economy, the largely neg-
lected opportunity to build an adequate research and development pro-
gram for quality improvement presents a vital problem. The cost of
such a program would be large, but it could be measured in millions of
dollars annually. If, through the lack of such a program, it becomes
necessary to make sharp reductions in the price of cotton, that loss will
have to be measured in hundreds of millions of dollars annually.

In this statement many factors which bear upon the competitive po-
sition of cotton have been omitted. In actual practice, the important
place of merchandising in the domestic market cannot be overlooked, nor
can the special nature of the export market. In the limited space avail-
able, however, attention has been concentrated upon the great economic
significance of quality improvement.

COTTON 11

III. PHYSIOLOGY OF THE COTTON PLANT

FRANK M. EATON
UJS. Department of Agriculture, College Station, Texas

1. Floral Initiation and Plant Development

a. Branching Habits of the Cotton Plant. Some of the most impor-
tant physiological responses of the cotton plant find their expression and
basis in the type of branches which are produced. According to condi-
tions of growth, the branches arising from the main stalk may be ex-
clusively vegetative branches or exclusively fruiting branches. In the
tixil of each leaf on the main stalk, and also on vegetative branches,
there are two buds. One of these buds, if it develops, will produce a
vegetative branch and the other a fruiting branch; both buds may de-
velop. Morphologically, the vegetative branches, or limbs, are like the
main stalk. Only fruiting branches develop flowers and bolls. A flower
bud, even though it may absciss while still a millimeter or two in
diameter, develops in the axil of each leaf of the fruiting branches.
Although there are various complexities, American Upland cottons, unless
planted too closely, typically develop from their main axes one or several
vegetative branches between the first and eighth nodes. Thereafter,
starting between the seventh to tenth nodes, only fruiting branches are
developed from the main-stalk nodes. In addition to developing from
the main stalk and from vegetative branches, vegetative branches may
also develop from fruiting branches. There are important early papers
by O. F. Cook on the morphology of the cotton plant.

Gaines (1947) has found that, in the absence of insect control, a loss
of 50 per cent of the floral buds during the first 30-day period of fruiting
is without effect on final yields. This is in conformity with earlier
agronomic and physiological observations showing that the loss of some
of the early floral buds aided in the maintenance of plant development
and that new buds were developed to replace those that were lost. Under
some conditions, the removal of early buds and flowers has resulted in
increased yields. Data are occasionally presented to show that the
flowers at fruiting-branch nodes near the main stalk are more apt to de-
velop into bolls than are those farther out. But if these first buds and
bolls are lost they are replaced by those that might otherwise have shed,
environmental conditions permitting.

The types of branches produced by the cotton plant are influenced
by temperature and by length-of-day (Fig. 1). Also the number of
vegetative branches may be influenced by closeness of spacing, defruiting,

FRANK M. EATON

darkening the tips of plants (Eaton and Rigler, 1948) and by treatments
with growth substances which cause buds to shed. Whether or not there
is some one chemical of hormone-like entity, which is responsible for the
determination of which type of branch shall develop under a given
circumstance, remains one of the most intriguing aspects of cotton
physiology. Some of the reactions of the cotton plant provide a basis
for regarding the fruiting branch as being homologous to the inflorescenses
of other plants.

Fig. 1. On the lefl are cotton plants with fruiting brands only, and on the right,
plants with a preponderance of vegetative branches. The latter plants had some
short fruiting branches with floral buds but so far none had produced bolls. These
plants all received 13-hour days and represent the difference between hot nights
(left) and cool nights (right) at San Diego, California, where the days are cool.
Certain photopcriodic cottons would give this same response to short days, left, but
produce only vegetative branches under long days. (Eaton, 1924).

b. Photoperiodism. No work on the photoperiodism of the cotton
more extensive than that of Konstantinov (1934) has appeared in the
literature. By that work it was shown that the length of day may alter
the fruiting activities of some, but not all, of the perennial arborescent
cottons, particularly those from equatorial regions, and to a slight extent,
also, varieties of Egyptian and of medium and late Uplands. He con-
cluded that the early (determinant) American Upland cottons, as well
as some of the wild forms from Mexico and elsewhere, were without
length-of-day reactions. The author states that when length-of-day
reactions were found, the basic change consisted in a lowering of the

COTTON 13

position of the first fruiting branches. All cottons exhibiting photo-
periodism are those requiring short days, i.e., cottons are unknown that
require long days for flowering.

c. Temperature. The striking influence that temperature may exert
on the kind of branches produced, and, therefore, on the fruiting of cot-
ton plants, is illustrated in Fig. 1. Dastur (1948) makes mention of
observing lower temperatures to be conducive to the development of
vegetative branches. At Shafter, California, where mean nightly tem-
peratures for the summer months averaged about 60F., Acala p-18-c
was observed in 1947 (unpublished data) to develop 10 times as many
vegetative branches per 20 feet of row as did the same variety similarly
spaced at Sacaton, Arizona, where the average minimum nightly tem-
peratures are about ten degrees higher. Laying a soil heating cable
under a dust mulch along the two sides of a row of cotton plants at
Shafter increased the nightly temperature by a few degrees 4 inches
above ground and in turn lessened the development of vegetative
branches.

d. 2,4-D an d Hormone Responses. Attention was first directed by
Staten (1946) to the high sensitivity of the cotton plant to wind-carried
traces of 2,4-dichlorophenoxyacetic acid and its derivatives. The out-
standing symptom of an excess of this material was shown by Staten and
by Dunlap (1948) to be the growth repression of the mesophyll of leaves
and involucral bracts which gave these organs a ligulate appearance in
which the veins were especially prominent. Brown et cd. (1948) il-
lustrate, also, the pronounced swelling of the stems of cotton plants at
ground level. Each of the foregoing investigators has shown that
dusts and fine mists of 2,4-D can be carried many miles in sufficient
concentration greatly to reduce cotton yields. Dunlap pointed out that
if the injury was not too severe the cotton plant could put out new
branches of normal appearance and develop late bolls. He also showed
that the seed of plants injured by 2,4-D might upon germination have
swollen hypocotyls and typical aberrations of the true leaves.

Ergle and Dunlap (1949) found that more than 0.002 mg. of 2,4-D
per plant reduced the yield and increased the height and number of vege-
tative branches of cotton plants. Changes were found in the concentra-
tions of several organic constituents of the leaves that were associated,
possibly, with the altered proportions of vein and mesophyll tissue. The
highest concentration used (0.04 mg,) appeared to reduce somewhat the
tensile strength of fiber.

Singh and Greulach (1949) concluded from a carefully planned green-

14 FRANK M. EATON

house experiment that sprays of a-naphthaleneacetic acid and a-naphtha-
leneacetamide, although altering several plant characters, caused no ef-
fects of agronomic significance.

In California, in either of two years, during periods when 60 to 70
per cent of the bolls were shedding, Eaton (1950) could find no evidence
of any effect on boll retention by dusting cotton plants with 1000 p.p.rn.
napthaleneacetic acid, with 100 p.p.m. sodium 4-chlorophenoxyacetate, or
with the two in combination. Of weekly sprays with 10 and 20 p.p.m.
4-chlorophenoxyacetate, (i-naphthoxyacetate, and a-naphthaleneacetate,
only the 20 p.p.m. concentration of 4-chlorophenoxyacetate altered
growth or fruiting. This latter material reduced significantly the number
of bolls per plant and bolls per 100 g. of fresh stems and leaves, and in-
creased significantly the height, and the number of main stalk nodes and
vegetative branches. The increased height and number of vegetative
branches were regarded as probably the result of reduced fruiting caused
by extra bud shedding. As a part of this work, attempts were made
both in winter and summer to alter the types of branches produced by
day-length sensitive and day-length neutral cottons by treatment with
various of the presently available synthetic growth substances. These
efforts were not successful, but the investigations are regarded as de-
serving of continued effort as new materials become available, particu-
larly any that influence floral initiaton or repression.

2. Mineral Nutrition

Knowledge of the mineral nutrition of plants has gained important
impetus during the past ten years from rapidly developing evidence and
views on the exchange of cations between the plant and soil. A recent
review by Wadleigh (1949) deals extensively with the relations repre-
sented. The order of ease of release of cations from soil colloids by
exchange reaction is headed by sodium which is released most easily
followed by potassium, magnesium, calcium and hydrogen. As measured
on clay membranes, the activity of sodium adsorbed on montmorillonitic
clay is 20 to 25 times that of calcium. The activity of adsorbed sodium
relative to calcium is always greater than the ratio of adsorbed sodium
to adsorbed calcium, but the calcium on kaolinitic clay may be 10 or
20 times as readily available to plants as that on montmorillonitic clay.
Calcium may be unavailable to plants if it constitutes only 50 per cent
or less of the bases retained by the clay, i.e., high levels of adsorbed
potassium or sodium may prevent calcium uptake. Hydrogen ions re-
leased from plant roots provide the critical exchange ion for the release
of calcium, potassium, etc., from clay. Carbon dioxide arising from root
metabolism is the antecedent agent in the transfer of hydrogen from root

COTTON 15

to clay. Of like recognized importance, but less well understood, are
the processes and intensities of adsorption of nutrient cations on root
surfaces and their relative rates of transference inwardly.

Investigations by Jacobson and Overstreet (1947) indicate that
energy arising from respiration is directly involved in the intake of anions
whereas the CO 2 product of respiratory activity functions in the hydrogen
transfer that is instrumental in cation accumulation by exchange.
Lundegardh's review (1947) deals extensively with this phase of mineral
nutrition. In the instance of cotton, Eaton and Joham (1944) found that
defruiting to increase sugar concentrations resulted in significant in-
creases in both bromine and potassium in the fibrous roots; there was
also an increase of both elements in the leaves, but the potassium in-
crease was slight and not significant.

a. Constant Sum of Cations. Recent papers by van Itallie (1948)
and Wallace et al. (1948b), who worked with oats and alfalfa, respec-
tively, have added support to a conclusion reached earlier that the sum
of the cation equivalents per unit of dry weight at a given stage of plant
development tends to be uniform even though the species is grown on
substrates of widely varied composition. Within what limits this con-
clusion can be extended to cotton is not yet clear. Cooper et al. (1948)
thought that it might not be applicable to cotton where there is a wide
variability in hydrogen-ion concentration. As grown on 7 plots at
Florence, South Carolina, the sum of equivalents of K, Na, Ca, and Mg
varied from 112 to 218. The two highest values were on limed plots
(pH 6.7 and 6.5). On plots with pH values of 5.1 and 5.2 respectively
the values were 159 and 112.

6. Sodium and Potassium. The interest that has been attached in
the South to the roll of sodium as a plant nutrient for cotton has applied
also in an important manner to other plants in other regions. Although
no one has assumed, or concluded, that sodium is an essential element,
there is now a wealth of evidence that it can make up in part, in vary-
ing degrees in different plants, for a deficiency in potassium supply.
Furthermore, in some plants, such as the beet (Sayre and Vittum, 1947),
sodium applications have resulted in yield increases over and above those
that could be obtained by potassium alone. Plants which tend to ac-
cumulate more sodium than potassium, such as beets, cabbages, carrots,
and spinach (Wallace et al., 1948a) tend also to be tolerant to sodium
and perhaps are more often benefited by sodium.

Collander (1941) has reported sodium to be higher in the roots than
in the shoots of a number of plants whereas K was higher in the shoots.

FRANK M. EATON

Cotton contains much less Na than K above ground, and, as found in
the expressed leaf sap (Eaton, 1942), there was only a fifth or less as
much Na as K (Table II). Cooper et al. (1947) have considered the

TABLE II

Influence of Sodium Equivalent to 100 Ibs. per Acre of Na 2 O on 10- Year Average

Cotton Yields and 3- Year Average Ca, K and Na Accumulations in Plants Equally

Supplied with Nitrate on Norfolk Sandy Loam a

Lbs. of K 2
and N source

Seed Cotton Yield,
Ibs. per acre

Cation concentration in plants
meq. per lOOg. dry weight

Actual

Na gain

Calcium

Potassium

Sodium

No potash
Cal-Nitro

306

70.00

19.63

1.42

Sodium nitrate

521

+ 215

60.00

17,43

11.27

15 Ibs. potash
Cal-Nitro

742

74.67

22.47

1.12

Sodium nitrate

943

+ 201

64.50

21.97

11.01

45 Ibs. potash
Cal-Nitro

1093

76.33

31.00

1.29

Sodium nitrate

1280

+ 187

60.83

32.57

10.75

60 Ibs. potash
Cal-Nitro

1201

74.85

35.57

1.69

Sodium nitrate

1383

+ 182

57.67

36.73

9.65

* Cooper and Garman (1942).

agreements and discrepancies between the order of accumulation of
mineral nutrients in higher plants and the order of the same ions when
arranged on the basis of their electrode potentials measured in equivalent
volts. Spiegelman and Reiner (1942) have called attention to the selec-
tive accumulation of potassium from K and Na mixtures by sand columns
and by myosin, and have suggested that considerations based on chemical
mechanisms offered more promise than those based on physical relations.
The possible fit of preferred ions in the lattice structure of the solid phase
has been pointed to as one explanation.

Potassium is customarily credited with promoting not only the utili-
zation of nitrogen in protein formation but with a catalytic activity in
the assimilation of carbon dioxide and the synthesis of carbohydrates
and oils. It is evident that sodium cannot perform all functions of potas-
sium in cotton, or in cotton does not accumulate in the right places in
sufficient concentrations. This is indicated by Volk's (1946) observation
that Na alleviated but did not eliminate cotton rust. Leaf rust is the

COTTON 17

important symptom of potassium deficiency in cotton. Biddulph's (1949)
radioautographs show K concentrations in cotton leaves to be much more
dense in and near the veins than outward in the more distant mesophyll.
Similar radioautographs of sodium distribution would be of interest. In
the Georgia Coastal Plain, Turner (1944) found that potash deficiency
great enough to cause marked leaf symptoms, heavy leaf loss, and a 25
per cent reduction in the yield of American Upland cottons decreased
the weight of seed per boll by only 10 per cent.

Gains from sodium applications have been common in field experi-
ments with cotton (Andrews and Coleman, 1939; Mathews, 1941; and
Holt and Volk, 1945). As yet there is insufficient evidence for conclud-
ing that the same yield increase might not have been gained from
additional potassium. When ample potassium was supplied, the last
mentioned investigators obtained no benefit from Na additions in green-
house tests, using both sand and potted soil cultures. In the field, how-
ever, they obtained gains of 98 to 213 Ibs. of seed cotton per acre in
plots supplied with sodium in addition to 24 to 48 Ibs. K 2 0. Mathews
(1941) found both sodium and potassium responses on Clarksville soils
in Georgia, but neither element gave a response on Decatur soil. The
availability of K (but not of Na) having been determined in both soils,
it was concluded that the lack of benefit from sodium on the latter soil
was due to the abundance of K. On the Clarksville soil, Na was esti-
mated as being worth 40 per cent as much as K as a fertilizer.

Data by Cooper and Garman (1942) are notable in showing nearly
uniform gains of approximately 200 Ibs. of seed cotton per acre from
applications of 100 Ibs. Na 2 per acre, when K 2 applications were in-
creased from none to 60 Ibs. per acre. At all K 2 levels the added sodium
caused nearly uniform increases in accumulation of Na from about 1.5
to 11 meq. per 100 g. of plant tissue. Adding but 60 Ibs. of K 2 was
sufficient to increase K accumulation from 19 to 35 meq. per 100 g. At
the high levels of supply, there was 3 times as much K as Na in the
above-ground portions of the plants.

Skinner et al. (1944) observed that extra K fertilization increased
the percentages of K and reduced the percentages of Mg and Ca, and
also of N and P in cotton plants. Like Cooper (1945), they concluded
that the requirements of the cotton plant for Mg is less than for K or
Ca and will be satisfied if dolomite is used in the manufacture of non-
acid forming fertilizers.

c. Phosphorus. In each of 5 years Brown and Pope (1939) reported
that heavy applications of phosphorus caused average increases of 30
per cent to 40 per cent in the proportion of flowers produced during the

18 FRANK M. EATON

first two weeks of the flowering period. With heavy P 2 5 applications,
there was also a large increase in the percentage of the seed cotton gath-
ered at the first picking. Potassium on the other hand appeared to de-
crease the determinateness of the plant and to increase ultimate yields.

Radio-phosphorus injected into a leaf vein by Biddulph and Markle
(1944) moved via the phloem to other parts of the plant. The downward
rate of 21 cm. per hour was thought too high to be accounted for by
diffusion. The upward movement was slower than the downward move-
ment. From 30 days before to 25 days after anthesis, Biddulph and
Brown (1945) found that the accumulation of both tagged and untagged
phosphorus in floral buds and bolls was at rates nearly proportional to
the gain in dry weight. Mason and Phillis (1944) supplied cotton plants
with phosphorus in amounts from that causing acute starvation to an
excess, and found that both the soluble and insoluble fractions in the
main-stalk leaves increased throughout the full range; the former in-
crease was linear whereas the latter tended to flatten.

In citrus, tomatoes, soybeans, pineapples and peanuts, various in-
vestigators have found high levels of nitrate to depress the uptake of
phosphate. Similarly, reduced growth on soils low in nitrate has resulted
from heavy phosphate fertilization. This also applies to cotton (un-
published work by H. E. Joham) .

The use of radio-phosphorus has permitted some significant con-
clusions on the availability to cotton of various types of phosphate
fertilizer. Measurements by Hall et al. (1949), showing the proportion
of accumulating phosphorus derived from the soil and from the tagged
fertilizer, have been made with a number of crops under various con-
ditions. On Norfolk sandy loam, cotton derived most phosphorus from
calcium metaphosphate and least from dicalcium phosphate, but the
source of phosphate was without effect on yield. In Alabama, Ensminger
and Cope (1947) concluded that, on old fertilizer plots, the responses to
various phosphates were dependent upon the calcium and sulfur defi-
ciences that had resulted from previous fertilizer practices.

d. Sulfwr. By classical interpretation, sulfur is essential to the syn-
thesis of proteins and when sulfur is deficient various plants become as
chlorotic as they do when nitrogen is deficient. With insufficient sulfur,
there is often an accumulation of N0 8 , as well as of other soluble forms
of N; starch and hemicelluloses also tend to accumulate. Little work
on the biochemical reactions of the cotton plant to insufficient sulfur has
been published, but the plant requirements are known to be fairly high.
In the leaf sap from cotton (Eaton, 1942), much more sulfur and much
less phosphorus were found than in that from the other plants examined.

COTTON 19

During the last war, the substitution of rock phosphates for superphos-
phate resulted in poor cotton yields in some localities. Willis (1936) has
reported finding that sulfur-free fertilizers produced crops equal to those
supplied with sulfur only on soils which had had heavy previous appli-
cations. Younge (1941) noted that sulfur deficiency reduced the number
and delayed the development of cotton bolls on a Coastal Plain soil.
Tests of cotton responses to sulfur at scattered locations in Florida by
Harris et al. (1945) show that there may be a widespread area that
would benefit from its inclusion in fertilizers. Were it not for the large
amount of sulfur in superphosphate and in the gases released to the
atmosphere by some industries, more information on the sulfur metabo-
lism of cotton might now be available.

e. Boron. This element, which is now thought to be involved in
oxidative enzyme systems, is essential to the formation of meristematic
tissues, and when deficient the fruiting branches of cotton are short and
the flower buds fail to develop. Boron has continued to be regarded as
an important constituent of cotton fertilizers under some conditions.
Coleman (1945) reports beneficial results from applying boron at the
rate of 20 Ibs. per acre to Grenada silt loam in Mississippi. Boll size
and number of bolls were increased, but no effects were found on per-
centage of oil in the seed. In representative Georgia soils with 0.05 to
0.55 p.p.m. of water-soluble boron, Olson (1942) failed to obtain in-
creased yields by adding boron.

/. Copper. Like iron, copper functions as a coenzyme in oxidation
and reducing systems. It has been shown by Manns et al. (1937) to
produce substantial increases in yields of cotton when added to fertilizers
in North and South Carolina and Virginia. Gaines et al. (1947) has
found copper applications to cotton in Texas to produce greater yield
increases when applied to the leaves as a dust with insecticides than
when applied in the soil. The extensive literature on copper as a nutri-
ent has been reviewed by Somner (1945), but no indication is afforded
as to how extensively benefits might accrue from its more general use
in cotton production.

g. Chemical Composition. There have been a good many investiga-
tions of the accumulation of minerals in the cotton plant by the various
State Experiment Stations in the South. One of the most recent is that
reported by Olson and Bledsoe (1942). Their data include 3 soils and
4 stages of growth. Relative to the seedling stage, the plants at maturity
were about half as rich in P 2 5 (0.44% on dry weight), CaO (2.08%)

20 IANK M. EATON

and MgO (0.99%), and four-tenths as high in percentage of N (1.60%)
and K 2 (1.39%). As calculated from data from plants on Cecil sandy
loam, the proportions of N, P, K, Ca and Mg in the mature cotton plants,
including squares and bolls, correspond closely with the proportions of
the mineral nutrients supplied by Hoagland's solution, which was de-
veloped on the basis of analyses of barley plants. The efficiency of the
cotton plant in fruiting activities is relatively high. The squares and
bolls of mature cotton plants were found to constitute 65.8 per cent of
the total dry weight of the plant and to contain 57.3 per cent of nitrogen,
78.7 per cent of the calcium, and 53.1 per cent of the magnesium.

According to data by Phillis and Mason (1942) the percentage com-
position of K, Ca, Mg, P, Cl and N in cotton leaves rises during the day
and falls at night through losses. Collections of dew on attached leaves
made about midnight contained an abundance of potassium and only
traces of calcium. The authors regard the results as being in harmony
with the view that the mineral elements enter the leaf in the wood, and,
with the exception of calcium, are translocated from it in the phloem.

8. Nitrogen

In the Sudan, Crowther (1934) found 60 per cent or more of the total
nitrogen of the cotton plant to be in the squares and bolls from the time
the first bolls had started to open until the plants were mature. Not-
withstanding continued leaf development, the movement of nitrogen from
leaves to buds and bolls was at a greater rate than the movement into
the leaves. This progressive exhaustion of leaf nitrogen continued from
the peak of flowering onward. In American Upland cotton, Olson and
Bledsoe (1942) found nearly the same proportion of the total nitrogen
to be in the buds and bolls at plant maturity as did Crowther. By
Wadleigh's (1944) extensive inquiry into the forms of nitrogen and
nitrogen metabolism of the cotton plant, it was shown that protein con-
stituted from two-thirds to three-quarters of the total nitrogen in all
the plant fractions examhied. At the stages of plant development
selected for sampling both total nitrogen and protein nitrogen were much
higher in leaves and immature seed than in other parts. This propor-
tion of protein is in accord with earlier results by Rigler et al (1937)
who studied the dialyzable constituents of entire plants. In Wadleigh's
experiment nitrate nitrogen varied from 3.1 per cent of the total nitrogen
in the leaves (low nitrogen plants) to 37.5 per cent in the fibrous roots
(high nitrogen plants). Mason and Phillis (1945) obtained a high linear
correlation between soluble and protein nitrogen in leaves until a rela-
tively high level of supply was reached, beyond which there was no
further increase in protein. Potassium and phosphorus starvation both

COTTON

caused reductions in the proportion of protein. The total nitrogen of
cotton fibers has been found to be correlated directly with the soluble
nitrogen of cotton leaves and inversely with the protein nitrogen (Eaton,
1947). Fine fibers are higher in nitrogen than are thick-walled fibers,
reflecting, probably, a higher ratio of protoplasmic residue to wall weight
in the former than the latter.

From a series of greenhouse comparisons of ammonium and nitrate
salts in water cultures, Holley and Dulin (1943) concluded that there
were no wide or very consistent differences in the yield benefits, or in
effects on growth, between the two forms of nitrogen. They pointed out
that ammonium fertilization initiated more flowers, but more of these
flowers were shed. Although these conclusions are valid, the data through
the 7 experiments reported in their Tables V, VII and XII show trends
in favor of the ammonium salts in fresh weight of plants and in bolls
per plant, and also in relative fruitfulness (computed by the writer)
that seemed worthy of testing by analyses of variance. As summarized
in Table III, ammonium salts produced a nonsignificant increase in
combined weight of leaves and stems, but highly significant increases in
bolls per plant and in relative fruitfulness; the latter amounted to 12
per cent. All of these effects are in the direction to be expected on the

TABLE III

Relative Fruitfulness of Cotton Plants in Several Experiments Involving Nitrogen

Nitrogen
source
meq. per 1.

Fresh stems
and leaves,
g.

Bolls per
plant,
number

Relative
fruit-
fulness *

Holley and Dulin (1943)

NH 4

1997

77.6**

4.4*

(7 experiments)

NOa

1868

70.3

3.9

Wadleigh (1944)

NOa 0.6

145 b

5.3 C

3.6

NOa 1.8

289 b

10.1

3.5

NOa 5.4

652 b

18.6 C

2.8

NOa-16.1

662 b

23.1 C

3.5

Eaton and Rigler (1945)

NOa 0.5

128

5.7

4.5

Low light

NOa 4.0

319

11.4

3.8

NOa 16.0

328

11.2

3.4

NOa-4.0

239

9.6

4.1

High light

NOa 1.0

107

7.2

6.8

NOa 4.0

328

22.9

6.4

NOs16.0

367

22.9

6.4

NOa 64.0

252

18.4

7.6

* Bolls per 100 g. of fresh stems and leaves.
b Green weight at time of second sampling.
c Number of bolls contributing to seed cotton.
*,** Significant at 0.05 and 0.01 level, respectively.

22 FRANK M. EATON

basis of the extra energy required for the reduction of nitrate ions. The
literature on nitrate and ammonium nutrition, as well as many other
features of nitrogen nutrition of green plants, has been extensively re-
viewed by Nightingale (1948).

4. Carbohydrates, Nitrogen and Fruitfulness

As fruit setting progresses, the cotton plant may become a victim of
its own morphological development. As fruiting progresses, nitrogen is
translocated to the bolls at a greater rate than it is taken up from the
soil. As the reserves within the plant are exhausted there is a yellowing
and a progressive reduction in the size of leaves and length of internodes ;
with heavy fruiting all terminal growth may stop. The diversion of
sugars to the bolls reduces the flow into the fibrous roots thereby reducing
nitrogen uptake as a consequence of the lessening of the essential meta-
bolic activity. The onset of nitrogen exhaustion is delayed as the ex-
ternal supply becomes more abundant, and also when the variety is
indeterminant in its growth habit. Wadleigh (1944) believed that high
respiratory activity and reduced carbohydrate supply was involved in
the low productivity of plants grown in a greenhouse with high tempera-
tures and low light intensity. The level of starch and dextrin decreased
with increased nitrate supply. Crowther (1944), in accord with some
of his earlier colleagues, concluded that the number of flowers produced
by the cotton plant depends on nitrogen supply, but that their continued
growth (i.e., boll retention) depends on carbohydrate supply.

Experiments by Eaton and Rigler (1945) were conducted with the
objective of learning whether in cotton there are particular relations
between nitrogen and carbohydrate levels that are conducive or non-
conducive to fruitfulness. Plants were grown in sand cultures supplied
with 1, 4, 16, and 64 meq. nitrate per liter: (1) in a greenhouse in the
winter where daily maximum light intensities were arranged to average
about 1000 foot-candles, and^ (2) freely exposed outdoors in the summer
where the light averaged about 10,000 foot-candles at midday. Between
the low and high light intensities there was an increase in sugar and
starch at all nitrate levels. In the plants supplied with 16 meq. N0 3 per
1. this increase was 4-fold in the leaves and 2-fold in the root bark. The
weight of leaves and stems in the high and the low light experiments
(Table III) were alike, but the plants under high light produced
twice as many bolls as did those under low light. Factors associated
with very low light thus caused decreased fruitfulness, i.e., influenced
the partition of growth materials between vegetative and fruiting activi-
ties in favor of the vegetative. Compared with light, the effects of level
of nitrogen supply on the partition of carbohydrate utilization between

COTTON 23

the two growth activities were found to be minor, i.e., in this experiment,
as in Wadleigh's (Table III), additional nitrogen, when not in excess,
caused proportional increases in growth and fruiting. At both very low
and very high nitrate levels relative fruitfulness was increased slightly.
This may be partially accounted for by a tendency toward smaller bolls
at these levels, but primarily by a repressed development of vegetative
branches. The plants supplied with 16 meq. N0 3 per 1. accumulated
more nitrogen and contained more starch than those supplied with 4
meq. N0 3 , but they set no more bolls.

In the foregoing experiment there was much floral bud shedding under
low light and much boll shedding under high light. An important but
as yet unanswered question arises from this experiment: If carbohydrate
deficiency is a general cause of boll shedding why were the supplies found
in the high light plants not more nearly utilized before these plants
started to shed? Any satisfactory explanation of the cause of boll shed-
ding on nutritional grounds will evidently need to go beyond the often
repeated carbohydrate and nitrogen theory. This is not to imply thq,t
an adequate and continuous supply of carbohydrate is not essential for
the maintenance of boll growth and for boll retention, but rather that
within the nutritional interpretation there are significant points of plant
composition including enzymes and hormones, that have not yet been
brought to attention by laboratory analyses. One of the most direct
supports for the nutritional interpretation of shedding is the fact that
defruiting of heavily laden cotton plants increases carbohydrate levels,
causes renewed vegetative growth, and renewed boll setting.

Rather extreme or prolonged reductions in light intensity have been
found by Dunlap (1945) to result in shedding, but he also noted that
short periods of heavy shading and longer periods of light shading did
not produce this result. It seems a little improbable that any but the
most unusually long or intense periods of cloudy weather are much of a
factor in shedding. This conclusion is in accord with early data by
E. C. Ewing in Mississippi and by Mason and Maskell in Trinidad. It
is even possible that short periods of overcast skies may favor carbo-
hydrate utilization and thereby boll retention. It is now generally recog-
nized that, as late as midsummer, an occasional burst of shedding may
be compensated for by the setting of new flowers. But in cotton, as in
other plants, periods of dark weather cannot be considered apart from
the rankness of the growth of the planting. Five per cent, or less, of the
light intensity at the top of the plants may be found near the ground
under a heavy growth of cotton.

24 FBANK M. BATON

5. Effect of Drought on Plant Composition and Fruiting

Drought, as studied by Eaton and Ergle (1948), was found to cause
an increase in hexose sugars in cotton leaves and large reductions in
starch. In the stems and roots, on the other hand, there were always
moderate to large increases in hexoses, sucrose, and starch. The data
show that the utilization of photosynthetic products in growth is cur-
tailed more by drought than is photosynthesis. Although the results
have as yet not been published, measurements by the same investigators
have shown that reduced water supply, even though decreasing vegeta-
tive growth by half and reducing boll periods and boll sizes, was without
appreciable effect on relative fruitfulness.

In an investigation of the organic acids of the cotton plant, Ergle
and Eaton (1949) found relatively high concentrations in the leaves and
lesser amounts elsewhere. The concentrations of these acids changed
little in leaves during prolonged respiration and they were not exten-
sively translocated. Drought caused an extensive reversible shift from
citric to malic acid. Little, if any, correlation was observed between
the organic acids and other determined organic constituents. Defruit-
ing had little effect on these acids.

6. Drought and Other Factors Affecting Boll Development and

Lint Properties

The effects of nitrogen and mineral deficiencies that are so marked
on the plant are not necessarily reflected in effects on fiber properties.
Instead, the influences on the latter are inclined to be minor or irregular.
Once the ovules have been fertilized and their growth has been initiated,
the boll occupies a favored nutritional position.

Increases in nitrogen supply under certain conditions (Wadleigh, 1944
and Nelson, 1949) have resulted in increases in length of fiber. This
response, however, does not seem always to occur (Sturkie, 1947), and
length reduction has been accredited to the use of nitrogen alone (Brown,
1946). Nelson (1949) reported that phosphate applications increase
boll size, but have little effect on the lint. Under conditions where potash
greatly increased yield, it also increased fiber length, weight per inch,
and x-ray angles, and there was an accompanying decrease in fiber
strength; yarn strength was reduced 5 to 15 per cent by 90 Ibs. per acre
of K 2 0. Fertilizer studies in Texas by Hooton et al. (1949) showed
that a high level of phosphorus increased fiber length in comparison
with a high nitrogen fertilizer, but not when compared with no fertilizer.

Drought, as reported by Barker (1946) and Sturkie (1947), has been
observed repeatedly to decrease the length of cotton fibers and usually,

COTTON 25

but not always, to increase their strength. Berkley et al. (1948) show
that the increase in tensile strength associated with drought is accom-
panied by a narrowing of the angle between the long axes of the cellulose
crystallites and that of the fiber, i.e., reduced x-ray angles.

According to investigations by Eaton et al. (1946) immaturity and
drought have similar effects on the composition of cotton seed. Both
drought and disease injury reduced substantially the percentage of oil
in seed and weight per seed, but left unchanged the percentage of protein.
Earlier investigators have shown that oil is synthesized in seeds during
the late stages of development whereas the input of nitrogen is continuous
from anthesis.

In a number of cottons, Berkley (1945) has found the fiber strength
per unit weight to increase only gradually after it is 35 days old and
that this increase is about what might be expected on the basis of changes
in the x-ray angles.

Anderson and Kerr (1943) have shown the enlargement of young
bolls to be uninhibited by severe wilting of the plant, but full size bolls
shrank during plant wilting and regained their size during the night.
They concluded that a lack of equilibrium between osmotic pressures
and diffusion pressure deficits in cottonseed was more apparent than
real. Kerr and Anderson (1944) concluded that imbibition is largely
responsible for water absorption by developing seed.

7. Oxygen Requirements for Root Growth

Leonard (1945), having observed marked correlations between oxygen
supply, texture and moisture content of Mississippi soils and the dis-
tribution of cotton roots, undertook more extensive controlled laboratory
experiments (Leonard and Pinckard, 1946). Young cotton plants were
grown with their roots extending into glass tubes of nutrient solutions
through which various gas mixtures were bubbled. The minimum oxygen
in the gas mixture required for elongation was between 0.5 and 1 per cent.
The optimum range was found to lie between 7.5 and 21 per cent. The
greatest root growth in any experiment was observed to be with 21 per
cent oxygen and 10 per cent carbon dioxide. The elongation of the tap
root was similar whether nitrate or ammonium nitrogen was supplied.
The absence of carbon dioxide did not affect root growth and 60 per cent
of this gas prevented growth.

26 JOHN T. PBESLEY

IV. DISEASES OF COTTON

JOHN T. PRESLEY

Mississippi Agricultural Experiment Station, State College, Mississippi

Cotton disease investigations have been in progress in the cotton-
growing areas of the United States since before the turn of the century.
Pammel (1888) reported that the root-rot disease of cotton was caused
by a fungus and was not a result of unfavorable soil conditions or of
chemicals or other materials in the soil. Atkinson (1892) reported that
cotton wilt was caused by a vascular-invading Fusarium. Orton (1900)
was the first to breed for wilt resistance in Upland cotton (Gossypium
hirsutum) and produced two wilt-resistant varieties, DILLON and DIXIE.

Since these early disease investigations in cotton, numerous workers
have studied the various diseases attacking the cotton plant and have
made notable contributions to an understanding of the disease problems
and to methods of control.

1. Seed Treatment

Workers in most of the cotton-growing states have cooperated dur-
ing the past several years in developing a uniform seed-treatment pro-
gram. As a result of this work, we are now able to recommend with
confidence the seed treatment that a farmer should use, regardless of
his location in the Cotton Belt. Those most commonly used are ethyl
mercury p-toluene sulfonanilide, 7.7 per cent active ingredient (Ceresan
M), at l l /2 oz. per bushel for fuzzy seed and 3 oz. per 100 Ibs. of delinted
seed, and zinc trichlorophenate, 50 per cent active ingredient (Dow 9-B),
at l l /2 oz. per bushel of fuzzy seed and 3 oz. per 100 Ibs. of delinted seed.
In addition to controlling pre-emergence and post-emergence damping-
off, these materials will disinfect the surface of the seed and eliminate
seed-borne diseases such as bacterial blight and anthracnose.

#. Phymatotrichum Root Rot

The first systematic work on the cause and control of cotton root
rot was done by Pammel (1888, 1889). Various causes for the disease
were suggested, such as certain chemical or physical conditions of the
soil, an excess of humic acid, an excess of lime, an excess of sulfuric acid,
insufficient drainage, or an impervious stratum of clay or limestone un-
derlying the plants that arrested the growth of the taproots.

Pammel (1889) isolated the causal organism from diseased cotton
plants and definitely established the fact that root rot is caused by a
fungus. A series of experiments was then set up to determine whether

COTTON 27

the disease organism was seed borne, and to find methods of control.
It was concluded that the disease organism is not seed borne. Recom-
mendations made for the control of the disease were : good cultural prac-
tices and rotation with nonsusceptible crops such as corn, sorghum, millet,
wheat, and oats; trench barriers, particularly in orchards and vineyards;
various chemicals; and heavy applications of barnyard manure.

Many methods of control have been tried by every worker since
Pammel but none of them has been satisfactory for the entire region
where the disease is a problem.

A variety of intensive chemical treatments ranging from common
table salt to kerosense oil have been used with varying success. Only
a few of the more promising ones will be mentioned here. King (1923)
found that a solution of formalin in dilutions of 1% to 2 per cent was
effective in eradicating the disease from small areas. The soil should
be saturated to a depth of more than a meter. Streets (1938) recom-
mended applications of ammonium sulfate or ammonium phosphate, at
a rate of one Ib. of ammonium salt to 10 square feet of surface, for
treating and protecting ornamental shrubbery and shade trees. Neal
et al. (1932) used ammonium hydroxide with good results. The soil
around affected plants was saturated with a 6 per cent solution by flood-
ing or by pressure methods.

Clean fallow, followed by deep tillage and rotation with nonsus-
ceptible crops, is effective in reducing the disease, particularly in Texas.
McNamara and Hooton (1930) state that, in a plot where more than
90 per cent of the plants were infected from 1919 to 1921 inclusive, no
diseased plants were found after a 2-year continuous fallow.

Numerous types of barriers have been developed to check the spread
of the disease. The more common types are: open trenches 12 to 20
inches in width and 18 to 30 inches in depth dug just in advance of the
front lines of infection; trenches filled with mixtures of sand and heavy
motor oils or sand and chemicals; strips of sheet metal or roofing paper
placed vertically in the soil; and as used by Taubenhaus and Ezekiel
(1935), barriers consisting of 2 or 4 rows of sorghum (a nonsusceptible
plant) planted in advance of the line of infection. The most effective
is the trench barrier containing mixtures of soil and heavy oils, salt,
ammonia, and sulfur.

Fertilizers high in nitrogen were shown by Jordan et al. (1939) to
reduce the incidence of root rot under some soil conditions. The results
of various cultural practices employed in Texas are summarized by
Jordan et al. (1948).

No promising results in the direction of disease resistance have been

JUIIN T.

reported from breeding experiments on cotton, but some of the selections
of grape and citrus in Texas appear to be resistant.

Control of root rot has been obtained by the use of stable and corral
manures where heavy applications of these and other organic materials
have been made in deep furrows during the fall and winter, and cotton
has been planted over them. King (1937) states that, in 1935, on plots
treated with manure 1.6 per cent of the plants died, while on the un-
treated plots 56.2 per cent of the plants died. For one acre, 20-40 tons
of green alfalfa or 8-15 tons of barnyard manure were used. More
recently, Lyle et al. (1948) reported a practical control for Phymato-
trichum root rot in Texas by use of a sweet clover and cotton rotation.

By the use of organic materials, soil conditions are created which
favor the rapid development of certain soil organisms which in turn
hinder the development of the root-rot fungus. Since these organisms
are present in almost every type of soil, the author believes that this
method of control will apply to all the areas affected by the disease.
The only problem will be to find the most efficacious method of applying
the organic materials for each locality or soil type.

3. Fusarium Wilt

A cotton wilt caused by a form of Fusarium was reported by Atkin-
son (1892) ; the disease was described as Fusarium vasinfectum n. sp.
It was observed to be of general occurrence on the lighter soil types of
the Southeastern cotton-growing states. At the present time, the disease
is known to occur in all parts of the world where cotton is grown.

Orton (1900) is credited with breeding the first wilt-resistant Upland
cotton. For a period of 8 years, healthy plants were selected from wilt-
infested fields where most of the plants had been killed. The selections
were tested under wilt conditions in the field and from these the two
resistant varieties, DILLON and DIXIE, were developed. Lewis and Mc-
Lendon (1917), working in Georgia, developed several resistant varieties.
Apparently most of the original resistant varieties were of the late-
maturing type since many of them have been discontinued. Also, as a
result of boll weevil damage to the late varieties, as Sherbakoff (1949)
points out, breeding work in the U.S. Dept. of Agriculture, became di-
rected to the production of earlier and more productive wilt-resistant
varieties by crossing resistant DILLON and DIXIE with the earliest and
most productive susceptible varieties such as TRIUMPH, COOK, COLUMBIA,

COKER, WEBBER, and FOSTER.

Shortly after the beginning of the wilt-resistance work in the United
States, Fahmy (1929) started the breeding of a resistant Sakel cotton
(Gossypium barbadense) in Egypt. About this same time Uppal et al

COTTON 29

(1941) began similar work in India. All of these workers, through
the results which they obtained and through the influence which their
work had upon subsequent investigations, have contributed materially
to the ultimate success in obtaining a satisfactory wilt-resistant cotton.
An outstanding example of this relatively recent work is the develop-
ment of COKER 100 WILT and COKER-4-IN-1 by the Coker Pedigreed Seed
Company. These cottons have been widely accepted and planted on
wilt-infested soil and as a result the losses from Fusarium wilt have been
reduced to a minimum. The most recent contribution in this field has
been the EMPIRE cottons developed by Smith and Ballard (1947) at the
Georgia Experiment Station. In respect to resistance, earliness, and
productivity EMPIRE stands near the top of all the commercial Upland
cottons.

4. Verticillium Wilt

Verticillium wilt of cotton caused by Verticillium albo-atrum R. & B.
was first reported by Carpenter (1914). No further mention of this
disease was made in the United States until it was reported by Sher-
bakoff (1929) as causing considerable damage to cotton in Tennessee.
Miles and Persons (1932) reported the disease as occurring on cotton
in the Mississippi Delta, and Herbert and Hubbard (1932) reported the
disease in cotton at the U.S. Field Station, Shafter, California.

Shortly thereafter, Brown (1937) reported the disease as occurring in
all of the cotton-growing areas of Arizona. At the present time, Verticil-
lium wilt is known to occur in cotton across the entire Cotton Belt from
South Carolina to California.

In the early stages of Verticillium wilt investigations, the disease
appeared to be more of a novelty than a potential destructive disease.
Beginning about 1937 to the present time, however, the disease has in-
creased in severity and in the total area affected to the point where it is
recognized as one of the major diseases of cotton. Certain areas of the
irrigated Southwest suffer losses that range up to 50 per cent, with over-
all losses ranging up to 20 or 25 per cent. Efforts were made by Rudolph
and Harrison (1939) to control this disease by the application of chem-
icals or soil amendments, but at the present time none of these practices
appears to be very promising. Breeding for disease resistance offers
the most practical method of control. Rudolph and Harrison (1939),
Presley (1946), and Barducci (1942) have contributed to this phase of
the problem. Salter (1946) reported the release of a resistant strain of
ACALA 1517 which was developed by Leding at the U.S. Cotton Field
Station, in New Mexico. Strains of COKER 4-iN-l and EMPIRE appear to
have considerable tolerance to the disease in Mississippi, whereas the

30 JOHN T. PRESLEY

cottons developed in the irrigated Southwest are very susceptible. The
breeding and selection work is being continued in California, Arizona and
New Mexico, as well as in Mississippi, where a variety, HARTSVILLE, has
been found to be highly resistant to wilt under conditions which obtain
in the Mississippi Valley. The late maturity and low yield of HARTSVILLE
renders it undesirable commercially, but it offers excellent possibilities
in a back-cross program for transferring resistance to the desirable com-
mercial types. A selection and hybridization program has been under
way for the past 4 years in Mississippi and notable progress has been
made. At the present time progenies are ready for yield trials in Missis-
sippi and for further resistance tests in other cotton-growing areas.

5. Bacterial Blight

Bacterial blight caused by Xanthomonas malvacearum (E. F. Sm.)
Dowson is one of the most common diseases of the cotton plant. It
attacks practically all varieties of Upland cotton and it is especially
severe on varieties of Sea Island and Egyptian cotton (G. barbadense).
The disease was first described by Atkinson (1891). Atkinson (1892)
published a rather complete description of the disease and called it
angular leaf spot. Smith (1901) reported that the disease was caused
by a bacterium. Xanthomonas is capable of affecting all above-ground
parts of the cotton plant and, according to the organ affected, the disease
is known as angular leaf spot, black arm, or bacterial boll rot. Although
the disease is present in all cotton-producing countries, crop losses vary
with varieties grown, seasonal conditions, and with the region in which
the cotton is produced. In the irrigated valleys of the Southwest, the
disease is especially severe and crop losses ranging up to 25 per cent are
not uncommon in certain areas.

Since the bacterium was found to be carried on the lint and on the
surface of the seed, Rolfs (1915) recommended the use of sulfuric acid
as a means of removing the lint and disinfecting the seed surface. Brown
and Streets, University of Arizona, perfected and patented a sulfuric
acid process for delinting cotton seed in 1934. Although the bacteria
which are carried on the lint and seed surface are removed by the sulfuric
acid treatment, carefully controlled experiments by many workers have
demonstrated that there is a certain amount of infection carried inside
the seed coat which is not controlled by the sulfuric acid or other seed
treatments. Results of the Uniform Seed Treatment Studies which have
been conducted for many years in most of the cotton-growing states
demonstrate that the recommended seed protectants are also effective in
controlling the surface-borne seed infections of bacterial blight. Seed
treatment reduces primary infections, but cannot be expected to entirely

COTTON 31

control the disease inasmuch as a small percentage of infection is carried
within the seed. The centers of infection resulting from the internal
infection may pass unnoticed in the field, but if favorable conditions
arise considerable spread may be expected. Much work has been done
on survival, dissemination, and spread of the causal organisms (Faul-
wetter, 1917; Hansford et al., 1933; Hare and King, 1940; Massey, 1930;
Rolfs, 1935). Stoughton (1933) has described the effects of environ-
mental conditions upon the disease.

It is obvious that the most practical approach to complete control
of bacterial blight is through the development of disease resistant
varieties. Fortunately, a resistant variety of Upland cotton has been
found. Simpson and Weindling (1946), working with many varieties of
cotton, found one selection of STONEVILLE to be resistant. This resistant
selection was designated STONEVILLE 20. Since the release of STONEVILLE
20 many workers have used it in transferring resistance to many other
varieties of Upland cotton. Blank, at College Station, Texas, has trans-
ferred resistance to the commercially desirable varieties of cotton which
are grown in Texas. Other workers in Tennessee, Mississippi, New
Mexico, Arizona and California are using STONEVILLE 20 in their breed-
ing programs in order that resistance to bacterial blight may be trans-
ferred to the new varieties which are being developed.

6. Root-Knot

Root-knot caused by Heterodera marioni (Cornu) Goodey was early
recognized by Gilbert (1914) as a serious disease of cotton in itself, and
also as a contributing factor to the losses in wilt-infested soils, due to
the fact that Fusarium wilt of cotton is much more severe where the
disease is associated with root-knot. Godfrey (1923, 1943) recommends
clean fallow, rotation with nonsusceptible crops, and summer plowing
for controlling root-knot. Watson and Goff (1937) and Watson (1945)
in Florida, reported control by use of mulches. Cuba (1932) suggests
the use of a carbon disulfide emulsion in controlling the disease and
recently almost innumerable reports have been made of the success of
other chemicals in controlling the disease. Jacks (1945) used formalde-
hyde, carbon disulfide, chloropicrin, and a mixture of 1-3 dichloropropane
and 1-2 dichloropropane (Shell D-D), and found that chloropicrin and
Shell D-D gave the most promising results. Smith (1948) found that
ethylene dibromide (Dow W-40) gave considerable increases in yield
when used on wilt-sick soil infested with nematodes. Presley (1949) also
obtained consistently high yields from plots treated with ethylene di-
bromide on wilt-infested soil also infested with root-knot nematodes.

Through the use of a simplified method of application and row treat-

32 J. C. GAINES

ment rather than blanket application, the chemical control of nematodes
is economically feasible. Studies conducted at the Mississippi Agri-
cultural Experiment Station have shown considerable carry-over effect
of the soil fumigant from one year to the next. Cotton grown in 1949
on plots fumigated in 1948 produced more than double the nonfumi-
gated plots. At the present price of cotton, soil fumigation should be a
profitable practice wherever nematodes are a serious problem.

7. Summary

Although many of the fundamental principles of disease control were
pointed out and investigated by earlier workers, it has been during the
past 10 to 15 years that outstanding progress has been made in develop-
ing control measures for the most important diseases of this crop. At
the present time practical control measures are available for the seedling-
disease complex commonly referred to as damping-off, for Phymato-
trichum root rot, for Fusarium wilt, for bacterial blight, and considerable
progress has been made in controlling Verticillium wilt. Controls have
also been developed for the root-knot nematode.

V. INSECT PESTS

J. C. GAINES

Texas Agricultural and Mechanical College System, College Station, Texas

Statistics released by the Bureau of Agricultural Economics and the
National Cotton Council indicate that insects cause an estimated average
annual loss of $208,727,000 to cotton planters of the South. The boll
weevil has been the most damaging insect for many years. Bollworms,
leafworms, fleahoppers, aphids, thrips, stinkbugs and other insects are
also responsible for much injury.

In recent years, research entomologists have made considerable
progress toward the development of chemical controls by employing
better laboratory techniques plus the latest designs for field experiments.
Newly developed insecticides are tested first in the laboratory, under
controlled conditions, and later are placed in field tests. The field tests
are designed for statistical analysis of the infestation data as well as
yields, in order to evaluate the materials.

1. Thrips

As soon as cotton emerges in the spring, injurious insects begin their
attack. Several species of thrips, Frankliniella tritici (Fitch), Frank-
liniella fitsca (Hinds) and Thrips tabaci, Lind. cause similar injury to

COTTON 33

cotton. Wardle and Simpson (1927) stated that thrips cause premature
and excessive defoliation. They found no evidence, however, that the
salivary secretion of the insect was toxic. Eddy and Livingston (1931)
showed that thrips retarded the growth of cotton seedlings and that un-
folded leaves were perforated and had marginal erosions. Gaines (1934)
observed that thrips transferred in large numbers from weed fields to
cotton early in the spring, and that later populations increased on cotton
at a rapid rate. Watts (1934, 1936) reported the biology of several
species of thrips known to attack cotton in South Carolina. Thrips
injure the tender leaves, destroy buds and retard fruit production. The
injury becomes so severe at times as to result in loss of stands.

Chapman et al. (1947), Fletcher et al. (1947) and Gaines et al. (1948)
have conducted thrips control tests. The results show that^ thrips may
be controlled with several organic insecticides; however, increases in
yield were not significant. Losses in stands can be prevented by use of
insecticides.

8. Cotton Aphid

Aphids, Aphis gossypii, Glov., also attack seedling cotton and are
especially injurious during cool damp weather. This pest retards growth
and fruit production and often results in loss of stand. Aphids may be
controlled with benzene hexachloride, but insecticidal control applied
to seedling cotton rarely proves economical except in cases where losses
in stands are prevented.

Infestations of aphids resulting from the use of calcium arsenate
later in the season cause excessive shedding of leaves, difficulty in gin-
ning and deterioration in the grade of the lint. R. C. Gaines et al. (1947)
reported that control of aphid infestations by the addition of nicotine
to the calcium arsenate used for weevil control proved economical.

3. Cotton Fleahopper

The fleahopper, Psallus seriatus (Reut.) also attacks cotton early in
the season. Howard (1898) reported this insect as a pest of cotton.
Hunter (1924) gave a brief account of severe injury to cotton in South
Texas caused by the then-called "cotton flea." Reinhard (1926a, 1926b,
1927) described the various stages of the insect and suggested the use
of sulfur to control the pest. The injury to the cotton plant is char-
acterized by an excessive blasting and shedding of small squares, a re-
duction in the number of fruiting branches, and either a tall whiplike
growth of the main stem or an increased number of vegetative branches.
Several workers (Gaines, 1933, 1942; Hunter and Hinds, 1904; Rein-
hard, 1926b) have shown that the fleahopper migrates from horsemint to

34 J- C. GAINES

cotton early in the spring and migrates from cotton to croton early in
July. Ewing (1929) and Painter (1930) investigated the possibility of
the fleahopper being a vector of a plant disease. They concluded that
the material injected into the plant by the insect did not spread far
from the point of injury and that both the disturbance and shedding of
squares was due to the multiplicity of bites. The fleahopper occurs
throughout the Cotton Belt, causing the greatest damage in the western
part of the area. Ewing (1931) noted that a mixture of Paris green and
calcium arsenate killed a higher percentage of adult fleahoppers than did
sulfur; however, sulfur proved the more effective in killing nymphs.
Later, Ewing and McGarr (1936, 1937) showed that Paris green-sulfur
(10:90) was more effective against this pest than sulfur alone.

Most of the newer organic insecticides have proven highly effective
against all stages of this pest. Low concentrations of either toxaphene or
DDT applied as sprays or dusts are effective and have proven practical
for the control of the fleahopper.

4. Boll Weevil

The boll weevil, Anthonomus grandis, Boh., crossed the Rio Grande
near Brownsville, Texas, on or before 1892, and by 1894 it had spread
through several surrounding counties in Southern Texas. The weevil
gradually spread to the northwest and eastward and by 1922 had cov-
ered practically the entire Cotton Belt. Several workers (Fenton and
Dunnam, 1928; Gaines, 1942; Hunter and Hinds, 1904; Hunter and
Pierce, 1912) have shown that there are two principal periods of dis-
persal and spread during the season. The first period occurs when the
hibernating weevils leave their winter quarters and go in search of food.
The second period is dependent on several factors: (1) large weevil
population; (2) abundance of fruit; (3) high percentage of infested fruit;
and (4) high temperatures. The hibernating weevils migrate to cotton
early in the season, feeding on the leaves and small squares. When the
squares become half -grown or larger, the female deposits an egg in a
cavity which has been formed by eating into the square or boll. The
cavity is then sealed by secreting a muccilaginous substance from ac-
cessary glands of her female organs. The feeding punctures are never
sealed, thus differentiating between those containing eggs. Either kind
of puncture soon causes the squares to flare and fall. Heavy weevil
infestations result in serious boll injury as well as square injury caused
by the grubs feeding on the contents of the fruit.

Newell and Smith (1909) recommended powdered arsenate of lead
for the effective control of weevils. Later, Goad (1918) and Coad and
Cassidy (1920) reported that calcium arsenate afforded good control

COTTON 35

of the weevil without causing injury to the cotton. Research work has
been conducted on the boll weevil in practically all Southern States and
numerous reports of this work have been issued by various Experiment
Stations and the Bureau of Entomology and Plant Quarantine. Calcium
arsenate, since discovery of its applicability, has been the recommended
control for this pest throughout the South. Since the use of this insecti-
cide often results in aphid infestations and is not effective against the
sucking type of insects which are injurious to cotton, an effort has been
made to find a desirable substitute by both manufacturers and entomol-
ogists. The development of organic insecticides during World War II
offered several possibilities. A number of workers (Becnel et al.,
1947; Dunnam and Calhoun, 1948; Ewing and Parencia, 1947, 1948;
Gaines and Dean, 1947, 1948; Gaines and Young, 1948; Ivy and
Ewing, 1946; Ivy et al., 1947; Parencia et al., 1946; Rainwater and
Bondy, 1947; Watts, 1948) have shown that toxaphene or a mixture of
benzene hexachloride and DDT were effective against the boll weevil,
as well as most of the other cotton insect pests. Calcium arsenate or
calcium arsenate mixed with an aphicide such as nicotine is still recom-
mended for weevil control in all the cotton states. However, 20 per
cent toxaphene-40 per cent sulfur or 3 per cent gamma benzene hexa-
chloride-5 per cent DDT-40 per cent sulfur are preferred to calcium
arsenate, particularly in those states where the bollworm causes injury.

Spraying cotton with arsenicals for weevil control has not proven
profitable. Results of tests conducted in several states in 1949 indicate
that some of the organic insecticides applied as spray emulsions at a
low pressure and volume per acre were effective against the weevil.
This method of application is being rapidly explored by cooperative
efforts of agricultural engineers and entomologists.

Several years ago, when the pink bollworm spread to the southern
counties of Texas, strict regulations were imposed on the planters re-
garding planting dates and early fall destruction of stalks. The fall
stalk destruction date was set sufficiently early to starve the boll weevil
before it hibernated. During the years following adequate fall destruc-
tion of stalks, the boll weevil was greatly reduced and the need for chem-
ical control was practically eliminated. Gaines and Johnston (1949)
reported the results of a fall stalk destruction program which was con-
ducted in Williamson County, Texas, during 1947. In this county, the
stalks were destroyed early by all planters and the weevil infestations
were greatly reduced the following year. Apparently a well executed
fall destruction program will greatly reduce the losses resulting from
the weevil.

The so-called "Florida Method" involved the removal and destruction

36 J. C. GAINES

of all infested squares early in June after the hibernating weevils had
emerged, followed by dusting the plants with calcium arsenate to destroy
the adults. Several factors made this method impractical. The pre-
square poisoning method was developed and used to some extent in the
eastern portion of the Cotton Belt and proved more effective in com-
munities where all the planters cooperated in the program. However,
neither method has been generally accepted by cotton planters.

Since several of the organic insecticides have proven effective both
in the form of sprays and dusts for controlling such insects as thrips,
aphids, fleahoppers and boll weevils which usually attack cotton early
in the season, the control of these insects appears both practical and
promising. Spray equipment is being developed which will allow the
insecticide to be applied when the cotton is cultivated. By combining
insecticidal applications with the usual cultivation practices the expense
of early-season control should be greatly reduced. In certain com-
munities where planters have cooperated in an early-control program,
the results have been favorable. During dry years this program should
prove profitable. In wet years and in areas where the bollworm causes
injury later in the season, however, several applications of insecticides
will be necessary to protect the cotton and the early season control may
be an added expense. In either case, early season control is a good
investment to insure early fruiting of cotton. The Texas Extension
Service suggested the early-season control program during 1949. In
some areas, it was well received and proved profitable. In other areas,
where bollworm infestations occurred, additional applications were neces-
sary to protect the crop during July and August. Cotton receiving the
late applications produced as much as the cotton receiving both early
and late applications. Additional research is being conducted to evalu-
ate the various programs in different areas.

^ 5. Bollworm

The bollworm, Heliothis armigera (Hbn.), occurs over the entire Cot-
ton Belt as a pest of many crops. This insect overwinters in the pupal
stage and the moth emerges early in the spring. The first brood feeds
on legumes and corn, while the second brood attacks corn causing consid-
erable injury to the ears. In the Southwest, corn matures early in July
and the moths emerging from the mature corn fields migrate to cotton
fields and severely damage the more succulent cotton. Sporadic oc-
currences of this pest have been reported throughout the entire Cotton
Belt. Riley (1885) studied the bollworm and recommended London
purple and Paris green for its control. Later Quaintance and Brues
(1905) recommended arsenicals, the use of trap crops, and cultural prac-

COTTON 37

tices to control the pest. Moreland and Bibby (1931), Gaines (1941,
1944) and Moreland et al. (1941) presented results of insecticidal tests
which indicated that calcium arsenate applied at the proper time gave
economical control of the pest. Parencia et al. (1946), Ewing and Par-
encia (1947), Gaines and Dean (1947) and Gaines et al. (1948) reported
the results of field tests in which DDT, 3 per cent gamma benzene hexa-
chloride-5 per cent DDT-sulfur and 20 per cent toxaphene-sulfur were
compared with calcium arsenate. The organic insecticidal mixtures
proved more effective in controlling the bollworm than calcium arsenate.
Unpublished results of tests conducted in 1949 indicate that toxa-
phene and toxaphene-DDT, when applied as spray emulsions, are effec-
tive in the control of the bollworm.

6. Pink Bollworm

The pink bollworm, Pectinophora gossypiella (Saund.) is a worldwide
pest of cotton and causes severe injury to the fruit. The greatest dam-
age caused by this pest is the destruction of bolls or rendering them
unfit for picking. Cotton produced under heavily infested conditions
is inferior in grade because of staining, shorter staple and less tensile
strength.

The pink bollworm was first described from specimens collected in
India and is believed to have spread by means of seed shipments to Egypt
around 1906. According to Hunter (1918, 1926b) the pink bollworm
was introduced into Mexico in 1911 through shipments of seed from
Egypt. The infestations in this country were introduced from Mexico
likewise through seed shipments, and also by moths drifting across the
border from heavily infested areas.

The biology of the pink bollworm was carefully studied in Mexico
by Loftin et al. (1921) and Ohlendorf (1926). Hunter (1926b) reported
on steps taken to prevent this pest from establishing itself in the United
States. The pink bollworm has been a potential danger to the cotton
industry in this country for the last 28 years.

Latest reports issued by the Bureau of Entomology and Plant Quar-
antine indicate that the pink bollworm has been found to occur in 127
counties in Western Texas, 8 in Oklahoma, 15 in New Mexico and 7 in
Arizona. It has been effectively controlled by the use of cultural prac-
tices including regulated dates for planting and destruction of stalks as
specified by the State Departments of Agriculture.

During the past few years, considerable research has been carried
on in Mexico by workers in the Bureau of Entomology and Plant Quar-
antine to effect a chemical control for the pink bollworm. The results

38 J. C. GAINES

of these tests indicate that DDT is effective in reducing the worm
population.

7. Hemipterous Insects

A number of hemipterous insects attack cotton, particularly in the
western areas of the Cotton Belt. According to Cassidy and Barber
(1939) the most important species of stink bugs causing injury to cotton
are Euchistus impectiventris Stal, Chlorochroa sayi Stal and Thyanta
custator (F.). Three species of plant bugs, Lygus hesperus Knight, L.
pratensis oblineatus (Say) and L. elisus Van D. are also injurious to
cotton. The most conspicuous injury resulting from hemipterous insects
is the blasting of young squares and bolls and the puncturing of bolls
followed by severe lint staining due to invading pathogenic organisms.

Increases in yields have been obtained with applications of Paris
green-sulfur or calcium arsenate-sulfur. Later work conducted by Stev-
enson and Kauffm&n (1948) proves that the organic insecticides are more
effective than the arsenical-sulfur mixtures. Either DDT, a mixture of
DDT-benzene hexachloride or toxaphene is recommended for control of
plant bug and stink bug.

8. Cotton Leafworm

The leafworm, Alabama argillacea (Hbn.) is one of the oldest known
insect pests of cotton. Since this insect cannot overwinter in any part
of the United States, infestations originate from flights of moths from
Central or South America. Almost every year this pest has been re-
ported from some section of the Cotton Belt. It was once, thought that
periods of maximum infestations occurred in 21 -year cycles, but this is
not now generally accepted. The conditions affecting the increase of this
pest in its native habitats, and the conditions existing in this country at
the time the moths appear, govern the injury produced.

This pest is primarily a^leaf feeder. After the leaves have been
destroyed, however, it may also devour the fruit. Due to its feeding
habit, the leafworm is easily controlled with almost any kind of arsenical.
It has been found that, with the exception of DDT, the organic in-
secticides used to control other cotton insects are also highly effective
against this pest.

Leaf worms have not developed to injurious numbers in the last few
years, due perhaps to the thorough dusting programs which have been
generally followed in the coastal area of Texas. However, unfavorable
conditions for leafworm development in South and Central America may
also have been a contributing factor.

COTTON

9. Spider Mites

Sporadic infestations of spider mites have been reported throughout
the Cotton Belt. This pest attacks the underside of leaves and causes a

TABLE IV

Insecticides Recommended by the Various States for Cotton
Insect Control during 1949

Rate

of application,

Insects

Insecticides Ibs. per acre

Thrips

10 per cent toxaphene

12 to 15

3-5-40 '

7 to 10

Aphids

3-5-40

7 to 10

Fleahopper

5 per cent DDT

10 per cent toxaphene

20 per cent toxaphene

3-5-40

Boll Weevil

Calcium arsenate alternated

with calcium arsenate 2 per cent

nicotine or 3-5-40

7 to 10

Calcium arsenate

7 to 10

3-5-40

10 to 15

20 per cent toxaphene

10 to 15

Bollworm

20 per cent toxaphene

10 to 15

3-5-40

10 to 15

10 per cent DDT

10 to 15

Calcium arsenate

10 to 15

Leafworm

Calcium arsenate

7 to 10

Lead arsenate

7 to 10

20 per cent toxaphene

3-5-40

Plant Bugs

5 per cent DDT

10 to 15

3-5-40

10 to 15

10 per cent toxaphene

10 to 15

Stink Bugs

2-5-40 b

15 to 20

3 per cent G. BHC

15 to 20

10 per cent DDT

10 to 15

3-5-40

10 to 15

20 per cent toxaphene

10 to 15

Red Spider

Sulfur

20 to 25

11 3 per cent gamma benzene

hexachloride 5 per cent DDT 40 per

cent sulfur.

b 2 per cent gamma benzene

hexachloride 5 per cent DDT 40 per

cent sulfur.

40 WILLIAM E. MEEK

discoloration and subsequent defoliation of the plants. Sulfur has been
recommended for a number of years as a control for this pest, but the
results obtained by planters have not always been favorable. In a recent
article, McGregor (1948) determined the species of spider occurring on
cotton in Texas as Septanychus sp. The two-spotted spider mite,
Tetranychus bimaculatus Har., is known to be widely distributed in the
southeastern section of the Cotton Belt.

When organic insecticidal mixtures were first used without sulfur,
a decided increase of spider mites was noticed, particularly in the South-
west. It was found that the addition of sulfur to the insecticidal mixtures
averted increases. Results of tests conducted by Iglinsky and Gaines
(1949) indicate that sulfur effectively controls the Septanychus sp. spider
mite.

10. General Recommendations for Chemical Control

The insecticides or insecticidal mixtures generally recommended by the
various states for cotton insect control during 1949 are given in Table
IV. The recommendations issued by the Extension Services of Arkansas,
Mississippi, Missouri, North Carolina, South Carolina and Tennessee do
not include sulfur in the organic mixtures. The entomologists in Ala-
bama, Arizona, California, Georgia, Louisiana, New Mexico, Oklahoma
and Texas suggest that the organic mixtures should contain at least 40
per cent sulfur to prevent red-spider increases.

VI. IMPROVEMENTS IN PRODUCTION PRACTICES

1. In Humid Areas

WILLIAM E. MEEK
U. S. Department of Agriculture, Stonevttle, Mississippi

The rapid advancement of mechanized practices in the humid areas
of the Cotton Belt, due to the shortage of farm labor and to the improve-
ment in equipment, has brought about many changes and improvements
in farming techniques. These can best be described in the order in
which the operations are carried on. Space will not permit any lengthy
comparisons between the most modern practices and those of past years.

a. Field Layout and Water Control. A water control program,
whether terracing for erosion control or a drainage system to remove
surplus water, or possibly a combination of the two, is an absolute
necessity where modern tractors with pneumatic rubber tires are used.
In order to obtain the best in control measures, it is necessary to give

COTTON 41

careful attention to field planning. Fields should be so arranged as to
give the longest rows possible with a minimum of short rows. It may
be well, therefore, to take certain parts of a field from row crops in order
to increase the efficiency of the operation. This field planning may pos-
sibly require the moving of tenant houses and other buildings from the
field, as well as relocation of roads and ditches. Low spots, or pockets,
lend themselves readily to "spot-plowing operations" and in areas where
heavy equipment is available land levelers are being used.

6. Disposal of Crop Residues. The modern power-driven stalk
shredders have greatly simplified the problem of crop residue disposal,
allowing the stalks to be cut to very small pieces, thereby increasing
coverage during the plowing operation and hastening the incorporation
of vegetation into the soil. Many of these same machines are used on
cover crops to shred thoroughly the green growth so that it can be com-
pletely and easily incorporated into the soil. In some instances the use
of the stalk shredder on cover crops will allow the planting date to be
advanced a week to 10 days.

c. Preparation of Seedbed. The manner of seedbed preparation
varies widely with the area; heavy disk harrows, mold board plows, disk
plows, wheatland plows, and middle breakers, are all standard imple-
ments. Regardless of the machine used, a good seedbed is essential.
Improvements in farm equipment enable the farmer not only to increase
efficiency, through the coverage of more acres per day, but improve the
seedbed produced in the plowing operation. Any plant, whether grown
in the garden as a shrub or in a 1000 acre field, retains its characteristics
and requires the same seedbed regardless of the acreage. Farmers realize
the importance of good seedbed preparation and increasingly are making
use of the modern equipment at their command.

d. Fertilization. The use of fertilizer has been on the increase for a
number of years and remarkable strides have been made in both ma-
chines and methods of application. In those areas where complete fer-
tilizer is used at planting time, equipment is now available in units for
one, two, or four rows which apply fertilizer according to the placement
recommendations, simultaneously with the planting operation. This not
only saves labor, but assures a more positive placement in relation to
the seed with better moisture conditions, thereby giving greater efficiency
and more complete utilization of the fertilizer by the plant. Throughout
the humid areas, nitrogen is the principal fertilizer used for sidedressing
the crop, and in some areas is the only fertilizer used. It is then applied

42 WILLIAM E. MEEK

both prior to planting and as a sidedressing. One of the newer nitro-
genous fertilizers is anhydrous ammonia, which in some areas is used in
extremely large quantities for direct application to the soil. Other areas
of the Cotton Belt, particularly in the Southeast, are using a complete
fertilizer at planting followed by anhydrous ammonia as a sidedressing
with excellent results. This particular form of fertilizer has advantages
for sidedressing, inasmuch as deeper placement is possible than when
granular fertilizers are used. The use of anhydrous ammonia is one of
the most important developments in cotton fertilization in recent years.

e. Planting. Increased efficiency in mechanical, acid, and gas delint-
ing of cotton seed has allowed the farmers to attain greater precision in
planting than ever before. Many farmers are now planting their cotton
to a stand, either by hill dropping or in some instances light drilling,
thereby eliminating the costly operation of thinning as well as attaining
a saving in seed planted per acre. Cross plowing or check-row planting,
while seemingly attaining popularity several years back, has become
less popular due to decreased efficiency of the mechanical picker. It
still retains its place, however, in areas badly infested with grass and
weeds, as a means of cheap control of these pests in the cotton crop.
Recent tests from several Experiment Stations indicate that the extreme
precision necessary in planting a crop such as corn is not necessary in
the growth of cotton. The various methods of planting do affect, how-
ever, the fruiting habits of the plant and increase or decrease, as the
case may be, the efficiency of the weed-control measures, which in turn
affect yield and operation of mechanical harvesters.

/. Cultivation. Cultivation rightfully is divided into 3 periods, early,
midseason, and late. It should be borne in mind that cultivation is
primarily for weed control and that the breaking of the land has been
accomplished during seedbed preparation. Rotary hoe attachments,
cited by Gull and Adams (1945) are mounted between the cultivator
gangs and independent of them, and have greatly speeded up and in-
creased the efficiency of early cultivation. Many farmers are setting
their machines on the floor of their shop by the line-diagram method,
and with the use of rotary hoes are attaining speeds up to 5 l /% miles per
hour during the early cultivation period, when timeliness is so impor-
tant. It should be pointed out that these speeds have been possible
through the cooperation of the Farm Equipment Industry in developing
and making available ground-working equipment suitable for high-speed
operations.

During the second stage in the cultivation of the cotton crop, flame

COTTON 43

cultivation is brought into the picture and is carried on simultaneously
with the regular shovel cultivation. Pioneer work with flame by Neely
and Brain (1944) was extended by Gull and Adams (1945). Through
the development of more efficient burners for the flame cultivator in
later years by Meek and other workers of the Mississippi Delta Branch
Station, it is now possible to begin the operation of these machines when
the cotton plant is much smaller than has heretofore been possible.
These new burners, using a standard spray nozzle for an orifice, allow
the size of the orifice to be changed at will, thereby permitting small
orifices to be used on young cotton, and the size of the opening increased
as the cotton plant increases in size. The burners also furnish a longer
exposure of the plant to the flame, which permits higher operating speeds
than has ever been obtainable before. Probably one of the greatest
advantages of these new burners is that they are set at an angle of 45
to the surface of the ground and when once set do not require further
adjustment. Clods and ridges do not particularly affect the action of
the flame and, as a consequence, the flame cultivator has been made
more adaptable to the Cotton Belt as a whole. The price of fuel, which
is either butane or propane, will be the determining factor in the use
of these machines. With flame cultivation controlling the grass and weeds
in the row and the regular shovels of the standard cultivator giving
control in the middles, a greater degree of efficiency is now attained for
humid regions than ever before. Mid-season control is, for the modern
farmer equipped with modern equipment, a comparatively simple matter.
Where mechanical harvesters are used, the late control of grass and
weeds in cotton is vitally important. The grass in particular must not
only be killed, but destroyed above ground. Farmers have found this
extremely difficult to do due to the size of the cotton plant. Great dam-
age has been done in the past by machines moving through the dense
vegetation. Wheel fenders for the tractor, developed at the Mississippi
Delta Branch Experiment Station, now allow cultivation by shovels and
flame much later in the season than ever possible before. In many areas,
it is necessary in the late stages of control to remove the standard shovel
cultivators and continue with the flame alone, particularly when rains
occur during the late growing season. These fenders, while invaluable
in the grass- and weed-control program, also allow the farmers to make
insecticidal applications, and then later to use his own tractor in applying
a defoliant, either dust or liquid.

g. Application of Insecticides. Insect control is of vital importance,
not only from the standpoint of yield, but also because it affects the
mechanical harvesting program. One of the newest developments is the

44 WILLIAM E. MEEK

attachment of spray rigs on supports of the regular shovel cultivator,
whereby organic insecticides are applied during the early stages of
growth, as a control for thrips and other insects. This early spraying
apparently causes the cotton to fruit earlier, which gives more uniform
distribution of the bolls on the plant and allows the mechanical har-
vesters to be started sooner in the fall and so do more work. Control
of boll weevil and other insects is carried on, where sprays are used, with
the same unit used in early applications. These machines are developed
to the stage that the gallonage per acre applied approximates that of
airplane application, and preliminary results indicate that the control
to be obtained by the two methods is comparable. By the use of the
fenders on the tractor wheels, this insecticide treatment may be applied
either as sprays or as dusts well toward the harvest season, or to a
point after which further control is not considered profitable. Insect
control, during the middle and latter part of the season, can greatly
affect the efficiency of the mechanical harvester. Where control is not
obtained, the plant has a tendency to grow tall and rank due to the
dropping of the fruit, and as a consequence, more vegetation must be
handled by the harvesting equipment and its efficiency is thereby
reduced.

h. Defoliation. Defoliation in the humid areas is carried on prin-
cipally with calcium cyanamid dust applied either by airplanes or ground
machines. This practice greatly increases the efficiency of the mechanical
harvester and allows cotton of a higher grade to be produced. Where
hand picking is practiced, defoliation permits the pickers to enter the
field earlier when heavy dews are prevalent, and also increases their
efficiency. The use of sprays in the humid areas appears to be limited.
In many instances, farmers defoliate their crop in order to remove the
leaves and allow sunlight to penetrate into the plant, thereby reducing
rot of the lower bolls. In other instances farmers may defoliate rather
early in order to reduce the damage from boll weevil or leaf worm, that
is, they discontinue their insect control program in favor of defoliation.

i. Harvesting. Cotton strippers are seldom used in the humid areas
due to the extreme difficulty of passing the large amount of vegetation
through the machines, and also to the inability of the gins to remove
the large amount of trash resulting from stripping the longer-staple
varieties common to the areas. The spindle- type pickers are the machines
used and are attaining great popularity under humid conditions. Their
operation at the present time is, however, restricted, as no machines are
made for entirely satisfactory performance in the Southeast in contoured

COTTON 45

fields, or where rocks are often encountered. On the more level lands,
however, the mechanical pickers are operating most satisfactorily and
large numbers of them are being used. Two spindle-type pickers were
in production commercially in 1949. There is every indication that at
least two more will be placed on the market in 1950, with possibly a
third being available.

Farmers are realizing that every phase in the production of the cotton
crop is of vital importance and that no one operation can be slighted
without affecting those which follow. From field layout and water con-
trol on through crop-residue disposal, seedbed preparation, fertilizing
and planting; and thence through cultivation for weed and grass control,
every single operation, including the insect control and defoliation pro-
gram, affects harvesting, and may have a decided effect both on yield
and quality of cotton produced. Better machines and better methods
are increasing the efficiency of the cotton farmer, permitting him to
produce a higher quality product with fewer man hours.

2. In Low-Rainfall and Subhumid Areas

HARRIS P. SMITH

Texas Agricultural Experiment Station, College Station, Texas

Low-rainfall cotton is that which is grown where the average annual
rainfall is less than 25 inches and where the rainfall distribution is such
that irrigation will materially increase yields. That grown west of a
line drawn from Corpus Christi, Texas, to Oklahoma City, Oklahoma,
will be, approximately, in the low-rainfall area. This area includes the
western halves of the states of Texas and Oklahoma and all of New
Mexico, Arizona and California. Bonnen and Thibodeaux (1937) show
that the approximate western limits of dry-land farming is a line drawn
from Raymondville, Texas, to Lovington, New Mexico.

Although a large part of the cotton produced in the subhumid area
is grown as dry-land cotton, approximately 3 million acres of cotton are
irrigated. The acre yields of dry-land grown cotton range from failures,
in exceedingly dry years, to a bale of 500 Ibs. per acre in years of ample
rainfall. Acre yields of irrigated cotton range from one-half to 3 bales
in some cases. Government estimates indicate that approximately 3*4
million bales of cotton were produced in 1949 from the irrigated acreage.

Cultural practices for low-rainfall cotton, both dry-land and irrigated,
differ from the practices of the humid areas; therefore, they are discussed
separately.

46 HARRIS P. SMITH

a. Field Layout. Fields that are to be irrigated must be carefully
surveyed, slopes determined, irrigation ditches located and lateral ditches
planned so that water can be properly applied, as discussed by Thomas
(1948). The land must be leveled with land-leveling machines and
"Fresnos" so that water will flow down the furrows in a uniform manner.
Borders are sometimes thrown up and the entire field flooded to facili-
tate planting and insure subsoil moisture.

b. Disposal of Crop Residues. In the High Plains area of Northwest
Texas and Western Oklahoma, crop residues such as cotton stalks and
sorghum stubble are left until January or February as a protection
against wind erosion. Just prior to listing the land, 4- and 5-row shop-
made rolling stalk cutters are used to cut the stalks. The dry stalks
and stubble do not decompose readily in the dry climate, and subsequently
cause frequent stoppages of cultural machines, particularly of mechanical
cotton strippers. In the Lower Rio Grande Valley area, governmental
regulatory measures require that all cotton stalks be cut and plowed
under by the first of September. Power-operated stalk cutters and
shredders are being used in this and other areas to chop and shred cotton,
sorghum and corn residues, and also green cover crops. A tandem disk
harrow hitched behind a rolling stalk cutter is a popular method of
cutting and disposing of crop residues throughout the low-rainfall area.
A knife arrangement, similar to that on a peanut digger, to cut the stalk
under the surface of the ground immediately above the root crown, makes
land preparation easier, and permits the use of modern mechanized tools,
such as the rotary hoe.

c. Preparation of Seedbed. Under dry-land farming conditions, land
is usually listed with tractor-mounted middlebreakers to form beds.
Under irrigation, however, Thomas (1948) states that land must be care-
fully leveled, broken flat with one-way plows, floated, ditches made and
often borders thrown up for an application of water before the final seed-
bed is prepared. Some farmers list forming a bed for each row, while
others throw up a wide bed to accommodate two rows. The wide double
beds are called "cantaloupe beds." The beds are harrowed down and
sometimes "boarded" off a few days before planting. In the High Plains
where cotton is planted in the furrow, the beds are "knifed" with long
knives just before planting to destroy weeds. Where cotton is planted
on low beds in Central Texas, rolling stalk cutters are frequently used
to chop the beds to break up clods and destroy weeds.

In some areas where a sandy topsoil is underlain by a clay subsoil,
chiseling and subsoiling is practiced to aid the infiltration of water.

COTTON 47

d. Fertilizer Responses. Fertilizer is not generally used in the low-
rainfall area because of the low crop response. Magee et al. (1944) state
that, in the High Plains of Texas, the natural soil fertility has not
been depleted to the extent that commercial fertilizers can be profitably
used. Hinkle and Staten (1941) found that the heavy soils of New
Mexico are inherently fertile and may be expected to produce satisfactory
cotton yields without fertilizers. Light soils, however, may give profit-
able responses when fertilizer is applied, especially manure. Staten and
Hinkle (1942) found that, where cotton followed two or more years of
alfalfa, considerably higher yields of cotton were obtained. Unpublished
results from Texas indicate that profitable increases in yields are ob-
tained in Central Texas when nitrogenous fertilizers are used.

e. Planting. Planting practices in the subhumid areas differ for dry-
land and irrigation farming. Jones (1948) discussed practices in the
High Plains where the seed is planted in the furrow with a lister type
2- or 4-row planter. A small amount of seed is drilled and little or no
thinning is done. Where rains wash excess soil over the seed, rotary-hoe
attachments mounted on sled cultivators are used to loosen the soil to
aid the emergence of seedlings. In Central and South Texas, planting is
done on low beds with 2- or 4-row tractor-mounted planters, as discussed
by Alsmeyer (1949). Presswheels are left off the planters and special
rollers are used a few hours after planting to compact the soil over the
seed. Planting on either flat-broken or bedded land is a general practice
under irrigated conditions. Where cotton is planted on flat land, many
farmers use a double-disk furrowing attachment in combination with a
runner or knife opener. Throughout the low-rainfall area most cotton
is drilled, but there is a trend toward the use of hill-drop attachments.
Rotary and electrically actuated valves are satisfactory for hill-dropping
cottonseed. There is also a definite trend toward the use of delinted seed.
One concern in Central Texas annually furnishes farmers 70,000 bushels
of certified mechanically-delinted and treated seed.

/. Thinning. Cotton farmers of the High Plains area of Texas and
Oklahoma have long followed the practice of planting to a stand, that is,
only enough seed is planted to give a stand of plants. Where cotton is
hill-dropped with an average of 4 to 5 seeds per hill, spaced 14 to 18
inches, no thinning is necessary. Where cotton is drilled to a thick
stand, the most common method of thinning is chopping with a hoe.
The average rate of pay for hand chopping in 1949 was $4.00 per day
per laborer. Many farmers use mechanical choppers and some practice
cross-plowing to reduce "chopping" costs.

48 HARRIS P. SMITH

g. Cultivation. Fairbank (1948) has pointed out that weed and grass
control is one of the major problems in cotton mechanization. Cotton has
been produced entirely by mechanical means under humid conditions, as
reported by Gull and Adams (1945). Limitations under dry-land and
irrigation culture, as given by Smith (1949a) emphasize the problem
of control of weeds and grasses in the crop row. Cultivation is not only
necessary to destroy weeds, but also to conserve moisture and to put
the soil in better condition for plant growth. The rotary-hoe cultivator
attachment is an excellent tool for breaking soil crusts arid for the
destruction of young weeds in the first two months of crop growth.
When the sweeps on each side of the rotary hoes are set flat and run
shallow, cultivation can be done 40 per cent faster than where the cus-
tomary fenders and sweeps, and sweep setting, are used. The rotary-
hoe attachment can be used either in the listed furrow or upon beds.
Current work shows that an integral-mounted 4-row tractor lister culti-
vator with rotary hoe attachment may reduce hoeing hours from 60 to
100 per cent in the High Plains area. The regular 2- and 4-row tractor-
mounted cultivators are used except where cotton is planted in the
furrow.

The number of cultivations required for weed control varies from an
average of 3 to 6 per season. Extremes may range from 1 to 10 per
season. The first cultivation of cotton may often be before seedlings
emerge and the last cultivation may be delayed until bolls begin to open.

Flame cultivation has not been used extensively in the low-rainfall
areas because of the different farming practices, and climatic conditions
and because the low moisture level in the surface soils retards germina-
tion of seeds of annual weeds.

h. Application of Insecticide*. Tractor-mounted dusters, and also air-
planes, are used for applying insecticides to cotton. There appears to be
a trend toward the use of spray equipment in some sections. Where the
pink boll worm is found, stalk cutters, disk harrows, and plows are im-
portant tools in cutting and burying of the crop residue. The distribu-
tion and relative importance of cotton insects is discussed in V.

i. Defoliation. Jones and Jones (1945) stated that defoliation of cot-
ton plants at harvest time in the subhumid areas is more difficult to
obtain than in the humid areas. Usually the weather is dry and hot,
the soil is deficient in moisture for plant growth and the plants are semi-
dormant, and in stress from lack of moisture. Under most conditions,
plants and foliage must be approaching maturity to obtain rapid de-
foliation. Smith (1949b) points out that an active plant and atmospheric

COTTON 49

moisture in the form of dew are two essential factors to complete de-
foliation of cotton.

Where cotton is defoliated, hand pickers can easily see and pick bolls
that would ordinarily be hidden by the foliage. The trash in mechani-
cally-picked cotton is dry and can be more easily removed, without
chlorophyll stain, by the cleaning equipment of the gin, and thus a higher
quality cotton obtained than where the plants have not been defoliated.

j. Harvesting. Smith et al. (1946a) discuss 4 methods of harvesting
cottoif in general use in the sublnimid areas. These methods are termed
hand-picking, hand-snapping, machine-stripping, and machine-picking.
The practice of snapping cotton began in the High Plains of Texas about
1912 and since that time has spread to all sections of Texas, and to other
states. Mechanical stripping also started in the High Plains area.
Home-made sled strippers wore used from 1914 to 1930, but were prac-
tically abandoned during the period 1931 to 1940. Two-row tractor-
mounted strippers were introduced in 1943. Cotton growers in Texas
and California are adopting the spindle-type picker, but some farmers
who grow the fine-textured, long staple cottons in New Mexico and Ari-
zona contend that this quality type of fiber is injured when machine
picked. The mechanical stripper requires a storm-proof type cotton
while the picker requires an open boll with fluffed looks, and a staple
length sufficient to allow the lint to wrap around the picking spindle.
Smith et al. (1939, 1946b) showed that varietal characteristics generally
affect field losses of mechanical harvesters more than the mechanical
factors. Williamson and Rogers (1948a, 1948b) point out that cultural
practices materially affect field losses of both the stripper and picker.
The quality of yarn manufactured from machine-stripped and picked
cottons of Texas production was not affected by the method of harvest
when compared with hand-harvested cottons, except for a slight lower-
ing of the appearance grade for the longer and fine-fibered cottons, as
discussed by Smith et al. (1946b) and Grimes (1947).

k. Irrigation. McDowell (1937) and also Barr (1949) have pointed
out that the proper use of water is perhaps the most fundamental
problem in cotton production in regions where the crop is grown under
irrigation. Irrigation practices are influenced by the nature of the soil,
amount and distribution of rainfall, temperature, and evaporation. These
factors or conditions vary widely in different areas and the practice for
one area will not apply, as a whole, to another area. The 3 principal
sources of water for irrigation are those impounded in reservoirs, pumped
from rivers, and pumped from wells. Under Arizona conditions, Harris

50 CHARLES A. BENNETT

(1947) found that the average amount of irrigation water required to
produce a crop of cotton ranges from 20 to 30 acre-inches. The amount
and distribution of the annual rainfall will influence the irrigation re-
quirements. Where cotton is grown under irrigation, there are 3 critical
periods of cotton growth: (1) planting to heavy fruiting; (2) fruiting;
and (3) the maturity period. Experiments in Arizona by Harris and
Hawkins (1942) and Harris (1947) show that, in general, the more
rapid the growth of the cotton plant prior to heavy fruiting, the higher
the final yields.

The application of water to cotton late in the season will * delay
maturity. Cotton plants require and use more water during July, August,
and September than at other times. The cost of producing cotton under
irrigation is higher than under rain-grown conditions and yields must be
higher to obtain profitable returns.

VII. IMPROVEMENTS IN GINNING PRACTICES

CHARLES A. BENNETT

17. S. Department of Agriculture, Stoncvillc, Mississippi.

The ginning processes are of vital importance to all who are engaged
in the cotton industry, because the few-minute period that is required
to gin a bale determines the grade and staple sample for selling the bale,
and this sample may reward or punish those who have contributed to its
production.

The United States leads in improvements in ginning, and competitors
quickly copy them, either by purchase of better machinery, or by efforts
to provide home-made substitutes. This is evidenced by some 17 patents,
with others pending, pertaining to ginning machinery and processes.

"Ginning" includes several important stages of processing, each hav-
ing a special function and bearing upon the end result, namely, the qual-
ity of the finished bale. Without being too technical, it may be said that
there are about seven of these stages, usually operated in the following
order: (1) conditioning; (2) cleaning and extracting; (3) distributing
and feeding to the gin stands; (4) separation of the fiber from the seed,
either by saws or rollers within the gin stands; (5) lint cleaning; (6)
disposition of freshly ginned seed and foreign matter; and (7) packag-
ing the lint into bales.

L Regulation of Moisture

Regulation of optimum moisture content has not yet been satisfac-
torily achieved. There are various regional conditions that call for

COTTON

different treatment of the seed cotton and the fiber. The restoration of
moisture in the arid regions is a much slower and more difficult operation
than is the mere drying necessary in humid areus. Approximately 30
Ibs. of seed cotton must reach each saw-gin stand per minute, and con-
ditioning must therefore be done in bulk streams, or in the final distri-
bution to the gin stands. Modern gins attempt to do both, insofar as
drying is concerned, but neither methods nor controls have yet been
devised for optimum regulation of the moisture content of the seed cotton

Seed cotton inlet

Sealed
dropper

Spreader
Stripper
Brush

6. Ginning

8. To Bale

7 Lint Cleaning

Fig. 2. Diagrams of the several processes of cotton ginning, except pressing and
packaging. In normal order of sequence, beginning at upper left hand corner, they
are (1) drying; (2) cleaning; (3) extracting, which is usually followed by more
cleaning; (4) distribution; (5) feeding; (6) ginning or the separation of the fiber
from the seed; and (7) lint cleaning during which the freshly ginned fiber is centri-
fuged and scrubbed.

52 CHARLES A. BENNETT

(principally the fiber) in the high-speed transit of the material through
the several processes ahead of the actual ginning or fiber separation from
the seed.

Figure 2 depicts in diagram form the several preliminary processes of
drying, cleaning, and extracting; it also shows the cleaning of the ginned
fiber or lint immediately after it leaves the gin stand. No lint cleaning
devices are as yet available for roller-types of cotton gins.

It will be noted from Fig. 2 that the drying involves a pneumatic
method; that the cleaning involves a threshing method which employs
both beaters and screens (or grids of some sort) ; and that the extracting
depends upon a carding method which is accomplished by toothed cyl-
inders and strippers. Drying fluffs up the cotton and liberates the
ordinary small particles of foreign matter, and at the same time it
appears, in rain-grown regions, to enhance the subsequent action of the
cleaners and extractors. If means can be developed for adding moisture
quickly to the seed cotton as it passes through gins in the arid regions,
much will be accomplished.

Drying is now a multistage process in itself at the more completely-
equipped gins, because several driers may be used in succession. They
may be all of one type, or mixed, and intermingled with them may be
auxiliary jets of hot air to the cleaners in lesser air volumes than are
employed in the regular dryers. Seed cotton may be conveyed through
the dryers by pneumatic or mechanical means. The Government design
(Fig. 2), developed by the U. S. Department Agriculture Ginning Lab-
oratory, is pneumatic, and consequently somewhat automatic in action
because the cotton moves in proportion to its fluffiness and dryness,
rather than by positive rate of travel from a belt or auger.

2. Cleaning

The "cleaners" have a unique function in the modern cotton gin,
whether they achieve the maximum removal of finer trash or not. First,
they open and fluff up the cotton as it comes from the dryer, often com-
prising the receiving chamber for this material in a very simple and
effective way. Fig. 2 shows this in a modern "blow-in" delivery of dried
cotton to the cleaner. Second, this fluffing is a mechanical agency that
loosens larger pieces and fragments of burs, hulls, leaves, and other
unwieldly foreign matter, so that the portion passing to the extractors
may be the more readily removed. That part of the foreign matter that
can be screened out, is, of course, usually discharged in the cleaners, but
not always at the first one. After the main extracting stage, the finishing
cleaners perform a "cleanup" job. All-in-all, cleaners generally take out
about 30 per cent of the foreign matter that cornes into the gin with the

COTTON

seed cotton, although sometimes 20 or more cylinders are required to
accomplish this.

3. Extraction and Interrelated Processes

The extractors exist for the purpose of extracting sticks, stems, burs,
hulls and larger pieces of foreign matter. Like the dryers and cleaners,
they too involve a series of individual extractions, interspersed here and
there between cleaners and driers, or between cleaners and gin stands.
Foreign matter removal by extractors is also approximately 30 per cent,
taken as a whole for all extraction processes that may be employed in
the modern gins, beginning with overhead machinery and ending right
within the huller fronts of the gin stands themselves. Huller fronts are
primitive extractors, from which the large modern counterpart has arisen.
The U. S. Department of Agriculture is now conducting research on better
removal of green, unopened bolls, and of heavy plant sticks and stems
that are very troublesome when machine-stripped cotton is being ginned.

Figure 3 depicts a cross-section of a modern cotton gin with a chain
of processes frequently used. Any similarity between the diagram and

Fig. 3. Diagram of a section through a modern cotton gin equipped to handle
machine-picked and roughly hand-harvested cottons. This diagram is for informa-
tive purposes only and does not conform in cross section to the majority of cotton
gins that have suction telescopes at the front of the stands. The equipment, how-
ever, is comparable.

CHARLES A. BENNETT

special brands or makes of machinery is strictly coincidental and does
not comprise Governmental endorsement. It will be noted (Fig. 3) that
some of the various processes are repeated, except those within the gin
stands and lint cleaners, and that the seed cotton now does rather ex-
tensive traveling on its way from storage bin or vehicle to the bale press
(not shown in the figure) .

Figure 4 depicts a somewhat different arrangement of dryers, cleaners
and other units than used in Fig. 3, both being representatives of im-
proved types of cotton gins now in use in humid and arid regions of the
Cotton Belt.

Fig. 4. Elevation diagram of a modern cotton gin with drying, cleaning and
extracting equipment on left side of the ginning aisle; and with a conventionally
equipped, standard gin for hand-picked cotton on the right hand. The "Annex," or
equipment to the left, is usually employed on machine-picked and other roughly
harvested cottons.

* From previous statements, it will be noted that the combined action
of cleaners and extractors, under optimum conditions, takes out about
60 per cent of the foreign matter from the seed cotton. The lint cleaners
remove up to about 13 per cent more. There is still room for improved
efficiency and simplification in both cleaning and extracting.

There is fairly definite information on moisture removal, but little
data on restoration because of the many variables involved. Usually
the first drying process can remove approximately 3 per cent of the
moisture in the seed cotton, most of it coming from the fiber rather than

COTTON 55

from the seed. In succeeding stages of drying, which may be concurrent
with cleaning as is indicated in Fig. 4, somewhat less moisture may be
removed than in the first stage, but the final result may bring the fiber
to 5 per cent moisture content, or less. Static electricity is then likely
to appear, with an attendant sequence of troubles. Likewise, too much
heat and too many cylinders of cleaning may cause the fibers to become
brittle, "ncppy," "nappy" and roped. Rapid indicators and controls arc
now needed to improve the present crude regulation of moisture content
to percentages best suited for each succeeding process.

Drying aids the cleaning process, and accordingly the moisture con-
tent of damp cottons should be reduced to 10 per cent. Cleaning assists
the extracting; extracting enhances the feeding and the ginning; and lint
cleaning "pronounces the benediction" on processing, if other steps in
processing have been good. With these modern processes, machine-
picked cotton may be brought very close in grade to hand-picked cottons,
but the gin cannot turn out a bale of machine-picked cotton as quickly
as it can gin clean hand-picked cotton; neither can it do it as cheaply,
nor with as small amount of machinery and labor.

Preservation of pure seed is possible at the most modern gins, if
correct methods of handling and clean-up are employed. Sterilization
of seed and incineration of trash can also now be accomplished without
causing community nuisances.

4- Summary

The modern cotton gin frequently represents an investment of many
thousands of dollars. True progress and improvements are coming
through better trade association practices and methods, through more
scientific appraisal of the engineering and economic aspects of the busi-
ness, and through the cooperation of scientists and collaborators in all
the elements of the cotton industry, from producer to cotton mill and
spinner.

The 7 processes, listed at the beginning of this section, are all work-
ing in the modern cotton gin to enhance and preserve the coordinated
efforts of the agronomist, geneticist, agricultural engineer, and cotton
technologist to produce cottons more acceptable to the trade and cottons
of maximum performance. The team-work of those engaged in research
on the ginning process is paying large dividends, as mechanized produc-
tion increases. Improved ginning practices and methods make for better
cotton and greater profit and will continue to contribute to the com-
petitive position of cotton as a fiber, food or feed.

56 HENRY D. BARKER

VIII. FIBER PROPERTIES AND THEIR SIGNIFICANCE

HENRY D. BARKER

[7. 8. Department of Agriculture, Beltsvttle, Maryland

Significant recent advances have been made in this country in de-
veloping instruments and applying techniques for measuring the fiber
properties of cotton. Some of these advances arc so new that their sig-
nificance is not generally understood. Many workers have made impor-
tant contributions to these advances. The objects of this brief summary
are to outline some of the recent advances and to emphasize their
significance to cotton breeders who are striving to develop better cottons,
and to others who are interested in more effectively utilizing the various
combinations of fiber properties that exist in this remarkable natural
fiber.

It has been known for a long time that the spinning performance
of cotton depends upon certain fiber properties such as length, strength,
structure, and fineness. Within the past few years a great deal more has
been learned about each of these properties, and how their interrelation-
ships affect yarn and fiber quality. As modern technology widens the
range of raw materials and as improved methods of processing these
raw materials into textiles are developed, it is inevitable that the trend
will be more and more toward the selection of cotton on the basis of
fiber quality. As evidence of this trend, attention is called to the increas-
ing number of fiber laboratories with modern equipment for measuring
fiber properties which are being installed by merchants and spinners.

1. Fiber Structure and Development

Basic to a better understanding of fiber properties, their interrela-
tionships, and their significance is the recent work on the origin and
nature of cotton fibers and the factors that affect fiber properties.

Each fiber, as reviewed by Anderson and Kerr (1938), is an out-
growth of a single epidermal cell of the cotton seed. This cell first
elongates as a thin-walled tubular structure to its maximum length and
the fiber wall is then thickened by deposition of cellulose, within, until it
matures.

The elongation phase of fiber growth begins on the day the flower
opens and pollination takes place. Elongation continues for a period
of 13 to 20 days depending on variety and growth conditions. It is slow
at first, more rapid for a few days, then slows down near the end of the
growth period. The short-fibered varieties generally elongate over a
shorter period of time than the long-fibered varieties. Length of the

COTTON 57

fiber is influenced also by environmental factors, especially water stress
within the plant. A slowing down in the rate of elongation results in
the production of short fibers. The thin membrane which encloses the
protoplasm is known as the primary wall. It is the outer layer of the
mature fiber, and is made up of waxes, pectins, and cellulose.

The thickening phase of fiber-wall growth begins after elongation has
ceased. Deposition of the secondary wall in most varieties takes place
within a period of 25 to 40 days. It may continue over a longer period
for long-staple varieties, as is true for elongation. Cellulose in the sec-
ondary wall of the fiber is laid down in a spiral formation. The spirals
in the first layer of the secondary thickenings are conspicuous because
they lie in a direction opposite to those in succeeding layers (Kerr, 1946).
Some details of minute fibrillar structure of this wall are visible through
the ordinary microscope (Fig. 5). Examination of a prepared cross
section of the matured fiber shows alternating light and dark rings in the
inner thickening. According to Anderson and Kerr (1938), each pair of
adjacent light and dark rings represents the cellulose deposited during
one day's growth. Although variety largely determines characteristics
of a secondary-wall structure, environment may modify them consider-
ably. Water stress, for example, which slows down the elongation of
the primary wall during the first 13 to 20 days, if continued through the
phase of secondary thickening, results in thinner, more dense layers in
the secondary wall.

Many workers (Barre et a/., 1947; Berkley et al., 1948; Hermans and
Weidinger, 1949; Hessler et al., 1948; Nelson and Conrad, 1948) have
contributed to providing the answer as to why the cotton fiber has a
tensile strength greatly exceeding that of man-made fibers. It has been
shown that this is determined by the character of the minute structure
of the secondary wall. This structure is determined by the chemical
constitution of the fiber and the location of the various components in the
different parts of the wall.

Cellulose makes up 88 to 96 per cent of the mature cotton lint and
forms the skeletal framework of the fibers of average thickness. The
cellulose is composed of many glucose anhydride units arranged in the
molecule in a threadlike chain. There are at least 3,000 to 4,000 glucose
units in a molecule of cellulose. These straight chains are deposited in
the wall parallel to the protoplasmic surface and more or less parallel to
each other. The ends of the chains overlap. In many regions of the wall,
50 to 100 chains may be truly parallel to each other. In this case the
molecules are held in a fixed position and are kept together by so-called
valence bonds. Such groups of parallel chains behave as minute crystals
and are called crystallites. Not all chains are truly parallel. When the

58 HENRY D. BARKER

distance between the cellulose molecules exceeds the effective distance,
the molecular structure is said to be "amorphous." Even a single chain
may pass from a region of crystallinity to a region that is amorphous in
nature. In a dried cotton fiber, at least 70 per cent of the total cellulose
is in the form of the minute crystallites. Groups of crystallites in turn
form the fibrils. The long axis of the molecules, the crystallites, and
the fibrils coincide.

Sisson (1937), Berkley et al. (1948), and others have shown that the
properties of the cotton fiber may be profoundly influenced by: (1) the
general direction of the chains and crystallites with respect to the fiber
axis; (2) the average length of the chains; and (3) the relative percentage
of crystalline and amorphous cellulose. While it is not possible to see
the molecular chains, the direction of the molecules may be determined
accurately by means of x-ray diffraction patterns.

Though the chains themselves cannot be seen, a large number of
parallel crystallites form the so-called fibrils that are visible under the
microscope. The direction of these fibrils to the fiber axis affects the
strength of the fiber. The more nearly they parallel the long axis of
the fiber, the stronger is the cotton. The organized crystalline cellulose
makes up the more dense parts and its arrangement is associated with
the tensile strength of the fiber. The amorphous, or unorganized, parts
are less dense and probably account for the flexibility or the efficiency
with which the fibers bend and form themselves into position.

The average length of the chains, as they are deposited within the
cotton hair, is apparently about 3,000 glucose units. Variations in the
average length of cellulose chains resulting from growth conditions have
not been shown to affect the fiber properties significantly. When mature
cotton is exposed in the field and damaged by the weather, the average
length of the chains is usually reduced. The action of ultraviolet light
reduces chain length and may affect fiber strength. Microorganisms that
frequently develop during field exposure leave many of the chains un-
affected, as judged by fluidity tests, but apparently produce localized
destruction of a high percentage of the chains with consequent reduction
in fiber strength.

S. Fiber Length

The oldest and most widely known property for evaluating cotton
fiber is fiber, or "staple," length. Until recently, spinners who wanted
yarns of a given strength specified the length of staple that should be
used to produce the desired strength.

One of the early methods of measuring fiber length was to sort out
the various length groups from a sample and measure them with a rule.

COTTON 59

A mechanical device that gives a more nearly accurate measurement is
the Suter-Webb duplex porter, developed by Webb (1932). The fibers
are removed from the sorter in order of length and placed in groups dif-
fering one-eighth inch in length. Each group of the same length is
weighed. On the basis of these data, it is possible Lo compute the length
of the longest 25 per cent of the fibers by weight and also to calculate
the average length of the entire sample. Though accurate, the Suter-
Webb duplex sorter is slow and tedious and therefore expensive to
operate.

A device that measures fiber length more rapidly has now come into
general use. This instrument, known as the Fibrograph, was developed
by Hertel and Zervigon (1936). It is a photoelectric device for scan-
ning a fiber sample and tracing a length-frequency distribution curve
from which fiber lengths are readily obtained. In operation, a small
quantity of ginned lint is placed upon one of a pair of combs. Several
transfers from comb to comb parallel the fibers. The combed sample
then consists of parallel fibers evenly distributed over the length of the
tw T o combs, with the fibers caught at random points by the teeth, and
extending from the combs. The two combs bearing the fibers are placed
in the Fibrograph, and the length distribution curve is traced directly
upon a card by manipulation of the instrument. Tangents drawn to the
resulting curve give two fiber-length measures: (1) the average length
of the longer half of the fiber population, upper half mean, which com-
pares roughly w r ith staple length as judged by commercial classers; and
(2) the mean length of all fibers in the sample.

3. Fiber Strength

Though methods for determining fiber strength with considerable
precision have been known for several years, most of them have been,
until recently, extremely time-consuming. The most laborious of the
traditional methods, the breaking of individual fibers, is now used only
for specialized research.

The Chandler bundle method, using the pendulum-type Scott tester,
was the most widely used procedure for testing fiber strength for many
years. In operation, as described by Richardson et al. (1937), a bundle
of parallel fibers about a millimeter in diameter is wrapped with thread
of known size and length so the cross-sectional area can be calculated
from measurement of the circumference. The wrapped bundle is placed
in the test jaws of the breaker, in which the weight required for breaking
is determined. The tensile strength of the sample is calculated as pounds
per square inch of cross-sectional area.

A rapid breaking-strength method using the Pressley (1942) strength

60 HENRY D. BARKER

tester, an instrument that can be operated much more easily and rapidly,
has largely replaced the Chandler bundle method. The Pressley breaker
is essentially a lever system activated by a weight rolling down a gradu-
ated inclined plane. A parallel tuft of fibers is placed in two jaws lying
against each other, and the protruding ends of the fibers are cut off. One
of the jaws is held fast and the other is pulled away by the rolling weight
on the lever. The weight automatically locks in position when the
bundle breaks. A scale indicates the load required to break the sample.
The broken tuft is weighed and the strength of the sample is computed
in Ibs. per milligram of cotton of a given length. The method is reason-
ably accurate, even though there is inadequate control over the speed
at which the breaking load is applied, and the portion of tuft between
the two jaws where the break occurs is short.

4. Fiber Fineness

The term "fiber fineness" is well established in cotton literature. Un-
fortunately it is a loose term even though useful in describing the number
of fibers that can be packed into a yarn of a given count. More precisely
it depends on two properties: (1) fiber perimeter, which is largely an
inherited characteristic; and (2) fiber wall thickness, which is dependent
on both genetic and environmental influences.

Cotton technologists in this country have, for many years, expressed
fiber fineness in terms of weight per inch as described by Richardson
et al. (1937). A sorter is used to make a length array. Usually 200
fibers are counted from each length group and weighed. The weights are
then used to prorate the various length groups and, finally, a weighted
mean for the entire sample is computed and expressed as "micrograms
per inch of fiber." This method, though difficult and time-consuming,
gives an accurate measure that represents the composite effects of peri-
meter and cell-wall development.

A much more rapid method for determining a different composite effect
of perimeter and cell-wall thickness was developed by Sullivan and Hertel
(1940); the instrument is culled the Arealometer. It provides a meas-
urement of the surface area of a given mass or volume of fiber. Because
it is so much more rapid than the weight per inch method, it or some
other air-flow method, such as that developed by Pfeiffenberger (1946)
or Elting and Barnes (1948), is now widely used in research and indus-
trial laboratories.

The Arealometer uses an aerodynamic principle governing the flow
of gas through a porous medium. The measures that were used up to
the fall of 1949 were as follows: in operation, a line sample of 100 mg.
is rolled into a plug and compressed in a tube of standard bore until

COTTON 61

resistance to air flow is equal to a standard resistance. The amount of
compression indicated by the length of the plug is read directly from
the instrument. By means of a calibration chart, plug lengths can be con-
verted to the specific surface in square centimeters per milligram. Al-
though the Arealometer is rapid and provides a very useful composite
measure of fineness, it, like weight per inch, does not provide information
on cell perimeter and cell-wall thickness.

Fiber perimeter varies with different varieties and species of culti-
vated cottons. At present, no rapid precise method has been worked
out for measuring fiber perimeter. Pearson (1950) developed a micro-
scopic technique for obtaining the comparative perimeters of different
varieties. In this technique, the fibers are measured in the primary-wall
stage and thus the errors and complications associated with attempts to
measure mature fibers are avoided. Tufts of young fibers are stained,
fanned out on a glass slide and allowed to dry. As there apparently is
no readily detectable shrinkage during the drying process, the widths of
the collapsed dried fibers as viewed longitudinally are equal to half the
fiber perimeter at the position measured.

Kerr (unpublished paper presented at the April, 1949, meeting of the
Fiber Society) has devised an indirect method of calculating perimeter
from surface area and weight-per-inch data. The weight-per-inch data
are recalculated in terms of weight (in micrograms) per centimeter and
the reciprocal of this quantity is divided into the Arealometer reading
(surface/mg.). In using this method, data from the "Annual Varietal
and Environmental Study of Fiber and Spinning Properties" for the crop
years 1943, 1944, 1945, and 1946, are compared. Strangely enough, the
characteristic perimeter of different varieties was found to be little af-
fected by varying environmental influences.

A newly designed Arealometer has been released to a number of
laboratories. Prior to September 20, 1949, results obtained with the
Arealometer have been reported in terms of cm. 2 /mg. Because of the
difference in density among textile fibers, it has been deemed desirable to
change the unit in the new instruments so that the same reading will be
reported on all fibers having the same geometrical size. The new unit
selected is "square millimeters per cubic millimeter."

Since the average density of cellulose is 1.52, and the new calibration
data give results that are three per cent higher, a cotton heretofore reported
with a specific surface of 2.00 cm. 2 /mg. will now read 314 mm. 2 /mm. 3 in
the new units (2.00 X 1-03 X 152 - 2.00 X 157 = 314). Results prior to
September 20, 1949, must therefore be multiplied by 157 to give results
in terms of the new unit which was adopted September 20, 1949. The
factor 157 is based on the average of 12 readings for each of 18 cottons

62 HENRY D. BARKER

covering a wide range of specific areas, and may be used as an average
conversion factor.

Several fiber characters other than those mentioned above are known
to exist and may be important in spinning performance but, in general,
are not well understood. Luster, drag, convolutions, reversals, and bends
are some of the properties that have not been extensively measured and
evaluated. For genetic variability in spinning performance, good pre-
dictions may now be made from data on fiber length, fiber strength, and
surface-area measurements. Measurements of environmentally-induced
variations in fiber properties, however, result in less accurate predictions.
This is interpreted as evidence that important properties other than
length, strength, and fineness may be associated genetically with one of
these three properties, but varies independently when modified by en-
vironment.

5. Significance of Fiber Properties

The significance of fiber properties in relation to end-use value has
been exceedingly difficult to determine. This is due to many causes. One
of these is the complexity of fiber-property interrelationships. Another
is that, in processing cottons of varying properties, many processing ad-
justments are required to obtain optimum results. It is difficult, there-
fore, to ascertain whether an adjustment that was made on the basis of
experience in dealing with a given property, such as length, is actually
optimum for some other property such as perimeter or cell-wall develop-
ment. In studying the interrelationships of fiber properties and their
significance in long draft spinning, Barker and Pope (1948) presented
correlation data on 447 samples.

Multiple-correlation studies established that about 80 per cent of the
varietal differences in skein strength may be accounted for by four
measured fiber properties obtained from the Fibrograph, Pressley breaker,
and Arealometer. For environmental effects, however, only about 55
per cent of the skein strength was accounted for by these 4 properties.

From the standpoint of the practical breeder whose main interest is
in genetic differences, R 2 is nearly as high for two properties, upper-half
mean length and Pressley index, as it is when all 4 or 5 fiber measure-
ments are evaluated. That Arealometer measurements can usually be
dispensed with, in estimating varietal differences in skein strength, is
apparently due to the fact that genetic differences in fiber length are
rather closely associated with differences in surface area.

For environmentally induced differences in fiber and spinning prop-
erties, however, a very different condition exists. In the first place, mean
length supersedes upper-half mean length as the important length meas-

COTTON 63

urement, and surface-area measurements become of greater importance.
The latter cannot be omitted without causing appreciable reduction in
R 2 values.

For yarn appearance grades, multiple-correlation coefficients showed
that upper-half mean length and surface area differences significantly
affect yarn appearance grade.

Fig. 5. Spiral structure of outer layer of secondary thickening. Where the fibrils
change direction is referred to as the spiral reversal (photomicrograph). X 850.

IX. BREEDING AND IMPROVEMENT

T. R. RICHMOND

Texas Agricultural Experiment Station, College Station, Texas

1. General

The cotton plant bears complete flowers. Cultivated varieties gen-
erally are placed in the "usually" self-fertilized category with respect to
crossing habit under natural conditions, but the per cent of crossing may
vary from less than 5 to approximately 50 (Kime and Tilley, 1947; Simp-
son, 1948a). The amount at any one location is proportional to the
number of wild and domesticated bees which visit the field. Since the
cottons of commerce are propagated by seeds, there is more or less of a
tendency toward inbreeding, depending on the breeding methods em-
ployed, the isolation of increase plots, the precautions taken against
mechanical mixtures, and many other factors.

Vavilov (1927) has shown that stable populations of crop plants exist
(or existed) in certain limited areas called "centers of origin," that varia-
bility in such centers is high, and that variability diminishes toward the
periphery of the distribution. There are at present two centers of origin
for American cultivated species, one for American Upland, G. hirsutum,
in Southern Mexico and Central America and one for Sea Island and

64 T. K. BICHMOND

related types, G. barbadense, in the Andean region of Peru, Ecuador and
Colombia. The American species have proved to be remarkably plastic
genetically and it has been possible through selection to develop agricul-
tural varieties of a significant range of types, many of which have been
widely adapted geographically.

The advances in breeding and improvement in cotton, as presented in
this section, will be confined almost entirely to the American cultivated
cottons and wherever possible, to American Upland, G. hirsutum, which
is the predominant cultivated variety in the United States.

2. The Breeding Problem

The observation that cotton varieties or strains often lose in vigor
and productivity under a regimen of selfing or close inbreeding, such as
that resulting from the progeny-row breeding systems, in vogue soon
after the rediscovery of Mendel's papers, has led many farmers and
breeders to consider "running out" of strains to be characteristic of cot-
ton. Excluding genetic abnormalities such as balanced lethals, homo-
geneity is a natural consequence of prolonged inbreeding. The observed
fact of reduction in productivity following inbreeding in a given strain
is attributable to a number of factors chief among them being: (1) the
degree of heterogeneity of the original parent stock; (2) the mathematical
probability against accumulating all (or even most) of the favorable
genes for yield in one homozygous line; (3) mechanical mixtures and
cross pollination with other varieties; and (4) selection for one (or a
small number) of characters without regard to other characters which
have an important function in the genetic complex. The last factor is
of utmost importance and often determines the success or failure of a
breeding program.

On theoretical grounds there is no reason to suspect that pure lines
of cotton are different in behavior from pure lines of any other crop.
The practical consideration, of a pure line involves both uniformity and
superior performance. Both cannot be sought rapidly and at the same
time. Selection for high yield "on a broad genetic base," which is neces-
sary if decreases in production are to be avoided, involves the simultane-
ous handling of a number of characters over a relatively long period;
such a procedure does not lead rapidly to homogeneity. Failure to ob-
serve the "broad base" concept and the desire for rapid development of
uniformity in one character at the expense of all others has taught many
cotton breeders the severe lesson that the probability of obtaining a
"uniformly bad" strain is much higher than that of obtaining a "uni-
formly good" one.

COTTON 65

It has been shown that the germ plasm, which constitutes the genetic
reservoir, from which present agricultural varieties have arisen, has
yielded a number of productive types which are well adapted to their
respective geographic areas of growth. Particularly has this been true
of American Upland stocks. Reselection within varieties, and even within
the progeny of varietal hybrids, over a period of many years, inevitably
has resulted in severe inbreeding and the elimination of many beneficial
as well as deleterious genes which were present in the native stocks. Since
American Upland varieties in the United States are all interrelated and
probably descended from not more than a dozen original introductions,
it is doubtful that future requirements of special fiber properties, disease,
insect and drought resistance, mechanical harvesting, and other special-
ized uses and properties can be met by the usual selection methods
entirely within present cultivated varieties (Richmond, 1947).

Two approaches or combinations of two approaches to further prog-
ress and improvement come to mind immediately: (1) development of
more precision in the breeding program through refinements in method
and design to provide more discriminatory statistical tests, and the es-
tablishment of indices which will measure the genetically potential per-
formance rather than the actual end-result behavior; and (2) introduction
of new germ plasm into the breeding material. In the first, the goal is
to determine the amount of genetic variability in the material and to
distinguish this variability from the variability due to environment. In
the second approach, genetic variability is increased purposely by the
addition of new genes. Genetic variability outside the range of that now
present in current agricultural varieties is available from 3 principal
sources: (1) obsolete agricultural varieties; (2) primitive stocks in or
near the center of origin; and (3) the wild species of the world. In the
United States, under a regional project in cotton genetics made possible
by the Research and Marketing Act of 1946, the Mississippi Agricultural
Experiment Station has the major responsibility for collecting and main-
taining stocks from the first source, and the Texas Agricultural Experi-
ment Station is responsible for the stocks from the last two sources.

The importance of genetic variability in the primary breeding material
cannot be over-emphasized, for it is axiomatic that the breeder cannot
bring out in his selections anything more than is present in the raw stock.
Modern breeding methods are designed to preserve and control genetic
variability, to guard against serious loss of favorable genes through the
restricted selection of only a few superior plants or progenies in any one
season. It follows that such methods must keep genetic variability high
in relation to environmental variability, particularly in the early stages
of the breeding program.

66 T. R. RICHMOND

S. Breeding Systems

Less than 20 years ago Cook (1932) advocated renewed emphasis on
"type" as the prime consideration in cotton breeding and in the main-
tenance of seed stocks. Under that "type selection" system, groups of
progenies, instead of single progenies, were propagated and maintenance
of the stock was carried out by reselection within the groups. Selection
and testing of progenies under different conditions "as a means of pre-
serving the adaptive characters of varieties which otherwise may be lost
even without being recognized" was recommended. The value of stable
mixtures of strains or varieties in providing "greater flexibility of re-
sponse" was recognized by Hutchinson (1939). In his genetic interpre-
tation of plant breeding problems, Hutchinson (1940) observed the
association of rapid degeneration with the more closely bred varieties and
recommended that "the effort at present devoted to achieving purity may
profitably be used to increase the efficiency of selection."

Hutchinson and Panse (1937) introduced randomization and replica-
tion into the progeny-row system of breeding. The system, which has
been designated as the "replicated progeny-row method," provides all
the information on means of progenies, and means of plants within
progenies, obtainable by earlier progeny-row methods and, at the same
time, makes possible additional valuable information. In the design
employed by Hutchinson and Manning (1943), the progeny of each
selected plant was tested in 10 randomized blocks, each plot of which
contained 5 plants. Where strains from a number of families were to be
tested in one lay-out, "compact family blocks" were arranged within the
main experiment to provide more precision in the interfamily compari-
sons. Not only does the design reduce the environmental contribution
to the variance, but it makes possible the partitioning of the total vari-
ance into its genetic and environmental components, thus, minimizing
"environmental fluctuations .while maintaining genetic contrasts." The
writers point out that selection on progeny means is as many times more
efficient than mass selection in the same material as there are plants
per progeny.

To avoid some of the dangers inherent in progeny-row breeding when
rigid selection is practiced on relatively few characters, Harland (1943,
1949) inaugurated a breeding system which he terms "mass-pedigree
selection." In practice, the system involves: (1) the growing of progeny
of a large number of selected plants; (2) determining the mean of each
progeny for the characters under consideration; (3) arraying the progeny
means for each character and selecting progenies whose means fall on
a certain segment of the distribution curve (the segments to be chosen

COTTON 67

by the breeder on the basis of the relative importance of one character as
compared to the others, and to the original variability of the material,
etc.) ; and finally, (4) massing of all the selected lines to form a bulk
planting from which another selection cycle may be started. According
to Harland (1949) "continuous selection by this method for any meas-
urable character tends to produce a system of gene frequencies resulting
in the manifestation of the character at a higher level through the elimi-
nation of alleles, the combinatory effects of which are ordinarily antag-
onistic to the standards laid down for the character." The "mass-pedigree
selection" system makes full use of the principle of progeny testing, and
at the same time is designed to preserve genetic variability through the
use of a large number of lines and a broad adaptation base by propagat-
ing massed lines under varying seasonal and other environmental con-
ditions. Furthermore, the method would preserve certain genes for vigor
as heterozygous loci, a condition which, in Harland's view, would give
the stock an advantage over strains in which the same genes were homo-
zygous. Used with considerable success in rehabilitating Tanguis cotton,
the method is recommended by Harland for wider application and use
instead of pure line selection systems. The system is similar in principle
to the older "type selection" methods, but recognizes, defines, and meas-
ures the component characters of the type, and provides a much more
critical progeny test. Actually, the method is not at odds with the
progeny-row method as employed by many contemporary plant breeders,
in which a number of tested lines of generally similar characteristics arc
massed at certain stages in the testing procedure, and the seed stock
distributed as an agricultural variety. The "mass-pedigree selection"
method obviates detailed records of families and lines, while the "rep-
licated progeny-row" method would seem to give a more precise test
of the all-important primary selection.

The American Upland cotton-breeding program of the Texas Agri-
cultural Experiment Station largely employs varietal hybrids. Because
of the wide range of soil and climatic conditions in Texas, the breeding
program must emphasize particularly the conservation and maintenance
of genetic variability in progeny tests extending over a number of years.
Probably in no other cotton-growing region is the maintenance of a
"broad adaptation base" so important. The system involves: (1) selec-
tion of single plants in F 2 ; (2) "duplicate progeny-row" evaluation in
F 3 ; (3) replicated tests of within-family bulks from F 3 to F 6 -s; and
finally, (4) reselection within families followed by massing of seed from
similar lines of superior performance.

Breeders have taken renewed interest in the native American Upland
cottons of Southern Mexico and Central America as a source of new germ

68 T. R. RICHMOND

plasm. Inasmuch as present cultivated varieties were derived from these
native stocks, important new economic characters if found in native
cottons, should be relatively easily transferred to cultivated stocks. In
1946, T. R. Richmond and C. W. Manning made a preliminary explora-
tion trip to the area. The next year S. G. Stephens, who was then em-
ployed by the Empire Cotton Growing Corporation, explored the area;
and in 1948, J. 0. Ware and C. W. Manning made a rather extensive sur-
vey of a wide area including parts of San Salvador, Guatemala and Mex-
ico. As a result of these recent expeditions, more than 640 stocks have
been collected.

The backcross method of breeding is clearly indicated, when the ob-
jective is to transfer a character which is conditioned by one simple gene,
or a small number of such genes, from one variety or type to another
without deleteriously affecting the desirable characters of the latter.
Only a small number of plants per progeny is required in each backcross
cycle when such characters are available. Unfortunately, in cotton, only
a very few such simply inherited and easily distinguished characters have
been recognized. The genes controlling resistances to certain cotton
diseases are the best examples, but more will doubtless be discovered as
the work of "sifting" recent foreign introductions proceeds.

The backcross method has been employed in varietal hybrids in at-
tempts to transfer characters which are inherited according to a quanti-
tative scheme. This approach should not be entirely discouraged, but
in its application the low probability of success should be pointed out.
Knight (1945) goes so far as to state that the backcross "system is
valueless in intraspecific (varietal) crosses because such hybrids, by the
first backcross, are normally very similar vegetatively to the backcross
parent (thus) the hybrid may look like the backcross parent but it
still contains a large proportion of the donor genotype and is unlikely to
breed true for the various qualitative and quantitative characters de-
sired."

Certain modifications or adaptations of the backcross and the
straight hybrid systems, or combinations of systems, may prove useful
in American Upland cotton breeding. Richey's (1927) "convergent im-
provement" method is a case in point. Another method suggested by
Sprague (1946) and recommended for use in certain cotton experiments
by Richmond (1949) is called "cumulative selection." In this method,
lines bearing the character under study, from as many diverse sources
as possible or practical, are selected in F 2 . After isolating relatively
good complexes, but without carrying on selection in each line to its ulti-
mate conclusion (complete uniformity), the selected lines are immediately
crossed in all possible combinations and carried to a new F2 in bulk.

COTTON 69

Selection for the character is then practiced again, and the cycle repeated
until the level of acceptability is reached.

The greatest range in variability in cotton exists among the wild and
cultivated taxonomic species; these are fairly well distributed over the
tropical and sub-tropical areas of the world. The two cultivated Amer-
ican species, G. hirsutum and G. barbadense cross readily and give
fertile progeny, as do the two cultivated Asiatic species, G. arboreum
and G. herbaceum. Though, literally, thousands of attempts have been
made, there has not as yet been selected from the progeny of a species
cross, a strain in which two quantitatively inherited characters have
been combined in the full expression of their original parental form. It
is extremely difficult, in fact, to find good examples of satisfactory inter-
mediate expressions of quantitative characters. The work of Harland
(1936) has shown that stable complexes of interrelated genes are built
up within each species. When such species are crossed, the "genetic
balance" is disrupted in the FU generation giving a maze of abnormal
and unbalanced types. Stephens (1950) presents evidence to show "that
multiple gene substitution, such as that suggested by Harland, is not
sufficient to explain the cytological, genetic and breeding phenomena
encountered in critical studies of fertile interspecific hybrids and their
progenies in Gossypium." It is stated that "recent evidence from studies
of amphidiploids which casts doubt on the validity of 'normal' chromo-
some pairing and hybrid fertility as indices of structural homology."
The structural (cryptic) differences to which Stephens refers are con-
sidered to involve much smaller 'pieces' of chromosome than those
involved in the gross structural changes which may be recognized cyto-
logically, and which may cause partial or total sterility in the hybrid
progeny.

The process of cotton improvement by breeding from species hybrids
will unquestionably be of long duration. Future requirements of the
cotton industry are likely to be such that they can be met more readily
by the introduction of characters outside the range of present cultivated
varieties. Not only does it seem important to continue research in those
inter-specific hybrids which cross readily, but methods of utilizing the
wealth of new germ plasm in the wild species of the world should receive
special attention.

Crosses between many of the cotton species with 13 pairs of chromo-
somes and all of the crosses of 13 by 26 paired species, if successful at
all, gave sterile hybrids, until recently. Little more than 10 years ago
Beasley (1942) induced fertility in a cross of Arizona Wild (G. thurberi
Tad. by G. arboreum) by doubling the chromosomes of the sterile
hybrid by treatment with colchicine. The resulting amphidiploid (n =

70 T. R. RICHMOND

26) was partially fertile with American Upland (G. hirsutum, n =
and a high degree of fertility was reached after the first backcross to
Upland. A vast new field of cotton improvement was thus opened up,
but it is, as yet, only barely explored. When this backcross was made,
it was found that fiber strength increased materially over anything pre-
viously known in the cultivated Upland cottons. This, of course, was
quite unforeseen, as the Asiatic parent was by no means outstanding in
respect to fiber strength and the American Wild parent had no spinnable
fibers at all. Subsequent studies have shown that the fibers of the new
hybrid have narrow cross-sectional areas (narrow perimeters), a char-
acter introduced from the apparently worthless wild cotton from Arizona.
Hence, in species crosses, the apparent valuable characters available for
transfer are probably only a fraction of the important qualities yet to
be discovered.

Breeding experiments with the so-called triple hybrid (Asiatic x
American Wild (doubled) x American Upland) have been in progress
at the Texas and North Carolina Stations from the time of Beasley's
first induced amphidiploid. Eventual success in these and other studies
involving species hybrids appears to rest on the not altogether vain
expectation that a favorable crossover in a given differential segment
will occasionally occur. Knight's (1946a) transference of blackarm
(bacterial blight) resistance from one species of cotton to another is
sufficient to demonstrate the tremendous benefits to be enjoyed when
success is finally achieved.

4- Hybrid Vigor in F t and Advanced Generations

The substantial increases in yield and improvement in other economic
characters obtained in first generation and double crosses in corn, and
the almost universal acceptance of hybrid seed as the propagating ma-
terial for commercial corn production, has led some cotton investigators
to re-examine the possibilities of similar methods in cotton. In recent
years several experiments, designed primarily to study hybrid vigor, have
been undertaken. Three inbred lines from varieties of American Upland
cotton and the 6 possible FI and F 2 hybrids from the lines were studied
by Kime and Tilley (1947) for a period of 3 years. The FI seed was
produced by hand pollination and the Fo by selfing. Significantly higher
yields were reported, for a 3-year period, for each of the hybrids as com-
pared to its highest yielding parent. The important economic considera-
tion was, however, the finding that by no means did all of the hybrids
excel significantly in the single years. Clearly, the significant increases
obtained for the 3-year means resulted, in all but one case, from the
additive effects of small single-year differences which were nearly always

COTTON 71

in favor of the hybrid. Significant increases in the yield of the advanced
generation (F 2 ) over the most productive parent were recorded for only
two crosses, and these occurred in only one year.

Simpson (1948a) reported on the heterosis exhibited in the progeny
of varieties and strains propagated by open pollinated seed from a pro-
duction area in which natural cross pollination approximated 50 per cent.
The progenies of 7 varieties were tested, seed of which were produced
under two conditions, i.e., (1) open pollination (crossed) in a 25 entry
variety tost, and (2) in isolated blocks (inbred). It was reported that
the yields of the 7 progenies from the crossed seed exceeded those of
"inbred" seed by 5.7 to 44.2 per cent, or an average of 15.4 per cent, The
practical significance of the data lies not so much in the average increase
of 15.4 per cent attributed to hybrid vigor, but in a comparison of the
"crossed" stocks with the highest yielding agronomic variety. When
such comparisons are made, it is seen that significant yield differences in
favor of the "crossed" stocks occur in only an occasional instance. Simp-
son (1948b) also conducted an experiment to measure the amount of
heterosis resulting from natural crossing in test plots at several locations
in the Cotton Belt.

The great handicap to the practical utilization of hybrid vigor in
cotton is the difficulty of producing the hybrid seeds. Two methods have
been proposed by Simpson (1948a), both of which require natural cross-
ing by bees. For certain conditions in India, BaUisubrahmanyan and
Narayanan (1947) have proposed vegetative cuttings as a method of
propagating FI cotton hybrids on a commercial scale. Probably the most
hopeful method now on the horizon is the use of male steriles as "mother"
plants. Male-sterile stocks, when interplanted with normal lines in areas
of high natural cross-pollination, would yield hybrid material of known
composition. The difficulty so far has been the discovery of a suitable
male-sterile type. Recently several apparently male-sterile lines have
been studied and found to be only partially, or periodically, so.

5. Special Phases

a. Fiber Properties. The development of suitable instruments and
methods for testing and predicting fiber and spinning properties, as dis-
cussed under VIII, has given great impetus to breeding for improved or
special fiber properties.

The data from regional variety studies have pointed to one salient
fact; that is, the variety (genetic constitution) is the single most im-
portant consideration in the determination of quality of cotton fibers.
That considerable genetic variability for fiber properties must have been
present in certain of the relatively modern varieties of American Upland

72 T. E. RICHMOND

cotton is evident by the fact that breeders have made significant improve-
ments in fiber strength and related characters by selection within varieties
and varietal hybrids. The value of the wild and primitive cottons of the
world as a source of fiber properties which are outside the range of Amer-
ican Upland types already has been emphasized. Strains with fibers 20
to 30 per cent stronger than the better Upland types have been extracted
from species hybrids at the Texas Station, and continuing yield trials
show that perceptible, though slow, progress is being made in the transfer-
ence of great fiber strength to acceptable Upland stocks. A wealth of
untried fiber characters remains in the base material.

b. Resistance to Disease, Insects, and Unfavorable Environments.
The important problems in breeding for resistance to cotton diseases have
been referred to under IV, and will not be further discussed here.

Though insects annually take a toll of millions of dollars from the
American cotton crop, scientists in this country have given scant atten-
tion to the extremely important problem of breeding for insect resistance.
Evidence obtained to date from a number of sources supplies a basis for
optimism as to the distinct possibilities for cotton improvement which
lie in this long neglected field. According to Knight (1946b), who re-
ferred to previous reports, "G. thurberi appears, at Shambat, to be im-
mune to pink bollworm (Pectinophora gossypiella) , and G. armourianum
itself shows very marked pink bollworm-resistance." British cotton in-
vestigators have worked extensively on Jassids (Emposca ficialis) and
have found resistance to be associated with plant pilosity, the more resist-
ant types showing the greatest degree of "hairiness" on the under side of
the leaf (Parnell et a/., 1949). Dunnam (1936) and Dunnam and Clark
(1939) found that, under dry conditions, hairy varieties retained signifi-
cantly more calcium arsenate dust than glabrous varieties; on undusted
cotton, the aphid population increased in direct proportion to the number
of hairs on the lower leaf sQrfaces.

Probably no more remunerative use can be made of the extensive
collections of wild and primitive cottons now available to breeders than
to subject them and certain of their hybrids to critical study for resist-
ance to major cotton insect pests. The same material should prove use-
ful as a source of new or unfamiliar genes for drought resistance, cold
tolerance, high oil and low gossypol content of the seeds, and many other
economic characters.

c. Adaptation to Mechanical Harvesting. Harvesting is an expensive
cotton production operation. Currently, about 90 per cent is performed
by hand. The recent development of machines which will perform this

COTTON 73

operation represents a great technological advance in the mechanization
of cotton production. As discussed under VI, mechanical cotton har-
vesters are of two general types: (1) picker; and (2) stripper.

The choice of variety is a very important consideration in planning
a mechanized-production program. Fortunately, several of the modern,
rapid fruiting, early maturing, American Upland Varieties are fairly
well adapted to spindle-type picking, particularly when grown under
planned systems of spacing and culture. Future work is being directed
along two lines: (1) the achievement of more mechanical efficiency in
picking, i.e., obtaining a higher per cent of the cotton harvested as com-
pared to the total available open cotton; and (2) obtaining raw cotton
which is equal or superior in grade and other lint qualities to that of
hand-picked cotton, i.e., freedom from leaf, stem, and bract trash and
other foreign matter, and with a minimum of discoloration and staining
due to exposure to weather.

The present "type ideal" for spindle picking seems to be a plant (1)
that will grow in a more or less upright position but at the same time be
early in fruiting habit and fairly determinate in growth habit, (2) that
will set its fruit in an evenly spaced manner all over the plant but be-
ginning well off the ground, (3) that will have bolls which will allow
the cotton to fluff and at the same time cause it to stick in the burr
strongly enough for good storm resistance, (4) that will mature its fruit
early and in a very short space of time, and (5) that will shed its leaves
readily when the major portion of the bolls have matured. Even a casual
consideration of such an "ideal" plant will reveal several physiological
and morphological antagonisms. The idea of obtaining early fruiting
and at the same time a set of bolls well off the ground, may be men-
tioned as one.

The "ideal" plant type for stripper-machine harvesting is a dwarf
or semi-dwarf, wilh short to medium fruiting branches, which will fruit
rapidly and mature its fruit early and well off the ground. The last
allows the lifters of the machine to slip under the lowest branches and
engage the cotton without picking up dirt and extraneous plant materials.
The seed cotton should be closely held in the boll at maturity as all or
most of the bolls on the plant must be mature before the stripper enters
the field; but the locks need not fluff. Reference has been made else-
where in this paper to a mutant boll character in which the seed cotton
is closely held in bolls which open only partially. According to Lynn
(1949), this stormproof-boll type suggests a complex of modifiers that
operate in connection with the main gene to cause varying expressions
of the characters. Uniform lines have been extracted which range from
the extreme mutant expression to types indistinguishable from the FI.

74 T. B. RICHMOND

The stormproof-boll character has been established in high yielding
strains adapted to stripper-type harvesting by workers at the Texas
Agricultural Experiment Station's substations at Lubbock and Chilli-
cothe.

Regardless of the type of mechanical harvester employed, any new
character, or refinement of existing characters, which would result in less
foreign matter in the harvested cotton and ginned lint would be a valu-
able contribution to the problem of mechanical harvesting. Varieties
with many large, spiny plant and leaf hairs give lower grades of ginned
lint from machine-picked cotton than those with fewer and shorter hairs.
A variety, designated as DELTA SMOOTH LEAF, which is almost free of leaf
and stem hairs has been developed by workers at the Mississippi Delta
Branch Station. The variety gives significantly higher grades of
machine-picked cotton and work is under way at several locations to
improve its yield and other agronomic properties, or to transfer the
character to other varieties. The prominently-toothed bracts (brac-
teoles), characteristic of Upland cotton, are known to contribute mate-
rially to lint trash. Attempts are being made through breeding to reduce
the size of the bracts or to eliminate them entirely. Two sources of
breeding material are worthy of mention: (1) the small, almost toothless
bracts of some members of the marie galante group of Upland cottons;
and (2) a deciduous-bract type, first studied at the Mississippi Delta
Branch Station, in which the bracteoles fall from the boll at maturity.

REFERENCES
These citations are grouped according to the major divisions of the article.

III. PHYSIOLOGY OF THE COTTON PLANT

Anderson, D. B., and Kerr, T. 1943. Plant Physiol. 18, 261-269.

Andrews, W. B., and Coleman, R. 1939. Proc. Assoc. Southern Agr. Workers 40,

61.

Barker, H. T3. 1946. f/.S. Dcpt. Agr. Tech. Bull. 931.
Berkley, E. E. 1945. Textile Research J. 15, 460-467.
Berkley, E. E., Woodward, O. C., Barker, H. D., Kerr, T., and King, C. J. 1948.

f/.*S. Dcpt. Agr. Tech. Bull. 949.

Biddulph, O. 1949. Paper presented Mineral Nutrition Symposium, Madison, Wis.
Biddulph, O., and Brown, H. D. 1945. Am. /. Botany 32, 182-188.
Biddulph, O., and Markle, J. 1944. Am. J. Botany 31, 65-70.
Brown, A. B. 1946. Louisiana Agr. Expt. Sta. Bull. 406.
Brown, G. A., Holdeman, Q. L., and Hagood, E. S. 1948. Louisiana Agr. Expt

Bull. 426.

Brown, H. B., and Pope, H. W. 1939. Louisiana Agr. Expt. Sta. Bull. 306, 15.
Coleman, R. 1945. Better Crops with Plant Food 29, No, 4, 18-20, 48-50.

COTTON 75

Collander, R. 1941. Plant Physiol 16, 691-720.

Cooper, H. P. 1945. Soil Sci. 60, 107-114.

Cooper, H. P., and Garman, W. H. 1942. Soti Sci. Am. Proc. 7, 331-338.

Cooper, H. P., Mitchell, J. H., and Page, N. R. 1947. Soil Sci. Am. Proc. 12, 364-

369.
Cooper, H. P., Paden, W. R., Garman, W. H., and Page, N. R. 1948. Soil Sci. 65,

75-96.

Crowther, F. 1934. Ann. Botany 48, 875-918.
Crowther, F. 1944. Ann. Botany FN.S.] 8, 213-257.

Dastur, R. H. 1948. Scientific Mono. No. 2, Indian Central Cotton Comm. (Re-
vised 2nd ed.) (Bombay).

Dunlap, A. A. 1945. Texas Agr. Expt. Sta. Bull. 677.
Dunlap, A. A. 1948. Phytopathology 38, 638-644.
Eaton, F. M. 1924. Botan. Gaz. 57, 311-321.
Eaton, F. M. 1942. /. Agr. Research 64, 357-399.
Eaton, F. M. 1947. Textile Research 17, 568-575.
Eaton, F. M. 1950. Botan. Gaz. Ill, (3), 314-319.
Eaton, F. M., and Ergle, D. R. 1948. Plant Physiol. 23, 169-187.
Eaton, F. M., and Joham, H. E. 1944. Plant Physiol. 44, 507-518.
Eaton, F. M., Lyle, E. W., Rouse, J. T., Pfeiffenberger, G. W., and Tharp, W. H.

1946. /. Am. Soc. Agron. 38, 1018-1033.

Eaton, F. M, and Rigler, N. E. 1945. Plant Physiol. 20, 380-411.
Ensminger, L. E., and Cope, J. T., Jr. 1947. J. Am. Soc. Agron. 39, 1-11.
Ergle, D. R., and Dunlap, A. A. 1949. Texas Agr. Expt. Sta. Bull. 713.
Ergle, D. R., and Eaton, F. M. 1949. Plant Physiol. 24, 373-388.
Gaines, J. C. 1947. J. Econ. Entomol. 40, 434.
Gaines, J. C., Owen, W. L., Jr., and Wipprecht, R. 1947. J. Econ. Entomol. 40,

113.
Hall, N. S., Nelson, W. H., Krantz, B. A., Welch, C. D., and Dean, L. A. 1949.

Soil Sci. 68, 151-157.
Harris, H. C., Bledsoe, R. W., and Calhoun, P. W. 1945. /. Am. Soc. Agron. 37,

323-329.

Holley, K. T., and Dulin, T. G. 1943. Georgia Expt. Sta. Bull. 229.
Holt, M. E., and Volk, N. J. 1945. J. Am. Soc. Agron. 37, 821-827.
Hooton, D. R., Jordan, H. V., Porter, D. D., Jenkins, P. M., and Adams, J. E.

1949. U.S. Dept. Agr. Tech. Bull. 979.
van Itallie, TH.B. 1948. Soil Sci. 65, 393-415.

Jacobson, L., and Overstreet, R. 1947. Am. J. Botany 34, 415-420.
Kerr, T., and Anderson, D. B. 1944. Plant Physiol. 19, 338-349.
Konstantinov, N. N. 1934. Maskva Eng. Summary, pp. 72-74.
Leonard, O. A. 1945. J. Am. Soc. Agron. 37, 55-71.
Leonard, 0. A., and Pinckard, J. A. 1946. Plant Physiol. 21, 18-36.
Lundegardh, H. 1947. Ann. Rev. Biochem. 16, 503-528.
Manns, T. F., Churchman, W. L., and Manns, M. M. 1937. Delaware Agr. Expt.

Bull. 207, 45-46.

Mason, T. G., and Phillis, E. 1944. Ann. Botany [N.S.] 7, 399-408.
Mason, T. G., and Phillis, E. 1945. Ann. Botany [N.S.I 9, 335-343.
Mathews, E. D. 1941. Georgia Expt. Sta. Circ. 127.
Nelson, W. L. 1949. J. Am. Soc. Agron. 41, 289-293.
Nightingale, G. T. 1948. Botan. Rev. 14, 185-221.

76 J- E. ADAMS

Olson, L. C. 1942. Proc. Assoc. Southern Agr. Workers 43, 78.

Olson, L. C., and Bledsoe, R. P. 1942. Georgia Expt. Sta. Bull. 222, 16.

Phillis, E., and Mason, T. G. 1942. Ann. Botany [N.S.] 6, 437-442.

Rigler, N. E., Ergle, D. R., and Adams, J. E. 1937. Soil Sci. Soc. Proc. 2, 367-374.

Sayre, C. B., and Vittum, M. T. 1947. J. Am. Soc. Agron. 39, 153-161.

Singh, S., and Greulach, V. A. 1949. Am. J. Botany 36, 646-651.

Skinner, J. J., Futral, J. O., and McKaig, N. 1944. Georgia Agr. Expt. Sta. Bull.

235.

Somner, A. L. 1945. Soil Sci. 60, 71-79.

Spiegelman, S., and Reiner, J. M. 1942. Growth 6, 367-390.
Staten, G. 1946. J. Am. Soc. Agron. 38, 336-544.
Sturkie, D. G. 1947. Alabama Agr. Expt. Sta. Bull. 263.
Turner, J. H. 1944. J. Am. Soc. Agron. 36, 628-698.
Volk, N. J. 1946. J. Am. Soc. Agron. 38, 6-12.
Wadleigh, C. H. 1944. Arkansas Agr. Expt. Sta. Bull. 446, 138.
Wadleigh, C. H. 1949. Ann. Rev. Biochem. 18, 658-678.
Wallace, A, Toth, S. J., and Bear, F. E. 1948a. Soil Sci. 65, 249-258.
Wallace, A., Toth, S. J., and Bear, F. E. 1948b. Soil Sci. 65, 477-486.
Willis, L. G. 1936. Soil Sci. Soc. Am. Proc. 1, 291-297.
Younge, O. R. 1941. Soil Sci. Soc. Am. Proc. 6, 215-218.

IV. DISEASES OF COTTON

Atkinson, G. F. 1891. Alabama Agr. Expt. Bull. 27, 1-16.

Atkinson, G. F. 1892. Alabama Agr. Expt. Bull. 41.

Barducci, T. D. 1942. Estac. cxpt. agr. LaMolina (Lima, Peru) Bol. 23.

Brown, J. G. 1937. Plant Disease Reptr. 21, 368.

Carpenter, C. W. 1914. Phytopathology 4, 393.

Cuba, E. F. 1932. Massachusetts Agr. Expt. Sla. Bull. 292.

Fahmy, T. 1929. Ministry Agr. Egypt, Leaflet No. 11, 1-16.

Faulwetter, R. C. 1917. J. Agr. Research 10, 639-648.

Gilbert, W. W. 1914. U.S. Dept. Agr. Farmers Bull. 625.

Godfrey, G. H. 1923. U.S. Dept. Agr. Farmers Bull. 1345.

Godfrey, G. H. 1943. Texas Agr. Expt. Sta. Progress Rept. 837.

Hansford, C. G., Hosking, H. R,, Stoughton, R. H., and Yates, F. 1933. Ann.

Applied Biol. 20, 404-420.

Hare, J. F., and King, C. J. 1940. Phytopathology 30, 679-684.
Herbert, F. W., and Hubbard, J. W^. 1932. U.S. Dept. Agr. Circ. 211, 7.
Jacks, H. 1945. New Zealand J. Sci. Technol. 27(2), 93-97.
Jordan, H. V., Nelson, A. H., and Adams, J. E. 1939. Soil Sci. Soc. Am. Proc. 4,

325-328.
Jordan, H. V., Adams, J. E., Hooton, D. R., Porter, D. D., Blank, L. M., Lyle,

E. W., and Rogers, C. H. 1948. 7.5. Dept. Agr. Tech. Bull. 948.
King, C. J. 1923. J. Agr. Research 23, 525-527.
King, C. J. 1937. 7.5. Dept. Agr. Circ. 425.
Lewis, A. C., and McLendon, C. A. 1917. Georgia Entomology Board Bull. 46,

1-34.
Lyle, E. W., Dunlap, A. A., Hill, H. O., and Hargrove, B. D. 1948. Texas Agr,

Expt. Sta. Bull. 699.

McNamara, H. C., and Hooton, D. R. 1930. 7.5. Dept. Agr. Circ. 85.
Massey, R. E. 1930. Empire Cotton Growing Rev. 7, 185-195.

COTTON 77

Miles, L. E., and Persons, T. D. 1932. Phytopathology 22, 767-773.

Neal, D. C., Wester, R. E., and Gunn, K. C. 1932. Science 75, 139-140.

Orton, W. A. 1900. U.S. Dept. Agr., Division Vegetable Physiology Path. Butt.

27, 1-16.

Pammel, L. H. 1888. Texas Agr. Expt. Sta. Bull. 4.
Pammel, L. H. 1889. Texas Agr. Expt. Sta. Bull. 7.
Presley, J. T. 1946. Mississippi Farm Res. 9(11), 1-2.
Presley, J. T. 1949. Mississippi Farm Res. 12(4), 1.
Rolfs, F. M. 1915. Cornell Univ. Agr. Expt. Sta. Mem. 8:998. 413.
Rolfs, F. M. 1935. Phytopathology 25, 971. (Abstract).
Rudolph, B. A., and Harrison, G. J. 1939. Phytopathology 29, 753.
Salter, R. M. 1946. Rept. Chief. Bur. Plant Industry Soils Agr. Eng. f U,S. Dept.

Agr. 35. ! ;

Sherbakoff, C. D. 1929. Tennessee Agr. Expt. Sta. Circ. 24, 2.
Sherbakoff, C. D. 1949. Botan. Rev. 15(6), 377-422.
Simpson, D. M., and Weindling, R. 1946. J. Am. Soc. Agron. 38, 630-635.
Smith, A. L. 1948. Phytopathology 38, 943-947.

Smith, A. L., and Ballard, W. W. 1947. Phytopathology 37, 436-437.
Smith, E. F. 1901. U.S. Dept. Agr., Division Vegetable Physiology Path. Bull. 28,

153.

Stoughton, R. H. 1933. Ann. Applied Biol 20, 590-611.
Streets, R. B. 1938. Univ. Ariz. Ext. Circ. 103.
Taubenhaus, J. J., and Ezekiel, W. N. 1935. Texas Agr. Expt. Sta. 48th Ann.

Rept., p. 86.
Uppal, B. N., Kulkarni, Y. S., and Ranadive, J. D. 1941. Rev. Applied My col.

20, 256.

Watson, J. R. 1945. Proc. Florida Acad. Sci. 7 (Nos. 2-3).
Watson, J. R., and Goff, C. C. 1937. Florida Agr. Expt. Sta. Bull. 311.

V. INSECT PESTS

Becnel, I. J., Mayeux, H. S., and Roussel, J. S. 1947. J. Econ. Entomol. 40, SOS-
SIS.

Cassidy, T. P., and Barber, T. S. 1939. J. Econ. Entomol. 32, 99-104.

Chapman, A. J., Richmond, C. A., and Fife, L. C. 1947. J. Econ. Entomol. 40,
575-576.

Goad, B. R. 1918. U.S. Dept. Agr. Bull. 731.

Goad, B. R., and Cassidy, T. P. 1920. U.S. Dept. Agr. Bull. 875.

Dunnam, E. W., and Calhoun, S. L. 1948. J. Econ. Entomol. 41, 22-25.

Eddy, C. 0., and Livingston, E. M. 1931. South Carolina Agr. Expt. Sta. Bull.
271.

Ewing, K. P. 1929. J. Econ. Entomol. 22, 761-765.

Ewing, K. P. 1931. J. Econ. Entomol. 24, 821-827.

Ewing, K. P., and McGarr, R. L. 1936. /. Econ. Entomol. 29, 80-88.

Ewing, K. P., and McGarr, R. L. 1937. J. Econ. Entomol. 30, 125-130.

Ewing, K. P., and Parencia, C. R., Jr. 1947. J. Econ. Entomol. 40, 374-381.

Ewing, K. P., and Parencia, C. R., Jr. 1948. /. Econ. Entomol. 41, 558-563.

Fenton, F. A., and Dunnam, E. W. 1928. /. Agr. Research 36, 135-149.

Fletcher, R. K., Gaines, J. C., and Owen, W. L. 1947. /. Econ. Entomol. 40, 594-
596.

Gaines, J. C. 1933. J. Econ. Entomol. 26, 963-971.

78 J- E. ADAMS

Gaines, J. C. 1934. /. Econ. Entomol. 27, 740-743.

Gaines, J. C. 1941. J. Econ. Entomol. 34, 505-507.

Gaines, J. C. 1942. Iowa State Coll. J. Sci. 17, 63-65.

Gaines, J. C. 1944. /. Econ. Entomol. 37, 723-725.

Gaines, J. C., and Dean, H. A. 1947. /. Econ. Entomol. 40, 365-370.

Gaines, J. C., and Dean, H. A. 1948. J. Econ. Entomol. 41, 548-554.

Gaines, J. C., Dean, H. A., and Wipprecht, R. 1948. /. Econ. Entomol 41, 510-512

Gaines, J. C., and Johnston, H. G. 1949. Acco Press, June Issue, pp. 15-18.

Gaines, R. C., and Young, M. T. 1948. J. Econ. Entomol. 41, 19-22.

Gaines, R. C., Young, M. T., and Smith, G. L. 1947. J. Econ. Entomol. 40, 600-

603.

Howard, L. O. 1898. U.S. Dcpt. Agr. Div. Entomol. Bull 18, 101.
Hunter, W. D. 1918. UJS. Dcpt. Agr. Bull. 723.
Hunter, W. D. 1924. J. Econ. Entomol. 17, 604.
Hunter, W. D. 1926a. U.S. Dept. Agr. Dept. Circ. 361.
Hunter, W. D. 1926b. UJS. Dept. Agr. Bull. 1397.

Hunter, W. D., and Hinds, W. E. 1904. U.S. Dept. Agr. Div. Entomol. Bull. 45.
Hunter, W. D., and Pierce, W. D. 1912. U.8. Dept. Agr. Bur. Entomol. Bull. 116.
Iglinsky, W., Jr., and Gaines, J. C. 1949. J. Econ. Entomol. 42, 702-705.
Ivy, E. E., and Ewing, K. P. 1946. J. Econ. Entomol. 39, 38-41.
Ivy, E. E., Parcncia, C. R., Jr., and Ewing, K. P. 1947. J. Econ. Entomol. 40,

513-517.
Loftin, U. C., McKinney, K. B., and Hanson, W. K. 1921. UJS. Dcpt. Agr. Bull.

918.

McGregor, E. A. 1948. J. Econ. Entomol. 41, 684-687.
Moreland, R. W., and Bibby, F. F. 1931. J. Econ. Entomol. 24, 1173-1181.
Moreland, R. W., Ivy, E. E., and Ewing, K. P. 1941. J. Econ. Entomol. 34, 508-

511.

Newell, W., and Smith, C. D. 1909. Louisiana State Crop Pest Comm. Circ. 33.
Ohlendorf, W. 1926. UJS. Dept. Agr. Bull. 1374.
Painter, R. H. 1930. J. Agr. Research 40, 485-516.
Parencia, C. R., Jr., Ivy, E. E., and Ewing, K. P. 1946. J. Econ. Entomol 39,

329-335.

Quaintance, A. L., and Brues, C. T. 1905. UJS. Dept. Agr. Bur. Entomol Bull 50.
Rainwater, C. F., and Bondy, F. F. 1947. J. Econ. Entomol 40, 371-373.
Reinhard, H. J. 1926a. Texas Agr. Expt. Sta. Bull 339.
Reinhard, H. J. 1926b. Texas -Agr. Expt. Sta. Circ. 40.
Reinhard, H. J. 1927. Texas Agr. Expt. Sta. Bull 356.
Riley, C. V. 1885. U.S. Entomol Comm. Kept. 4, 355-384.
Stevenson, W. A., and Kauffman, W. 1948. /. Econ. Entomol. 41, 583-585.
Wardle, R. A., and Simpson, R. 1927. Ann. Applied Biol 14, 513-528.
Watts, J. G. 1934. J. Econ. Entomol 27, 1158-1159.
Watts, J. G. 1936. South Carolina Agr. Expt. Sta. Bull 306.
Watts, J. G. 1948. J. Econ. Entomol 41, 543-547.

VI. IMPROVEMENTS IN PRODUCTION PRACTICES

Alsmeyer, H. L. 1949. Proc. 3rd Ann. Beltivide Cotton Mechanization Con].

(National Cotton Council).

Barr, G. W. 1949. Arizona Agr. Expt. Sta. Bull 220.
Bonnen, C. A., an4 Thjbodeaux, B, H t 1937, Texas Agr. Expt. Sta, Bull 544,

COTTON 79

Fairbank, J. P. 1948. Proc. 2nd Ann. Beltwide Cotton Mechanization Conf. (Na-
tional Cotton Council).

Grimes, M. A. 1947. Texas Agr. Expt. Sta. Bull. 697.

Gull, P. W., and Adams, J. E. 1945. Mississippi Agr. Expt. Sta. Bull. 423.

Harris, K. 1947. Arizona Agr. Expt. Sta. Bull. 210.

Harris, K., and Hawkins, R. S. 1942. Arizona Agr. Expt. Sta. Bull. 181.

Hinklc, D. A., and Staten, G. 1941. New Mexico Agr. Expt. Sta. Press Bull. 923.

Jones, D. L. 1948. Proc. 2nd Ann. Beltwide Cotton Mechanization Conf. (Na-
tional Cotton Council).

Jones, D. L., and Jones, W. L. 1945. Texas Agr. Expt. Sta. Progress Rept. 949.

McDowell, C. H. 1937. Texas Agr. Expt. Sta. Bull. 543.

Magee, A. C., Bonnen, C. A., and Thibodeaux, B. H. 1944. Texas Agr. Expt. Sta.
Bull. 652.

Neely, J. W., and Brain, S. G. 1944. Mississippi Agr. Expt. Sta. Circ. 118.

Smith, H. P. 1949a. Progressive Farmer 64(5), 18.

Smith, H. P. 1949b. Progressive Farmer 64(8), 9.

Smith, H. P., Killough, D. T., Jones, D. L., and Byron, M. H. 1939. Texas Agr.
Expt. Sta. Bull. 580.

Smith, H. P., Killough, D. T., and Jones, D. L. 1946a. Texas Agr. Expt. Sta. Bull.
686.

Smith, H. P., Rouse, J. T., Killough, D. T., and Jones, D. L. 1946b. Texas Agr.
Expt. Sta. Bull. 683.

Staten, G., and Hinkle, D. A. 1942. New Mexico Agr. Expt. Sta. Press Bull. 941.

Thomas, W. I. 1948. Arizona Agr. Expt. Sta. Bull. 214.

Williamson, M. N., and Rogers, R. H. 1948a. Texas Agr. Expt. Sta. Progress
Rept. 1111.

Williamson, M. N., arid Rogers, R. H. 1948b. Texas Agr. Expt. Sta. Progress
Rept. 1134.

VIII. FIBER PROPERTIES AND THEIR SIGNIFICANCE

Anderson, D. B., and Kerr, T. 1938. Ind. Eng. Chem. 30, 48-54.

Barker, H. D., and Pope, O. A. 1948. U.S. Dept. Agr. Tech. Bull. 970.

Barre, H. W., Barker, H. D, Berkley, E. E., Doyle, C. B., Marsh, P. B., Pearson,

N. L., and Ware, J. O. 1947. Better Cottons. In U.S. House Comm. on Agr.

Hearings (Study of Agricultural and Economic Problems of the Cotton Belt,

Pt. 2), 80 Cong., 1st session.
Berkley, E. E., Woodward, 0. C., Barker, H. D., Kerr, T., and King, C. J. 1948.

C/.S. Dept. Agr. Tech. Bull. 949.

Elting, J. P., and Barnes, J. C. 1948. Textile Research J. 18, 358-366.
Hermans, P. H., and Weidinger, A. 1949. J. Polymer Sci. 4, 135-144.
Hertel, K. L., and Zervigon, M. G. 1936. Textile Research J., 6, 331-339.
Hessler, L. E., Merola, G. V., and Berkley, E. E. 1948. Textile Research J. 18,

679-683.

Kerr, T. 1946. Textile Research J. 15, 249-254.

Nelson, M. L., and Conrad, C. M. 1948. Textile Research J. 18, 155-164.
Pearson, N. L. 1950. Textile Research J. 20(3), 152-164.
Pfeiffenberger, G. W. 1946. Textile Research J. 16, 338-343.
Pressley, E. H. 1942. Am. Svc. Testing Materials Bull. 118, 13-17.
Richardson, H. B., Bailey, T. L. W., Jr., and Conrad, C. M. 1937. VJ5. Dept. Agr.

Tech. Bull. 545.

80 J. E. ADAMS

Sisson, W. A. 1937. Textile Research J. 7, 425-431.

Sullivan, R. R., and Hertel, K. L. 1940. Textile Research J. 11, 30-38.

Webb, R. W. 1932. Am. Soc. Testing Materials Proc. 32, 764-771.

IX. BREEDING AND IMPROVEMENT

Balasubrahmanyan, R., and Narayanan, N. G. 1947. India Cotton Growing Rev.

2, 125-129.

Beasley, J. O. 1940. Am. Naturalist 74, 285-286.
Beasley, J. O. 1942. Genetics 27, 25-54.
Cook, O. F. 1932. U.S. Dept. Agr. Tech. Bull. 302.
Dunnam, E. W. 1936. /. Econ. Entomol. 29(6), 1085-1087.
Dunnam, E. W., and Clark, J. C. 1939. J. Econ. Entomol. 31(6), 663-666.
Harland, S. C. 1936. Biol. Revs. 11, 83.
Harland, S. C. 1940. Trop. Agr. (Trinidad) 17, 53-54.

Harland, S. C. 1943. Soc. Nac. Agr., Inst. Cotton Genetics (Lima, Peru) Bull. 1.
Harland, S. C. 1949. Empire Cotton Growing Rev. 26, 163-174.
Hutchinson, J. B. 1939. Chronica Botan. 5, 4/6. ,

Hutchinson, J. B. 1940. /. Genetics 40, 271-282.
Hutchinson, J. B., and Manning, H. L. 1943. Memo. Cotton Breeding Station,

(Trin.) Series A. No. 20. (Empire Cotton Growing Corp., London).
Hutchinson, J. B., and Panse, V. G. 1937. Indian J. Agr. Sci. 7, 531.
Hutchinson, J. B., Silow, R. A., and Stephens, S. G. 1947. The Evolution of

Gossypium and the Differentiation of the Cultivated Cottons. Oxford Univ.

Press, London.

Kime, P. H., and Tilley, R. H. 1947. /. Am. Soc. Agron. 39, 307-317.
Knight, R. L. 1945. J. Genetics 47, 76-86.
Knight, R. L. 1946a. Empire J. Exptl. Agr. 14, 153-174.

Knight, R. L. 1946b. Empire Cotton Grouting Rev., Rep. from Expt. Sta., p. 55.
Lynn, H. D. 1949. M. S. Thesis, Texas A. and M. Col. Library.
Parnell, H. E., King, H. E., and Ruston, D. F. 1949. Empire Cotton Growing

Corp., Research Mem. No. 7.
Richey, F. D. 1927. Am. Naturalist 41, 430.
Richmond, T. R. 1947. Rayon Textile Monthly 63, 373-375.
Richmond, T. R. 1949. Texas Agr. Expt. Sta. Bull. 716.
Simpson, D. M. 1948a. /. Am. Soc. Agron. 40, 970-979.
Simpson, D. M. 1948b. Proc. 4^k Ann. Assoc. Southern Agr. Workers.
Skovsted, A. 1937. J. Genetics 34, 97-134.
Sprague, G. F. 1946. Biol. Revs. 21, 101-120.
Stephens, S. G. 1950. Botan. Rev. 16(3), 115-149.
Vavilov, N. I. 1927. Bull. Applied Botany, Genetics, Plant Breeding (Leningrad)

16(2), 420-428.

Soil Nitrogen

L. E. ENSMINGER AND R. W. PEARSON

Alabama Agricultural Experiment Station, and U.S. Department of
Agriculture, Auburn, Alabama

CONTENTS

Page

I. Introduction 81

II. Factors Affecting Nitrogen Content of Soils 83

1. Temperate Soils 83

2. Tropical Soils 85

III. Nature of Organic Nitrogen in the Soil 87

IV. Nitrogen Transformations 89

1. Nitrification 89

2. Symbiotic Nitrogen Fixation 91

3. Nonsymbiotic Nitrogen Fixation 93

V. Effect of Cropping Practices on Nitrogen Level 94

VI. Nitrogen Economy of Eroded Soils 98

VII. Commercial Nitrogen vs. Barnyard Manure and Green Manures . . . 100

VIII. Nitrogen Trends in Various Parts of the U. S 104

1. General Trends 104

2. Southeast 106

3. Midwest 107

4. Great Plains 108

5. Irrigated Regions 108

References 109

I. INTRODUCTION

Records of ancient Chinese civilization, as well as early Greek and
Roman writings, refer to the use of animal manure as a means of increas-
ing plant growth. The general acceptance of this practice as well as the
growing of leguminous green manure crops offers strong evidence that
nitrogen has been a deficient element in the cultivated soils of the world
since the beginning of agriculture. Since the time of Liebig and Lawes,
there has been an increasing awareness of the acute nature of this de-
ficiency in many soils. A better understanding of the problem and the
availability of commercial forms of nitrogen has resulted in a rapid in-
crease in the use of nitrogen in fertilizers during the past few decades.
The world-wide use of nitrogen fertilizers, which according to Parker
(1946) increased from 720,000 tons of nitrogen in 1913 to 2,766,000 tons

L. E. ENSMINGER AND E. W. PEARSON

in 1937, is also a result of a changing agriculture. An indication of this
trend is shown in Pierre's (1949) statement that there has been an in-
crease of over 2 million acres of row crops in Iowa alone since 1941.
Such changes have been accompanied by a decrease in legume-grass
acreage which has resulted in an accelerated decline in soil fertility and
a marked increase in response obtained from additions of nitrogen. The
increasing use of hybrid corn, Zea mays, L., and improved varieties of
other crops has further widened the gap between the soil's nitrogen supply-
ing power and potential crop yield. Recent research in the South has
shown that corn responds to much higher levels of nitrogen applications
when plant population and fertility level are properly balanced than has
been previously recognized. These findings have emphasized the critical
deficit in the nitrogen economy of southern soils.

NITROGEN IN SURFACE FOOT
OF SOIL

Fig. 1. Soil nitrogen map of the United States. (Mehring, 1945.)

The total nitrogen content of most of the soils of the United States
(Fig. 1), exclusive of the Prairie and Chernozem groups, amounts to less
than 3,000 Ibs. in the surface acre layer. Large areas in the Southeastern
and far Western parts of the country contain less than 1,000 Ibs. per sur-
face acre. Although present knowledge is inadequate to permit an ac-
curate evaluation of the optimum quantity of soil nitrogen for any
specified set of conditions, it is known that less than 4 per cent of the total,
depending upon the crop grown and the cultural practices used, generally
becomes available for plant use during a season. It is probable that the

SOIL NITROGEN ' 83

amount of soil nitrogen that would have been considered adequate for
a particular soil two or three decades ago would no longer be considered
sufficient. More nitrogen is necessary now to take advantage of improved
crop varieties and better cultural methods. The level at which soil
organic matter and nitrogen can be maintained is determined, within
limits, by the operator's choice of cropping system and management
practices. A sound management program including the judicious selec-
tion of crop rotations that include legumes, conservation of crop residues,
proper care and utilization of farm manure, and adequate protection
against erosion will favor the maintenance of a desirable nitrogen balance
and materially decrease the need for additions of nitrogen fertilizer.

II. FACTORS AFFECTING NITROGEN CONTENT OF SOILS

1. Temperate Soils

Under virgin conditions the organic matter and nitrogen content of
a soil approaches an equilibrium value the magnitude of which depends
primarily upon climate, vegetation, and the physical characteristics of
the soil. Jenny (1930) indicated the importance of soil-forming factors
in influencing the nitrogen content of medium textured soils in the United
States in the following order of decreasing importance: (1) climate, (2)
vegetation, (3) topography and parent material, and (4) age. Many
soils were originally relatively well supplied with nitrogen, and as a result
they produced satisfactory yields during the early years of cultivation.
A new set of conditions are established, however, when soils are brought
into cultivation, resulting almost invariably in a much lower organic
matter and nitrogen level. The decline is a gradual process, being most
rapid during the first few years of cultivation. These changes are illus-
trated by the data presented in Fig. 2. These data from Myers et al.
(1943) indicate that the equilibrium of soil nitrogen under cultivated
conditions occurs at much lower levels but in the same order as in the
virgin state.

The great importance of climatic factors in determining the soil
organic matter and nitrogen level has been shown by Jenny (1930) who
found for grassland soils an inverse relationship between the latter and
the mean annual temperature. His data show that in the U. S. soil
nitrogen decreases from North to South and that for each fall of 10C.
in mean annual temperature the average nitrogen content increases two-
or three-fold. Jenny further found that grassland soils along an annual
isotherm increased in nitrogen content with increasing rainfall and humid-
ity. These data emphasize the difference between the operating levels

L. E. ENSMINGER AND R. W. PEABSON

0220

15 20 25

YEARS OF CULTIVATION

Fig. 2. Effect of cultivation on soil nitrogen content at Hays, Colby, and Garden
City, Kansas. (Myers et al, 1943.)

of nitrogen that are practical of attainment in soils of the Northern and
those of the Southern part of the United States.

A direct relationship between precipitation and the nitrogen content
of Washington soils was reported by Sievers and Holtz (1923). Soils
receiving about 8 inches of rainfall annually were found to contain only
about 25 per cent as much nitrogen in the surface 6 inches as soils re-
ceiving 20 inches per year. This general relationship was confirmed by
Fowler and Wheeting (1941) , whose data further showed that the increase
in organic matter and nitrogen is accompanied by a widening C:N ratio.

Russell and McRuer (1927) examined a series of homogenous types
and found the nitrogen to vary with topography as well as with rainfall.
The nitrogen content increased with increasing rainfall and, under the
same precipitation, level types contained more nitrogen than rolling types.
Russell (1927) reported work done on Minnesota, Kansas, and Nebraska

SOIL NITROGEN 85

soils showing a close correlation between nitrogen and texture, the finer
textured soils being highest in nitrogen. It is generally recognized that
it is more difficult to build up or maintain the nitrogen content of coarse
textured soils than it is to build up or maintain nitrogen in fine textured
soils.

2. Tropical Soils

On the basis of the data presented by Jenny (1930) for the United
States, the soils of the tropics should be very low in organic matter and
nitrogen. Jenny, Bingham, and Padilla-Saravia (1948) showed that the
nitrogen and organic matter contents of the equatorial soils of Colombia,
South America, are related to climate. The nitrogen-climate surface of
the Colombian soils was found to be very similar in shape to the nitrogen-
climate surface for soils of the Great Plains area. A comparison, how-
ever, of the Colombian and North American soils with equal annual
temperature and moisture values shows the Colombian soils to be much
higher in nitrogen and organic matter than the North American soils.
They point out that many of the light colored soils from the hot and
humid areas are much higher in nitrogen and organic matter than their
color indicates.

Dean (1930) found that in an unselected group of Hawaiian soils
the mean nitrogen content of the 223 samples examined was 0.31 per cent.
The author concluded that the nitrogen content of Hawaiian soils did
not agree with the data presented by Jenny for the United States and
that to agree the average nitrogen content should be less than 0.1 per cent.
In discussing the soils of Cuba, Bennett and Allison (1928) state that the
supply of organic matter is higher in well-drained tropical soils than is
generally supposed by numerous writers who have assumed that with
good porosity and aeration organic matter in tropical soils is soon dis-
sipated. Jenny, Gessel, and Bingham (1950) found that Colombian and
Costa Rican soils were rich in nitrogen and organic matter as compared
with Californian soils of the Sierra Nevada mountains.

It appears evident that some tropical soils are higher in nitrogen and
organic matter than expected on the basis of available data for the soils
of North America. On the other hand, a few investigators have empha-
sized the low nitrogen and organic matter content of certain tropical soils.
Mohr (1944) states that in the low, hilly lands of the tropics conditions
are optimum for mineralization of plant residues, and, as a result, little
organic matter can remain either in the original or transformed state.
He points out, however, that at elevations above 1000 meters, with good
moisture and aeration, the humus content increases with elevation. He
suggested the predominance of molds over bacteria in the soil at the

86 L. E. ENSMINGEB AND B. W. PEARSON

higher elevations as the reason for greater accumulations of organic mat-
ter. Corbet (1935) also emphasized the low nitrogen content of tropical
soils in the Dutch East Indies.

There is little information in the literature as to why some tropical
soils are abnormally high in nitrogen and humus. It has been suggested
by some writers that the tropics are high in humus because of the large
amount of vegetative growth produced. Data by Jenny et al. (1950)
show that tropical forests dropped 8,000 to 11,000 Ibs. of leaves and twigs
per acre per year, whereas forests of the Nevada Sierra mountains of
California produced only 800 to 3,000 Ibs. It has also been suggested
that organic matter accumulates in tropical soils because of an inactive
microbial population. Jenny et al. (1950) have calculated the rate of
decomposition of forest floors in Colombia to be higher than in Cali-
fornia. They also found that alfalfa leaves placed in natural soils de-
composed faster in tropical soils than in temperate soils. These findings
would lead one to expect low contents of nitrogen and organic matter in
tropical soils even though litter fall is large. Jenny (1950) has attempted
to bring these two conflicting observations into harmony. His data show
that the decomposition of the forest floor, which rests on the mineral soils,
proceeds rapidly and that observations indicate a considerable portion
of the decomposition products infiltrate into the mineral soil. The rate of
decomposition of the infiltrated products appears to be slow and as a
result humus accumulates rapidly to a high level. Jenny suggested favor-
able climatic conditions and high annual rate of nitrogen fixation, largely
by leguminous trees, as primary causes of luxuriant growth.

It is difficult to understand why the infiltrated products decompose
so slowly. Kelley (1915) has emphasized the inert nature of the unculti-
vated soils of Hawaii and has suggested the lack of aeration as one of
the causes. An examination of British Guiana sugar-cane soils by Hardy
and Hewitt (1948) revealed a low nitrifying capacity in these soils. The
authors attributed the low nitrifying capacity to absence of proper nitri-
fying organisms. They were not certain that the absence of nitrosomonas
and nitrobacter was the only reason for the low nitrifying capacity since
they did not determine the effect of inoculation. They also suggested
a shortage of available phosphate as a partial cause of the low nitrifying
capacity. It is apparent that organic matter of tropical soils, especially
the portion mixed with the mineral soils, decomposes more slowly than
one would expect on the basis of climatic conditions. It is evident, there-
fore, that for some reason soil conditions are unfavorable for active
microbial decomposition. It is generally known that many tropical soils
are deficient in certain mineral elements such as phosphorus. The avail-

SOIL NITROGEN 87

able phosphorus content of these soils may be too low to support an
active microbial population.

III. NATURE OF ORGANIC NITROGEN IN THE SOIL

Although organic matter is a very important part of the soil, little
is known about the chemistry of this material. Such information is
needed in order to evaluate properly the role of organic matter in soil
fertility. The C:N ratio is often used as a means of characterizing
organic matter, but we do not know why soils of one area may have a
different ratio than soils of another area. Many of the methods of ap-
proach have been empirical in nature and as a result have added little
to our basic knowledge of soil organic matter.

As plant residues decompose a considerable quantity of microbial cell
substance is formed. According to Norman (1942), one-third to one-half
of the organic fraction of the soil may be microbially derived if decom-
position takes place under aerobic conditions. During the process of
decomposition, the more resistant constituents of plant residues accumu-
late along with microbial substances. It is now generally believed that
lignin or lignin-derived material makes up a substantial part of the or-
ganic fraction of soils. Gottlieb and Hendricks (1945) have presented
evidence to show that plant lignin in the soil is altered considerably by
decomposition. Their studies on the hydrogenation of "alkali lignin"
indicate that lignin in the soil undergoes a similar type of change as does
lignin treated with alkali, but at a much slower rate.

Soil organic matter also contains nitrogenous complexes that are im-
portant from the standpoint of fertility. The biological resistance of
organic soil nitrogen is indicated by the fact that less than 4 per cent
becomes available in any one season. Soil nitrogen is usually considered
to be proteinaceous in character, but for some reason it is much less
available than the nitrogen in ordinary proteins. Waksman and Iyer
(1933) have presented data to support their theory that the availability
of the soil protein is reduced through combination with lignin. If protein
nitrogen is stabilized by association with lignin this phenomenon may
furnish a partial explanation for the lack of build-up of soil nitrogen
following continued use of green manure crops.

Another explanation of the unavailability of organic nitrogen has
been put forward by Ensminger and Gieseking (1942). They demon-
strated that mixtures of proteins with certain clays are more resistant
to hydrolysis by proteolytic enzymes than proteins alone. Previous work
by them (1939) had shown that proteins are adsorbed within the "001"
lattice spacing of montmorillonitic clays. The adsorption of proteins
was found to increase with increasing pH, which indicates that the pro-

88 L. E. ENSMINGEB AND K. W. PEARSON

teins react as bases. Unpublished data by Ensminger show that the
addition of montmorillonitic clays to proteins and to Florida peat was
effective in rendering these materials less susceptible to microbial decom-
position under laboratory conditions as measured by the rate of CO 2
evolution or by the quantity of nitrates formed. Recent work by Allison
et al. (1949) show that the nature and amount of colloid present in
sand-colloid mixtures in which organic materials are decomposing is an
important factor in determining the quantity of residual carbon after a
year of decomposition. The effect of the colloid was marked where
readily decomposable materials were added, but there was no effect where
peat, sawdust and cellulose were used. MontmorillonHe exerted the
greatest effect in holding carbon and kaolin exerted the ieast. Carring-
ton colloid, which is mainly a mixture of montmorillor.iite and hydrous
mica, gave an intermediate effect. In some cases the addition of 10
per cent bentonite to sand nearly doubled the quantity of residual carbon.
It is interesting to note in this regard that work by Goring and Bartholo-
mew (1949) shows bentonite to be effective in decreasing the rate of
mineralization of organic phosphorus. They also found the effect of
kaolinite to be less marked. As previously stated it is a common obser-
vation that it is easier to build up or maintain organic matter in a heavy
soil than it is in a sandy soil. On the basis of present knowledge it is
not possible to determine if this is due wholly or in part to the protective
action of the clay in heavy soils. Poor aeration in heavy textured soils
is often given as the reason why they hold more carbon than sandy soils.
Allison et al. (1950) believe this explanation to be inadequate and point
out that, where water logging is not a problem, the oxygen supply in
both sand and clay soils is sufficient for rapid and complete aerobic de-
composition of plant materials. Even so, evidence presented by Myers
(1937), Tyulin (1938), Springer (1940) and others further emphasizes
the probable interaction between inorganic and organic colloids in the
soil. Springer (1940) believes that the chernozems contain rather stable
inorganic-organic complexes. A new field of research is opening up as a
result of these studies and basic information about these complexes may
add much to our knowledge of the nitrogen economy of soils.

It appears that soil organic matter is not only a mixture of organic
constituents ranging from unchanged plant residues to that component
designated as humus, but is also so closely united with the inorganic
fraction as to be changed in properties as a result of the union. Although
environmental conditions affect the nature of soil organic matter, the
organic matter of different soils is similar in certain respects. The ques-
tion is often asked as ^o what influence cropping systems have on the
nature of soil organic matter. This question can not be fully answered

SOIL NITROGEN 89

until better methods have been developed for characterizing the organic
fraction. It is interesting to note in this connection that Peevy and
Norman (1948) conclude from their studies that alkaline hypoiodite may
be useful in detecting differences in the nature of soil organic matter.

IV. NITROGEN TRANSFORMATIONS

During the course of a year there is a considerable intake and outgo
of nitrogen in most soils including a number of complex reactions. These
reactions, which are collectively termed the "nitrogen cycle," are largely
biological, and closely parallel a similar series of reactions of the organic
carbon. Although the nitrogen cycle has been studied intensively for
many years and is understood in general, our knowledge of it is still
seriously lacking in certain respects. Nitrogen occurs in soils principally
in organic complexes of microbial origin. The original source of this
combined nitrogen was the elemental nitrogen in the earth's atmosphere
which has been tapped through fixation, by lightning discharge, and by
certain groups of microorganisms. The micro and macroflora debris
which is returned to the soil is the material from which the reserve soil
nitrogen supply is derived.

1. Nitrification

As organic residues decompose, there is a narrowing of the C:N ratio
until an equilibrium ratio of about 12:1 is reached. During the process
of decomposition, the nitrogen in the residues becomes less and less avail-
able. Broadbent and Norman (1946) working with the stable isotope
of nitrogen, N 15 , found that the addition of energy supplying material
increased the rate of release of nitrogen from the soil organic matter.
They state that the unavailability of organic soil nitrogen may be due
to the lack of enough energy material to support an active microbial
population rather than to the formation of resistant complexes. If the
availability of organic soil nitrogen is increased by the addition of such
energy materials as plant residues, it is obvious that the carbon in soil
organic matter should also be more readily utilized by the microorgan-
isms.

Nitrogen availability is known to be strongly influenced by the decom-
position of plant residues in the soil. Parberry and Swaby (1942)
studied the release of nitrogen from different organic materials added to
soils and found that sufficient nitrogen for crop needs was liberated in
one season only from materials containing an initial nitrogen content of
greater than 2.5 per cent. No nitrogen was liberated in this period from
materials having less than 1.5 per cent nitrogen. Waksman and Tenney
(1927) found that 1.7 per cent nitrogen was adequate for microbial needs.

90 L. E. ENSMINGEB AND B. W. PEARSON

Bledsoe (1937) studied a number of weeds common to Florida in relation
to the rate of nitrification, and concluded that the water-soluble nitrogen
content appeared to be the most important factor involved in the nitri-
fication of green and dried plants, followed by total nitrogen and degree
of hydration or moisture. His data show that if the water-soluble nitro-
gen content is 0.5 per cent or above, favorable nitrate accumulation
occurs, even though the total nitrogen content is less than 1.7 per cent.
Whiting (1926) also found that the water-soluble nitrogen content of
materials determined to a large extent the rate of nitrification and that
there was usually a direct relationship between water-soluble nitrogen
and total nitrogen. On the basis of these investigations it is improbable
that many materials with a nitrogen content of less than 1.5 per cent
would give a positive release of nitrogen in one season. This means that
only legumes and certain nonlegumes in the early stages of growth will
give a positive release of nitrogen. On the other hand a quick release
may be expected from materials with a nitrogen content of more than
about 2.5 per cent.

When soil organic matter decomposes, ammonia is liberated and
under favorable soil conditions it is converted to nitrate. Nitrification
has long been considered to be a biological process, but in recent years
Dhar (1933) and others working in India have contended that the process
is photochemical, especially in the tropics. Waksman and Madhok
(1937) investigated the effect of light and heat on nitrate formation in
soils and concluded that biological oxidation must still be considered as
the important process in nitrate formation. Fraps and Sterges (1935)
repeated some of the experiments of Dhar and concluded that photo-
nitrification was of little or no importance in normal soils.

The total soil organic nitrogen has been regarded as an important
factor in determining the amount of available nitrogen that a soil will
supply. Work by Waksman (1923), Burgess (1918), Brown (1916),
Gowda (1924), Gainey (1936), Fraps (1920, 1921), and Fraps and Sterges
(1947) has shown that usually the more fertile soils produce the greatest
amounts of nitrate, but they also found many exceptions. Gainey (1936)
and Allison and Sterling (1949) found that nitrate formation was directly
related to total soil nitrogen. There are a number of other factors, how-
ever, that affect the production of nitrates in field soils. One factor is
soil treatment, such as the addition of limestone, phosphorus, and potas-
sium. A second factor is tillage operations, such as plowing, cultivation,
fallowing, and mulching. Third, such climatic conditions as temperature
and moisture are very important.

Sewell and Gainey (1932) and Call (1914) have shown that in Kansas
July plowing for wheat, Triticum vulgare } has produced approximately

SOIL NITROGEN 91

10 bushels per acre more than September plowing. They found that the
acre yield of wheat was proportional to the nitrates present at seeding
time. Beneficial effects of fallowing in arid regions appears to be due
to an increase in nitrates as well as to an increase in moisture. Buckman
(1910) working with Montana soils having an annual rainfall of 12 to
17 inches concluded that an increase in moisture meant an increase in
nitrates if other factors were favorable. Harper (1945) stated that
available nitrogen in the Southern Great Plains area may be low at the
time of planting fall crops because of dry weather and unfavorable
conditions for tillage during summer months. Smith and Vandecaveye
(1946), in a study of the productivity of Palouse silt loam, reported
that nitrogen was the principal factor affecting crop yields under a
continuous wheat system, but nitrogen was not a factor under a wheat-
fallow system. Moisture was also an important factor, since continuous
wheat plus nitrogen fertilizers did not yield as high as wheat-fallow
plus nitrogen.

It is a generally accepted fact that the greatest accumulation of
nitrates takes place during the summer months and the least during
winter months. Usually nitrates increase from winter to summer and
decrease from summer to winter. These seasonal variations in nitrates
are closely associated with such climatic factors as moisture and tem-
perature. Part of the increase from winter to summer may be due to the
rather intensive period of cultivation in spring and summer, Lyon and
Bizzell (1913) presented evidence to show that freezing and thawing
increases the subsequent formation of nitrates.

It is apparent that the rate of nitrification in soils containing a
reserve of organic nitrogen can be increased by various treatments
and cultural practices. At the same time it is evident that nitrification
can occur only at the expense of the total soil nitrogen supply and,
therefore, cannot be maintained at a high level unless the reserve supply
of nitrogen is maintained or increased.

2. Symbiotic Nitrogen Fixation

Nitrogen fixation has been and will continue to be one of the most
important factors in the nitrogen balance of soils. The importance of
nitrogen fixation is indicated in the report on the balance sheet of plant
nutrients in the United States by Lipman and Conybeare (1936). For
the year of 1930, they estimate that of the 16,253,862 short tons of
nitrogen added to the agricultural soils of the United States, 9,830,736
tons were added as a result of biological nitrogen fixation. The quan-
tity of nitrogen fixed can be increased considerably by making conditions

92 L. E. ENSMINGER AND R. W. PEARSON

more favorable for biological fixation and by increasing the acreage of
legumes.

Rhizobia in the nodules of living leguminous plants fix nitrogen.
How much nitrogen will be fixed by legume bacteria depends on a number
of factors. Such soil conditions as aeration, available nitrogen, moisture,
and the amount of active calcium are very important. Also, some
strains of the various species of Rhizobium are not effective in fixing
nitrogen. This indicates the importance of inoculating with commercial
cultures of known effectiveness. Certain legumes fix much more nitro-
gen than others. The nitrogen content of most inoculated legumes does
not vary too widely, so it follows that the quantity of nitrogen fixed is
more or less proportional to the total growth of the legume. The amount
of vegetative growth obtained will vary with the legume used and of
course the growth of any particular legume will depend upon the fer-
tility of the soil.

Bracken and Larson (1947) calculated the nitrogen increase from
alfalfa on 20 farms in Cache Valley, Utah. The 20 farms showed an
average fixation of 246 Ibs. per acre per year. These workers based their
calculations on the nitrogen in the hay plus that accumulated in the
soil. This means that some of the 246 Ibs. of nitrogen resulted from
nonsymbiotic fixation. In a 10-year experiment at Ithaca, New York,
Lyon and Bizzell (1934) found that continuous alfalfa gave an apparent
fixation of 268 Ibs. of nitrogen per acre per year. With barley, rye, or
oats, annual nonsymbiotic fixation amounted to 17 Ibs. Assuming that
erosion, leaching, and volatilization losses were the same for the two
cropping systems this would leave 251 Ibs. acquired symbiotically.
Duggar (1899) found that the air dry weight of vines, roots, and stubble
of hairy vetch, Vicia villosa, Roth, cut April 19 was 3,967 Ibs. per acre
and contained 137 Ibs. of nitrogen. Vetch cut May 9 weighed 6,870
Ibs. and contained 202.8 Ibs. of nitrogen. Although these figures do not
show the amount of nitroge"h fixed by vetch since part of the nitrogen
was taken from the soil, they do indicate that fixation increases as the
plants get older. Lipman and Conybeare (1936) estimated that legumes
fixed an average of 87.6 Ibs. of nitrogen per acre per year. Gustafson
(1948) states that 80 Ibs. is a conservative estimate of the average an-
nual fixation of nitrogen by legumes. These are average figures for all
legumes under a variety of conditions and are lower than may be expected
for certain legumes under favorable conditions. It is evident then that
an increase in legume acreage will go a long way in providing the extra
nitrogen needed to maintain high yields. Also, an increase in the acreage
of legumes would help to prevent excessive losses of soil nitrogen by

SOIL NITROGEN 93

erosion and leaching as well as increase the addition of nitrogen to the
soil.

3. Nonsymbiotic Nitrogen Fixation

Certain organisms living in the soil obtain nitrogen from the air and
use soil organic matter as a source of energy. Since these organisms are
not directly associated with higher plants, the transformation of ele-
mental nitrogen to a fixed form is known as nonsymbiotic fixation.
The amount of nitrogen fixed by this process depends on such factors as
the supply of readily available energy material, supply of available
nitrogen, and pH. Soils high in available nitrogen probably fix little or
no nitrogen in this manner. The Azotobacter group of bacteria, which
is largely responsible for free fixation, is very sensitive to pH. Work by
Gainey (1948) shows that pH 6.0 or above is favorable. An available
supply of mineral nutrients, especially calcium and phosphorus, favors
vigorous fixation.

The data on the quantity of nitrogen fixed nonsymbiotically are
rather meager. Lyon and Wilson (1928) report that land kept in grass
for 10 years without nitrogen additions gained 415 Ibs. of nitrogen.
Probably about 6 Ibs. was added by rainwater, but the loss by drainage
and volatilization was probably greater than the addition in rain. It
would appear that the annual gain of 41.5 Ibs. of nitrogen would be a
conservative figure ascribed to fixation under the conditions of this
experiment. Lyon and Buckman (1947) state that at least 25 Ibs. of
nitrogen is fixed nonsymbiotically in a representative arable soil. It
should be pointed out, however, that conditions in many soils are not
favorable for this type of fixation. Lipman and Conybeare (1936)
reported 6 Ibs. per acre per year as the average quantity of nitrogen
fixed nonsymbiotically. Although the amount may not be large, it is
an important source of soil nitrogen and in any sound soil management
program consideration should be given to making conditions as near
optimum as possible for this type of fixation.

A comprehensive review of the literature on inoculation of crops by
Allison (1947) reveals that Russian workers have reported rather large
yield increases from the inoculation of soils with Azotobacter species.
Allison et al. (1947) studied the effect under greenhouse conditions of
inoculating two soils with Azotobacter and "Azotogen." They found no
significant effect of inoculation on the growth or nitrogen content of such
crops as barley, Hordenum sp., Sudangrass, Sorghum vulgare Sudanese,
kale, Brassica oleracea acephala, rape, Brassica sp., rye, Secale cereale,
and Swiss chard, Beta cicla. Work by Gainey (1949) showed no evi-

94 L. E. ENSMINGEB AND B. W. PEABSON

dence that the maintenance of Azotobacter in a soil for 20 years had any
effect on crop growth or the nitrogen balance of the soil.

V. EFFECT OF CROPPING PRACTICES ON NITROGEN LEVEL

Under natural conditions there exists an equilibrium between the ad-
dition of organic matter by vegetation and its decomposition by micro-
organisms. As shown previously, the equilibrium level of nitrogen is
determined largely by climatic conditions. Cultivation of soils usually
results in a decrease in nitrogen content from that in the virgin state
by speeding up microbial decomposition and by subjecting the land to
greater losses of nitrogen by erosion and leaching.

Numerous studies have been made of the effect of cropping on both
rate of decline and final nitrogen content of soils in the wheat-growing
regions. Shutt (1910) found that a soil of Indian Head, Saskatchewan,
had lost almost one-third of its nitrogen after nearly a quarter of a cen-
tury of cultivation. Crop removal accounted for 700 Ibs. of the loss,
leaving unaccounted 1,486 Ibs. For Minnesota conditions Snyder (1905)
calculated that, out of 7,700 Ibs. at the beginning of the experiment,
continuous wheat plots had lost 2,039 Ibs. of nitrogen per acre foot in
12 years. The crops removed 450 Ibs., indicating a loss of about 1,600
Ibs. by other means. A rotation consisting of wheat, clover, wheat, oats,
and corn with a dressing of farm manure every 5 years showed a loss
of nitrogen of about 18 per cent. Harper (1945) reported that 11 Okla-
homa Panhandle soils had lost 14.8 per cent of their nitrogen after 15
years of cropping. Brown et al. (1942) reported that in Alberta the
largest loss of nitrogen occurred with soils originally high in nitrogen.
The nitrogen content of the surface 6 inches was 17 to 22 per cent lower
after 30 years of cultivation. It appears that these soils lose approxi-
mately one per cent of their nitrogen for every year of cultivation.
Dodge and Jones (1948) studied the effect of long-time fertility treat-
ments on nitrogen and carbon content of plots at Manhattan, Kansas,
and concluded that the fertilizer treatments and cropping systems had
only a slight influence on the trend of nitrogen or carbon, but may have
an influence on the speed at which equilibrium is reached as well as
the ultimate level. This conclusion is supported by the data of Myers
et al. (1943) for western Kansas, who show that soils cropped to con-
tinuous small grain and alternate small grain and fallow tend to reach
an equilibrium state quicker and at a higher nitrogen level than do sys-
tems that include row crops. The soils at Hays, Colby and Garden
City were approaching a state of equilibrium with respect to nitrogen
after 30 to 35 years of cultivation (Fig. 2). The approaching equilibrium
level was highest at Hays and was lowest at Garden City.

SOIL NITROGEN 95

Bracken and Greaves (1941) surveyed the nitrogen losses on farms
in two areas of Utah. A study of 9 dry farms in Cache Valley, northern
Utah, showed the first foot of virgin land to be 15.9 per cent higher in
nitrogen than adjacent wheat land. Twelve farms in Juab Valley, cen-
tral Utah, were found to be 14.5 per cent lower in nitrogen than virgin
soils. They considered the equilibrium levels of Cache Valley and Juab
Valley soils to be approximately 0.17 and 0.09 per cent nitrogen, re-
spectively. Severely eroded areas had lost 58.5 per cent of their nitrogen.

Salter and Green (1933) attempted to estimate to what extent various
crops have increased or decreased the organic carbon and nitrogen con-
tent of soils. They give the following yearly changes in percentage of
the total nitrogen present in the soil: corn, 2.97; wheat, 1.56; oats,

1.45; legume-grass hay in a 5-year rotation (predominantly timothy),
+2.87. It is evident from these data that rotations containing hay
crops, especially legumes, tend to conserve soil nitrogen. White et al.
(1945) reported that, at the end of 72 years, unfertilized grassland soils
had a nitrogen content 68.2 per cent above an unfertilized plot in a
4-year grain rotation, and 40.0 per cent higher than a PK treated plot
in the same rotation. These data show that phosphorus and potassium
are important in conserving soil nitrogen.

Jenny (1933) investigated the effect of cropping on decline of soil
nitrogen in the Middle AVcst. Reduction in soil nitrogen was greater
in the earlier years of cultivation, as shown by the following: first 20
years, 25 per cent; second 20 years, 10 per cent; and third 20 years,

7 per cent. Data by Myers et al. (1943) show the nitrogen trend of
dry land soils to be of the same pattern as is characteristic for the more
humid soils of the Middle West. The fact that soils under clean cultiva-
tion tend to approach a new nitrogen equilibrium at which point a finite
value is maintained points toward nonsymbiotic nitrogen fixation as a
contributing factor.

Some of the plots of the Rothamsted Experiment Station in England
have been in continuous wheat for more than a century except for fallow
every fifth year since 1925. Crowther (1947) gives a summary of the
nitrogen changes for the period 1865 to 1945. He concluded that the
unmanured plots had reached a substantial equilibrium within the first
20 years. The plots receiving farmyard manure increased annually in
nitrogen content for the first 40 years and then leveled off. In 1945
plots that had received farmyard manure annually since 1843 were over
twice as high in nitrogen as the untreated plots.

In the Southeast much of the land has been in cultivation for a. long
period, and humidity and temperature have been favorable for decompo-
sition. As a result of these conditions, many of the soils have attained

96 L. E. ENSMINGER AND R. W. PEARSON

a cultivated state of equilibrium with respect to organic matter and
nitrogen. Cropping systems that include winter or summer legumes,
or both, will usually provide a satisfactory turnover of nitrogen and the
total may actually be increased by such systems. Tidmore and Volk
(1945) found an increase in soil nitrogen amounting to 30 per cent over
a 9-year period by turning under soybeans, Glycine soja, every other year.
Work by Moser (1942) showed the average organic matter content of
the Piedmont section of South Carolina to be about one per cent. It
was found, however, that the organic matter level could be raised to
1.5 per cent by using lespedeza, vetch and crimson clover, Trifoliwn
incarnatum, in the cropping system. Lespedeza grown continuously
raised the organic matter content to 2.7 per cent, indicating that amounts
as high as 2.7 per cent would be difficult to maintain under practical
farming operations. The results of Holley et al. (1948) in Georgia show
a slight upward trend in soil nitrogen for rotations with and without
legumes. Apparently the commercial nitrogen applied to the crops in the
no-legume rotation was sufficient to maintain as much soil nitrogen as
the use of legumes. Jones (1942) concluded from a study of various
systems of green manure crop management using lysimeters that the
nitrogen content of the soils was maintained at a constant level for 4
years when 225 Ibs. of sodium nitrate was used. There was a net gain
in nitrogen when summer legumes were turned under, but the net gain
was highest when vetch was grown as a source of nitrogen.

The decline of soil nitrogen under cultivation is due to several causes.
One of the most easily measured losses of soil nitrogen is that by crop
removal. The magnitude of this loss will vary with the crop as well
as with the yield. Russell (1927) believes some gaseous nitrogen product
is formed as a result of cultivation. He referred to one of the Broad-
balk wheat plots that received 14 tons of farmyard manure annually,
containing 200 Ibs. of nitrogen. This plot lost nitrogen amounting to 70
per cent of the quantity added. The no-manure plot alongside, in spite
of leaching, showed no apparent loss of nitrogen. The work of Shutt in
Saskatchewan (1910), Snyder in Minnesota (1905), and Swanson and
Latshaw in Kansas (1919) showed that much of the nitrogen lost by
cultivation could not be accounted for by crop removal. In attempting
to account for nitrogen losses from dry-land soils in Utah by means
other than crop removal, Bracken and Greaves (1941) believed that
slight losses occurred through leaching and erosion. They thought that
the major loss was due perhaps to chemical and biological changes re-
sulting in volatilization. Lysimeter investigations by Collison et al.
(1933) showed an unaccounted loss of nitrogen amounting to as much as
118 Ibs. per acre per year. The authors conclude that such losses must

SOIL NITROGEN 97

be due to volatilization. In the greenhouse at high nitrogen levels, Pinck
et al. (1945) observed a nitrogen loss amounting to 14 per cent of that
added. They suggested that much of the loss may have been due to
metabolic processes occurring within the plant.

All studies do not indicate such a high loss of unaccounted for nitro-
gen. Lysimeter studies by Smith (1944) in Arizona showed that a
Mohave clay, which had been double cropped to wheat and hegari,
Sorghum vulgare, for 12 years increased in nitrogen from 0.052 per cent
to 0.067 per cent. A Gila clay under the same conditions showed a
decrease in nitrogen from 0.085 per cent to 0.063 per cent. Except for
one year all crops were removed from the tanks. All plots had a positive
nitrogen balance when crop removal was considered.

The unaccounted-for-loss of nitrogen under field conditions just men-
tioned may have been high, since losses by leaching and erosion were
not known but assumed to be unimportant. These losses, especially by
erosion, may have been greater than assumed. Even though only a small
amount of runoff occurred, it is possible that considerable nitrogen was
lost. Martin (1941), working with Collington sandy loam in New Jersey,
showed that the eroded material contained from 3 to 8 times as much
organic matter and nitrogen as the soil itself. Rogers (1941) also re-
ported a higher nitrogen content of eroded material than of original
soil. The greater concentration of organic matter and nitrogen in the
eroded material may be due in part to the floating off of pieces of organic
matter not thoroughly incorporated with the inorganic fraction of the
soil. Eroded material is usually higher in clay than the parent soil, and
Kardos and Bowlsby (1941) showed the percentage of organic matter
to be higher and the C:N ratio to be lower in the clay fraction than in
the whole soil.

Some of the earliest work dealing with losses of plant nutrients under
various cropping systems as a result of sheet erosion was reported by
Miller and Krusekopf (1932). Their data show that such cropping sys-
tems as continuous corn, continuous wheat, and corn-wheat-clover ro-
tations lost respectively 66, 32, and 26 Ibs. of nitrogen per acre annually.
A plot in the same experiment plowed 4 inches deep and kept fallow
lost 118 Ibs. as compared with a loss of 0.6 Ib. of nitrogen from a con-
tinuous bluegrass, Poa pratensis, L., sod. According to Uhland (1947),
a Shelby loam at Bethany, Missouri, which had been cropped to corn
continuously for 10 years, lost 50.9 tons of soil per acre per year. Land
in a 3-year rotation of corn, wheat, and hay lost 7.51 tons of soil. There
was only a trace of soil lost from land in continuous alfalfa or bluegrass.
At Clarinda, Iowa, continuous corn plots lost 5.32 times as much soil
as a 3-year rotation of corn-oats-clover. These data show that soil losses

98 L. E. ENSMINGEB AND E. W. PEARSON

may be decreased considerably by following the proper cropping system.
Rotations including clovers result in less erosion than continuous corn,
but the effect cannot be explained entirely by the greater cover provided
by the rotation. For example, data presented by Miller and Krusekopf
(1932) show that during the 6-month growing period continuous corn
plots lost more soil from erosion than the rotation plot in corn. They
suggested improved soil granulation as a reason.

These and other data point to the large losses of surface soil accom-
panying certain cropping practices. Usually loss of surface soil means
proportionate losses of nitrogen and organic matter. In this connection
Slater and Carleton (1938) examined a Shelby silt loam and a Marshall
silt loam and observed a linear function between erosion losses and
organic matter depletion. The organic matter content of the soils dropped
0.002 per cent for each ton of soil eroded. They estimated that on a
fallowed plot erosion had increased the depletion of organic matter to 18
times that normally lost by oxidation.

There is little doubt as to the influence of erosion on the rate of
organic matter and nitrogen depletion. If erosion is kept to a mini-
mum, the job of maintaining a satisfactory level of soil nitrogen will
be a much easier one. The detrimental effect of erosion, ns measured
by crop yields, will obviously appear much quicker on soils with a shallow
surface horizon.

VI. NITROGEN ECONOMY OF ERODED SOILS

Decline in soil nitrogen, whether it be due to increased oxidation as
a result of cultivation, crop removal, erosion, or a combination of
these, is usually accompanied by a decrease in crop yields, especially
of nonlegumes (Fig. 3). Results from Zanesville, Ohio, reported by
Borst et al. (1945) show a close correlation between erosion, organic
matter, and corn yields for continuous corn. Uhland (1947) reported
results of a study of corn yields in relation to depth of surface soil for 18
fields near Fowler, Indiana. Yields were in direct proportion to depth
of topsoil and varied from 19.8 bushels where no topsoil remained to 69.5
bushels where the topsoil measured 12 inches.

Although all of the decrease in crop yields resulting from erosion is
probably not caused by loss of nitrogen, there are data at hand that
show nitrogen to be a major factor. Rost (1939) studied the relative
yields of oats in pot tests using successive soil layers of 6 profiles and
observed a decrease in yields from the surface downward. The addition
of a nitrogen fertilizer largely or completely removed any differences in
the productivity of various layers. The relative yields of red clover
decreased for the second 6-inch, and second and third foot sections, but

SOIL NITROGEN

rose for the fifth foot. The addition of phosphate or phosphate and
potash completely removed any differences in productivity except for
one layer of one profile. Hays et al. (1948) found that the loss of 3 or
4 inches of the surface from a Fayette silt loam did not permanently

tr
o

a
o

* 995

LOUISIANA

o' i ' L

ARKANSAS

MISSOURI
I |

IOWA

005

010 015 020

AVERAGE TOTAL NITROGEN CONTENT OF SOIL

Fig. 3. The average corn yield per acre in relation to the average total nitrogen
content of the soil. (Jenny, 1930.)

impair its productivity if managed properly. This means that practices
are needed that will build up the organic and nitrogen content, such
as long rotations including 3 years or more of alfalfa-grass hay and ap-
plications of barnyard manure. Fayette silt loam is a young soil and
the main difference between surface and subsoil is in the organic matter

100 L. E. ENSMINGER AND B. W. PEARSON

and nitrogen contents. Results of an experiment carried out at Bethany,
Missouri, from 1932-1942 and reported by Smith et al. (1945) show the
effects of cropping systems and treatments on the productivity of an
exposed Shelby loam subsoil. Corn was planted on all plots in 1942 to
determine the producing capacity of the subsoil as a result of past treat-
ments. The surface soil without treatment produced 43 bushels of corn.
A rotation of corn, oats, and sweet clover, Melilotus sp., with sweet
clover turned under the second year plus an original application of lime
and a 4-12-4 fertilizer to oats yielded 44 bushels. The same rotation and
lime treatment with superphosphate applied to oats plus 8 tons of barn-
yard manure plowed under before all previous corn crops yielded 64.6
bushels. Continuous grass-legume meadow for 11 years with an original
treatment of lime and 4-12-4 fertilizer yielded 44.2 bushels. A rotation
of corn, wheat, and meadow 2 years on limed subsoil with 4-12-4 fer-
tilizer applied to oats produced 34.6 bushels of corn. It appears from
these data that in addition to lime and 4-12-4 fertilizer, organic matter
and nitrogen are necessary to bring the producing capacity of the sub-
soil up to that of the untreated surface soil.

VII. COMMERCIAL NITROGEN vs. BARNYARD MANURE AND
GREEN MANURE

The question often arises as to the relative merits of commercial
nitrogen and such sources as barnyard manure and green manures for
crop production. In fact a group of popular writers have gone so far
as to state that manured crops are superior to commercially fertilized
crops for human consumption. Although the value of manure and crop
residues of various kinds as soil amendments for crop production is gen-
erally recognized, it is difficult if not impossible in many instances to
obtain enough of these materials for maximum yields. Under ordinary
farming conditions, it is often necessary to supply a part of the nitrogen
needs of nonlegumes by addition of commercial nitrogen, if maximum
yields are to be obtained.

At the Rothamsted Experiment Station wheat has been grown con-
tinuously since 1844 and yield records have been kept since 1852. Yield
data for a 95-year period are reported by Bear (1949). These data show
that plots receiving as much as 1,392 Ibs. of fertilizer outyielded a plot
receiving 15.7 tons of farmyard manure annually. Data by Thorne
(1930) of the Ohio Experiment Station reveal that chemicals, applied in
amounts equivalent to the N, P2C>5, K 2 in manure, were just as effective
as manure in increasing crop yields. Thorne wrote the following on the
subject of manures: "When manure has been compared with chemical
fertilizers the manure usually has been used in such amounts as to carry

SOIL NITROGEN 101

far larger quantities of the essential elements of fertility than those given
in the chemicals and, without stopping to consider this point, the carbona-
ceous matter of the manure has been credited with the superior effect
produced." As reported by Smith (1942) the 50-year average of con-
tinuous wheat plots on Sanborn Field shows that 6 tons of manure pro-
duced 18.8 bushels as compared with 20.3 bushels for a complete
fertilizer. The complete fertilizer contained nitrogen, phosphorus, and
potash equivalent to that contained in a 40-bushel wheat crop.

Green manuring has been practiced from early times with variable
results. The efficiency of the practice seems to vary with such circum-
stances as soil, climate, and crop. It is often stated that green manuring
is not effective in regions having less than 20 inches of rainfall. It is
generally believed that the main benefit of plowing under green manure
crops results from the addition of nitrogen and organic matter in the
soil. Other benefits that may result from the green manures are im-
proved physical properties and mobilization of plant nutrients. It is
apparent that many of the benefits materialize only when decomposition
takes place. Green manures are subject to rapid decomposition, which
means that the succeeding crop would receive most of the benefits and
that little nitrogen reserve would be built up. Indeed, results of longtime
fertility experiments have been disappointing with respect to build-
ing up organic matter and nitrogen reserves. In fact, results by Broad-
bent and Norman (1946), using isotopic nitrogen, show that the addition
of Sudangrass accelerated the decomposition of soil organic matter.
Thus, not only are green manures rapidly consumed, but they also seem
to increase the rate of decomposition of the organic matter already in the
soil.

Although the practice of green manuring may not be successful in
raising the nitrogen reserve of soils, real proof of the value of the practice
will depend on its effect on crop yields. There are numerous publications
showing the effect of green manuring on crop yields. There are few publi-
cations, however, showing the relative value of legume nitrogen turned
under in the form of green manure and a comparable amount of commer-
cial nitrogen. In many of the experiments reported, the quantity of
nitrogen turned under by legumes has been much greater than the quan-
tity of commercial nitrogen applied. Unless this point is considered the
wrong impression may be obtained as to the relative value of the two
forms of nitrogen.

At the Georgia Agricultural Experiment Station on Cecil sandy
loam Hale (1936) conducted an 8-year field test in which winter legume
green manure and nitrate of soda were compared for production of cotton,
Gossypium hirsutum. Hairy vetch and Austrian winter peas, Pisum

102 L. E. ENSMINGEB AND B. W. PEARSON

sativum arvense, turned under 2 weeks before planting cotton produced
1,044 Ibs. of seed cotton as compared to 951 Ibs. where 100 Ibs. of nitrate
of soda was used. Plots that received 200 Ibs. per acre of nitrate of
soda produced 1,154 Ibs. of seed cotton or approximately 100 Ibs. more
than the green manure plots. According to the author, at least one ton
of air-dry material was turned under, which should have been equivalent
to 60 to 80 Ibs. of nitrogen. Under conditions of this experiment com-
mercial nitrogen was applied at a rate somewhat lower than that fur-
nished by the winter legumes, but was more efficient pound for pound
for production of cotton.

A 12-year average of results in Louisiana reported by Haddon (1941-
1942) show that turning under 13,671 Ibs. of Oregon vetch, Vicia sativa,
increased the yield of seed cotton 932 Ibs. per acre. In the same test,
500 Ibs. per acre of nitrate of soda increased the yield of seed cotton
920 Ibs. Although the nitrogen content of the vetch is not given, it
would probably be safe to say that more nitrogen was turned under in
the form of vetch than was supplied by 500 Ibs. of nitrate of soda. The
increases were rather large in both cases due to the steady decline of
check plots over the 12-year period.

Krantz (1946) studied the effect on corn yields of rate of application
of nitrogen with and without cover crops. His data show that Austrian
winter peas produced an average of 88.6 bushels of corn at two locations
and that 90 Ibs. of commercial nitrogen produced 82.7 bushels. Also,
there was no significant difference between corn produced by 180 Ibs.
of commercial nitrogen and that produced by Austrian winter peas plus
90 Ibs. of commercial nitrogen.

Blair and Prince (1940) investigated the comparative values of green
manures and nitrate of soda for the growth of wheat and rye. Cow-
peas, Vigna sinensis, the first few years and soybeans thereafter, \vcrc
sown immediately after grain harvest. The soybean crops contained
about 69 Ibs. of nitrogen and it was assumed that two-thirds came di-
rectly from the air. Certain plots were top-dressed with 160 Ibs. per
acre of nitrate of soda. Average yields for 14 years show that nitrate
of soda produced 22.8 bushels of wheat as compared with 21.2 bushels
for green manuring. In the case of rye nitrate of soda produced 24.5
bushels and green manuring 18.6 bushels.

It is a common practice among the vegetable growers of the Atlantic
Coast to fertilize small grain cover crops with commercial nitrogen. If
the cover crop is utilized as a green manure, the question arises as to the
efficiency of the nitrogen thus applied in terms of increased yields of the
succeeding crop. Unpublished data by the Alabama Agricultural Ex-
periment Station show a greater efficiency of nitrogen as measured by

SOIL NITROGEN 103

corn yields when applied directly to the corn crop. For example, 40
Ibs. of nitrogen applied to oats at planting time plus 40 Ibs. at time of
turning oats produced 32.6 bushels of corn as a 6-year average. Eighty
Ibs. of nitrogen, applied to corn as a side-dressing produced 44.4 bushels.
Also, 40 Ibs. of nitrogen applied to oats at planting time and 40 Ibs. to
corn as a side-dressing produced 37.8 bushels, which is 6.6 bushels less
than the same amount of nitrogen applied directly to corn. Results of
greenhouse experiments by Pinck et al. (1948) show the largest yields
and the greatest recovery of added nitrogen where no green manure crop
was turned under.

Cover crops seeded for the purpose of turning under are used most
extensively in the eastern and southeastern states. Along the Atlantic
Coast catch crops are extensively used and in the southern states winter
legumes are commonly used. In either case the cover crops do not in-
terfere with regular cash crops. It is doubtful if cover crops can be
grown profitably unless they can be grown in off seasons. In the North,
legume crops in the regular rotation are in effect green manures.

The extent to which any individual farmer should use green manure
crops will depend on a number of factors. Cover crops that are turned
under in the late spring present a problem in getting land prepared in
time for the crop to follow. Farmers with tractors can usually handle
this problem without much difficulty. Land subject to serious erosion
should certainly be planted to a cover crop if at all possible in order to
lessen the loss of valuable topsoil. The cost of commercial nitrogen is
another factor to be considered. When nitrogen prices are high, use of
legume cover crops becomes more attractive. Also because of physical
properties, certain soils may be more responsive to green manuring than
others. On the basis of present data, it would seem wise to grow green
manure crops, especially legumes, wherever it is possible to turn them
under in time to plant the succeeding crop at its best seeding date, and
to use commercial nitrogen where legumes are not grown. Results by
Hale (1936) show that winter legumes plus 100 Ibs. of sodium nitrate
produced 104 Ibs. more seed cotton than winter legumes alone. Krantz
(1946) obtained an increase in yield of corn by the use of commercial
nitrogen with winter legumes over winter legumes alone. Research
workers in Illinois (1948) state that the lack of nitrogen is holding down
corn yields on many Illinois soils. They believe, however, that where
deep-rooting, "stand over" legumes occupy the land one-fourth of the
time, applications of commercial nitrogen are not economical. It is
apparent that the successful farmer must use considerable skill in com-
bining the best fertilizer and cropping practices to furnish adequate

104 L. E. ENSMINGEB AND B. W. PEARSON

nitrogen and other nutrients as well as to maintain a good physical
condition.

VIII. NITROGEN TRENDS IN VARIOUS PARTS OF THE U. S.

1. General Trends

Soil analyses show that many of our soils are lower in nitrogen and
organic matter than they were under virgin conditions, and that a state
of equilibrium has not been reached. If we are to look ahead with the
idea of maintaining the fertility of our soils, we should familiarize our-
selves with some general trends. Liprnan and Conybeare (1936) have
prepared a balance sheet of plant nutrients for the soils of the United
States and their estimates for nitrogen are as follows:

Tons

Losses (harvested crops, grazing, erosion, leaching) 22,899,046
Additions (fertilizer, manures, rainfall, irrigation

water, seeds, nitrogen fixation) 16,253,862

Net annual loss 6,645,184

According to their estimates, 25.09, 24.2, and 23.0 Ibs. per acre of nitro-
gen are lost annually by removal in harvested crops, erosion and leaching,
respectively. Mehring and Parks (1949) have estimated the amount of
nitrogen removed in harvested crops as compared with the amount added
in fertilizers and manures for various sections of the country. Their
estimates are presented in Fig. 4. Although in some sections more nitro-
gen is added by fertilizers and manures than is removed in crops, the data
for the United States as a whole show that about twice as much nitro-
gen is removed by crops as is added by fertilizers and manures. Since
most crops are grown to be harvested and removed, little can be done
about this loss other than to return as much of the plant residue as pos-
sible. Erosion losses are being reduced by better soil conservation prac-
tices, and leaching losses can be minimized by keeping the land covered
as much as possible and by applying commercial nitrogen to crops at
the proper time. Such practices can go a long way in decreasing the
spread between losses and additions.

A nitrogen balance may also be approached by increasing additions.
Additions of nitrogen to the soil may be increased considerably by in-
creasing the acreage of legumes and by increasing the amount fixed per
acre. The latter may be accomplished by proper inoculation, liming,
and fertilization. Soil treatments that favor symbiotic fixation will in
most cases be favorable for nonsymbiotic fixation. Efficient handling

SOIL NITROGEN

105

and use of manure would also favor a balance between additions and
removal of nitrogen. It is well known that much of the manure pro-
duced is never returned to the land. Lyon and Buckman (1947) esti-
mate that only 40 per cent of the nitrogen removed in crops will reach
the soil again. Mehring and Parks (private communication) estimate a
yearly production of 1,371,059,000 tons of manure in the United States
and utilization of only 221,790,000 tons. This amounts to an annual
application of manure nitrogen of 1,448,790 tons. It is obvious from

REMOVED IN HARVESTED CROPS
APPLIED IN FERTILIZERS
APPLIED IN MANURES

NEW MIDDLE SOUTH EAST WEST EAST WEST MOUNTAIN PACIFIC UNITED
ENGLAND ATLANTIC ATLANTIC NORTH NORTH &UTH SOUTH MUUNIAIN KAUNU STATES
CENTRAL CENTRAL CENTRAL CENTRAL

Fig. 4. Average annual removal of nitrogen in harvested crops and the amounts
replaced in fertilizers and manures. (Mehring and Parks, 1949.)

these figures that proper use of farmyard manure could materially in-
crease our nitrogen additions to soil. Commercial nitrogen is being
used in ever increasing amounts. Figures presented by Scholl and Wal-
lace (1949) show that 824,482 tons of commercial nitrogen were applied
to our soils in the year ending June 30, 1948. This discussion of nitro-
gen balance should not be taken to mean that a balanced nitrogen sheet
necessarily means sufficient nitrogen for maximum yields under all con-
ditions. It does indicate, however, the nitrogen trend for the country,
and points to the need of establishing a balance if we are to maintain
a high productivity for the country as a whole. Improved crop varieties
outyield old varieties and as a result remove more nutrients per acre.
Unless these nutrients are replaced, yields and perhaps quality of crops
will decline. The gradual decrease in protein content of corn grain dur-

106 L. E. ENSMINGER AND B. W. PEARSON

ing the past number of years may be due in part to the decline of soil
nitrogen.

2. Southeast

Fertility trends with respect to nitrogen vary somewhat from region
to region. In the Southeast soils have always been low in nitrogen and
organic matter. It is probable that some of the cultivated soils that have
been well managed are higher in nitrogen than under virgin conditions.
Conditions are favorable for microbial decomposition for a large portion
of the year, which makes it difficult to build up a reserve supply of
nitrogen. A turnover of nitrogen and organic matter can be maintained
by use of crop residues and summer and winter legumes. Even though
winter legumes may be grown successfully without interfering with regu-
lar cash crops, it is difficult for a farmer to turn more than one- fourth of
his crop land late in the spring with mule power. Until more farms are
mechanized, this puts a rather definite limit on the extent of winter
legume acreage. If Alabama is representative of the Southeast that
limit has not been approached. One-fourth of the cultivated land in
Alabama would amount to a little over 2,000,000 acres, whereas in 1948
there were approximately 1,000,000 acres planted to winter legumes. It-
is interesting to note, however, that the winter legume acreage in Ala-
bama is on the increase. In 1918 there were less than 1,000 acres planted
as compared with 1,000,000 acres planted in 1948. Assuming an average
fixation of 60 Ibs. of nitrogen per acre, legumes would furnish 30,000
tons annually. The fixed nitrogen plus slightly more than 40,000 tons of
commercial nitrogen gives a total of about 70,000 tons of nitrogen for
Alabama crops. Experimental results show that the cotton and corn
acreage alone should receive close to 100,000 tons. These figures serve
as an indication of the need for nitrogen in this part of the country if
crops are to be produced most economically. Studies in the Southeast
have revealed rather consistent returns from the application of nitrogen
to corn. In this area 2 Ibs. of nitrogen can be expected to produce a
bushel of corn. The extra bushels from nitrogen are economical bushels,
since they are produced at little added cost other than the cost of the
nitrogen. The same is true for cotton where a pound of nitrogen will
produce about 12 Ibs. of seed cotton. Cummings (1949) estimates that
an additional 165,000 tons of nitrogen would be required to increase corn
yields in nine southern states by 10 bushels per acre.

In the Southeast winter legumes play an important part as far as the
nitrogen economy of row crop land is concerned, but other legume crops
are becoming prominent in the overall nitrogen economy. Alfalfa and
Lespedeza sericea are becoming important forage crops, and, when grown

SOIL NITROGEN 107

on land suited to row crops, they can be used effectively in cropping
systems. This is pointed out by the work of Krantz (1949). He ob-
tained a yield of 127.2 bushels of corn per acre in 1947 following a 4-year
stand of alfalfa on a Cecil loam. The plot received in addition 500 Ibs.
per acre of 4-10-10 fertilizer. Results by Mooers and Hazelwood (1945)
show that land in high-yielding sericea for 3 or more years can be ex-
pected to produce large yields of corn for several successive years. The
acreage of improved permanent pastures is increasing rapidly. These
pastures are usually legume-grass mixtures. The legumes serve as forage
and supply some nitrogen for the grass.

3. Midwest

Many of the soils of the Midwest were rich in organic matter and
nitrogen when first brought under cultivation. Cultivation has caused
a decrease in the nitrogen and organic matter content of these soils with
a resulting decrease in fertility. Even so the average corn yield in Iowa,
i'or example, increased from 39.3 bushels per acre for the period 1929-
1933 to 54.4 bushels for the period 1939-1943. This increase in yield
of 15 bushels was brought about in spite of the decline in fertility, pri-
marily by the widespread use of hybrid corn and to some extent by better
cultural practices. This increased yield means an additional removal
of nutrient elements, which will have to be supplied in order to continue
to reap the benefits of hybrid corn. During the past 10 to 15 years there
has been an increase in acreage of such soil-depleting row crops as corn
and soybeans. This has been accompanied by a decrease in acreage of
legume-grass meadows and pastures. Thus, the more rapid decline in
soil fertility in recent years has resulted in greater increases in yields
from fertilizers, especially nitrogen.

Although the midwestern states are not using large quantities of
nitrogen when compared with consumption in some of the southern states,
the percentage increase in recent years has been large. For example,
nitrogen consumption in the West-North-Central States was 11,392 tons
for the year ending June 30, 1946, and was 37,580 tons for the year end-
ing June 30, 1948. Pierre (1949) stated that the North Central states
should be using more fertilizer and estimated that amounts used in Iowa
at present should be more than trebled.

The extent to which commercial nitrogen consumption will increase
in the Midwest will depend on several factors. Many of the soils of this
region are still relatively high in nitrogen, and if properly managed, a
high nitrogen fertility level may be maintained with little or no com-
mercial nitrogen. The climate of the Midwest favors the use of farm-
produced nitrogen. Temperature and rainfall are favorable to production

108 L. E. ENSMINGER AND B. W. PEARSON

and use of organic matter as a source of nitrogen for crops. The
type of farming carried on in the area is also favorable to farm-produced
nitrogen. Large quantities of feed crops are grown and on many farms
the production of these crops is accompanied by livestock production.
This type of farming not only encourages but necessitates the use of
legumes and sod crops in the cropping system and at the same time pro-
vides for the fixation of atmospheric nitrogen in the soil. These legume
and sod crops not only conserve and add nitrogen, but promote a better
physical condition.

4. Great Plains

The average annual rainfall of the Great Plains area varies from
slightly less than 13 inches in the western part to 22 inches in eastern
North Dakota. While rainfall is higher in the southern part of the
region it is relatively less effective in producing crops. Much of the
rainfall in the Great Plains occurs during the summer months.

According to Myers et al. (1943), moisture is the overall limiting
factor in crop production in the dry-land areas of the Great Plains. Such
conservation practices as terracing and contour farming can do much
to increase yields. Maintaining a high level of organic matter will in-
crease the efficiency of rainfall by increasing the rate of infiltration.
Especially in the western part of the region, alternate small grain and
fallow is practiced as a means of using the existing moisture to better
advantage.

A review of literature by Harper (1945) on the response of crops to
nitrogen fertilization showed that in most cases nitrogen increased yields
little or not at all. Land that has been summer fallowed will seldom
respond to nitrogen fertilization because of the accumulation of nitrate
nitrogen brought about by fallowing. Also, plowing as far ahead of
wheat sowing time as possible will usually bring about sufficient nitrate
accumulation to eliminate the need for commercial nitrogen.

5. Irrigated Regions

Most of the arid and semi-arid lands were low in nitrogen and organic
matter when first put under irrigation. It has been surprising that many
of these soils with little or no commercial nitrogen have produced profit-
able yields of grains and vegetable crops in rotation with alfalfa. As
a general rule about one-fourth the land is kept in alfalfa. McGeorge
(1949) believes that in Arizona crops are getting nitrogen from several
natural sources, such as irrigation water, especially pump water, and
nonsymbiotic nitrogen fixation. Analyses show 10 to 50 p.p.m. of nitrate
in many of the pump waters. It is evident that an acre foot of water

SOIL NITBOGEN 109

would contain a significant amount of nitrate. Smith (1944) working
with a red desert soil of Arizona found strong evidence of considerable
nonsymbiotic nitrogen fixation.

Even though these soils have produced good crops without the use
of much nitrogen fertilizer, recent work has shown that nitrogen fer-
tilizer may be used to advantage. Work by the Division of Soil Man-
agement and Irrigation, U.S. Department of Agriculture (private
communication) shows that the lack of nitrogen is limiting yields of non-
legumes in the Columbia River Basin. Nitrogen was found to increase
the efficiency of a limited supply of irrigation water. Also, an increased
supply of irrigation water was used even more efficiently with increasing
amounts of nitrogen. Salter (1949) has revealed that exceptional crop
yields may be produced if water as well as other factors are controlled.
Such yields as 162 bushels of corn, 174 bushels of grain sorghum and
almost 600 bushels of potatoes per acre have been obtained under irri-
gation.

Where the supply of irrigation water is limited, it is probable that
most of the needed nitrogen can be produced on the farm by growing
legumes in the rotation. Where a more ample supply of water is avail-
able, it is very probable that commercial nitrogen may be used
economically.

REFERENCES

Allison, F. E. 1947. Soil Sci. 64, 413-429.

Allison, F. E., Caddy, V. L., Pinck, L. A., and Armingcr, W. H. 1947. Soil Sci.
64, 489-497.

Allison, F. E., Sherman, Mildred S., and Pinck, L. A. 1949. Soil Sci. 68, 463-478.

Allison, F. E., and Sterling, L. D. 1949. Soil Sci. 67, 239-252.

Bear, F. E. 1949. Chemurgic Digest 8, 4-6, 26.

Bennett, H. H., arid Allison, R. V. 1928. The Soils of Cuba. Monumental Print-
ing Co., Baltimore, Maryland.

Blair, A. W., and Prince, A. L. 1940. New Jersey Agr. Expt. Sta. Bull. 677.

Bledsoe, M. R., Jr. 1937. J. Am. Soc. Agron. 29, 815-821.

Borst, H. L., McCali, A. G., and Bell, F. G. 1945. U.S. Dept. Agr. Tech. Bull. 888.

Bracken, A. F., and Greaves, J. E. 1941. Soil Sci. 51, 1-15.

Bracken, A. F., and Larson, L. H. 1947. Soil Sci. 64, 37-45.

Broadbent, P. E., and Norman, A. G. 1946. Soil Sci. Soc. Am. Proc. 11, 264-267.

Brown, A. L., Wyatt, F. A., and Newton, J. D. 1942. Sci. Agr. 23, 229-232.

Brown, P. E. 1916. J. Agr. Research 5, 855-869.

Buckman, H. O. 1910. J. Am. Soc. Agron. 2, 121-138.

Burgess, P. S. 1918. Soil Sci. 6, 449-462.

Call, L. E. 1914. J. Am. Soc. Agron. 6, 249-259.

Collison, R. C., Beatie, H. G., and Harlan, J. D. 1933. New York State Agr. Expt.
Sta. Tech. Bull. 212.

Corbet, A. S. 1935. Biological Processes in Tropical Soils. W. Heffer and Sons,
Cambridge.

110 L. E. ENSMINGEB AND B. W. PEARSON

Crowther, E. M. 1947. Rothamsted Expt. Sta. Rept. for 1947 , p. 31.

Cummings, R. W. 1949. Plant Food J. 3, 24-25.

Dean, L. A. 1930. Soil Sci. 30, 439-442.

Dhar, N. R. 1933. Soil Sci. 35, 281-284.

Dodge, D. A., and Jones, H. E. 1948. J. Am. Soc. Agron. 40, 778-785.

Duggar, F. 1899. Alabama Agr. Expt. Sta. Bull. 105.

Ensminger, L. E., and Gieseking, J. E. 1939. Soil Sci. 36, 57-68.

Ensminger, L. E., and Gieseking, J. E. 1942. Sott Sci. 53, 205-209.

Fowler, R. H., and Wheeling, L. C. 1941. J. Am. Soc. Agron. 33, 13-23.

Fraps, G. S. 1920. Texas Agr. Expt. Sta. Bull. 259.

Fraps, G. S. 1921. Texas Agr. Expt. Sta. Bull. 283.

Fraps, G. S., and Sterges, A. J. 1935. Soil Sci. 39, 85-94.

Fraps, G. S., and Sterges, A. J. 1947. Texas Agr. Expt. Sta. Bull. 693.

Gainey, P. L. 1936. Soil Sci. 42, 157-163.

Gainey, P. L. 1948. J. Agr. Research 76, 265-269.

Gainey, P. L. 1949. J. Agr. Research 78, 405-411.

Goring, C. A. I., and Bartholomew, W. V. 1949. Soil Sci. Soc. Am. Proc. 14, in press.

Gottlieb, S., and Hendricks, S. B. 1945. Soil Sci. Soc. Am. Proc. 10, 117-125.

Gowda, R. N. 1924. Soil Sci. 17, 333-342.

Gustafson, A. F. 1948. Using and Managing Soils. McGraw-Hill Book Co., New

York.

Haddon, C. B. 1941-1942. Biennial Rcpt. of the Northeast Louisiana Expt. Sta.
Hale, G. A. 1936. J. Am. Soc. Agron. 28, 156-159.

Hardy, F., and Hewitt, C. W. 1948. Tropical Agr. 25, Nos. 1-12. 38-40.
Harper, J. Horace, 1945. Soil Sci. Soc. Am. Proc. 10, 16-22.

Hays, 0. E., Bay, C. E., and Hull, H. H. 1948. ,/. Am. Soc. Agron. 40, 1061-1069.
Holley, K. T., Stacy, S. V., Bledsoe, R. P., Bogess, T. S., Jr., and Brown, W. L.

1948. Georgia Agr. Expt. Sta. Bull. 257.

Illinois Agr. Expt. Sta. 1948. Nine-Year Report (1938-1947).
Jenny, H. 1930. Missouri Agr. Expt. Sta. Bull. 152.
Jenny, H. 1933. Missouri Agr. Expt. Sta. Bull. 324.
Jenny, H. 1950. Soil Sci. 69, 63-69.

Jenny, H., Bingham, F. T., and Padilla-Saravia, B. 1948. Soil Sci. 66, 173-186.
Jenny, H., Gessel, S. P., and Bingham, F. T. 1950. Soil Sci., in press.
Jones, R. J. 1942. J. Am. Soc. Agron. 34, 574-585.
Kardos, L. T., and Bowlsby, C. C, 1941. Sott Sci. 52, 335-349.
Kelley, W. P. 1915. Hawaii Agn. Expt. Sta. Bull. 37.
Krantz, B. A. 1946. National Joint Committee on Fertilizer Applications. Proc.

Twenty-second Annual Meeting, 114-120.

Krantz, B. A. 1949. North Carolina Agr. Expt. Sta. Bull. 366.
Lipman, J. G., and Conybeare, A. B. 1936. New Jersey Agr. Expt. Sta. Bull. 607.
Lyon, T. L., and Bizzell, J. A. 1913. J. Am. Soc. Agron. 5, 45-46.
Lyon, T. L., and Bizzell, J. A. 1934. J. Am. Soc. Agron. 26, 651-656.
Lyon, T. L., and Buckman, H. O. 1947. The Nature and Properties of Soils.

Macmillan, New York.

Lyon, T. L., and Wilson, B. D. 1928. Cornell Univ. Agr. Expt. Sta. Mem. 115.
McGeorge, W. T. 1949. Commercial Fertilizer 78: 20-21, 44, 46-48.
Martin, J. P. 1941. Soil Sci. 52, 435-443.
Mehring, A. L. 1945. Ind. Eng. Chem. 37, 289-295.
Mehring, A. L., and Parks, R. Q. 1949. Agr. Chemicals 4, No. 10, 36-40.

SOIL NITROGEN 111

Miller, M. F., and Krusekopf, H. H. 1932. Missouri Agr. Expt. Sta. Res. Bull. 177.
Mohr, E. C. J. 1944. The Soils of Equatorial Regions. Edwards Bros., Inc., Ann

Arbor, Michigan.

Mooers, C. A., and Hazelwood, B. P. 1945. Tennessee Agr. Expt. Sla. Bull. 197.
Moser, F. 1942. /. Am. Sac. Agr on. 34, 711-719.
Myers, H. E. 1937. Soil Sci. 44, 331-354.
Myers, H. E. 1940. Soil Sci. Soc. Am. Proc. 5, 237.
Myers, H. E., Hallsted, A. L., Kuska, J. B., and Haas, H. J. 1943. Kansas Agr.

Expt. Sta. Tech. Bull. 56.

Norman, A. G. 1942. Soil Sci. Soc. Am. Proc. 7, 7-15.

Parberry, N. H., and Swaby, R. J. 1942. Agr. Gaz. New South Wales 53, 357-361.
Parker, F. W. 1946. /. Am. Soc. Agron. 38, 283-298.
Peevy, W. J., and Norman, A. G. 1948. Soil Sci. 65, 209-226.
Pierre, W. H. 1949. Plant Food J. 3, 23-24.

Pinck, L. A., Allison, F. E., Gaddy, V. L. 1945. Soil Sci. Soc. Am. Proc. 10, 230-234.
Pinck, L. A., Allison, F. E., and Gaddy, V. L. 1948. Soil Sci. 66, 39-52.
Rogers, H. T. 1941. Soil Sci. Soc. Am. Pruc. 6, 263-271.
Rost, C. O. 1939. Soil Sci. Soc. Am. Proc. 4, 281-287.
Russell, E. J. 1927. Soil Conditions and Plant Growth. Longmans, Green and

Co. Ltd., London.

Russell, J. C. 1927. J. Am. Soc. Agron. 19, 380-388.
Russell, J. C., and McRuer, Wm. G. 1927. Soil Sci. 24, 421-452.
Salter, R. M. 1949. Plant Food J. 3, 10, 23.

Salter, R. M., and Green, T. C. 1933. J. Am. Soc. Agron. 25, 622-630.
Scholl, W., and Wallace, Hilda M. 1949. Agr. Chemicals 4, No. 6, 34-39.
Sewell, M. C., and Gainey, P. L. 1932. J. Am. Soc. Agron. 24, 221-227.
Shutt, F. T. 1910. J. Agr. Sci. 3, 335-357.

Sievers, F. J., and Holtz, H. F. 1923. Washington Agr. Expt. Sta. Bull. 176.
Slater, C. S., and Carleton, E. A. 1938. Soil Sci. Soc. Am. Proc. 3, 123-128.
Smith, D. D., Whitt, D. M., Zingg, A. W., McCall, A. G., and Bell, F. G. 1945.

U.S. Dept. Agr. Tech. Bull. 883

Smith, G. E. 1942. Missouri Agr. Expt. Sta. Bull. 458.
Smith, H. V. 1944. Arizona Agr. Expt. Sta. Tech. Bull. 102.
Smith, H. W., and Vandecaveye, S. C. 1946. Soil Sci. 62, 283-291.
Snyder, H. 1905. Minnesota Agr. Expt. Sta. Bull. 89.
Springer, V. 1940. Bodenkundc Pflanzenernahr. 18, 129-167.
Swanson, C. 0., and Latshaw, W. L. 1919. Soil Sci. 8, 1-39.
Thorne, C. E. 1930. The Maintenance of Soil Fertility. Orange Judd Publishing

Co., Inc., London, 241.

Tidmore, J. W., and Volk, N. J. 1945. /. Am. Soc. Agron. 37, 1005-1010.
Tyulin, A. T. 1938. Soil Sci. 45, 343-357.
Uhland, R. E. 1947. Science in Farming. U. S. Dept. Agr. Yearbook Agr. (1943-

1947) pp. 527-540.

Waksman, S. A. 1923. Soil Sci. 16, 55-67.
Waksman, S. A., and Iyer, K. R. N. 1933. Sot'Z Sci. 36, 57-68.
Waksman, S. A., and Madhok, M. R. 1937. Soil Sci. 44, 361-375.
Waksman, S. A., and Tenney, F. G. 1927. Soil Sci. 24, 317-333.
Wheeting, L. C. 1937. Soil Sci. 44, 139-149.

White, J. W., Holben, F. J., and Richer, A. C. 1945. J. Am. Soc. Agron. 37, 21-31.
Whiting, A. L. 1926. J. Am. Soc. Agron. 18, 854-875.

Vegetable Production

J. E. KNOTT AND O. A. LORENZ

University of (California, Davis, California

CONTENTS

Page

I. Introduction 114

II. Fertilization 116

1. Liquid and Gas 116

2. New Materials 117

3. Fertilizer Placement 118

4. Nitrogen Sprays 118

5. Plow Sole Application 119

6. Starter Solutions 119

III. Trace Elements 120

1. Sodium 120

2. Molybdenum 120

IV. Development of New Vegetable Varieties 121

Resistance to Disease 122

V. Utilization of Heterosis 128

1. Sweet Corn 128

2. Tomato 129

3. Pepper 131

4. Carrot 132

5. Onion 132

6. Eggplant 132

7. Squash 133

8. Cantaloupe 134

9. Cucumber 134

10. Cabbage 134

11. Asparagus 135

VI. Growth Control Techniques 135

1. Plant Production and Handling 135

2. Fruit Set 137

3. Growth Inhibition 141

VII. Labor Saving Devices 141

1. Direct Field Seeding 142

2. Use of Herbicides 146

3. Harvesting Machinery 149

VIII. Possible Future Developments 151

References 152

113

114 J. E. KNOTT AND O. A. LORENZ

I. INTRODUCTION

Increase in population and changes in diet have greatly stimulated
vegetable growers to increase their acreage. Not only are more acres
being planted to vegetables, but the average yield per acre, in the case
of a considerable number of crops, has improved fairly steadily during
the past decade or more.

Thirty years ago the amount of land planted to crops for processing
exceeded that for the fresh market. Since strictly market garden produc-
tion, however, does not enter into statistical summaries, the difference
may not have been as great as the data indicate. From 1926 to 1942,
the fresh market acreage exceeded that for processing. This increase
was due to the development of large scale vegetable production in areas
at great distances from the ultimate markets. Since then the acreages
for processing and for fresh market have been about equal, first, because
of the increased demand for processing crops during the war years, and
second, because of the expansion of the frozen food industry.

Fresh market production of the 25 most important vegetables, omit-
ting potatoes and sweet potatoes from consideration, utilized an average
of 1,694,000, 1,677,000, and 1,892,000 acres annually for the three 5-year
periods: 1934-38, 1939-43, and 1944-48, inclusive. The 11 important
processed crops averaged 1,383,000, 1,633,000 and 1,916,000 acres during
the same periods.

The production of vegetables for processing, in the main, represents
cash crops in general farm rotations, and is conducted in areas where the
climate and soil provide a good yield at low cost. The vegetable crops
destined for the fresh market, on the other hand, are grown where the
climatic conditions are favorable for high yields and where the produce
may be sent to market under optimum price conditions. As indicated
above, there was little difference in the acreages of the two types of pro-
duction during the period 1944 to 1948 inclusive, yet the farm value
averaged $590,401,000 annually for the vegetables sold fresh and only
$215,569,000 for those processed.

The six most important states in the production of fresh vegetables
are California, Texas, Florida, New York, New Jersey, and Arizona. A
comparison of the average annual acreages for the periods 1944-48 and
1939-43 shows for California a gain of 13.0 per cent; for Florida, 29.2
per cent; for New York, 9.9 per cent; and for New Jersey a loss of 0.5
per cent. Acreage increase was greatest in Arizona (48.6 per cent) with
Texas next (41.8 per cent). During the past 5 years the average annual
acreage devoted to fresh vegetable production in Texas was 351,760,
compared to 382,290 acres in California. The expanding acreage in the

VEGETABLE PRODUCTION 115

lower Rio Grande Valley is aided by a favorable freight rate compared
to that from California. The upward trend in Texas coupled with the
shift from vegetables to field crops in certain districts of California
might make Texas the leading fresh vegetable producing state in the
near future.

The ranking states in the growing of vegetables for processing are
Wisconsin in the lead, followed by California, Indiana, Minnesota, New
York, and Illinois. The average annual acreage from 1944 to 1948 was
37.8 per cent greater in Wisconsin than in the previous five years. Cali-
fornia and Minnesota showed a 30 per cent gain; Illinois and New York
11.2 per cent and 15.6 per cent, respectively; while Indiana had a 4.4
per cent decrease. Although the total acreage is rather minor compared
to that of sweet corn grown for processing in Wisconsin, Minnesota, and
Illinois, a recent development in Idaho and Oregon has been the almost
three-fold increase, from an average of 4,000 acres each during the period
1937 to 1946 to over 11,500 acres in 1948. Washington had about the
same acreage in 1948 as Idaho and Oregon, although the percentage in-
crease was not so great. The processing pea acreages in these three
northwestern states has shown a somewhat similar expansion. Climatic
conditions in the northwest are such that quality is not rapidly lost as
these crops mature. Since a very considerable proportion of the sweet
corn and the pea crops there are quick-frozen, this maintenance of quality
is very important. The most spectacular acreage change among the
processing crops is that of green lima beans in California, this also being
due to the climatic adaptation of the area to a crop grown for freezing.
In 1948 there were 21,700 acres of green lima beans harvested, compared
to an average of only 3,220 acres a year during the period 1937 to 1946.

During the 10 year period 1939 to 1948, the national farm value of
all tomatoes for both fresh market and processing averaged annually
$139,105,000, while that of lettuce, its nearest rival, was $69,841,000.
Snap beans were next with $49,148,000; onion, celery, and peas were
closely grouped around the $40,000,000 figure.

The losses in weight and quality of fresh vegetables during the mar-
keting period are great, especially at the retail level. To reduce this
waste much research has been conducted; recent developments in con-
sumer packaging may be helpful. Space will not be available in this
review to cover these recent trends in marketing. Neither will it be
possible to review the voluminous literature concerning the new materials
coming to the fore in the control of insects and diseases.

Much has been accomplished in recent years in the improvement of
varieties, in the fertilization of vegetables, and in techniques to increase
yields or reduce unit costs. All of these are important to the future

116 J. E. KNOTT AND O. A. LORENZ

expansion of vegetable production and constitute some major advances
from recent research.

II. FERTILIZATION

Adequate fertilization is perhaps more important in the culture of
vegetables than with most other crops. The development of a high qual-
ity product is as great a consideration as is the maintenance of high
yields. The intensive production of vegetable crops on high-priced land
requires heavy fertilization. For their potentialities to be fully realized,
the new hybrids and other high yielding varieties may well require even
heavier and more efficient fertilization than has been practiced in the
past. The use of new and improved materials for the control of insects
and diseases will tend to justify an increased use of fertilizers. Therefore,
there is much research under way wherever vegetables are grown to
determine for such local conditions the best fertilizer materials, the
optimum rates at which they should be supplied, and the most suitable
methods of applying them.

1. Liquid and Gas

New products and new methods of fertilizer application are becoming
increasingly available for use by many vegetable growers. Important
new materials were described by Merz (1940). Probably the most sig-
nificant change in the fertilizer industry has been the use of liquid and
gaseous materials. These may be applied in the irrigation water, in-
jected into the soil, or sprayed directly on the soil, as described by King,
Newcomb, and Chenoweth (1943). McCollam and Fullmer (1948) esti-
mated that about 6 per cent of the total tonnage of fertilizer sold in
California in 1947 was applied in liquid form.

The material having the greatest increase in use has been anhydrous
ammonia (NH 3 ). Rosenstein (1936) called the method of applying this
material in the irrigation water "nitrogation." Cylinders of anhydrous
ammonia containing 150 Ibs. each are connected by steel tubes. The
gas is metered into the irrigation water through calibrated orifices,
the orifice size depending upon the temperature of the ammonia and the
rate of delivery desired. Sawyer (1948) stated that the concentration of
ammonia in the water should be between 50 and 75 p.p.m., with 110
p.p.m. as the upper limit.

Chapman (1944) showed the effect of soil permeability, temperature,
and agitation on the losses of ammonia from the irrigation water. He
concluded that under most field conditions losses will be less than 10 to
12 per cent but that with high water temperatures and low soil permea-
bility they might be over 26 per cent.

VEGETABLE PfiODUCTION 117

Beaumont and Larsinos (1932), in field and plot tests, found am-
monia solution to be only about 70 per cent as effective as equivalent
amounts of nitrogen derived from ammonia sulfate or sodium nitrate.
Chapman (1944) reported aqueous ammonia to be comparable with other
sources of nitrogen. Lorenz and Doneen (unpublished) have found
anhydrous ammonia as good a source of nitrogen for potatoes and onions
as ammonium sulfate.

Jenny, Ayers, and Hosking (1945) concluded that the adsorption of
ammonia from ammonium hydroxide was largely dependent upon soil
texture. The fine textured soils adsorbed more nitrogen than coarse
textured soils. They found that acid soils tend to adsorb more nitro-
gen from ammonium hydroxide than from ammonium sulfate, while the
reverse was true on alkaline soils.

"Nitrojection," described by Sawyer (1948), refers to the direct injec-
tion of anhydrous ammonia into the soil. By this method ammonia under
its own pressure is injected into the soil in furrows made by chisels or
other furrow openers. Andrews, Edwards, and Hammons (1948) state
that in Mississippi there are 200 machine.* for applying ammonia by this
method. Their work showed that anhydrous ammonia and solutions
of ammonia have crop-producing values equal to or superior to am-
monium nitrate placed in the soil in the conventional manner.

Jackson and Chang (1947) demonstrated the tremendous power of
the soil to fix gaseous ammonia and concluded that gaseous loss of NH 3
from the soil is not an important factor in the use of anhydrous am-
monia as a fertilizer. They showed that even coarse-textured soils have
good efficiency in retaining gaseous ammonia. Finding that the moisture
content of the soil had only a slight effect on ammonia retention, they
concluded that this factor could probably be neglected in field practice.

Jenny, Ayers, and Hosking (1945) found that NH 3 adsorbed on dry
clays was loosely held and could be removed by aeration, indicating that
direct injection of NH may not be advisable under certain soil
conditions.

8. New Materials

To meet the need for cheaper water-insoluble forms of nitrogen sev-
eral urea-formaldehyde materials have been produced. By varying the
proportion of these ingredients, fertilizers containing practically all
degrees of solubility can be prepared. McCool (1941) described one of
these, urea-ammonia-liquor-37, which in tests on tomatoes and corn
gave yields fairly comparable to those with nitrogen derived from cotton-
seed meal.

118 J. E. KNOTT AND 0. A. LORENZ

3. Fertilizer Placement

The latest general recommendations on fertilizer placement have
been summarized in a report by the National Joint Committee on Fer-
tilizer Placement (1948). These vary somewhat for the different vege-
tables, but for practically all of those planted on the level, the committee
recommends that the fertilizer be applied in bands 2 to 3 inches to the
side and 1 to 2 inches below *the seed. With large amounts of fertilizer
(1,000 Ibs. or more per acre) the bands should be placed at the greater
distance from the seed. Also, when exceptionally large amounts of fer-
tilizer are used, part or even all of it may be applied broadcast and
then plowed or disked under prior to planting.

Vegetables planted on beds and grown with furrow irrigation, as in
Arizona and California, should be fertilized by placing bands in the
shoulder of the bed, approximately 4 inches deep and 2 inches from the
plant row. Lorenz (1949) showed that with bed-grown, transplanted
onions, best results were obtained by placing the fertilizer directly under
the plant row before setting the plants. A single band in the center of
the bed 5 or 6 inches from the plants was inferior to one under each
row or in each shoulder of the bed. If applied approximately two months
after transplanting, fertilizers in the shoulder of the bed gave much
better results than in the center. Such direct-seeded crops as spinach,
onions, and peas, which are irrigated before emergence, experienced
harmful results when the fertilizer was placed 2 to 3 inches directly under
the plant row.

4- Nitrogen Sprays

Hamilton, Palmiter, and Anderson (1943) suggested the use of ura-
mon in foliage sprays as a means of regulating the nitrogen supply of
apple trees. This method of application has been applied recently to
certain vegetable crops. Results summarized by the Du Pont Company
(1948) show that potatoes have responded to midseason sprays contain-
ing 20 to 25 Ibs. of urea per 100 gallons of water. Plants vary, however,
in their susceptibility to injury from nitrogen foliage sprays and for
tomatoes it seems safe to use only 5 Ibs. of urea per 100 gallons of spray
as compared to 4 or 5 times this amount for potatoes. Tests conducted
in the greenhouse and in the field in California (Lorenz, unpublished) us-
ing rates as high as 60 Ibs. of urea for 100 gallons of water have shown
very little benefit on lettuce, spinach, and potatoes. If enough urea was
applied to give increased foliage color or growth, burning of the foliage
usually occurred.

VEGETABLE PRODUCTION 119

5. Plow Sole Application

"Plowing down" fertilizers has become popular in some of the Eastern
states. With that method, Sayre (1942), in New York, reported an
increase of approximately 2 tons of tomatoes per acre during dry seasons.
Rahn (1943), in Pennsylvania, reported slightly better results from fer-
tilizers plowed down in bands than from broadcast applications.

The benefits from plow sole applications are most frequently attrib-
uted to several causes. Since the fertilizers are located 6 to 10 inches
below the surface in moist soil, the roots can readily absorb them, The
plant nutrients, such as potassium and phosphorus, are in an area where
they are less readily fixed by the alternate wetting and drying of the
soil. Moreover, the crop is less likely to be injured by fertilizer burn-
ing due to high concentration of soluble salts.

There are two common methods of plowing down fertilizers: broad-
casting before plowing and applying as a single band in the bottom of
each furrow. The latter method appears to be the better, since it should
result in less fixation of certain plant nutrients. Inasmuch as the yield-
benefits reported from plow sole applications are small, it is doubtful that
this method of fertilizer application offers many possibilities. On the
other hand, it does present another method of band placement.

6. Starter Solutions

The common term for plant food dissolved in water and used to
hasten recovery of transplanted seedlings is starter solution. It is pre-
sumed that the application of fertilizer in the immediate area of the root
stimulates early root development. This is often reflected in more rapid
early growth and possibly in earlier maturity (Natl. Joint Comm. on
Fert. Appl., 1948). Sayre (1941) reported that starter solutions gave
good results on transplanted tomatoes, cabbage, celery, pepper, eggplant,
and muskmelon, and on certain seed-sown crops such as beans, peas, sweet
corn, and beets. Rahn (1943), in Pennsylvania, also noted that starter
solutions were most effective in increasing the early yield of tomatoes
and sweet corn but that somewhat less favorable results were obtained
on the early yields of cabbage and snap beans. The total yields of
tomatoes, snap beans, and cabbage were increased but the final yield of
sweet corn was not.

To prepare starter solutions, fertilizer mixtures of soluble materials
and high analysis, such as 25-52-0, 12-52-17, or simple compounds, such
as 11-48-0 Ammo-phos, sodium nitrate, or potassium nitrate, are dis-
solved in water at rates approximating 5 Ibs. per 50 gallons of water.
One pint to one quart of the starter solution is applied to each plant.

120 J. E. KNOTT AND O. A. LORENZ

Damage to the young seedlings may result if the fertilizer materials are
concentrated much above 5 Ibs. per 50 gallons of water.

III. TRACE ELEMENTS

The need for supplying small quantities of certain trace elements for
vegetable production on some soils has long been recognized. Such use
has become a part of normal commercial practice wherever the need for
these exists. Recently sodium and molybdenum applications to vege-
tables have been demonstrated as being of importance in some situations.

1. Sodium

Hartwell and Damon (1919) mentioned the possible value of sodium
in the growth of certain plants. Recent work by Harmer and Benne
(1941), Sayre and Shafer (1944), and Sayre and Vittum (1947) showed
distinct value from the use of this element in the production of certain
crops. Sayre and his coworkers demonstrated that sodium applications
from several sources corrected certain deficiency symptoms and increased
the yield of beets grown in New York. They believed that sodium func-
tioned as an independent element and did not partially substitute for
potassium. Raleigh (1948), using solution cultures, found that chlorides
gave more consistent increases in the growth of table beets than did
additions of sodium.

On the basis of field experiments, Harmer and Benne (1941) recom-
mended using 500 to 1,000 Ibs. of sodium chloride per acre for celery,
table beets, and turnips grown on the muck soils in Michigan. They did
not, however, recommend its use for certain other crops such as lettuce,
corn, parsley, potatoes, and tomatoes. Similar benefits were obtained
with sodium sulfate and sodium chloride, indicating that sodium, and
not chloride, was responsible for the yield increase. In fact, they ob-
tained increases in yield only when sodium was applied with potassium.

Holt and Volk (1945), using nutrient solution, found sodium to in-
crease the growth of 7 different crops when potash was omitted. They
stated that the beneficial effect of sodium decreased as the potash level
of the nutrient media increased but not in direct proportion to it. They
concluded that for some plants sodium may be an essential element for
maximum growth and that sodium can be substituted for potassium to a
variable extent, depending on the kind of plant.

2. Molybdenum

On certain soils in Australia and New Zealand, cauliflower and broc-
coli plants show a "whiptail" distortion and interveinal chlorosis. Davies

VEGETABLE PRODUCTION 121

(1945) found that the addition of 3 Ibs. per acre of sodium molybdate
to the soil eliminated the interveinal chlorosis of cauliflower plants.
Mitchell (1945) noted that the whiptail distortion could be corrected by
applications of 5 to 20 Ibs. of ammonium molybdate per acre. Since
then, Waring, Shirlow, and Wilson (1947) and Wilson and Waring (1948)
found that even smaller concentrations of molybdenum were effective
and suggested using rates of from one-fourth Ib. to two Ibs. per acre of
sodium or ammonium molybdate for the control of this disease. Wilson
(1948) obtained a yellowing of the outer leaves, marginal leaf burning,
and reduced plant size of lettuce grown on soils that were deficient in
molybdenum and where whiptail had been observed on cauliflower. This
disorder in lettuce was also corrected by the addition of molybdenum to
the soil. Warington (1946) has also shown that molybdenum is essential
in the nutrition of lettuce.

Wilson and Waring (1948) found that molybdenum deficiency symp-
toms were most prominent on plants receiving heavy nitrogen fertiliza-
tion and suggested that the function of molybdenum in the plant is
related to nitrate reduction. Their results agree with the work of Stein-
berg (1937) on Aspergillus niger. This indicated that molybdenum is
essential for activation of nitrate reductase in the reduction process,
whereby nitrates are reduced to ammonia.

IV. DEVELOPMENT OF NEW VEGETABLE VARIETIES

It would be impracticable to list all the new vegetable varieties that
have appeared in recent years. In some instances they have been short-
lived; in others, they will remain in use for many years. Some varieties
have very wide adaptation. Many, however, are fairly restricted in their
climatic requirements and do well only at certain seasons of the year
or in particular places. Thus the widely scattered vegetable industry
requires varieties with specific adaptation, and plant breeders have at-
tempted to fill this need.

The GREAT LAKES variety of lettuce (Lactuca sativa, L.), developed
by Barrens and Whitaker (1943), has caused widespread disruption in
the lettuce industry. While it was selected primarily for summer har-
vest in Michigan, its characteristics of hard heading, slow bolting, and
considerable tipburn resistance have enabled it to replace the better
quality Imperials formerly planted and have led to the production of head
lettuce in areas and at times of the year not previously possible. Various
improved strains of GREAT LAKES are now appearing, as well as varieties,
such as PENNLAKE (Lewis, 1949), which have GREAT LAKES in their
parentage.

122 J. E. KNOTT AND O. A. LORENZ

Thompson (1948) has released the PROGRESS lettuce variety designed
to replace IMPERIAL 44 which, though of high quality, was quite suscepti-
ble to tipburn and slime. PROGRESS is the result of a cross between
IMPERIAL 44 and an unnamed hybrid. When planted as an early season
lettuce, it is quite resistant to tipburn. The color is dark green and the
quality is good.

Examples that characterize the multitudinous objectives of modern
vegetable improvement are listed below:

The SUMMER PROLIFIC tomato (Lycopersicon esculentum, Mill.) de-
scribed by Denman (1948), which sets well in hot weather in Texas
and resists splitting after thunder storms; the high yielding MAGNOLIA
PICKLING cucumber (Cucumis sativus, L.) for Mississippi (Anderson,
1949) ; the IOWA YELLOW GLOBE 44 of Peterson and Haber (1949), a high-
yielding storage onion (Allium Cepa, L.) for peat or muck soils; the very
uniform, sure-heading MICHILI, a selection of CHIHILI Chinese cabbage
(Brassica pekinensis, Rupr.) by Drewcs (1948); the TRIUMPH bush lima
bean (Phaseolus lunatics, L.) developed by Magruder and Wester (1948)
for processing; the V-l cantaloupe (Cucumis Melo, L.), produced by
John Moran of the Ferry-Morse Seed Co. and discussed by Cuthbertson
(1948), which is sulphur resistant and thus may be dusted for the control
of powdery mildew (the latest development is SULFUR RESISTANT 91);
FLAGSHIP, a FI hybrid sweet corn (Zea Mays, L.) (Eto, 1948), which
has few suckers, the ears borne at a uniform height, and strong roots,
all of which adapt it to mechanical harvesting; WISCONSIN GOLDEN 800
(Andrew, 1948), a FI hybrid sweet corn with a high degree of cold re-
sistance in the spring; SAN JOAQUIN, an early-maturing, high-yielding,
nonbolting onion developed by Davis and Jones (1946) for the South-
west; and EXCEL, a Bermuda type onion for the South, the work of Jones,
Perry, and Davis (1947), a variety which produces 35 per cent more
marketable onions as much as 14 days earlier than YELLOW BERMUDA,
because it is fairly resistant to bolting, doubling, and splitting.

Resistance to Disease

The major portion of the breeding work with vegetables is aimed
at the development of resistance to disease, or in some cases, to insects.
The only sure avenue of control in combatting some diseases attacking
vegetables is inherited resistance, or at least considerable tolerance.
Evidence is accumulating that wheel injury to some crops when spray
and dust equipment moves through a field may be considerable. The
application of fungicides and insecticides by airplane is not entirely
satisfactory, because weather conditions may upset what needs to be a
carefully timed application. Moreover, resistance is particularly impor-

VEGETABLE PRODUCTION 123

tant in eliminating virus diseases, because even though insecticides may
be applied to control vectors, insects may inoculate the plants before
the insecticide kills them. Thus resistance is an advantage even for the
control of a disease or insect that may be held in check by the applica-
tion of some material to the foliage.

Wilt-resistant watermelons (Citrullus vulgaris, Schrad.) developed
elsewhere have been of little value in southeastern Missouri. Starting
with a resistant plant, probably a natural hybrid, in DIXIE QUEEN, Hib-
bard (1947) produced MISSOURI, QUEEN, which is highly resistant to most
strains of fusarium wilt. Epps and Sherbakoff (1949) report the release
of the MILES watermelon for use on wilt-infested soils of Tennessee. It
is the result of a cross between DIXIE QUEEN and the wilt-resistant KLON-
DIKE R-7 developed in California. MILES is a very productive, high qual-
ity melon, showing 12.1 per cent soluble solids in some tests.

The CONGO watermelon described by Andrus (1949) has considerable
resistance to anthracnose but not to downy mildew or to fusarium wilt.
An African melon was crossed with IOWA BELLE, and an inbred line from
this was crossed with GARRISON. The resulting melon has dark green
striping and a hard rind. A comprehensive review of watermelon breed-
ing has been prepared by Parris (1949) covering the inheritance of plant
and fruit characters and placing particular emphasis on the development
of disease resistance.

Many varieties of peas (Pisum sativum, L.) have been released for
use in the fresh market, cannery, or freezing plant. Each use requires
different characteristics as to color and size of berry, and concentration
of pod maturity, but all may need resistance to certain diseases. An-
derson (1948) reports the release of SHOSHONE for canning, a wilt-resistant
pea, resulting from a cross of PRIDE by ROGERS GIANT HAMPER. It has out-
yielded all other varieties of the PERFECTION type. VICTORY FREEZER with
the same parentage is designed for quick freezing^

Zaumeyer (1949) has developed RIVAL, a round-podded green bean
(Phaseolus vulgaris, L.), resulting from a cross between u. s. NO. 5 REFU-
GEE and FULL MEASURE. Tests in many locations have shown its wide
adaptation for market, canning, and freezing. The pods are borne fairly
high on the plant and are quite concentrated in time of maturity. They
are stringless and fiberless. RIVAL is resistant to the common bean
mosaic and the new strain, New York 15 mosaic. Both these viruses are
seed-borne. The average yield of RIVAL in all states where tests were
run was 9,080 Ibs. of green beans per acre, compared to 5,580 Ibs. for

TENDERGREEN.

Another Zaumeyer (1950) development is TOPCROP, a sister line of
RIVAL, from the u. s. NO. 5 REFUGEE by FULL MEASURE cross. It is immune

124 J. E. KNOTT AND O. A. LORENZ

to the common bean mosaic and the virus disease "greasy pod." The
entirely stringless green pods are round in cross-section, fairly straight,
and meaty, and are without fiber. TOPCROP is adapted to most parts of
the country for canning, freezing, and fresh market. It has the vigor
and hardiness to develop plants of good size under adverse conditions.

The SUPERGREEN snap bean by Anderson (1948) came from a cross
between IDAHO REFUGEE and FULL MEASURE. It shows considerable tol-
erance to both forms of the common bean mosaic and outyields TENDER-
GREEN and FULL MEASURE. It also has a fairly concentrated pod set.
A bean designed for the southern states is LOGAN, from a cross of u. s.
NO. 5 REFUGEE and BLACK VALENTINE. This variety described by Wade
(1943) is highly resistant to the common bean mosaic and to powdery
mildew. It shows tolerance to rust and to bacterial blight. One very
favorable characteristic is its ability to set pods during hot weather.

In some sections of the United States, curly top is a problem in bean
growing. The PIONEER, developed from a series of crosses of BURTNER
and BLUE LAKE (Dana, 1944), has resistance to the curly top virus and
that of the common bean mosaic. In the snap bean stage, the pods arc
of good quality.

Australia inclines to varieties stemming from the old CANADIAN WON-
DER. Shirlow (1947) has released RICHMOND WONDER from a cross of
CLARENDON WONDER by WELLINGTON WONDER. It is a heavy producer of
straight pods, 9 to 10 inches in length and oval in shape. These arc
fleshy but have fairly heavy strings and fibers as they age. The RICH-
MOND WONDER has comparatively good resistance to halo blight and angu-
lar leaf spot.

The greatest hazard to cucumber production in the southern states
is downy mildew (Pseudosperonospora cubensis). PALMETTO, a dark
green slicing variety, developed by Barnes (1948), is highly resistant
but not immune to this disease. Immunity to downy mildew (Perono-
spora spinaciae) has been introduced by Smith (1949) into the HOLLANDIA
and VIROFLAY varieties of spinach (Spinacea oleracea, L,). The immunity
was identified as a single dominant gene in a variety of spinach from
Iran.

Probably greater effort is being devoted to breeding tomatoes for
disease resistance than any other vegetable crop. Young and Mac-
Arthur (1947) have cataloged and * described 49 characters with gene
symbols and more than 60 other characters without gene symbols. This
work is an aid in recognizing the good, mediocre, and undesirable charac-
ters to be watched for in any developmental program.

The Southern Tomato Exchange Program (STEP) has been estab-
lished by the workers in the southern states to evaluate critically new

VEGETABLE PRODUCTION 125

tomato varieties arising from the breeding work in that area. Material
from other sections of the country is sometimes included. Yarnell (1948)
has discussed the method of operation and the advantages which have
accrued from the wide-spread observations on the disease-resistance and
productivity of new varieties prior to their official release. Other co-
operative vegetable trials are handled in the same general manner.

The steps in the development of WISCONSIN 55 tomato have been de-
scribed by Walker, Pound, and Kuntz (1948). This variety was designed
for canning purposes. Its tolerance of early blight delays defoliation.
WISCONSIN 55 is classed as intermediate in its resistance to fusarium
wilt.

The SOUTHLAND reported by Andrus (1948) is a general purpose
variety. It has resistance to collar rot and early blight. Moreover, it is
moderately resistant to phytophthera and alternaria blights, and is al-
most immune to fusarium wilt. SOUTHLAND is claimed to be relatively
resistant to fruit cake, puff, and blossom-end rot.

In addition to wilt resistance, unless a tomato for the South has the
ability to set fruit during hot weather, its seasons of production will be
limited. One that combines these two factors is the ALL SEASON developed
in Louisiana (George, 1949). The pink color of its fruit, however, is a
drawback on some markets.

For winter and spring production, where the variety must set fruit
under cool temperatures and low light intensity, the LAKELAND described
by Skirm (1948) may have a place. It is not as susceptible to fusarium
wilt or mosaic as is RUTGERS, which it might replace, but the fruits are
free from radial cracking and almost free from circumferential cracks.

The spotted wilt virus is under intensive study in Australia and the
United States. So far no highly resistant varieties have appeared. The
most ambitious breeding program for spotted wilt resistance in the
tomato is in Hawaii. There the multiple objective is for resistance to
spotted wilt, gray leaf spot, nematode, tobacco mosaic, and fusarium wilt,
along with high-yielding ability, top quality, and high vitamin C content.
Frazier, Kikuta, and Hendrix (1947) reported their progress in this
program. The first variety to be released from this work by Kikuta,
Hendrix, and Frazier (1945) is PEARL HARBOR. This was from a cross
between BOUNTY and sc-10 developed by the California Experiment Sta-
tion. BC-10 was a F 6 selection of 133-6 x L. pimpinellifolium back-
crossed to 133-6. PEARL HARBOR shows a high degree of resistance to
spotted wilt in Hawaii. Since this resistance is transmitted to the FI
progeny of crosses, the authors suggest that FI hybrid seed with PEARL
HARBOR as one parent might be useful. This may not prove satisfactory,
however, because Hutton and Peak (1949) have shown that the character

126 J. E. KNOTT AND 0. A. LORENZ

for spotted wilt resistance is not only highly recessive but also de-
pendent on multifactorial inheritance. The linkage between the re-
sistance and the L. pimpinellifolium growth characters is such that the
resistance is lost as succeeding generations approach desirable commer-
cial characteristics.

Smith and Gardner (1950) have pointed out that under severe epi-
demic conditions, so-called spotted wilt resistant varieties become
diseased. When spotted wilt is less epidemic, they make satisfactory
growth and set fruit, while susceptible varieties are severely damaged.
This confused situation may be due to the existence of several strains
of virus. Holmes (1948) accounted for the susceptibility of PEARL HAR-
BOR to spotted wilt in New Jersey by the presence of a strain of the
causative virus different from that occurring in Hawaii.

According to Norris (1946), the spotted wilt virus is a complex of 5
distinct strains: the Tip Blight, Necrotic, Ringspot, Mild, and Very
Mild. In his studies, L. peruvianum appeared to be immune to the last
4 strains of the spotted wilt virus and highly resistant to the Tip Blight
form. Therefore, the development of a resistant variety undoubtedly
will have to come from the crossing of L. peruvianum with L. esculentum,
a cross which is facilitated by the embryo culture technique of the
hybrid embryo as demonstrated by Smith (1944).

Cabbage yellows is an important disease on many soils, particularly
in the northern states. Research at the Wisconsin station has shown
that resistance is inherited in two ways. One type is a quantitative
character controlled by a number of genes and affected by high tem-
peratures and the nutrition of the plant. The other is a qualitative
character controlled by a single gene, JK, completely dominant to the
susceptible gene, r. The latter type is the better and has been incor-
porated in each of five early and midseason varieties of cabbage
(Brassica oleracea var. CAPITATA, L.) by Walker and Jolivette (1948)
who compared these with susceptible varieties of the same maturity
groups. The resistant varieties are WISCONSIN GOLDEN ACRE, RESISTANT

DETROIT, RACINE MARKET, MARION MARKET, and GLOBE. The single factor

resistance was also introduced into late-maturing varieties. Walker
et al (1948) described the IMPROVED WISCONSIN BALLHEAD and compared
its growth characteristics with that of late susceptible varieties. Work on
WISCONSIN HOLLANDER to which the single gene for resistance has been
added is still in progress.

Resistance to nematode, H. marioni, is one of the great needs in vege-
table production, especially with warm-season crops planted in the
southern and western United States. It is one of the most difficult prob-
lems facing the vegetable breeder. More progress has been made with

VEGETABLE PRODUCTION 127

the tomato than with any other vegetable crop. Watts (1947) obtained
cuttings from a F 3 plant developed by Smith (1944) from a cross between
MICHIGAN STATE FORCING (L. esculentum) and L. peruvianum, P. I. 128,-
657 through the use of the embryo culture technique. The FI clone is
self-sterile. Watts succeeded in crossing the FI clone with commercial
varieties and obtained one strongly resistant plant, which was self-
fertile. His results indicate that the nematode resistance is controlled
by two dominant factors.

Frazier and Dennett (1950), using Watts' material, note that re-
sistance is not a matter of preventing nematode entrance but of counter-
acting gall formation. A high degree of resistance was passed on to the
F 4 progeny of crosses with commercial varieties. A fruit diameter of 2
inches or more has been obtained, but further back-crossing is necessary
to get commercial types. The authors suggest that the high dominance
of resistance may make possible the early commercial use of FI hybrid
material.

In 1938 a new race of Erysiphe cichoracearum, D C threatened the
Imperial Valley cantaloupe industry. All commercial varieties of canta-
loupes, including POWDERY MILDEW RESISTANT NO. 45, were susceptible.
Pryor, Whitaker, and Davis (J946) described the steps taken to meet
this new race of mildew. Returning to some of the material originally
developed in the course of breeding of POWDERY MILDEW RESISTANT NO.
45, they found a gene for resistance to this second race of mildew in the
Cucumis Mela imported from India. This was incorporated into No. 45
and other commercial varieties so that by successive crossing and back-
crossing, three new varieties, No. 5, 6, and 7 have been developed all
highly resistant but not immune to both mildew races. Of these, No. 6,
is the most popular commercially.

Ivanhoff (1945) outlined the development of TEXAS RESISTANT NO. 1
cantaloupe, which is high-yielding in the absence of pests and shows
considerable resistance to downy mildew and aphids.

Resistance to a form of pepper mosaic occurring in Puerto Rico has
been introduced into the CALIFORNIA WONDER variety of pepper (Capsicum
frutescens, L.) by Riollano, Adsuar, and Rodriquez (1948) by crossing
it with a Mexican hot pepper, CAIARESMENO. Pungency and disease re-
sistance were found to be inherited as single genetic factors. A number
of resistant lines with acceptable market qualities have been developed
but are as yet unnamed.

128 J. E. KNOTT AND 0. A. LORENZ

V. UTILIZATION OF HETEROSIS

The potentialities in the use of heterosis appear unlimited, but each
vegetable crop presents certain problems in the production of FI seed
and in the commercial advantages which may be obtained by its use.
Ashton (1946) has given a comprehensive review of the research having
to do with the exploitation of heterosis with vegetable crops among
others.

The widespread employment of FI hybrids in sweet corn production
and the extensive developmental program with this crop have stimulated
interest in the possibility of capitalizing on this genetic phenomenon
in the production of other vegetables.

1. Sweet Corn

The FI hybrid sweet corns (Zea Mays, L.) have almost eliminated
the older open-pollinated varieties in the commercial production of sweet
corn in the United States. Perhaps the relative ease of producing hybrid
seed and its consequent minor cost have supplemented the natural ad-
vantages to be derived from its use. New hybrid sweet corns continue
to appear yearly. Singleton (1948) has reviewed the historical back-
ground, particularly in relation to the program of the Connecticut Agri-
cultural Experiment Station, which has been a leader in this work. He
points out that the great need now is a comprehensive study of quality
in inbreds. Doty et al. (1945) have demonstrated that inbreds differ
in their sugar content at harvest time and in the rate at which the sugar
is changed into polysaccharides during marketing or holding.

Earliness is an important economic factor in growing sweet corn for
market in many parts of the country. This involves both the adaptation
of a given variety to planting as early in the spring as possible, even
while growth conditions are sub-optimum, and the possession of the
ability to make the maximum rate of growth in order to reach market
maturity quickly. An approach to this latter problem has been suggested
by Haskell and Singleton (1949), who tested 17 lines of sweet corn,
mostly inbreds. They compared germination when seed was held in
moist soil at 50F. for 32 days before removal to a warm greenhouse
with that of seed planted directly in the field at the earliest possible date.
They found a significant correlation. The controlled temperature method,
they believe, is more severe than the field conditions but may serve as
a means of pretesting for cold resistance and may reduce the number of
lines which would need to be planted in the field. There is considerable
variation between lines. The ability to germinate at low temperatures

VEGETABLE PRODUCTION 129

seems to depend more on the genetic constitution of the embryo than on
the sugar-starch relationship of the endosperm.

The release of a new sweet corn inbred, Oh55, has been announced
by Park (1949). It has been crossed with Connecticut inbred C53 to
give the BROOKHAVEN, and with C68 to give the PERSHING varieties.
These two Y^ hybrids have shown outstanding resistance to the corn ear-
worm in four years' trials in southern Texas.

8. Tomato

Hybrid vigor in tomatoes has been extensively investigated and the
older literature well reviewed. Already there are a number of FX hybrids
named and available through commercial channels. Several problems
remain to be solved, if the use of hybrid tomato seed is to assume im-
portance commercially. The shift toward the use of FI hybrids in
tomato has not been as spectacular as with sweet corn.

Powers (1945)' found in his studies that the greatly increased yield
of his best FI hybrid over the best yielding variety, DENMARK, was due
to an increase in earliness. He concluded that all of the high producing
FI hybrids had at least one parent derived from crosses between Lyco-
persicum esculentum, Mill, and L. pimpinellifolium, (Jusl.) Mill.

Munger (1947) compared for two successive years the yielding ability
of FI hybrids of EARLIANA x VALIANT and EARLIANA x RUTGERS with the
standard early varieties, KARLIANA, VALIANT, and VICTOR. The FI of
EARLTANA x VALIANT gave significantly greater early yields than any of
the others. The other FI was better than EARLIANA or VALIANT. There
was little difference in the size of marketable fruits (over 3 ounces) be-
tween the varieties.

Shifriss (1945b) announced the release of the BURPEE HYBRID tomato.
Since then the BURPEE'S BIG BOY, BURPEEANA EARLY HYBRID, FORDHOOK HY-
BRID, and CLINTON HYBRID, all F/s, have been introduced and are avail-
able commercially. He pointed out that there may be a remarkable
increase in the number of fruits set on each plant, compared to the
parents. These fruits may be mutually inhibitive, or carbohydrate pro-
duction may be inadequate. As a consequence smaller sized fruits are
produced than would be anticipated in the FI progeny, which should
have fruits about intermediate in size compared to the two parental
forms. The counteraction of this size effect may call for radical changes
in fertilizer and other cultural practices. Shifriss (1947) believes that
the greater yield of hybrids is due to the greater absorption of nutrients
and to the ability to utilize them to advantage. They would therefore
have to be well fertilized if their ultimate potential were to be realized.

Spacing influences yields and would be related also to the need for

130 J. E. KNOTT AND O. A. LORENZ

optimum supplies of nutrients and moisture. Larson and Currence
(1944) studied the effect of 2-, 3-, 4-, and 6-foot spacings of plants in
4-foot rows, on the early yield, total yield, and fruit size of 4 hybrids
and the PRITCHARD variety. Not all the strains responded in the same
way to a change in spacing, depending apparently on their inherent plant
size. Early yield per acre was significantly greater at a 2-foot spacing
than at the other spacings. The total yield per acre was significantly
less at the 6-foot spacing than at the others. The 2- and 3-foot spacings
gave significantly smaller fruit than those at 4 and 6 feet. Hybrids which
developed their maximum size at a close spacing failed to increase it
at wider spacings.

Not only do cultural practices influence the results with FI hybrid
tomatoes, but soils are important also, as shown by Larson and Mar-
chant (1944). They compared three FI lines on two soil types in loca-
tions where the mean temperatures were approximately the same. Since
the hybrids did not behave the same on the two soils, the authors con-
cluded that it will probably be necessary to develop FI lines for specific
soil types. This might make the use of FI hybrids as complicated as is
the production of the seed itself.

According to Larson (1948), Currence has compiled data on the rela-
tionship between the best yielding hybrid and the best yielding standard
variety in trials in eight scattered geographical locations in the eastern
and central United States. The average increase in yield in all the states
was 32 per cent. This increase in yield in FI hybrids which usually is
accompanied by earliness would well justify the expense of the more
costly hybrid seed, provided fruit size is adequate.

It is toward the development of techniques for producing hybrid
seed at a minimum cost that considerable recent research has been
directed.

The use of complete male sterility in contrast to the semi-sterility
suggested by Currence (1944) has been recommended by Rick (1944)
as a means of avoiding the necessity for emasculation. In further studies
(Rick, 1945), male sterile mutants were found fairly readily, about 5
per cent of all unfruitful plants or 0.005 per cent of all plants; in all
varieties investigated Rick (1948) noted that male sterility in each
mutant was due to a different single recessive gene. No mutants for
female sterility or cytoplasmic male sterility were found. All but one
of the mutants so far studied yield 50 per cent male sterile individuals
by backcrossing while that one, ms5, produces functional pollen and so
is reproducible in 100 per cent of the progenies obtained by self-
fertilization.

Although the tomato is normally self-pollinated, variations in the rate

VEGETABLE PRODUCTION 131

of natural cross-pollination might be utilized with male-sterile mutants
to avoid hand pollination as well as emasculation. Rick (1947) studied
the effect of the planting arrangements of male-sterile and female plants
on the seed production in male-sterile plants through the activity of
native solitary bees at Davis, California. The highest yield of hybrid
seed for vector transmission of pollen was 4 per cent of that from fertile
plants. Vector activity in different regions of California measured in
the same manner has been reported more recently (Rick, 1949). Rates
of natural cross-pollination in different localities fluctuated greatly, the
yield of hybrid seed varying from 2 to 47 per cent of that of the fertile-
plant yields. Whether or not the natural vectors alone could be relied
upon to produce satisfactory crops of hybrid seed on male-sterile plants,
it is demonstrated that they can substantially supplement yields pro-
duced by hand pollination.

Roevcr (1948) reported a type of functional sterility in which the
anthers fail to dehisce. This was due to a simple recessive gene, which
could he easily incorporated in other desirable parent lines. It can be
maintained as a pun* line by hand pollination for use as a female parent.
Larson and Paur (1948) described this recessive mutant in greater detail.
It is a gamopdalous type with Hie extremity of the corolla coalesced,
preventing the discharge of the pollen.

With the use of some one of these forms of sterility in lines with good
combining ability and the employment of the mechanical pollen collector
of Cottrell-Dormer (1945), the time necessary to produce an ounce of
seed might well be reduced to 30 or 45 minutes. This would make hybrid
seed more reasonable in cost and lead to its extensive use in commercial
tomato production.

Larson and Currence (1944) compared the yielding ability of FI
and Fo lines of tomatoes with those of the parents. The average increase
in yield of the FT over the parental average was 39 per cent, while that
of the Fo over the parental average was 23 per cent. The use of F 2 seed
would be another way of avoiding the high cost of FI seed, provided the
F 2 populations are not undesirably variable.

3. Pepper

Martin (1949) reported the discovery of a male-sterile strain of
cayenne pepper (Capsicum frutescens, L.) which may expedite the pro-
duction of FI hybrid seed. Since the cost of obtaining FI hybrid seed
by hand pollination is almost prohibitive, he studied the potentialities
of F 2 progenies of fertile inbreds the seed of which could be produced
at a reasonably low cost. The F 2 seed of certain lines gave larger yields
than the inbreds but not equal to that from the FI seed. In the most

132 J. E. KNOTT AND O. A. LORENZ

desirable line, the F 2 gave an increase of 428 Ibs. of dried pepper per
acre over the inbred, and the FI 270 Ibs. over the F 2 .

4. Carrot

Welch and Grimball (1947) found a male-sterile plant in the carrot
variety TENDERSWEET (Daucus Carota, L.). This has potential use for
the easy production of F t hybrid seed which should result in uniform
sized roots and tops, and uniform color of roots which would be of mate-
rial advantage in the commercial production of bunched carrots.

5. Onion

The utilization of hybrid vigor in onions (Allium Cepa, L.) is devel-
oping rapidly. Some FI hybrids are available commercially, being
produced by the use of male-sterility, which has been described by Jones
and Clarke (1943). All the plants with normal cytoplasm, N, produce
viable pollen. When the sterile type of cytoplasm, S, is in combination
with the recessive gene for male sterility, ms, no viable pollen is produced,
but if it is present with the dominant gene, Ms, then viable pollen is
produced.

In order to maintain the materials for the production of FI hybrid
seed without emasculation and hand pollination, it is necessary to carry
along certain lines: a male-sterile line with the genotype S ms ms and a
fertile line with the genotype N ms ms. All the progeny from crossing
these two will be male sterile. The male-sterile line so maintained is
then crossed by natural vectors with the particular male-fertile inbred
parent showing the best combining ability. It may have the genotype
N Ms Ms, N Ms ms or N ms ms.

Jones and Davis (1944) report that the male sterile character has
been incorporated in almost all important commercial varieties. Re-
sistance to thrips and various diseases is readily introduced into the
FI hybrids in some cases. Among the advantages gained by the use of
FI hybrids can be great size, as in CALIFORNIA HYBRID RED NO. 1, if that
is desired. Far more important, however, are uniformity of maturity and
size of bulb. These factors are important in the mechanical harvesting
of onions and in reducing the grading necessary for market preparation.

6. Eggplant

A beneficial effect on yield from the planting of FI hybrids of the
eggplant (Solanum Melongena, L.) has been demonstrated by Odland
and Noll (1948). The size of fruit produced by the FI plants was about
the same as the mean of the fruit size of the parents. The increase in
yield was due, therefore, to the setting of more fruits per plant. They

VEGETABLE PRODUCTION 133

noted also a greater uniformity in size of fruits of the FI plants than
on the parental lines. The early yield on the hybrids was greater than
on the parents, and this was correlated with the larger total yield. The
mean yield of all 16 hybrids used in their studies was 62 per cent greater
than the mean yield of all the parents. The highest yielding hybrid
produced 17.25 tons more fruit per acre than the highest yielding parent.

NEW HAMPSHIRE HYBRID (not a FI) X FLORIDA HIGHBUSH gave FI progeny

which outyielded significantly all parent varieties and all hybrids except

EARLY LONG PURPLE X NEW HAMPSHIRE HYBRID.

7. Squash

Male sterility in squash (Cucurbita maxima, Duch.) was reported by
Scott and Riner (1946). They found that a single recessive gene con-
ditioned the abortion of the androecium in the bud stage with consequent
absence of pollen. This factor is inherited as a single recessive. Thus no
pollen is produced. This characteristic should make the production of
FI hybrid seed a simple process. It may be worthwhile to explore the
possibility of hybrid vigor in squash with this male sterile character,
as Hutchins and Croston (1941) found in their study of 10 FI hybrid
lines that increased yield over that of the higher yielding parent occurred
in 7 cases, while in the other 3 the yields were not greatly different.
This was a combined effect of maturity, weight per fruit, and number
of fruit per plant. It appeared that this heterosis effect was more pro-
nounced when the parents differed considerably in their readily observ-
able characteristics than when the parents were closely related.

Curtis (1942) described the YANKEE HYBRID, a yellow straightneck
summer squash (Cucurbita Pepo, L.), which is a FI resulting from a
cross between EARLY PROLIFIC STRAIGHTNECK and an inbred, CONNECTICUT
NO. 10. The very early production of female blossoms leads to a profit-
able large early yield, although the total yield through the whole grow-
ing season may not differ much from that of the EARLY PROLIFIC. Curtis
recommended that the 2 inbred parental lines be planted in alternate
rows, isolated at least 2 miles from other squash plantings. All male
blossoms should be removed from the plants in the row which is to be the
female parent.

Shifriss (1945c) discussed a form of male sterility in C. Pepo, similar
to that reported later in C. Maxima by Scott and Riner (1946). To get
the hybrid seed, all the male fertile plants in a backcross population are
removed in one roguing. Since the staminate blossoms appear 7 to 14
days ahead of the female blossoms, it is simple to take one male blossom
from each plant and determine by quick examination whether the androe-
cium is shrivelled. If this form of male sterility were introduced into

134 J. E. KNOTT AND O. A. LORENZ

the horticultural varieties of C. Pepo, it would facilitate the production
of FI hybrid seed at low cost.

8. Cantaloupe

Hybrid vigor in the cantaloupe or muskmelon (Cucumis Melo, L.)
was studied by Hunger (1942). Three FI hybrids between a fusarium
resistant selection No. 13 and BENDER, HONEY ROCK, WEAVER SPECIAL, or
QUEEN OF COLORADO were included in 4 field experiments in New York.
Considering his experiments as a whole, the hybrids as a group yielded
more fruit than their parents. The yield of fruits from the hybrids was
about the same or a little less than that of BENDER, the popular variety
of the area. When yield of flesh and sugar were used as the criteria,
the hybrids produced a higher proportion of flesh and much more sugar
than BENDER. The hybrids usually led in the production of early fruit
although in most cases the differences were not great. Munger suggested
that where disease resistance is dominant, hybrids may provide stop-gap
control until true breeding varieties are developed. For example, a FI
between No. 13 and POWDERY MILDEW RESISTANT NO. 45 carried resistance
to both fusarium wilt and powdery mildew. It was of good quality and
appearance. As with other crops, however, the cost of FI hybrid seed
may be a deterent to its use.

A recent development in this field was the discovery by Bohn and
Whitaker (1949) of a simple recessive gene for male sterility in the
muskmelon. Meiosis is apparently normal, but the development of the
mother pollen cells seems to end at the tetrad stage. This male sterile
mutant could be useful in breeding work as well as simplifying the pro-
duction of FI hybrid seed for commercial plantings.

9. Cucumber

Shifriss (1945a) developed the BURPEE HYBRID cucumber (Cucumis
sativus, L.) from an inbred line of a temperate zone cucumber crossed
with that of one from the tropics. While not immune to disease, it
shows considerable resistance to downy mildew, mosaic and wilt.

10. Cabbage

In a comparison of 7 standard varieties of cabbage (Brassica oleracea
var. capitata, L.) with an equal number of Fj hybrids, Odland and Noll
(1950) found that the yield of the hybrids was 31 per cent greater than
that of the standards. To take advantage of this hybrid vigor, they
have outlined a six-step procedure for the production of FI seed without
the necessity for hand pollination.

Four inbred lines, two each of two varieties, all homozygous for self-

VEGETABLE PRODUCTION 135

incompatibile genes are combined in a manner similar to that used
in the production of double-crossed corn. The inbred lines and the FI
hybrid in each variety constitute lines which are completely self-incom-
patible, yet entirely cross-compatible in both directions. This plan
should make practical the commercial growing of hybrid seed, once
the proper inbreds with good combining ability are developed.

11. Asparagus

The dioecious condition in Asparagus officinalis, L. might be utilized
to circumvent hand emasculation and hand pollination in the production
of FI hybrids. Randall and Rick (1945) suggested a method in which
homozygous pistillate lines could be developed from haploids that occur
naturally in twin seedlings and inbred staminate lines from the occasional
seeds produced by self-pollination in certain staminate plants, described
by Rick and Hanna (1943).

VI. GROWTH CONTROL TECHNIQUES

There is always an interest on the part of those concerned with vege-
table production in any procedure that will simplify operations or make
more certain that the ultimate goals of maturity date, quality and yield
are attained.

L Plant Production and Handling

In southern areas, where the season is earlier, the vegetable industry
has made an extensive practice of growing and shipping plants for trans-
planting into fields in the North. This is very common on the East
coast and to a lesser extent in the western states. Problems exist in the
transportation and handling of such plants and in the employment of
techniques which will aid them in resuming growth when put in their
final position.

More than 10,000 acres were devoted to tomato plant beds in Georgia
alone during 1946, according to Miller et al. (1949), who studied various
methods of shipping the plants to northern growers by rail and air.
They concluded that enough bunker ice should be used in the refrigerator
cars to reduce plant temperatures to the range of 50 to 70F.; higher
or lower temperatures were found injurious to subsequent growth. This
cooling would require 5,000 Ibs. of ice per car if outside temperatures
were between 70 and 80F., and 6,000 Ibs. if they ranged between 80
and 90F. Diagonally opposite vents should remain open. Fan cars
iced in the same way are preferred if available and should move with
vents closed. Shipment by airplane cuts the time in transit to about

136 J. E. KNOTT AND O. A. LOBENZ

one-sixth of that required for rail movement. The plants for air ship-
ment should be placed on wet peat moss in crates.

How the plants should be held if unfavorable weather conditions faced
a grower when he received his shipment was another aspect of this
problem that has been studied by Thomas and Moore (1947). Plants
were sent from Tifton, Georgia, to Lafayette, Indiana. Three shipments
arrived with lapses of 2 days, and the fourth after 3 days en route. All
were put in the field when the last one arrived. The other three had
been held 3, 5, or 7 days in a dry room at 70F. There were 8 bundles
of 25 plants, each packed with peat moss in a % bushel hamper. One
hamper of each shipment was moistened by standing in 3 inches of water
for 5 to 10 minutes daily. A significant decrease in stand and in the
average number of leaves left on each plant was evident at the end of
7 days in the field, when the plants were held longer than 3 days in
storage. After 16 days in the field, the plants in both lots held 7 days
and in that held 5 days without moistening had fewer leaves than the
other lots. Early yield, but not total yield, of fruit was significantly
greater in the plants set immediately or held for only 3 days than in
those held 5 or 7 days.

One explanation for the results of Thomas and Moore may have been
the exhaustion of the carbohydrate reserves during the period in transit
and storage. Went and Carter (1948) experimented with the application
of solutions containing 10 per cent sucrose and 0.025 per cent sulf anil-
amide with a little drene as a wetting agent. When this mixture was
sprayed on leaves of tomato plants held in the dark, apical growth
measurements showed that the sugar was absorbed more readily through
the lower epidermis than through the upper one. Growth, even in the
latter case, was almost twice that of the unsprayed controls. As a result
of this and other experiments, Went and Carter suggested that if tomato
plants were sprayed with a sugar solution just before shipping, they
would not suffer the setback^so commonly observed.

These authors noted also beneficial effects of sugar applications to the
foliage of tomato plants growing at high temperatures and low light in-
tensities. Smith (unpublished) utilized this technique in California on
tomato plants at the time they were trans])! anted from the plant bed to
the field. It very definitely reduced the number of plants lost where
field setting was done at high temperatures. No beneficial effect was
noted under moderate temperatures.

Another possible means of avoiding this loss of plants is suggested
by the work of Burgis (1948). He reported that spraying the tops of
tomato plants with a 10 per cent aqueous solution of Geon 31 X, a syn-
thetic latex, before they were pulled from the plantbed, or dipping them

VEGETABLE PRODUCTION 137

in such a solution as they were pulled, caused the plants to be stiff and
easy to handle in the transplanting operation. The treated plants showed
much less wilting the day after being transplanted to the field than did
those not given the latex treatment.

In his desire to place his crop on the market at the earliest possible
moment under his local climatic conditions, the grower usually starts
his plants under some sort of protection. Even though they suffer a
check when transplanted to the field, a gain in earliness of maturity is
still usually obtained. Cauliflower is one of the most difficult crops to
handle in this way because of its tendency to "button/' or apparently
to develop a curd too soon, with accompanying restricted leaf growth.
Carew and Thompson (1948) studied the factors responsible for this
development. They concluded that the curds in the so-called prematurely
heading plants are no further advanced than in those hidden by normally
developed foliage. "Buttoning," however, was increased if the plants
were held in flats beyond the best stage of development for transplant-
ing to the field and if the soil in the field had a low level of available
nitrogen thus preventing vigorous vegetative growth. Contrary to com-
mon belief, checking the growth of the young plants by exposure for
about 3 weeks to a temperature range of 40 to 50F., or keeping the
soil moisture low in the plant bed effected a decrease in later
"buttoning."

0. Fruit Set

It has long been known that tomato blossoms fail to pollinate them-
selves when temperatures are too low or too high, when there is no wind
to agitate them, or when certain physiological processes are at a mini-
mum. Went (1944) made extensive studies of thermoperiodicity, that
is, diurnal fluctuations in temperature, in relation to the growth and
fruiting of tomatoes. He placed particular importance on the night tem-
perature. Fruit set failed to occur, or was reduced, if the night tempera-
tures were below 59F. or above 68F. Varieties differed somewhat in
their critical minimum night temperature.

Inasmuch as temperature is important in the setting of vegetable
fruits and little can be done about it aside from making some adjust-
ments in planting dates, investigators have sought artificial means of
aiding fruit set under field conditions. A delayed planting date might
give the grower a yield of tomato fruits, but would likely cause him to
miss the high premium usually obtained for the first fruit to reach the
market early in the season. Mitchell (1947) stated that growth-regu-
lating substances have proved valuable in setting tomato fruits under
greenhouse conditions, but no consistent effect has been obtained out-

138 J. E. KNOTT AND 0. A. LORENZ

of-doors. No improvement in yield of MARGLOBE tomatoes grown in the
field was noted by Murneek (1947) in Missouri. He used /?-naphthoxy-
acetic acid at 20 to 100 p.p.m. and p-chlorophenoxyacetic acid at 10 and
20 p.p.m., applied at weekly intervals from mid-June to late September,
some concentrations on whole plants and others on the flowers only.
The only effect observed was a maximum increase of 15 per cent in fruit
size by application of the most concentrated solutions of the two growth
substances. He concluded that hormone sprays are of no value for
field-grown tomatoes except possibly where sunlight is subnormal.

On the other hand, Wittwer, Stallworth, and Howell (1948) found
that the minimum night temperatures during June in Michigan are below
the optimum range of 59 to 68F. suggested by Went (1944). Under
these conditions, a 25 p.p.m. p-chlorophenoxyacetic acid spray applied
to the first flower clusters of VICTOR and RUTGERS varieties planted for
early market gave a significant increase in number and size of early
fruits. Even with RUTGERS, grown for canning from southern-grown
plants and not commencing to bloom until July 9, spraying the first,
second, and third flower clusters increased the early and total yields
and the fruit size by about a half ounce over the controls.

During 1945 to 1947, Mann and Minges (1949) conducted 29 experi-
ments in widely scattered areas in California where market tomatoes
flower under cool temperatures. They used /J-naphthoxyacetic acid,
p-chlorophcnoxyacetic acid and its sodium salt, and 2, 4-dichlorophenoxy-
acetic acid, all applied as aerosols, water sprays, or in dust carriers to
the flower clusters as a whole. Each growth substance was used through
a range of concentrations, which were not necessarily the same for all
four materials. In all tests the treated blossoms produced larger fruits
than those of the check plants but size varied with the chemical used,
its concentration, and the carrier. This greater fruit size was considered
in part the reason, at least in some of the experiments, for the increased
early yields on treated plants. The growth-substance treatment seemed
to shift the yield of fruit to an earlier part of the season rather than to
increase the total yielding capacity of the plants. All combinations of
growth substances and carriers were effective in setting fruit. A single
spraying of a solution of 50 p.p.m. of 4-chlorophenoxyacetic acid gave
the most consistent results.

That the application of growth-regulator sprays may be effective
when night temperatures fail to drop below the 77F. maximum noted by
Went (1944) is evident from the work of Mullison and Mullison (1948).
In Caracao where the minimum night temperature was above 78F. and
the minimum day temperature 83F., rising to 90F. or above, flower
abscission is a frequent cause of poor fruit set. They grew by the gravel-

VEGETABLE PRODUCTION 139

culture technique 3 nondeterminate varieties, INDIANA BALTIMORE, MICHI-
GAN STATE FORCING, and PAN AMERICA, with 3 determinates, PEARSON,
VICTOR, and PRITCHARD. Of the growth stimulants used, p-chlorophenoxy-
acetic acid at 75 mg./l. gave the best fruit set and fruit size, with a very
considerable increase over the controls.

Some growth stimulants have failed to give beneficial results, as was
the case in the experiments of Paddock (1948). He made 3 applications
to tomato plants in Texas during the month of April. The growth
substances were either atomized on the inflorescences or applied to the
whole plant in the pest control spray containing Copper Hydro and lead
arsenate. Alpha-2, 4, 5-trichlorophenoxypropionic acid was used at 10
mg./l. while a 2-chlorophenoxypropionic acid was given at a 25 mg. rate.
Yields were reduced markedly by applying either hormone with the
copper and lead spray to the whole plants. Alpha-2, 4, 5-trichloro-
phenoxypropionic acid applied in this way caused considerable deformity
of the plants. The yields from the plants receiving the atomized treat-
ment of either of these two materials were not significantly different
from those of the plants having the pest control spray without a growth
substance.

An explanation for the sometimes conflicting results obtained by dif-
ferent investigators has been offered by Hemphill (1949). He found
that the growth-regulating substances applied to young flower buds may
delay their opening and also stimulate the development of rough, puffy
fruits. Successful results can be obtained if the spray containing the
growth substance is directed away from the terminal portion of the plant
which bears young buds. Hemphill thought that the application should
be delayed until 3 or 4 flowers in each of the several lower clusters have
opened.

The effect on their composition by the hormone setting of tomatoes
was studied by Holmes et al. (1948). They used PRITCHARD tomatoes set
by the use of /J-naphthoxyacetic acid at 50 p.p.m. There was no sig-
nificant difference in the mineral or vitamin constituents between fruits
set with the growth substance, with or without prior emasculation of
the blossoms, and the fruits from untreated plants.

A warning as to a possible detrimental effect on the tomato fruits set
by the use of growth substances has been given by Hewlett (1949).
Such fruits produced in greenhouses have a tendency toward premature
softening 1 or 2 days after picking. Whether this softening will be a
factor in field-grown tomatoes set with growth substances is not yet
known.

Blossom and pod drop in pole, bush, and lima beans are sometimes
a result of unfavorable conditions, especially of hot dry weather or of

140 J. E. KNOTT AND O. A. LORENZ

insect activity. The results in the use of growth substances as correctives
have been conflicting.

Wester and Marth (1947) obtained no significant effect in Maryland
on the yield of pods of 13 varieties of bush lima beans from a-naphtha-
leneacetic acid applied in various concentrations in dusts and in sprays.
The number of applications and their timing had no influence nor did
the inclusion of boron. The natural setting of pods was good at the
time they made their tests. Plants of the HENDERSON, PEERLESS, and
FORDHOOK 242 bush lima varieties were treated by Clore (1948), in the
State of Washington, with water sprays containing 0.5 per cent Carbo-
wax 4000 and 5, 50, 100 or 1,000 p.p.m. of a-naphthaleneacetic acid.
The yield from the 5 p.p.m. rate was not better than that of the untreated
plot but as the concentration was increased above this, the yields were
reduced very strikingly.

In Missouri, Wittwer and Murneek (1946) sprayed or dusted snap
bean plants of several varieties 3 to 5 days after the appearance of the
first flowers. In their experiments the treatments were given 1 to 5
times at weekly intervals. Para-chlorophenoxypropionic acid was more
promising than a-o-chlorophenoxypropionic acid, 0-naphthoxyacetic acid,
or 2, 4-dichlorophenoxyacetic acid. The best concentrations of this mate-
rial was 2 p.p.m. in a water spray, or 25 p.p.m. in a dust if 50 Ibs. of
dust were used per acre. Large benefits in yield were obtained when
flowering occurred in hot weather. Even when conditions were favorable
for fruit setting, there was a gain by the use of the growth regulator.
The authors expressed the belief that growth regulators are not likely
to be effective where seed formation and development are important, as
in peas, lima beans, and dry shell beans.

Fisher, Riker, and Allen (1946) in Wisconsin used IDAHO REFUGEE
and ROUND POD KIDNEY WAX beans. Their most effective growth stimu-
lant was a-naphthaleneacetic acid applied in dust form to the wax beans;
it was not helpful, however, on the green beans. Sprays were usually
detrimental. Dusting twice with 70 p.p.m. and 140 p.p.m. gave a slight
increase in yield in their first year's tests. This yield might have been
better had not rain followed each dusting. The next year, without rain
interference, a 24 per cent increase in yield was obtained by the use of
40 p.p.m. and 12 per cent from 80 p.p.m. ; 160 p.p.m. decreased it. The
authors concluded that the increase in yield from the treated plants was
due to an increase in the number of beans produced rather than to any
effect on size.

Para-chlorophenoxy acetic acid at 2 p.p.m., sprayed on TENDERGREEN
BEAN plants twice a week during August, in New York, by Randhawa
and Thompson (1948) for a total of 5 applications, increased the total

VEGETABLE PRODUCTION 141

yield. Beta-naphthoxy acetic acid and a-o-chlorophenoxypropionic acid
gave no significant increases, while 2, 4, 5-trichlorophenoxyacetic acid
depressed the yields. The second year of the experiments, when the
beans flowered in July, the early yields were increased but not the total
yields by 3 sprayings at weekly intervals in the case of all the materials.
Not all of the concentrations used gave increases in yield. The beneficial
effect seemed to be the darker green color, uniformity, and increased
length of pods compared with those of the untreated plots. There was
no difference in the number of seeds. No significant difference in the
ascorbic acid content of the beans from the sprayed and check plants
was found.

3. Growth Inhibition

Growth substances have been used also to retard the metabolic activi-
ties of vegetables. In an attempt to extend the marketing period of
cauliflower, Carolus, Lee, and Vandermark (1947) used the methyl ester
of a-naphthaleneacetic acid to check the formation of an abscission layer
in the leaf petioles around the curd to delay the yellowing of the petioles;
and to reduce loss in weight of the head. When the cauliflower heads
were held at 32F. and 80 to 90 per cent relative humidity, 100 mg. of
the chemical sprayed on the leaves, or placed on shredded paper in a
paper bag enclosing the head, was markedly effective in retarding the
undesirable changes.

Isbell (1948) studied the effect of dusts containing 2.2 per cent of
the methyl ester of naphthaleneacetic acid or of methyl 1 -naphthalene-
acetate on kohlrabi, turnips, potatoes, and sweet potatoes, using one Ib.
of dust to 8 to 10 bushels of the vegetable. The sprouting of potatoes
was delayed by the treatment. While sprouting of sweet potatoes ap-
peared to be delayed, there was some evidence of internal injury. Un-
treated lots of kohlrabi and turnips had more usable individuals at the
end of the storage period than did the treated lots.

On the other hand, Smith (1946) reported the successful retardation
of root and shoot growth of carrots, beets, rutabagas, and turnips dur-
ing storage by the use of the methyl ester of naphthaleneacetic acid, ap-
plied by either the dust or the impregnated shredded paper techniques.

VII. LABOR SAVING DEVICES

In addition to planting varieties that will yield well in spite of the
presence of pests, and to fertilizing and irrigating the plants so that the
maximum yield can be obtained, the grower is able also to cut his unit

142 J. E. KNOTT AND 0. A. LORENZ

cost of production by the utilization of mechanical aids and cultural
techniques.

Hand labor is usually expensive, either because of the hourly wage
or because of inefficiency. Thus any technique that will increase the
hourly output per worker helps the grower.

1. Direct Field Seeding

Certain vegetable crops have been started customarily in some form
of plantbed where they could be provided with close supervision while
awaiting suitable seasonal conditions in the field. This has been true
particularly in the case of tomato and celery. Transplanting does check
the growth of the plants and may reduce the yield.

Recently there has been a decided swing to the direct-field seeding of
these crops. Half the acreage of cannery tomatoes in certain of the
midwestern states is currently so planted and much of that in California
both for canning and for the fresh market in the fall and winter months.
More than half of the celery for late fall harvest in the central coastal
district of California has in recent years been seeded directly in the
field where the crop is to mature.

Several problems arise in direct-field seedings. One is the matter of
weed competition with the small seedlings. The pre-emergence spraying
of the planted field with a material which does not have a deleterious
influence on the germination of the crop seed has greatly reduced the
weed problem. Furthermore, selective oil sprays can be used to elim-
inate weeds from among the young celery plants.

Providing a proper moisture supply under field conditions is some-
times difficult. Doneen and MacGillivray (1943) have classified
vegetable seeds into 4 groups, depending on their ability to extract
moisture from the soil in the process of germination. Tomato seed was
able to absorb water fairly well when the soil moisture was close to the
permanent wilting percentage, although germination would occur sooner
if the soil moisture were higher than that. Of the 21 vegetables studied
celery seed, on the other hand, was the least able to extract moisture
from the soil. The moisture content had to be well above the permanent
wilting percentage of the soil in order to get any germination at all,
and close to the field capacity if satisfactory germination was to be
obtained. Thus with celery seed great care must be exercised in direct-
field seedings to see that the soil moisture content is held at a high
level.

Another characteristic of celery seed is its slow rate of germination
compared to many other vegetable seeds. Efforts have been made by
Taylor (1949) to overcome this delay. He developed a technique for

VEGETABLE PRODUCTION 143

prespouting hypochlorite-treated celery seed at an alternating tempera-
ture of 48F. for 16 hours and 70F. for 8 hours. After 8% days, about
10 per cent of the seeds showed sprouts. When planted in the field, such
seed began to emerge in 3% days. This was approximately 2 weeks
earlier than would have been expected from unsprouted seed.

Thinning of vegetable plants from seed planted directly in the field
is an expense. Moreover, if the plants in the seeded row are too thick,
or if thinning is delayed too long, there is likely to be damage to the root
systems of the plants left after thinning. This may produce a check in
growth that would be of some economic importance where speed of
maturation from seeding to harvest is critical. The accurate spacing of
each seed in the row does much to alleviate this condition and to make
thinning easier, or to eliminate it altogether. This can be accomplished
either by the use of drills which meter quite well those seeds that are
naturally more or less spherical in shape, or by adding a material to the
seeds which will make them spherical for subsequent use in a precision
planter. The precision planter is the more important of the two. There
is no point in making each seed into a spherical ball, if these are then
to be planted at random in the row by the dispersion system.

Pelleted onion and carrot seeds can be sown where the plants are to
mature. With crops such as lettuce, tomato, and others in which each
plant should finally be 10 to 24 inches from its neighbors in the row,
the job of cutting out the extra plants with a hoe is greatly simplified
if a pelleted seed is dropped every 2 to 3 inches than if unpelleted seed
were drilled at random. Moreover, the finger work to remove one of
two plants standing very closely adjacent is eliminated.

Bainer (1947) described a number of precision planters, both of com-
mercial and of modified types used in his tests: Cobbley, Rassmann,
Milton, International No. 40, and John Deere Nos. 55 and 66. The
planters were equipped with horizontal or vertical plates with a given
size of cell for the size of seed to be planted, a cutoff to prevent more
than one seed staying in each cell, and a knockout or ejector to make the
seed leave the cell at the proper time. An important characteristic of a
precision planter must be a smooth small tube (*/ to %") through
which the seed falls from the metering device to the furrow. Any
roughness in this tube will retard the fall and the seed will not be evenly
spaced.

The Ventura bean planter with vertical rotors has been modified
recently for precision planting of pelleted seed. Since no cutoff is needed,
there is no cracking or injury of the built up coating of the seeds. A
plexiglass tunnel or guide over the top of the cups enables the operator
to see that the machine is feeding properly. As the plate passes the apex

144 J. E. KNOTT AND O. A. LORENZ

of its rotation, the seed falls out of the cup against the back of the pre-
ceding cup. It is then ready to drop free at the proper moment. The
plexiglass tunnel holds the seed from dropping sideways off the plate.

Some confusion exists in the use of the terms coating and pelleting.
Some writers take the view that coating is building up one seed in size,
while pelleting involves putting several seeds into a ball as for range
reseeding.

The term coating might be more appropriate in cases such as that
recommended by Newhall (1945), in which onion seed received a light
treatment with methyl cellulose and then equal weights of seed and a
dust containing Arasan or Thiosan were mixed together to give the seed
protection from the soil-borne onion smut organism. The seed size was
not increased greatly.

Pelleting, on the other hand, more aptly describes the procedure in
which materials are added to a seed regardless of its original shape until
it becomes more or less spherical. Two general processes have been
used. By the method described by Vogelsang, Schupp, and Reeve (1947)
the seed is alternately wet with a methyl cellulose solution as it revolves
in a pan, and dusted with feldspar or preferably 65 per cent feldspar
and 35 per cent flyash. The methocel binds the material to the seed,
and the whole is built into a pellet. The other method, Burgesser (1949),
involves the use of a special montmorillonite which is adhesive when
dam}), forming a hard pellet as it dries.

The increase in weight of each individual seed depends on the original
shape. An onion seed can be built into a pellet 8 to 10/64 of an inch in
diameter, with an increase of about 8 times its weight. On the other
hand, the weight of each carrot seed is increased about 22 times to get
a pellet the same size as that of the onion, and a lettuce seed 60 times
for a pellet 10 to 12/64 of an inch in diameter. This increase in bulk
makes for more expense in handling. Inasmuch as pelleted seeds should
be accurately spaced, however, this means that far less seed is planted
to the acre. The saving in seed may just about offset the cost of pelleting
and the extra handling charges.

Linn and Newhall (1948) have compared onion seed coated with
methyl cellulose plus a fungicide with pellets made by use of methyl cellu-
lose plus feldspar and a fungicide. The latter were very hard balls,
while the former were more or less mealy. It required plate hole 20 in
a Planet Jr. No. 300 seed drill to plant about the same number of seeds
per foot of the feldspar pellets, as were distributed by plate hole 10 for
the coated seed, and plate hole 8 for the untreated seed. This was not
precision planting. The authors noted that when this conventional type
of seed drill was used, there was considerable cracking, splitting, and

VEGETABLE PRODUCTION 145

crushing of the pellets. The two types of covering of the seed plus either
Arasan or Tersan were about equally effective in the control of onion
smut. Leach (1948) found that when a quantity of Arasan equivalent
to 75 per cent of the seed weight was included in the coating of onion
seed, emergence was delayed by 50 per cent. Coating onion seed with
Arasan at 5 per cent of the seed weight, or without it, had no effect on
germination or rate of emergence. In his studies, the coating of tomato
seeds resulted in a slight delay in emergence but did not affect the per-
centage of germination. The addition of Arasan or Phygon to the coat-
ing material increased the protection, but organic mercury retarded and
reduced emergence.

Vogelsang, Schupp, and Reeve (1947) experimented with additions
to the pelleting material of superphosphate or of a 2-5-5 fertilizer at the
rate of 5 to 10 per cent of the weight of the seed. The 10 per cent
addition of superphosphate had the least injurious effect on emergence.
They stated that fungicides would have to be applied to the seed greatly
in excess of the recommendations of the fungicide manufacturers, if pre-
emergence and post-emergence damping-off were to be controlled. In
contrast, Burgesser (1949) stated that the results with the inclusion of
fertilizing materials in the pellets have not been encouraging, largely
because the quantity which could be incorporated without being toxic
is but a few ounces per acre. He remarked also that growth -promoting
substances, such as hormones and vitamins, have shown no consistent
value.

Other problems have been pointed out by Carew (1949). Sometimes
as much as 30 per cent of the pellets were "dummies," i.e., contained no
seed; the pellet had been formed around a bit of dirt or chaff by the
methocel process. In other instances each pellet contained 2 or 3 seeds.
These manufacturing details should be possible of correction since this
condition has not been the general experience elsewhere. Carew cited
the fact that there were sometimes 2 or 3 days delay in emergence of
lettuce seedlings from pelleted seed, and as emergence was uneven, it
tended to prolong the harvest period.

Bishop (1948) studied the effect of pelleting by the montmorillonite
method on the seed and its germination. Lettuce, tomato, and onion
seed were used. The pellets were crushed to remove the seed, or the clay
was washed off with a stream of water. The freed seed was then com-
pared with uncoated seed of the same lot, which was washed in an equiva-
lent quantity of water. There was no difference in the germination
behavior of the 3 lots, thus indicating that the pelleting process in itself
had not harmed the seed. In uncovered cold frames and in field tests,
the pelleted seeds germinated just as well, but at a slightly slower rate

146 J. E. KNOTT AND O. A. LORENZ

than the uncoated seeds. The author believed this delay was of no com-
mercial importance. The standard laboratory germination tests of naked
and pelleted seed showed some reduction in germination due to pelleting.
Vogelsang, Schupp, and Reeve (1947) have indicated that, under normal
temperature conditions, the rate of emergence is the same, but when it is
cool the pelleting may delay emergence by several days.

Plantings made with pelleted seeds fit well with the use of pre-
emergence sprays and selective herbicides. If the pelleting delays
emergence a few days, it may simplify the timing of a pre-emergence
spray. The systems used by carrot growers have been described by
Taylor (1949). Precision planting of seed and spraying with an oil
spray after emergence means that the rows can be placed close together
without danger of crowding the developing carrots and without the neces-
sity of having a wide uncropped area between every two broad-banded
rows. The practice of chiseling or deep loosening of the soil between
the rows, which is so ingrained in the minds of many growers, can be
eliminated. The tops are less leggy when the seed is pelleted and planted
with precision machinery, because crowding is avoided. Even then,
however, growers of carrots for bunching cannot count on having all
their plants produce good, smooth roots of marketable size. Mann and
MacGillivray (1949) noted that some carrot plants having plenty of
space in the row, as would be the case if pelleted seed were planted,
still developed roots too small for marketing. This small size was due
in part to hereditary factors. The most important reason, however, was
delayed germination. The percentage of carrot seeds with small or
weak embryos ran high in some samples, even though the seed size was
satisfactory. Slowly germinating seedlings suffered in competition with
the foliage or roots of the more vigorous plants.

Carew (1949) suggested that wider use of chemical weed killers
might be possible if pelleted seeds were planted. Activated carbon placed
in the material surrounding the seed might protect it from the harmful
effects of 2,4-D (2,4-dichlorophenoxyacetic acid), applied to the soil as a
pre-emergence spray.

2. Use of Herbicides

Since it has been demonstrated that the principal benefit from the
cultivation of the soil around growing plants is the control of weeds,
methods have been sought to reduce the labor involved. Chemical weed
control cannot completely replace, but should supplement, usual tillage
operations. Crafts and Harvey (1949b) have reviewed the literature
relating to the new weed control techniques.

This means of weed control has reached the stage of development

VEGETABLE PRODUCTION 147

where it is possible for research institutions to publish specific recommen-
dations for the use of herbicides on vegetables for example, those of
Carew (1949), Dunham, Grim and Heggeness (1949), and Crafts and
Harvey (1949a). There are 2 main methods employed in chemical weed
control in crops pre-emergence and post-emergence applications. In
the latter, selectivity must be considered, lest the vegetable plants as
well as the weeds be killed. With some crops, so-called "stem sprays"
can be employed. Here the spray is directed at the weeds and the
very base of the plants, thus avoiding the foliage of the crop. Much
work is in progress adapting the methods to species and varieties and
evaluating the potentialities of the new materials, such as maleic hydra-
zide and others that are becoming available.

Research has shown that 2,4-D has varying effects on vegetable
plants and seeds, depending on environmental conditions, especially tem-
perature and moisture, and on the time of applications in relation to
planting. Under moist soil conditions, the 2,4-D gives an effective con-
trol of weeds with a rapid disappearance of toxicity, whereas in dry soil
the toxicity remains for a long period, according to Warren and Hernan-
dez (1948) and Crafts (1949). That heavy rains following the appli-
cation of 2,4-D made above germinating seed may cause injury to the
developing seedlings is apparent from the studies of Dearborn, Sweet,
arid Havis (1948). Danielson (1948) obtained good control of the weeds
in an over-wintered crop of spinach in Virginia by the preplanting use of
1.4 Ibs. of the sodium salt of 2,4-D in 100 to 500 gallons of water per
acre. When the spinach was planted 4 days after treatment of the soil
in early November, the stand was poor; when seeding was delayed until
12 days after treatment, there was no injurious effect on the stand.

Sweet corn varieties differ in their susceptibility to 2,4-D injury.
Dearborn, Sweet, and Havis (1948) applied one Ib. of the ammonium
salt of 2,4-D to small sweet corn plants of 8 varieties 14 days after
planting. The foliage of LINCOLN and IOANA showed the least injury,
while that of SENECA DAWN and NORTH STAR was most severely affected.
The resistance of 18 varieties of sweet corn to 2,4-D was studied by
Ellis and Bullard (1948). They used 0.7 Ib. of the acid per acre as a
sodium salt when the plants were 15 to 18 inches tall. The spray was
directed toward the base of the plants, thus avoiding the leaves. The
corn had received two prior cultivations. No significant difference was
observed between the check and treated plots as to stand, yield of ears,
or time of maturity. The 2,4-D had apparently caused brittleness in
some varieties for a wind storm which occurred a week after the spraying,
snapped off more plants of COUNTRY GENTLEMAN, WHITE HYBRID 3321, and
HURON than of other varieties. Zink (1949b) gave 1, 2, and 3 Ibs. per

148 J. E. KNOTT AND O. A. LORENZ

acre applications of the sodium salt of 2,4-D to 12-inch high plants of
12 hybrid sweet corn varieties. The sprays were applied at the base
of the plant. He thought that the more advanced physiological age of
the quicker maturing varieties or the fact that not all varieties normally
produce brace roots to the same extent would be responsible for varietal
differences in resistance to development of collar effect about the stalk
and of abnormal brace roots. Brace roots would provide for easy access
of the 2,4-D. SENECA 60, MARCROSS, CARMEL CROSS, LINCOLN, and SENECA
CHIEF were least injured.

The butyl ester of 2,4-D was used by Alban and Keirns (1948) for
pre-emergence experiments with 25 different vegetable crops. An appli-
cation of 0.66 Ib. per acre held weeds in check for about 3 weeks; less
than that amount failed to give good weed control. Moreover, 2, 3, or 4
times the 0.66 Ib. rate eliminated or controlled all weeds for at least 6
weeks. Only sweet corn, snap bean, mung bean, potato, and asparagus
developed satisfactorily following a pre-emergence application of 1.33
Ibs. of the butyl ester per acre. These crops, plus tomato, peas, cucumber,
and lima bean, grew well where 0.66 Ib. of the ester were applied per
acre. The other 16 crops did not tolerate a concentration which would
control weeds.

Great reliance has been placed on oils as herbicides. Quite specific
recommendations are being made for their use as pre-emergence sprays
or as selective sprays. For pre-emergence work, it appears important
that the crop seed be planted deep enough to allow maximum germina-
tion of the weeds to be killed off before the crop seedlings emerge. This
has been stressed by Nixon and Smith (1949) for the elimination of
weeds in tomato plant beds and would hold true, of course, for any
planting where a pre-emergence herbicide treatment is to be made.

Havis (1948) studied the effects of 31 pure hydrocarbons of the aro-
matic, olefin, and paraffin series on peas, lettuce, spinach, carrot, onion,
and timothy. These hydrocarbons had boiling ranges from 80 to
300C. Those with a boiling "range between 150 and 275C. were in
general more toxic than those with a higher or a lower boiling range.
Stoddard Solvent, 4 aromatic distillates, and dinitro ortho secondary
butyl phenol (Dow Contact) were compared by Sweet and Havis (1948).
Application was made 2 days after planting beet and radish seed. The
tar distillates with high boiling ranges were not effective weed killers.
These authors also tested various nonselective herbicides applied about
the bases of tomato, cabbage, and broccoli plants 3 weeks after they had
been transplanted to the field. Even though no spray hit the foliage, the
ammonium salt of 2,4-D was injurious to the plants. Neither it nor the
methyl ester of naphthaleneacetic acid gave good control of the weeds.

VEGETABLE PRODUCTION 149

Dow contact, Stoddard Solvent, a heavy aromatic naphtha, and an aro-
matic distillate all gave good weed control. As machinery is developed
to make possible the directive application of herbicides as was done in
this work a whole new field of use for some of the oil fractions seems
probable.

Other materials are finding a place. Lachman (1948) has shown
that isopropyl n-phenyl carbamate as a pre-emergence spray at 5 or 10
Ibs. per acre killed grasses with little harm to the germination of the
seed of broadleaved plants. Snap bean, beet, spinach, and onions from
sets appeared especially resistant.

One of the newer materials is potassium cyanate. Hedlin (1948)
used 0.5 per cent and one per cent sprayed on seedling onions at a rate
of 80 gallons per acre. Weeds were effectively controlled, especially if
small. When the onions were larger, he used 1, 2, 3, and 4 per cent
solutions. Observations on weed kill indicated that a 2 per cent spray
was strong enough to control the weeds, if small, without injury to the
onion tops or the resultant yield. Zink (1949a) found that the degree
of control of weeds with potassium cyanate decreased rapidly as the
weeds became older. Effective control necessitated that the weeds be
dry when the herbicide was applied and have several hours of dry
weather following the application. A 0.9 per cent solution at 80 gallons
per acre applied under low pressure (25 to 50 Ibs. per square inch) to
onions 2 to 3 inches high did not injure the onions and controlled broad-
leaved weeds. Concentrations of 1.8, 2.4, or 3.6 per cent caused burning
of the foliage of seedling onions when used with or without a wetting
agent and applied at a rate of 60 gallons per acre. He found that garlic
was more tolerant of the stronger concentrations than was the onion.

Observations by growers and research workers indicate that when
weeds develop under dry conditions they may be hard to kill with potas-
sium cyanate even though they are small. Warren and Ellis (1950)
believed that the rate of application may have to be determined by the
succulence of the onion plant rather than by its size or stage of devel-
opment. In their studies, potassium cyanate did not injure onion plants,
if there had been considerable dry weather previous to the treatment.
On the other hand, if the plants were succulent, the same rate of appli-
cation used without injury on the somewhat hardened plants could prove
harmful.

8. Harvesting Machinery

Much of the developmental work on equipment for expediting the
harvesting of vegetable crops is done by the manufacturers of farm
machinery. Portable viners, which will cut the vines and depod peas

150 J. E. KNOTT AND O. A. LORENZ

and lima beans as they operate through a field, and machines for lifting
and topping carrots and beets to be stored or used in processing, are
pieces of equipment which are coming into general use.

Attempts to adapt for sweet corn harvesting the regular field corn
picker which snaps or pulls the ear from the stalk have proved only
moderately successful. Often there is too much bruising and husking
of the sweet corn ears. In addition, up to 25 per cent of the ears may
be left in the field. The use of pickers also breaks down the stalks, pre-
venting their use as silage. Burr (1949) surveyed the use of 131 me-
chanical sweet corn harvesters of 5 different types used in Wisconsin
and Minnesota in 1948. There was a wide range in the degree of success
obtained with them, depending in part on the tonnage of green matter
to be put through the machine. Twist (1949) described one type of im-
proved sweet corn harvester which harvests 2 rows of corn at a time.
The knives cut the stalks into segments as these are drawn down through
the picker heads. As the shank of the ear is cut, the ear falls to one side.

Not all sweet corn varieties are successfully harvested by machine,
according to Huelsen (1948). The COUNTRY GENTLEMAN hybrids are
fairly well adapted, EVERGREEN hybrids less well, and practically all of the
yellow hybrids poorly adapted. Both Huelsen and Twist (1949) have
indicated some of the characteristics which must be incorporated into
a sweet corn variety before it can be considered well adapted for me-
chanical harvesting. These are: relatively few leaves, a stiff strong
stalk and root system, few suckers, tight heavy husks, ears borne at least
18 to 24 inches above the ground surface, and a medium shank to the
ear. Huelsen thought that an ear w r ith a shank which will snap off
readily at a weak node just below the ear, as in the case of Country
Gentleman, would be best. Thus a machine which gave the stalk a
vigorous shake might work.

The method of planting has a bearing also on the success in the
use of mechanical harvesters.^ Drilled corn seems to be handled more
readily by the machine than check-rowed corn.

The harvesting of snap beans is an expensive hand operation and one
that presents difficulties if it is to be done mechanically. A variety with
a concentrated pod set would be best adapted for machine harvesting.
Yeager (1949) described a picking machine, developed by J. W. Ward
of Vernon, N. Y., which effectively harvests the pods from 2 rows of
beans with very little bruising. The leaves are removed automatically.
Then even the very small beans are taken from the vines by the teeth
on the revolving drums and deposited in bags.

Various types of machines have been designed to lift and top onions.
The elevating mechanism of most of these operates much the same as

VEGETABLE PRODUCTION 151

that of a potato digger. A different principle is incorporated in a ma-
chine designed and built by Lorenzen (1950). A narrow, wedge-shaped
blade cuts the roots, thus loosening the bulbs. The onion plants are car-
ried upward and to the rear by a pair of round rubber belts, which grasp
them by their necks. Disc knives cut off the tops, allowing the bulbs
to be conveyed to the sacks. This apparatus will harvest 2 acres in a
10-hour day. Between 87 and 99 per cent recovery of the bulbs has
been obtained, depending on the weed population in the field. This
harvesting machine can be used readily on early onions, which in many
cases are harvested before many of the tops have commenced to ripen
at the neck, or in fields where 50 to 60 per cent of the tops are down.
A planting made with Fj hybrid onion seed would be likely to reach the
maturity stage for harvesting more uniformly than would ordinary
stocks, a factor which would be an advantage for mechanical harvesting.
Thus with all 3 of these crops, sweet corn, snap bean, and onion, it
is obvious that a breeding program is desirable in order to develop
varieties with those particular characteristics which will fit them for
mechanical harvesting. This same procedure will have to be followed
in the mechanical harvesting of asparagus and tomatoes, which is already
in the testing stage. Modifications in spacing and other cultural tech-
niques must accompany the development of adaptive varieties, if hand
labor is to be eliminated in the harvesting of these two crops.

VIII. POSSIBLE FUTURE DEVELOPMENTS

One may expect that the trends indicated in the preceding discussion
will continue in much the same directions in the near future. There is
likely to be constant emphasis on the development of methods and tech-
niques to reduce production costs per unit of the harvested crop. Herbi-
cidal sprays may change our whole conception of the spacings that
should be given to vegetable plants. Closer spacing than is now com-
mon may permit a greater return per unit area for the usually high-
priced land on which vegetables are so often grown. Such a change will
necessitate a re-evaluation of fertilizer and irrigation practices.

The future appears bright for the development of new vegetable
varieties of high quality. Many of these will have improved resistance
to some of the diseases difficult to control otherwise.

The vegetable industry as a whole can benefit materially from the
development of new uses for vegetables. Furthermore, the distribution
system by which vegetables are marketed needs to be studied critically,
both from the economic aspect and from that of maintaining quality.

152 J. E. KNOTT AND O. A. LORENZ

It may be that consumer-size packaging, now being widely discussed,
will play an important part in altering the distribution of fresh vegetables.

REFERENCES

Alban, E. K., and Keirns, V. E. 1948. Proc. Am. Soc. Hort. Sci. 51, 526-532.

Anderson, M. E. 1948. Western Canner and Packer 40 (2), 51.

Anderson, W. S. 1949. So. Seedsman 12 (2), 14.

Andrew, R. H. 1948. Market Grow. J. 77 (3), 23.

Andrews, W. B., Edwards, F. E., and Hammons, J. G. 1948. Mississippi Agr. Expt.

Sta. Bull. 451.

Andrus, C. F. 1948. So. Seedsman 11 (9), 16.
Andrus, C. F. 1949. So. Seedsman 12 (11), 13, 43.

Ashton, T. 1946. Imp. Agr. Bur. Plant Breeding and Genetics, Cambridge, England.
Bainer, R. 1947. Agr. Eng. 28, 49-54.

Barnes, W. C. 1948. Proc. Am. Soc. Hort. Sci. 51, 437-441.
Barrons, K. E., and Whitaker, T. W. 1943. Quart. Bull. Michigan Agr. Expt. Sta.

25, 252-254.

Beaumont, A. B., and Larsinos, G. J. 1932. Am. Fertilizer 76, 9-10, 28, 30.
Bishop, J. C. 1948. California Agr. 2 (8), 6, 16.

Bohn, G. W., and Whitaker, T. W. 1949. Proc. Am. Soc. Hort. Sci. 53, 309-314.
Burgesser, F. W. 1949. Fruit Veg. Rev. 11 (1), 18-19.
Burgis, D. S. 1948. Market Grow. J. 77 (8), 21, 36.
Burr, H. R. 1949. Canning Trade 71 (38), 7-8.

Carew, H. J. 1949. New York Agr. Expt. Sta. Farm Res. 15 (2), 16.
Carew, J. 1949. Cornell Agr. Ext. Bull. 769.

Carew, J., and Thompson, H. C. 1948. Proc. Am. Soc. Hort. Sci. 51, 406-414.
Carolus, R. L., Lee, S. H., and Vandermark, J. S. 1947. Proc. Am. Soc. Hort. Sci.

49, 367-369.

Chapman, H. D. 1944. Proc. Natl. Comm. on Pert. Appl, 18-23.
Clore, W. J. 1948. Proc. Am. Soc. Hort Sci. 51, 475-478.
Cottrell-Dormer, W. 1945. Queensland J. Agr. Sci. 2, 157-169.
Crafts, A. S. 1949. Hilgardia 19, 141-158.

Crafts, A. S., and Harvey, W. A. 1949a. California Agr. Ext. Circ. 157.
Crafts, A. S., and Harvey, W. A. 1949b. Advances in Agron. 1, 289-320.
Currence, T. M. 1944. Proc. Am* Soc. Hort. Sci. 44, 403-406.
Curtis, L. C. 1942. Connecticut Agr. Expt. Sta. Circ. 152.
Cuthbertson, D. C. 1948. West. Feed and Seed 3 (5), 8, 20.
Dana, B. F. 1944. Seed World 55 (8), 46-47.
Danielson, L. L. 1948. Proc. Am. Soc. Hort. Sci. 51, 533-535.
Davies, E. B. 1945. Nature 156, 392-393.
Davis, G. N., and Jones, H. A. 1946. So. Seedsman 9 (7), 17.
Dearborn, C. H., Sweet, R. D., and Havis, J. R. 1948. Proc. Am. Svc. Hort Sci.

51, 536-540.

Denman, T. E. 1948. Seed World 62 (7), 12.

Doneen, L. D., and MacGillivray, J. H. 1943. Plant Physiol. 18, 524-529.
Doty, D. M., Smith, G. M., Roach, J. R., and Sullivan, J. T. 1945. Indiana Agr.

Expt. Sta. Bull. 503.
Drewes, H. 1948. Seed World 62 (8), 36-37.

VEGETABLE PRODUCTION 153

Dunham, R. S., Grim, R. F., and Heggeness, H. G. 1949. Minnesota Agr. Ext.

Pamphlet 168.

Du Pont de Nemours, E. I., and Co. 1948. Du Pont Nugreen Fertilizer Com-
pound Spray Nitrogen.

Ellis, N. K., and Bullard, E. T. 1948. Proc. Am. Soc. Hort. Sci. 51, 505-508.
Epps, J. M., and Sherbakoff, C. D. 1949. Seed World 64 (6), 14, 16.
Eto, W. 1948. Seed World 62 (7), 26.

Fisher, E. A., Riker, A. J., and Allen, T. C. 1946. Phytopathology 36, 504-523.
Frazier, W. A., and Dennett, R. K. 1950. Proc. Am. Soc. Hort. Sci. 55, in press.
Frazier, W. A., Kikuta, K., and Hendrix, J. W. 1947. Proc. Am. Soc. Hort. Sci. 49,

235-240.

George, L. V. 1949. So. Seedsman 12 (1), 13.
Hamilton, J. M., Palmiter, D. H., and Anderson, L. C. 1943. Proc. Am. Soc. Hort.

Sci. 42, 123-126.

Harmer, P. M., and Benne, E. J. 1941. /. Am. Soc. Agron. 33, 951-979.
Hartwell, B. L., and Damon, S. C. 1919. Rhode Island Agr. Expt. Sta. Bull. 177.
Haskell, G., and Singleton, W. R. 1949. J. Agron. 41, 34-40.
Havis, J. R. 1948. Proc. Am. Soc. Hort. Sci. 51, 545-546.
Hedlin, W. A. 1948. Proc. Am. Soc. Hort. Sci. 51, 501-504.
Hemphill, D. D. 1949. Missouri Agr. Expt. Sta. Res. Bull. 434.
Hibbard, A. D. 1947. Missouri Agr. Expt. Sta. Bull. 502.
Holmes, A. D., Smith, C. T., Kuzmeski, J. W., and Lachman, W. H. 1948. Food.

Technol. 2, 252-255.

Holmes, F. O. 1948. Phytopathology 38, 467-473.
Holt, M. E., and Volk, N. J. 1945. J. Am. Soc. Agron. 37, 821-827.
Hewlett, F. S. 1949. Proc. Am. Soc. Hort. Sci. 53, 323-336.
Huelson, W. W. 1948. West. Feed and Seed. 3 (10), 10, 17-18.
Hutchins, A. E., and Croston, F. E. 1941. Proc. Am. Soc. Hort. Sci. 39, 332-336.
Hutton, E. M., and Peak, A.R. 1949. J. Australian Inst. Agr. Sci. 15 (1), 32-36.
Isbell, C. L. 1948. Proc. Am. Soc. Hort. Sci. 52, 368-374.
Ivanhoff, S. S. 1945. So. Seedsman 8 (2), 11, 28.

Jackson, M. L., and Chang, S. C. 1947. J. Am. Soc. Agron. 39, 623-633.
Jenny, H., Ayers, A. D., and Hosking, J. S. 1945. Hilgardia 16, 429-457.
Jones, H. A., and Clarke, A. E. 1943. Proc. Am. Soc. Hort. S<ci. 43, 189-194.
Jones, H. A., and Davis, G. N. 1944. U.S. Dept. Agr. Tech. Bull. 874.
Jones, H. A., Perry, B. A., and Davis, G. N. 1947. So. Seedsman 10 (3), 13, 57, 60.
Kikuta, K., Hendrix, J. W., and Frazier, W. A. 1945. Hawaii Agr. Expt. Sta. Circ.

24.
King, A. S., Newcomb, G. T., and Chenoweth, O. V. 1943. Oregon Agr. Ext.

Bull. 626.

Lachman, W. H. 1948. Proc. Am. Soc. Hort. Sci. 51, 541-544.
Larson, R. E. 1948. Veg. Grow. Asso. Am. Ann. Rept. f 103-115.
Larson, R. E., and Currence, T. M. 1944. Minnesota Agr. Expt. Sta. Tech. Bull.

164.

Larson, R. E., and Marchant, W. L. 1944. Proc. Am. Soc. Hort. Sci. 45, 341-347.
Larson, R. E., and Paur, S. 1948. Proc. Am. Soc. Hort. Sci. 52, 355-364.
Leach, L. D. 1948. Abst. in Phytopathology 38, 916.
Lewis, M. T. 1949. Pennsylvania Agr. Expt. Sta. Bull. 502. Suppl. 2, 3.
Linn, M. B., and Newhall, A. G. 1948. Phytopathology 38, 218-221.
Lorenz, O. A. 1949. Fertilizer Rev. 24 (2), 5-7, 11-12.

154 J. E. KNOTT AND O. A. LORENZ

Lorenzen, C. 1950. Agr. Eng. 31 (1), 13-15.

McCollam, M. E., and Fullmer, F. S. 1948. Better Crops With Plant Food 32

(6), 6-8, 46-47.

McCool, M. M. 1941. Boyce Thompson Inst. Contrib. 11 (6), 393-401.
Magruder, R., and Wester, R. E. 1948. Seed World 62 (5), 12-14.
Mann, L. K, and MacGillivray, J. H. 1949. California Agr. 3 (10), 9, 13.
Mann, L. K., and Minges, P. A. 1949. Hilgardia 19, 309-337.
Martin, J. A. 1949. South Carolina Agr. Expt. Sta. Unnumbered Mimco.
Merz, A. R. 1940. U.S. Dept. Agr. Circ. 185.
Miller, E. V., Moore, W. D., Schomer, H. A., and Vaughan, E. K. 1949. VJS. Dep:.

Agr. Circ. 805.

Mitchell, J. W. 1947. U.S. Dept. Agr. Yearbook Agr. 1943-1947, 256-266.
Mitchell, K. J. 1945. New Zealand J. Sci. Tech. 27, 287-293.
Mullison, W. R., and Mullison, E. 1948. Botan. Gaz. 109, 501-506.
Munger, H. M. 1942. Proc. Am. Soc. Hort. Sci. 40, 405-410.
Munger, H. M. 1947. New York Agr. Expt. Sta. Farm Res. 13 (3), 1.
Murneek, A. E. 1947. Proc. Am. Soc. Hort. Sci. 50, 254-262.
National Joint Committee on Fertilizer Application. 1948. Natl. Pert. Axsoc.

Pamphlet 149.

Newhall, A. G. 1945. New York Agr. Expt. Sta. Farm Res. 11 (1), 18, 20.
Nixon, P. P., and Smith, G. E. 1949. Proc. Am. Soc. Hort. Sci. 53, 347-348.
Norris, D. O. 1946. Australian Council Sci. Ind. Res. Bull. 202.
Odland, M. L., and Noll, C. J. 1948. Proc. Am. Soc. Hort. Sci. 51, 417-422.
Odland, M. L., and Noll, C. J. 1950. Proc. Am. Soc. Hort. Sci. 55, in press.
Paddock, E. F. 1948. Proc. Am. Soc. Hort. Sci. 52, 365-367.
Park, J. B. 1949. Ohio Agr. Expt. Sta. Farm Home Res. 34, 62-63.
Parris, G. K. 1949. Econ. Botany 3, 193-212.

Peterson, C. E., and Haber, E. S. 1949. Seed World 64 (4), 14, 16.
Powers, W. L. 1945. Botan. Gaz. 106, 247-268.
Pryor, D. E., Whitaker, T. W., and Davis, G. N. 1946. Proc. Am. Soc. Hort. Sci.

47, 347-356.

Rahn, E. M. 1943. Pennsylvania Agr. Expt. Sta. Bull. 443.
Raleigh, G. J. 1948. Proc. Am. Soc. Hort. Sci. 51, 433-436.
Randall, T. E., and Rick, C. M. 1945. Am. J. Botany 32, 560-569.
Randhawa, G. S., and Thompson, H. C. 1948. Proc. Am. Soc. Hort. Sci. 52, 448-

452.

Rick, C. M. 1944. Science 99, 5.43.
Rick, C. M. 1945. Genetics 30, 347-362.
Rick, C. M. 1947. Proc. Am. Soc. Hort. Sci. 50, 273-284.
Rick, C. M. 1948. Hilgardia 18, 599-633.
Rick, C. M. 1949. Proc. Am,. Soc. Hort. Sci. 54, 237-252.
Rick, C. M., and Hanna, G. C. 1943. Am. J. Botany 30, 711-714.
Riollano, A., Adsuar, J., and Rodriquez, A. 1948. Proc. Am. Soc. Hort. Sci. 51,

415-416.

Roever, W. E. 1948. Science 107, 506.
Rosenstein, L. 1936. Shell Chemical Company Bull. 1.
Sawyer, F. G. 1948. Chem. Eng. News 26, 3258-3260.
Sayre, C. B. 1941. New York Agr. Expt. Sta. Farm Res. 7 (2), 1, 13.
Sayre, C. B. 1942. New York Agr. Expt. Sta. Farm Res. 8 (3), 8-10.
Sayre, C. B., and Shafer, J. T. 1944. Proc. Am. Soc. Hort. Sci. 44, 453-456.

VEGETABLE PRODUCTION 155

Sayre, C. B., and Vittum, M. T. 1947. J. Am. Soc. Agron. 39, 153-161.

Scott, D. H., and Riner, M. E. 1946. Proc. Am. Soc. Hort. Sci. 47, 375-377.

Shifriss, O. 1945a. So. Seedsman 8 (2), 15, 26.

Shifriss, O. 1945b. So. Seedsman 8 (4), 15, 26, 30.

Shifriss, O. 1945c. J. Heredity 36, 47-52.

Shifriss, O. 1947. Market Chow. J. 76 (8), 10, 30, 31.

Shirlow, N. S. 1947. New So. Wales Agr. Gaz. 58, 459.

Singleton, W. R. 1948. Connecticut Agr. Expt. Sta. Bull. 518.

Skirm, G. W. 1948. So. Seedsman 11 (11), 14.

Smith, O. 1946. Down to Earth 2 (2), 5-8.

Smith, P. G. 1944. Proc. Am. Soc. Hort. Sci. 44, 413-416.

Smith, P. G. 1949. California Agr. 3 (8), 13-14.

Smith, P. G., and Gardner, M. W. 1950. Phytopathology 40, in press.

Steinberg, R. A. 1937. J. Agr. Research 55, 891-902.

Sweet, R. D., and Havis, J. R. 1948. Proc. Am. Soc. Hort. Sci. 51, 509-514.

Taylor, C. A. 1949. Plant PhysioL 24, 93-102.

Taylor, F. J. 1949. Country Gentleman 119 (4), 19, 87, 89, 90.

Thomas, H. R., and Moore, W. D. 1947. Proc. Am. Soc. Hort. Sci. 49, 264-266.

Thompson, R. C. 1948. So. Seedsman 11 (10), 13, 53.

Twist, G. F. 1949. Food Packer 30 (8), 49-50.

Vogelsang, P., Schupp, A. A., and Reeve, P. A. 1947. Proc. Amer. Soc. Sugar

Beet Technol. 1946, 603-609.

Wade, B. L., 1943. Seed World 53 (5), 12-13, 40-41.
Walker, J. C., and Jolivette, J. P. 1948. U.S. Dept. Agr. Circ. 775.
Walker, J. C., Larson, R. H, Foster, R. E., and Kuntz, J. E. 1948. U.S. Dept. Agr.

Circ. 776.
Walker, J. C., Pound, G. S., and Kuntz, J. E. 1948. Wisconsin Agr. Expt. Sta. Bull.

478.
Waring, E. J., Shirlow, N. S., and Wilson, R. D. 1947. ./. Australian Inst. Agr. Sci.

13, 187-188.

Wanngton, K. 1946. Ann. Applied Biol. 33, 249-254.

Warren, G. F., and Ellis, N. K. 1950. Proc. Am. Soc. Hort. Sci. 55, in press.
Warren, G. F., and Hernandez,. T. P. 1948. Proc. Am. Soc. Hort. Sci. 51, 515-525.
Watts, V. 1947. Proc. Am. Soc. Hort. Sci. 49, 233-234.
Welch, J. E., and Grimball, E. L., Jr. 1947. Science 106, 594.
Went, F. W. 1944. Am. J. Boiavy 31, 135-150.
Went, F. W., and Carter, M. 1948. Am. J. Botany 35, 95-106.
Wester, R. E., and Marth, P. C. 1947. Proc. Am. Soc. Hort. Sci. 49, 315-319.
Wilson, R. D. 1948. J. Australian hist. Agr. Sci. 14, 180-187.
Wilson, R. D., and Waring, E. J. 1948. J. Australian Inst. Agr. Sci. 14, 141-145.
Wittwer, S. H., and Murneek, A. E. 1946. Proc. Am. Soc. Hort. Sci. 47, 285-293.
Wittwer, S. H., Stallworth, H., and Howell, M. J. 1948. Proc. Am. Soc. Hort. Sci.

51, 371-380.

Yarnell, S. H. 1948. Proc. Am. Soc. Hort. Sci. 52, 375-382.
Yeager, A. F. 1949. Market Grow. J. 78 (9), 24.

Young, P. A., and MacArthur, J. W. 1947. Texas Agr. Expt. Sta. Bull. 698.
Zaumeyer, W. J. 1949. Seed World 64 (8), 10, 12.
Zaumeyer, W. J. 1950. Seed World 66 (3), 8, 62-63.
Zink, F. W. 1949a. California Agr. 3 (2), 8.
Zink, F. W. 1949b. California Agr. 3 (6), 15-16.

Prairie Soils of the Upper Mississippi Valley

GUY D. SMITH, W. H. ALLAWAY AND F. F. RIECKEN

U. 8. Department of Agriculture, Ames, Iowa, and the Iowa Agricultural
Experiment Station, Ames, Iowa *

CONTENTS

Page

I. Introduction 157

II. Characteristics of a Modal Prairie Soil 159

III. Variability of Prairie Soils as Functions of Soil-Forming Factors .... 166

1. Biotic Factors 167

2. Climate 173

3. Parent Materials 175

a. Physical Characteristics 175

b. Mineralogical and Chemical Characteristics 176

c. Stratified Materials 176

d. Progressively Changing Materials 178

e. Thickness of Soluiu 180

4. Time 182

5. Topography 186

6. Summary of Concepts of Prairie Soil Genesis 190

IV. Classification of Prairie Soils 192

1. In Higher Categories 192

2. In Lower Categories 194

a. Into Series 194

b. Into Families 194

V. Distribution of the Prairie Soils 196

VI. Crop Yields from Prairie Soils 197

Changes under Cultivation 198

References 203

I. INTRODUCTION

In "Soils and Men/' the 1938 Yearbook of the U.S. Department of
Agriculture, prairie soils were defined as "the zonal group of soils having
a very dark brown or grayish brown surface horizon, grading through
brown soil to the lighter colored parent material at 2 to 5 feet, developed
under tall grasses, in a temperate, relatively humid climate. The term
has a restricted meaning in soil science and is not applied to all dark-
colored soils of the treeless plains but only to those in which carbonates
have not been concentrated in any part of the profile by the soil-forming

* Journal Paper No. J1784 Project 1151 of the Iowa Station.

157

158 GUY D. SMITH, W. H. ALLAWAY AND F. F. RIECKEN

processes." The purpose of this paper is to review the characteristics,
concepts of genesis and the geography of the Prairie soils of the upper
Mississippi Valley and to reexamine the concepts of Prairie soils in rela-
tion to the present system of soil classification.

The concept of Prairie soils was introduced by Marbut (1927) as a
dark-colored soil in whose maturely developed profile no higher per-
centage of lime carbonate is found than in the parent material beneath
and in which either a shifting or an accumulation of sesquioxide or both
has taken place. Marbul grouped the Prairie soils with the Pedalfers,
and in a later publication (1935) described the Prairie soil in more detail.

He said, "the typical Prairie soil profile has a very dark brown or black surface
horizon, or layer, the blackness being caused by the presence of a high percentage of
organic matter. . . . This layer is underlain by a brown horizon which is little if any
heavier in texture than the surface horizon but differs from the surface layer in the
much lower percentage of organic matter and in the brown color. . . . This layer, m
turn, is underlain by parent materials ranging widely in character/'

"The profile just described is the ideal and somewhat theoretical profile. The
soil over most of the area is less dark than the typical soil, is slightly acid at the sur-
face, the colloids in the surface are slightly deflocculated, and very slight eluviution
has taken place. The soil is in the earliest stages of podzolic development. In soil
survey work a soil in the prairies is accepted as a member of the prairie group if the
surface horizon has a well-defined dark color and is 8 or more inches thick. Most of
the Prairie soils are slightly degraded."

Barshad (1946) concluded that the Prairie soils can be defined more
precisely by basing the definition on soil properties which show small
variations between profiles.

He stated, "These are, besides dark brown to black surface color, granular struc-
ture, and absence of lime accumulation in the profile, the predominance of Ca and Mg
ions among the exchangeable bases, the presence of exchangeable H* throughout the
profile, the three-layer type of crystal lattice of the clay minerals, the nature of the
correlation with depth a diffusion pattern of the pH, of the percentage unsatura-
tion, of the C, of the N, and of Jlie C-N ratio, and also the nature of the variation
with depth in the composition of the clay. The common feature with respect to the
clay distribution with depth is that the pattern of the distribution is the result of
extensive clay migration."

The experience of many people working in soil survey over the years
has indicated that Marbut's arid Barshad's concepts of Prairie soils re-
quire some modification. Barshad's requirement that there be a diffusion
type of relation between depth and such properties as pH and the per-
centage unsaturation would eliminate from the Prairie soils a number of
the soils of the midwestern United States which are much more closely
related to that group than to any other. Profile data which show that
the lower part of the A horizon may have a lower base saturation and

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 159

that the B 2 may have a lower pH than either the horizon above or below
are common. Likewise, BarshacPs requirement that the clay minerals
of Prairie soils have a three layer type of crystal lattice may prove
unwise. Soils develop from kaolinitic shales or from other parent mate-
rials which cannot form significant quantities of clays with a three layer
crystal lattice. Under a humid climate and a grass vegetation such soils
might develop all of the other physical and chemical characteristics of
the Prairie soils.

The authors believe in the light of their present knowledge that the
only characteristics common to all the Prairie soils of the United States
are as follows:

1. A dark colored surface horizon, 6 inches or more in thickness in virgin soils
with Munsell color values in the immediate vicinity of 10YR 3/1, 3/2, or 2/2
when moist. The content of organic carbon in available analyses of unculti-
vated samples ranges from about 0.5 per cent to 6 per cent. Organic carbon
contents decrease gradually with depth. Carbon-nitrogen ratios of the surface
horizons are approximately 11 it 2 and decrease gradually with depth to about
7 2 in the lower B horizon.

2. Subsoil colors of brown, yellowish brown or greyish brown predominate, fre-
quently with mottles or incipient gleying.

3. The exchange complex contains IT but usually in smaller amounts than
the combined Ca** and Mg + *. No horizon has yet been found where the
H* exceeded the combined amounts of the bases. Although such soils might
exist, it is doubtful that a soil should be considered a Prairie soil if the H +
greatly exceeds the total Ca* + , Mg"*" 1 " and K*. The pH measurements of Prairie
soils show a range of 4.5 to 7.0 in the surface.

4. Horizons are not sharply separated but have diffuse boundaries. The transi-
tional horizons are usually several inches thick.

Further experience may show this list to be too inclusive.

No general statement concerning the structure, clay distribution, type
of clay mineral present, or presence or absence of a zone of lime accumu-
lation in the Prairie soils can be safely made at the present time. The
reasons why generalizations regarding these properties cannot be made
will be discussed later.

II. CHARACTERISTICS OF A MODAL PRAIRIE SOIL

In presenting the current concept of the Prairie soils it seems desirable
to select some example which may typify the "modal" or "ideal" Prairie
soil and consider the deviations of other Prairie soils from this soil as
functions of the various soil forming factors. For this example the
authors have selected the Tama silt loam as found in central and eastern
Iowa.

However, before entering a detailed description of the Tama series

160 GUY D. SMITH, W. H. ALLAWAY AND F. F. BIECKEN

it is desirable to mention the overall range of the Prairie soils so that
the Tama soils will be seen in proper perspective. The parent materials
of Prairie soils range from sands on one hand to clays on the other.
The thickness of the AI l horizon ranges from about 6 to over 20 inches,

1 Letter designations for horizons used in this paper have the following meanings
as applied to Prairie soils: A. A major surface horizon which is the horizon of
eluviation of clay and/or the horizon of maximum organic matter accumulation.
The Ai horizon is dark in color. The As horizon is transitional to B, but more like
A than B; B. The horizon of illuviation of clay and/or an intermediate horizon be-
tween the A and C differing from them in color and structure. The Bi is transitional
to the A, but more like B than A. The Ba is that part of the B having the greatest
illuviation of clay and may be absent. The Bs is transitional to C, but more like B
than C; C. The parent material which underlies the solum, as distinguished from
accidental substrata. It may be either oxidized and leached or oxidized and un-
leached in Prairie soils.

and the thickness of the solum ranges from about 25 to as much as 100
inches. The pH of the AI horizon ranges from about 4.5 to 7. Where
the distribution of clay is the result of genetic processes, the ratio of
the clay contents of the AI and B 2 horizons ranges from about 1.1 to
about 0.5. The mode is sometimes, but not always, at the center of the
ranges. The modal texture of the parent material is medium textured,
a silt loam or loam, because of the extensive loess deposits and extensive
till deposits of loam texture. The thickness of the AI horizon is about
14 inches and of the solum is about 36 inches. The pH is in the neigh-
borhood of 5 to 6. The modal ratio of clay content of the AI to the
B 2 horizons is about 0.9. While the location of the mode is a geologic
accident, the Tama profile was selected because it lies near the mode of
many properties.

Of the properties listed above the degree of development of the tex-
ture profile, that is the ratio of the percent of clay in the AI and B 2
horizons is considered of outstanding importance from the point of view
of classifying the Prairie soils. As will be pointed out later, this ratio
is believed to be a reflection of the degree of weathering of the soil, and
the variations in a considerable number of other properties are correlated
with variations in the degree of development of the textural profile.

The suggestion of Thorp and Smith (1949) that minimal, medial,
and maximal subgroups be established for the various great soil groups,
based on the relative degree of horizon differentiation, has been adopted
in this paper. In the Prairie soils the development of textural profile is
considered of more significance on the average than any other property
which might be chosen as the basis for subdivision into minimal, medial,
and maximal subgroups. While the subgroups will be discussed in more

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY

161

detail later it may be said in general that the minimal Prairie soils show
no textural difference between A and B horizons, the medial Prairie soils
have a B horizon which is slightly heavier in texture than the A, and
the maximal Prairie soils have a B horizon which is considerably heavier
than the A horizon. From this point of view the Tama profile is a medial
Prairie soil. As was pointed out by Riecken (1945), however, the distri-
bution of soils with different degrees of genetic textural horizon develop-

Ltqtnd

Hi Minimol Prolrit toll*

H Mtdiol Proirit toils

B Maximal Proirit toils

1 Glacial drift

2 Lotss

3 Rtsiduum

Fig. 1. Distribution of minimal, medial and maximal Prairie soils and dominant
parent materials in upper Mississippi Valley.

ment is quite largely a matter of geologic accident. If no additional
glacial deposits had been made following the Kansan glacial age and if no
catastrophic erosion had occurred, the Prairie series chosen as a middle-
of-the-range representative of the group would probably have been what
is now considered a maximal Prairie soil. On the other hand if the
Prairie soils were coextensive with the late Wisconsin glacial deposits
there would probably have been no maximal Prairie soils. The general
distribution of the minimal, medial, and maximal Prairie soils is shown
in Fig. 1.

The selection of the Tama soil to represent the "modal" Prairie soil

162

GUY D. SMITH. W. H. ALLAWAY AND F. F. RIECKEN

has therefore been governed by several factors. In most of its character-
istics, and especially its physical properties, it is modal. Furthermore,
in the mapping history of the Tama soil, it has been somewhat more
specifically defined than has been the case for some other prominent
Prairie soil series, such as Marshall, Carrington or Grundy, which might
have been used. The choice was also favored by the fact that a con-
siderable body of quantitative information concerning the Tama soils
was available to the authors.

brownish

black

surface

moderately
permeable
subsoil

leached
medium
textured loess

Muscatine

Gorwin

Tama

black
surface

moderately
permeable
subsoil

leached
medium
textured loess

dark brown
surface

moderately
permeable
subsoil

leached
medium
textured loess

Prairie
(slightly qleyed)

Wiesenboden
(Humic-qlei)

Prairie

Fig. 2. Relationship of soil characteristics to topography in Tama and Grundy
Counties, Iowa.

In central and eastern Iowa the Tama soils are developed from loess
on gently undulating to rolling upland sites (Fig. 2). The loess is be-
lieved to have been deposited during or following the retreat of the lowan
ice sheet. In the Tama soils area this loess deposit has a maximum
thickness of 125 inches or more. The native vegetation on the Tama
soils was tall grass prairie (chiefly Andropogan furcatus). According to
Kincer et al. (1941) the area has an annual rainfall of about 34 inches
of which about 24 inches comes during the warm season. Average
monthly temperatures range from about 18F. in January to 74F. in
July with a frost free period of about 160 days.

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 163

The morphological characteristics of a Tama silt loam profile
sampled near Gladbrook, Iowa by R. W. Simonson are as follows:

Ai i 0-6" The furrow slice, a mellow, very dark grayish brown (Munsell

color value 10YR 3/2 moist) heavy silt loam with indistinct me-
dium and fine crumb structure. (Nomenclature for soil structure
after Templin et al. 1945). The soil mass is permeated by pin-
holes, and occasional worm holes (1 mm. in diameter). Plant-
roots abundant.

Ai 2 6-12" Very dark grayish brown (10YR 3/2 moist) friable light silty clay

loam with well-developed medium granular and fine crumb struc-
ture. Plant roots and pinholes are abundant, and worm holes
(2-3 mm.) are common. Horizon boundaries are diffuse.

A 8 <fe Bi 12-18" Variegated dark brown (10YR 4/3 moist) and very dark grayish
brown (10YR 3/2 moist) friable light silty clay loam having fairly
distinct medium granular and fine crumb structure. Individual
aggregates are largely one color or the other, but crushed mass is
very dark greyish brown. Plant roots are common and pinholes
abundant. Worm holes passing vertically through the horizon are
common (20-60 per sq. ft.). Horizon boundaries are diffuse.

B 2 i 18-27" Dark brown (10YR 4/3 moist) silty clay loam having occasional
distinct nut -like aggregates (2-4 cm.) and many indistinct frag-
ments. Aggregates crush to fine granules if broken down carefully.
Plant roots and worm holes are common, and pinholes abundant.
Horizon boundaries are diffuse.

Bg 2 27-36" Dark yellowish brown (10YR 4/4 moist) grading into yellowish
brown (10YR 5/4 moist) silty clay loam that digs out in coarse
irregular blocks which break readily to medium granular and
blocky structure. The crushed color is yellowish brown. Plant-
roots are present but not common. Pinholes are common and
worm holes less common than horizon above. Horizon boundaries
are diffuse.

B 3 36-48" Yellowish brown (10YR 5/4 moist) light silty clay loam which

crushes to light yellowish brown (10YR 6/4 moist). Numerous
mottles of reddish yellow (SYR 6/5 moist) and small black iron
concretions are present. Structure is indistinct in places but soil
mass breaks down readily into medium blocky structure when re-
moved. The faces of the blocks are commonly coated with dark
yellowish brown. Plant roots and worm holes are scarce, but pin-
holes are common. Horizon boundaries are diffuse.

C 48" + Very pale brown (10YR 7/3 to 7/4 moist) silt loam with numer-

ous reddish yellow mottles, and occasional small iron concretions.
Structure is similar to horizon above but aggregates are larger. A
few dark yellowish brown coatings are present on surface of ag-
gregates. Plant roots and worm holes are scarce, but pinholes are
common. Grades into calcareous loess between 60 and 100 inches.

The profile described above was situated in a cultivated field on a
slope of about 4 to 5 per cent on the side of a low ridge.

164 GUY D. SMITH, W. H. ALLAWAY AND F. F. RIECKBN

Some of the chemical and physical properties of samples from this
same profile are shown in Table I.

The mechanical analysis of this Tama profile shows a slight accumu-
lation of clay in the B horizon. The zone of clay enrichment is rather
broad and no semblance of an abrupt claypan is evidenced. The sand
fraction as determined in this analysis consists of some concretionary
material.

The carbon and nitrogen contents are highest in the surface and
decrease gradually with depth. In no case does a layer have a higher
content of carbon or nitrogen than the layer above it in the profile.

Calcium is the dominant replaceable cation. The surface layer
showed the lowest replaceable calcium content of any horizon in the
profile. The amount of replaceable magnesium is about one-third of the
replaceable calcium in layers near the surface, and increases to about
one-half the replaceable calcium in the lower layers. In contrast to
calcium and magnesium, the content of replaceable potassium is highest
in the surface layers. The replaceable hydrogen content decreases grad-
ually from the surface downwards.

Replaceable sodium was not determined in this profile, but results
from similar soils indicate that the replaceable sodium content of the
Tama soils would probably not be greater than 0.2 or 0.3 milli equivalents
per 100 g. in any horizon.

The pH value of the surface layer is the lowest in the profile, with
the samples from the middle of the B horizon being slightly more acid
than those from the A 3 or BI. The least acid samples were those from
the C horizon.

The phosphorus content is highest in the C horizons and is at a mini-
mum in the upper B horizon. If the parent material is assumed to have
had a uniform phosphorus content, the present distribution of phosphorus
would indicate some downward movement of phosphorus. The acid
soluble phosphorus in this profile has been studied by Pearson, Spry
and Pierre (1940). They found that the percentage of the total phos-
phorus that is soluble in dilute acid is very low in the surface soil and
in the upper B horizon, but increases markedly in the lower B and in
the C horizons. The nature of the phosphate compounds in the C hori-
zon of this profile has been studied by Stelly and Pierre (1942), who
concluded that a mixture of apatite and aluminum or ferrous phosphates,
with the apatite form predominant, was present in this horizon.

The Tama silt loam profile, for which the description and analysis
are reported above, has not been subjected to total chemical analysis
by fusion procedures. However, Marbut (1935) has reported the total
analysis of a Tama silt loam profile taken just north of Newton, Iowa.

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY

165

^ 2

i-3 d

OQ O

*ts* N

H o

O
(?

2 s

t^ CD

fe a

^ d

f-d-g

5 d w

A 5 *-<

r \QQ CU

__rt 00

!J3 cu

t>. CO 00 VH N CO i>- 00 -<

CO t* t OO 00 OO OO 00 Oi

HC^rH CO. rH Tt< OQ O

CO CO CN rH

CO CO t^ CO !> O CO O 00

OQ CO O Tf| -3J TJ< IO *O TtJ CS|

- -' ---- o - o * o * o

rHCOOOTHCOrH l^ 00 pC^

co Jt**

cocor-jTjjcqoi p oo i-jop ^;

I-H CO CO CO <N* C^" IO rh tQ CO CO

OO O5 t^^

iqcQrHrHpoOCOOitq
^H^'^H^OOOaoi

t>. rH CM Tt< Tt< rH t>.

<N CO CO CO CO CO (M

co, c^ oq o a> 05
-^ -^ ^i co t^ l>

OS O 00 CO CN -*H 00 CM CM O rH

T^ *4 1> io co c? oo t > * co TJ^ co

CM CN rH rH rH rH O O P O O

o o

d
o

1 b -

o r ,

qU-s

_* ni u

ooooo

38888

ooooo

OOOOO

COrH rH 0010

ooooo
pop pp

O O i-H r-H ^H

00 Oi CO <D CO
O5 l^ OQ 1^* Oi

ooooo

CO TH O^ CO Tf

ooooo

?2S?Sg.

rH CN T*" CO CO

ooooo

I
I

166 GUY D. SMITH, W. H. ALLAWAY AND F. F. RlECKEN

The Tama soils of this Newton area are very similar to the Tama soil
described here. The analyses reported by Marbut are shown in Table
II. The samples used in this analysis represent rather broad layers in
the profile and consequently may not reveal certain trends that would
become evident if thinner layers were analyzed individually. Neverthe-
less, there is apparently a downward movement of iron and aluminum
either as free oxides or as a part of the clay. The alkaline earth and
alkali bases were present in rather high amounts. The lower layers were
generally higher in magnesium than the upper layers. The contents of
Ca, K, and Na did not show any pronounced trends in vertical distribu-
tion. The analysis reported by Marbut shows a somewhat higher per-
centage of nitrogen than the analysis shown in Table I. This suggests
that the Tama profile used for Marbut's study may have come from a
more gentle slope than the one used in Table I. The surface layer was
omitted in Marbut's report.

The minerals in the clay fraction of the surface layer of the Tama
profile from Tama County have been studied by Russell and Haddock
(1940). On the basis of differential thermal studies, chemical analysis,
and cation exchange capacity, these workers concluded that the clay
fraction of this sample was dominated by minerals of the montmorillonitc
and illite groups, with very minor amounts of kaolinitc. Peterson (1944)
has also reported thermal studies of certain horizons of a Tama profile
from the same locality. Peterson's work indicates that kaolinite is
probably not a major constituent of the samples (6 to 7 inches and
26 to 36 inches) which he studied. Although the clay mineral data for
this soil profile are quite incomplete it seems likely that the colloids in
this Tama soil are predominantly 2:1 lattice type clay minerals. This
is in accord with the criteria of Barshad (1946) on the types of clay
minerals in Prairie soils. Ross and Hendricks (1945) have also pointed
out that the soils derived from loess and glacial till in this region contain
largely 2:1 lattice type clays.

From studies of soil development on loess (Smith, 1942) it appears
likely that at least half of the clay (<2[x) in the Tama soil has been
formed from coarser material subsequent to the deposition of the loess.

III. VARIABILITY OF PRAIRIE SOILS AS FUNCTIONS OF
SOIL-FORMING FACTORS

The Tama profile described in the preceding pages was selected as an
example of the modal or middle-of-the-range Prairie soils. To describe
the Prairie soils it is also necessary to describe the ways in which other
Prairie soils differ from the Tama profile. Since any profile variations

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY

167

are due to variations of one or more of the soil forming factors, the cur-
rent concepts of the roles of the individual soil forming factors in the
genesis of Prairie soils are presented.

1. Biotic Factors

Throughout most of the prairie region of the Middle West the native
prairies and forests existed side by side in an intricate but orderly
pattern controlled in large part by slope. The forests occupied belts of
strongly sloping lands along the streams and were spreading slowly over

IOWA ORIGINAL FOREST COVER

WISCONSIN
ILLINOIS

LEGEND
i FOREST

MISSOURI

Fig. 3. Original forest cover of Iowa.

the more gently sloping portions of the uplands. They occupied gentler
slopes in the eastern part of the region than in the western part. In
the eastern part of the Prairie region (western Indiana and eastern
Illinois) most slopes in excess of six to eight per cent were forested, and
many gentler slopes on narrow and broad ridges were forested. In the
central part (eastern Iowa), slopes steeper than 8 to 10 per cent were
usually forested as well as some of the gentler slopes on narrow ridges.
In western Iowa slopes in excess of twenty per cent were generally for-
ested only if they faced north or east or were in protected coves. A
similar relationship of slope and vegetation in northern Missouri has
been described by Shrader (1946). The belts of forest were wider in
general to the east of the streams than to the west even though slopes

168 GUY D. SMITH, W. H. ALLAWAY AND F. F. EIECKEN

were comparable. The intricate pattern of prairie and forest for Iowa
is shown in Fig. 3. The heavily forested areas in northeastern and
southeastern Iowa are hilly, and the level or gently sloping broad
ridges commonly had a grass vegetation. Equally hilly areas in western
Iowa along Missouri River were largely prairie, although the forests
have spread rapidly since settlement in any uncultivated fields.

There has been much speculation as to why the grass persisted in a
climate humid enough to support a mixed hardwood forest vegetation.
Prairie fires, grazing by buffalo, soil and moisture conditions, drought
cycles, and the competitive ability of the tall grasses have all been sug-
gested as reasons for the presence of the prairie (McComb and Loomis,
1944). The original prairie consisted chiefly of big bluestem (Andro-
pogan furcatus) with an admixture of a considerable number of other
grasses, legumes, and various other forbs. The grasses commonly
reached a height of 6 to 8 feet, and formed a very tough sod. With the
sod to keep tree seeds from reaching the soil, and the intense competition
for moisture and light between the tall grasses and the tiny seedlings
of the trees, the competitive ability of the grasses must have been very
great. Nevertheless, the soil conditions indicate that in many areas the
trees were slowly invading the prairie prior to settlement. A few sites
still remain where a dense mature forest, chiefly elm, oak and hickory,
is growing on soils which arc indistinguishable in the field from the
adjacent Prairie soils. Starting in such an area and studying the mor-
phological changes in the soil with distance toward the nearest area
of Gray-brown Podzolic soil, it is often found that the changes are more
or less continuous over a distance of one-sixteenth to one-quarter mile.
The first perceptible morphologic change from the Prairie to the Grey-
brown Podzolic soil is either an increase in the degree of development
of the structural aggregates in the B horizon in the well-drained soils or
the appearance of light and then heavy grey coatings in the lower part
of the A horizon of moderartely well-drained soils. The most rapid de-
crease in organic matter seems to occur in the lower part of the A horizon
where the normal differences between Prairie and Gray-brown Podzolic
soils are greatest. The next stage is the development of the Gray-brown
Podzolic AI and A 2 horizons overlying a B horizon which has nearly
the normal color for the Prairie soils. Following this the organic content
decreases in the upper part of the B horizon. The most persistent evi-
dence of a former prairie vegetation is the presence of organic coatings
on the structural aggregates in the lower part of the B horizon. This
transition is shown schematically in Fig. 4. The chemical and invisible
physical changes which occur during the change of a Prairie soil into a
Gray-brown Podzolic soil have not been studied in any detail, nor can

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY

169

any good estimate be made of the time required for the complete trans-
formation, other than that it is at least several hundred and probably
well over a thousand years. The reverse process, the regradation of a
Gray-brown Podzolic soil into a Prairie soil, is theoretically possible so
long as chemical weathering has not been so severe that the primary
calcium bearing minerals have been destroyed. There are extensive
areas of Prairie soils in eastern Iowa and northwestern Illinois which
are now correlated with the Tama series. In these the B horizon has

Direction of forest encroachment
Fig. 4. Schematic diagram of transition from Gray-brown Podzolic to Prairie soil.

both the structure and the moderatc-to-heavy coating of light grey silt
on the structural aggregates which arc normally considered characteristic
of Gray-brown Podzolic soils. No better explanation has been advanced
for these grey coatings than that they are relict characteristics of a
former Gray-brown Podzolic soil which has been converted into a Prairie
soil by the encroachment of grass on the forest. The reasons for a shift
from forest to grass vegetation are not clear unless the change was as-
sociated with the so-called "climatic optimum" discussed by Flint (1947),
a period about 6000 to 4000 years ago when the climate was thought to
have been warmer and drier than that of today.

It is possible to form an opinion as to which of the properties of the
Prairie soils are due to the grass vegetation by comparing adjacent
Prairie and Gray-brown Podzolic soils developed on comparable slopes
from the same parent materials. For this purpose the Tama profile
(Prairie), Table I, may be compared with a profile of Fayette silt loam
(Gray-brown Podzolic) taken in the same county. The data for the
Fayette profile are given in Table III.

170

GUY D. SMITH, W. H. ALLAWAY AND F. F. EIECKEN

o
O

EH o

3 1

H W

.S 03 -H

a ^>

v ^ o <i>

-D
<D

d d

A * ID

Ace &

S S 2 S

oooooooooo

CM. oo

o o o o o o* o o o o

i- o o o o o o* o o o o o* o o o o*

CO CO

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY

171

Study of the two sets of data would indicate that the differences are
ones of degree as much as of kind. The organic carbon content of the
Tama profile is approximately twice that of the Fayette profile. There
is evidence of stronger leaching in the Fayette, for the A2 horizon has a
much higher percentage of exchangeable H. There is also evidence of
either greater clay movement or a more rapid rate of formation and
destruction of clay under forest vegetation, for the Fayette A horizon
is lower in clay and the B higher in clay than the Tama. The ratios
of exchange capacity and clay contents of the two B horizons are very
similar, indicating a similarity in the kind of clay minerals present.
Peterson (1946), using x-ray diffraction, thermal analysis and exchange
capacities was unable to find significant differences between the clay
minerals in the Prairie and Gray-brown Podzolic soils formed from
loess in Iowa.

The differences in amounts and distribution of organic matter have
been explained by Marbut (1928) and others as being the resultant dif-
ferences of organic matter additions on and in the soil. The theory has
been that forest leaf litter decomposes on the surface, while the fibrous
grass roots decompose in the soil. This theory is strengthened by Cline's
(1949) observation that where leaf litter is carried into a nearly neutral
soil by earthworms to decompose in the soil, a thick AI horizon com-
parable to that of the Prairie soils has been formed. Accurate measure-
ments of the comparative return of organic matter by grasses and trees
have not yet been made. While leaf return can be measured, the num-
bers of roots which die and decompose each year have not yet been satis-
factorily measured. The best comparative estimates available are given
in Table IV.

TABLE IV

Estimates of Annual Organic Matter Production by Grass and Trees

Grass
Trees

Roots,

Ibs. per acre

0-6 inches

312-540 '

375-625 b

Tops,
Ibs. per acre

1600-4400
4000-6000 d

Total,
Ibs. per acre

1912-4940
4375-6625

'Thorp (1948) for big bluestem (Andropogan furcatus).
b Scholtes (1940) for shortleaf and loblolly pine.
Jenny (1941) for tall grass Prairie.

d Chandler (1941) estimates of leaf litter from deciduous hardwood forest, multi-
plied by two to allow for woody growth.

The annual additions of organic matter to Prairie and Gray-brown
Podzolic soils are apparently of the same order of magnitude, and the

172 GUY D. SMITH, W. H. ALLAWAY AND F. F. RIECKEN

differences in organic matter content cannot be ascribed to different rates
of organic matter addition. This was the conclusion reached by Jenny
(1941).

Differences in acidity between the Prairie and Gray-brown Podzolic
soils have been advanced as an explanation for the differences in organic
matter. This explanation seems untenable. The vegetation and organic
matter differences of the Chernozem and Gray-wooded soils of Alberta
and Saskatchewan and western Minnesota parallel the relationships
between the Prairie and Gray-brown Podzolic soils of the corn belt. Yet
the Grey-wooded soils have the same reaction profile as the Chernozem?,
according to the data of Newton, Ward and Bentley (1948). If reaction
is the controlling factor it would seem that the Gray-wooded soils should
resemble the Chernozems rather than the Gray-brown Podzolic soils.

Without additional data it is not safe to draw conclusions as to the
reasons for the differences in organic matter contents between the
Prairie and Gray-brown Podzolic soils except that they are related in
some way to the native vegetations.

The differences in the degree of saturation of the Tama and Fayette
profiles are characteristic of the differences found between Prairie and
Gray-brown Podzolic soils where all soil forming factors other than
vegetation are held constant. The lower saturation of the Fayette profile
reflects the greater leaching of the Gray-brown Podzolic soils. Smith
(1942) showed that 50 per cent more carbonates had been leached from
Gray-brown Podzolic soils than from comparable Prairie soils in Illinois.
The greater proportion of exchangeable hydrogen and the greater loss of
carbonates under forest conditions probably represent a combination of
a lower return of bases in plant residues, a higher acidity of the water
entering the soil after passing through the leaf litter, and interception of
water for transpiration of greater depths by the trees. Quantitative
data are not available to permit comparison of the grass and forest vege-
tations on these points.

Since volume weight measurements were not made on the Tama and
Fayette profiles, it is only possible to make a rough comparison of the
total clay contents. From the data available on volume weights in sim-
ilar soils, it would appear that the total amount of clay in the two soils
is almost identical. The significant difference is in the distribution of
the clay, the Tama having more clay in the A and less in the B horizon
than the Fayette. The differences are thought to represent differences in
eluviation under grass and forest, but the reasons for the differences are
unknown.

There are undoubtedly faunal differences among the various Prairie

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 173

soils, but no studies have been reported. Changes in the Prairie soils
induced by man are discussed later.

2. Climate

In considering variations in the properties of Prairie soils that have
been brought about as a result of climate, certain characteristics of the
climate of the upper Mississippi Valley are of prime importance. For
one thing the area shows no regions of rapid change in any of the ordi-
narily measured components of climate. The climate of the region as a
whole is controlled by factors effective over long distances rather than
by local features such as topography or bodies of water.

The climatic range over the area of Prairie soils in the upper Missis-
sippi valley is wide. Some of the more significant climatic features of
weather stations around the periphery of the Prairie soil area are shown
in Table V. Along with this situation of very gradual changes in climate
within the prairie region the parent materials of the area are variable,
with no single uniform parent material covering a broad expanse.

In spite of these difficulties Jenny (1930) has been able to establish
certain relationships between climate and the content of soil organic
matter and nitrogen in this region. These relationships became evident
when large numbers of samples were considered in relation to gradients
in climate. In Jenny's work the organic matter and nitrogen in a wide
variety of soils was found to decrease exponentially with mean annual
temperature, and to increase logarithmically with the humidity factor
(N.S.Q.).

It should be noted in connection with the work done by Jenny on
the effect of temperature upon soil organic matter, that differences in
the age of the parent materials tend to parallel differences in temperature
in most of the regions studied. Through central United States where
the temperature increases from north to south, the northern part of the
area is covered by late Pleistocene deposits, the central part is dominated
by early Pleistocene materials and the soils of south central United
States are predominantly formed from deposits of pre-Pleistocene origin.

A compilation of data available from several sources provides cer-
tain comparisons of the effect of rainfall on the properties of soils de-
veloped from loess under grass vegetation. In Fig. 5 the pH values of
two Chernozem soils and three Prairie soils, developed under varying
amount of annual rainfall from loess of reasonably similar texture are
shown. While the age and mineralogical composition of the loess from
which these soils formed is undoubtedly subject to some variation, it is
felt that it is more nearly uniform than any other parent material cover-
ing a similar range of climatic conditions in the upper Mississippi valley.

174

GUY D. SMITH, W. H. ALLAWAY AND F. F. RIECKEN

TABLE V

Climatic Features of Weather Stations on the Periphery of the
Midwestern Prairie Soil Area *

Precipitation inches

Station and

January

July

Growing

November

April

location in av. temp.,

av. temp.,

season,

through

prairie area

days

March

Oct.

Annual

Columbus, Nebraska

22.3

76.4

160

4.42

23.37

26.79

(western limit)

Alexandria, Minnesota

7.5

69.7

140

3.06

19.06

22.12

(northern limit)

Sedalia, Missouri

29.7

76.9

179

10.55

29.42

39.97

(southern limit)

Lafayette, Indiana

27.5

76.2

168

13.12

24.88

38.00

(eastern limit)

KincereJaZ. (1941).

The topography was also fairly similar for all types, the sites being on
gently rolling upland divides. Since these profiles fall almost on an east-
west line mean temperatures do not differ greatly. The pH values indi-
cate that the two soils developed under the highest rainfall levels are the
most acid, and the two soils developed under the lowest rainfall are the
least acid of the 5 profiles considered. The loess from which the Tama
profile from Illinois developed is apparently higher in calcium carbonate
than the loess from which any of the other profiles were formed. There-

I '

1. Hastings, (Chernozem), Sher-
man Co. Nebr., Ann. Rainfall
23.7"

2. Holdrege, (Chernozem), Phelps
Co. Nebr., Ann. Rainfall 23.2"

3. Marshall, (Prairie), Shelby Co.
Iowa, Ann. Rainfall 29.3"

4. Tama, (Prairie), Tama Co.
Iowa, Ann. Rainfall 33.8"

5. Tama, (Prairie), Menard Co.
Illinois, Ann. Rainfall 34.6"

4.0

5.0

6.0
PH

7.0

8.0

Fig. 5. pH values of selected soils developed from loess under tall grass vegeta-
tion with varying rainfall.

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 175

fore, the differences in parent material may have tended to minimize,
rather than accentuate the differences in pH due to rainfall. Since the
differences are not extreme, the limitations of pH values as criteria for
characterizing soil types must be kept in mind in appraising their
significance.

Although climatic conditions over the north central prairie area of
United States are relatively uniform, it appears that soils with very
similar profile characteristics can be formed under other combinations of
climatic factors. For example, Barshad (1946) has shown that the
Prairie soils of California are very similar to the Prairie soils of the
Midwest, even though climatic conditions vary. The California Prairie
soils are formed under a lower annual rainfall differing in seasonal dis-
tribution from that of the Midwest prairie area. Winter temperatures
tend to be higher and summer temperatures lower in the California
Prairie soil area than in the Midwest prairie area. Since the major part
of the rainfall in the California Prairie soil area comes in the cool, but
frost free, winter months, the effect of this rainfall on soil weathering is
greater than that of the summer rainfall of the Midwest.

3. Parent Materials

a. Physical Characteristics. One of the early important Prairie soil
series was the Marshall series, developed from glacial materials which
included both loess and till materials. This broad series was soon sub-
divided into many other series, at first the division into "loess" and
"drift" soils being expressed. Later the physical character of the loess
or drift was recognized as being very important in influencing the kind
of soil formed. In Illinois, Stauffer (1935) and Wascher and Winters
(1938), on the basis of laboratory and field studies, concluded that the
Prairie soil derived from Wisconsin drift and earlier classified as "Brown
Silt Loam" could be subdivided into the following series on the basis of
the mechanical composition of the till: the Clarence series, derived from
a clayey till, the matrix of which has a high content of Maquoketa shale
whose principal clay mineral is illite; the Saybrook series, derived from
a till much lower in clay; and the Swygert and Elliott series, derived
from tills of intermediate clay contents. As pointed out by Odell (1947)
these series, under similar management, vary in productivity.

According to studies of Walker and Brown (1936) the sandy-textured
Prairie soils accumulate less organic matter than similar soils with silt
loam or silty clay loam surface and subsoil textures. Their analyses of
the surface horizons of 6 soils formerly considered types within a series
are given in Table VI. Presumably the soils in this study had adequate
surface drainage. Therefore it is likely that the differences in organic

176 GUY D. SMITH, W. H. ALLAWAY AND F. F. RIECKEN

matter content were a resultant of better fertility and moisture holding
status of the finer textured soils.

TABLE VI

Average Phosphorus, Nitrogen and Carbon Content of Surface Layers of
Some Prairie Soil Types *

Lbs. per acre
Surface texture (2,000,000 Ibs. of soil)

Phosphorus

Nitrogen

Carbon

Sand

800

550

8,000

Fine sand

750

850

11,500

Sandy loam

864

2,000

25,000

Fine sandy loam

895

2,133

26,380

Loam

1,147

3,756

44,300

Silt loam

1,288

4,606

53,580

* Based on data for soils formerly mapped as types of the Carrington series.

While mechanical composition has been considered important in
classifying Prairie soils into series, other physical properties such as
porosity and compaction must be considered. For example, the Clarion
and Monona soils are minimal Prairie soils without textural profiles.
The Monona soils are developed from nearly sand-free loess having a
volume weight (apparent specific gravity) of 1.3. The Clarion soils
have developed from Wisconsin till having about the same clay content
as the loess but 30 to 50 per cent sand and a volume weight of 1.5. The
difference in volume weight, or compaction, results in a lower aeration
porosity and a lower permeability in the till below the Clarion solum
than in the loess below the Monona solum. The substratum porosity is
important in the design of terrace systems for runoff control. Level
terraces can be used with safety on Monona soils because of the high
permeability of the substratum, but it is doubtful if conventional level
terraces can be safely used on Clarion soils because of the greater com-
paction and lower permeability of the till substratum. Thus, it is neces-
sary to distinguish between the Clarion and Monona series on the basis
of the substratum porosity even though there are other differences of
importance.

b. Mineralogical and Chemical Characteristics. Since the vast ma-
jority of Prairie soils are developed in loess or glacial drift with a very
similar and very highly mixed mineralogical composition, the importance
of mineral composition on soil character is difficult to assess, except in
a few outstanding instances. Although Graham (1943) and Springer

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY

177

(1948) have shown that Prairie soils series vary in minerals important
in plant nutrition, these differences are more likely a result of weathering
intensities than of original differences in the mineral composition.

The principal mineralogical differences are found in soils developed
from water-sorted materials, where high percentages of clay minerals
have been segregated or removed. In such cases covarying factors such
as porosity and permeability are associated with the mineralogy.

The mineralogical composition of the loess and tills of Iowa are es-
sentially the same except for size distribution of the particles according
to Kay and Graham (1941). Nevertheless, according to available data
by Hutton (1948) and Riecken, Allaway, and Smith (1947) the Monona
series developed from loess has a larger amount and probably a higher
supplying power of available K than does the Clarion series developed
from till. This is thought to be due to the larger content of silt sized
minerals in the Monona series than in the Clarion series. In the Monona
profile there is almost no sand size fraction, but over 70 per cent silt size
fraction. In the Clarion profile about 30 to 50 per cent of medium
sized sand is usually present.

c. Stratified Materials. It has long been the practice in the classifi-
cation of soils into series to give considerable emphasis to the geological
origin of the material in the lower horizons. This has been and is the
practice in the Prairie region. In Table VII there are listed several

TABLE VII

Examples of Prairie Soil Series Derived from Mixed (Stratified) Parent Materials

Series

Nature of parent material

Main agricultural significance
of substratum

Sac

Thin well-sorted loess over me- Water storage capacity too low

dium plastic calcareous glacial till. for use of level terraces.

Dodgeville Thin loess over jointed limestone. Soil is droughty and rate of soil-
loss by erosion should be kept
at minimum.

Lacona Thin loess over kaolinitic shale. Permeability and fertility of

shale is low.

O'Neill Thin loess over gravel. Soil is droughty.

Prairie soil series derived from stratified materials. There are many
other such series where the parent material of the upper layers is of
distinctly different character than in the lower horizons. As discussed
by Riecken and Smith (1949a) the solum, including what is commonly

178 GUY D. SMITH, W. H. ALLAWAY AND F. F. KIECKEN

referred to as the A and B horizons, may be formed from one kind of
material, while the substratum is of an entirely different lithology and/or
physical character. In many of these instances, a sample of the parent
material from which the solum has been formed is no longer present.
From the agricultural point of view, it is sometimes quite important to
identify the nature of the material in the substratum below the solum.
As pointed out by Riecken and Smith (1949a), the solums of such
series as the Sac and Galva series is derived from similar material, while
the most significant differences occur below the solum. In most of the
series used for illustration in Table VII, the solums are usually different
in one or more properties, but the differences, such as degree of acidity
or small differences in texture, may not be as important as the character
of the substratum material. It is the intention here to stress the fact
that in practice the character of the substratum material below the solum
is given important weight in classifying the Prairie soils in series when
the substratum lies within the range of roots of common crops and its
characteristics influence the choice of crops, soil management practices,
or the productivity of the soil.

d. Progressively Changing Materials. Where the parent material is
changed progressively by slow deposition, the thickness and nature of
the various horizons are usually somewhat different from those in soils
the parent material of which was deposited at one time. In Fig. 6, there
is illustrated the common occurrence of soils derived from very thick
loess in western Iowa. The Napier and Castana series are derived from
colluvial and/or talus deposited materials. Where rapid recent or mod-
ern deposition has taken place the Hornick series, an Alluvial soil, is
found. In Table VIII, the thickness and relative organic matter content
of the Napier, Hornick and Monona series is compared. The Monona
series, developed on slopes which had slow geologic erosion, has a moder-

TABLE VIII

Comparison of Horizon Thickness and Organic Content of
Napier, Monona and Hornick Series

Series Thickness of
organic layer,
inches

Relative organic
matter
content

Parent
material
changes

Napier (Prairie)
Hornick (Alluvial)
Monona (Prairie)

15-35
absent *
9-12

high
low
medium high

Slow deposition
Rapid deposition
Slow erosion

* Organic content low and horizon of its accumulation not apparent.

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY

179

ately thick organic layer. The Napier series, developed on slopes which
had slow geologic deposition, has a much thicker organic layer with a
higher organic matter content throughout. The higher organic matter
content is considered due in part to the somewhat less well-aerated but
more favorable moisture conditions. The thicker organic layer is con-
sidered to be due to slow deposition of new parent material on the A
horizon of the soil, with the organic matter accumulating again in the
new material. In the Napier series, the texture profile is no more de-

maximal
Lithosol

minimal
Lithosol

minimal
Prairie

Fig. 6. Relationship of soil characteristics to topography and parent material in
western Iowa.

veloped than in the Monona profile, and Hutton (1948) has shown that
the latter has no textural B horizon. It would seem that new parent
material in the Napier series is added at a more rapid rate than the clay
can form or accumulate to develop a textural B horizon.

Where truncation occurs after soil formation has been initiated, the
soil profile often acquires some characteristics not common to soils where
truncation has not occurred. Iron oxide mottlings of various hues of
yellow and strong brown are present in the middle solum of the Shelby
series, as recently defined in the Livingston County, Missouri, Soil Sur-
vey. The Shelby is one of the Prairie soils derived from Kansan and/or

180 GUY D. SMITH, W. H. ALLAWAY AND F. F. BIECKEN

Nebraskan glacial till. Such mottlings in soils often are interpreted
as being caused by poor internal aeration. In the Shelby series, however,
these mottlings are more probably a relict mottling from a former profile
of weathering. In certain areas of southeastern Iowa and western
Illinois where loess-mantled Illinoisan glacial deposits occur, post-loess
dissection of the landscape may be the cause of the presence of mottlings
in the solum of some Prairie series developed in the loess. Through trun-
cation iron oxide mottlings developed in deeply buried materials are
brought close to the surface and tend to persist in any newly developed
solum.

e. Thickness of Solum. In Fig. 7, the thickness of the solums of 4
representative Prairie soil series are shown graphically. The thickness
of the solum increases with increasing coarseness of texture and/or per-
meability of the parent material. The thickness of the solums of the 4
series is 25, 30, 48 and 90 inches. Stauffer (1935) studied the Clarence,
Elliott and Saybrook series in some detail. He found that the CaC0 3
equivalent of the unleached till from which each of these soils was de-
rived is about 20 per cent. The till from which the Clarence soil was de-
rived contains about 50 per cent 5 micron clay, and the till from which
the Elliott soil was derived contains about 40 per cent 5 micron clay.
The thickness of the solum of the Clarence is about 25 inches at which
depth primary carbonates were found. The solum of the Elliott is about
30 inches thick, with primary carbonates at 30 inches. The solum of
the Saybrook series is about 36 inches thick. It w r as formed from loam-
textured till of about 16 per cent 5 micron clay content, with CaC0 3
equivalent of about 20 per cent. The till of these three series is of the
same geological age, occurring intermixed on the same ground moraines
in the Tazewell drift area in northeastern and east central Illinois. There-
fore, the thickness of the solum of the Clarence, Elliott and Saybrook
series is related to the texture and the permeability of the till.

In Fig. 7, the Muscatine series, with a solum of about 48 inches, is
derived from a calcareous loess containing about 80 per cent silt, and the
Thurman series, with a solum of about 90 inches, is formed from cal-
careous medium-to-fine eolian sand.

These examples indicate that the thickness of the solum of Prairie
soils depends in part upon the character of the parent material. Of the
many properties of parent material affecting solum thickness, perhaps
the most important is the porosity of the material, although other factors
such as lime content, height of ground water,, slope characteristics and
time of weathering are also important. For the series represented in

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY

181

Fig. 7, the porosity of the parent materials has been the most important
factor in causing differences in thickness of solum.

For the Thurman series, illustrated by the thickest solum in Fig. 7,
the B horizon is considered to be truly genetic. Studies of the colloid
extracted from the so-called B horizon of the Thurman soil indicate that
it is made up of 2:1 lattice clays with about 8 per cent free iron oxide.
The fine clay 0.2[i) has an exchange capacity of 64 meq. per 100 g.
Except for the fact that this colloid is higher in free iron oxide it seems

DEPTH

Increasing particle size of parent material

Clay Silty clay loam

Soil Series Clarence Elliott

Sit loam
Muscat ine

Sand
Thurmon

Fig. 7. Relation of particle size to solum thickness.

to be very similar to the colloid extracted from B horizons of other
Prairie soils. The clay layers found at about 72 inches depth tend to
parallel the surface, cut across bedding planes, and are unrelated to the
water table level. The clay is therefore considered by the authors to
have formed in the solum and accumulated in layers as illustrated in
Fig. 7.

It is of interest to speculate on the factors causing accumulation of
the clay in thin layers at depths of 60 to 80 inches in the solum of the
Thurman series. This clay is chiefly less than 0.2[i in size, and because
the water table is low and the pores large it would likely have migrated
still deeper in the absence of a flocculating agent. The clay contains
about 8 per cent free iron oxides as determined by the method of Jef-
fries (1946), while the clay in the B 2 layers of other Prairie soils such
as the Monona and Seymour series contains from 215 to 6 per cent free

182 GUY D. SMITH, W. H. ALLAWAY AND F. F. EIECKEN

iron oxides, according to Hutton (1948). The authors consider that
the clay in the Thurman is probably flocculated by the iron oxides. It
is likely that the concentration of iron oxides above 60 or 70 inches is
too low to cause flocculation at shallower depths, and that sufficient con-
centration of iron oxides is reached only at depths of 60 to 80 inches.

4. Time

In considering the variations in Prairie soils that have resulted from
differences in the time during which soil development has taken place,
it must be emphasized that the majority of the Prairie soils are developed
from relatively young parent materials. The vast majority if not all of
the Prairie soils of the upper Mississippi Valley have been formed since
the Wisconsin glaciations, which occurred late in the Pleistocene period.
This fact was pointed out by Norton (1933) when he attributed the
existence of most of the Prairie soils to the youth and unweathered con-
dition of the parent material. Because of this it is difficult to find groups
of Prairie soils which form a chrono-sequence, in Jenny's (1946) terminol-
ogy, covering an extended time interval.

Perhaps the best example of chrono-sequences in the Prairie great
soil group are found in soils formed from loess of varying thickness, but
deposited at the same time and from the same source. The difference in
the "effective age" of such sequences of soils has been pointed out by
Smith (1942) as being due to the differences in the age of that portion
of the loess from which the solum has developed. Smith compared two
"end members" of such a sequence, the Jasper silt loam (now called
Richview) and the Tama silt loam, both of which had developed under
grass vegetation on sites with good surface drainage. The Jasper soil,
representing the older member of the sequence, showed a more highly
developed textural B horizon, had a lower percentage base saturation,
and was lower in organic carbon than was the Tama soil, the young
member of the sequence. It was concluded from these observations that
the Prairie soils are not in equilibrium with their environment and that
the direction of their development is toward the Planosols.

Another study of such a sequence of Prairie soils developed from
varying thicknesses of loess has been reported by Hutton (1947). The
tendency toward increased differentiation of textural horizons, accom-
panied by an increase in replaceable hydrogen was also evident in this
work.

These trends in the development of Prairie soils had been recognized
by Marbut (1928) who pointed out that saturation with bases decreases
with age in the Prairie soils. He also indicated that an increase in the

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 183

degree of eluviation and illuviation should coincide with this decrease in
percentage base saturation.

A detailed study of the formation and movement of clay in one Prairie
soil, the Grundy silt loam of Missouri, has been reported by Haseman
and Marshall (1945). They used the zircon in the soil as an inert and
immobile reference material for the calculation of changes in other con-
stituents. By this criterion the Grundy soil was developed from loess
which was essentially uniform to a depth of 69 inches. The clay content
of the whole profile was found to have been increased about 20 per cent
by the processes of soil formation. Definite evidence of movement of
clay in the profile was obtained. The A horizon (0-17 inches) had lost
66 per cent of the clay it contained originally. This represents a mini-
mum amount of translocation since some clay was formed in this horizon
during the development of the profile. On the other hand, the clay
content of certain layers in the B horizon had approximately doubled
during the development of the profile. Along with the formation and
movement of clay the various layers in the profile had changed in volume.
The A horizon had shrunk but the B horizon showed marked swelling,
for a net increase in volume in the entire profile of about 8 per cent.
An increase in weight of about 6 per cent for the whole profile was at-
tributed to oxidation and hydration of the minerals and accumulation
of organic matter.

Investigations of the actual mechanism by which differentiation of
textural horizons takes place in Prairie soils are rather limited. How-
ever, this same process as it occurs in the level soils associated with the
Prairie soils of Illinois has been investigated in detail by Bray (1934,
1935). It is highly probably that the mechanism involved in the for-
mation of claypans in these soils is nearly the same as that occurring
in the Prairie soils developed on sites with somewhat better surface
drainage. The process as outlined by Bray is essentially a formation
of 2:1 lattice type clay in both A and B horizons and the translocation
of a part of the clay formed in the surface layer to the B horizon. The
formation of clay from coarser materials is inaugurated before the profile
section is leached free of calcium carbonate, and, in the Illinois soils
studied, the total amount of fine clay in the profile reached a maximum
while the soil was only slightly acid. The movement of clay from the
surface was attributed to transport by drainage waters moving through
channels in the soil mass in turbulent flow. Deposition of clay in the
B horizon was attributed to a decrease in the rate of flow of the per-
colating water.

In a laboratory sttidy of claypan formation, Smith (1934) found
evidence that deposition of clays in the B horizon may be brought about

GUY D. SMITH, W. H. ALLAWAY AND F. F. BIECKEN

by the flocculation of the clay by electrolytes in the ground water or by
iron oxide colloids carrying a charge opposite that of the clay. It seems
likely that this second effect, namely the flocculation of clays by iron
oxides, may tend to stabilize the colloids in the surface layers of well
drained soils, and contributes to the fact that claypans form more slowly
in soils with good surface drainage than in soils developed on level sites.
In Button's (1947) work it was found that among the older soils the
heaviest layer in the B horizon tended to occur closer to the surface as
the soils became more highly weathered. The amount of 2 (A clay in
the heaviest layer increased regularly as the soils became more highly
weathered. These observations might be interpreted as indicating that
the B horizon develops by a sieve action, with mobile clay from the A
horizon being stopped by the impenetrable B horizon. Or, it might be
interpreted as representing the influence of a finer textured parent mate-
rial which was discussed earlier.

Bray (1937) concluded that the clay formed in the Illinois soils he
studied was largely "superfine" clay, that is clay with particles having
equivalent diameter less than 0.06 jx. One process proposed for the for-
mation of the material consisted of "flaking off" of particles from the
edge of weathering mica, along with a loss of K from the particles to
form a beidellite-nontronite type of mineral. As weathering progressed
the Si0 2 percentage in the coarse clay increased through loss of other
constituents. Decomposition of the fine clay was apparently not signifi-
cant until the soils had reached stages of profile development which would
exceed those included in the Prairie great soil group.

Most of the mineralogical investigations of Prairie soil colloids, such
as those by Hendricks and Fry (1930), Russell and Haddock (1940),
Peterson (1944), Hutton (1948), and Barshad (1946) have shown that
2:1 lattice type minerals dominate the clay fractions of these soils.
Minor amounts of quartz and iron oxides are usually present. Barshad
concluded from his studies and from a consideration of other work that
the dominance of 2:1 lattice type clays in Prairie soils could be used as
a criterion for defining this great soil group. Exceptions to this rule,
although rare, merit some consideration. For example, Alexander, Hen-
dricks, and Nelson (1939) estimated that kaolinite comprised almost
one-half of the material finer than 0.3 [x in the B 3 horizon of a Prairie
soil (Carrington) from Iowa. Subsequent investigations by the authors
indicate that kaolinitic clays are an important constituent of the coarse
2.0 0.2 [x clay in several Carrington profiles. However, since kaolinite
is present in the relatively unweathered till at the base of these profiles
it appears that it is not a product of soil-forming processes. Nevertheless,
it may be preferable in defining the Prairie soils to require that only the

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 185

clay formed during soil development be dominantly 2:1 lattice clay
minerals, and to exclude from the definition any clay inherited from the
parent material.

A weathering sequence of clay sized minerals has recently been pro-
posed by Jackson et al. (1948). A consideration of available information
on the clays in Prairie soils indicates that stages 6 through 9 (quartz,
illite, mica intermediates, montmorillonite) are most common. The
absence of major amounts of stage 5 minerals (albite and other feldspars)
is indicated by the low amounts of nonreplaceable Na and Ca ordinarily
found in Prairie soil colloids (Marbut, 1935, and Hutton, 1948). Aside
from instances in which the parent material contained kaolinite, such
as mentioned above, there is little information pointing to kaolinite in
Prairie soils. In fact, several studies (Whiteside and Marshall, 1944,
and Larson et a/., 1946) of Planosols which may be considered post-
Prairie soils have failed to indicate that major amounts of stage 10
minerals (1:1 lattice clays) have been formed even at these advanced
stages of profile development. In accord with the ideas of Jackson et al.
(1948) concerning the effect of particle size upon rate of weathering, the
coarse clay (2 to 0.2 [A) of Prairie soils tends to be largely the quartz
and illite minerals of stages 6 and 7 in their weathering sequence, whereas,
the fine clay tends to be dominated by mica-intermediates and mont-
morillonites of stages 8 and 9. An apparent deviation from the weather-
ing sequence has been found to occur in the coarse clay fractions of sur-
face horizons of older Prairie soils and Planosols by several investigators
(Whiteside and Marshall, 1944; Bray, 1934; Larson et al., 1946). In
these instances quartz seems to become dominant in the coarse clay
(2-0.2 (A) of such soils. This may be due to the fact that since the rate of
solution of quartz is a function of its particle size, quartz in the coarse
clay fraction is more resistant, relative to other minerals, than is quartz in
the fine clay fraction. Another factor that must be kept in mind is that
under some conditions certain minerals may break down to finer particle
sizes and be eluviated from the surface horizons at an early stage of
weathering, whereas under conditions where no clay movement was pos-
sible these minerals might persist in the surface horizons for a greater
period of time. In instances where redistribution of clay is possible the
entire profile must be considered in evaluating the stage of weathering.

To summarize the effects of time and the course of weathering in Prairie soils,
some of the principal processes may be listed as follows:

1. The removal of free carbonates from the profile is inaugurated before the soil
has reached a stage of development that would entitle it to be considered as a Prairie
soil, and this process is completed while the profile is still in the minimal or medial
stage of Prairie soil development.

186 GUY D. SMITH, W. H. ALLAWAY AND F. F. BIECKEN

2. The removal of bases from the profile is inaugurated in very young Prairie
soils and continues throughout all stages of Prairie soil development.

3. The amount of organic matter in the profile of a Prairie soil increases during
the early stages of profile development, reaches a maximum, and then declines at
advanced stages of profile development.

4. The formation of 2:1 lattice type clays is started in young Prairie soils and
may in some cases continue through most of the stages of profile development within
the ranges of Prairie soils. In cases where a maximum clay content within the
profile is reached at an earlier stage of profile development, this maximum may be
due to a balance between clay formation and clay decomposition.

5. The redistribution of clay within the profile starts in the medial Prairie soils
and continues through and beyond the maximal Prairie soils.

6. The decomposition of primary minerals, especially the feldspars, takes place
throughout all stages of Prairie soil development.

5. Topography

The early concepts of Prairie soils implied that these soils should be
found on gently undulating upland sites. In such sites the removal of
soil by erosion under natural conditions was supposed to balance the
effects of weathering and leaching and result in a "steady state" which
was considered to be a normal soil in equilibrium with its environment.
In light of the discussion of the effects of time on the genesis of Prairie
soils as given in earlier sections of this paper it appears that only a few
Prairie soils are actually found in such a steady state ; most of them seem to
be tending to develop into Planosols. The effect of topography on the
Prairie soils is most evident as it regulates the relative rate at which
different segments of the landscape advance through the Prairie stage
of soil development. Variations in the amount of water entering and
moving through the soil profile account for much of the effect of topog-
raphy upon the rate of soil weathering.

The role of topography on the development of the various portions
of a Prairie soil landscape developed from a single parent material is
affected by a number of factors. Among these are, degree of slope gradi-
ent, length of slope, direction of slope, curvature of slope, nature of
parent material, age of soil, and position of permanent free water table.

While the effects of topography are evident in nearly all soil charac-
teristics, the distribution of organic matter and clay in the profile can
be used to typify many of the changes taking place as a result of topo-
graphic variations.

A detailed study of the accumulation of nitrogen in the profiles of
some minimal Prairie soils and Lithosols as affected by the direction,
length and percentage of slope has been reported by Aandahl (1948).
A series of profiles arranged on traverses across hills in a virgin prairie
area in western Iowa was investigated. The parent material was deep,

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY

187

silty loess, containing about 5 per cent calcium carbonate equivalent
and being quite permeable to water. The permanent water table was
not a factor in limiting the downward movement of water. The general
findings of this study are shown in Fig. 8. The effects of length and
direction of slope upon accumulation of nitrogen and leaching of CaC0 3
are evident from this figure. Profiles from the lower parts of west facing
slopes were higher in nitrogen and more deeply leached than were soils
near the top of these slopes. In fact, profiles on the crest of such slopes

01%

t=i

i i i

ill I i I I

AVERAGE PERCENT NITROGEN OF THE 3-24 INCH LAYER

SLOPE
14%

Fig. 8. Per cent nitrogen profiles and average per cent nitrogen of the 3-24 inch
layer of virgin soil profiles on different slope positions of the Ida-Monona soil area
of Ida County, Iowa.

showed essentially no leaching of CaCO 3 . On east facing slopes the
CaCO 3 was leached to greater depths even though the slope gradient was
greater. This general relationship is shown in Fig. 6. Observations of
the vegetation on virgin areas in this region indicate that east slopes are
universally more densely covered by grasses or brush than are west and
south facing slopes. This may be due to the protection from hot winds
and to partial shading afforded by east facing slopes.

The relation of topography to the distribution of soil types in a prairie
area having somewhat different geomorphological features is also shown
in Fig. 2. This figure shows the distribution of the Tama, Muscatine and
Garwin soils developed from a relatively thin deposit of sjlty loess over
lowan till in Tama County, Iowa. Slopes in the area are longer and not
as steep as in the Monona-Ida area considered by Aandahl (1948). The
Tama soils described earlier in this paper are found on undulating topog-

188 GUY D. SMITH, W. H. ALLAWAY AND F. F. BIECKEN

raphy, especially near the convex crests of long slopes. The Muscatine
soils, which have thicker A horizons and somewhat more mottled subsoils
are found on very gently sloping divides and near the base of long slopes.
The Garwin soils, which are deep black soils with highly mottled subsoils
(Wiesenboden or Humic-glei soils) usually occupy flats, depressional
areas and drainageways. The differences in organic matter and clay
distribution between Tama and Muscatine are shown in Table IX.

TABLE IX
Comparison of Organic Carbon and Clay Contents of Tama and Muscatine Soils *

Tama P-27, 4 per cent slope Muscatine P-32, b 1 per cent slope

Depth, Organic carbon, Clay, Organic carbon, Clay,

inches per cent per cent per cent per cent

0-6

2.86

27.5

3.38

28.2

6-9

2.44

3.29

9-12

1.98

31.7

2.67

34.2

12-15

1.74

2.35

15-18

1.46

32.4

2.04

33.9

18-21

1.12

1.84

21-24

0.91

34.0

1.76

36.7

24-27

0.71

1.36

27-30

0.59

34.2

0.82

38.9

"Samples collected and determinations made under the supervision of Roy W.
Simonson.

b Muscatine profile sampled is a borderline profile transitional to Garwin.

With increased weathering a different effect of slope is seen. In
southeastern Iowa where maximal Prairie soils are extensive, the slopes
of 1 per cent are occupied by the Edina series, a Planosol developed
under grass, while slopes of 4 per cent arc occupied by the Seymour series,
a maximal Prairie soil. Data on these soils show that the Seymour soils
have a higher content of organic carbon than do the gentler sloping Edina
soils. The clay content of the Edina subsoil is higher than is that of the
Seymour.

Smith (1941) pointed out the changing relationships between Prairie
and Wiesenboden soils on the one hand and Planosols on the other hand
as the soils are subjected to increased weathering. These relationships
are shown in Fig. 9. In young parent materials, minimal Prairie soils
are likely to occupy level areas, gentle slopes, the base of long slopes and
protected coves, while steeper and more exposed areas are occupied by
Lithosols. Planosols are very rare and are found only in distinct depres-
sions over a low water table. At somewhat more advanced stages the

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY

189

level areas are occupied by very deep Prairie soils or Wiesenboden soils.
The more sloping sites, which in more youthful situations were occupied
by Lithosols, are now occupied by Prairie soils. The subsoils of these
Prairie soils are only slightly mottled. Planosols are found in slight
depressions and are more common than in the young area (Muscatine

Slope

Erosion during
development

illiliffll
Wiesenboden Prairie

Plonosol Prairie-- Wiesenboden
Planosol Planosol
transition transition

Fig. 9. Relationship between topography and soil character in southwest central
Illinois: Menard, Sangamon, Christian, and Fayettc Counties.

and Ipava catenas, Fig, 9). In areas where the soils are still older and
more highly developed, the Prairie soils on the more sloping sites are
found to have mottled subsoils and to show marked accumulation of
clay in the B horizons. Planosols occupy level areas as well as depres-
sions, and may even be found on gentle slopes (Harrison catena, Fig. 9).
Thus it appears that as the soils grow older, the percentage of the land-
scape covered by Planosols shows a steady increase at the expense of

190 GUY D. SMITH, W. H. ALLAWAY AND F. F. RIECKEN

Prairie and Wiesenboden soils. With continued weathering the Planosols
tend to occupy all areas which have not been rejuvenated by geologic
erosion, or where a high water table has not inhibited the translocation
of clay (Cowden and Cisne catenas, Fig. 9). In general, geologic erosion
has rarely been a factor on slopes of less than about 7 per cent. In the
eastern part of the Prairie region slopes of much over 7 per cent have
usually been forested. In the western part of the Prairie region slopes
from about 7 to 15 per cent are usually occupied by Prairie soils, but
slopes much in excess of 15 per cent have suffered such rapid geologic
erosion that the areas are occupied by Lithosols rather than Prairie soils.
The effects of topography upon soil development are also dependent
upon the nature of the parent material and the position of the water
table. In general wherever the water table is low, porous parent materials
tend to minimize the effects of slope gradients upon the soil profiles,
whereas impervious parent materials accentuate the effects of topography.

6. Summary of Concepts of Prairie Soil Genesis

Since the soil is the resultant product of the interaction of all the
various soil forming factors, it is not possible to isolate and study the
influence of any one factor except in a frame of reference of all the other
factors. Given one set of conditions a particular soil forming factor may
produce effects in one direction, while under a slightly different set of
conditions it may produce effects in the opposite direction. For example,
it was shown that the relations between slope and organic matter are
sometimes reversed as the soil becomes more weathered.

The Prairie soils of the upper Mississippi Valley can be considered
to have developed under relatively uniform climatic and biotic influences.
They have developed from a variety of parent materials ranging from
an occasional sand deposit high in quartz through the dominant calcare-
ous alluvial and glacial sediments of highly mixed mineralogical com-
position to the occasional residuum from illitic and kaolinitic shales.
Before they became Prairie soils they were Lithosols or Alluvial soils,
and possibly some may have been young Gray-brown Podzolic soils. In
most years some water moved down through the soil on its way to under-
ground outlets. Further, before they could become Prairie soils some
organic matter had to accumulate, and carbonates had to be removed
from the surface horizons. By this time the formation of 2:1 lattice clay
minerals had begun. With increased time the carbonate zone moved
downward and the AI horizon thickened until it attained the normal
thickness for the AI of a Prairie soil. This is now considered the maximal
stage of the Lithosol although there is some reason to classify this as the
pre-minimal stage of the Prairie soils. With still more time the carbonate

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 191

zone is separated from the zone of organic matter accumulation and the
soil enters the minimal stage of the Prairie soil. The time required to
reach this stage will vary with the slope factors, the parent material and
the climate.

In the minimal stage the clay content of the soil is nearly uniform
with depth. It is thought that this means that the clay has been largely
formed in place during and shortly after the removal of the carbonates.
The primary feldspars have not been strongly weathered, and appreci-
able amounts of calcium and sodium are being released. The micas and
illites are releasing potash rapidly. Nevertheless, exchangeable hydro-
gen is present in small amounts in the exchange complex.

With additional weathering evidence of actual clay movement can
be found, and the soil enters the medial stage. In general only the fine
clay (<0.2 (JL) appears to move in the soil. Movement is believed to
start as a result of turbulent flow of water. Movement of the clay stops
either when the downward movement of the water is stopped when a
fine pore is reached, or when the clay particles are flocculated by elec-
trolytes or by positively charged hydrated iron oxide particles. In fine
textured materials the clay accumulates at shallow depths (18 to 24
inches or less). In coarse textured materials (sands) the clay moves
to depths of five or more feet very commonly, and is believed to stop
as a result of the flocculating effect of hydrated iron oxides.

During the medial stage the organic matter content appears to reach
a maximum. Weathering of the silt size calcium feldspars in the A
horizon is largely but not entirely completed. Sodium and potassium
feldspars are the least weathered of the feldspars. With continued leach-
ing and reduced release of bases, base saturation of the surface horizon
drops to the neighborhood of 50 to 70 per cent, but saturation in the
lower horizons remains high. While there is little positive evidence, it
seems probable that clay formation continues in both A and B horizons,
possibly at a reduced rate, and that decomposition of some of the clays
begins.

From a geological point of view, the time required to pass through
the medial stage into the maximal stage is often short, not much more
than the time since the lowan glacial age. In the western part of the
Prairie region there seems to be a narrow range of slopes where slow
but continuous erosion will prolong the medial stage. This range seems
to lie in the neighborhood of 7 to 15 per cent slopes. In the eastern part
of the Prairie region slopes steep enough for erosion to be active under
grass sod have usually been invaded by forests.

With continued leaching, and accompanying formation and move-
ment of clay in medium and fine textured materials, a sharp textural

192 GUY D. SMITH, W. H. ALLAWAY AND F. F. EIECKEN

contrast is developed between the A and B horizons, and the transitional
horizons (A 3 and BI) are narrowed to an inch or two. The pores in the
B horizon are clogged with clay and conditions are favorable for occa-
sional waterlogging and gleying. There is considerable doubt that this
stage will be reached if the matrix is coarse textured or if the solum
is underlain at shallow depths with coarse textured materials. Under
these conditions it appears that there may actually be a complete re-
moval of a considerable portion of the clay either as clay or as decom-
position products.

At this stage in the finer materials, when gleying occurs, the soils are
grading into Planosols and are considered maximal Prairie soils. They
are considered to become Planosols when the B horizon becomes so
slowly permeable that an intermittent water table develops above the
B horizon and produces a distinct bleached A 2 horizon with base satura-
tion well below 50 per cent.

In the coarse materials, where the solum remains too permeable for
gleying, the direction of development has not been studied. However,
the reddish colored B horizons of such soils suggest that they may be-
come Reddish Prairie soils. The classification of such borderline soils
will have to await more precise definition of the Reddish Prairie soils.

The Prairie soils of the upper Mississippi Valley are relatively young,
with few exceptions either being formed in materials deposited during
the Wisconsin glacial age or formed from older deposits exposed by
erosion during or since Wisconsin time. The Prairie soils would have
only minor extent in the Mississippi Valley were it not for the extensive
recent glacial deposits.

IV. CLASSIFICATION OF PRAIRIE SOILS
1. In Higher Categories

Marbut's (1927) classification of all soils into Pedalfers and Pedocals
centered attention on the presence or absence of a zone of CaCO 3 accu-
mulation as the differentiating criterion between Prairie soils and Cher-
nozems. The analytical data available to Marbut indicated that this
was a valid criterion. Jenny and Leonard (1934) studied the depth to
the layer of carbonate concretions along a transect extending from
Colorado through Kansas to Missouri. They noted an increase in the
depth to the zone of lime accumulation with increase in rainfall but
reported the presence of zones of lime accumulation in the Prairie zone
under rainfall between 35 and 40 inches. In general with the higher
rainfall the zone of lime accumulation lay somewhere between 60 and
120 inches.

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 193

Factors other than rainfall also appear to influence the presence or
absence of a zone of lime accumulation as well as the depth to the con-
cretionary zone. The lime content of the original parent material is,
within limits, as important as is rainfall. Soils developed in non-
calcareous materials or in materials having very low lime contents are
apt to lack a zone of CaCOa accumulation. Very sandy soils such as
the Thurman which straddles the Chernozem-Prairie zone lack a horizon
of lime accumulation according to present information. On the other
hand the Clarion soils found in central Iowa and the Clarence soils
found in eastern Illinois have carbonates at depths as shallow as 25 to
35 inches. In a Clarion profile sampled near Ames, Iowa, the upper few
inches of the carbonate zone were white and had a calcium carbonate
equivalent of approximately 39 per cent. The till directly below this
had a calcium carbonate equivalent of only 19 per cent. A Clarence
profile reported by Stauffer (1935) from eastern Illinois showed some
evidence of a zone of calcium carbonate accumulation. The Saybrook
and Elliott soils from eastern Illinois reported by Stauffer showed
definite evidence of having a zone of lime accumulation at a depth of
about 4 to 5 feet, having both a maximum of secondary and of total
carbonates, followed by a decrease with depth.

The age of the soil is another factor which can influence the presence
or absence of a zone of CaCO 3 accumulation or its depth. Because of
cyclical or seasonal fluctuations there are years when the rainfall is more
than 50 per cent above average. For example, the mean annual rainfall
for Nebraska is given as 22.3 inches but in 1905 the rainfall was 31.5
inches and in 1915, 35.6 inches, according to the 1941 Yearbook of the
U. S. Department of Agriculture. Given enough of such seasons of above
average rainfall the soils will tend to take on some of the characteristics
associated with the most humid periods. Most of the Chernozem soils
of northern United States and Canada, such as the Prairie soils, have
developed since the start of Wisconsin time and are relatively young soils.

Viewed as a whole the soils of the Chernozem region more commonly
have a zone of CaC0 3 accumulation than do the soils of the Prairie
region, and the depth to the zone is shallower. Nevertheless, because of
the lack of a zone of lime accumulation in some of the soils in the Cher-
nozem zone and the presence of the zone of lime accumulation in some
of the Prairie soils at the eastern or most humid limit of the Prairie
region, there appears to be reason to question the validity of the separa-
tion of Prairie and Chernozem soils as different zonal great soil groups
if the zone of CaCOs accumulation is the differentiating criterion. No
other criterion has been suggested which would be more valid.

When Baldwin et al. (1938) established great soil groups for the

194 GUY D. SMITH, W. H. ALLAWAY AND F. F. RIECKEN

intra-zonal soils associated with the Prairie soils it became important
to develop criteria for distinguishing between the Prairie soils and the
other great soil groups. Riecken (1945) pointed out the necessity for
considering borderline series in defining a great soil group. If the Prairie
great soil group is too narrowly defined either new great soil groups must
be established for the borderline series or they will be left outside any
group. He pointed out that incipient gleying must be permitted in those
Prairie series which constitute intergrades with Wiesenboden (Humic-
glei) soils. He also pointed out the Prairie group must accommodate
intergrades with the Planosols which have mottled heavy textured genetic
B horizons. The Muscatine and Seymour series were selected as examples
of Prairie soils which are intergrades with the Wiesenboden (Humic-
glei) and Planosols respectively.

Riecken (1945) also pointed out that it is difficult to establish hard
and fast rules for drawing the boundaries between different great soil
groups and that different men are apt to draw the boundaries in different
places. However, once the criteria are agreed upon all series having
similar features should be grouped readily into the same great soil group.
For example, if it is decided that the Muscatine series is to be classed
as a Prairie soil it follows that the Floyd, Mahaska, Ipava and Lisbon
series must also be classed as Prairie soils. The problems involved in
determining the borderline between the Prairie soils on the one hand
and the Chernozem, Wiesenboden (Humic-glei), Planosol, Lithosol, Al-
luvial soil, Reddish Prairie, and Gray-brown Podzolic soils on the other
hand cannot be discussed in detail in this paper, but will be discussed
elsewhere.

8. In Lower Categories

a. Into Series. In classifying the Prairie soils into different series,
the 10 fundamental features of the soil profile listed by Marbut (1927)
are taken into consideration. *

In a recent review of the concept of the series, Riecken and Smith
(1949a) have emphasized that at times the character of the substratum
material is important in classifying soils into series. This was discussed
earlier. Perhaps, therefore, the thickness of soil profile significant to
agriculture should be added to the list of fundamental soil features. It
is also possible that certain physical properties such as porosity and
permeability should also be added to the list.

For the Midwest, more than 120 different Prairie soil series have been
recognized up to the present. Some of these series, such as the Clarion,
occupy an extensive area, whereas others, such as the Minden may be

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY

195

quite minor in extent. No attempt will be made here to list all the
Prairie series recognized.

Phases of soil types, as discussed by Riecken and Smith (1949a),
have properties of the profile within the defined range of the soil type
and series. For Prairie soils, the most common phases mapped are those
indicating slope and accelerated erosion. Generally, the number of
phases mapped is dependent on the degree of detail needed for the
objective in mind.

6. Into Families. The concept of family accepted here is essentially
that outlined by Baldwin, Kellogg and Thorp (1938), namely, that a
family is a category in soil classification between series and great soil
group and composed of one or more distinct series having similar profiles.
As indicated by Riecken and Smith (1949a), little published information
is available for guidance in grouping of series into families.

As the family category is intermediate in generalization between the
Great Soil Group and series, it is obvious that not all of the fundamental
soil features can be used if a large number of families is to be avoided.
Some of the features which seem to be important enough to warrant
consideration as family criteria are: thickness of the A horizon and
of the solum, degree of gleying, texture and mineralogy of the solum,
and degree of horizon development. Examples of possible families with
the major morphologic criteria are given in Table X.

TABLE X

Examples of Families, with Major Morphologic Criteria

Family

Major morphologic criteria

Napier
Gravity

Tama

Muscatine

Grundy

Monona

Thurman

Clarence

Ai thick, 20 inches or more ; slight B development ; silty ; not gleyed.

Ai thick, 20 inches or more; slight B development; silty; slightly

gleyed.

Average Ai medium B development; silty; not gleyed.

Average Ai medium B development; silty; slightly gleyed.

Average Ai strong B development; silty clay; slightly gleyed.

Average Ai slight B development ; silty ; not gleyed.

Average Ai thick A 3 -Bi ; slight B development ; sand ; not gleyed.

Average Ai medium B development; clay; slightly gleyed.

In classifying Prairie soils at categorical levels between the great
soil group and the series, there are two other considerations. One is
whether the nomenclature should be connotative or abstract, and the
other is the number of categories needed between the great soil group
series to express the varjpu features. If the' soil family i$ to be

196 GUY D. SMITH, W. H. ALLAWAY AND F. F. RIECKEN

used for making statements about the agriculture, the tendency would
be to have a rather large number of families. However, if the family
category is to be used chiefly to express fundamental soil characteristics,
few families would be needed. After further study it might be decided
that two categories would be more useful, so that agricultural predictions
could be made at the lower level.

V. DISTRIBUTION OF THE PRAIRIE SOILS

Because of limitations of scale, it is not possible to give here the
details of the current nomenclature of the Prairie Soil Series in map
form. Soil association maps showing the general distribution of Prairie
Series have been published for Indiana by Bushnell (1944), for Illinois
by the University of Illinois Agricultural Experiment Station (1949),
for Wisconsin by Muckenhirn and Dahlstrand (1947), for Minnesota
by McMiller (1947), for Iowa by Riecken and Smith (1949b), for north-
ern Missouri by Shrader (1946), and for Kansas and Nebraska by Thorp
et al. (1948). A revised association map for the United States is in
preparation by the Division of Soil Survey, Bureau of Plant Industry,
Soils, and Agricultural Engineering.

Fig. 1 is a generalized map showing the major areas of occurrence
of Prairie soils in the Middle West compiled from the sources listed
above. The boundaries of the Prairie soils and the original prairies were
not the same. There were large areas of prairie in northeastern Missouri
and southern Illinois, where the soils are considered Planosols because
they have a low base saturation, a distinctly bleached A 2 horizon, and
an abrupt boundary between the A horizon and the clay pan.

While Fig. 1 cannot show detailed soil association areas, it is possible
to show in a general way the main groups of parent materials according
to origin, and the regional distribution of the minimal, medial and maxi-
mal Prairie soils. The dominance of loess and glacial drift will be noted.
Where more than one kind of parent material is important, the most
extensive is listed first.

The classification into minimal, medial and maximal Prairie soils is
based on the profile characteristics of the upland Prairie soils having
slopes from about 3 to 8 per cent. Obviously in the region of maximal
Prairie soils, there are younger and less developed soils on some of the
low terraces and some of the steeper slopes. The map therefore shows
only the degree of horizon differentiation for the most extensive upland
soils on slopes ranging from 3 to 8 per cent. There is some question yet
about the proper classification of some of the areas called maximal in
central Kansas.

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 197

The boundaries on the north and east between the Prairie and Gray-
brown Podzolic soils, while highly generalized, are sharp, and the transi-
tions are usually not more than one-quarter mile in width. The
boundaries between the Prairie soils, Reddish Prairie soils and Planosols
on the south are gradational, and the transition takes place over a dis-
tance of 10 to 20 miles or more. The boundary to the west between the
Chernozem and Prairie soils has been discussed elsewhere. The boundary
as drawn represents the common presence or absence of a distinct zone
of lime accumulation in soils developed since Late Wisconsin time from
permeable highly calcareous parent materials having slope gradients of
two to seven per cent with a low water table. With these restrictions the
transition still occupies a very broad zone, perhaps more than 200 miles
in width, and may possibly be a separation of little or no validity.

VI. CROP YIELDS FROM PRAIRIE SOILS

The outstanding agricultural characteristic of the Prairie soils has
been the high yields of grain crops produced without use of fertilizers
although the use of fertilizers has been increasing rapidly in recent years.
The period of exploitive agriculture of the Prairie soils appears to be
drawing to a close. While continued cultivation will in time produce
mineral nutrient deficiencies which could cause declines in yields, the
first hundred years of cultivation have not witnessed any appreciable
decline in average yields. Declines in the amount of available nitrogen
released through decomposition of the soil organic matter have been
compensated for by increased growth of legumes, improved varieties, and
improved cultural methods. For the period 1940-1944, census data shows
that average acre yields of corn (maize) for some whole townships (36
square miles) have been as high as 70 bushels on some of the best Prairie
soils. These yields were reported in Tama and Grundy Counties, Iowa,
where Tama and Muscatine soils are dominant. During the same period,
corn yields in comparable townships in Lyon County, Iowa, on the border
between Prairie and Chernozems soils (Moody soils) were 50 bushels
per acre, or 20 bushels less. In these townships, areas of Intrazonal and
Azonal soils are minor in extent and probably have had little influence
on the average yields.

Yields from specific soils have been reported by Odell (1947). He
reports that individual farmers on Tama-Muscatine soils have obtained
acre corn yields as high as 85 to 90 bushels for the whole period 1937-
1944. These are probably about the top farm yields obtained over a
period of years in the corn belt. They were secured by farmers using
essentially a corn-corn-oats-clover or corn-corn-oats-clover-clover rota-

198 GUY D. SMITH, W. H. ALLAWAY AND F. F. BIECKEN

tion with applications of barnyard manure and limestone, but without
significant amounts of commercial fertilizers. While the highest farm
yields exceed these values in favorable years, no farm is known to the
authors where harvested yields for the entire planted acreage have aver-
aged as much as 100 bushels of corn per acre over a 5-year period. Corn
yields from the best Wiesenboden (Humic-glei) and Alluvial soils may
approach those of the most productive Prairie soils, but probably do not
exceed them.

While the most productive Prairie soils under the best of the prevailing
systems of management have been producing the corn yields mentioned
above, average yields on the other Prairie soils have been considerably
lower. Odell (1947) reported farm yields of only 33 bushels of corn
from Hagener loamy sand, a Prairie soil developed from eolian sand, and
29 bushels from Clarence silt loam and Rowe silty clay, Prairie and
Wiesenboden (Humic-glei) soils respectively developed from till having
a clay texture. These yields were secured by farmers who had only one-
eighth of their land in legumes each year. Obviously corn yields are
still lower if the Prairie soils are truncated by erosion and no provision
is made for supplying nitrogen to the corn.

The range in average farm yields for the 1937-1944 period for crops
other than corn reported by Odell (1947) from farms on Prairie soils
are as follows: soybeans, 11 to 24 bushels per acre; winter wheat, 16 to
28 bushels; oats, 19 to 49 bushels; spring barley, 27 to 39 bushels; alfalfa
hay, 2.0 to 2.8 tons. These figures cover the range between the poorer
soils under about average management to the better soils under better
than average management. Each figure is an average of the yields
obtained by a number of farmers so the extreme range is somewhat
greater than the figures given.

With respect to productivity, the Tama series, cited earlier as a modal
Prairie soil, lies well above the middle of the range of the Prairie soils
and probably is somewhat "higher in productivity than modal because
of its higher rainfall.

Changes under Cultivation

A few studies have been made of the changes which have occurred
in the Prairie soils during cultivation. In most cases the virgin site was
represented by a field which had never been plowed, but which had been
used for meadow or pasture. Samples from this field were compared
with samples from an adjacent cultivated field.

In reviewing the literature it is often difficult if not impossible to be
sure exactly what has happened. Not only is there always the possibility
of original differences in the soils compared, but erpsjoji ma y h$ve re-

PRAIHIE SOILS OF THE UPPER MISSISSIPPI VALLEY 199

moved layer of unknown thickness from the cultivated soils. In that
case different layers are compared when) the surface 6 inches are studied.
Organic matter may be lost thrcmgli decompositions iiu the cultivated fields
or through erosion, and the effects should be separated if possible.

Whiteside and Smith (1941) compared virgin and cultivated samples
of Flanagan silt loam, a moderately well drained medial Prairie soil
developed from thin loess over Wisconsin till of Tazewell age. The
cultivated field was first plowed in the late 1850's. No fertilizers were
ever added. The cultivated field showed a loss of 32 per cent of the
original 3.66 per cent of organic carbon in the surface 5 inches, and
15 per cent loss in the 5 to 12-inch layer. No differences were observed
below 12 inches. About 30 per cent of the exchangeable bases in the up-
per foot were lost, but there was little change in the percentage of satura-
tion, apparently because the loss of organic matter reduced the exchange
capacity. No change was found in mechanical composition.

Ulrich (1949) compared the surfaces of virgin and cultivated samples
of Minden silt loam, a medial Prairie soil developed from thick loess in
western Iowa. He found reductions of 33 per cent in the original 3.34
per cent carbon, and 31 per cent in nitrogen in the surface 6 inches. The
volume weight of the surface had increased from 0.96 to 1.19 and aeration
porosity had been reduced by about one-third.

There was no evidence of erosion from either the Flanagan or Minden
soils nor was there much opportunity because they were nearly level.
The loss of one-third of the original organic matter in these studies must
have been primarily by decomposition and oxidation. The loss represents
a movement toward a new equilibrium where the level of organic carbon
will be determined largely by the rotation, the amount and quality of
crop residues returned, and the amount of manure added.

Anderson (1949) showed nitrogen decreases of 31 and 36 per cent in
the surface 6 inches of Marshall and Grundy soils, medial and maximal
Prairie soils, developed from loess in southwestern Iowa. In the 6- to
12-inch layers of these soils he found nitrogen decreases of 21 and 35
per cent, respectively. However, the mechanical analysis of the Grundy
soil indicates that some erosion had occurred on the cultivated soil.

Anderson found no consistent change in pH, but there was a consistent
decrease in exchangeable potassium in the surface 6 inches of the culti-
vated soils and in the percentage of large (greater than 0.25 mm.)
aggregates. He found decreases in aeration porosity in the Prairie soils,
the decrease ranging from 4 per cent of the original aeration porosity
in a soil developed from glacial drift to 58 per cent in the Marshall
samples.

Rost and Rowles (1940) in Minnesota have reported comparisons or

200 GUY D. SMITH, W. H. ALLAWAY AND F. F. RIECKEN

6 cultivated Carrington silt loam and 6 Tama silt loam surfaces with
an equal number of samples from pastures which were presumed to have
been uncultivated. They showed a 12 per cent loss of the original 2.66
per cent organic carbon in the upper surface of Carrington silt loam and
a 5 per cent loss of the organic carbon in the lower A. In Tama silt
loam they found that 13 per cent of the original 2.84 per cent organic
carbon had been lost from the upper A (0 to 4 or to 6 inches) but that
no change had occurred below 4 or 6 inches. The mechanical analyses
of the samples suggest that some slight erosion of the cultivated fields
had taken place.

Hide and Metzger (1939) studied the effects of cultivation on some
of the Prairie soils of Kansas. Twenty sites were selected where com-
parisons could be made between virgin soils, and soils which had been
cultivated for 30 years or more across the slope and up and down the
slope. The differences represent the combined effects of loss of organic
matter by erosion and by accelerated decomposition. On the average
they found a loss of 37 per cent of the carbon and 32 per cent of the
nitrogen in the surface layer where cultivation was across the slope.
Where cultivation was up and down the slope, 44 per cent of the carbon
and 37 per cent of the nitrogen had been lost from the surface. Of the
soils studied, 4 were on slopes of 10 per cent or more, 7 were on slopes
of 5 to 10 per cent and 9 were on slopes of 4 per cent or less. The greatest
total loss occurred on soils having the greatest total amount of organic
carbon, but the percentage losses were not related to total amounts.

Several studies have been made of long time experiments to deter-
mine the effects of rotations and soil treatments on the Prairie soils.
From these studies it is evident that the organic matter content of culti-
vated Prairie soils must still be declining. The level at which the organic
matter will be eventually stabilized appears to depend upon the rotations
used, the applications of crop residues and manure, the degree to which
erosion is controlled and to a less extent the application of fertilizers.

The Morrow plots in Illinois, established in 1876, appear to offer the
best opportunity to measure long time changes in organic matter as in-
fluenced by rotations and soil treatment. On these plots it is possible
to compare the effects of continuous corn since 1876 with rotations of
corn and oats since 1876, and of corn, oats and clover since 1901.
The corn, oats and clover plots from 1876 to 1901 were in a rotation of
corn, corn, oats, meadow (3 years) .

Each of the 3 plots has been split into a north and south half. The
north half has had no soil treatment and all crop residues have been
removed. The south half of each plot has received animal manure
applications ahead of the corn equal in weight to the dry weight of all

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 201

crops produced. In addition the south half of each plot has received
applications of lime and phosphate.

The organic carbon was reported by DeTurk et al. (1927) for samples
collected in 1913 and 1923. Stauffer et al. (1940) reported data for
samples collected in 1938. These data are given in Table XI.

TABLE XI

Organic Carbon Content of Surface Layer of Morrow Plots as
Influenced by Crop Rotation *

Year

Continuous corn

Corn-oats

Corn-oats-clover

sampled

MLP b

North

South

North

South

North

South

1913

2.11

2.48

2.37

2.64

2.63

2.92

1923

1.94

2.35

2.21

2.68

2.57

2.88

1938

1.74

2.09

2.14

2.44

2.28

3.35

* 1913 and 1923 samples were 6% inches; 1938 sample was 6 inches.
b Barnyard manure, lime and phosphate.

The soil type on the Morrow plots has not been determined by the
Illinois Experiment Station because it does not exactly fit any established
type, but it is a very close relative of the Flanagan series studied by
Whiteside and Smith (1941). The organic carbon content of the virgin
Flanagan silt loam was 3.66 per cent for the surface 5 inches. The Mor-
row plots could not have been greatly different from this value. By
the time the first samples were collected in 1913 the experiment had
been under way for 36 years and the effects of the different rotations
and treatments were already apparent. The next 25 years of the experi-
ment showed a further decline in the organic carbon for every rotation
without treatment. With treatment, only the corn-oats-clover rotation
failed to show a decline. However, in every case the decline during the
25 year period was less than it must have been in the preceding period.
The low carbon value of the continuous corn plots, according to Stauffer
et al. (1940), has been influenced in part by erosion despite the gentle
slope of one to two per cent. They also point out that the sod borders
of the plots which have been in Kentucky bluegrass (Poa pratensis) sod
since 1904 have an organic carbon content of 3.2 per cent for the surface
6 inches. It would appear that a rotation of corn, oats and clover plus
manure, lime and phosphate is at least as effective in maintaining the
organic carbon content of the Prairie soils as is bluegrass.

Peevy, Smith, and Brown (1940) report similar general conclusions
from the study of long time rotation and manurial treatments in Iowa

202 GUY D. SMITH, W.* H. ALLAWAY AND F. F. RIBCKEN

on Clarion loam, a minimal Prairie soil. Dodge and Jones (1948) also
drew similar conclusions from a study of rotation and manurial experi-
ments on a Prairie soil, Geary silty clay loam in Kansas.

The effect of erosion induced by cultivation of the Prairie soils is to
remove organic matter. Thus, when erosion is active the Prairie soils
tend to lose their principal distinguishing characteristic and the farmers
lose one of their principal assets. While it is possible in some instances
to determine with reasonable accuracy the thickness of the layer removed
by erosion, in other instances it is not now possible to make even a rea-
sonable estimate. Until the relation of the slope factors to organic matter
content is known no valid estimates of the effect of erosion can be made.
Figure 8 taken from Aandahl (1948) illustrates conditions in a virgin
Prairie area. It was pointed out earlier that slope gradient is not the
only factor affecting the topsoil thickness. The curvature, length and
direction of the slope at times appear more important than gradient.
Estimates of erosion based on the assumption that the original topsoil
thickness was uniform or varied only with slope gradient have little
validity over areas of much extent. Such estimates have a pronounced
tendency to overestimate the amount of erosion which has taken place.

The influence of erosion on yields of crops, principally corn, has re-
ceived some attention, but again the conclusions are clouded by the diffi-
culties of determining the amount of loss, and in controlling such factors
as shape of slope, slope gradient, and soil management, all of which may
have independent effects on yields.

Where erosion does not bring to the surface material with greatly
different physical and chemical properties the principal effect of sheet
erosion is to reduce the nitrogen supplying power of the soil. Experi-
ments with uesurfaced plots on Tama silt loam at Dixon, Illinois, accord-
ing to Bauer et al. (1945), have shown that a rotation including legumes
plus fertilizers enabled the Tama subsoil to produce as much corn as the
soil with its original AI horkon but without fertilizers. Yields of alfalfa
hay were not greatly affected by desurfacing. Oat yields were reduced
but no nitrogen was reported to have been applied to the oats. It seems
probable that the effects of erosion on minimal and medial Prairie soils
developed in medium textured parent materials can be largely overcome
by use of fertilizers. The main damage therefore lies in increased costs
of production, or in reduced yields if proper rotations and fertilizers are
not used.

The maximal Prairie soils, and the minimal or medial Prairie soils
developed in heavy textured parent materials, present a different picture.
The physical properties of the B horizons of such soils make cultivation
difficult. Seedbeds are difficult to prepare, and aeration is poor. Con-

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 203

sequently the loss of the AI horizon of such soils would seriously reduce
future crop yields even though fertilizers are used. Future study may
show methods of overcoming these handicaps but no methods are known
now. The areas of maximal Prairie soils shown in Fig. 1 are therefore
the principal areas where erosion probably will permanently reduce the
productive capacity. Areas of soils developed from heavy textured par-
ent materials can also be permanently damaged by erosion. Except for
northeastern Illinois such areas are usually of only local importance.
Subsequent to the completion of this manuscript, and the setting of the type, the
authors agreed to propose the name "Brunigra" as a substitute for "Prairie soils."
The need for changing the name of Prairie soils has been apparent for many years
because of the confusion between the meaning of "Prairie soils," a great soil group,
and "prairie soils," referring to all soils developed under grass vegetation. The long
established usage of the latter term gives it precedence. Delay in coining a new
name seemed wise until it could be established that there were distinct enough dif-
ferences between Prairie soils and Chernozems to warrant two great soil groups.
The relation between the Brunigra soils and Chernozems will be discussed in more
detail in a subsequent paper.

REFERENCES

Aandahl, A. R. 1948. Proc. Soil Sci. Soc. Am. 13, 449-454.

Alexander, L. T., Hendricks, S. B., and Nelson, R. A. 1939. Soil Sci. 48, 273-279.

Anderson, M. A. 1949. M. S. Thesis, Iowa State College.

Baldwin, M., Kellogg, C. E., and Thorp, J. 1938. U.S. Dept. Agr. Yearbook "Soils

and Men! 1 pp. 979-1001.
Barshad, I. 1946. Soil Sci. 61, 423-442.

Bauer, F. C. et al. 1945. Univ. Illinois Agr. Expt. Sta. Bull 516, 162.
Bray, R. H. 1934. Am. Soil Survey Assoc. Bull. 15, 58-65.
Bray, R. H. 1935. Am. Soil Survey Assoc. BuU. 16, 70-75.
Bray, R. H. 1937. Soil Sci. 43, 1-14.

Bushnell, T. M. 1944. Purdue Univ. Agr. Expt. Sta. Special Circ. 1.
Chandler, R. F., Jr. 1941. /. Am. Soc. Agron. 33, 859-871.
Cline, M. C. 1949. Soil Sci. 68, 259-272.
DeTurk, E. E., Bauer, F. C., and Smith, L. H. 1927. Univ. Illinois Agr. Expt. Sta.

Bull. 300.

Dodge, D. A., and Jones, H. E. 1948. /. Am. Soc. Agron. 40, 778-785.
Flint, R. F. 1947. Glacial Geology and the Pleistocene Epoch. Wiley, New York,

pp. 487-499.

Graham, E. R. 1943. Soil Sci. 55, 263-273.

Haseman, J. F., and Marshall, C. E. 1945. Missouri Agr. Expt. Sta. Bull. 387.
Hendricks, S. B., and Fry, W. H. 1930. Soil Sci. 29, 457-478.
Hide, J. C., and Metzger, W. H. 1939. J. Am. Soc. Agron. 31, 625-632.
Button, C. E. 1947. Proc. Soil Sci. Soc. Am. 12, 424-431.
Hutton, C. E. 1948. Ph. D. Thesis, Iowa State College.
Jackson, M. L., Tyler, S. A., Willis, A. L., Bourbeau, G. A. and Pennington, R. P.

1948. /. Phys. Chem. 52, 1237-1260.
Jeffries, C. D. 1946. Soil Sci. Soc. Am. Proc. 11, 211-212.

204 GUY D. SMITH, W. H. ALLAWAY AND F. F. BIECKEN

Jenny, H. 1930. Missouri Ayr. Expt. Sta. Research Bull. 152.

Jenny, H. 1941. Factors in Soil Formation. McGraw-Hill, New York.

Jenny, H. 1946. Soil Set. 61, 375-394.

Jenny, H., and Leonard, C. D. 1934. Soil Sci. 38, 363-381.

Kay, G. F., and Graham, J. B. 1941. loiw Geol Survey Annual Rept. 38, 183.

Kincer et al 1941. VS. Dept. Agr. Yearbook. Climate and Man. 685-1200.

Larson, W. E., Allaway, W. H., and Rhoades, H. F. 1946. Soi7 Set. Soc. Am. Proc.

11, 443-447.

McComb, A. L., and Loornis, W. E. 1944. Bull. Torrey Botan. Club 71, 41-76.

McMiller, P. R. 1947. Minnesota Agr. Expt. Sta. Bull. 392.

Marbut, C. F. 1927. Proc. 1st Intern. Cong. Soil Sci. Com. V, 1-31.

Marbut, C. F. 1928. Lectures to the Graduate School of the U.S. Dept. Agr.

(mimeo.).

Marbut, C. F. 1935. Atlas of Am. Agr. U.S. Govt. Printing Office, pp. 62-69.
Muckenhirn, R. J., and Dahlstrand, N. P. 1947. Soils of Wisconsin (leaflet).
Newton, J. D., Ward, A. S, and Bentley, C. F. 1948. Univ. Alberta Bull. 21.
Norton, E. A. 1933. Am. Soil Survey Assoc. Bull. 14, 40-42.
Odell, R. T. 1947. Univ. Illinois Agr. Expt. Sta. Bull. 522.

Pearson, R. W., Spry, R., and Pierre, W. H. 1940. J. Am. Soc. Agron. 32, 683-695.
Peevy, W. J., Smith, F. B., and Brown, P. E. 1940. J. Am. Soc. Agron. 32, 739-753.
Peterson, J. B. 1944. Soi7 Set. Soc. Am. Proc. 9, 37-48.
Peterson, J. B. 1946. Soil Sci. 61, 465-475.
Riecken, F. F. 1945. Sot7 Sci. Soc. Am. Proc. 10, 319-325.
Riecken, F. F., Allaway, W. H., and Smith, G. D. 1947. Soil Sci. Soc. Am. Proc.

12, 432-440.

Riecken, F. F., and Smith, G. D. 1949a. Soil Sci. 67, 107-116.

Riecken, F. F., and Smith, G. D. 1949b. Principal Upland Soils of Iowa. Iowa

State College Agr. Expt. Sta. (mimeo.).

Ross, C. S., and Hendricks, S. B. 1945. U.S. Geol. Survey Prof. Paper 205B
Rost, C, O., and Rowles, C. A. 1940. Soi7 Sci. Soc. Am. Proc. 5, 421-433.
Russell, M. B., and Haddock, J. L. 1940. Sot7 Sci. Soc. Am. Proc. 5, 90-94.
Scholtes, W. H. 1940. M.S. Thesis, Duke University.
Shrader, W, D. 1946. Soil Sci. Soc. Am. Proc. 11, 458-463.
Smith, G. D. 1934. Missouri Agr. Expt. Sta. Research Bull. 210.
Smith, G. D. 1941. Soil Sci. Soc. Am. Proc. 6, 78-82.
Smith, G. D. 1942. Univ. Illinois Agr. Expt. Sta. Bull. 490.
Springer, M. E. 1948. Soil Sci. Soc. Am. Proc. 13, 461-467.
Stauffer, R. S. 1935. J. Am. Soc. Agron. 27, 885-894.
Stauffer, R. S., Muckenhirn, R. J., and Odell, R. T. 1940. J. Am. Soc. Agron. 32,

819-832.

Stelly, M., and Pierre, W. H. 1942. Soil Sci. Soc. Am. Proc. 7, 139-147.
Templin, E. H., Hearn, W. E., Lyford, W., O'Neal, A. M., and Roberts, R. C. 1945.

Soi7 Sci. Am. Proc. 10, 479-481.

Thorp, J. 1948. U.S. Dept. Agr. Yearbook "Grass," pp. 55-66.
Thorp, J., and Smith, G. D. 1949. Soil Sci. 67, 117-126.
Thorp, J., Williams, B. H., and Watkins, W. I. 1948. Soil Sci. Soc. Am. Proc. 13,

438-445.
Univ. of Illinois Agr. Expt. Sta. 1949. Principal Soil Association Areas of Illinois.

(mimeo.).
Ulrich, R. 1949. Ph. D. Thesis, Iowa State College.

PRAIRIE SOILS OF THE UPPER MISSISSIPPI VALLEY 205

U.S. Dept. of Agr. Yearbook. 1938. Soils and Men, p. 1174.

Walker, R. H., and Brown, P. E. 1936. Iowa Agr. Expt. Sta. Bull. 203.

Wascher, H., and Winters, E. 1938. Am. J. Sci. 35, 14-21.

Whiteside, E. P., and Marshall, C. E. 1944. Missouri Agr. Expt. Sta. Research

Bull. 386.
Whiteside, E. P., and Smith, R. S. 1941. J. Am. Soc. Agron. 33, 765-777.

Ladino Clover

GILBERT H. AHLGREN AND R. F. FUELLEMAN

Rutgers University, New Brunswick, New Jersey, and University of
Illinois, Urbana, Illinois

CONTENTS

Page

I. History and Distribution 208

1. History 208

2. Distribution 209

II. Characteristics and Adaptation 209

1. Description 209

2. Adaptation 210

a. Cold Resistance 1 210

b. Slimmer Temperatures and Moisture 212

c. Soil Conditions 212

3. Breeding and Genetics 212

4. Diseases and Insects 213

III. Establishment and Management 213

1. Effect of Lime and Soil pH 213

2. Fertilizing New Seedings 214

3. Fertilizer Tests 214

4. Topdressing Established Stands 216

5. Suitable Grass Companions 217

a. In the Northeast 217

b. The Southeast 218

c. The Midwest 219

d. The West 219

6. Seeding Practices 220

a. Shallow Planting 220

b. Inoculation 220

c. Seeding Dates and Companion Crops 221

d. Seeding Methods 222

e. Seed Characteristics 222

7. Use in Renovation 223

8. Management 223

a. Care of the New Seeding 224

b. Rotational Grazing 224

c. Ladino Pasture Yields 225

d. Carrying Capacity of Ladino Pastures 225

e. Fall Grazing 225

IV. Utilization 226

1. Pasture 226

2. Hay and Silage 227

207

208 GILBERT H. AHLGREN AND R. F. FUELLEMAN

3. Seed Production 227

4. Other Uses 229

V. Summary 230

References 230

I. HISTORY AND DISTRIBUTION

Seed of ladino clover was first planted in the United States in the
spring of 1891 at the North Carolina Experiment Station at Raleigh.
McCarthy and Emery (1894) described it as follows: "This is an im-
proved variety of the common white clover. Our seed came from
Fratelli Ingegnoli, Milan, Italy. The plant is much more robust and has
larger leaves than the common species, but produces very little seed.
If it seeded more freely it would undoubtedly supersede the common
white clover, as it gives more than twice as much herbage and seems
as hardy as the other." In 1950 the seed problem has been largely
solved and the prediction that ladino would supersede common white
clover has come to pass.

1. History

The name "ladino" appears to have been derived from Lodi, a town
in the Province of Lombardy in northern Italy. There in the upper valley
of the Po it is thought to have developed by natural selection from the
common white clover.

According to Madson and Coke (1937), seed was obtained for experi-
mental purposes by the United States Department of Agriculture in
1903. Trials made in the Northeastern states were failures, but new
tests were made in Idaho in 1911 where initial success with both forage
and seed production was obtained. The first field planting for commer-
cial seed production was undertaken in southern Idaho in 1918. Ac-
cording to Badley (1930), Jones (1931), Medeck (1929), Pickett (1933),
Storgaard (1930), and Tesche (1929), ladino readily became accepted
in California and Oregon and was soon widely and enthusiastically
grown in those states.

Although immediate acceptance of ladino as a valuable crop was not
forthcoming, the Pacific Coast states recognized its merits and it became
established there in the period from 1920 to 1930. During the following
decade, experiments and field performance tests proved its adaptation
to the Northeastern United States, and trials from 1940 to 1950 have
shown its value in the Great Lake States and in many of the South-
eastern states. Now it is grown to some extent in every state except

LADING CLOVER 209

North Dakota, and national estimates indicate about 4 million acres
planted to ladino and ladino-grass mixtures.

This rapid acceptance of ladino under different climatic and soil
conditions indicates that it is widely adapted if modifications 'in local
environment are made to meet its requirements. It means, further, that
the crop is productive, nutritious and economically practical to grow
over large areas of the United States. In 1942 Hollowell said, "Ladino
is rapidly becoming the foundation of an intensive grassland agriculture in
the Northeastern states. " His statement has since proved true and it
might now be applied to areas of the Great Lake States, some of the
Southeastern states, and to irrigated ladino valleys in the West.

2. Distribution

Ladino alone and in mixed plantings is grown most intensively in
California and Oregon, where about 800 thousand acres are produced
under irrigation. The crop is generally planted in the Northeastern
states. Ohio and New York each grow about one-half million acres in
mixed plantings primarily for pasture. Already 300 thousand acres are
indicated in Wisconsin and such Southeastern states as Virginia, North
Carolina, and Tennessee are growing well over 150,000 acres each. The
crop is increasing in acreage and significance in all except one or two
states. Table I gives the acreage of ladino clover by states as estimated
for 1949.

Acceptance of ladino in the cornbelt states has been indicated by
Hollowell (1946) and in the Southeastern states by Lovvorn (1947, 1949).
That it is more extensively grown than is generally supposed is shown
in Table I.

II. CHARACTERISTICS AND ADAPTATION
1. Description

Ladino clover, Trifolium repens L., is a large form of common white
clover. The variety name LATUM has been added to it by several writers,
as, for example, Madson and Coke (1937), and Schoth (1936, 1944), but
this is believed to be incorrect.

According to a study by Ahlgren and Sprague (1940) ladino com-
pared to common white clover is several times larger in all plant organs
but otherwise similar in appearance. It possesses thick, fleshy stolons
which root readily at the nodes and under favorable conditions spread
vigorously, since plants may cover an area of 9 to more than 16 square
feet. The stolons are the most stable feature of the plant and are most
useful in distinguishing it from common forms of white clover (Fig. 1).

210

GILBERT H. AHLGREN AND R. F. FUELLEMAN

The petioles may range in height from 6 to 30 inches, depending on
growth conditions, and each petiole bears a single leaf composed of three
leaflets. The plant does not usually flower so profusely as white clover.
The seed cannot be distinguished from white clover, being similar in size,
color, and shape. The root system is shallower than that associated with
red clover, Trifolium pratense, or alsike clover, Trifolium hybridum.

Fig. 1. A typical ladino clover plant showing general growth habit. (By per-
mission of McGraw-Hill Book Co.)

3. Adaptation

This clover has a wide climatic tolerance, yet its spread is limited by
extreme cold, drought, or heat. Its broad climatic adaptation was pre-
dicted by Hegnauer as early as 1931.

a. Cold Resistance. Ahlgren and Burcalow (1949) have reported
that "ladino appears to be as winter hardy as medium red clover but it
is less hardy than common white clover." Under Connecticut condi-

LADING CLOVER

211

TABLE I

Estimated Ladino Clover Acreage in the United States, 1949

State

Acreage

State

Acreage

Alabama

2,000

Nebraska

100

Arizona

1,000

Nevada

2,000

Arkansas

5,000

New Hampshire

30,000

California

587,000

New Jersey

80,000

Colorado

1,500

New Mexico

10,000

Connecticut

75,000

New York

500,000

Delaware

10,000

North Carolina

175,000

Florida

400

North Dakota

Georgia

63,000

Ohio

500,000

Idaho

33,000

Oklahoma

7,500

Illinois

135,000

Oregon

111,000

Iowa

5,000

Pennsylvania

87,500

Indiana

15,000

Rhode Island

500

Kansas

7,500

South Carolina

50,000

Kentucky

50,000

South Dakota

500

Louisiana

7,500

Tennessee

375,000

Maine

45,000

Texas

1,000

Maryland

50,000

Utah

2,000

Massachusetts

9,000

Vermont

100,000

Michigan

10,000

Virginia

167,500

Minnesota

500

Washing! on

7,000

Mississippi

20,000

West Virginia

1,000

Missouri

1,000

Wisconsin

300,000

Montana

5,000

Wyoming

Total in United

States 3,646,050 acres

tions, Brown (1937) found that hiclino was about as hardy as alfalfa.
Investigations by Fuellcman (1948) on reported failures due to winter-
killing in Illinois have shown these failures to be related to improper
fall management. It was a common impression, as indicated by Bruins
(1930), that ladino was not so hardy as red clover or alfalfa and, there-
fore, it could not be grow r n in the colder northern latitudes. Its Italian
origin is also partly responsible for this early conclusion. Adequate
hardiness to withstand the severe winters of Maine has been indicated
by Fink (1943). Pure stands of ladino are subject to winterkilling by
heaving, according to Brown and Munsell (1941), but this difficulty
is largely eliminated when ladino is grown in association with a grass.
This conclusion was also reached by Schoth (1944).

Smith (1949) has shown that ladino has less ability to survive en-
cased in ice than does common white clover. Presumably this means that

212 GILBERT H. AHLGBEN AND B. F. FUELLEMAN

survival under ice sheets during severe winter weather is also less
certain.

b. Summer Temperatures and Moisture. In the Southeast the high
temperatures and shorter day periods result in reduced blooming. Ac-
cording to Hollowell (1948), some ladino dies in summer and fails to
reseed. In parts of the Southeast this species frequently behaves as an
annual. Madson and Coke (1933) have also noted that excessive high
temperatures restrict its growth and injure the crop.

Ladino requires a good supply of moisture for normal growth. Thus
on the grazing lands of the West it makes its best growth under irriga-
tion, as indicated by Hegnauer (1931), Jones (1930), Jones and Brandt
(1930), Miller and Booher (1949), and Whitney (1939). Cool and moist
weather stimulates vegetative growth in the humid regions. Bright,
warm, and dry weather increases flowering and seed set, and under very
dry conditions the crop becomes almost dormant or the stand may be
injured. The difficulties experienced in growing the crop west of the
Mississippi River, except under irrigation, are for the most part related
to lack of available moisture and high temperatures.

c. Soil Conditions. Most investigators agree that ladino is best
adapted to moist well-drained soils and not to dry, sandy, or very wet
soils. This has been indicated by Ahlgren (1946), Donaldson (1939),
Haddock (1943), Johnstone (1947), Phillips (1943), Rumler (1945),
Snyder (1949), Van Alstine (1943), and many others. No evidence has
been presented to indicate adaptation on dry or light soils. It will grow
fairly well on droughty and sandy soils in seasons of abundant rainfall,
but under these conditions it is not reliable.

Ladino appears to be more tolerant of wet soils than is alfalfa or red
clover but slightly less so than alsike clover. Madson and Coke (1937)
indicated that it is rather intolerant of alkali.

3. Breeding and Genetics

No specific investigations on the breeding or genetics of ladino
clover have been reported, although considerable work on common white
clover may be found. It is probable that ladino is similar to white clover
in most phases of breeding and genetic behavior.

Some important characteristics of white clover that may be applicable
are the presence or absence of glucosides and enzymes, as indicated by
Williams (1939), Milville and Doak (1940), and Corkill (1940); the
presence of a high degree of sterility, as shown by Williams (1931) ; and
the almost complete self-compatability, as described by Atwood (1942),

LADING CLOVER 213

Natural crossing by bees is undoubtedly the primary means of pollina-
tion for ladino clover.

4. Diseases and Insects

Only a few diseases are known to attack ladino. A leafspot disease,
Cercospora zebrina, has been reported by Garber and others (1946).
An inherited disease similar to a virus has been described by Atwood
and Kreitlow (1946). 8clerotinia trifoliorum or common root rot was
reported by Kreitlow (1949) and has been found extensively in the At-
lantic coastal states. A virus referred to as "yellow patch " and given
the trinomial Marmon medicaginis H. var. Ladino n. var. has been
described by Kreitlow and Price (1949).

Such insects as the potato leafhopper, Empoasca fabae (Harv.), the
lygus bug, Lygus sp., the alfalfa plant bug, Adelphocoris lineolatus
(Goeze), and grasshoppers injure ladino in Wisconsin according to Ahl-
gren and Burcalow (1946). The lygus and alfalfa plant bugs cause
the flowers to blast, thus reducing seed yields. Leafhoppers and grass-
hoppers consume and injure the leaves. Hollowell (1942) has reported
the flea beetle, Halticus citvi (Ashm.) as very destructive in the North-
east. It makes small whitish spots scattered over most of the leaf sur-
face. Thatcher, Dodd, and Willard (1948) have indicated that the clover
leaf weevil, Hypera punctate (Fab.) often damages ladino by feeding on
the leaves in May.

III. ESTABLISHMENT AND MANAGEMENT
1. Effect of Lime and Soil pH

Best response from ladino is usually found on soils in which the pH
ranges between 6.0 and 6.5. In North Carolina Lovvorn and Dobson
(1947) suggested the use of 1 to 1% tons of lime per acre for land not
recently limed. In Connecticut Brown and Munsell (1941) have studied
the effect of lime and soil pH on ladino. On a soil of pH 5.64 the ladino
yield was 2,290 Ibs. per acre; at 5.77 it was 2,624; at 5.82 it was 2,953,
and at 5.96 it was 3,160 Ibs. per acre. There was thus a difference of
about 40 per cent in yield in favor of the highest pH value. It is ap-
parent that ladino, like other legumes, is favorably affected by addition
of adequate quantities of lime to the soil. Although it does respond to
lime, it has been said by Brown and Munsell (1941) to be more tolerant
of acid soils than is alfalfa. Hollowell (1942) has indicated that it "will
grow well on medium to slightly acid soils." Fink (1943) said that it "will
grow on acid soils provided the soil contains a minimum of 1500 pounds
of available calcium per acre."

214 GILBERT H. AHLGREN AND R. F. FUELLEMAN

2. Fertilizing New Seedings

For making new seedings of ladino various fertilizer treatments are
suggested. Thus Eby (1941) has suggested broadcasting 400 Ibs. of a
5-10-10 fertilizer per acre for a ladino-grass seeding or 800 Ibs. of an
0-12-12 material for a ladino-alfalfa mixture. Thatcher, Dodd, and
Willard (1948) suggested 500 Ibs. per acre of 0-20-0 or 0-12-12. Liberal
applications of phosphate and potash fertilizer have been advocated by
Owens (1945) but no nitrogen, since this is said to stimulate grass com-
panions to the detriment of the ladino. In Maryland, Kemp, Kuhn, and
Magruder (1943) suggested 300 to 800 Ibs. per acre of 0-14-6 or 0-12-12,
depending on the natural fertility of the soil. In Maine (1943) plowing
under 10-15 tons of manure fortified with superphosphate has been sug-
gested. The use of 300 Ibs. of 4-12-4 or its equivalent is a practice indi-
cated in Virginia by Hutcheson (1942). Prince (1945a) wrote that the
most successful ladino growers in New Hampshire are using a complete
fertilizer for seeding down. Light applications of nitrogen fertilizer
or manure together with phosphate and potash have been suggested by
Kenney, Fergus, and Henson (1949). Ahlgren and Burcalow (1949)
have reported that 200 to 400 Ibs. of commercial fertilizers such as
0-20-10, 0-20-20, 0-10-20, 0-14-7, and 0-9-27 per acre are commonly
used in the Midwest. In North Carolina, Lovvorn and Dobson (1947)
advocated from 700 to 1000 Ibs. of a 2-12-12 fertilizer per acre. In
Oregon (1944) about 300 Ibs. of superphosphate has been suggested per
acre and if the soils are poor the addition of nitrogen and potash is said
to be advisable.

In seeding down to ladino and ladino-grass mixtures many workers
suggest plowing under 5 to 10 tons of manure per acre. Supplemental
mineral fertilizers, especially phosphates, are also added. In the North-
east and the Southeast complete fertilizer mixtures containing moderate
amounts of nitrogen are suggested. In the Midwest straight mineral
fertilizers are generally supplied. Under irrigated Western conditions
if fertilizer is supplied it is usually superphosphate with smaller quan-
tities of nitrogen and potash.

3. Fertilizer Tests

The most extensive fertilizer tests on ladino have been conducted by
Brown and Munsell (1941). Their work shows that mineral fertilizers
used in conjunction with lime will double and treble the ladino yields
under Connecticut conditions. The application of lime alone or phos-
phate fertilizer alone was unsatisfactory. Similarly plots to which a
combination of phosphate and potash was applied gave inferior re<-

LADING CLOVER 215

sponses to similar plots to which lime was also added. Addition of nitro-
gen resulted in reduced yield from the ladino. This is in agreement with
recent results reported by Garb&f'et ai. (1946). Further studies by
Brown and Munsell (1941) showed that the spring addition of 100 Ibs.
of potash per acre stimulated ladino so that, by October, it occupied
61 per cent of the treated area as compared to only 29 per cent on un-
treated plots. Furthermore, the second cutting of ladino was increased
60 per cent by the potash treatment. The use of lime together with
phosphate and potash fertilizers invariably maintained better stands of
ladino than did any of these treatments used singly.

Brown and Munsell pointed out that a good ladino-grass crop will
remove 40 Ibs. of phosphoric acid per acre and that a basic application
of 200 to 300 Ibs. of 20 per cent superphosphate or its equivalent must
be added to replace this loss. Also, that 70 to 100 Ibs. of potash are
contained in a good crop, and that an annual acre treatment of 150 to
200 Ibs. of 50 per cent muriate of potash is needed to offset this. They
conclude that there is no justification for using nitrogen fertilizers on
fields having 30 per cent or more of their areas occupied by ladino.

Fink (1943) indicated that 8 tons of manure fortified with super-
phosphate (80 Ibs. N, 80 Ibs. P 2 O 5 , and 80 Ibs. of K 2 per acre) will
maintain satisfactory stands of a ladino-timothy mixture in Maine. He
also suggested that the annual application of 80 Ibs. of phosphoric acid
and 80 Ibs. of potash will provide satisfactory yield and survival of this
mixture. Further nitrogen is not needed to maintain ladino-timothy
stands and its use for this mixture will depend on economic conditions.
According to Fink, about 35 Ibs. of phosphoric acid and 169 Ibs. of pot-
ash are removed per acre by a good crop of ladino. This larger quantity
of potash, compared with that indicated by Brown and Munsell, can
mean only that more was available under the conditions of these Maine
experiments.

Under Vermont conditions Midgley (1938) increased ladino yields
by lime and fertilizer treatments. The use of iy 2 tons of lime per acre
on Woodbridge loam resulted in a 3-year average yield of ladino of
3,326 Ibs. dry matter per acre. Lime plus 80 Ibs. of phosphoric acid and
100 Ibs. of potash produced 5,670 Ibs.; lime and double the amount of
phosphoric acid and potash yielded 7,323 Ibs. per acre. Adding the
equivalent of 300 Ibs. of ammonium sulfate per acre to the last treatment
increased the yield to 8,097 Ibs. Substituting 8 tons of barnyard manure
for the potash and nitrogen gave 6,046 Ibs. Midgley concluded that lime
in conjunction with heavy mineral fertilizer treatment was most satis-
factory for ladino.

In New York, Blaser and McAulift'e (1949) showed that phosphate

216 GILBERT H. AHLGREN AND B. F. FUELLEMAN

applications gave increased yields of ladino and grass and that drilling
this fertilizer was superior to broadcasting it. Work reported from Wis-
consin by Ahlgren and Burcalow (1949) showed an average 2-year in-
crease in ladino yields of 4.2 to 6.7 tons per acre from an annual
application of 300 Ibs. of 0-20-20 fertilizer per acre.

At the U.S. Regional Pasture Laboratory, workers (1946) concluded
that the use of nitrogen fertilizer on an orchard grass-ladino association
reduced the amount of clover in the mixtures because of increased stimu-
lation of the grass and subsequent competition from it.

Several investigators have emphasized the value of potash in rela-
tion to productiveness and persistence of ladino. Thus Ahlgren (1941)
has shown that yields increase with increasing amounts of potassium
from 1 to 256 p.p.m. supplied in nutrient solution. Sprague and Eby
(1948) reported 6 years' results on ladino maintenance with varying
potash treatments. Beginning with a good stand of ladino their annual
potash treatments gave the following results: (a) no treatment, 20.6
per cent ladino, (b) 50 Ibs. K 2 O per acre, 37.3 per cent, (c) 100 Ibs.
K 2 O, 47.8 per cent, and (d) 200 Ibs. K 2 0, 60.0 per cent ladino. The 200
Ibs. of muriate of potash per acre treatment thus resulted in a ground
coverage of 40 per cent more clover at the end of 6 years than the plot
receiving no potash.

For ladino in New Jersey, Sprague and Eby (1948) have suggested
10 to 20 Ibs. borax per acre in a few soils too low in boron. Use of
boron has also been suggested for certain soils in Georgia by Brooks
and Buice (1947).

4. Topdressing Established Stands

Most authorities agree that mineral fertilizer treatments are best for
maintaining productive stands of ladino. Sprague and Eby (1948) sug-
gested the use of 300 to 400 Ibs. of 0-19-19 or 0-20-20 fertilizer per acre
as an annual treatment io New Jersey. Small annual applications of
phosphoric acid and potash in Georgia have been recommended by
Brooks and Buice (1947). In Rhode Island, Shaw (1944) recommended
light topdressings of manure in fall followed by the equivalent of 600
Ibs. of 0-14-14 in spring. He suggested spring topdressings of 800 Ibs.
of 5-10-10 per acre if manure is not added. In Pennsylvania Grau
(1944b) recommended 8 to 10 loads of superphosphated manure or 300
Ibs. of 0-14-7 or 0-12-12 per acre, applied annually. About 500 Ibs.
per acre of 0-20-0 or 0-12-12 was suggested by Thatcher, Dodd, and Wil-
lard (1948) in Ohio. Ahlgren and Burcalow (1949) reported that mod-
erate applications of barnyard manure fortified with phosphate fertilizers,
or 200 to 250 Ibs. of a phosphate and potash mixture per acre, are ap-

LADING CLOVER 217

plied in early spring every 2 or 3 years in the Midwest. In New York
Van Alstine (1943) suggested 500 Ibs. of 0-20-10 or 0-20-20 per acre
annually. For North Carolina the recommendation made by Lovvorn
and Dobson (1947) is 400 to 600 Ibs. of 0-10-20, 0-12-12, or 0-9-27 per
acre.

Thus on established ladino stands most workers suggest the use of
a light manure topdressing in early fall followed by a mineral fertilizer
treatment in spring. The application of such complete mineral fer-
tilizers as 0-14-14 or 0-10-20 is most common at rates of 200 to 500 Ibs.
per acre per year.

S. Suitable Grass Companions

The question of whether to grow ladino clover alone or with a grass
or grasses was first propounded by Brown and Munsell (1941). From
their research on this question, begun about 1932, they concluded that
suitable grasses tend to increase total yield and to protect ladino against
winterkilling. Other points favoring the inclusion of a grass are: (a)
reduction of laxative effect of ladino on cattle; (b) decreased danger
from bloat when ladino is pastured; (c) increased ease of mowing and
curing; and (d) enhancement of the chances of obtaining a satisfactory
crop.

a. The Northeast. The choice of satisfactory plant associations
depends on the use to be made of the crop, and on regional conditions
affecting the adaptation of various grasses and legumes. Thus Fink
(1943) reported that "none appear so widely adapted to Maine condi-
tions as is timothy." A similar suggestion was made by Talbot and
Miner (1946). Brown and Munsell (1941) and Owens (1945) expressed
the belief that, in Connecticut, orchard grass, Dactylis glomerata, is most
satisfactory in ladino mixtures used primarily for pasture, and timothy,
Phleum pratense, for those seeded chiefly for hay. The addition of ladino
to red clover-timothy hay mixtures has also been suggested. Haddock
(1943) in New Hampshire advocated mixtures with red clover, alsike
clover, alfalfa, Medicago sativa, or timothy for hay and with timothy,
orchard grass, 1 or smooth bromegrass, Bromus inermis, for pasture.
Smith (1949) in Vermont has propounded similar pasture mixtures except
that orchard grass is not recommended for either hay or pasture. For
New Jersey, Sprague and Eby (1948) suggested at least 1 Ib. of ladino
in all pasture mixtures that include such species as alfalfa, red clover,
alsike clover, orchard grass, bromegrass, reed canary grass, and timothy
(Fig. 2). Hollowell (1942) indicated that ladino-timothy mixtures are
especially valuable in the Northeast on productive soils and where sum-

218

GILBERT H. AHLGREN AND K. F. FUELLEMAN

mer temperatures permit the continuous growth of timothy. Mixtures
of ladino with bromegrass, meadow fescue, Festuca elatior, and reed
canary grass, Phalaris arundinacea, are said to be valuable. In Pennsyl-
vania Grau (1944) suggested orchard grass, bromegrass, Alta fescue,
Festuca elatior var. ARUNDINACEA, tall oatgrass, Arrheratherum elatuis,
reed canary grass, or perennial ryegrass, Lolium perenne.

Fig. 2. A mixture of ladino clover and bromegrass is a favorite for pasture pur-
poses. (Photograph by New Jersey Agricultural Experiment Station.)

b. The Southeast. In the Southeastern states the choice of grass
companions is somewhat different from those in the Northeast. Thus
Lovvorn and Dobson (1947) indicated orchard grass for well-drained
soils and dallis grass, Paspalum dilatatum, for poorly drained areas.
Redtop, Agrostis alba, is also valuable and tall fescue promising. Brooks
growth of timothy.

LADING CLOVER 219

c. The Midwest. In the Midwest the most popular combinations,
according to Ahlgren and Burcalow (1946), are alfalfa, red clover, alsike
clover, and timothy or bromegrass. For Ohio, Thatcher, Dodd, and
Willard (1948) suggested simple mixtures with timothy, smooth brome-
grass, or orchard grass. They also felt that ladino adds to the value of
Kentucky bluegrass, Poa pratensis, pastures although they recognized
that it is difficult to maintain a good stand of ladino in a bluegrass sod.
Fuelleman (1948), in Illinois, listed alfalfa, red clover, timothy, orchard
grass, bromegrass, or meadow fescue as suitable companions.

d. The West. In the West for irrigated pastures Miller and Booher
of California (1949) suggested mixtures containing domestic ryegrass,
Lolium multiflorum, perennial ryegrass, tall fescue, orchard grass, or
dallis grass. They indicated that occasionally some alfalfa, bur clover,
Medicago sp., birdsfoot trefoil, Lotus corniculatus, or Harding grass,
Phalaris stanoptera may be added. Schoth (1944) believed that the
best associations in western Oregon are redtop, domestic ryegrass, peren-
nial ryegrass, timothy, orchard grass and Alta fescue. In Washington
(1946) such grasses as Alta fescue, domestic ryegrass, perennial rye-
grass, Kentucky bluegrass, meadow foxtail, Alopecurus pratensis, orchard
grass, and creeping red fescue, Festuca rubra, are said to be most satis-
factory. Alsike and red clover are also mixed.

A study of the grass and legume associations suggested for ladino,
as given above, indicates that the erect hay types of crops are recom-
mended as companions. Few authorities advocate growing ladino with
such sod formers and shorter growing species as Kentucky bluegrass,
redtop, colonial bentgrass, red fescue, or Bermuda grass. The mixtures
suggested in the Midwest are similar to those of the Northeastern states.
The Southeast is most interested in tall fescue, orchard grass, and dallis
grass, and under irrigated Western conditions species common to both
the Southeast and Northeast are useful. In addition, domestic and peren-
nial ryegrass and meadow foxtail are considered valuable. Sod type
grasses are evidently less valuable than those giving an open type growth
or possessing a bunch habit.

Unfortunately, few findings are available concerning the merits of one
mixture compared to another. This entire phase of grass-legume com-
panions best suited for growth with ladino clover remains to be clarified.
Ladino is seldom sown alone except for seed or for poultry and bee
pastures.

220 GILBERT H. AHLGREN AND R. F. FUELLEMAN

6. Seeding Practices

This crop is seeded at the rate of one-half Ib. per acre with mixtures
used primarily for hay; one Ib. if the mixture is for pasture; and 2 to 4
Ibs. per acre for straight ladino stands. A single pound of ladino con-
tains about 750,000 seeds or enough for 15 seeds per square foot in an
acre of land. Brooks and Buice (1947) studied rates of seeding plant-
ing from 2 to 15 Ibs. per acre and concluded that 2 to 3 Ibs. was as satis-
factory as larger rates.

a. Shallow Planting. Most authorities agree that seed must be
planted very shallow for best success. Thus Owens (1945) and Smith
(1949) suggested from one-half to one inch. Haddock (1943) indicated
that the seeds should be barely covered. Sprague and Eby (1948) recom-
mended planting no deeper than one-quarter to one-half inch, and Lov-
vorn and Dobson (1947) said that the seed should be covered lightly.
Less than one-half inch was suggested by Miller and Booher (1949) and
by Schoth (1944). A firm compact seedbed obtained by early prepara-
tion, harrowing, rolling, or cultipacking is advocated by all authorities.
Planting at a depth of one inch or more would appear to be too deep,
and a light covering of one-quarter to one-half inch most generally
advisable.

b. Inoculation. It is believed that few growers inoculate seed of
ladino clover, but several authorities suggest that this procedure be con-
sidered. This legume belongs to the same cross-inoculation group as do
common white clover, red clover and alsike clover. Thatcher, Dodd, and
Willard (1948) have expressed the belief that even though ladino be
planted on fields that have previously grown successful crops of these
clovers the seed should always be inoculated, especially in view of the
recent discovery of superior strains of symbiotic bacteria. Fuelleman
(1948) also advocated seed inoculation for fields that have not grown
successful crops of clover recently and also if ladino is seeded on
infertile and acid soils. Brooks and Buice (1947) reported that "inocula-
tion is very necessary." Hollowell has said that "inoculation is unneces-
sary, but to insure a crop, it may be advisable, particularly at the first
seeding." According to Sprague and Eby (1948), inoculation is advis-
able on most soils unless successful clover crops have been grown recently.
Haddock (1943) indicated no need to inoculate on many soils, but sug-
gested that other soils require inoculation and when ladino is grown on
a field for the first time it should always be inoculated. Other workers
make no mention of the need for inoculation. It would seem safest to

LADING CLOVER 221

inoculate the seed until ladino is successfully grown on the area in
question.

c. Seeding Dates and Companion Crops. Owens (1945), in Connecti-
cut, suggested seeding in spring with oats or on top of fall-sown rye and
pasturing of the companion crops. He indicated that seedings may also
be made in early spring alone and in late summer until about August 10.
In Maine, Fink (1943) has indicated that seeding with spring oats is
most common but planting with canning peas is also practiced. Good
stands are also obtained by planting with sudan grass or millet in late
spring or early summer. Haddock (1943) has said that, in New Hamp-
shire, early spring seedings on top of land seeded to grass the previous
fall or plantings in late July or early August without a companion crop
are suggested. Seeding in early spring or late summer has been recom-
mended in New Jersey by Sprague and Eby (1948). The early spring
seedings are made on top of a winter grain or with spring oats. It is
suggested that the companion crop be mowed for hay or silage or pas-
tured. If oats are to be used for grain, the seeding rate should be re-
duced to two-thirds the normal rate.

Lovvorn and Dobson (1947) expressed the belief that in North Caro-
lina seeding from September 1 to 15 is best, except in the mountains,
where August or April seedings appear more practical. Brooks and
Buice reported (1947) late August or early September seedings to be best
in Georgia except in the mountains, where good stands are obtained
from March or April seedings.

In Wisconsin, according to Ahlgren and Burcalow (1946), spring
seedings in oats are considered superior and August or early September
seedings somewhat hazardous. Pasturing the companion grain crop or
making it into hay is desirable to reduce danger from shading or lodging.
Fuelleman (1948) has suggested early planting in March or April in
Illinois. Early spring seedings in Ohio are recommended by Thatcher,
Dodd, and Willard (1948), who have indicated that midsummer seedings
often fail because of unfavorable high temperatures.

Miller and Booher (1949) have indicated that seedings are made
in California from October 1 to early March because fall and winter
temperatures are more favorable than those of spring and summer.
Seedings in Oregon are most successful from May to July, according to
Schoth (1944). Fall seedings are not generally recommended because
of winterkilling and damage from slugs. Spring seedings from April to
May are preferred by Law and Ingham (1944) in Washington, but some
early fall seedings are also made.

In northern and central latitudes the most satisfactory time to plant

222 GILBERT H. AHLGREN AND R. F. FUELLEMAN*

ladino appears to be in early spring with oats as the companion crop
or on top of fall-sown winter grains or grasses. Most authorities sug-
gest that the rate of planting the companion oat crop be reduced or
used for silage, hay, or pasture. In southern latitudes early autumn
seedings without a companion crop are preferred except at high latitudes,
where early spring appears best.

d. Seeding Methods. Throughout the Northeast and the Southeast,
seeding is accomplished primarily by means of cyclone and wheelbarrow
seeders or the grass-seeding attachments on grain drills. Most workers,
as, for example, Ahlgren and Burcalow (1946), Cox (1940), Dickey
(1946), Lovvorn (1949) and Serviss (1945a), recommend firming the soil
by cultipacking either before or after seeding. Some suggest cultipacking
both before and after as extra insurance of a stand. Dilution of the
seed with corn meal or other suitable carriers helps to ensure uniform
distribution.

Law and Ingham (1944) have found that the best stands of ladino
in Washington are obtained by planting with an alfalfa or grass seed
drill. Rolling with a corrugated roller prior to seeding is advocated if
such drills are not available. This operation is then followed with a
light harrowing or another rolling to cover the broadcast seed. Most
seedings made in Oregon, according to Schoth (1944), are broadcast and
covered by a light harrowing. Miller and Booher (1949) have described
several methods used in California, namely, (a) endgate broadcast
seeders, (b) seeder and roller combined, and (c) airplanes with special
seeding attachments for such tiny seeds as those of ladino.

e. Seed Characteristics. Seed of ladino is indistinguishable from
common white clover seed. To derive the greatest benefit pure ladino
seed must be used in planting. Most purchasers are cautioned to buy
certified seed or seed of known origin from reliable seed dealers.

Some seeds of ladino, "according to Thatcher, Dodd, and Willard
(1948), are considered "hard" because they are fairly impermeable to
absorption of water. This results in delayed germination from as little
as a few days to more than a year. Hollowell (1942) has expressed the
belief that the presence of 10 to 40 per cent hard seeds in ladino is not
objectionable. The delayed germination of such seed may ensure a
stand if part of the early seedlings are killed by unfavorable weather
conditions.

Hard clover seeds pass through the digestive tracts of animals on
ladino pastures or of those consuming ladino hay. As the manure is
spread it becomes a source of volunteer ladino plants.

LADING CLOVER 223

7. Use in Renovation

Unproductive pastures may often be improved by adding lime and
fertilizer, disking and harrowing, and seeding with ladino alone or with
suitable grass partners. Fink (1943) indicated that many low, moist
soils, high in organic matter and available calcium, in Maine, can be
converted to ladino by pimply fertilizing and seeding on top of the old
sod in early spring.

Thatcher, Dodd, and Willard (1948) conducted 50 establishment
trials with ladino. Analysis of their findings at the close of the 1944
pasture season led to the following conclusions: (1) Where no soil prep-
aration was made and neither lime nor fertilizer applied, only 10 per
cent of the trials were successful. (2) Application of lime and fertilizer
but no soil preparation gave 23 per cent good stands. (3) Addition of
lime and fertilizer together with disking resulted in 41 per cent successes.
Of 26 tests conducted in 1945 and 1946, only 8 were failures on land
that had been limed, fertilized, and disked. These were explained by
late seeding, too little seedbed preparation, overgrazing, or unfavorable
weather. In contrast to other legumes tried, ladino was outstanding, as
shown in Table II. These results indicate that ladino is readily estab-
lished by common renovation practices in rundown pastures. Under such
conditions, a stand of ladino is as easily obtained as one of red or alsike
clover.

TABLE II

A Comparison of the Establishment of Legumes in
18 Renovation Tests on Ohio Farms a

Per cent of

Per cent showing

good or

Crops

perfect stand

excellent growth

at the

end of the first

season

Ladino clover

Alfalfa

Sweet clover

Birdsfoot trefoil

Lespedeza sericea

none

'Ohio Agr. Expt. Sta. Bull. 684 (1948).

8. Management

The popularity of ladino clover has increased rapidly with increased
knowledge of the techniques of managing the crop. The yield and per-
sistence are closely related to the techniques applied in growing the
crop.

224 GILBERT H. AHLGREN AND R. F. FUELLEMAN

a. Care of the New Seeding. The seedling stage is said to be the most
critical in the life of the plant. It is common belief that excessive shad-
ing may reduce seedling vigor and even destroy the plants and also that
early and too close grazing will limit the growth and development of
the young plants.

Ladino spreads very little during the first 6 or more weeks of its
growth, most of its development being in the form of root and primary
top growth. After this initial period, growth of creeping stems becomes
rapid if conditions are favorable. To encourage such spread the removal
of companion grain crops by grazing or as hay is especially valuable,
according to Sprague and Eby (1948). Removal of a companion crop
of oats by Brown and Munsell (1941) at various dates, however, did
not affect the final stand. The oats were removed on June 13, June 29,
and July 6, and the respective ladino stands on October 16 were 97, 98,
and 96 per cent.

It has been shown by Thatcher, Dodcl, and Willard (1948) that straw
from grain crops, if left on the field after combining, may smother
ladino.

b. Rotational Grazing. Brown and Munsell (1941) were the first
to demonstrate that ladino clover stands were injured under close, con-
tinuous clipping and also that frequent and close clipping reduced the
yields. Their best management treatment over a period of years was
cutting at a 10-inch height back to 4 inches. They concluded that "con-
tinuous or close grazing, especially in late fall, are not practices con-
ducive to the well-being of ladino."

Thatcher, Dodd, and Willard (1948) reported that "too close grazing
and regrazing before the new growth is well started will quickly thin the
stand." They suggested that ladino should not be grazed continuously
closer than 3 to 5 inches. Sprague and Eby (1948) have recommended
rotation grazing, permitting the clover to be consumed until 3 or 4
inches of top growth remains.

Four years 7 results by Ahlgren and Burcalow (1949) comparing mod-
erately to closely grazed ladino-bromegrass pastures showed that average
yields of dry forage were 1.3 tons per acre greater under moderate graz-
ing. They expressed the belief that the crop ought not to be grazed
until it is 8 to 10 inches high and that the grazing should then be down
to 4 or 5 inches only. Rotational grazing is the only method suggested
for utilizing ladino clover pastures.

It has been pointed out by Thatcher, Dodd, and Willard (1948) and
by Sprague and Eby (1948) that the feeding value of ladino changes
very little from the young to the bloom stage of growth, and this is an-

LADING CLOVER 225

other reason why grazing can easily be deferred. The protein content
over the entire range of maturity, however, is said to vary from 12 to
30 per cent.

c. Ladino Pasture Yields. Ladino yields reported by Brown and
Munsell (1941) indicate a range from 2,500 to 5,500 Ibs. of dry matter
per acre. In comparisons of ladino-grass mixtures, Sprague and Eby
(1948) obtained the following results: ladino-orchard grass, 7,629 Ibs.;
bromegrass-ladino, 6,555 Ibs.; Kentucky bluegrass and white clover, 2,594
Ibs. Thus the ladino-grass mixtures were shown to be several times more
productive than white clover-bluegrass pastures. According to Ahlgren
and Burcalow (1946), ladino-bromegrass yields have been as high as
6,800 Ibs. per acre. Fuelleman (1948) reported yields ranging from 3,200
to 7,577 Ibs. per acre. In comparative tests in mixtures, the following
results were obtained: ladino alone, 4,110; ladino-bromegrass, 5,195;
ladino-timothy, 4,426; and ladino-Alta fescue, 4,286. According to Lov-
vorn and Dobson (1947), ladino pastures produce 3,500 to 5,000 Ibs. of
dry matter per acre in North Carolina.

d. Carrying Capacity of Ladino Pastures. Brown and Munsell
(1941) have indicated that one acre of good ladino is sufficient to provide
200 cow pasture days. Ahlgren and Burcalow (1949) reported that "one
acre will provide pasturage for one cow, or 10 to 12 hogs, or 12 to 14
sheep, or 125 turkeys, or 400 to 600 growing chickens or 300 to 350
laying hens." According to Sprague and Eby (1948), ladino will support
500 to 600 young birds per acre. Owens (1945) suggested one acre per
cow. Fink (1943) expressed the belief that about two acres of ladino-
timothy are needed per cow for hay, silage, and pasture. Fuelleman's
(1948) experiments indicated a carrying capacity of one to one and a
half animal units per acre.

e. Fall Grazing. All authorities agree that close pasturing in the fall
will damage the stand of ladino. According to Schoth (1944), this is
more nearly true of ladino alone than when it is grown in grass mixtures.
Haddock (1943) expressed the belief that ladino ought not to be grazed
in New Hampshire after September 20. Sprague and Eby (1948) sug-
gested leaving 4 to 6 inches of top growth in the field for winter protec-
tion, and a similar recommendation was made by Ahlgren and Burcalow
(1949).

226 GILBERT H. AHLGREN AND R. F. FUELLEMAN

IV. UTILIZATION

Ladino is used in many so-called "multiple" or "triple" purpose mix-
tures. It is considered valuable for pasture, hay, and silage and for
seed production.

1. Pasture

The primary use for ladino clover is as a constituent of pasture mix-
tures. All kinds of livestock, including dairy and beef cattle, sheep,
hogs, and poultry, relish the tender succulent leaves and their high
palatability.

There is some danger that cattle and sheep may bloat on ladino pas-
ture. This hazard may be reduced by growing ladino with one or more
grass companions. In this regard, Miller and Booher (1949) have said
that 40 to 50 per cent legume has been found to meet the grazing pref-
erences of cattle and sheep with a minimum hazard of bloat. Feeding
dry hay or allowing easy access to dry hay, straw, or other grass pas-
tures is an additional safeguard. Withholding cattle and sheep from
ladino pastures when the leaves are wet or covered with dew is further
protection against bloat.

Studies by Ewalt and Jones (1939) have shown that 65 per cent of
the nutrient requirements of high-producing dairy cows could be pro-
vided by irrigated ladino pastures.

According to Snyder (1946), ladino is especially valuable for sheep,
since a high-yielding, high-protein, nutritious feed is needed. This crop
is said to help reduce labor in growing sheep and to permit maximum
use of pasture as a source of feed.

In Pennsylvania (1946) ladino pastures were shown to be particularly
valuable for growing pigs. Feeding hogs for a 70-day period on ladino
pasture with a standard concentrate ration resulted in an average daily
gain of 1.33 Ibs. per animal.* The group fed for the same period in drylot
only averaged 1.24 Ibs. daily. The investigators concluded that an acre
of ladino grazed for 70 days was worth $29.92 in terms of gain made by
the pigs.

Reduction in feed costs for growing birds on ladino pasture is usually
indicated as 15 to 20 per cent. Furthermore, the mortality rate of birds
on clover range is often lower than that of birds raised in confinement.
Pullets are said to prefer ladino to any other pasture crop. Flock (1946)
expressed the belief that a 20 per cent saving on feed can be made by
use of ladino range. Sprague and Eby (1948) showed a 14 per cent
saving of feed costs on pullets raised on range compared to those grown
in confinement. Kennard, Thatcher, and Chamberlin (1947) have

LADING CLOVER 227

shown feed cost reductions in raising pullets for egg or meat purposes on
ladino range. They also reported better egg production by pullets raised
on ladino range than by those in confinement.

2. Hay and Silage

Ladino is considered valuable for hay or silage especially when it is
grown in mixtures and carefully managed. If mixed seedings are used
continuously for hay, the ladino will soon be lost, because it cannot with-
stand excessive shading.

Maximum yields are obtained by cutting when ladino is in full bloom.
If properly cured, the hay will be of high quality, rich in minerals, vita-
mins, and protein.

Because of its tangled mass of leaves, the crop is difficult to mow and
to cure. Its moisture content (80 to 85 per cent) adds to the curing
problem. The tendency for mowed ladino to pack tightly in the windrow
also increases the problem of drying. Even when ladino is apparently
adequately cured for the mow, spoilage frequently occurs in storage.
The use of grasses for companions makes both mowing and curing easier.

Tests conducted by Thatcher, Dodd, and Willard (1948) on ladino
for hay have shown that three cuttings a year, averaging 18 to 20 per
cent protein and 3 tons of dry matter an acre, may be obtained. Such
yields are usually less, however, than comparative hay yields from al-
falfa, red or alsike clover mixtures.

A good quality ladino clover-grass silage may be obtained if partial
wilting after mowing is permitted and if the proper amount of preserv-
atives is used. Good ladino-timothy silage has been produced by Fink
(1943).

3. Seed Production

Until recently, the production of ladino seed has not been adequate
to meet the high demands. Heavy demands compared to supplies have
resulted in high prices per pound. Though most of the ladino seed is
produced in California and Oregon, there is an increasing interest in
seed production elsewhere, especially in the North Central states.

In 1944 the high prices for ladino seed led Grau (1944) to suggest
use of a dandelion rake in the harvest of the seed from poultry ranges.
Suggestions for the growing and harvesting of ladino seed in the humid
East have been made by Kenny, Fergus, and Henson (1949), by Sheldon
and Dexter (1948), and by Burcalow, Ahlgren, and Smith (1948). The
advice given by these men includes (a) the use of certified ladino seed
for planting, (b) seeding on a weed-free firm seedbed, and (c) special
techniques for harvesting and threshing the seed crop.

228 GILBERT H. AHLGREN AND R. F. FUELLEMAN

In the East the first crop is commonly pastured, or it may be removed
for silage and the second growth used for seed production. Pasturing
or mowing is best completed by early June so that the seed crop may
develop during the hot, dry weather of July and August. The crop is
mowed for seed when most of the seedheads are brown. It may be left
in the swath, placed in windrows, or allowed to dry and then placed in
stacks. Ladino is often combined directly from the windrow or swath,
but large losses of seed result from this practice and yields of as low
as 10 to 50 Ibs. per acre result.

Recently Sheldon and Dexter (1948) developed a vacuum type har-
vester containing a rotating beater and a suction nozzle. This machine
is outstandingly successful in preventing seed losses, as may be noted in
Table III. It was more effective than combining directly from windrows

TABLE III
Comparison of Methods of Harvesting Ladino Clover Seed *

Pounds per acre of
Methods of harvest seed recovered

Combining from windrows 100

Combining from swaths 60

Vacuum harvests from swaths 160

'Michigan Agr. Expt. Sta. Quart. Bull. 31 (1948).

and far superior to combining from the swaths (1943). Use of the
vacuum type harvester is recommended in Michigan and Wisconsin.

In California, Oregon, and other Western states nearly all the seed
is raised under irrigation. The period of greatest blooming is during
late June. Early spring growth is grazed as in the East, and mowing is
preferred because it gives greater uniformity of growth of the seed crop.
In all areas if pasturing is'not practiced, the first cutting is usually re-
moved as hay, since it often contains many grasses and weeds. The seed
crop is ready for harvest in late August or early September when nearly
all the seedheads are brown. A period of two to two and a half months
after mowing a hay crop or grazing is required for the ladino seed to
reach maturity.

In the West the common practice is to mow, windrow, and shock the
ladino in the field for a few days prior to threshing. The seed is threshed
by means of a stationary threshing machine with clover-hulling attach-
ments. Seed yields range from 50 to 300 Ibs. per acre with average yields
about 100 Ibs.

Three problems common to ladino seed production in both the East

LADING CLOVER 229

and West are (a) ensuring adequate pollination, (b) elimination of weed
seeds, and (c) maintenance of seed purity. To set seed, the flowers must
be pollinated by bees. Ford (1948) has indicated that ladino produces
plenty of nectar and is attractive to bees. Newell and Mead (1945)
have expressed the belief that it is far superior to red clover but probably
not so good as alsike in its attractiveness for bees. According to Wood
(1944), it is a valuable honey plant and attractive to bees. In the ladino
seed-growing areas, common recommendations include provision of one
to two hives of bees per acre of ladino raised for seed.

Madson (1945) was one of the first to point out the problems in
producing ladino seed. He recognized that the buyer wanted seed free
from undesirable weed seeds and also of the pure ladino type. Such
weed seeds as buckhorn, Plantago lanceolata, and dock, Rumex crispus,
are particularly troublesome. Yellow trefoil, Medicago lupulina,
birdsfoot trefoil, sorrel, Rumex acetosella, cinquefoil, Potentilla sp., witch-
grass, Panicum capillare, and alsike clover are other common contam-
inants.

According to Madson (1945), white Dutch clover and Louisiana white
clover have probably been sold occasionally as ladino. This practice is
recognized as unfair and it will eventually be corrected through certifica-
tion or other means. The need to seed ladino on fields free of white
clover is also apparent, since the latter cross-pollinates readily with
ladino. It is also evident that adequate isolation of ladino seed fields
from white clover is necessary to prevent cross-pollination.

The production of ladino seed is increasing steadily according to
Edler (1947). In 1936 only 1,220 acres were raised for seed, from which
came 106,000 Ibs. More recent statistics are as follows: 1940: 7,000
acres and 560,000 Ibs. of seed; 1945: 16,800 acres and 990,000 Ibs. of
seed; 1949: 25,000 acres and 2,400,000 Ibs. of seed. Fairly large imports
were also made from Italy, so that probably nearly three and three quar-
ter million Ibs. of seed were raised in or imported into the United States
in 1949. Thus at last the seed supply seems to have caught up with
demand.

4- Other Uses

Ladino is being used successfully as a cover crop in orchards according
to Haddock (1943), Sprague and Eby (1948) and Thatcher, Dodd, and
Willard (1948). The dense mat of low-growing, nitrogenous materials
protects the soil and helps conserve moisture. Because it is a perennial
and seeds freely, ladino is easy to maintain in orchards. For orchard
cover and improvement it is grown alone or with grass companions such
as timothy or bluegrass.

230 GILBERT H. AHLGREN AND R. F. FUELLEMAN

It is also useful for growing as orchard mulch and can stand some
cultivation. As a competitor with the trees for moisture during dry
periods it is much less severe than is alfalfa or sweet clover.

V. SUMMARY

Ladino clover is undoubtedly one of the most important pasture
legumes in the United States. The characteristics of this plant which
have led to its rapid rise and acceptance by livestock farmers are
numerous. A brief recapitulation here appears in order.

As a species, ladino is widely adapted to different soil and climatic
conditions. A temperate climate and moist fertile soils favor its growth
and development. It will also grow well on poorly drained and mildly
acid soils, but it is not adapted to poor droughty soils or extremely wet
soils.

It possesses a perennial habit of growth and often establishes itself
by natural reseeding. As a companion for the hay type grasses and
legumes, it is reasonably satisfactory and far superior to common white
clover. It is nutritious and palatable, begins growth early in spring, and
recovers rapidly after grazing or mowing. As a pasture crop, it is highly
productive.

Among the disadvantages of ladino are its inability to withstand
close or continuous grazing and its tendency to winterkill or summerkill.
It is difficult to mow for hay and also to cure, and hay yields are usually
not comparable to those from alfalfa or red clover.

From a national standpoint there is increasing interest in ladino, pri-
marily as a pasture crop and secondly for hay or silage. The acreage
of ladino alone or with grass companions is already around four million,
and increases are evident wherever the crop can be grown successfully.

REFERENCES

Ahlgren, G. H. 1941. Soil Sci. 52, 229-235.

Ahlgren, G. H. 1946. Hoard's Dairyman 91, 332.

Ahlgren, G. H., and Sprague, H. B. 1940. New Jersey Agr. Expt. Sta. Bull. 676.

Ahlgren, H. L., and Burcalow, F. V. 1946. Wisconsin Agr. Expt. Sta. Ext. Circ.

367.

Ahlgren, H. L., and Burcalow, F. V. 1949. Crops and Soils 2, 5-7.
Atwood, S. S. 1942. J. Am. Soc. Agron. 34, 353-364.
Atwood, S. S., and Kreitlow, K. W. 1946. Am. J. Botany 33, 91-100.
Badley, J. E. 1930. Pacific Rural Press 120, 155.
Blaser, R. E., and McAuliffe, C. 1949. Soil Sci. 68, 145-150.
Brooks, 0. L., and Buice, G. D. 1947. Georgia Agr. Expt. Sta. Circ. 153, 12 pp.
Brown, B. A. 1937. Rural New Yorker 96, 312.

LADING CLOVER 231

Brown, B. A., and Munsell, R. I. 1941. Starrs (Connecticut) Agr. Expt. Sta. Bull.

235.

Bruins, J. F. 1930. Hoard's Dairyman 75, 566.
Burcalow, F. V., Ahlgren, H. L., and Smith, D. C. 1948. Wisconsin Agronomy Dept.

Mimeograph.

Corkill, L. 1940. New Zealand J. Sci. Tech. 22, 65B-67B.
Cox, H. R. 1940. New Jersey Agr. 22, 1.

Dickey, J. B. R. 1946. Pennsylvania Agr. Ext. Circ. 296, 1-10.
Donaldson, R. W. 1939. Massachusetts Agr. Ext. Leaflet 144, 4.
Eby, C. 1941. New Jersey Agr. Expt. Sta. Circ. 408, 7.
Edler, G. C. 1917-1947. UJS. Dept. Agr., Bur. Agr. Economics.
Ewalt, H. P., and Jones, I. R. 1939. Oregon Agr. Expt. Sta. Bull. 366, 1-25.
Fink, D. S. 1943. Maine Agr. Expt. Sta. Bull. 415.
Flock, M. V. 1946. Poultry Triv. 52 (13), 14, 29-31.
Ford, W. H. 1948. Gleanings Bee Cult. 76, 679.
Fuelleman, R. F. 1948. Ill Agr. Ext. AG 1358, 1-11.
Garber, R. J., Myers, W. M., Sprague, V. G., Sullivan, J. T., Robinson, R. R., and

Kreitlow, K. W. 1946. U.S. Dept. Agr. Misc. Pi/6. 590.
Grau, F. V. 1941. Hoard's Dairyman 86, 195.
Grau, F. V. 1944a. Am. Poultry J. 75 (7), 28, 29.
Grau, F. V. 1944b. Pennsylvania State Coll Agr. Ext. Circ. 261, 8.
Haddock, J. L. 1943. N. H. Agr. Ext. Circ. 254.
Hegnauer, L. 1931. Washington Ext. Circ. 17, 2.
Hollowell, E. A. 1942. C/./S. Dept. Agr. Farm Bull. 1910, 9.
Hollo well, E. A. 1946. Successful Farmer 44, 64-67.
Hollowell, E. A. 1948. U. Dept. Agr. Yearbook Agr., 360-363.
Hutcheson, J. D. 1942. Better Crops 26, 17-18.
Johnstone, W. F. 1947. The Progressive Farm. 62 (1), 13.
Jones, B. J. 1931. Pacific Rural Press 121, 391.
Jones, I. R. 1930. Hoard's Dairyman 75, 718.
Jones, I. R., and Brandt, P. M. 1930. Oregon Agr. Ext. Bull. 264.
Kemp, W. B., Kuhn, A. 0., and Magruder, J. W. 1943. Md. Farm. 27 (7), 2, 13.
Kennard, D. C., Thatcher, L. E., and Chamberlin, V. D. 1947. Ohio Agr. Expt.

Farm Home Res. Bull. 246, 112-116.

Kenney, R., Fergus, E. N., and Henson, L. 1949. Kentucky Agr. Ext. A. Mim. 194.
Kreitlow, K. W. 1949. Phytopath. 39, 158-166.
Kreitlow, K. W., and Price, W. C. 1949. Phytopath. 39, 517-528.
Law, A. G., and Ingham, I. M. 1944. Washington State Coll. Ext. Circ. 83, 4.
Lovvorn, R. L. 1949. South. Seedsman 12 (2), 18, 43.

Lovvorn, R. L., and Dobson, S. H. 1947. North Carolina Agr. Ext. Circ. 301, 1-4.
McCarthy, G., and Emery, F. E. 1894. North Carolina Agr. Expt. Sta. Bull. 98.
Madson, B. A. 1945. California Dept. Agr. Bull. 34 (1), 15-16.
Madson, B. A., and Coke, J. E. 1933. California Agr. Ext. Circ. 81, 1-16.
Madson, B. A., and Coke, J. E. 1937. California Agr. Ext. Circ. 81, 1-15.
Maine. 1943. Maine Agr. Ext. Circ. 172, rev. 12 pp.
Medeck, C. H. 1929. Pacific Rural Press 117, 54.
Midgley, A. R. 1938. Vermont Agr. Expt. Sta. Bull. 431.
Miller, M. D., and Booher, L. J. 1949. California Agr. Expt. Sta. Ext. Circ. 125.
Milville, J., and Doak, B. W. 1940. New Zealand J. Sci. Tech. 22, 67B-71B.

232 GILBERT H. AHLGREN AND R. F. FUELLEMAN

Newell, R. E., and Mead, R. M. 1945. Am. Bee J. 85, 433.

Owens, J. S. 1945. Connecticut Agr. Coll. Ext. Folder 2, 6 pp.

Pennsylvania. 1946. Pennsylvania Agr. Expt. Sta. Bull. 480.

Phillips, C. E. 1943. Delaware Univ. Agr. Ext. Mimeog. Circ. 17, 4 pp.

Pickett, J. E. 1933. Pacific Rural Press 126, 71.

Prince, F. S. 1945a. Better Crops with Plant Food 4, 6.

Prince, F. S. 1945b. New Eng. Homestead 118 (4), 6-8.

Rumler, R. H. 1945. Hoard's Dairyman 90, 142.

Schoth, H. A. 1936. Oregon Agr. Expt. Circ. 117, 1-8.

Schoth, H. A. 1944. Oregon Agr. Expt. Sta. Circ. 161, 12.

Serviss, G. H. 1945a. Am. Agr. 142 (1), 5, 8.

Serviss, G. H. 1945b. Am. Agr. 142, 109.

Shaw, R. S. 1944. Rhode Island State Coll Ext. Circ. 33, 6.

Sheldon, W. H., and Dexter, S. T. 1948. Michigan Agr. Expt. Sta. Quart. Bull. 31,

215-218.

Smith, D. 1949. J. Am. Soc. Agron. 41, 230-234.
Smith, L. H. 1949. Vermont Agr. Ext. Circ. 116.
Snyder, H. 1946. The Sheepman 17 (8), 320-321.
Snyder, H. J. 1949. Hoard's Dairyman 94, 211.

Sprague, M. A., and Eby, C. 1948. New Jersey Agr. Expt. Sta. Bull. 736.
Storgaard, L. H. 1930. Pacific Rural Press 120, 113.
Talbot, R. F., and Miner, B. B. 1946. Maine Ext. Circ. 172.
Tesche, W. C. 1929. Pacific Rural Press 118, 344.
Thatcher, L. E., Dodd, D. R., and Willard, C. J. 1948. Ohio State Agr. Expt. Sta.

Res. Bull. 684.
Thatcher, L. E., Dodd, D. R, and Willard, C. J. 1948. Ohio State Res. Bull. 251,

48-53.

Van Alstine, E. 1943. New York Agr. Coll. (Cornell) Ext. Bull. 569.
Vander Meulen, E. 1943. Michigan Agr. Expt. Sta. Quart. Bull. 25, 329-333.
Whitney, D. J. 1939. CaL Cultivator 86, 292.
Williams, R. D. 1931. Welsh PI. Breeding Sta. Bull. 412, 181-208.
Williams, R. D. 1939. J. Genetics 38, 357-365.
Wood, A. D. 1944. Gleanings Bee Cult. 72, 539.

The Control of Soil Water

E. C. CHILDS AND N. COLLIS-GEORGE
School of Agriculture, University of Cambridge, England

CONTENTS

Page

I. The Scope of the Review 234

II. Research Methods 234

III. The Basic Approach 235

1. The Soil Moisture Characteristic 235

a. The Concept 235

b. Experimental Methods 236

c. Interpretation of Moisture Characteristics 240

d. Soil Moisture Constants 243

e. The Equivalence between the Moisture Characteristic and the

Static Moisture Profile 243

f. The Soil Moisture Characteristic as a Tool for Measuring Soil

Moisture Content 244

2. Soil Permeability 246

a. Definition 246

b. The Measurement of the Permeability of Saturated Soil Samples 247

c. The Measurement of the Permeability of Unsaturated Soil . . . 249

d. The Permeability of Soil to Air *. . 253

e. The Relationship between the Permeability and the Physical

Constitution of the Soil 253

3. The Diffusion of Water in Soil 253

IV. Drainage and Irrigation 258

1. The Soil Water Balance Sheet 258

2. Irrigation 258

a. Statement of the Problem 258

b. The Fundamental Solution 258

c. The ad hoc Solution 260

d. The Arbitrary Solution 260

3. Drainage 262

a. The Nature of the Problem 262

b. The Fundamental Solution 264

c. Some Typical Problems 264

d. Field Experiments 266

e. Field Measurements of Permeability 267

4. Engineering Aspects 268

References 269

233

234 B. C. CH1LDS AND N. COLLIS-GEORGE

I. THE SCOPE OF THE REVIEW

A review of work in the field of soil moisture in all its aspects would
cover practically the whole subject of soil physics, and it is necessary
in the first place to define for the reader the scope of this survey. The
decision to control soil water, whether by drainage or by irrigation, or
both, is ultimately made upon economic considerations. Control works
will only be entered upon by the practising farmer if he has reason to
suppose that, as a consequence, the increase in the annual agricultural
income will exceed the equivalent annual cost of the measures. The
costs of works are in the main determined by purely physical matters
such as the design of the drainage or irrigation system. Purely physical
matters, such as the effect of the system upon the water table on the one
hand or on the distribution of irrigation water on the other, will enter
into the first stage of the estimation of returns. We propose to deal with
such physical aspects in detail, since the bulk of recent research on
drainage and irrigation has been in this field.

Biological questions are raised when we have to consider the effect
upon crops, and hence upon income, of the water control measures
adopted. We shall find rather little to say upon such topics; one can
but present as facts the little information which is available. In any
case it must not be forgotten that the advantages experienced may not
be biological at all; for example, benefits of drainage may be felt through
improvement of farm transport or facilitation of cultivation, both pri-
marily physical matters.

Finally, the farmer is concerned with economic matters such as the
market for his increased production. Under the economic heading must
also be considered such questions as whether the improved land will
attract farmers good enough to exploit the improvements to the full.
It is not everybody, for example, who will choose to farm fen peat
(muck), fertile as it may be when reclaimed. These economic-cum-
psychological, imponderable questions fall quite outside our province;
they are mentioned only in order to place the review in its proper agri-
cultural perspective.

II. RESEARCH METHODS

The approach to problems may commonly be on of two quite dif-
ferent kinds. We may need to know urgently, for practical purposes,
the effect of particular treatments applied to soils and crops. In such
straits it is enough to have ready at hand the results of field experiments
in which just such treatments have been applied. This is the method
commonly adopted by the extension or advisory service. It has the

THE CONTROL OF SOIL WATER 235

evident advantage of giving the limited amount of information which
is specifically sought, and this information can be obtained by any en-
lightened agronomist who knows what he wants. It has serious disad-
vantages. Insight into the fundamental processes at work is rarely
obtained, so that the information gained is relevant solely to the particu-
lar conditions of the experiment and may be applied to other condi-
tions only with more or less risk. In relation to the area benefited, the
costs of the investigation may therefore be high, and this is especially
the case with engineering projects such as are associated with water
control measures.

The alternative method is to make a fundamental analysis of the
processes involved in terms both of precisely defined and measurable
soil and plant properties and of equally precisely specified conditions.
Among the former, for example, we might include the permeability and
moisture characteristics of the soil profile and among the latter the
rate at which water is being sprayed on the surface or the design of the
drainage system which has been installed. Measurement of the circum-
stances and fundamental properties in any particular case would then
enable us to forecast events with accuracy. The advantages of this
method are, firstly, its capacity for wide application and, secondly and
consequentially, its ultimate economy of effort and, therefore, its cheap-
ness. The disadvantages are that it is slow; it cannot progress at com-
mand but must rely on individuals having from time to time flashes of
inspiration; it is not necessarily able to answer any practical question
on demand; and it requires the services of scientists who are specialists
in their own fields.

A third method, in some respects to be regarded as lying between
those already described, consists in seeking arbitrarily defined soil and
plant properties capable of ready measurement and empirical correla-
tion with field properties of importance which themselves defy precise
definition. This concept will be made clearer when it is dealt with in its
proper place. All methods have their proper places. We shall present
basic ideas as a logically developed skeleton upon which to build a body
of observed fact.

III. THE BASIC APPROACH

1. The Soil Moisture Characteristic

a. The Concept. All forms of control of soil water content rely upon
the control of water movement in soil. This movement is caused by a
hydraulic .potential gradient, and is limited by a soil property known as
permeability. The latter will be defined under III-2.a., while of the

236 E. C. CHILDS AND N. COLLIS -GEORGE

former it is enough to say at the moment that the total potential, which
is usually written <f>, is the sum of two components. If a body of water
is high up, it tends to fall to a lower level; height is a measure of the
gravitational component of potential. Again, in a system where water
is at different pressures at different places, the water tends to move from
the higher pressure to the lower pressure zones ; pressure is a component
of potential. The total potential is obtained simply by adding the
components algebraically (i.e., taking their signs into account) , a feature
which makes potential a more amenable factor to deal with than is the
force on soil water, since force components have to be added by geo-
metrical constructions proper to vectors.

For water to move in soil, there must be differing potentials, and in
general this implies separately differing gravitational and pressure com-
ponents. Since the pressure of the water affects the moisture content of
the soil, a varying pressure implies a soil of varying properties. It is
basic to the study of the movement of water to reach an understanding
of the relation between soil water pressure and soil water content. Here
we must observe that the pressure is commonly measured relative to at-
mospheric pressure as zero. A "head" of water in excess of atmospheric
pressure is called a hydrostatic pressure, and a diminution of pressure
below atmospheric is called sometimes a suction and sometimes a pres-
sure deficiency. The word "suction" is open to some academic objec-
tions, but it has the advantage of brevity and of conveying a well
understood meaning to the general agronomist, and we shall therefore
adopt it here.

Methods of measuring the soil moisture content at a chosen value of
water pressure or suction are described under III-l.b.; if we have the
results of such measurements and plot them as a curve, as in Fig. 1,
this curve describes the whole course of the pressure-moisture content
relationship, and has been called the "moisture characteristic" by Childs
(1940). Often it appears that the moisture characteristic can have dif-
ferent forms for the same soil, one form when the experiment begins
with wet soil and the suction is steadily increased, the other when the
soil has initially little water and the suction is steadily relaxed. There
has been some difference of opinion as to how far this hysteresis, as it
is called, is a real effect, and the matter will be further discussed under

m-i.c.

b. Experimental Methods. To determine the moisture content of soil
presents no difficulty, but to measure the pressure, and particularly the
suction, is another matter. A direct method uses a water manometer in
contact with the soil sample via a permeable membrane. Suppose, in

THE CONTROL OF SOIL WATER

237

Fig. 2 (a), we have a funnel with a floor in it pierced with holes, in each
of which is sealed a fine capillary tube T, of circular cross-section (radius
r), the neck of the funnel being connected by a t/-tube of rubber to a
burette B. Let us fill the apparatus with water and adjust the height
of the burette until the level of water in it is the same as that of the floor

o
o

Saturation

Hydrostatic Pressure

Fig. 1. A hypothetical moisture characteristic. The arrows indicate whether the
experiment was carried out with increasing or decreasing moisture content.

of the funnel. Then we know that the capillary tubes will suck water
to a height H above the free level in the burette, where

H = 2S/gpr (III.l)

S being the surface tension and p the density of water. If now we lower
the burette, the water level in the capillaries will also fall, until, when
it is level with the floor of the funnel, the water level in the burette will
be H below the floor, (Fig. 2, b). We could now cut off the capillaries
flush with the floor, leaving mere holes (Fig. 2, c), so that these holes are
in effect supporting the water beneath them which is subjected to the
suction H, (the free water level in the burette being at zero hydrostatic
pressure). A further lowering of the burette would, of course, be an
attempt to lower the capillary water level below the floor of the funnel,
i.e., out of the tubes altogether, when air would rush through the holes

238

E. C. CHILDS AND N. COLLIS-GEORGE

and allow the level of water in the funnel to fall to that in the burette.
A filter paper or sintered glass filter is, in effect, just such an array of
very fine capillary holes, and provided we do not exceed the suction
which the largest hole can support, water is held up to the underside of
the filter and is therefore in contact with the water in any soil sample
which we may place on top. Such a soil sample is thus subjected to the

(a)

Soil
Sample

Fig. 2. This explains the action of a membrane in the simple manometer ap-
paratus for measuring moisture characteristics. The capillary meniscus supports the
suction H, which is transmitted to the soil sample in (d).

suction indicated by the level of the burette water below the sample
(Fig.2,d).

Filter papers and sintered glass filters are vulnerable, since a single
accidental large hole will let air through at low suctions. Recent devel-
opments have been directed to the choice of membranes such as cello-
phane and sausage skin, which will withstand large suctions. A second
disadvantage is that, even with adequate membranes, this simple ap-

THE CONTROL OF SOIL WATER 239

paratus is limited to suctions less than the height of the water barometer,
since at this suction the water in the funnel would fall away from the
under side of the "floor" even though no air entered. This limitation
has been removed by making the funnel of steel and brass, providing it
with an airtight sealed lid, and applying pressure to the space above
the "floor" instead of suction to the space below it. In effect this in-
creases the atmospheric pressure suffered by the soil sample, so that
the room air pressure on the under side represents a corresponding rela-
tive suction; provided the pressure difference on the two sides of the
membrane is maintained, it does not matter whether there is pressure on
the upper side or suction on the lower. These developments are the
work of L. A. Richards and his collaborators, and the very complete
list of references which he has given in his recent paper (1949) enables
us to avoid quoting them separately. "Pressure plates," as they have
been called, have been built to work at pressures up to 180 atmospheres.

The centrifuge has been used to apply high suctions to soil samples.
It will be more convenient to discuss the basis of such uses of the centri-
fuge later (III-l.e), but it may be said here that Russell and Richards
(1938) have obtained moisture characteristics in this way.

Various indirect methods have in the past been devised to measure
high suctions, and they will be described here only in such measure as is
necessary to the understanding of recent criticisms. The best known
indirect methods utilize the fact that suction exerted upon water has
an effect upon its freezing point and upon the pressure of its vapor in
contact with it. Dissolved salts also have just such an effect, so that
suction is not uniquely determined unless the other contributory factors
are known. The freezing point method cannot be interpreted without
certain assumptions, such as that the ice formed is free of the soil water
suction (Schofield and Botelho da Costa, 1938). These authors find
that suctions measured in this way agreed usually with those measured
by a rough direct method, a result confirmed by Richards and Campbell
(1948) using thermistors as thermometers. However, Schofield and Bo-
telho da Costa themselves found one soil where agreement was not ob-
tained. More recently Davidson and Schofield (1942) described a soil
for which agreement was obtained between the freezing point and vapor
pressure methods. Day (1942), on the other hand, finds the freezing
point method technically unsatisfactory. Anderson and Edlefson (1942c)
use a method of progressive freezing of an initially saturated soil in a
dilatometer, and Edlefson and Anderson (1943) interpret these results
on the assumption that the strong attractive force which solid surfaces
exert on water molecules in their immediate neighborhood introduces a
positive pressure. Alexander et al. (1936) describe an electrical method

240 E. C. CHILDS AND N. COLLIS-GEORGE

of detecting the freezing point which seems capable of great precision.
We have mentioned that allowance must be made for osmotic pres-
sure when calculating suction from the freezing point or the vapor
pressure of soil water. Edlefson and Anderson (1943) require the same
correction to be made in the direct membrane method, while Schofield
categorically states the contrary, when he says that a "suitable mem-
brane" gives the suction directly (1948). In our opinion it is a mistake
to suppose that mechanical potential is only a term in the more general
free energy of the water, and that, in consequence, an osmotic pressure
difference must always be associated with a hydrostatic pressure dif-
ference even in the absence of a semi-permeable membrane. If there
should be a momentary mechanical equilibrium between the water on
one side of a permeable membrane and the soil solution on the other,
with a difference of salt concentration, this is not a true state of thermo-
dynamic equilibrium, which will subsequently be slowly achieved by a
process of diffusion, as Edlefson and Anderson themselves recognize.
Thermodynamic arguments are based upon states of thermodynamic
equilibrium, and it is only where such equilibrium is maintained, for
example, by a truly semipermeable membrane that the osmotic pressure
in the soil solution must be compensated by a suction on the water in the
funnel, giving an indicated suction which is not related to the hydro-
static suction experienced by the soil water. With a truly permeable
membrane the suction is obtained without error. To be sure there is
no certainty whether a membrane is permeable or semipermeable, since
absence of diffused solute in the manometer may mean either that such
solute is absent in the soil (no osmotic pressure) or that it is present
but excluded by the membrane, when there would be an osmotic com-
ponent of suction. In a private communication, Schofield suggests very
reasonably that a filter coarse enough to break down at suctions of less
than 20 atmospheres (capillary radius 750 A) must be permeable to
particles small enough to contribute sensibly to osmotic pressure. Such
a membrane would therefore measure suction correctly, but error might
arise with the use of membranes such as cellophane and sausage skin
(Reitemeier and Richards, 1943).

c. Interpretation of Moisture Characteristics. Soil may lose water by
one or both of two different mechanisms. If loss of water is not attended
by shrinkage, air must enter. If loss of water is attended by exactly
equivalent shrinkage, no air enters, but the particles are drawn more
closely together. The interpretation of the moisture characteristic will
be different in the two cases.

The entry of air into a "cell" in the voids amounts to an air-water

THE CONTROL OF SOIL WATER 241

interface being drawn through the channel of entry. Since water wets
the solid surfaces of the channel, the interface must be curved, and will,
therefore, support a pressure difference on the two sides given by the
well known expression, really just another form of equation III.l.,

p = 2S/r (111.2)

where r is the radius of the cylindrical channel and the pressure on the
convex or water side of the interface is less by the amount p than that
on the other or air side; i.e., p is the suction on the water. If the channel
is not cylindrical no simple expression can be used, but we can calculate
a value r from a knowledge of p and refer to it as the effective channel
radius, in much the same way as one sometimes calculates the effective
radius of a solid particle from a knowledge of the velocity of its fall in
water, using Stokes law, even though it be known that the particle is not
spherical.

If we increase the suction on a nonshrinking soil from p t to p* and
find that a volume V of water is released in the process, we can say that
this volume is contained in channels of size less than r/ (since otherwise
the water would have been released at a suction lower than pi) but
greater than r% (since otherwise the water would not have been released
at a suction as low as p 2 ). The moisture characteristic of a nonshrinking
soil can therefore be interpreted as reflecting the pore size distribution.
That this interpretation is essentially correct has been directly demon-
strated by Swanson and Peterson (1942), who compared a direct measure-
ment of pore size distribution under the microscope with that inferred
from the moisture characteristic.

This interpretation has provided us with a powerful tool for soil
structure investigation, comparable in importance with and complemen-
tary to mechanical and aggregate analysis. Donat (1937) has in this
way followed the course of tilth formation by frost, while Childs (1940;
1942) has used the method to assess the stability of clay soils to wetting
from the dry state and to show the dominant role of organic matter in
determining this stability. Feng and Browning (1946) have found this
a useful tool for the study of the degree of instability of some soils in
Iowa. It may be of interest here to remark that Russell's use of the
soil moisture characteristic (1945) to estimate the amount of water re-
quired to raise a water table by a given amount is an example of the
direct utility of the moisture characteristic without reference to inter-
pretation. The same author's study of soil structure (1941) relies upon
the interpretation in terms of pore size distribution. We may cite Brad-
field and Jamieson (1938) and Learner and Lutz (1940) as others who
have examined soil structure in this way.

242 E. C. CHILDS AND N. COLL1S -GEORGE

When loss of water is accompanied by an exactly compensating
shrinkage of the soil, there is clearly no air entry. In this case we have
to deal with colloidal or surface active soil particles, and the increasing
suctions which are required for the progressive extraction of water are,
in fact, required to draw the solid particles more closely together against
the "cushioning" of the Gouy diffuse layers of ions which surround the
particles. The distance to which such a layer extends from the solid
surface when it is quite free to develop in an infinite amount of external
solution has long been known (Gouy, 1910), but the calculation of the
equilibrium conditions when the layer development is restricted by the
presence of a neighboring particle has but recently been carried out
(Langmuir, 1938; Verwey and Overbeek, 1948; and Schofield, 1946).
In effect the increased concentration of the dissociated ions in a layer
which is thinner than that which would develop freely introduces an
osmotic pressure component which, for equilibrium, must be balanced
by a suction; release the suction and water will be drawn in to allow
the particles to repel each other, i.e., to permit the mass of soil to swell.
The further development of this interpretation of the moisture character-
istic of a swelling soil is an important field of future research.

A satisfactory interpretation of the moisture characteristic should
account for observed hysteresis. In fact observers are not agreed that
hysteresis is always demonstrable. Haines (1930), Schofield (1935),
S. J. Richards (1938), and Richards and Fireman (1943) are among
those who have reported hysteresis, while Christensen (1944) has in-
ferred it from the variations of the permeability of soil at constant suc-
tion. On the other hand, among authors who have not been able to
satisfy themselves that the phenomenon was demonstrable are Rogers
(1935a) and Edlefson and Smith (1943). On theoretical grounds it seems
probable that a nonshrinking soil of irregular pore size should exhibit
hysteresis. Without necessarily subscribing to Haines's quantitative
exposition, we may recognize that if an air-water interface is drawn into
a larger cell via a narrower neck, it will require a greater suction to
draw it through the neck than is required to prevent it climbing back to
the neck through the cell when the suction is released by stages; i.e.,
the suction required to empty the cell is greater than that which will
just permit the cell to refill. Such cells and necks are a feature of
loosely packed sands, which typically exhibit hysteresis. Structural
cracks of more or less uniform calibre might not show such an effect
markedly. When we turn to shrinking soils, we might not expect hyster-
esis, since the thickness of a restricted Gouy layer is determined uniquely
by the suction. However, it might well be that one effect of suction
would be to orientate the particles irreversibly into positions of closer

THE CONTROL OF SOIL WATER 243

packing. It may be that Schofield referred to such a mechanism when
he ascribed hysteresis to "micro-plastic forces" (1935); he shows the
similarity in this respect between the moisture characteristic and the
pressure volume curves for clays in a later paper (1938).

However one may account for hysteresis, the fact remains that if
it occurs, there is a possibility of a drying soil remaining in contact with
a wetting soil of lower moisture content yet at equal suction, and being
in equilibrium. Schofield (1935) has sought to account for the phe-
nomenon of "field capacity" on these grounds alone .....

d. Soil Moisture Constants. In general, soil moisture characteristics
are smooth curves, and present no evidence that water is present in
sharply defined groups, bound with forces of quite different kinds. At
very high suctions, of course, such sharply defined groups may exist and
are, in fact, used for diagnosis of clay mineral types. The so-called
moisture constants, e.g., hygroscopic moisture, capillary and noncapillary
moisture, can only be defined arbitrarily either by specifying a suction
pressure or, as in Baver's definition of capillary and noncapillary water
(1938), by arbitrarily specifying a feature of the moisture characteristic
as the point of inflexion, where it exists.

e. The Equivalence between the Moisture Characteristic and the
Static Moisture Profile. A column of soil with its lower end in a reser-
voir of water at zero hydrostatic pressure will draw water upwards or,
if saturated to begin with, will hold some of the water back and not
let it drain wholly out. At a height H cm. above the free water in the
reservoir the suction is H cm. of water. Thus at equilibrium, which
may possibly be attained only after a long time, the moisture content
at height H cm. is that appropriate to a suction of H cm. of water.
The plot of moisture content against height (the moisture profile) is,
therefore, in such circumstances a repetition of the plot of moisture
content against suction (the moisture characteristic), as Buckingham
pointed out long ago (1907). If, in the field, water is found to settle
out into a borehole to give a free water surface at a definite depth, this
free surface is clearly at zero hydrostatic pressure; it is known as the
water table. Soil water below this is under pressure and above it is
under suction. It is the nature of most moisture characteristics that little
water is lost until. a certain minimum suction is applied, since even the
largest pores are commonly not very large and require a suction of a
few .centimeters of water to empty them. Hence the zone immediately
above the water table experiences increasing suction with height but
remains sensibly saturated up to such a height that suction is effective

244 B. C. CHILDS AND N. COLLIS-GEORGE

in dewatering. This zone of saturation is known as the capillary fringe.
It is, from its nature, incapable of precise definition; indeed, it may
sometimes be incapable of precise location, since if the moisture char-
acteristic does not exhibit a sudden onset of dewatering at a more or less
well-defined suction, the soil will not exhibit a more or less well-defined
capillary fringe (Childs, 1945a.).

If we imagine the force of gravity to be very much increased (say a
thousandfold), the suction of a given column of water of height H is
increased in proportion, namely a thousandfold. In absolute units the
suction is gpH for a column of height H above the water table, where
g is the appropriate gravitational acceleration and p the density of water.
Hence the moisture profile which extended over a height // under gravity
would be telescoped into a height of only ///1000 in a field of 1000 g.
Such very strong fields may be achieved in the centrifuge, where cen-
trifugal force takes the place of gravity and is quite under control up
to the limits of the instrument. Russell and Richards' method of ob-
taining the moisture characteristic (1938) is a method of sampling such
a condensed profile at a constant height above the water table and with
increasing field strengths. The moisture equivalent (see IV-2.d) is the
average moisture content in such a condensed profile with a standardized
height (1-6 cm.) of soil column and a standardized field strength of
1000 times gravity. For a critical and detailed examination of the cen-
trifuge method see Schaffer et al. (1937).

/. The Soil Moisture Characteristic as a Tool for Measuring Soil
Moisture Content. For many practical purposes it is desirable to know
the soil moisture content rather than the pressure or suction. The
simplest method other than the extraction and drying of a sample is to
measure the soil water suction and to infer the moisture content by
reference to the moisture characteristic of the soil. The apparatus now
customarily called a tensiometer is a permeable membrane and mano-
meter similar in principle to that described under III-l.b. The mem-
brane is usually a tubular or conical probe of unglazed ceramic ware,
strong enough to be pushed into a hole prepared for it. When the water
in the manometer is in equilibrium with the soil water it reads the pre-
vailing suction. It is usual to calibrate the apparatus directly by
measuring the suction recorded at known moisture contents, using the
soil from the site where the probe will be buried. Among workers who
have developed such apparatus are Rogers (1935a), Richards (1942),
Kenworthy (1945), Colman et al. (1946), and Hunter and Kelley (1946).

A less direct method relies upon the moisture characteristic of an ab-
sorbent body buried in contact with the soil. Such absorbers have com-

THE CONTROL OF SOIL WATER 245

monly been plaster blocks and more recently fiberglass or nylon cells.
The soil will have a suction appropriate to its moisture content and
will transmit this suction to an absorber with which it is in equilibrium.
The plaster block will then have a moisture content appropriate to this
suction, in accordance with its own moisture characteristic. The mois-
ture content of the block may then be obtained by raising it to the
surface for weighing, afterwards replacing it (Davis and Slater, 1942),
or by measuring the electrical resistance or capacitance between elec-
trodes buried in the block, as recommended by Fletcher (1939), Ander-
son and Edlefson (1942a, 1942b), Bouyoucos and Mick (1940, 1947,
1948), Edlefson et al. (1942) ? Bouyoucos (1947, 1949), and Colman
and Hendrix (1949). These measured quantities are interpreted in
terms of soil moisture content by the aid of an initial direct calibration.
Slater and Bryant (1946) have compared some of these methods with
direct sampling, and report that each indirect method has its useful
range, tensiometers being useful for high moisture contents, weighed
plugs being accurate over a fairly wide range, while the resistance blocks
us developed at that time had a use for large surveys, where accuracy
was not & prime consideration. Thome and Russell (1947) have found
electrical capacitance a most unreliable criterion of soil moisture con-
tent, while Childs (1943) 1ms offered objections to this method which
do not appear to have been adequately answered by those who pro-
posed them (Anderson, 1943).

The general conditions affecting the accuracy of all soil moisture
measurements which depend upon the moisture characteristic are out-
lined below. Moisture content cannot be inferred with accuracy from a
measurement of suction pressure when considerable changes of the latter
have but little effect on the former, such as in the regions AB and CD
of the hypothetical curves of Fig. 1. This is a limitation imposed by
the soil and cannot be circumvented. In addition, if the absorbent block
is working in a region where its moisture content is not much affected
by changes of suction, insensitivity is imposed by the block. The solu-
tion is, of course, to choose a material, whether plaster of paris, fiber-
glass, nylon or what you will, suited to the range of suctions likely to
be encountered.

It may be mentioned here that other indirect methods of soil moisture
measurements have been proposed which do not involve the moisture
characteristic. Thus Shaw and Baver (1939) use thermal conductance
and Anderson (1943) and Wallihan (1945) the electrical conductance
of the soil itself to indicate moisture content. Haise and Kelley (1946)
and Cummings and Chandler (1940) have assessed the value of the
thermal method in comparison with plaster blocks, and agree in regarding

246 E. C. CHILDS AND N. COLLIS-GEORGE

it as being somewhat insensitive and limited to suctions of less than
four atmospheres.

2. Soil Permeability

a. Definition. If, in a column of soil of unit cross section, (e.g., one
square centimeter) water is flowing at a rate of Q ml. per sec., it may
be said to have an effective velocity of Q cm. per sec. If the column
occupies only a part of the length of a tube in which it is contained,
Q is the actual velocity with which the water travels in the approach
section before it enters the soil. The true velocity in the soil is not
amenable to discussion; we can never learn much about it and refer
always to the measurable effective velocity, v .

It requires a difference of potential between two points in such a
column to drive water from one to the other. It is known from experi-
ment that the effective velocity of flow in the column is usually propor-
tional to the potential difference, A</>, between two cross-sectional planes,
and inversely proportional to the distance, /, separating those planes.
These relationships may be written

v = -Kk<t>/l (III.3)

The negative sign indicates that the flow is in the direction opposite to
that in which </> increases. The quantity A</>/J is the potential gradient,
or the amount by which </> changes per unit distance along the column.
This law is called Darcy's law after its discoverer (1856). The constant
K gives the flow velocity for unit potential gradient, i.e., is a measure
of the readiness with which the soil permits the flow of water. It is
called the permeability of the soil to water.

Water flow is not always restricted to a single direction by confining
tubes, and for that reason a more general definition of permeability is
often desirable, of which equation III.3 is a particular case. In an un-
confined body of ground water we may connect together in imagina-
tion all points which have the same selected potential, and we shall find
that we have drawn a surface in the soil body; such a surface is called
an equipotential surface. A cross section of the column described above
is a case of such a surface. If we choose a slightly different potential
we shall draw another surface slightly separated from the first. Thus
we may choose intervals of potential and draw a series of separated
surfaces rather like the layers of an onion; these surfaces will be closer
together at some places than at others. These surfaces being at different
potentials, water will flow from one to another, and just as water flowing
down sloping ground will take the direction of steepest slope, so in the
ground water body it will take the direction of steepest potential gradient.

THE CONTROL OF SOIL WATER 247

This is naturally the direction in which it will reach the next equipoten-
tial surface in the shortest distance, namely at right angles to the surfaces.
The potential gradient at a point is the potential difference between
successive equipotentials divided by the shortest distance between those
equipotentials at that point, in the direction of that nearest distance.
That is what is meant by the vector notation grad<t>. The generalised
expression for Darcy's law then becomes

v = - K grad < (III.4)

It is necessary to point out that Darcy's law is a result of experiment
and is not universally true. If, for example, the potential gradient is
very high, the water may have a turbulent motion in the pores instead
of a laminar or streamline motion. Then, just as in a crowd that is
jostling instead of orderly, the flow velocity is less than it ought to be
according to Darcy's law. Soil water movements are rarely likely to
be the result of such excessive potential gradients. Then again, some
soils may have structural features which encourage flow in one direction
at the expense of others. Such soils are called anisotropic, and in these
the average direction of flow may be in a direction other than that of
the mean potential gradient. Such properties are fortunately no bar
to the solution of problems although they present complications. For an
exhaustive treatment of permeability the reader may be referred to
Muskat's book (1937).

6. The Measurement of the Permeability of Saturated Soil Samples.
Permeability is simply and precisely defined, and would, at first sight,
appear to be capable of simple and precise measurement. We have but
to pack the soil into a tube of known cross section, clamp it in position,
as in Fig. 3, and connect the ends to a source of water and a sink
respectively. To maintain saturation the pressure must be positive
everywhere, and to maintain flow we must have a higher potential at
the inflow end than at the outflow. In Fig. 3, the input potential is the
sum of pi (equals gph^ and the height component gpHi, while the out-
flow potential is similarly the sum of gph* and 0pff*. The difference
between these potentials divided by the length of the column, Z, is the
potential gradient; while the rate of collection of water at the sink in
the steady state, divided by the cross section of the tube, gives us v ,
and we require nothing more for the calculation of K. Engineers, indeed,
report permeability measurements in this way as a matter of routine.
In fact, many difficulties arise, and considerable literature is devoted to
the interpretation of measurements of this kind.

In the first place the material, if it is unconsolidated, must usually

248

E. C. CHILDS AND N. COLLIS-GEORGE

be retained between permeable diaphragms (gauzes, filters and the
like) and these absorb a part of the measured potential difference. The
junction between the sample and the diaphragm may develop a quite
anomalous resistance to flow. This difficulty may be surmounted by in-
serting manometers at points within the soil column, defining a new
length of path over which we know the potential gradient. In our ex-
perience such manometers are insensitive and sluggish on the one hand,
and capricious on the other. Then the measured flow rate for an un-

Constant

Head
Reservoir

Outflow

Fig. 3. Diagrammatic representation of an apparatus for measuring the per-
meability of a porous material, which is packed in the flow tube.

changing potential gradient shows a tendency to change. Smith et al.
(1944), for example, report a steady decrease of permeability due to
fungal growth in the voids, which they reduced by introducing toluene
into the flow system. Christiansen (1944), after a review of the litera-
ture, comments upon the errors introduced if care is not taken to elim-
inate trapped air. This air, it is claimed, is not swept out bodily by
the flowing water, and gives rise not only to false values of permeability
but also to changing values as the trapped air subsequently dissolves.
Later, Christiansen et al. (1946), used C0 2 to displace air before the
wetting of the soil, and while in this way they eliminated a rise of per-
meability to a peak value after a run of a few days, they did not elim-
inate a steady fall of permeability. Such a fall may reduce the
permeability to one thousandth of its initial value, as Christensen found

THE CONTROL OF SOIL WATER 249

(1944). This author also lists some factors which might be expected
to cause such a change. Fireman (1944) found that permeability is so
dependent upon soil structure and, therefore, upon changes during flow
which affect structure, that he regards permeability as among the best
criteria of soil structure. He includes trapped air, arbitrary packing,
the washing out of electrolytes and consequent modification of colloidal
properties, transport of the clay fraction and the growth of micro-
organisms as being among important causes of variation of permeability
during the course of its measurement. Pillsbury and Appleman (1945)
and Smith and Browning (1946) endorse the general opinion as to the
errors due to entrapped air and its consequent solution in the flowing
water. George (1948), on the other hand, emphasizes the errors due to
water releasing dissolved air and thus steadily unsaturating an initially
saturated soil sample.

A difficulty of quite another kind is that the preparation of a soil
column for a permeability measurement inevitably disturbs a structured
soil, so that the results may have little relation to permeability in the
field. Various implements have been described purporting to extract
undisturbed samples. It will be enough to refer here to that of Goode
and Christiansen (1945) and to the elaborate device of Donnan et al.
(1943). At best one is quite at the mercy of the soil type; a very few
stones are sufficient to invalidate the use of such tools. This difficulty
can be overcome in one way only, namely by the development of methods
of measuring permeability in the field, thereby eliminating the necessity
of collecting samples. Such methods might properly be discussed next,
but for the fact that the one most recent and least open to objection
cannot be described before we have dealt with the theory of drainage.
(See IV-3.e.)

c. The Measurement of the Permeability of Unsaturated Soil. The
permeability of saturated soil is relevant only to the flow of ground water,
i.e., the water at positive pressure below the water table together with
that small amount in the capillary fringe which is at small suctions.
Such conditions arc paramount in drainage problems. When we have
to discuss the infiltration of water in conditions of suction, such as ac-
company most phases of irrigation and even some phases of drainage,
then in general we have to deal with the passage of water through soil
which is only partially saturated.

Since water can only pass through pores which are full of water, an
unsaturated soil necessarily offers greater resistance to flow than a sat-
urated soil. This statement sometimes seems to surprise those who
know from experience that a dry soil usually soaks up water more readily

250 E. C. CHILDS AND N. COLLIS -GEORGE

than a wet one, but we have to remember that the wetting of an initially
dry soil involves the passage of water along a moisture profile, which may
create its own steep potential gradient. The permeability, which may
be low, is the velocity of flow per unit potential gradient, and the high
velocity of infiltration into a dry soil is due to the high potential
gradient, not to a high permeability. Confusion between overall infil-
tration rate and permeability is far too common. We may refer to a
paper by Horton (1940) for a discussion of the relationship between
the two concepts.

One may attempt to measure the permeability of unsaturated soils
by a steady flow method similar to that described under III-2.b, but
certain special difficulties present themselves. Since it is necessary to
have suction everywhere to ensure unsaturation, the inflow r and outflow
must be designed for the supply and withdrawal of water under suction.
This may necessitate the enclosure of the soil sample between mem-
branes of the type described under III-l.b, which will maintain the
sample in contact with water even under suction; such a solution of
the problem was adopted by Richards in his pioneering work (1931).
Then, since a potential gradient is necessary to produce flow, and since
such a gradient may involve a pressure gradient (in this case a suction
gradient), it would appear that a moisture content gradient is unavoid-
able, moisture content being determined by suction in accordance with
the moisture characteristic. We then have difficulty in determining the
moisture content at the particular spot where permeability is being
investigated, since one cannot sample the column without upsetting the
state of flow. Then too, one must be able to measure the potential
gradient over short lengths at reasonably uniform moisture content, since
otherwise the observed permeability cannot be related to moisture con-
tent. Richards (1931) used short soil columns at sensibly constant
uniform moisture content and measured the potential gradient with two
tensiometers inserted at fairly close spacing.

Moore's method (1939) was to achieve a steady flow state in an
upward direction; the soil column standing in free water at constant
pressure and losing water by evaporation from the upper end into a
room with controlled temperature and air circulation. Apertures were
provided in the cans containing the soil, both for the insertion of
tensiometers at frequent height intervals and for sampling for moisture
content. The rate of flow was necessarily slow, so that sampling did not
cause a serious disturbance of the flow state. Flow was permitted to
proceed until the steady state was reached, as indicated by constant
readings of the tensiometers, a proceeding which lasted several days for
the heavier soils used. At the steady state the following quantities

THE CONTROL OF SOIL WATER

251

were measured: (a) the rate of supply of water to the base, (b) the suc-
tions recorded by the tensiometers and (c) the soil moisture contents of
samples extracted from known heights. Thus at any height of the col-
umn, information was obtained both as to moisture content and
permeability.

Childs and George (1948, 1950) used a steady state method with down-
ward flow, and with fair control over the rate of flow, the moisture con-
tent and the potential gradient. They made use of the fact, pointed out

Computed

(Childs and George)

5 10 15

Permeability, C.G.S. x I0 9

Fig. 4. Experimental and computed curves of the variation of permeability of a
porous material with its moisture content.

by Childs (1945a), that, provided the soil column is long enough, a con-
siderable part of it is at uniform moisture content and suction, and is
therefore subjected only to the uniform gravitational potential gradient.
The uniformity of moisture content permits the use of simple means for
indicating it electrically, using the change of apparent capacitance of a
condenser of which the soil forms the dielectric. Water was supplied
at the upper surface at a controlled rate, this rate determining the
moisture content of the column at the steady state. By slanting the col-
umn away from the vertical various potential gradients were obtained,
down to one half the gravitational gradient. As an example of the de-
pendence of permeability upon moisture content, found in this way, we
present Fig. 4. The chief feature of all such curves is the rapid fall of

252 E. C. CHILDS AND N. COLLIS-GEORGE

permeability with quite small degrees of unsaturation ; permeability is
sensibly zero while there is still an appreciable moisture content.

The difficulties of steady state methods have resulted in attempts
to interpret the stages of transience, when the advance of water into
dry soil is introducing changes of moisture content everywhere. Bodman
and Colman (1943) have developed such a method, elaborated later by
Colman and Bodman (1944). They used a soil column contained in a
tube built up of short sections and provided with thin sliding diaphragms.
Water was supplied at the top and percolated downwards. At any de-
sired stage of penetration the diaphragms, which were intially withdrawn
to leave the column continuous, were quickly pushed in, and in this way
divided the column into a series of isolated sections. The moisture con-
ditions were thus "frozen" and could be determined at leisure, the
moisture content being determined directly and the suction being in-
ferred from a knowledge of the moisture characteristic. In the silty and
sandy loams studied water was found to move in such a way as to
produce an ever increasing depth of soil, wetted to an almost constant
percentage of saturation, separated by a well-marked water front from
soil not yet wetted. Thus the mean rate of application of water repre-
sented the mean rate of flow through this "transmission" zone, since, once
wetted, this zone had no further effective storage capacity. From the
nature of the experiment it was possible to find the permeability only
at the one degree of unsaturation (70 or 80 per cent of saturation) which
the transmission zone happened to present. However, if these authors
had calculated the rate of flow at a given section by computing the rate
of storage beyond it, they could have measured permeability over the
whole range of moisture contents in the flow column. They report that
the permeability of the sandy loam at 70 per cent saturation was only
one-fifth of that at saturation, but that the silt loam was very little
affected down to 80 per cent of saturation.

Bradfield and Jamieson~(1938) and Bendixcn and Slater (1946) have
used as flow columns the samples placed in sintered glass Buchner fun-
nels for moisture characteristic determinations. Noting that each stage
of moisture removal requires an appreciable time for attainment of
equilibrium, and that the time increases with each stage of clewatering,
they have related the permeability of the sample to the moisture content
at the stage under consideration. The method does not seem to lend
itself to precision, but indicates the general trend of the permeability-
moisture content relationship in a convenient way.

Christensen (1944) has compared Richards' and Moore's methods
with a radial flow method which he devised, using three Prairie soils as

THE CONTROL OF SOIL WATER 253

his materials, but perhaps the chief interest of this paper lies in its dis-
cussion of inconstancy and hysteresis exhibited by permeability.

d. The Permeability of Soil to Air. Certain causes of the change of
permeability of soil during the course of experiments may be avoided
by employing air as the flowing fluid, the relationship between the per-
meability to water and to air for the same flow path being expounded
in the standard texts. Richards (1940) has criticized such methods on
the valid grounds that the effects of water on soil are essential factors
affecting the flow of water in soil, and not just bothersome complications
of technique to be avoided. Smith and Browning (1947) make the same
point. In any case we must point out that experiments of this sort must
be carried out with precisely the same flow paths; if we are concerned
with the flow of water in saturated soil, we must measure the permeability
of quite dry soil to air (i.e., air saturated), and it may well be quite im-
possible to maintain the internal geometry of the wet soil when it is
dried.

The measurement of the air permeability of field soils is a different
matter; such permeability has importance for plant growth. It is com-
plementary to water permeability, air flowing in air-filled pores concur-
rently with water flow in water-filled pores. Such measurements hardly
fall within the scope of this review, but we may refer to recent advances
reported in a paper by Kirkham (1946).

e. The Relationship between the Permeability and the Physical Con-
stitution of Soil. The permeability of soil is a property conferred upon
it by a certain geometrical configuration of its solid, liquid and air com-
ponents ; if these are completely specified then the permeability is deter-
mined as a consequence. It therefore becomes important to be able to
relate permeability to more elementary soil properties, not only because
in this way we can be assured that we understand flow processes but
also because it might well happen that to infer the permeability from
a knowledge of the physical composition is an easier matter than an ex-
perimental determination of permeability. It is certainly to be admitted
that engineers prefer the converse proposition, namely that it is simpler
to measure the permeability and to infer, say, the specific surface of the
sand bed concerned, than to measure the specific surface directly (Car-
man, 1938, 1939; Carman and Arnell, 1948; Arnell and Hennebury,
1948). Such an approach is, however, fundamentally unsound (Childs
and George, 1948, 1950) , since one can conceive different models of differ-
ent specific surface having the same permeability; permeability is one of
the consequences of a specified physical makeup, not vice versa. Insofar

254 B. C. CHILDS AND N. COLLIS-GEORGE

as this converse thesis may give satisfaction in practice, it does so be-
cause of the somewhat limited range of porous bed types encountered,
these being usually structureless loose sands. As we shall see, the prop-
osition breaks down utterly when we have to deal with structured
materials such as soil.

We shall have little to say about the classical formulae relating per-
meability to physical constitution; these are in the textbooks. It is
common experience that permeability is related to soil texture, and that
soil texture may often by satisfactorily indicated by the mechanical
analysis or a related parameter, such as the specific surface U. Various
authors have produced related expressions for the dependence of perme-
ability on these particle characteristics; among those best known are
Kozeny (1927), Zunker (1933), Terzaghi (1925), and Fair and Hatch
(1933). Engineers have given considerable attention in recent years
to elaborating formulae of this type, taking into account such factors
as slip of the fluid at the solid surfaces, and we may refer readers to
papers by Carman (1947), Rigden (1943, 1947) and Lea and Nurse
(1939) without describing them in greater detail here. The inadequacy
of them all to account for the properties of soils, as distinct from struc-
tureless sands, may be demonstrated by reference to a single example.
Fair and Hatch's formula is

In this expression / is the porosity (volume of voids per unit apparent
volume of porous material), U is the specific surface (surface developed
by unit true volume of solids due to fineness of division) and A is a con-
stant of proportionality. The porosity / does not depend upon the fine-
ness of division of the solids, but only upon the closeness of packing,
and since close packing is resisted more by small particles than by large,
particularly if the smaller particles are surface active (i.e., colloidal), the
porosity of clay soils is usually rather higher, but not very much higher,
than that of sands. The specific surface depends very much upon the
size of the particle, since subdivision of a unit volume of solids clearly
increases the exposed surface. Hence, according to equation III.5 the
permeability of a clay should be negligible in comparison with a sand.
If now we compare two clay soils, one with well developed structural
fissures and the other a structureless mass, the structural fissures of the
former will increase the porosity but slightly and the specific surface
not at all. There should therefore be but little difference between the
two permeabilities according to equation IIL5. In fact, however, the
fissures are dominant in conferring permeability and that of the struc-

THE CONTKOL OF SOIL WATER 255

tured clay may be many hundreds or thousands of times that of the
puddled mass.

The failings of these early formulae are due to a misplaced emphasis
on the solid phase of the porous medium. Soil scientists, from the nature
of their material, have been much more aware of the dominant role
played by the voids, and since we have means of studying the voids
which are at least as simple as mechanical analysis, progress in relating
permeability to the configuration of the voids has been relatively rapid.
Baver (1938) made rough measurements of the permeabilities of a
variety of sands and soils with known moisture characteristics. He at-
taches great significance to the point of inflection of the moisture char-
acteristic; in his view it divides the capillary (higher suction) from the
noncapillary (lower suction) water. The larger pores holding non-
capillary water are significant for permeability, whilst the suction at the
point of inflection indicates whether the soil is characterized by large
or small pores. Thus an appreciable proportion of relatively larger
pores favors permeability, other things being equal, while a high suction
at the flex point inhibits permeability. He therefore defines the porosity
factor as the radio of noncapillary water to the logarithm of the suction
at the point of inflection, and shows experimentally that the permeability
of his samples was a function of this factor. Nelson and Baver (1940)
later elaborated this work, confirming earlier results and, in addition,
finding a correlation between permeability and the volume of pores
drained at a suction of 40 cm. of water, these, of course, being rather
large pores. Further empirical relations of this kind were proposed by
Smith et al. (1944), who sought to assess the contribution to permeability
of the various groups of pores present in a soil, instead of seeking to find
a particularly -significant group. They divided the voids into pores
emptied at suctions of less than 10 cm. of water (porosity contribution
/i), those emptied at suctions between 10 and 40 cm. of water (porosity
contribution / 2 ), those emptied at suctions between 40 and 100 cm. of
water (porosity contribution / 3 ), and finally all smaller pores, which had
no contribution to make to permeability. Their porosity factor, in the
Baver sense, is given as /j + j 2 /4 + 1s/ 10, the several terms indicating
the various weights to be assigned to the given pore groups. Bendixen
and Slater (1946) return to the idea of a correlation between permeability
and the volume of water drained at a single suction, introducing a time
factor. Thus they obtain a correlation with the water drained at a
suction of 60 cm. of water in a period of one hour.

The most recent development is due to Childs and George (1948,
1950). They used the whole range of the moisture characteristic to ob-
tain the pore size distribution, and from this calculated the probability of

256 B. C. CHILDS AND N. COLLIS-GEOHGE

occurrence of sequences of pore pairs of all the possible combinations of
sizes. Making an application of Poiseuille's equation which is perhaps
overbold, but is certainly customary in such circumstances, they calculated
the contribution to permeability of each group of sequences, and the over-
all permeability by summing up these contributions. This treatment per-
mitted the calculation of the permeability at any chosen moisture con-
tent, since for this purpose one has only to omit from the summation
all those contributions involving pores larger than the upper limit of
those filled with water at the selected moisture content. The variation
of permeability with moisture content which they computed agreed
reasonably with the results of their experiments, a test of their theory
which is rather rigorous since the type of pore size distribution changes
drastically with progressive unsaturation. The comparison of theory
with experiment which they present is reproduced in Fig. 4, and includes
also a comparison with a Kozeny type of expression as developed for the
purpose. These authors also discuss the ability of their treatment to
deal with structural and anisotropic permeability, but leave this field
for future study.

Recently, Brinkman (1947, 1948) calculated the permeability of
porous material on quite novel premises. He calculates the viscous force
exerted upon a particle embedded in a porous and permeable material
consisting of a packed mass of similar particles. The expression for per-
meability derived in this way is claimed to be satisfactory over a wide
range of porosities, and to cover the case where the particles themselves
have a finite permeability. The pore geometry is not explicitly involved.

3. The Diffusion of Water in Soil

When observing water movements the agronomist is usually more
conscious of a moisture gradient than of the potential gradient, which
is the fundamental cause of movement. To express the rate of water
movement in terms of the gradient of the moisture profile is to use the
language of diffusion ; the rate of movement per unit gradient of moisture
content is the coefficient of diffusion.

The coefficient of diffusion may be expressed in terms of concepts
which have already been established. The moisture profile gives in-
formation about the moisture content and the moisture gradient at any
point. The moisture content determines the permeability of the particu-
lar soil (see III-2.C and III-2.e) insofar as that is a function of moisture
content, and also the suction component of potential; thus the moisture
gradient settles the suction potential gradient. The permeability and po-
tential gradient uniquely determine the rate of water movement, which
is therefore determined in terms of the moisture gradient at the point.

THE CONTROL OF SOIL WATER

257

Hence the coefficient of diffusion may be calculated and expressed by a
curve as a function of moisture content. Fig. 5 shows curves for a sand
fraction and for a slate dust fraction (Childs and George, 1948, 1950).
For light soils the diffusion coefficient varies from high values for wet
soils, (implying a ready admission of water by the necessarily wetted sur-
face soil) to very low values for dry soils. As a consequence, the soil can
sustain steep moisture gradients for a long time when dry but can main-
tain only slight moisture gradients when wet. This is one of the reasons

I 40

J 30

S 20

Sand

ICT 7 KT* I<T S IO' 4 IO" 3 I<T 2 IO' 1

Diffusion Coefficient

10 IO 2

Fig. 5. Curves showing how the diffusion coefficients of porous materials vary
with moisture content.

why a deep soil settles down quickly after watering to a state of very
slow movement (an apparent equilibrium), characterized by a layer of
wet surface soil with a slight moisture gradient separated by a steepen-
ing moisture gradient (water front) from the lower dry soil; the moisture
content of the wet layer is a characteristic of the soil type and is only
slightly affected by other factors. The agronomist recognizes this
moisture content as the field capacity; hysteresis may also play a part
in this phenomenon (Schofield, 1935).

The coefficient of diffusion of water in heavy soils, complicated by
considerations of structure, has not been studied yet, but is low in all
circumstances of moisture content. For this reason, such soils even

258 E. C. CHILDS AND N. COLLIS-GEORGE

when wet can maintain steep moisture gradients for a long time, imply-
ing a very slow penetration of water applied at the surface. One should
certainly use caution in making assumptions as to the constancy of the
diffusion coefficient. Childs (1936, 1938) made such an assumption for
soils of silty loam and clay texture, based on some small experimental
evidence, and was able in this way to account for the main features of
the slow redistribution of water in a profile. Recently Kirkham and
Feng (1949) have shown that water penetration in lighter soils cannot
be accounted for by assuming a uniform diffusion coefficient, which, in
view of the above discussion, need occasion no surprise.

IV. DRAINAGE AND IRRIGATION
1. The Soil Water Balance Sheet

Water incident on the soil surface, whether in the form of rain, irri-
gation water or melting snow, penetrates in the appropriate circum-
stances described under III-3 to moisten a certain depth to the field
capacity. The nature and condition of the soil set a limit to the maxi-
mum possible rate of infiltration, and if the rate of application exceeds
this limit the excess will remain on the surface and flow either to natural
surface drainage channels or to local depressions. If this natural drain-
age is adequate, such excess will be lost to the soil.

In the intervals between the arrival of water there will be evapora-
tion from the surface. This will dry the surface and thereby bring it
to the condition in which it can maintain a steep moisture gradient
without much water movement (see III-3), so that the rate of evapora-
tion will soon be reduced to small proportions. Thus water which may
have entered readily, leaves by direct evaporation only reluctantly even
without considerations of gravity. The remainder of soil water will
be used by vegetation. If the whole of the stored water is thus used
before the next subsequent Catering, the soil starts again in its original
dry condition ; and if such a state of affairs persists, the region tends to
aridity, supporting only the vegetation for which the prevalent water
supply suffices. If vegetation with a greater water requirement is de-
sired, then there is a need for irrigation.

If, on the other hand, the water stored from one rainfall is not en-
tirely used before the arrival of the next, water will penetrate to greater
and greater depths. In such circumstances there is commonly at some
depth or other an impeding layer upon which a body of ground water
builds, bounded at its upper limit by the water table and capillary
fringe (see III-l.e). This ground water will tend to drain via natural
channels just as will surface water, but if, in spite of this tendency,

THE CONTROL OF SOIL WATER 259

the water table rises sufficiently near the surface to waterlog the root
zone of growing crops, then there is a need for drainage. It is also pos-
sible for persistent surface water to produce a drainage need because
of inadequate powers of infiltration even though the ground water be at
a safe depth. An area which presents a drainage problem one season
may well need irrigation at another, and both needs must be considered
on their merits. It is no solution to neglect drainage on the grounds
that there may be drought later.

In areas of marginally deficient rainfall, use may be made of the
nonreturn nature of soil water, i.e., of the relative readiness to accept
and reluctance to evaporate. By fallowing for a year to prevent water
use by growing vegetation, it is possible to store nearly the whole of that
year's rainfall and thus use 2 years' rainfall for one crop. This is the
dry farming system.

2. Irrigation

a. Statement of the Problem. Insofar as one may generalize about
any agricultural matter, the irrigation problem may be stated in simple
terms. One has a limited range of crops in mind, and water resources
which are not under complete control but are apportioned in accordance
with a policy. One must so use the soil properties as to temper the
resources to the crop. The crop is characterized by a certain water
requirement and a certain root range, which is to some extent under
control by choice of variety. At best, therefore, the depth of soil in
which water may usefully be stored is limited by the crop itself; at
worst it might be additionally limited by the presence of impeding soil
horizons or by a general soil impermeability which prevents adequate
penetration of water. The question that arises is, will the storable water
suffice to tide the crop over the waterless interval?

b. The Fundamental Solution. The progress outlined under III
enables us to explain the occurrence of a recognizable moisture content
known as the field capacity, but until the diffusion problem has been
completely solved we shall not be able to calculate what this moisture
content is, nor in what time a given depth of soil will attain it. Nor,
indeed, are we able to report any attempt to account on fundamental
lines for a lower limit of useful moisture content, the wilting point.
Nevertheless it is in this direction that the fundamental solution must
lie. When the solution is complete we shall be able to forecast, from a
knowledge of the moisture characteristic profile alone, the whole course
of penetration and redistribution of irrigation water in given circum-
stances, since the moisture characteristic reflects the pore size distri-

260 E. C. CHILDS AND N. COLLIS-GEORGE

bution, which in turn determines the permeability and diffusion
coefficient. Given also a knowledge of the wilting point, we shall be
able to forecast the storable water content. This concept of usable or
available water has been introduced by Haynes (1948).

c. The ad hoc solution. The complete fundamental solution is not yet
to hand. The alternative is to carry out experimental irrigations on
plots in the problem areas, and it was on these lines that the earlier
experiments such as those of Israelson (1918), Harris andTurpin (1917)
and Greene (1928) were carried out. Such experiments clearly presup-
pose the existence of a water supply; the expense of bringing water to
the area must be undertaken before the benefits can be assessed. The
method is therefore more suited to the expansion of an existing irriga-
tion scheme than to the inception of quite new schemes.

d. The Arbitrary Solution. The basic approach explains the phe-
nomenon of the field capacity. The lower limit, the so-called wilting
point, is not so explained, but its existence is recognized in a general
way by agronomists. While we cannot yet accurately deduce the field
capacity on fundamental grounds, considerable attention has been paid
to the search for arbitrary treatments which can be shown empirically
to reduce the soil to the field capacity. This approach to the problem
is quite analogous to the chemists' arbitrary nutrient extraction methods
(water, citric acid, buffered acetic acid, neutral salt solutions, etc.) for
estimating the nutrient status of a soil ; there is no suggestion that plants
really do extract their nutrients by employing such reagents, but only
a claim, based on experience, that the amounts of nutrient extracted by
these reagents can be correlated with crop welfare. In just the same
way there is no suggestion that arbitrary imposed conditions such as
controlled centrifuging or controlled suctions have any fundamental con-
nection with the phenomenon of field capacity; we have seen that the
matter is not so simple. But there is a suggestion that we may be able
to hit upon a treatment which will in fact bring the soil to a moisture
content which is near enough to the field capacity for the method to be
diagnostically useful.

The first of the arbitrary treatments was the well-known centrifuga-
tion of Briggs and McLane (1907), which defined the moisture equivalent,
a state which Veihmeyer and Hendrickson (1931) identified with the
field capacity. Later Veihmeyer and Hendrickson (1949) recognized
that the field capacity is not an equilibrium state. For that matter the
moisture equivalent is not an equilibrium moisture content but a moisture
profile (see III-l.e), with an average moisture content appropriate to a

THE CONTROL OF SOIL WATER 261

mean suction of 0.8 atmosphere, Veihmeyer admits that the field capaci-
ties of sandy soils are not well represented by the moisture equivalent;
but it must be remarked that neither very sandy soils nor the heaviest
clays are well suited to irrigation in any case.

Among those who use controlled suctions to achieve the field capacity,
Richards and Weaver (1943) recommend a suction of % atmosphere
if the moisture equivalent exceeds 22 per cent and % atmosphere for
soils of lower moisture equivalent. Colman (1947) has studied the
field capacities and moisture characteristics of a wide range of soils, and
shows that there is no universally applicable suction which will achieve
field capacity. Browning (1941) shows how the ratio of field capacity/
moisture equivalent varies with the moisture equivalent. More serious
than the variability of the correlation according to soil texture is the
fact that the field capacity of a soil of given texture varies according
to the circumstances ; if the soil is uniform to great depth the field capac-
ity is lower than if a shallow surface layer is underlain by, say, coarse
sand or gravel (Moore, 1939). The reason is clear. Water movement
slows as soon as the lower material reaches its field capacity, which being
coarse, it does at relatively low suctions. The upper layer, in equilibrium
with this low suction, has a higher moisture content than it would have
had if it could have reached its own characteristic field capacity cor-
responding to a higher suction. The layer of coarse material acts as an
impediment to penetration which may be almost as effective as an in-
herently impermeable layer. This effect is present to an extreme degree
in lysimeters as commonly designed. Colman (1946) has pointed out
that, unless the base is in the form of a large tensiometer, it may impose
a condition of zero suction and consequent saturation not existing in
the field. Furthermore the field capacity seems also to depend upon
the moisture content ruling before irrigation (Kraebel and Sinclair, 1940)
and upon the depth of penetration of water (Colman, 1944).

The question of the very existence of a definite wilting point seems
a vexed one. On the one hand Veihmeyer and his school regard the plant
as being able to extract water with undiminished ease right down to a
well-defined soil moisture content, at which permanent wilting sets in
(Veihmeyer et al., 1943) while Rogers (1935b), growing precisely the
same variety of strawberry, observes distress for lack of water at quite
low suctions, a result which is allied to earlier work of Powers (1922).
One can only say that there is here a contradiction of evidence which
indicates some undiscovered difference of technique, and such contra-
dictions must be resolved before discussion can serve a useful purpose.
Breazedale and McGeorge (1949) have recently eliminated much am-
biguity from wilting point determinations. Then again Wadleigh

262 E. C. CHILDS AND N. COLLIS -GEORGE

et al. (1947) show that a plant may be "trained" to withstand high
soil suctions without wilting by growing it in a saline solution at con-
trolled osmotic pressure. Magistad and Reitemeier (1943) have also
studied the effect of varying salt concentration at the wilting point.
The accepted belief is that the uptake of water by plants becomes
increasingly difficult as the soil suction increases, that wilting sets in
as soon as transpiration at the leaves exceeds the rate of water intake
at the roots, and that the range of suctions over which wilting becomes
evident or serious, corresponds to a soil moisture content range which is
narrow, the slope of the moisture characteristic being gradual for most
soils at large suctions. Thus wilting may set in gradually in terms of
suction, but suddenly in terms of decreasing moisture content, so that
a more or less well-defined moisture content may be recognized as a
"wilting point." It is in this sense that Botelho da Costa (1938) defined
a wilting point in terms of suction. It is with the moisture content at
the wilting point that the irrigation farmer is concerned. The difference
between his soil moisture content and the wilting point moisture content
is his "available water" (Allyn and Work, 1941). Reimann et al. (1945)
have also investigated the "availability" of soil water throughout crop
life. Arbitrary suctions for achieving the wilting point have been put
at about 16 atmospheres by Botelho da Costa (1938) and at 15 atmos-
pheres by Richards and Weaver (1943).

It may be mentioned here that infiltration of water into soil sup-
porting crops is not the only problem which may confront the irrigation
engineer. In some areas, as in parts of California, irrigation water is
drawn from wells sunk in alluvial fans, the ground water being replen-
ished by diverting hill streams into water-spreading areas. The paper
by Allison (1947) describes some problems presented by the decreasing
permeability of such spreading grounds.

3. Drainage

a. The Nature of the Problem. In its physical basis, the drainage
problem is simpler than the irrigation problem, since it is concerned
with the flow of ground water, i.e., in saturated soil with sensibly uniform
permeability. The sorts of problem that arise are presented in a general
way in Fig. 6, which indicates diagrammatically a layer of permeable
material underlain by an impermeable bed. The ground water body de-
scribed under I V.I is built up in the permeable layer, the water table be-
ing indicated by the line AB. The pressure built up under the water table,
together with gravity, forms a potential field which directs flow to an
exit BC where water emerges at zero hydrostatic pressure; positive pres-
sures cannot be built up since there is free flow over this surface, while

THE CONTROL OF SOIL WATER 263

there can be no suction or water would not emerge. Such a surface is
called a surface of seepage, and the rate at which water emerges at such
a "spring line" is determined by the configuration of the ground water
body and by the permeability of the layer. If rainfall exceeds this rate
of emergence the water table rises, while if the opposite is the case the
water table falls.

If the natural drainage is deficient, the water table may rise danger-
ously high everywhere, and this is particularly likely if the permeable

Fig. 6. A hypothetical drainage situation, a body of permeable material resting
on an impermeable bed. Rainfall produces ground water bounded by the water table
AB and surface of seepage BC. A drainage system at L controls local water and that
at M controls foreign water. The dotted line shows the controlled water table.

layer is thin. In such a case drains are laid in positions such as L to
control the rise of the water table in spite of local rainfall, and one may
describe such systems as local systems. On the other hand the water
table may only be embarrassing near the natural spring lines, where
water is inevitably near the surface, in which case one is more concerned
with cutting off the arrival of water from a distance than with guarding
against local rainfall ; one lays a line or two of drain, in a position such
as M . Such systems are often called interceptor drains, a poor descrip-
tion since all drains function by intercepting unwelcome water. We
prefer to call them foreign water drains. The problem is to determine,
for given soil and weather conditions and for a given drainage design,
the position and state of movement of the water table and the flow of

264 E. C. CHILDS AND N. COLLIS -GEORGE

water to the drains. The case of heavy clay land, which is characterized
commonly by a thin layer of structurally permeable topsoil over an im-
permeable bed at a depth of perhaps only a few inches, may be regarded
as a particular case of the more general circumstances described.

b. The Fundamental Solution. Darcy's law prescribes the flow con-
dition for a given distribution of potential, while it is a consequence of
the incompressible nature of water that it cannot be further stored in a
region which is already saturated; in any case it is not being stored if
the flow has reached a steady state. These two circumstances combined
result in a differential equation which the potential must obey. This is
the well-known Laplace's equation, about which we need only say here
that it cannot be solved at will for any particular problem, but that
certain well-known solutions of specific problems occur in all the text-
books. Further, while a solution is in general impossible by routine
analysis, it is simple to test whether a supposed solution is really valid.
The solution of a particular problem must provide us with a potential
distribution which satisfies Laplace's equation and which agrees with
the known distribution of potential or streaming conditions imposed at
certain boundaries by the particular circumstances of the problem. For
example, the water flux across the water table may be known and cer-
tainly the pressure here is zero; the potential at the drain may conven-
iently be the arbitrary datum. It is also known that the pressure at a
surface of seepage is zero, and that an impermeable bed coincides with
a streamline. These are all boundary conditions which a particular
problem might have to account for. When the potential distribution
is known, the streamlines may be known to provide a family of curves
which intersect the equipotentials at right angles, and may be drawn in
to complete the "flow net," whence, from a knowledge of the permeabil-
ity, the actual rates of flow everywhere may be computed.

c. Some Typical Problems. In Figure 7, we see an idealized case of
local drainage, and since an understanding of this case is a prerequisite
to understanding more complicated cases, it has been much studied. The
idealization lies in supposing the depth and permeability of soil to be
uniform, the depth, size and separation of drains to be similarly uniform,
and that the steady flow state has been achieved. Furthermore at this
stage it is necessary to make the unwarranted assumption that a uni-
formly distributed rainfall implies a uniformly distributed water flux at
the water table. The drastic assumptions made prior to 1934 have been
summarized in a paper by Russell (1934). We need only say that they

THE CONTROL OF SOIL WATER

265

amounted to assuming the final result, (the elliptical water table), which,
not surprisingly, failed to survive even the most casual scrutiny.

It is clear from the symmetry of the case that the section ABCDEF
in Fig. 7 is representative of the whole cross section. The shape of the
water table AF is unknown, but the rate of arrival of water there is
known ; DCBA is known to be a streamline and EF is another. This case
is certainly not one with a recognized solution, but it is at least pos-
sible to transform it into one with less unsettled boundaries. One way

Soil Surface

Impermeable Bed

Fig. 7. Diagram showing control of local water, such as at L in Fig. 6. In this
section three drains are shown of which ED is one. ABCD and FE are boundary
streamlines of the section ABCDEF, and the problem is to locate the water table
AF in any given circumstances.

to do this is by drawing the boundaries in a diagram which has for its
coordinates the flow velocity components in the x and y directions instead
of the position components x and y themselves, the result of being a
"hodograph" with boundaries of settled shape. Further transformations
result in a problem which is recognizable as a textbook solution provided
certain limitations are imposed. This is the method of Gustafsson
(1946). Childs (1947b) pointed out that the solution was less general
than was supposed. This, and other transformation solutions both for
drains and for open ditches, have been described by Wedernikov (1936,
1937, 1939) and by van Deemter (1949). Complete general solutions
have yet to be proposed.

One may apply a method of "images" provided the soil is taken as
being waterlogged to the surface (the only case in which the position of
the water table is known ab initio). The flow is then the same as would
be obtained if a second similar and similarly drained layer were placed
upside down on that shown in Fig. 7, water being fed in at the upper
drains and removed at the lower. This is a textbook case, and was pre-

266 E. C. CHILDS AND N. COLLIS-GEORGE

sented as a solution for artesian drainage by Kirkham (1940), and for
mixed rainfall and artesian water (Kirkham, 1945a). The latter result
confirmed sand model work by Harding and Wood (1941). Kirkham
(1947a) also discusses the effect of proximity of drains to the imperme-
able bed again with a waterlogged soil. Hooghoudt (1940) used the
treatment for the case of a water table not at the surface, which is hardly
justifiable.

For the complete solution of more general problems one may use
analytical methods to elucidate the boundary conditions and then proceed
by guided or inspired guessing, approaching a final true solution by trial
and error. Childs, in a series of papers (1943c, 1945a, 1945b, 1946, 1947a,
1948) , proceeded in this way, using the fact that Darcy 's law is formally
identical with Ohm's law as a basis for the construction of electrical
analogues of the drainage problems, these analogues being made to satisfy
the boundary conditions by trial and error. Such problems included
cases of foreign water drainage, nonsteady states, and mixed fresh and
salt groundwater zones, taking into account both the capillary fringe and
the flow through the upper unsaturated soil. The papers should be re-
ferred to for further information, since it is stressed that no general
formula can be presented; each problem must be studied on its merits.
Van Deemter (1949) has used Southwell's method of relaxation to achieve
a similar result by purely numerical methods, without appeal to experi-
mental techniques, and has taken into account the possible stratification
of the permeable layer. McClelland and Gregg (1944) use the graphical
reiteration method.

One may also proceed by building drained sand sections in boxes, and
examining the potential distribution by manometers. Hooghoudt (1937b)
presented some results of this kind, but appears to have identified the
pressure at a point with the height of the water table above. Insofar as
such pressure readings reflect the water table position, his results are in
agreement with the nonsteady state described by Childs (1947a). Some-
what similar experiments have been described by Donnan (1946) and by
Donnan et al. (1947), and, in connection with road subgrade drainage,
by McClelland (1943).

d. Field Experiments. For descriptions of ad hoc experimental drain-
age fields the reader may be referred to the Transactions of the Sixth
Commission of the International Society of Soil Science (1933,Z? and
1937,5) and to a paper by Weir (1928). Such experiments sought to
trace the variations of water table with weather, in plots with different
designs of drainage system. The water table observations were com-
monly made by the use of vertical perforated tubes forming well bores

THE CONTROL OF SOIL WATER 267

of small diameter. Childs (1945a) has shown that the insertion of such
tubes tends to upset the preexisting potential distribution and to impose
a local draw down on the water table, giving false results. The modern
practice is to insert several tubes to different depths, each being open
only at the bottom, and in this way to plot a potential distribution. The
complete flow net may thus be constructed, and the position of the water
table inferred, since it is the surface at which hydrostatic pressure is
just zero. Groundwater surveys of this type have been described by
Christiansen (1943), by Donnan and Christiansen (1944) and by Kirk-
ham (1947b). The latter studied the natural drainage of a hillside in
Iowa by such means. Such natural drainage surveys are likely to be
an important development of the future, since artificial drainage systems
can hardly be designed with certainty in the absence of information about
the water to be intercepted.

An increase of rainfall rate causes the water table to begin to rise,
and this rise constitutes a storage of ground water. The drainage rate,
which is determined by the ground water configuration, therefore increases
but slowly. Similarly, when the rainfall rate decreases, the drainage
rate follows the decrease but sluggishly. Ground water drainage is there-
fore characterized by a stability which is a consequence of the reservoir
action of the body of ground water. In very thin soils the reservoir
capacity is limited; in clay soils, which may be regarded as a few inches
of permeable soil over impermeable clay (see IV.Sa), there may be prac-
tically no stabilizing action and Childs (1943b) has described an experi-
mental drainage field with mole drains in clay, the outfall performance
having the characteristics of surface or storm water.

e. Field Measurements of Permeability. Again it is convenient to
refer to Russell's paper (1934) for an account of early work, much of
which was descriptive or, at best, semiquantitative. Modern methods
seem to be variations of one of these earlier methods, namely that which
makes use of the effect of pumped wells on the ground water. A pumped
well is essentially a drain, and the flow of soil water to it is a drainage
problem similar to those described above. Hooghoudt (1936, 1937a) has
been active in relating the rate of rise of water in the well after cessation
of pumping to the permeability of the soil in which it is sunk. The treat-
ment is rather an approximate one, but recently Kirkham (1945b) and
Luthin and Kirkham (1949) have studied the flow of ground water from a
tube open only at the end, giving a solution of Laplace's equation for a
spherical cavity, and discussing methods for solving other forms of cav-
ity. From this Kirkham gives formulae which are good approximations
to the relationship between soil permeability and the rate of flow of water

268 E. C. CHILDS AND N. COLLIS-GEORGE

from the tube to the ground water. The solutions are equally applicable
if the tube is first pumped dry and ground water then allowed to flow to
it. The method is thus analogous to the pumped well technique rather
than to surface infiltration techniques such as those of Freckman and
Janert and of Flodkvist, described by Russell (1934) , in that one is study-
ing flow in the zone of ground water and not in the overlying unsaturated
soil with unspecifiable conditions.

4. Engineering Aspects

This review would not be complete without a brief reference to some
of the problems which the flow of soil water presents to the engineer.
The fluctuations of the water table and of the volume of ground water
constitute a reservoir action. Where drainage consists of controlling the
ground water, drain performance is characterized by considerable stabil-
ity. Surface or storm water is, on the other hand, violently fluctuating,
and a sudden storm may throw loads on the rivers out of all proportion
to the total volume of water involved. Flood control and soil conserva-
tion arc therefore deeply concerned with infiltration, which contributes
to ground water at the expense of surface water. The soil conservation
stations of the problem areas of the United States, together with Experi-
ment Stations of the State Colleges and Universities, have been much
occupied with infiltration studies. The literature is so voluminous that
abstraction can be but arbitrary. We may refer to publications of the
Tennessee Valley Authority and to progress reports of the regional sta-
tions (see, for example, the report for the North-West Appalachian
Station, Zanesville, 1939). The general opinion is that surface water
is usually a result of an excessive rate of rainfall, there being ample pore
space for. its accommodation in time. Infiltration rate is suppressed by
a variety of factors, some of which we have not previously mentioned;
for example, some soils are not easily wetted (Jamieson, 1945). Then
again, unstable Foil may -suffer surface sealing due to lack of organic
matter, the sealing being due to the impact of rain on unprotected bare
soil. We refer to papers by Musgrave and Free (1936), Ellison and
Slater (1945), Duley and Domingo (1943), Kidder et al. (1943), Lewis
and Powers (1938), McCalla (1944, 1945), and to Parker and Jenny
(1945), without any pretence of doing more than pick at random. Some-
times infiltration cannot be expected to cope with rainfall, in spite of
the most favorable soil conditions. Wilson et al. (1946) have described
well designed terraces where such circumstances prevailed. Provision
must be made for watercourses protected against erosion ; such protection
may commonly be afforded by vegetation, and a comprehensive descrip-
tion of such means has been given by, among others, Ree and Palmer

THE CONTROL OF SOIL WATER 269

(1949). To deal more thoroughly with this aspect of the study of soil
moisture would be to invade the vast field of soil conservation.

REFERENCES

Alexander, L. T., Shaw, T. M., and Muckenhirn, R. J. 1936. Soil Sci. Soc. Am.

Proc. 1, 113-119.

Allison, L. E. 1947. Soil Sci. 63, 439-450.
Allyn, R. B., and Work, R. A. 1941. Soil Sci. 51, 307-319.
Anderson, A. B. C. 1943. Soil Sci. 56, 29-41.

Anderson, A. B. C., and Edlefson, N. E. 1942a. Soil Sci. 53, 413-426.
Anderson, A. B. C, and Edlefson, N. E. 1942b. Soil Sci. 54, 35-46.
Anderson, A. B. C., and Edlefson, N. E. 1942c. Soil Sci. 54, 221-232.
Arnell, J. C., and Hennebury, G. O. 1948. Can. J. Research 26A, 29-38.
Baver, L. D. 1938. Soil Sci. Soc. Am. Proc. 3, 52-56.

Bendixen, T. W., and Slater, C. S. 1946. Soil Sci. Soc. Am. Proc. 11, 35-42.
Bodman, G. B., and Colman, E. A. 1943. Soil Sci. Soc. Am. Proc. 8, 116-122.
Botelho da Costa, J. V. 1938. J. Agr. Sci. 28, 630-643.
Bouyoucos, G. J. 1947. Soil Sci. 64, 71-81.
Bouyoucos, G. J. 1949. Soil Sci. 67, 319-330.

Bouyoucos, G. J., and Mick, A. H. 1940. Soil Sci. Soc. Am. Proc. 5, 77-79.
Bouyoucos, G. J., and Mick, A. H. 1947. Soil Sci. 63, 455-465.
Bouyoucos, G. J., and Mick, A. H. 1948. Soil Sci. 66, 217-232.
Bradficld, R., and Jamieson, V. C. 1938. Soil Sci. Soc. Am. Proc. 3, 70-76.
Breazedale, E, and McGeorge, W. T. 1949. Soil Sci. 68, 371-374.
Briggs, L. J., and McLane, J. W. 1907. U.S. Dept. Agr. Bur. of Soils 45.
Brinkman, H. C. 1947. Applied Sci. Res. The Hague Al, 27-34.
Brinkman, H. C. 1948. Applied Sci. Res. The Hague Al, 81-86.
Browning, G. M. 1941. Soil Sci. 52, 445-450.
Buckingham, E. 1907. UJS. Dcpt. Agr. Bur. of Soils 38.
Carman, P. C. 1938. J. Soc. Chem. Ind. London 57, 225-234.
Carman, P. C. 1939. J. Soc. Chem. Ind. London 58, 1-7.
Carman, P. C. 1947. Nature 160, 301-302.

Carman, P. C., and Arnell, J. C. 1948. Can. J. Research 26A, 128-136.
Childs, E. C. 1936. /. Agr. Sci. 26, 527-545.
Childs, E. C. 1938. Soil Sci. 46, 95-105.
Childs, E. C. 1940. Soil Sci. 50, 239-252.
Childs, E. C. 1942. Soil Sci. 53, 79-92.
Childs, E. C. 1943a. Soil Sci. 55, 219-223.
Childs, E. C. 1943b. J. Agr. Sci. 33, 136-146.
Childs, E. C. 1943c. Soil Sci. 56, 317-330.
Childs, E. C. 1945a. Soil Sci. 59, 312-327.
Childs, E. C. 1945b. Soil Sci. 59, 405-415.
Childs, E. C. 1946. Soil Sci. 62, 183-192.
Childs, E. C. 1947a. Soil Sci. 63, 361-376.
Childs, E. C. 1947b. Acta Agr. Suecana 2, 353-356.

Childs, E. C. 1948. Proc. 2nd. Intern. Con/. Soil Mechanics 6, 150-153.
Childs, E. C., and Collis-George, N. 1950. Proc. Roy. Soc. London. 201 A, 392-405.
Childs, E. C., and George, N. C. 1948. Faraday Soc. Disc. 3, 78-85.

270 E. C. CHILDS AND N. COLLIS-GEORGE

Christensen, H. R. 1944. Soil Sci. 57, 381-391.

Christiansen, J. E. 1943. Agr. Eng. 24, 339-342.

Christiansen, J. E. 1944. Soil Sci. 58, 355-365.

Christiansen, J. E., Fireman, M., and Allison, L. E. 1946. Soil Sci. 61, 355-360.

Colman, E. A. 1944. Soil Sci. 58, 43-50

Colman, E. A. 1946. Soil Sci. 62, 365-382.

Colman, E. A. 1947. Soil Sci. 63, 277-283.

Colman, E. A., and Bodman, G. B. 1944. Soil Sci. Soc. Am. Proc. 9, 3-11.

Colman, E. A., Hanawalt, W. B., and Burck, C. R. 1946. /. Am. Soc. Agron. 38,
455-458.

Colman, E. A., and Hendrix, T. M. 1949. Soil Sci. 67, 425-438.

Cummings, R. W., and Chandler, R. F., Jr. 1940. Soil Sci. Soc. Am. Proc. 5, 80-85.

Darcy, H. 1856. Les fontaines publiques de la ville de Dijon. Dalmont, Paris.

Davidson, A. L. C., and Schofield, R. K. 1942. J. Agr. Sci. 32, 413-427.

Davis, W. E., and Slater, C. S. 1942. J. Am. Soc. Agron. 34, 285-287.

Day, P. R. 1942. Soil Sci. 54, 391-400.

van Deemter, J. J. 1949. Applied Sci. Res. The Hague A2, 33-53.

Donat, J. 1937. Trans. 6th Comm. Intern. Soc. Soil Sci. Zurich B, 423-439.

Donnan, W. W. 1946. Soil Sci. Soc. Am. Proc. 11, 131-136.

Donnan, W. W., and Christiansen, J. E. 1944. Western Construction News, Novem-
ber.

Donnan, W. W., Aronovici, V. S., and Blaney, H. F. 1947. Report on Drainage
Investigation in Irrigated Areas of Imperial Valley, California, U.S. Dept. Agr.
Soil Conservation Service.

Donnan, W. W., Aronovici, V. S., and Fox, W. W. 1943. Soil Sci. Soc. Am. Proc.
8, 367-371.

Duley, F. L., and Domingo, C. E. 1943. Soil Sci. Soc. Am. Proc. 8, 129-131.

Edlefson, N. E., and Anderson, A. B. C. 1943. Hilgardia 15, 133-134, 217-235,
277-283.

Edlefson, N. E., Anderson, A. B. C., and Marcum, W. B. 1942. Soil Sci. 54, 275-279.

Edlefson, N. E., and Smith, W. O. 1943. Soil Sci. Soc. Am. Proc. 8, 112-115.

Ellison, W. D., and Slater, C. S. 1945. Agr. Eng. 26, 156-157.

Fair, G. M., and Hatch, L. P. 1933. /. Am. Water Works Assoc. 25, 1551-1565.

Feng, C. L., and Browning, G. M. 1946. Soil Sci. Soc. Am. Proc. 11, 67-73.

Fireman, M. 1944. Soil Sci. 58, 337-353.

Fletcher, J. E. 1939. Soil Sci. Soc. Am. Proc. 4, 84-88.

George, N. C. 1948. Soil-Water Relationships. Thesis, Univ. Cambridge, England.

Goode, W. E., and Christiansen, J. E. 1945. Agr. Eng. 26, 153-155.

Gouy, M. G. 1910. J. de physique (4) IX, 457-468.

Greene, H. 1928. J. Agr. Sci. 18, 531-543.

Gustafsson, Y. 1946. Acta Agr. Suecana 2, 1-157.

Haines, W. B. 1930. J. Agr. Sci. 20, 97-116.

Haise, H. A., and Kelley, 0. J. 1946. Soil Sci. 61, 411-422.

Harding, S. W., and Wood, J. K. 1941. Soil Sci. Soc. Am. Proc. 6, 117-119.

Harris, F. S., and Turpin, H. W. 1917. /. Agr. Research 10, 113-135.

Haynes, J. L. 1948. J. Am. Soc. Agron. 40, 385-395.

Hooghoudt, S. B. 1936. Verslag. RijkslandbProefst., 's Grav. 42, 449-541.

Hooghoudt, S. B. 1937a. Trans. 6th Comm. Intern. Soc. Soil Sci. Zurich B, 42-57.

Hooghoudt, S. B. 1937b. Verslag. RijkslandbProefst., 's Grav. 43, 461-676.

Hooghoudt, S. B. 1940. Verslag. RijkslandbProefst., 's Grav. 46, 515-707.

THE CONTROL OP SOIL WATER 271

Horton, R. E. 1940. Soil Sci. Soc. Am. Proc. 5, 399-417.

Hunter, A. S., and Kelley, O. J. 1946. Soil Sci. 61, 215-217.

Israelson, O. W. 1918. J. Agr. Research 13, 1-36.

Jamieson, V. C. 1945. Soil Sci. Soc. Am. Proc. 10, 25-29.

Kenworthy, A. L. 1945. Soil Sci. 59, 397-404.

Kidder, E. A., Stauffer, R. E., and Van Doren, C. A. 1943. Agr. Eng. 24, 155-159.

Kirkham, D. 1940. Trans. Am. Geophys. Union 21, 587-594.

Kirkham, D. 1945a. Trans. Am. Geophys. Union 26, 393-406.

Kirkham, D. 1945b. Soil Sci. Soc. Am. Proc. 10, 56-68.

Kirkham, D. 1946. Soti S<-i. Soc. Am. Proc. 11, 93-99.

Kirkham, D. 1947a. Soil Sci. So<*.. Am. Proc. 12, 54-59.

Kirkham, D. 1947b. Soil Sci. Soc. Am. Proc. 12, 73-80.

Kirkham, D., and Feng, C. L. 1949. Soil S-i. 67, 29-40.

Kozeny, J. 1927. Sitzber. Akad. Wiss. Wien 136A, 271-306.

Kraebel, C. J., and Sinclair, J. D. 1940. Trans. Am. Geophys. Union 21, 84-90.

Langmuir, I. 1938. /. Chem. Phys. 6, 873-896.

Lea, F. M., and Nurse, R. W. 1939. J. Soc. Chem. Ind. London 58, 277-283.

Learner, R. W., and Lutz, J. F. 1940. Soil Sci. 49, 347-360.

Lewis, M. R., and Powers, W. L. 1938. Soil Sci. Soc. Am. Proc. 3, 334-339.

Luthin, J. N., and Kirkham, D. 1949. Soil Sci. 68, 349-358.

McCalla, T. M. 1944. Soil Sci. Soc. Am. Proc. 9, 12-16.

McCalla, T. M, 1945. Soti Sci. Soc. Am. Proc. 10, 175-179.

McClelland, B. 1943. Proc. Highway Research Board 23, 469-487.

McClelland, B., and Gregg, L. E. 1944. Proc. Highway Research Board 24, 364-376.

Magistad, O. C., and Reitemeier, R. F. 1943. Soil Sci. 55, 351-360.

Moore, R. E. 1939. Hilgardia 12, 383-426.

Musgrave, G. W., and Free, G. F. 1936. /. Am. Soc. Agron. 28, 727-739.

Muskat, M. 1937. Flow of Homogeneous Fluids Through Porous Media. Mc-
Graw-Hill, New York.

Nelson, W. R., and Baver, L. D. 1940. Soil Sci. Soc. Am. Proc. 5, 69-76.

Parker, E. R., and Jenny, H. 1945. Soti Sci. 60, 353-376.

Pillsbury, A. F., and Appleman, D. 1945. Soil Sci. 59, 115-123.

Powers, W. L. 1922. Soti Sci. 14, 159-165.

Ree, W. O., and Palmer, V. J. 1949. U. Dept. Agr. Soil Conserv. Service Bull. 967.

Reimann, E. G., Van Doren, C. A., and Stauffer, R. S. 1945. Soti Sci. Soc. Am.
Proc. 10, 41-46.

Reitemeier, R. F., and Richards, L. A. 1943. Soil Sci. 57, 119-136.

Richards, L. A. 1931. Physics 1, 318-333.

Richards, L. A. 1940. Soti Sci. Soc. Am. Proc. 5, 49-53.

Richards, L. A. 1942. Soil Sci. 53, 241-248.

Richards, L. A. 1949. Soil Sci. 68, 95-112.

Richards, L. A., and Campbell, R. B. 1948. Soti Sci. 65, 429-436.

Richards, L. A., and Fireman, M. 1943. Soil Sci. 56, 395-404.

Richards, L. A., and Weaver, L. R. 1943. Soti Sci. 56, 331-339.

Richards, S. J. 1938. Soti Sri. Soc. Am. Proc. 3, 57-64.

Rigden, P. J. 1943. /. Soc. Chem. Ind. London 62, 1-4.

Rigden, P. J. 1947. J. Soc. Chem. Ind. London 66, 130-136.

Rogers, W. S. 1935a. J. Agr. Sci. 25, 326-334.

Rogers, W. S. 1935b. Ann. Kept. East Mailing Res. Sta. England, pp. 111-120.

Russell, J. L. 1934. J. Agr. Sci. 24, 544-572.

272 B. C. CHILDS AND N. COLLIS-GEORGE

Russell, M. B. 1941. Soil Sci. Soc. Am. Proc. 6, 108-112.

Russell, M. B. 1945. Agr. Eng. 26, 292.

Russell, M. B., and Richards, L. A. 1938. Soil Sci. Soc. Am. Proc. 3, 65-69.

Schaffer, R. V., Wallace, J., and Garwood, F. 1937. Trans. Faraday Soc. 33, 723-734.

Schofield, R. K. 1935. Trans. 3rd Intern. Cong. Soil Sci. of Oxford II, 37-48.

Schofield, R. K. 1938. Trans. 1st Comm. Intern. Soil Sci. Soc. Bangor, Wales A,

38-45.

Schofield, R. K. 1946. Trans. Faraday Soc. 42B, 219-225.
Schofield, R. K. 1948. Discuss. Faraday Soc. 3, 129.

Schofield, R. K., and Botelho da Costa, J. V. 1938. J. Agr. Sci. 28, 644-653.
Shaw, B., and Baver, L. D. 1939. Soil Sci. Soc. Am. Proc. 4, 78-83.
Slater, C. S., and Bryant, J. C. 1946. Soil Sci. 61, 131-155.
Smith, R. M., and Browning, D. R. 1946. Soil Sci. 62, 243-253.
Smith, R. M., and Browning, D. R. 1947. Soil Sci. Soc. Am. Proc. 11, 21-24.
Smith, R. M., Browning, D. R., and Pohlman, G. G. 1944. Soil Sci. 57, 197-213.
Swanson, C. L. W., and Peterson, J. B. 1942. Soil Sci. 53, 173-185.
Terzaghi, C. 1925. Eng. News Record 95, 832-836.

Thorne, M. D., and Russell, M. B. 1947. Soil Set. Soc. Am. Proc. 12, 66-72.
Veihmeyer, F. J., and Hendrickson, A. H. 1931. Soil Sci. 32, 181-193.
Veihmeyer, F. J., and Hendrickson, A. H. 1949. Soil Sri. 68, 75-94.
Veihmeyer, F. J., Edlefson, N. E., and Hendrickson, A. H. 1943. Plant Physiol. 18,

66-78.
Verwey, E. J. W., and Overbeek, J. Th.G. 1948. Theory of the Stability of Lyo-

phobic Colloids. Elsevier, New York.

Wadleigh, C. H., Gauch, H. G., and Strong, D. G. 1947. Soil Sci. 63, 341-349.
Wallihan, E. F. 1945. Soil Sci. Soc. Am. Proc. 10, 39-40.
Wedernikov, W. W. 1936. Compt. rend. 202, 1155-1157.
Wedernikov, V. V. 1937. Z. angew. Math. Berlin 17, 155-168.
Wedernikov, V. V. 1939. Compt. rend. acad. sci. U.R.8.S. 23, 335-337.
Weir, W. W. 1928. Hilgardia 3, 143-152.
Wilson, H. A., Riecken, F. F., and Browning, G. M. 1946. Soil Sci. Soc. Am. Proc.

11, 110-118.
Zunker, F. 1933. Trans. 6th Comm. Intern. Soc. Soil Sci. Groninoen B, 18-43, 187.

Preservation and Storage of Forage Crops

R. B. MUSGRAVE AND W. K. KENNEDY

Department of Agronomy, Cornell University, Ithaca, New York

CONTENTS

Page

I. Introduction 274

II. Measurements of Changes in Quality During Preservation and Storage . 275

1. Energy and Protein 275

2. Factors Governing the Content of Energy Sources in Harvested

Forages 276

a. Stage of Growth 276

b. Respiration 276

c. Leaching of Nutrients 276

3. Correlations of Energy Sources with Protein, Fiber, and Carotene . 278

III. Silage 279

1. Problems of Silage Making 279

a. Characteristics of Silage 280

b. Principles of Silage Making 280

c. Comparison of Carbohydrate and Protein Crops for Silage . 281

d. Detrimental Effects of Poor Quality Silage 281

2. The Ensilage Process 282

a. Chemical Changes 282

b. Control of Undesirable Fermentation 284

(1) Lowering pH 284

(a) Lactic Acid Stimulation 285

(b) Direct Acidification 287

(2) Sterilization 288

(3) Complete Exclusion of Oxygen 289

c. Losses of Dry Matter and Nutrients 290

3. Questions Needing Further Research 291

IV. Field-Cured Hay 294

1. Factors Influencing Rate of Drying 294

a. Time of Day to Cut 295

b. Raking, Cocking, Crushing and Tedding 295

2. Field Losses of Dry Matter and Digestible Nutrients 296

3. Changes in Vitamin Content during Field Curing 297

4. Storage Losses 297

a. Carotene 297

b. Heating 297

c. Preservatives 299

V. Barn Hay Drying 299

1. Introduction 299

2. Basic Barn Drier Designs 300

972

274 R. B. MUSGRAVE AND W. K. KENNEDY

Page

3. Operation of Barn Driers 300

a. Moisture Content 300

b. Density 301

c. Supplemental Heat 302

4. Dry Matter and Nutrient Losses during Barn Curing 303

5. Fungicides 304

VI. Artificial Drying 304

1. Types of Driers 304

2. Losses 304

3. Limitations 304

VII. Experiments Comparing Silage, Barn-Cured and Field-Cured Hay . . 306

VIII. Conclusions 309

References 311

I. INTRODUCTION

Forage plants are more widely distributed than any other group of
crops. They are grown almost exclusively in extensive areas where other
crops either cannot be grown because of climatic conditions or are not
grown because of transportation costs and other economic factors. They
are grown elsewhere in competition with grain and other cash crops to
supply forage for needed livestock production. They are also grown to
maintain the yields of cash crops by reducing erosion, by increasing
organic matter and nitrogen, by improving soil structure, and by aiding
in the control of weeds, diseases, and insects.

Regardless of the reason why forage crops are grown, usually one or
more harvests per year must be removed and placed in storage if these
crops are to be a part of an economic system of farming. Storage is
necessary in order to provide winter feed and also because rapid deteri-
oration takes place in the field if crops are not harvested and preserved
at the proper stage. In -processing crops for storage, and during the
storage period, appreciable losses of feeding value usually occur. The
magnitude of these losses is usually much larger than the increased pro-
duction resulting from genetical improvement and the adoption of better
management practices; but more emphasis has been placed on research
leading to increases in the yield and quality of forage than to decreasing
preservation and storage losses.

It is the purpose of this paper to point out the relative importance
of the constituents of quality which must be examined in determining the
magnitude of losses incurred during preservation and storage. The aim
is also to describe the various methods of preservation and to evaluate
the degree to which each preserves the original quality.

PRESERVATION AND STORAGE OP FORAGE CROPS 275

II. MEASUREMENTS OF CHANGES IN QUALITY
1. Energy and Protein

Feeding standards for livestock are based on two groups of nutrients,
energy-supplying and protein-supplying. An energy source is required
for all animal body functions leading to maintenance, growth and pro-
duction. Therefore, it is needed in much greater quantities than is
protein. Ferguson (1949) and Lewis and Eden (1949), discussing dairy
rations for Britain, point out that the prime need in quality roughage
is a high content of energy sources.

The concentration of energy sources in hay or silage is important
because high producing animals require larger quantities of energy sup-
plying constituents, or total digestible nutrients, than are contained in the
quantity of good roughage they are able to consume. Therefore, hay or
silage must be supplemented with grain. Any decrease in the energy
value of the roughage must be compensated by feeding even more grain.
Energy in grain is more than twice as expensive as energy in hay, and
the cost of the grain necessary to replace a decrease in energy content of
the roughage must be used to evaluate the economic loss occurring during
preservation and storage.

Protein concentration and its production per acre have been used
much more frequently than has energy content in attempting to evaluate
the feeding value of roughages. Protein-rich concentrates frequently are
more expensive per unit weight than high energy feeds. Because larger
quantities of energy sources must be fed than protein, farmers paid on
the average more than six times as much for energy sources as for protein
in the United States during the 12-year period of 1937-1948. This state-
ment is based on the formula derived by Watson (1939) for determining
the relative value of protein and starch equivalents in two feeds, and on
U. S. prices as reported in Crops and Markets (1949). During the pre-
ceding 12-year period this ratio was even greater. On the English market
Watson (1939) found the starch equivalents, the European measure of
energy-bearing foodstuffs, to be fully equal in cost to protein equivalents.

Protein, as feeding standards indicate, is a necessary nutrient in the
ration of roughage-consuming ruminants and should be supplied in defi-
nite ratios with energy-bearing foodstuffs, the exact ratios depending on
the kind, age, and production level of the animal. Farm-produced feeds
are often too low in protein to meet ration specifications. However, this
deficiency usually can be corrected at less net expense than can a de-
ficiency in total digestible nutrients.

When forage crops are harvested at an early stage, both energy con-
stituents and protein content are high. The protein content may be high

276 B. B. MUSGBAVE AND W. K. KENNEDY

enough to eliminate the need for supplemental feeding of this nutrient,
but grain still is required to meet the energy needs of high-producing
animals. Hence, in determining the value of losses during preservation,
comparisons in energy content must be given more weight than protein
comparisons.

2. Factors Governing the Content of Energy Sources in
Harvested Forages

a. Stage of Growth. Forages which carry the highest percentage of
energy constituents come only from immature crops which are highly
digestible. Sotola (1941) found that the dry matter in bromegrass,
Bromus inermis, Leyss, contained 80 per cent of total digestible nutrients
when harvested at the 6- and 12-inch stages. This value was only 55 per
cent when the crop had matured to the late flowering stage. From growth
curve data (MacDonald, unpublished) and the appropriate digestible
nutrient content, it can be calculated that the amount of total
digestible nutrients per acre increases rapidly to just before heading.
The rate of increase then starts declining and no increase occurs in the
amount of total digestible nutrients during or after flowering. This in-
dicates that the general practice of harvesting forage crops at the flower-
ing stage secures the maximum yield of total digestible nutrients.

If the values of the acre yields at the various growth stages are
determined in a dairy ration, however, it is found that when the crop
is harvested just prior to heading its acre value is more than 50 per cent
higher than when harvested at the flowering stage. Since a 1,000-lb.
cow producing over 12 or 15 Ibs. of milk daily requires some grain (Wood-
ward, 1939) and the energy in grain is about 2.5 times more expensive
than the energy in hay, the value of the higher energy content of early
cut hay which replaces the grain in the ration more than offsets the
higher total energy production obtained by later cutting.

Woodman and Evans J 1935) have also shown that there is a great
decrease in digestibility of alfalfa, Medicago sativa L., as it matures.
The fact is inescapable that high quality roughage must come from an
immature crop. The acre yield of such forage may be lower but its feed-
ing value in a dairy ration is higher than that cut later if one considers
the amount of supplemental grain feeding required.

b. Respiration. Hay and silage may decrease in energy value in sev-
eral other ways besides being allowed to mature beyond the vegetative
stage. Respiratory activity present in the plant continues after it is
cut. The respiration of leaves from detached wheat, Triticum vulgare
Vill. during a 6-day starvation period, has been measured by Duff and

PRESERVATION AND STORAGE OF FORAGE CROPS 277

Forward (1949). Expressing their data on the rate of carbon dioxide
evolution in terms of pounds of dry matter respired per ton of 14 per
cent moisture hay per day, the rate of loss during the first hour of
starvation at 22.2C. was 130 Ibs., but this dropped to about 30 Ibs. at
the end of the first day. Thereafter, the rate rose slowly to the fifth day
when the dry matter loss was 50 Ibs. per day. The rate again fell to 30
Ibs. at the end of the sixth day when observations ceased. This experi-
ment was performed with green tissue maintained at its original high
moisture content in a darkened moist chamber during the starvation
period. The time curve of dry matter losses from detached leaves
described above is very similar in shape and magnitude to those reported
by Yemm (1935) for several other plant species. It is also similar to
loss curves obtained when hay plants were dried rapidly to 20 per cent
moisture and held at that level for 5 or 6 days. Rehydrating artificially
dried hay to 20 per cent permits respiration losses similar in quantity to
those upon dehydration to 20 per cent but the curve starts out at near
zero rate and builds up slowly, (very slowly if the mold spore population
is low) to a 30-60 Ib. peak (Musgrave and Dawson, unpublished data).
While the above losses were all observed under aerobic conditions, Phil-
lips et al. (1935) observed this type of loss curve when they subjected
moistened alfalfa to anaerobic conditions.

The respirational loss upon rehydration to 13.5-25 per cent moisture
is due to a microbial population composed chiefly of only 3 or 4 fungus
species (Galloway, 1935; Wright, 1941). The major loss in the case of
dehydration to 20 per cent moisture is very likely from the same source
but Duff and Forward (1949) attributed the losses from starving leaves
held at high moisture content entirely to the respiratory activities of
the plant cells. Although there is a remarkable similarity in the rates
and amounts of respiration from plant materials under the three different
moisture conditions, these workers have shown that the greatest deteriora-
tion will be found in the plants with the highest percentage of growing
tissue and the highest content of simple carbohydrates.

According to Yemm (1937) there is a steady transformation of in-
soluble nitrogen compounds to soluble compounds during the early stages
of the starvation of a leaf. When the carbohydrate suppty becomes
limiting for respiration there is a rapid accumulation of ammonia.
Fleischmann (1912) observed these same changes when whole plants
were slowly cured into hay, and measured high losses of nitrogen espe-
cially when leaching followed slow curing.

c. Leaching of Nutrients. The leaching loss from hay has been
measured under very few conditions where it could be separated from

278 B. B. MUSGRAVE AND W. K. KENNEDY

other losses. Guilbert et al. (1931) working with field-cured hays found
leaching losses of nitrogen-free extract to range from 6 to 35 per cent,
but organic matter losses were small. These losses were created by
laboratory methods approximating complete water extraction. Such leach-
ing has not been related to that from rainfall of various types and
intensities. Elliott (1947) found that one inch of simulated rain over a
one-hour period leached 1.9, 0.3, and 4.7 per cent of the dry matter from
rapidly-dried hay of timothy, Phleum pratense L., red clover, Trijolium
pratense L., and alfalfa, respectively. Although the timothy samples
showed no significant loss of dry matter, its digestibility as measured by
rabbits did drop 20.2 per cent due to leaching and subsequent artificial
drying at 70C.

Wiegner (1925) found that the dry matter lost during field weathering
was practically 100 per cent digestible. These losses included leaching,
leaf and respiration losses. Sotola (1933) has shown that leaves from
slowly cured alfalfa are about 66 per cent digestible. Putting Wiegner's
conclusions and Sotola 's findings together would indicate that leaching
may have harmful effects in addition to nutrient removal as is also sug-
gested by Elliott's data (1947).

Kellner (1915) and others have shown that as digestible nutrients
become less concentrated in a feed due to any cause the energy required
for digestion increases. In hay, according to Wiegner (1925), this loss
will amount to about one-third of the total of energy losses from all
field and stack sources.

3. Correlations of Energy Sources with Protein, Fiber, and Carotene

Although the energy derivable is the basis of determining the value
of feed it is not an easy determination to make. Expensive feeding trials
requiring considerable time are necessary. Therefore, many attempts
have been made to find a chemical determination which could be corre-
lated with feeding value -(Oampton and Jackson, 1944) . The use of
protein content as a measure of feeding value was found unsatisfactory
for naturally occurring hays (Watson and Ferguson, 1937). On the
other hand it is still used as an indicator of quality in artificially dried
grass in Britain (Ferguson, 1949). Crampton and Jackson (1944) work-
ing with green and dried pasture clippings found that protein content
was of no value in predicting the seasonal variation in digestibility of
pasture grasses. Fiber was equally poor as an index of nutritive value.
Yet for 167 European and Indian hays Watson and Ferguson (1937)
report fair inverse agreement between nutritive value and a figure ob-
tained by adding to the protein content twice the fiber content. A

PRESERVATION AND STORAGE OF FORAGE CROPS 279

straight line negative relationship exists but no statistical proof of the
significance of the regression coefficient is offered. Crampton and Jack-
son (1944) and others could not predict digestibility of forage from its
lignin content as determined by the Crampton and Maynard method
(1938). Ellis et al (1946), however, improved the method for lignin
determination and suggested the possibility of using lignin content as a
means of estimating the digestibility of forages. Using this improved
method Forbes et al. (1946), Swift et al (1947), and Kane et al (1949)
have found lignin to be indigestible and have had success in determining
the digestibility of forages from the ratio of lignin in the feces to that
in the forage consumed. Saltonstall (1948) found a close relationship
between digestibility of pasture herbage and its lignin content but the
recovery of the lignin in the feces was not complete. Since lignin as de-
termined by existing methods may be partially digestible, and since a
given level of lignin does not result in the same digestibility of different
forage crops or of the same forage crop grown under different environ-
mental conditions, its content is not an accurate index of the nutritive
value of forages.

Carotene has been tried as an indicator of nutritive value undoubtedly
because it has been found to be destroyed by the same agents as are the
energy-yielding constituents, namely, maturity, leafiness, weathering and
oxidation. One of these causes alone, however, such as oxidation in
storage, can reduce the carotene content to near zero while only a very
slight change in energy content takes place (Camburn et al., 1944). Fer-
guson (1949) concludes that although a high carotene content in dairy
hay will maintain the vitamin A content of dairy products through the
winter, the farmer is paid no premium for this practice at the present
time. Therefore, little value is placed on carotene content above that
needed for the health of the stock. This amount can be as low as 2 p.p.m.
and yet meet minimum requirements for milk cows building up a store
of vitamin A each year while on pasture.

III. SILAGE
1. Problems of Silage Making

The making of silage is a possible method of preserving excess pas-
ture and the first cutting from meadow crops when climatic conditions
are unfavorable for field curing of hay. It also is the most satisfactory
method of storing weedy crops and plants with coarse stalks, such as
corn or sorghums, which may result in considerable waste if fed as hay
or fodder.

280 B. B. MUSGRAVE AND W. K. KENNEDY

a. Characteristics of Silage. Silage is a high quality succulent feed
with a clean acid odor and taste. It is firm in texture and green to
brown in color. Morrison (1948) states that good silage is a very
palatable feed and animals will generally eat more roughage dry matter
if fed a mixture of silage and hay than if fed hay alone. This feeding
practice not only increases animal production but may result in a con-
siderable saving in the amount of concentrates required for good pro-
duction. Shepherd et al. (1948) listed the following standards upon
which silage can be judged. These were set up by the American Dairy
Science Association Committee on silage methods in 1942.

Very good: clean, acid odor and taste, no butyric acid, no mold,
sliminess or proteolysis, acid pH of 3.5 to 4.2, ammonia nitrogen
less than 10 per cent of total nitrogen. Good: acid odor and taste,
trace only of butyric acid, acid pH of 4.2 to 4.5, ammonia nitrogen
10 to 15 per cent of total nitrogen. Fair: some butyric acid, slight
proteolysis or some mold, acid pH 4.5 to 4.8, ammonia nitrogen 15 to
20 per cent of total nitrogen. Poor: high butyric acid, high proteoly-
sis, sliminess or mold, acid pH above 4.8, ammonia nitrogen about
20 per cent of total nitrogen.

6. Principles of Silage Making. The principles of silage making con-
sist of handling the material so that aerobic activity will be small and
undesirable anaerobic activity will be prevented. In order to decrease
aerobic losses the green fodder is placed in a suitable enclosure or stack
and compressed, thereby limiting the amount of entrapped oxygen and
excluding the entrance of oxygen into the interior of the silage. Greater
compression and exclusion of air is accomplished by fine chopping, care-
ful packing and the addition of more succulent material or water. Set-
tling and further compaction occurs with the death of the cells and as the
plant tissue loses its tugor and becomes flaccid. Wherever there is an
exposed surface, mold growth results. Within a few weeks the mold
growth permeates 6 to 12 inches into the silage before the shortage of
oxygen inhibits further fungal activity (Miller and Golding, 1949) . The
molds on the surface of compact silage have such a high requirement for
oxygen that any inward movement of oxygen by diffusion is prevented
by the seal set up by the molds (Watson, 1939).

If the entrance of air is not prevented, the aerobic organisms, espe-
cially molds, grow very rapidly and may result in such severe heating
that the hay will actually char, with tremendous losses in feeding value
(Am&s and Woodman, 1922).

To prevent undesirable anaerobic respiration resulting in putrefaction,
butyric acid formation and other undesirable fermentations, the pH of

PRESERVATION AND STORAGE OF FORAGE CROPS 281

the silage generally is lowered. This is accomplished by means of lactic
acid fermentation or the addition of inorganic acids (Watson, 1948).

c. Comparison of Carbohydrate and Protein Crops for Silage. Corn,
Zea mays L. and sorghum, Sorghum vulgare Pers., cut in the glazed to
early dent or similar stage of maturity usually produce a very good
silage. This has been attributed to the good supply of carbohydrates
and to a dry matter content (28-30 per cent) sufficiently high to result
in little or no excess moisture. By the nature of the material it normally
packs loosely enough for drainage of any excess liquid from the mass,
which prevents the silage from becoming excessively wet. These factors
encourage the formation of lactic acid which effectively lowers the pH
and inhibits the undesirable fermentations. Probably the chief problem
in large stalk carbohydrate-rich crops is having dry, light, and fluffy
material which packs poorly and becomes moldy (Watson, 1939).

If corn and sorghums are cut when immature they are lower in carbo-
hydrate and dry matter content. When the crop is ensiled it packs
tightly. Free water is released from the plant material by compression
during the ensiling process and by the subsequent settling. Silage from
immature corn frequently is high in pH and butyric acid content, and
putrefaction may result. The probability of this occurring is increased
if the excess liquid is not drained from the silage. The problem of con-
trolling these undesirable fermentations is even more acute in silage
made from protein-rich, immature hay and pasture forage crops in which
the carbohydrate content is low, the moisture content is frequently high,
and the fine stems facilitate tighter packing than in the case of corn.

d. Detrimental Effects of Poor Quality Silage. A moldy condition in
silage indicates that large losses of nutrients have occurred (Amos and
Woodman, 1922). Reed and Barber (1917) report that although most
of the molds in silage are nonpathogenic, some are pathogenic to animals.
Morrison (1948) points out that slightly moldy silage may not affect
cattle, but sheep and especially horses are more apt to be affected.
Moldiness may cause unpalatability in silage and stock usually refuse
to eat normal quantities of it.

The undesirable fermentations so difficult to control in protein-rich
crops usually result in a silage with an offensive odor. It is unpleasant
to handle and causes several feeding problems. It may be unpalatable
and the animals will not consume enough to constitute the normal quan-
tity of roughage. This must be compensated for by feeding more grain
or a decrease in production will result. If putrefaction has occurred,
according to Russell (1908), the proteins and amino acids are broken

282 R. B. MUSGKAVE AND W. K. KENNEDY

down into ammonia, amines and amides, some of which may be toxic
to animals and may cause digestive disorders when fed. van Beynum
and Pette (1940) state that when butyric acid silage is fed to animals
it is impossible to prevent contamination of the milk with butyric
organisms. Such milk may have an off-flavor and is unsuited for the
making of certain cheeses in which gas-producing organisms are
objectionable.

Thus not only does moldy or undesirably fermented silage probably
result in greater amounts having to be discarded as inedible, but the
remaining silage is less palatable, and is consumed in smaller quantities
which results in decreased production, increases the problem of producing
high quality milk and cheese, and may be toxic if fed to animals. Con-
siderable research has been devoted to determining methods of pre-
venting undesirable fermentation especially in the protein-rich crops.

2. The Ensilage Process

a. Chemical Changes. When a green crop is ensiled the plant cells
and aerobic microorganisms rapidly exhaust the oxygen in the entrapped
air and release carbon dioxide, water and heat. In a period of a few
hours, if the material is properly stored, aerobic respiration ceases, but
the enzyme systems still function under anaerobic conditions (Peterson
et aL, 1925).

Babcock and Russell (1900), Sherman and Bechdel (1918), LeClerc
(1939) and others have stated that protoplasmic respiration and enzyme
activity are the chief factors involved in preservation of silage. To
prove this theory Russell (1908) ensiled crops with no treatment, with
toluene to prevent microbial activity, and with heat to prevent both
microbial and enzymatic activity. Because the toluene treated material
was well preserved, even though it did not have a typical silage aroma
and appearance, he concluded that the bacterial action was of secondary
importance and merely ^complicated the process. Esten and Mason
(1912), Hunter and Bushnell (1916), and others took the opposite view
that the activity of microorganisms was the more important phase of
the process since the acidity developed by the organisms helped to pre-
vent undesirable fermentations. Peterson et al. (1925) showed that
sterilized corn inoculated with lactic acid organisms was preserved as
good quality silage. The activity of the plant cells is important but
unless there is complete sterilization of the mass, the action of the micro-
organisms cannot be overlooked because the types which develop deter-
mine the quality of the silage.

The major changes in the carbohydrate fraction of the ensiled crop
from anaerobic processes and the resulting losses in energy are due to the

PRESERVATION AND STORAGE OF FORAGE CROPS 283

formation of organic acids. The organic acids consist of the nonvolatile
acids, chiefly lactic, and the volatile acids, determined as acetic acid al-
though butyric acid is reported separately. The relative and total
amounts of these acids are so important in determining the quality of
silage that Kirsch and Hildebrandt (1930) established 5 classes of silage
on the basis of their organic acid content. The American Dairy Science
Association also considered acid content in their classification (Shepherd
et al, 1948).

Of the nonvolatile acids, lactic is present in the largest amounts with
formic, propionic, malic, succinic and other acids also reported as present
(Russell, 1908; Neidig, 1918; Peterson et al., 1925). Lactic acid usually
comprises over 50 per rent of the organic acids in silage classified as
very good, and generally amounts to 1 to 2 per cent of the fresh silage
(Kirsch and Hildebrandt, 1930). Since lactic acid has a much higher
dissociation constant (K = 1.38 X 10~ 4 ) than either acetic acid
(K - 1.75 X 10- 5 ) or butyric acid (K - 1.48 X 10~ 5 ) the pH of 3,5
to 4.2 which high quality silage attains is largely determined by lactic
acid.

Heineman and Hixson (1921) found that the bacteriological changes
in corn silage consisted of 3 phases, but in the final phase the lactic acid
bacteria dominate the fermentation. The hexoses and pentoses in corn
silage make an excellent medium for these organisms. Butyric acid
organisms attack more complex carbohydrates than lactic acid organisms,
and according to Archibald (1946), butyric acid formation may be high
in excessively wet silage. Acetic acid is a byproduct of the lactic and
butyric acid organisms, and may be formed in large quantities if the
lactic acid organisms lack an adequate supply of fermentable carbohy-
drates (Stone et al., 1943). While alcohols and aldehydes have been
isolated from silage they do not appear to play an important role in its
making (Hart and Willaman, 1912).

The degree of protein breakdown has been determined by comparing
the soluble nitrogen and ammonia nitrogen in the silage with that present
in the fresh crop. In all types of silage a portion of the protein under-
goes a breakdown to amino acids as a result of a tryptic enzyme present
in plants (Russell, 1908; Lamb, 1917; Hunter, 1921). Further changes
beyond the amino acids may be caused by other enzymes, but probably
the proteolytic bacteria of the Clostridium group play an important role
(Allen and Harrison, 1937; Rosenberg and Nisman, 1949). Several
species of soil bacilli are frequently found in silage and have the ability
to digest proteins and even the lactobacilli organisms may attack pro-
teins in the absence of carbohydrates (Hunter, 1917, 1918) . The organ-
isms break down protein compounds to organic acids, especially butyric

284 B. B. MUSGBAVE AND W. K. KENNEDY

acid and ammonia, amides, and amines (Russell, 1908; Schieblich, 1930,
1931). If growth of these organisms is not suppressed, undesirable char-
acteristics may develop in the silage and feeding problems may result.

Formation of nitric dioxide from fermenting silage has been reported
recently (Anon., 1949). The gas probably is released as nitric oxide
which is oxidized to the dioxide when it comes in contact with atmos-
pheric oxygen. The conditions which bring about this loss of nitrogen
are not understood, but it is due probably to the reduction of nitrates
by the anaerobic organisms (Wilson, 1943).

The mineral portion of the plant may undergo changes in combination
but will not be lost unless leached from the silage in the juice or as a
result of exposure to rain (Watson, 1939). Magnesium may be released
from the chlorophyll forming phacophytin, which gives silage a brown
color (Woodman, 1923). Due to the formation of organic acids and other
ether-soluble compounds, the ether extract fraction of silage is higher
than that of the fresh crop.

b. Control of Undesirable Fermentation. (1) Lowering pH. Virtanen
(1934) reported that lowering the pH to 4.2 prevented the growth of
butyric acid organisms and van Beynum and Pette (1936) showed that
these organisms were killed at pH 3.5. Virtanen also found that 1.5 per
cent of the total nitrogen of fresh clover was present as ammonia. In
the silage made therefrom at pH 3.6 ammonia constituted 2.0 per cent
of the nitrogen, at pH 4.3 12.0 per cent, and at pH 4.5 21 per cent.
Virtanen (1933) reported that below pH 4.0 both plant enzymes and
anaerobic microorganisms could not cause protein breakdown as measured
by soluble nitrogen and ammonia, but later work indicated that the
breakdown is not prevented except at pH of 3.0 or lower (Virtanen,
1934) . However, the growth of bacteria of the coli-aerogenes group and
butyric acid bacteria is prevented at pH 4.0. At pH 4.0 lactic acid
organisms produce appreciable quantities of lactic acid and some acetic
acid, and the proteins undergo marked hydrolysis to soluble nitrogenous
compounds, but the increase in ammonia nitrogen and decrease in digesti-
ble protein is small (Watson, 1939; Martos, 1941). Thus, rapidly low-
ering the pH to about 4.0 is an effective way, and is the method usually
used, for controlling butyric acid formation and undesirable proteolysis
in ensiled crops. Decreasing the pH to this desired level is usually ac-
complished by encouraging lactic acid fermentation or by direct acidifica-
tion with acids. In carbohydrate-rich plants, lactic acid fermentation
normally occurs with a decrease in pH to 4.0 or less, if the crop is har-
vested at the desired stage of maturity (Peterson et al., 1925) . In protein-
rich crops, such as legumes, active lactic acid fermentation may be

PRESERVATION AND STORAGE OF FORAGE CROPS 285

lacking. Legumes, in particular, are high in bases which form a strong
buffer system and cause more acid to be required to lower the pH (Wil-
son, 1935).

(a) Lactic Acid Stimulation. Lactic acid fermentation in silage is
encouraged, since by lowering the pH and possibly also by the antag-
onism of the lactic acid organisms, butyric acid and proteolytic bac-
teria are suppressed (Rodenkirchen, 1939; Waksman, 1947). Numerous
workers have conducted studies on the possibility of inoculating silage
crops wkh bacterial cultures to insure the proper fermentation. Occa-
sional benefits have been observed from this practice, but the general
conclusion is that inoculation is of little value. It has been concluded
that the green crop carries sufficient quantities of acid-forming bacteria,
but their growth and development is dependent upon the existing en-
vironment (Watson, 1948).

One of the factors which favors the growth of lactic acid organisms
is an adequate supply of carbohydrates, which are ordinarily lacking in
hay and pasture crops. Several early workers found that lactic acid
fermentation could be encouraged by mixing hay crops with high-carbo-
hydrate crops, such as corn or sorghum, or with high carbohydrate mate-
rials (Reed and Fitch, 1917). The use of molasses containing about 50
per cent sugars at rates of 15 to 100 Ibs. per ton of green forage gives
very successful results when ensiling meadow grasses and legumes. As
the amount of legume increases or as the material is less mature, the
amount of molasses needs to be increased (Watson, 1948).

Among other carbohydrate products which have been added success-
fully are crude sugar, ground grains, potato flakes, whey paste or dried
whey (Watson, 1939; Bender and Bosshardt, 1939). Fresh whey and raw
potatoes have been found too high in water for use unless the crop has
been wilted to a moisture content of 60 per cent or less (Kirsch et al.,
1934; Brouwer, 1937; Watson, 1939; Shepherd et al, 1946). Many other
materials including beet pulp, straw, cob meal, and hay have been used
with varying degrees of success. These latter products do not materially
increase the quantity of potentially fermentable carbohydrates but tend
to decrease the moisture content of the mass, and their effect is akin to
wilting (Watson, 1939).

Even if fermentable carbohydrates are added to succulent protein-
rich crops, a silage high in lactic acid will not be produced unless the
surplus water drains out of the silage (Crasemann and Heinzl, 1949).
Water-logged silage is high in acetic acid and may contain some butyric
acid (Archibald, 1946).

In contradiction to the hypothesis that certain crops, such as hay
and pasture forages, do not contain sufficient quantities of carbohydrates

K. B. MUSGRAVE AND W. K. KENNEDY

for lactic acid fermentation, it has been suggested that sufficient carbo-
hydrates are present but that the physical conditions are unsuitable for
the growth of the desirable acid-producing organisms. A good quality
lactic acid-type silage can be produced from protein crops with moisture
contents of 70 per cent or less without addition of supplements (Cooper,
1917; Woodward and Shepherd, 1936).

Immature forages which are cut, chopped and rapidly put in the
silo usually contain well over 75 per cent moisture at the time of ensiling.
The material packs well, little air is entrapped, and a large amount of
liquid is released. As a result, the amount of aerobic respiration is
small, the temperature of the mass does not rise materially, and the
formation of acetic and butyric acid and ammonia results, but little
lactic acid accumulates. On the other hand, if forages have a lower
moisture content because of partial wilting or approaching maturity,
or are stored unchopped in shallow layers, the conditions are reversed
and good lactic acid fermentation develops (Watson, 1939, 1948).

The success of the wilting method is attributed to the greater amount
of oxygen entrapped within the crop as it is ensiled. Most of the lactic
acid organisms do best at low concentrations of oxygen (mieroaerophilic)
or are facultative anaerobes (Martos, 1941). On the other hand many
butyric acid-forming bacteria are true anaerobes, although some are
facultative anaerobes (Watson, 1939). Acetic acid appears to be present
in larger relative amounts when butyric acid formation is favored (Archi-
bald, 1946).

Another factor in wilting which may encourage lactic acid formation
is the concentration of sugars (Wilson, 1948a). Although wilted forage
does not contain greater amounts of sugar at the time of ensiling than
unwilted material (dry matter basis), the resultant silage from wilted
crops is higher in sugars (Stone et al., 1943; Autrey et a/., 1947).

A number of workers have maintained that the temperature of the
mass during fermentation determines the type of silage produced. Tem-
perature increases, chiefly* caused by the amount of aerobic respiration,
can be controlled by regulating the stage of maturity, the degree of chop-
ping and packing, the moisture content and depth of the silage. Gen-
erally, material which is packed loosely and allowed to heat to 50C.
or higher (warm fermentation process) produces a sweet smelling silage
low in volatile acids with no production of butyric acid (Dijkstra, 1948;
Roseveare, 1948).

The cold fermentation practice is used in Germany because butyric
acid supposedly is formed if the temperature rises. The forage is chopped,
tramped as it is ensiled, and heavily weighted to exclude as much air
as possible. If properly handled the temperature does not rise above

PRESERVATION AND STORAGE OF FORAGE CROPS 287

20C. The moisture content must be 70 per cent or less because exces-
sively wet material will not make a silage that will keep satisfactorily,
even though a preservative is added (Crasemann and Heinzl, 1949).

In England silage is made successfully from immature forage by
ensiling the crop unchopped. It is added to the stack or silo in layers
of 5 to 8 feet in depth and allowed to heat so that all of the mass will
reach temperatures of 27 to 38C. This process also develops a desirable
fermentation yielding lactic acid in excess of acetic acid with no appre-
ciable amounts of butyric acid (Watson, 1939, 1948; Bohstedt, 1944).

The proponents of controlling fermentation by regulating the tem-
perature of the silage mass, whether it be below 20C., between 27 to
38C., or above 50C., explain the success of their methods on the basis
that lactic acid organisms are favored at the temperatures they advocate
while the growth of the undesirable organisms is suppressed. Other work
indicates that the optimum temperature for both the lactic acid and
butyric acid bacteria is variable, but near 37C. for most of the species
in both groups. Thus it would appear that the difference in tempera-
ture is not the chief reason for the presence of butyric acid in wet im-
mature silage (Watson, 1939).

(b) Direct Acidification. The addition of acids to decrease the pH
of silage to below 4.0 has been extensively used for crops which normally
do not produce a rapid lactic acid fermentation when ensiled. The addi-
tion of acids to silage dates back to 1885 when citric, tartaric and hydro-
chloric acids were compared (Giglioli, 1914). The two organic acids
were unsuitable, but the addition of hydrochloric acid was recommended.
Gerlach et al. (1929) found that the growth of undesirable organisms
in silage is suppressed materially at one per cent concentration of lactic
acid and prevented at the 1.5 per cent level. Due to its high cost the
direct addition of lactic acid is impractical but indirect sources such as
sour whey or molasses may be used.

In 1926 and 1927 Fingerling in Germany patented a process which
consisted of adding sufficient quantities of hydrochloric acid to produce
a pH of 2.0. The process resulted in excellent preservation, but was
exacting and the problem of feeding such an acid feed to livestock pre-
vented the method from being widely accepted (Watson, 1939). In
1929, Virtanen (1934) established the process, known as the A.I.V.
process, of direct acidification in the making of silage. Virtanen rapidly
lowered the pH of the crop to at least 4.0 but not lower than 3.0 by the
use of mineral acids, usually a mixture of hydrochloric and sulfuric acids.

The acid in the A.I.V. or other processes must be uniformly applied
to the crop at an accurate rate determined by the type of crop, its
moisture content and the soil upon which it is grown. Immature and

288 R. B. MUSGRAVE AND W. K. KENNEDY

legume-rich forages require more acid to lower the pH than do other
crops. Apparently the acid requirement for any particular lot of forage
is proportional to its dry matter content. Changes in the moisture con-
tent of the crop require an adjustment of the rate of acid applied. A
crop produced on a neutral or alkaline soil requires more acid to lower
its pH than one produced on an acid soil (Virtanen, 1934; Watson, 1939).

Shortly after the A.I.V. process was introduced, the Defu process,
which consisted of adding molasses equal to 0.20 per cent of the green
crop and hydrochloric acid containing one per cent phosphoric acid, was
suggested (LeClerc, 1939; Bender and Bosshardt, 1939; Watson, 1939).
Liquid phosphoric acid and solid acid materials such as phosphorus
pentachloride and sulfur trioxide have also been used to lower the pH
of the crop. All these materials successfully lower the pH if properly
used and produce a good quality silage from protein-rich crops (Virtanen,
1932; Wilson and Webb, 1937; Watson, 1939).

An effective way of counteracting the acidity and avoiding most of
the feeding problems of mineral acid silage is the feeding of legume-
rich hay, sodium bicarbonate or limestone (Lepard et at., 1940; King,
1943). However, Ingham (1949) reports that although the cattle were
fed 2 per cent ground limestone, phosphoric acid-silage was inferior to
molassess-silage in palatability, phosphorus retention, economy of milk
production and absence of physiological disturbances.

(2) Sterilization. A number of other chemicals which do not mate-
rially lower the pH of silage have been used to encourage desirable fer-
mentations or to prevent any fermentation. By aerobic and anaerobic
respiration the carbon dioxide content can increase to 70 to 90 per cent
(volume basis) of the gases in an airtight silo (Neidig, 1914; Peterson
et al., 1925; Watson, 1939). At these concentrations it has been the-
orized that both respiration and fermentation would be checked. The
Cremasco process, developed in Italy, consists of placing well-wilted
material containing 60 to* 70 per cent dry matter into airtight silos which
are heavily weighted and sealed. According to Samarani (1922) this
results in negligible bacterial activity with about 5 per cent loss of sugars
and very slight breakdown of proteins. Other work indicates that lactic
and acetic acid formation is of about the same magnitude as has been
observed in ordinary silage made by the wilted process (Schmidt, 1934;
Mikhin et al., 1937) . Butyric acid formation is low and surface spoilage
is essentially absent. Although a carbon dioxide atmosphere may prevent
undesirable fermentations in wilted forages it has little or no effect in
preventing undesirable fermentations in crops ensiled with over 70 per
cent moisture content (Mikhin et al., 1936; Albada, 1946).

Wilson (1948b) reviewed the literature and found that salt has no

PRESERVATION AND STORAGE OF FORAGE CROPS 289

sterilizing effect, and has little affect in preventing a rise in temperature.
Apparently salt increases the initial rate of drainage which may have a
slight beneficial effect on lactic fermentation.

Procopio (1942) treated fodder with sulfur dioxide as it was ensiled
and found that a concentration of 0.8 to 1.0 per cent sulfur dioxide in-
hibited all biological and enzymatic processes while 0.25 to 0.30 per cent
sulfur dioxide suppressed their action. No free sulfur dioxide could be
found after 5 months. Kvasnikov and Raev (1939) found that 0.15 per
cent sulfur dioxide was lethal in pure cultures of lactic, acetic, and buty-
ric acid bacteria. Workers in New Jersey (unpublished) treated silage
with concentrations of 2, 4, and 6 Ibs. of sulfur dioxide per ton of green
crop and found that the ether-soluble fraction was not materially in-
creased in the silage. This would indicate that little organic acid forma-
tion occurred. Overholser and Cruess (1923) and Sisakyan and VasiPeva
(1945) found that sulfur dioxide inhibits oxidizing enzymes, especially
peroxidase, and Voinovitch et al. (1949) report that sulfur dioxide in
combination with thiamine or nicotinic acid is an effective antioxidant.
A legume-grass silage treated with 0.25 per cent sulfur dioxide at the
time of ensiling is being fed to dairy cattle by Pennsylvania workers with
no adverse effects.

Other chemicals which have been used for controlling fermentations
in silage are sulfamic acid, acid salts, calcium hypochlorite, phenol, ben-
zoic acid, borax, salicylic acid, oxalic acid, carbon disulfide, formic acid,
and formaldehyde. Either the results have not shown them to be ad-
vantageous or the data concerning their value is questionable (Watson,
1939; Johnson et al., 1941b).

The application of heat to the ensiled crop has been tested with the
aim of producing a sweet silage similar to that obtained in the warm
fermentation process. Warm air, steam and electricity have been used
to heat the silage. Results indicate that the application of heat prob-
ably delays the fermentation. The resultant silage, however, contains
nearly the same amounts of lactic and acetic acids as are present in
untreated silage (Knisely, 1903). Butyric acid also has been found in
heated silage along with the usual amount of protein breakdown. It
appears that the use of heat has no practical application in silage making
(Hoffman, 1923; Watson, 1939).

(3) Complete Exclusion of Oxygen. The possibility of storing silage
in a vacuum has been studied by Schmidt (1934), but sufficient carbon
dioxide is released to result in essentially the same system and type of
silage as obtained following the addition of carbon dioxide or enclosure
in an air-tight container in which normal respiration processes exhaust
the oxygen.

290 R. B. MUSGRAVE AND W. K. KENNEDY

The application of pressure is an effective means of compressing the
silage, excluding air and preventing the subsequent entrance of air. By
limiting the amount of air, excessive heating and aerobic losses are pre-
vented. Pressure per se, however, at least over a large range, has no
influence on the silage processes (Dijkstra, 1945; Crasemann and Heinzl,
1949).

c. Losses of Dry Matter and Nutrients. The losses of dry matter,
crude protein and carotene which occur in the preservation of hay crops
' are discussed in VII. The losses which occur after the crop is ensiled
are due to surface spoilage, drainage and fermentation. In general, the
surface losses can be reduced to essentially zero in tower silos which are
air-tight or have properly protected surfaces, but approach 30 per cent
for small stack silos (Watson, 1939; Rogers, 1949). Additional losses
due to aerobic respiration occur on the exposed surface of opened silos.
Spoilage due to mold growth is prevented when at least a l a /2 to 2 inch
layer of silage is fed each day during the winter and when somewhat
larger quantities are fed during the summer (Morrison, 1948), but the
magnitude of the dry matter losses which occur even when visible mold
growth is prevented is not known. Probably it is considerably higher
during the summer than during the winter.

The loss of dry matter due to drainage usually does not exceed 3
per cent and approaches zero if the dry matter content of the ensiled
material is 30 per cent or higher (Godden, 1923; Archibald and Gun-
ness, 1945; Monroe et aL, 1946). Internal losses of dry matter due to
fermentation have been reported as being from 2 to 30 per cent, but
probably a 7 to 10 per cent loss is the minimum that can be expected
(Taylor et al., 1940; Camburn et al., 1942; Shepherd et al., 1947, 1949).
Rapid lowering of the pH by the addition of acids or fermentable carbo-
hydrates and partial wilting of the crop appears to result in less dry
matter loss than when na preservatives are added to unwilted material.
Severe heating of silage with the temperature of the mass rising as high
as 70 to 75C. may cause charring and excessively high losses of dry
matter with over 80 per cent loss of digestible crude protein (Woodman
and Hanley, 1926).

The carbohydrate undergoes the greatest loss of any nutrient, with
20 to 30 per cent of the starch and pentosan fractions disappearing
(Peterson et al., 1925) . Regardless of whether silage is stored unwilted,
with or without a preservative such as molasses or mineral acids, or
after partial wilting, the forage undergoes fermentation. Normally lactic
acid fermentation results in less loss in feeding value than butyric and
acetic acid fermentations. The loss of energy is less than 3 per cent in

PRESERVATION AND STORAGE OF FORAGE CROPS 291

converting glucose to lactic acid, but is about 23 per cent when butyric
acid is formed, and 38 per cent when acetic acid is the acid product
(Buchanan and Fulmer, 1928). According to Watson (1939) the loss
in starch equivalent is slightly less than 25 per cent from silage receiv-
ing preservatives and slightly higher than 33 per cent from unwilted
material receiving no preservatives. These losses appear to be high in
view of the recent investigations by Hodgson (1949) and Shepherd et al.
(1949) discussed in VII.

The loss of digestible crude protein is greater if acidification or rapid
lactic acid fermentation does not take place (Brouwer et al., 1933; Wat-
son, 1939). In all silages the amount of soluble nitrogen materially
increases during the ensiling process, although if amino acid breakdown
does not occur the percentage of digestible protein is not markedly
decreased (Peterson et al., 1937; Watson, 1939).

The "grass juice f actor " also is retained in legume and grass silage
especially when molasses or acids are added (Johnson et al., 1941a).
Vitamin C is of little importance and is almost completely lost through
the action of the peroxidases in untreated silage, but the addition of
mineral acids may cause the retention of 50 per cent of the vitamin C.
Vitamin B and D are not known to undergo any appreciable change,
but are generally present in small quantities (Watson, 1939).

3. Questions Needing Further Research

It has been contended that the important chemical changes occurring
in silage are the result of cell respiration and enzyme activity, but
microbial activity may be equally or even more important. All widely
used methods of silage production are based on the principle of con-
trolling undesirable fermentation and encouraging or at least not limiting
desirable fermentations.

In controlling undesirable fermentations, the importance of lactic
acid production has been attributed to its causing a decrease in the pH
of the silage. The control of butyric acid organisms at a pH of 4.0 to
4.4 when lactic acid organisms are present (van Beynum and Pette,
1936) alternatively may be accounted for by the antagonism existing be-
tween the two groups of microorganisms, as indicated by Rodenkirchen
(1939). Regardless of the reason, vigorous lactic acid fermentation
appears to control undesirable fermentation. Gorini (1942) points out
that the "native" lactic acid organisms are not present in sufficient quan-
tity and quality on many crops to promote a vigorous lactic acid fer-
mentation that will effectively lower the pH and permanently control
the butyric acid and proteolytic bacteria.

The practice of inoculating silage with lactic acid bacteria has not

292 B. B. MUSGBAVE AND W. K. KENNEDY

been universally successful; however, there has been sufficient success
with inoculation to indicate that it might have a place in silage making
(Watson, 1939). Gorini (1942) has contended that inoculation of silage
has not been more successful because the lactic acid cultures were un-
suited to "vegetal sugars" and to the environmental conditions which
exist in a particular silage making process. Special emphasis needs to
be placed on the selection of an organism which will grow vigorously
on succulent, immature protein-rich crops and suppress undesirable fer-
mentations by increasing the acidity, or by other forms of antagonism.
Since legumes, in particular, have a strong buffer capacity and the pH
is difficult to lower, the control of undesirable organisms by means other
than lowering the pH would be desirable.

In addition to the search for more efficient strains of bacteria to
control undesirable fermentations, a detailed study of the relationship
of environment and undesirable fermentations is needed. Why do un-
desirable fermentations occur when high moisture crops are ensiled? Is
it due to the amount of the liquid phase per se y the amount and type of
the gaseous phase, or the amount and type of the solid phase? Does the
carbon-nitrogen ratio or the insoluble-soluble carbohydrate ratio of the
crop have any significance in determining the type of fermentation?
Does the wilting process merely concentrate the chemical constituents
of the plant or does it alter their physical and chemical properties? Does
the addition of insoluble carbohydrates only alter the physical condi-
tion of the mass and thus encourage lactic acid fermentation of existing
sugars, or are they hydrolyzed to fermentable carbohydrates which lactic
acid organisms can attack? Is the temperature of the mass during fer-
mentation important or is it merely a modifying factor? There is a
lack of fundamental information from which conclusive answers to the
above questions can be derived.

The problems of ensiling immature protein-rich crops have not been
adequately solved. The wilting process is widely advocated but requires
going over the field two or more times, requires considerable experience
in judging the proper degree of wilting, and is not possible in rainy
weather. The addition of molasses, other carbohydrate-rich materials,
or acids also has several limitations. Obtaining the materials is usually
an extra expense, the addition of the materials requires time and super-
vision, molasses may cause clogging of the blower pipe, and the acids may
corrode machinery and damage clothing. If preservatives are not ap-
plied uniformly, the under-treated portions of the silage may undergo
undesirable fermentation. Even though a very small portion of the
silage undergoes butyric acid fermentation, van Beynum and Pette
(1940) point out that when it is fed, the milk becomes contaminated

PRESERVATION AND STORAGE OF FORAGE CROPS 293

and cannot be used in making certain high quality cheese. These diffi-
culties often discourage many farmers from adding preservatives even
though mechanical devices are available for simplifying the process.
Can better methods of applying the existing preservatives be developed?
Are there other preservatives which are easier to apply?

Frequently the roughage is low in protein due to the low amount
of legumes in the hay and pasture crops. Although generally the protein
content of the roughage is best increased by the growing of more legumes,
the possibility of improving the protein level of low protein silage needs
further investigation. Brigl and Windheuser (1931), Windheuser et al.
(1935), and Cullison (1943) have added nonprotein nitrogen to silage
with the expectation that the silage organisms would convert the nitro-
gen into proteinaceous material. Urea, ammonium carbonate and am-
monium bicarbonate have been used. The protein level appears to be
improved without decreasing the palatability; however, more data are
needed before the addition of nonprotein nitrogen to silage can be
recommended.

Silage is a very palatable feed and one of the most satisfactory
methods of preserving roughages, but the losses in feeding value are
nevertheless considerable. Nearly 15 per cent loss in feeding value
occurs due to the ensiling process even when the best accepted methods
of making silages are used. In other words, 6 or less tons of total digesti-
ble nutrients are fed for every 7 tons ensiled. The loss due to surface
spoilage can be reduced to a negligible amount if due care is taken in
topping off and sealing a well constructed silo; however, considerable
spoilage frequently results from the accidental entrance of air. Little is
known about the magnitude of the surface losses which occur when a
silo is opened and fed during warm weather. The surface of the silage
is an excellent medium for mold growth which is exemplified by the
speed with which the silage must be fed to prevent visable mold growth.
Even though mold growth is not observed, aerobic respiration is taking
place and an appreciable loss of nutrients is probably occurring.

Pentzer et al. (1933) report the use of sulfur dioxide to control mold
growth and respiration in grapes. The possibility of treating the surface
silage with sulfur dioxide as a fungicide to prevent surface spoilage is
under investigation at Ohio State College. Dawson et al. (unpublished)
are studying the prevention of mold growth in hay by the use of certain
phenol compounds. These may be applied to prevent surface spoilage
in silage.

Recently a steel, glass coated tower silo which is air tight has been
offered for sale in the United States. In this aerobic losses are limited
to those supported by the entrapped air. The silo is sealed when the

294 R. B. MUSGRAVE AND W. K. KENNEDY

filling operation ceases and no air can enter the silo except that neces-
sary to compensate for the decrease in volume as the silage is removed
by a mechanical unloader operating in the bottom of the silo. The
unloader digs its way into the silo after it is filled and is then sealed into
position. Such a silo eliminates the topping off and sealing required
in conventional silos.

While improved methods of controlling surface spoilage have been
developed, the losses within the mass due to fermentation have been
classified as "unavoidable losses," yet they are frequently greater than
surface spoilage. Rapid lowering of the pH and wilting appear to
decrease fermentation losses, but what effect does the external tempera-
ture during fermentation have upon the preservation of a crop and the
dry matter losses which occur? No loss in dry matter occurs with com-
plete sterilization which causes the cessation of cell respiration, enzyme
activity, and growth of microorganisms, but to date, workable procedures
for such sterilization have not been developed.

The possibility of using bactericides which will inhibit silage fermen-
tations and still not be toxic to the flora of the rumen when the silage
is ingested should be investigated. The use of sulfur dioxide may de-
crease the amount of "unavoidable losses" in silage by inhibiting bac-
teria and by repressing certain of the enzymic changes without causing
the silage to be toxic to animals. Perhaps other chemicals would be
of equal or more value.

Advances in methods of producing high quality silage from protein-
rich crops are needed, and until some of the problems of ensiling such
crops are eliminated the proper conservation of much of the surplus pas-
ture and first cutting hay crop will not be accomplished. The validity
of the contention that losses of approximately 15 per cent of the digestible
nutrients are unavoidable needs to be questioned and further attempts
made to prevent them.

IV. FIELD-CURED HAY

1. Factors Influencing Rate of Drying

The final quality of field-cured hay as it is removed from storage
is largely determined by the extent of the losses taking place in the
field. Field losses are directly related to the length of time required
for drying. Unfortunately the highest quality forage as cut usually
requires the longest curing period. This has been attributed to poor
drying weather prevailing at the time hay crop plants reach their maxi-
mum vegetative growth. However, high moisture content (Archibald
et al.j 1946) and hygroscopicity of the plant tissue (Dexter et aL, 1947)

PRESERVATION AND STORAGE OF FORAGE CROPS 295

have been proposed but not proven to be as important as weather in de-
termining the rate of drying of early cut hay,

a. Time of Day to Cut. Henson (1939) noting hourly variation in
moisture content of some plants, determined the moisture content of
alfalfa at hourly intervals from 8 a. m. to 8 p. m. and concluded that in
this plant and in probably all common hay crops there is not sufficient
change during the day to justify using plant moisture as a guide to time
of day to mow. Hartwig (1942) found that hay cut in the morning with
the dew on it was usually as dry at the end of the first day of curing as
was hay cut after the dew had dried off. Hay cut in the late afternoon
traps the day's production of photosynthate in the tops (Curtis, 1944).
This may result in hay of higher nutritive value if the method of curing
controls subsequent respiration.

b. Raking, Cocking, Crushing and Tedding. Swath curing in most
trials has been more rapid than windrow curing (Henson, 1939; Higgins,
1932; Kiesselbach and Anderson, 1927). However, swath curing for
1 to 4 hours followed by windrowing has compared well with complete
swath curing as to rate and has produced hay sufficiently better with
respect to leafiness and color that it is recommended for alfalfa.

Raking should be done before the plants dry to below 40 per cent
moisture content because leaves begin to shatter at this point (Zink,
1936). He observed, however, that alfalfa under some conditions may
be dried to 20 per cent moisture content and yet retain its leaves during
handling. He considers the problem needs further study to delineate
the controlling factors.

Cocking wilted hay delays the time for storage still more than wind-
rowing (Rather and Morrish, 1935). Watson (1939) reviewing his own
work and that of others shows that in good drying weather cocking fre-
quently produces hay of poorer quality than faster methods but decidedly
better hay under poor drying conditions.

Crushing the stems of hay between rollers as it comes off the cutter-
bar decreases the field drying time under nearly all conditions. Crushed
soybean Glycine max (L.) Merrill hay retained its leaves and pods and
was stored with much higher carotene content than uncrushed hay in tests
by Jones and Dudley (1948). Uncrushed hays of sudan grass, Sorghum
vulgare, Pers., var. SUDANENSE, Piper, and Johnson grass, Sorghum hale-
pense, L., absorbed more water during the night than crushed hay in con-
tradiction to the usual statement that crushed hay not only dries out
faster but will absorb greater quantities of dew or rain (Terry, 1948).

Trials with tedding in the United States have shown very little ad-

296 K. B. MUSGRAVE AND W. K. KENNEDY

vantage in speeding up the curing, and farmers have largely discon-
tinued the practice (Henson, 1939).

2. Field Losses of Dry Matter and Digestible Nutrients

Field losses of carbohydrates, crude protein, minerals and vitamins
occur as a result of respiration, leaching and leaf shattering (Hodgson
et al., 1948). Field losses of dry matter from 8 lots of hay made in
England by hand methods amounted to over 16 per cent (Watson et al.,
1937). This compares with more recent studies in the United States
where drying conditions may be better but where the haymaking job is
done mechanically. Mixtures containing a large proportion of alfalfa,
made into hay with the use of a tractor power mower, side delivery rake
and hay loader, have given an average dry matter loss of 25 per cent
during 4 trials in as many years (Shepherd et al., 1947, 1949).
Grass and legume crops when made into hay lost an average of 11.7
per cent of dry matter in 8 trials in Vermont (Camburn et al., 1942,
1944).

Respiration accounts for a loss of up to 10 per cent of the hay during
field curing. Mechanical losses including shattering and leaching make
up the remainder (Wiegner, 1925). If the drying is rapid only the
simpler carbohydrates are lost through respiration, while with prolonged
drying, nitrogen also is lost especially when there is leaching.

Very few data have been reported showing the total loss of leaves
during field curing. Kenney (1916) reports an average loss of 12.43 per
cent from 41 lots of alfalfa. Henson (1939) working also with alfalfa
found leaf losses of 5 to 10 per cent during the loading operation follow-
ing various curing practices. In the Vermont studies, the average total
field losses from 3 alfalfa, one red clover and 4 timothy crops were
16.2, 32.5, and 4.2 per cent of dry matter respectively (Camburn et al.,
1942, 1944). If a small respirational loss is assumed for the timothy
crop the mechanical loss n^ust have been practically nil. Gerlach found
that grass hay lost 3 per cent while Bokhara clover, Melilotus alba,
Desr., M. ; lost 28 per cent of its dry weight as a result of leaf shattering
in 4 trials (Watson, 1939). It is generally recognized that grass crops
suffer less loss through shattering than legume crops.

The seriousness of leaching, mechanical and respiration losses, as
has been pointed out in II-3c, is evident when it is recognized that the
dry matter lost may be 100 per cent digestible. Under certain conditions
an approximate percentage decrease in the nutritional value of hay
during field curing may be obtained by doubling the percentage of dry
matter lost because the dry matter of hay is only about one-half
digestible.

PRESERVATION AND STORAGE OF FORAGE CROPS 297

S. Changes in Vitamin Content during Field Curing

Vitamin D develops in hay during the field curing period. Wallis
(1944) found that this development is as rapid in hay raked into small
windrows as when left to cure in the swath. Raking into large windrows
or cocking slows down the rate of development. Turning the windrows
before the final period of curing increased the vitamin D content ap-
proximately 10 per cent. Good increases occurred even in periods of
cloudy, rainy weather since the intensity of the ultraviolet rays is not
decreased as much as the longer light rays. The content of vitamin D
ranged between 300 and 1,000 International Units per Ib. when the
alfalfa hay was sufficiently dry for storage. This content was more than
doubled by a prolonged field exposure of 6 days.

Vitamin C and several vitamins in the B group are well preserved
in field-cured hay but are considered to add little to the value of hay
(Watson, 1939). The influence of curing conditions on the content of
tocopherols in hay has not been studied to date.

Carotene loss varies directly with the amount of exposure to light,
and the length of time required for curing and temperature. Enzymatic
action (Mills and Hart, 1945) photolysis and oxidation (Guilbert, 1935)
are the processes causing the loss of carotene. The loss is usually at
least 50 per cent, and when hay is exposed to extreme weathering es-
sentially 100 per cent of the carotene is lost (Camburn et al., 1942, 1944).

4. Storage Losses

a. Carotene. As would be expected from the processes causing its
destruction carotene continues to diminish after hay is stored. Generally
the higher the carotene content the greater is the absolute storage loss.
In the studies by Camburn et al. (1942, 1944) losses during one year
of storage have ranged from 46 to 78 per cent of the content as stored.
Greater losses occur in summer than in winter (Kane et al., 1937). Hay
which heats badly in storage undergoes nearly a complete loss of carotene.

b. Heating. The amount of heat which accumulates in stored hay
depends on the moisture content, nature of the crop, density, size and
shape of the storage mass. Hay which is stored at a moisture content
under 13 per cent will not heat (Wright, 1941). Under climatic condi-
tions prevailing in most of the hay-producing regions this low content
is rarely attained for two reasons. Curing periods with relative humidi-
ties low enough to obtain such moistures are infrequent, and farmers
avoid such extreme drying for fear of excess leaf shattering and a
brash, unpalatable product. With moisture contents between 13 and 25

298 R. B. MUSGRAVE AND W. K. KENNEDY

per cent the heat given off in respiration is sufficient to volatilize the
water in excess of 13 per cent. The vapor escapes from normally-packed
long hay; drying is accomplished and thereby respiration and heating
are checked. Hay with 25 to 35 per cent moisture can dry down without
serious heating provided it is placed in a bin with a relatively large
amount of surface exposure, which increases radiation losses and facili-
tates the escape of moisture. Roethe (1937) found that for undercured
long hay the bin should not be wider than 12 or 14 feet.

Early cut hay made with a small amount of field loss may undergo
greater heating in storage than that cut at the ordinary time (Watson,
1939). It is not clear whether this is due to a higher percentage of
respirable materials or to the greater density resulting from the better
packing of a more finely divided and pliable product. Greater density,
besides placing a larger quantity of respirable material in a unit of
space, also slows down the drying by inhibiting vapor and air movement.

Both chopped and baled hay attain higher temperatures when stored
at a given moisture content than does long hay. Miller et aL (1934)
and Woodward and Shepherd (1936) found dry matter losses of about
5 per cent whether the hay was chopped and stored at near 25 per cent
or near 15 per cent moisture content. This indicates very similar quan-
tities of heat produced yet the maximum temperatures attained varied
from 41 to 66C., and density, as measured by the storage space per
ton, varied from 227 to 152 cubic feet. Both moisture content and
fineness of chopping affected the density but neither significantly affected
the loss of dry matter.

Hay stored above 30 to 35 per cent moisture, through the processes
of microbial respiration and chemical oxidation, may heat to the point
of ignition. According to Hoffman (1940) heating up to about 66C.
is caused primarily by microorganisms. Such heating causes the pro-
duction of oxygen-absorbing materials which generate more heat as they
are oxidized. He warns that hay which has heated to the 66 to 110C.
range may rise very rapidly to about 226 C. at which ignition takes
place. Dobie (1948) has developed an inexpensive temperature probe
with which hot zones can be located quickly when excessive heating is
suspected. Keeney (1941) has found that forcing carbon dioxide into
zones of overheated hay lowers the temperature long enough for the hay
to be removed.

The heating of undercured hay is utilized in the production of brown
hay. According to Watson (1939) this method of hay making originated
in England. The heat produced by bacteria and fungi growing on hay
stacked at about 50 per cent moisture is sufficient to dry it down to the
equilibrium level. At the same time sufficient carbon dioxide is pro-

PRESERVATION AND STORAGE OF FORAGE CROPS 299

duced to destroy the green color, and heat develops a brown color which
may grade into black at very high temperatures (Henson, 1939). Brown
hay may also develop in the center of large stacks of hay containing
much less than 50 per cent moisture and is frequently found in the centers
of bales of undercured hay.

The amount of dry matter lost in the successful production of brown
hay varies enormously. Woodward and Shepherd (1936) found a loss
of 6.5 per cent but Hoffman and Bradshaw (1937) observed a maximum
loss of 22 per cent. The losses in digestible nutrients are very high
because the heating, besides consuming a large part of the digestible
nutrients, renders the remainder less digestible (Watson, 1939).

Hay stored under 25 to 30 per cent moisture usually retains its color
(Henson, 1939). The storage loss of dry matter from such hay is re-
markably constant at about 5 per cent. The carbohydrates found in the
nitrogen-free extract constitute most of the loss according to Camburn
et al. (1944).

c. Preservatives. Numerous experiments with salting undercured
hay as stored have rarely shown any beneficial effects other than delaying
the time at which the maximum temperature is reached (Henson, 1939;
Roethe, 1937). Musgrave and Dawson (1946) found that preservatives
containing sodium bicarbonate as the active ingredient had no inhibiting
effect on the respiration of undercured hay.

V. BARN HAY DRYING
1. Introduction

Hay for barn drying is gathered from the field and placed in storage
at much the same moisture content as is brown hay. Unlike brown hay
it is kept cool during the storage drying period by forcing air through it.

This method of preservation as evolved by Weaver (1937) elim-
inates at a moderate cost most of the usual losses occurring in field cured
hay. It takes advantage of the rapid removal of moisture occurring in
the first part of the field drying cycle without incurring undue weather
hazard. Raking, loading, handling arid hauling are performed before
the crop is dry enough for the leaves to shatter. The system is simple,
and requires only a small capital investment; labor requirements and
operating costs are low and it eliminates the danger of fire from spon-
taneous combustion (Finn-Kelcey, 1948).

300 B. B. MUSGBAVE AND W. K. KENNEDY

H. Basic Barn Drier Designs

A fan which will deliver large quantities of air against only low
static pressure is used to force air into a duct system. This system may
consist of a large main duct vented to laterals or to a raised slatted
floor through which the air is distributed under the hay. The floor on
which the drier is built must be air tight. By terminating the laterals
or slatted floor 4 to 6 feet from the edges of the mow a greater percentage
of the air is caused to pass up through the hay. The main duct and open-
ings from it are of such dimensions as to create an air velocity of about
1,600 feet per minute when handling sufficient air to yield, as a rule,
not less than 10 nor more than 20 cubic feet per minute per square foot
of mow floor area. Under these conditions the air in the system is under
a static pressure of approximately % inch of water. With 10 or 15
feet of long hay over it the static pressure goes up to between % and
1 inch of water (Schaller et al., 1945) .

The fan is either powered by an electric motor or internal combustion
engine, the former having the advantages of automatic operation with
little fire hazard, while the latter is damaged less in case of overloading,
its speed is easily varied and it may be useful in contributing waste heat
to the drier (Terry, 1947). Only a general description has been given
because the engineering features of barn driers vary so much with the
numerous sizes and shapes of mows, barn designs, total depths of loading,
and types of hay.

3. Operation of Barn Driers

a. Moisture Content. Experience with driers for long hay has estab-
lished the most desirable moisture level at about 35 per cent (Duffee,
1947). This figure assumes an air flow of 10 to 15 cubic feet per min-
ute per square foot of mow area and a normal 8 to 10 foot loading depth.
In actual practice when partially cured hay is threatened with rain it is
often placed over the driers at much higher moisture contents. Frudden
(1946) points out that hay wetter than 35 per cent can be placed on the
drier with no danger from spontaneous combusion, but that if moldy hay
is to be avoided the amount of water going on the drier should not exceed
that which can be evaporated in 7 days. However, the time required to
evaporate a given amount of water with unheated air will depend some-
what on atmospheric conditions which cannot be predicted with certainty.
Numerous reports indicate that most drying cycles required at least
10 days and that the majority of cycles permit some molding. This may
be in the form of a light, fine mold scattered throughout the mass or it
occurs in heavy concentrations in spots receiving inadequate air flow.

PRESERVATION AND STORAGE OF FORAGE CROPS 301

Since drying progresses slowly from the bottom of the mow to the top,
the upper layers remain moist until drying is complete. This gives an
easy means of determining when the mow of hay is dry but at the same
time provides conditions favorable to top molding.

A thin layer of moist hay will cool the incoming air to very near
its wet bulb temperature (Frudden, 1946). This cooling, together with
the small amount of water vapor added, creates relative humidities of
85 to 95 per cent. This suggests that during periods of high humidity
occurring at night and in cloudy, rainy weather no drying will take
place, but Jennings (unpublished data) showed that when air is blown
through hay continuously the rate of drying is only slightly reduced at
such times. Likewise, Davis (1947) found that drying goes on in the
hay above the thin layer which exhausts all of the air-borne heat avail-
able for drying. Dawson and Musgrave (1946), using a laboratory tech-
nique, explained these occurrences by demonstrating that over 60 per
cent of the heat consumed during a drying cycle is produced by the res-
piratory activity of microorganisms growing on the moist hay. Hendrix
(1947) found that heat produced in the hay accounted for over 60 per
cent of the drying of chopped hay even when the air was blown con-
tinuously at the high rate of 25 cubic feet per minute per square foot.

b. Density. Density of the hay over the drier influences the rate of
drying by its effect on the amount and path of air flow. Increasing the
density increases the static pressure and thereby cuts down the fan
capacity per unit of power. It also causes greater side losses of air.
Where the density of loading is variable air will tend to channel up
through the loose zones resulting in poor efficiency (Kalbfleish et aL,
1947). This condition is noticeable when the hoist and fork or sling are
used to load the drier. Dropping the hay develops a packed zone under
the track. This can be eliminated by dropping the hay first to an ele-
vated platform from where it can be scattered over the system in small
loose bunches (Finn-Kelcey, 1948).

Hay with more than 40 to 45 per cent moisture will become quite
dense when it settles. According to Jennings (unpublished data) set-
tling occurs as fermentation softens the stems. His data also show that
better air flow can be maintained through an 8-foot depth of hay which
is built up at the rate of 2 feet per day than when all of it is loaded
on the drier at one time.

After one 8 or 10-foot layer has been built up and completely dried
it is possible to repeat the process with another layer on top of the
first (Weaver and Wylie, 1939). As the depth of hay increases the
static pressure goes up and increasing losses of air occur by leakage out

302 B. B. MUSGRAVE AND W. K. KENNEDY

the sides of the lower part of the mow. This can be lessened by closing
off the laterals and opening vents along the sides of a centrally located
main duct (Bruhn, 1947a).

Baled hay has been successfully dried over air duct systems by spac-
ing the bales 2 inches apart on the sides and ends and placing successive
layers at right angles to the ones below (Weaver et al. 1947). This
creates an extreme example of variable density but apparently the chief
function of the air is to remove the vapor from between bales while
the energy released by fermentation vaporizes the water. Weaver et al.
(1947) found very little difference in the rate at which heat could be
dissipated from bales spaced 2 inches apart and from a single bale
through which air was forced. The moisture content of baled hay .for
barn drying should not be over 35 per cent and air at a rate of 20 cubic
feet per minute per square foot should be provided (Shedd and Barger,
1947).

Coarsely chopped hay (cutter set at 2 inches) is well adapted to barn
drying. Its density is slightly higher than long hay. Therefore the
maximum depth desirable is about one-fifth less than that for long hay
(Frudden, 1946) . Field chopping from the windrow followed by loading
the drier with a blower and distributor pipe provides the best means
of mechanizing the handling of moist hay for barn drying (Whisler,
1947).

c. Supplemental Heat. In some northern regions with short growing
seasons the hay crop must be harvested rapidly in order to catch it at
the optimum stage of maturity. Cold air driers do not dry hay fast
enough to keep pace with the developing crop. To correct this deficiency,
Bruhn (1947b) and Strait (1944), working in Wisconsin and Minnesota,
respectively, have tried out the addition of heat to the system. As men-
tioned above gasoline engine drives provide waste heat and will raise
such air as they can mov 3 to 5C. Where special heating units are at-
tached to cold air systems usually only a 5 to 8C. temperature rise is
attempted. With a greater differential more air must be circulated to avoid
condensation and molding on top of the mow. Providing extra forced ven-
tilation of the mow space above the hay will also lessen the amount of
condensation. Heated air systems work efficiently early in the drying
cycle and then become very inefficient as dry flues develop in loosely
packed zones through which much of the heated air escapes (Kalbfleish
et al.y 1947). Calculations from experimental data reveal that the ef-
ficiency of utilization of the heat in the fuel to vaporize water, ranges
from 10 to 40 per cent, depending on such factors as the initial moisture
content, uniformity of packing, insulation and the type of heat exchanger

PRESERVATION AND STORAGE OF FORAGE CROPS 303

(Kalbfleish, et al., 1947; Musgrave and Loosli, unpublished data; Bruhn,
1947a; Davis and Barlow, 1948).

In spite of such low efficiency however, many heated systems have
produced dry hay for as little fuel and power expense as cold air systems.
The latter are producing hay with 100 to 200 kilowatt-hours per ton of
dry hay. The cost of the electricity amounts to about one-half of the
overall expense of barn drying.

4. Dry Matter and Nutrient Losses during Barn Curing

The dry matter losses during barn curing arise primarily from fermen-
tation apparently accompanying infestation with a mixed population
of microorganisms. The amount of dry matter consumed or destroyed
by these organisms will depend primarily on the rate of drying and there-
fore to some extent on the initial moisture content. Temperature, rate
of air flow, relative humidity, and nature of the crop are modifying
factors. The average loss as determined to date is about 10 per cent
(Blaser and Turk, unpublished data; Musgrave and Loosli, unpublished
data; Shepherd et a/., 1947, 1949). The hays in which this average loss
was observed went onto cold air driers at moisture contents slightly
above the currently recommended 35 to 40 per cent range. Where
losses were measured on heated air driers the initial moisture was higher
but the dry matter loss was only 5 per cent.

The protein of hay suffers a much higher loss in heated air driers
than does dry matter. The total protein content was reduced 9 to 10
per cent when heated air was used, and 11 per cent without heat, which
was only slightly higher than the dry matter loss. Three New York
tests of drying baled hay with air heated 25C. at moisture contents
ranging from 26 to 64 per cent resulted in an average total protein loss of
18.6 per cent while the dry matter loss was only 4.2 per cent. The varia-
tions in initial moisture had no measurable effect on the protein losses
(Musgrave and Loosli, unpublished data).

Elliott (1947) measured the loss of digestible nutrients during barn
drying of bloom stage timothy. Compared to quick drying at 70C.
barn curing caused a loss of 29.2 per cent of the total digestible nutrients.
The ether-extract and crude fiber fractions suffered the greatest per-
centage decreases but they made up only a small fraction of the digestible
dry matter as determined by rabbits. Absolute losses in the different
nutrient groups contributed equally to the digestible dry matter loss.
Dexter et al. (1947) comparing the feeding value of hay dried at various
depths on a barn drier found that the quality decreased from bottom to
top.

As would be expected by the favorable conditions for oxidative and

304 B. B. MUSGBAVE AND W. K. KENNEDY

enzymatic activity during barn drying, carotene undergoes considerable
destruction. However, numerous reports indicate that this loss, serious
as it is, leaves the hay with satisfactory levels, ranging roughly from
10 to 50 micrograms per gram (Shepherd et al., 1947, 1949). This level
is further decreased during subsequent storage at a rate comparable
to that following other drying techniques. Additional estimates of the
preservation of nutrients by barn driers can be found under VII.

5. Fungicides

Efforts have been made to eliminate the losses occurring in barn cured
hay by killing the microorganisms or inhibiting their respiration with
gas. Sulfur dioxide, ammonia, carbon dioxide, chlorine and formalde-
hyde have been tried without much success (Weaver and Wylie, 1939;
Curtis, unpublished data). Curtis did succeed in cutting the respirational
loss to about one-half with either sulfur dioxide or ammonia but expe-
rienced difficulty in barn drying hay so treated, presumably because an
important source of heat was lost. Some organic fungicides have been
found to be effective in inhibiting respiration but these demand further
study to determine their effect in the rumen (Dawson etal., unpublished
data). It would seem that a fungicide would be beneficial only in a
heated air system where the loss of respirational heat would not seriously
prolong the drying cycle.

VI. ABTIFICIAL DBYING
1. Types of Driers

The process of artificial drying, as it is generally considered, differs
in several ways from barn curing with supplemental heat. Artificial
driers are used to process green forage which remains in the drier just
long enough to evaporate the water. Barn driers on the other hand are
large batch driers. The continuous flow of forage through artificial driers
is accomplished automatically by endless belt, rotary drum, or air blast.
Hand transfer is used with tray driers. Thus the forage is agitated, and
rapid, uniform drying occurs without the possibility of channeling and
loss of hot air through dry zones as is experienced with barn driers. The
temperature of the air blast can be much higher in the artificial drier
because the forage is discharged as soon as it is dry, with little chance
for heat damage.

The efficiency of the utilization of the heat in fuel to evaporate water
from forage during artificial drying is about 50 per cent. This can be
raised to about 80 per cent by employing additional expensive equip-
ment to reclaim the heat of condensation of water in the outgoing air.

PRESERVATION AND STORAGE OF FORAGE CROPS 305

8. Losses

The loss of dry matter during artificial drying usually does not exceed
5 per cent. There is practically no field loss because the standing crop
is transferred directly to truck or trailer by a field chopper which cuts,
chops, and elevates it. The only place where a loss can occur is between
the elevator and the conveyance and then only in the absence of adequate
sideboards and screen or cloth hood. As the time between mowing and
drying is short very little respiration loss takes place.

At the drier, however, even when care is taken to clean up all of the
fine material which collects in and around the machine, a loss of dry
matter always occurs (Barr, 1933). Camburn et al. (1942, 1944) ex-
perienced drier losses ranging from about 1 to 12 per cent. The higher
losses apparently occurred when the final product contained moistures
above 10 or 11 per cent. This is opposite to most of the work reviewed
by Watson (1939), which showed that the greatest dry matter loss
occurred when there was overdrying of the forage. He concluded that
the dry matter loss should not exceed 5 per cent when the drier is operated
carefully.

No loss in the digestibility of the nutrients occurs during artificial
drying so long as the dry forage is not exposed to temperatures above
176C. (Hodgson et a/., 1935). Protein digestibility was seriously re-
duced when an exhaust gas temperature of 205C. was used.

Artificial dehydration preserves carotene well and is the method used
to produce carotene-rich leaf meals. Carotene is lost rapidly in storage
and therefore the superior preservation during drying should not receive
too much weight in estimating the relative merits of various methods of
preservation (Ferguson, 1949).

There is practically no loss of nutrients during storage because there
is insufficient moisture in the dried product to permit fermentation.
Wright (1941) has shown that rehydration in storage may cause a loss
unless precautions are taken to keep the relative humidity of the storage
shelter below 65 per cent.

3. Limitations

The capacity of an artificial drier is constant. The growth curves of
crops for drying fluctuate enormously. In order to secure economical
production from a drier continuous operation is necessary to minimize
a very high overhead expense. This problem is accentuated by the fact
that only the very highest quality forage justifies the minimum cost of
artificial drying. As was pointed out under II grass and legume crops
remain in optimum condition for short periods only. Thus for a drier

306 R. B. MUSGRAVE AND W. K. KENNEDY

to operate economically a 'sequence of crops must be made available.
This type of cropping is in itself expensive and difficult to manage.
Probably the most practical arrangement, if any exists, is to combine
other methods of preservation with a drier of moderate to small capacity.
Such a drier can be kept in operation rather steadily by growing crops
or mixtures with varying maturity dates, and by using nitrogen and other
fertilizers where needed for aftermath cuts. These cuts can be spread
out fairly well over the remainder of the season because of the variety
of crops, and good distribution of initial times of cutting which can be
attained with the various methods of preservation (Moskovitz, 1941).
Partially drying a crop in the field will increase the output of a drier
because its capacity in terms of dry produce is determined by the amount
of water to be evaporated. This practice, however, has the two serious
drawbacks of permitting field curing losses and of making the manage-
ment of the drier difficult. Crops which contain both dry and wet mate-
rial, a condition which usually exists following partial field curing, may
be either variable in moisture content when coming from the drier or
have part of the material damaged by scorching.

VII. EXPERIMENTS COMPARING SILAGE, BARN-CURED
AND FIELD-CURED HAY

With the different methods of harvesting, preserving, and storing
forage crops used at the present time, the loss of nutrients is inevitable,
but the magnitude and type of losses are governed by the particular
method employed. In general, dry matter losses are smallest for dehy-
drated forages; are intermediate in rank and about the same for barn-
cured hay and silage; and greatest for field-cured hay, with most of it
occurring during harvesting (Table I). Harvesting losses are lower for
barn-cured than field-cured hay because by removing the forage from
the field when it has a higher moisture content, fewer leaves are lost by
shattering, and weather damage is decreased. The magnitude of dry
matter losses in field-cured hay is largely governed by the amount of loss
which occurs in the field. Harvesting losses in field-cured hay under
unfavorable conditions were over 75 per cent higher than when hay was
cured without rain (Shepherd et al., 1947, 1949) . The amount of har-
vesting loss occurring in silage probably increases as the degree of field
wilting increases, however, wilting usually improves the quality of the
silage and decreases fermentation losses. Little field loss occurs in de-
hydrating forages if the crop is harvested by a direct cut and loading
machine. Hodgson (1949) reported that barn-curing hay with heated

PRESERVATION AND STORAGE OF FORAGE CROPS

307

air resulted in about 6 per cent less dry matter loss in the mow than
when unheated air was used.

Protein in forage crops is best preserved by ensiling but undergoes
appreciable loss during harvesting when field cured (Table II). Protein
losses during the actual dehydration process are considerably larger than
the dry matter losses. The higher the temperature of the material during
drying the larger the loss of protein as well as the decrease in digestibility
of the protein. Hodgson (1949) showed little benefit in protein preserva-
tion from using heated air as compared with unheated air in the barn
curing of hay.

The percentage of carotene retained is usually higher in silage than

TABLE I

Mean Dry Matter Losses in Roughages Resulting from Different
Methods of Preservation

Dry matter losses reported by:

Method Camburn et al. Turk Shepherd Shepherd

of (1938, 1942, 1944) and et al. et al.

Preservation Newlander et al. Blaser (1947) (1949)

(1938, 1940, 1942) (Unpublished)

Silage

Harvesting

7.7

6.3

2.0

Spoilage 3.3 a

5.0 b

Fermentation 8.1

8.9

10.7 c

11.8 c

Total 11.4

21.6

17.0

13.8

Field Cured

Harvesting

21.3

20.3

36.4

Storage

7.4

4.7

3.2

Total 15.0

28.7

25.0

39.6

Barn Cured

Harvesting

9.4

12.0

4.0

Storage

8.7

7.0

4.5 d

Total

18.1

19.0

8.5

Dehydrated

Harvesting

1.6

Dehydrating process

1.8

Total 9.0

3.4

* Dry matter loss calculated for silo 36 feet high.

b Dry matter loss was 5.0 per cent for top spoilage and 9.1 per cent for side
spoilage. The amount of spoilage probably would be smaller in a tight-wall silo.
Including spoilage and fermentation losses.
d Heated air was used.

308 R. B. MUSGRAVE AND W. K. KENNEDY

in other forms of preservation (Table III). However, the method of
silage making apparently is an influential factor, for Camburn et al.
(1944) reported only 4 per cent preservation of carotene in silage made
with no preservative, but 49 per cent retention with silage preserved
with mineral acids. Carotene is well preserved by dehydration of forage
crops, but field curing and barn curing result in extremely low carotene
retention. According to Hodgson et al, (1947) the butter fat from cows
fed alfalfa silage contained approximately twice as much vitamin A
during January, February and March as the butterfat from cows receiv-
ing U.S. No. 2 alfalfa hay.

TABLE II

Mean Protein Losses in Roughages Resulting from Different Method?

of Preservation

Protein losses reported by:

Method of Camburn, et al. (1938, Shepherd Shepherd

Preservation 1942, 1944) ct at. (1947) ct al (1949)

and

Newlander et al (1938,
1940, 1942)

Silage

Harvesting

107

8.1

Spoilage ft

Fermentation 6.6

4.3 "

8.5 "

Total

15.0

16.6

Field Cured

Harvesting

29.7

47.2

St orage

1.3

3.4

Total 22.8

31.0

50.6

Barn Cured

Harvesting

16.0

7.8

Storage

9.0

9.1 c

Total -

25.0

16.9

Dehydrated

Harvesting

Dehydrating process

10.2

Total 22.1

16.4

a Protein losses due to spoilage were high m their small experimental silos, but
of the same magnitude as the dry matter losses. The loss of dry matter was 3.3
per cent for a silo 36 feet high.

b Includes spoilage and fermentation losses.

c Heated air was used.

PRESERVATION AND STORAGE OF FORAGE CROPS 309

The pounds of total digestible nutrients recovered for each 100 Ibs.
dry matter ensiled or hayed reported by Newlander et al. (1940) and
Camburn et al. (1942, 1944) are: 42.2-45.8 Ibs. for silage, 51.4-60.2 Ibs.
for dehydrated hay, and 42.2-52.0 Ibs. for field cured hay. On the acre
basis, Shepherd et al. (1947) found that in comparison with field-cured
hay, milk production was higher by 12.5 per cent for silage, 16.0 per
cent for barn-cured hay with heated air, and 8.1 per cent for barn-
cured hay with unheated air. They also reported a 20 per cent reduction
in milk production if the hay was damaged by rain while being field
cured. In another study Shepherd et al. (1949) found that in comparison
with poor quality field-cured hay, milk production was higher by 40.3
per cent for silage, 48.2 per cent for barn-cured hay with heated air, and
49.6 per cent for dehydrated hay.

In comparing the feeding values, Turk and Blaser (unpublished) did
not observe any significant difference for unit weight of dry matter pre-
served as silage, barn-cured hay or field-cured hay. Shepherd et al.
(1949) and Kane et al. (1949) reported that field cured hay was slightly
lower in feeding value than silage, barn-cured and dehydrated forages.
Thus, in comparing the four methods of preservation, the loss of dry
matter from the entire crop accounted for the major losses in feeding
value.

TABLE III
Carotene Retention in Roughages Resulting from Different Methods of Preservation

Carotene preservation reported by:

Method of Camburn et al. Camburn et al. Shepherd Shepherd

Preservation (1942) (1944) et al. (1947); et al. (1949)

Hodgson, (1949)

Silage

4-49

3.8

Field Cured

0.6

Barn Cured

5.3

Dehydrated

22.1

VIII. CONCLUSIONS

The feeding value of a forage crop per acre is at a maximum for only
a very short time during its development. Therefore, in order to retain
the full value of a crop, a system of handling, preservation and storage
must be employed which removes the crop from the field rapidly, and at
the same time preserves it so that minimal loss will occur. These opera-
tions often have to be done when weather conditions are not ideal for

310 B. B. MUSGRAVE AND W. K. KENNEDY

the preservation of the forage as dry roughages. Delayed cutting will
result not only in an inferior quality of roughage, but it also may result
in the elimination of certain legumes such as Ladino white clover, Tri-
folium repens L., from legume-grass mixtures. If the legume is elim-
inated from the mixture the yield and quality of the succeeding crops
usually are reduced. Thus the early harvest and storage of forages,
although adverse weather conditions may persist, are necessary for pre-
serving the quality of the existing crop as well as ensuring that of suc-
ceeding ones.

Artificial drying results in the best preservation with least storage
loss, but the low capacity of economically sized units necessitates the
use of other methods when large volumes of forage need to be handled
within a short time. Ensilage is the best system for this purpose because
of its high efficiency for timely handling, though it causes nearly a 20
per cent deterioration of the feed value in the product. Rapid crop
removal may also be accomplished when hay is field cured, but large
field and storage losses result especially with immature, succulent crops.
Barn curing is intermediate with respect to both losses and handling
capacity.

Although silage-making appears to offer the best method of preserv-
ing forages during inclement weather, many unsolved problems contribute
to the uncertainty of the process and hinder its adoption by farmers.
In the future, however, probably a higher percentage of roughage grown
in humid regions will be stored as silage because of the progress being
made by research in solving these problems, the limitations of artificial
heat for drying roughage,' and the difficulty of speeding the field curing
process to a point where it can be accomplished during short periods of
fair weather.

Regardless of the difficulties experienced in producing high quality
field-cured hay in humid regions, hay probably will continue to be pro-
duced. Some hay is a valuable and almost essential component of the
roughage ration of high producing animals, even though large quantities
of silage are fed. Hay is also a more concentrated form of roughage in
terms of digestible nutrients per pound of roughage than is silage. Un-
like silage, it can be transported to areas where stored roughage is scarce
and can be handled more easily when mechanical equipment is lacking.

It is essential that improved preservation practices prevent the dilu-
tion of the energy content of roughages through the loss of the more
digestible portions of the crop as well as the overall loss of dry matter.
More emphasis should be placed on the concentration of roughage
nutrients than on total yield of nutrients. Roughage quality should be
evaluated in terms of the amount of supplemental grain feeding required

PRESERVATION AND STORAGE OF FORAGE CROPS 311

when the energy content of the roughage is reduced by faulty preserva-
tion methods or by later harvesting with the purpose of obtaining higher
yields. The preservation and storage of high quality roughage is the
most serious problem in the field of forage crop improvement and
management.

REFERENCES

Albada, M. 1946. Verslag. Landbouwk. Onderzoek. 52, 112-205.

Allen, L. A., and Harrison, J. 1937. J. Agr. Sci. 27, 271-293.

Amos, A., and Woodman, H. E. 1922. /. Agr. Sci. 12, 337-362.

Anonymous. 1949. Hoard's Dairyman 94, 803.

Archibald, J. G. 1946. J. Agr. Research 72, 277-287.

Archibald, J. G., and Gunness, C. I. 1945. J. Dairy Sci. 28, 321-324.

Archibald, J. G., Bennett, E., and Kuzmeski, J. W. 1946. /. Dairy Sci. 29, 795-800.

Autrey, K. M., Knodt, C. B., and Williams, P. S. 1947. J. Dairy Sci. 30, 775-785.

Babcock, S. M., and Russell, H. L. 1900. Wisconsin Agr. Expt. Sla. 17th Ann.
Rep., pp. 123-141.

Barr, H. T. 1933. Agr. Eng. 14, 131-132.

Bender, C. B., and Bosshardt, D. K. 1939. J. Dairy Sci. 22, 637-651.

van Beynum, J., and Pette, J. W. 1936. Zentr. Bakt. Parasitcnk., Abt. II 94, 413-
433.

van Beynum, J., and Pette, J. W. 1940. Verslag. Landbouwk. Onderzoek. 46C,
397-407.

Bohstedt, G. 1944. Agr. Eng. 25, 388-392.

Brigl, P., and Windheuser, C. 1931. Biedermanns Zcntr. B. Tiererndhr. 3, 220-242.

Brouwer, E. 1937. Biedermanns Zcntr. B. Tiererndhr. 9, 508-524.

Brouwer, E., deRuyter deWildt, J. C., and Dijkstra, N. D. 1933. Verslag. Land-
bouwk. Onderzoek. 39C, 401-463.

Bruhn, H. D. 1947a. Agr. Eng. 28, 202-204, 207.

Bruhn, H. D. 1947b. Agr. Eng. 28, 251-253.

Buchanan, R. E., and Fulmer, E. I. 1928. Physiology and Biochemistry of Bac-
teria. Vol. I, Williams and Wilkins, Baltimore, pp. 387-388.

Camburn, O. M., Ellenberger, H. B., Newlander, J. A., and Jones, C. H. 1938.
Vermont Agr. Expt. Sta. Bull. 434.

Camburn, O. M., Ellenberger, H. B., Jones, C. H., and Crooks, G. C. 1942. Ver-
mont Agr. Expt. Sta. Bull. 494.

Camburn, 0. M., Ellenberger, H. B., Jones, C. H., and Crooks, G. C. 1944. Ver-
mont Agr. Expt. Sta. Bull. 509.

Cooper, T. P. 1917. North Dakota Agr. Expt. Sta. Ann. Rept. 28, 16.

Crampton, E. W., and Jackson, I. R. C. 1944. J. Animal Sci. 3, 333-339.

Crampton, E. W., and Maynard, L. A. 1938. J. Nutrition 15, 383-395.

Crasemann, E., and Heinzl, 0. 1949. Rept. 5th Intern. Grassland Congr. Nether-
lands 32, 1-8.

Cullison, A. E. 1943. Mississippi Farm Research 6, No. 4, p. 8.

Curtis, O. F. 1944. J. Am. Soc. Agron. 36, 401-416.

Davis, R. B., Jr. 1947. Agr. Eng. 28, 289-290, 293.

Davis, R. B., Jr., and Barlow, G. E., Jr. 1948. Agr. Eng. 29, 251-255.

Dawson, J. E., and Musgrave, R. B. 1946. Agr. Eng. 27, 565-566.

312 R. B. MUSGRAVE AND W. K. KENNEDY

Dexter, S. T., Sheldon, W. H,, and Huffman, C. F. 1947. Agr. Eng. 28, 291-293.

Dijkstra, N. D. 1945. Versing. Landbonwk. Onderzoek. 51, 1-38.

Dijkstra, N. D. 1948. Empire J. Exptl. Agr. 16, 231-236.

Dobie, J. B. 1948. Agr. Eng. 29, 160-161.

Duff, G. H., and Forward, D. F. 1949. Can. J. Research 27, 125-145.

Duffee, F. W. 1947. Agr. Eng. 28, 294-296.

Elliott, R. F. 1947. The Effect of Curing Methods on the Nutrient Losses from
Hays. Thesis, Cornell University, Ithaca.

Ellis, G. E., Matrone, G., and Maynard, L. A. 1946. J. Animal tici. 5, 285-297.

Esten, W. M., and Mason, C. J. 1912. Connecticut (Starrs) Agr. Ex pi. tita. Bull.
70.

Ferguson, W. S. 1949. /. Ministry Agr. (Engl.) 55, 517-522.

Finn-Kelcey, P. A. 1948. J. Ministry Agr. (Engl) 55, 348-354.

Fleischmann, F. 1912. Landw. Vcrs. N/a. 76, 236-447.

Forbes, E. B., Elliott, R. F., Swift, R. W., James, W. H., and Smith, V. F. 1946.
J. Animal Set. 5, 298-305.

Frudden, C. E. 1946. Agr. Eng. 27, 109-111.

Galloway, L. D. 1935. ./. Textile ln*l. 26, T123.

Gerlach, M., Gunther, E., and Seidel, C 1929. Lutuhr. Jahrb. 69, 559-639.

Giglioli, I. 1914. Tram. 3rd Intern. Congr. Trojt. Agr. 2, 662-690.

Godden, W. J. 1923. ,/. Agr. Sri. 13, 462-466.

Gorini, C. 1942. Intern. Rev. Agr. Part 1 Agr. Sri. 33, 98T-109T.

Guilbert, H. R. 1935. J. Nutrition 10, 45-62.

Guilbert, H. R., Mead, S. W., and Jackson, H. C. 1931. Hilgardui 6, 13-26.

Hart, E. B., and Willaman, J. J. 1912. J. Am. Chem. Sor. 34, 1619-1625.

Hartwig, H. B. 1942. J. Am. Soc. Agron. 34, 482-485.

Heineman, P. G., and Hixson, C. R. 1921. J. Bad. 6, 45-51.

Hendrix, A. T. 1947. Agr. Eng. 28, 286-288.

Henson, E. R. 1939. Iowa Agr. Expt. Sta. Research Bull. 251.

Higgins, F. L. 1932. Minnesota Agr. Expt. Sta. Tech. Bull. 83.

Hodgson, R. E. 1949. US. Dept. Agr. Bur. Dairy Jnd. Memorandum Inf. 71.

Hodgson, R. E., Davis, R. E., Hosterman, W. H., and Heinton, T. E. 1948. US.

Dept. Agr. Yearbook (Grass), pp. 161-167.

Hodgson, R. E., Knott, J. C., Graves, R. R., and Murer, H. K. 1935. J. Ayr. tie-
search 50, 149-164.
Hodgson, R. E., Shepherd, J. B., Hosterman, W. H., Schoonleber, L. G., Tysdal,

H. M., and Wagner, R.-E. 1947. Agr. Eng. 28, 154-156.
Hoffman A. H. 1923. California Agr. Expt. Sta. Kept., 1922-1923, 50-89.
Hoffman, E. J. 1940. J. Agr. Research 61, 241-257.

Hoffman, E. J., and Bradshaw, M. A. 1937. ./. Agr. Research 54, 159-184.
Hunter, C. A. 1921. J. Agr. Research 21, 767-789.
Hunter, O. W. 1917. J. Agr. Research 10, 75-83.
Hunter, O. W. 1918. J. Agr. Research 15, 571-592.

Hunter, O. W., and Bushnell, L. D. 1916. Kansas Agr. Expt. Sta. Tech. Ball. 2.
Ingham, R. W. 1949. Grass Silage and Dairying. Rutgers Univ. Press, New

Brunswick, p. 31.

Johnson, B. C., Elvehjern, C. A., and Peterson, W. H. 1941a. ,/. Dairy Sci. 24, 86.
Johnson, B. C., Peterson, W. H., Hegsted, D. M., and Bohstedt, G. 1941b. J. Agr.

Research 62, 337-348.
Jones, T. N., and Dudley, R. F. 1948. Agr. Eng. 29, 159, 161.

PRESERVATION AND STORAGE OF FORAGE CROPS 313

Kalbfleish, W., Clarke, M. F., and Ripley, P. 0. 1947. Sri. Agr. 27, 609-624.
Kane, E. A., Wiseman, H. G., and Gary, G. A. 1937. J. Agr. Research 55, 837-847.
Kane, E. A., Jacobson, W. C., and Moore, L. A. 1949. UJS. Dept. Agr. Bur. Dairy

Ind. Memorandum Inf. 83.
Keeney, L. G. 1941. Agr. Enq. 22, 176.

Kellner, O. 1915. The Scientific Feeding of Animals. Duckworth and Co., London.
Kenney, R. 1916. Kansas Slate, Board Agr. (Quart.) Rept., June, pp. 241-247.
Kiesselbach, T. A., and Anderson, A. 1927. J. Am. Soc. Agr on. 19, 116-126.
King, W. A. 1943. /. Dairy Sri. 26, 975-981.

Kirsch, W. 1933. Imp. Bur. Plant Genetics: Herbage Plants Bull. 8.
Kirsch, W., Feeder, K. E., and Lukaczewicz, J. 1934. Biedermanns Zentr. B.

Tiererndhr. 6, 149-158.
Kirsch, W., and Hildebrandt, H. 1930. Die Silofutterhercitung nach den Kalt-

garverfahren. Paul Parey, Berlin.

Knisely, A. L. 1903. Oregon Agr. Expt. Sla., 15th Ann. Rept., pp. 34-38.
Kvasnikov, E. I., and Raev, Z. A. 1939. Microbiology (USSR) 8, 479-480. Ab-
stract. 1941. Chcm. Abstract* 35, 846.
Lamb, A. R. 1917. Iowa Agr. Expt. Ma. Bull. 40.

LeClerc, J. A. 1939. U.S. Dept. Agr. Yearbook (Food and Life), pp. 992-1016.
Lepard, (). L., Page', E., Maynard, L. A., Rasmussen, R. A., and Savage, E. S.

1940. ./. Dairy SVt. 23, 1013-1022.

Lewis, W., and Eden, A. 1949. J. Ministry Agr. (Engl.) 56, 12-15.
Mart os, V. F. 1941. J. Bad. 42, 140-141.

Mikhin, A. M., Fokm, V. M., and Tupikova, A. A. 1936. Problems Animal Hus-
bandry (USSR) 5, 74-94.
Mikhin, A. M., Fokin, V. M., and Tupikova, A. A. 1937. Problems Animal Hm-

bandry (USSR) 7, 142-153.

Miller, D. D., and Golding, N. S. 1949. ./. Dairy tici. 32, 101-110.
Miller, R. C., Silver, E. A., and Willard, G. J. 1934. Ohio Agr. Expt. Sta. Bull.

532, 89-90.

Mills, R. C., and Hart, E. B. 1945. ,7. Dairy Sei. 28, 1-13.
Monrop, G. F., Hilton, J. H., Hodgson, R. E., King, W. A., and Kraus, W. E. 1946.

,/. Dairy Sci. 29, 239-256.
Morrison, F. B. 1948. Feeds and Feeding. 21st ed., Morrison Publishing Co.,

Ithaca.

Moskovitz, I. 1941. Intent. Rev. Agr. 32, 213-279.
Musgrave, R. B., and Dawson, J. E. 1946. Farm Research 12, 18-19.
Neidig, R. E. 1914. Iowa Agr. Expt. Sta. Research Bull. 16.
Neidig, R. E. 1918. J. Agr. Research 14, 395-409.
Newlander, J. A., Ellenberger, H. B., Camburn, O. M., and Jones, C. H. 1938.

Vermont Agr. Expt. Sta. Bull. 430.
Newlander, J. A., Ellenberger, H. B., Camburn, O. M., and Jones, C. H. 1940.

Vermont Agr. Expt. Sta. Bull. 459.
Newlander, J. A., Ellenberger, H. B., and Jones, C. H. 1942. Vermont Agr. Expt.

Sta. Bull. 485.
Overholser, E. L., and Cruess, W. V. 1923. California Agr. Expt. Sta. Tech.

Paper 7.
Pentzer, W. T., Asbury, C. E., and Hamner, K. C. 1933. Proc. Am. Soc. Hort.

Sci. 30, 258-260.
Peterson, W. H., Bird, H. R., and Beeson, W. M. 1937. J. Dairy Sci. 20, 611-623.

314 R. B, MUSGRAWS AND W. K. KENNEDY

Peterson, W. H., Hastings, E. G., and Fred, E. B. 1925. Wisconsin Agr. Expl.

Sta. Research Bull 61.
Phillips, M., Goss, M. J., Beavens, E. A., and James, L. H. 1935. J. Agr. Research

50, 761-776.
Procopio, M. 1942. Chim. Ind. Agr. Biol. 18, 240-248. Abstract. 1944. Chem.

Abstracts 38, 3743.

Rather, H. C., and Morrish, R. H. 1935. Michigan Agr. Expt. Sta. Quart. Bull. 17.
Reed, G. M., and Barber, L. 1917. Missouri Agr. Expt. Sta. Bull. 147, 29.
Reed, O. E., and Fitch, J. B. 1917. Kansas Agr. Expt. Sta. Bull. 217.
Rodenkirchen, J. 1939. Milchw. Forsch. 20, 82-94.
Roethe, H. E. 1937. Agr. Eng. 18, 547-554.

Rogers, C. F. 1949. Ohio Farm and Home Research 34, 124-128.
Rosenberg, A. J., and Nisman, B. 1949. Biochim. et Biophys. Acta 3, 348-357.
Roseveare, G. M. 1948. J. British Grassland Soc. 3, 63-66.
Russell, E. J. 1908. /. Agr. Sci. 2, 392-410.
Saltonstall, L. 1948. The Measurement of the Quantity and Quality of Pasture

Herbage Consumed by Sheep. Thesis, Cornell University, Ithaca.
Samarani, F, 1922. Hoard's Dairyman 63, 806.
Schaller, J. A., Mitchell, Nolan, and Dickerson, W. H., Jr. 1945. Agr. Eng. Pub.

No. 6, Tennessee Valley Authority, Knoxville.
Schieblich, M. 1930. Tiererndhr. 2, 367-375.

Schieblich, M. 1931. Biedermanns Zentr. B. Tiererndhr. 3, 437-449.
Schmidt, K. 1934. Biedermanns Zentr. B. Tiererndhr. 6, 481-497.
Shedd, C. K., and Barger, E. L. 1947. Agr. Eng. 28, 257-258, 260.
Shepherd, J. B., Woodward, T. E., and Melin, C. G. 1946. U.S. Dcpt. Agr. Tech.

Bull. 914.
Shepherd, J. B., Hodgson, R. E., Ellis, N. R., and McCalmont, J. R. 1948. J7..S.

Dept. Agr. Yearbook (Grass), pp. 178-190.
Shepherd, J. B., Hodgson, R. E., Schoenleber, L. G., Tysdal, H. M., Wagner, R. E.,

Hosterman, W. H., Sweetman, W. J., Wiseman, H. G., Melin, C. G., Moore,

L. A., Hein, M. A., and Heinton, T. E. 1947. U.S. Dept. Agr. Bur. Dairy

Ind. Memorandum Inf. 43 Revised.
Shepherd, J. B., Schoenleber, L. G., Hosterman, W. H., Tysdal, H. M., Wagner,

R. E., Sweetman, W. J., Wiseman, H. G., Melin, C. G., Moore, L. A., and

Hein, M. A. 1949. UJS. Dept. Agr. Bur. Dairy Ind. Memorandum Inf. 72.
Sherman, J. M., and Bechdel, S. I. 1918. J. Agr. Research 12, 589-600.
Sisakyan, N. M., and VasU'eva, N. A. 1945. Biokhimiya 10, 117-124. Abstract.

1945. Chem. Abstracts 39, 4164.
Sotola, Jerry. 1933. /. Agr. Research 47, 919-945.
Sotola, Jerry. 1941. J. Agr. Research 63, 427-432.
Stone, R. W., Bechdel, S. I., McAuliffe, H. D., Murdock, F. R., and Malzahn, R. C.

1943. Pennsylvania Agr. Expt. Sta. Bull. 444.
Strait, John. 1944. Agr. Eng. 25, 421-422.
Swift, R. W., Thacker, E. J., Black, A., Bratzler, J. W., and James, W. H. 1947.

J. Animal Sci. 6, 432-444.
Taylor, M. W., Bender, C. B., and Russell, W. C. 1940. New Jersey Agr. Expt.

Sta. Bull. 683.

Terry, C. W. 1947. Agr. Eng. 28, 141-144.
Terry, C. W. 1948. Agr. Eng. 29, 208-209, 214.
Virtanen, A. I. 1932. Schweiz. landw. Monatsh. 10, 257-269.

PRESERVATION AND STORAGE OF FORAGE CROPS 315

rtanen, A. I. 1933. Empire J. Exptl Agr. 1, 143-155.

rtanen, A. I. 1934. Contribution from the laboratory of Valio, No. 2, Helsinki.

linovitch, I., Cheftel, EL, Durocher, J., and Kahane, E. 1949. Compt. rend. 228,

1823-1824.
aksman, S. A. 1947. Microbial Antagonisms and Antibiotic Substances. 2nd ed.,

Commonwealth Fund, New York.

allis, G. C. 1944. South Dakota Ayr. Expt. Sta. Circ. 53.
atson, S. J. 1939. The Science and Practice of Conservation. Grass and Forage

Crops. The Fertilizer and Feeding Stuffs Journal, London, Vols. I and II.
atson, S. J. 1947. J. Roy. Agr. Soc. Engl. 108, 130-142.
atson, S. J. 1948. Edinburgh and East Scotland Coll. Agr. Misc. Pub. 12.
atson, S. J., and Ferguson, W. S. 1937. J. Ministry Agr. (Engl.) 44, 247-260.
atson, S. J., Ferguson, W. S., and Horton, E. A. 1937. J. Agr. Sci. 27, 224-258.
eaver, J. W., Jr. 1937. Agr. Eng. 18, 25-27.
eaver, J. W., Jr., Grinnells, C. D., and Lovvorn, R. L. 1947. Agr. Eng. 28, 301-

307.

eaver, J. W., Jr., and Wylie, C. E. 1939. Tennessee Agr. Expt. Sta. Bull. 170.
hisler, P. A. 1947. Agr. Eng. 28, 497-499.
iegner, G. 1925. Mitt. deut. Landw.-Ges. 40, 321-332.
ilson, J. K. 1935. J. Dairy Sci. 18, 317-325.
ilson, J. K. 1943. J. Am. Soc. Agron. 35, 279-290.
ilson, J. K. 1948a. ./. Am. Soc. Agron. 40, 541-552.
ilson, J. K. 1948b. ./. Am. Soc. Agron. 40, 901-907.
ilson, J, K., and Webb, H. J. 1937. ./. Dairy Sci. 20, 247-263.
indheuser, C., Hoffman, O., and Ohlmer, E. 1935. Bicdcrmanns Zentr. B. Tier-

crndhr. 7, 372-381.

oodman, H. E. 1923. J. Agr. Sci. 13, 240-242.
oodman, H. E., and Evans, R. E. 1935. ,/. Agr. Sci. 25, 578-597.
oodman, H. E., and Hanley, F. 1926. J. Agr. Sci. 16, 24-50.
oodward, T. E. 1939. U.S. Dcpt. Agr. Yearbook (Food and Life), pp. 592-596.
oodward, T. E., and Shepherd, J. B. 1936. J. Dairy Sci. 19, 697-706.
right, N. C. 1941. J. Agr. Svi. 31, 194-211.
imm, E. W. 1935. Proc. Roy. Soc. London B117, 504-525.
imm, E. W. 1937. Proc. Roy. Soc. London B123, 243-273.
ak, F. J. 1936. Agr. Eng. 17, 329-330.

The Reclamation of Coal Mine Spoils*

HELMUT KOHNKE

Purdue University Agricultural Experiment Station, Lafayette, Indiana

CONTENTS

Page

I. Introduction 318

II. The Condition of the Spoil Banks 319

1. Topography and Erosion 319

2. Soil Material 322

a. Physical Conditions 322

b. Chemical Conditions 323

c. Soil Profile Development 325

3. Water 326

4. Volunteer Vegetation 328

5. Plantings 329

6. Wild Life and Fish 330

III. Methods of Testing Spoil Bank Materials 330

IV. Land Use Capabilities 332

1. Land Use Capability; Classification of Ungraded Spoils 332

2. Land Use Capability; Classification of Graded Spoils 333

a. Rotation Crops and Orchards 334

b. Improved Pastures 334

c. Forests 334

d. Other Uses 335

V. Methods of Revegetation 335

1. Forest Trees 335

2. Pastures 336

a. Unleveled Spoils 336

b. Graded Spoils 337

3. Rotation Crops and Orchards 338

VI. Grading 338

1. Reasons for Grading Spoils 338

2. How to Grade Spoils 339

3. Objections to Grading of Spoils 341

VII. Economic Aspects 341

VIII. Legislation 344

IX. Discussion 344

1. Establishing a Reclamation Policy 344

2. Future Research Needs 346

3. Contributions of Spoil Bank Research to Science 346

X. Summary 347

XI. Acknowledgments 348

References 349

* Approved by the Director as Journal Paper number 445.

318

HELMUT KOHNKE

I. INTRODUCTION

More than a fifth of all the coal mined in the United States in recent
years has been produced by the open cut method. (Fig. 1). This "strip
mining" of coal started about the time of the Civil War but remained
insignificant until the beginning of the twentieth century. In 1914 only
0.3 per cent of the total coal production in the United States was mined by
stripping. By 1948 this proportion had climbed to 23.3 per cent (Young
et aL, 1949). Less labor is needed to produce a ton of coal by stripping
than by underground mining. Fewer accidents occur in stripping (Adams
and Geyer, 1944; Sinks, 1946). As a rule strip mining recovers over

Spoil Bonk*

v .

Coal Removed- X- ---- ^ ----- x

o 25 so 75 100 feet

Fig. 1. Schematic cross section of abandoned strip coal mine.

95 per cent of the coal in place while underground mining recovers only
50-60 per cent, largely because of the necessity of leaving pillars for
roof support. The equipment needed to remove overburden economically
to greater depths has only become available during the last 30 years.

Strip mining leaves the overburden in steep and jagged ridges that
make the landscape appear desolate and unproductive (Figs. 2 and 3).
About a quarter million acres have been stripped for coal so far (January,
1950) in the United States and at least half again this area has been
covered with overburden or otherwise made unfit for farming purposes.
The end of strip mining for coal is not in sight. No accurate estimates
of the potential coal field acreage exist but there is little doubt that the
half way mark has not yet been reached. The states in which strip min-
ing is of importance are Indiana, Illinois, Ohio, West Virginia, Pennsyl-
vania, Missouri, Kansas, Oklahoma, Kentucky, Alabama, Arkansas and
Iowa.

This article is written in an attempt to clarify the physical conditions

THE RECLAMATION OF COAL MINE SPOILS

319

of coal mine spoil banks as they affect plant growth and consequently
land use and to discuss the reclamation of these areas.

Fig. 2. Coal stripping operations. (Courtesy Indiana Coal Producers Association.)

II. THE CONDITION OF THE SPOIL BANKS

To plan reclamation work of the coal mine stopping's a detailed
knowledge of their conditions is necessary. In the following paragraphs
topography, erosion, soil material, water, vegetation and wild life will be
considered as they are found on worked out strip mines.

1. Topography and Erosion

Topographically, coal mine stoppings consist of a series of parallel
ridges and troughs and a deep, long cut with one very steep slope on one
side and a somewhat" flatter one on the other. The origin of these fea-
tures is illustrated in Fig. 1. The ratio between these two components

320

HELMUT KOHNKE

depends largely on the original slope of the terrain. In steep country
only one or two cuts of coal are possible because of the increasing thick-
ness of the overburden. The final cut takes up as much space as the
series of ridges but generally the area covered by the troughs and ridges
greatly predominates.

Where the overburden has been removed with a power shovel the
ridges are of approximately the same height for some distance; where a
drag-line is used they are composed of a series of hillocks. As these ma-
chines are operated to deposit the overburden as far from the exposed coal

Fig. 3. Fresh coal mine spoils. (Courtesy Indiana Coal Producers Association.)

THE RECLAMATION OF COAL MINE SPOILS 321

as possible, no uniform spreading of the material over the area is attained.
Each bucket load is dropped on or near the previous one and the earth
and stones slide down on both sides forming a ridge. The natural angle
of repose of this material is from 80 to 90 per cent. The more stony
the overburden the steeper is the angle of repose. During the first year
or two considerable settling occurs. Since in the center of the ridge the
depth of material and therefore the total depth of settling is greater
than in the troughs, the slopes of the older spoils may be reduced to 60
and 70 per cent. Very shaley spoils may stay as steep as 80 per cent
after as much as 25 years.

Erosion is intense on fresh spoils, as long as they are not covered by a
mantle of vegetation. No water stable aggregates exist in this material
which is practically devoid of organic matter. Only the many rock and
shale fragments protect the banks from erosion. In spite of the fact
that the ridges are too short for runoff to concentrate these are frequently
riddled by gullies. Since the steepness of the slopes permits any soil
material that is detached from the ground to be transported downhill
by splash or run-off the sandy spoils erode faster than the banks which
contain enough clay to give some cohesiveness. Under the less sloping
conditions of an ordinary field, sand is usually less easily eroded than
clay loam. Frequent observations and actual measurements (Stiver,
1949) show that erosion rounds off the tops of the ridges and takes away
sheets of soil material along the sides but does not actually reduce the
slopes of the ridges. The eroded material accumulates in the troughs and
creates a nearly flat colluvium composed of sand, silt and clay. In many
cases erosion in the troughs is sufficiently active to prevent such an ac-
cumulation, and the eroded material is carried on down into the larger
depressions within the spoil area. In flat or slightly rolling country it is
an exception for coarse eroded material to be washed more than 50 feet
beyond the slopes of the spoil banks on to unmined land. Some of the
clay and the finer silt is sometimes carried in the streams issuing from a
mined area but no damage resulting from this has been reported or
observed. In hilly land erosional debris sometimes covers creek bottoms,
overloads small streams and increases flood hazards (Tyner and Smith,
1945).

The distance from ridge to ridge varies with the stripping method
even within the same operational area. It ranges from 35 to 70 feet
with an average near 50 feet. The differences in elevation between the
tops and the troughs depend on the distances from ridge to ridge, on
the steepness of the slopes and on the amount of previous smoothing of
the ridges and filling of the troughs. Differences of from 5 to 30 feet have
been measured. Most of them are between 10 to 20 feet.

322 HELMUT KOHNKE

The final cut from which coal is taken remains open, since no more
overburden is moved. This pit is about 50 feet wide at the bottom. The
side toward the spoil banks has the normal angle of repose of unconsoli-
dated material. The other side, facing the unmined land, is much steeper.
It is called the "high wall." Usually enough sloughs off from this wall
so that the lower 5 or 10 feet are covered with rock and earth debris of
about 80 per cent slope. The area from which the coal was removed and
which has not been covered again is frequently nearly horizontal since
most coal deposits which are strip mined do not have much of a tilt.
The depth of the final cut depends entirely on the depth of overburden
removed. This varies from a few feet to as much as 90 and even 100 feet.

2. Soil Material

The overburden which has been deposited represents potential soil
material. Definitions for soil vary, but here it is considered as the
unconsolidated covering of the earth capable of supporting plant growth.
On this basis the bulk of the material that makes up the spoils is soil.
Since it is rock material that is just beginning the normal processes of soil
formation, it may be thought of as soil at zero time. Some of the spoil
material that is initially so acid that plants cannot exist on it may be
called soil material, or potential soil, since it too will eventually sup-
port vegetation.

a. Physical Conditions. The outstanding characteristic of the spoil
material is its rockiness. It is difficult to make quantitative statements
on the proportion of rocks and "fine earth" (material that can pass a
2 mm. sieve) that are of any general application. Some spoils are cov-
ered with an abundance of large and small rocks, while elsewhere only a
few small stones indicate that the site is not just another field. The
amount of rocks at the surface depends on the geologic strata overlying
the coal, on the excavation methods employed and on the time elapsed
since the mining operations. All coal is deposited between layers of
shale, sandstone, or sometimes limestone. Occasionally soil has been
formed directly from these rocks. In other cases much of these sedimen-
tary rocks has been eroded and replaced by glacial drift alluvium or loess.
The uncovering of the coal is done in the most economical manner. Oc-
casionally this calls for depositing the large rocks at the bottom of the
growing pile of spoil material in order to prevent slipping of the spoil
upon the unexcavated coal. In other mines the earthy material of the
top 6 to 12 feet of overburden is pushed into the pit after the coal is
taken out and the rocks of the lower strata are then deposited on top.

The fine earth material on the spoil is predominantly loam, with

THE RECLAMATION OF COAL MINE SPOILS 323

smaller areas consisting of clay or of sand (Limstrom, 1948; Stiver,
1949). An interesting feature is the complete absence of any soil hori-
zons in the fresh spoil banks. The structure of spoil material is generally
loose initially but as time goes on it settles to a rather compact mass.
The high concentration of electrolytes, especially calcium sulfate, in the
spoil tends to keep the clay in a flocculated condition. The dearth of
organic matter prevents the formation of truly water-stable aggregates.

Of particular importance are the moisture conditions of the spoils,
since they determine to a large extent the plant growth potential. The
rather loose structure and the resulting large pores existing in many
spoils allow water to accumulate in the soil without being pulled down
by the combined effects of capillarity and gravity that regulate the mois-
ture content of field soils. In spoils the capillarity is broken in many
points of the soil body and the result is a local accumulation of nearly
tension-free water. On hot, sunny days, a week after rain, it is not
difficult to find water drops at 5 or 6 inches depth in shaley spoils. All
observations indicate that spoils are well supplied with soil moisture.
While surface runoff is large on fresh spoil, little water is lost by trans-
piration, and after vegetation is well established the infiltration capacity
increases. The surface crust of soil on bare, exposed spoil banks dries
out considerably during the summer months. The growth of such water
tolerant plant species as black willow, Salix nigra, Marsh, river birch,
Betula nigra, L., and large smart weeds, Polygonum sp., on the crests of
many of the ridges is clear evidence of the abundance of water. Stagnant
water occurs only in some of the depressions of the troughs.

Temperature conditions in spoils are similar to those in field soils.
This is especially true after vegetation has been established. On bare
spoils summer temperatures on the sides facing the sun exceed by 5 to
8C. corresponding field soil temperatures (Stiver, 1949). This causes
difficulties in the establishment of young seedling plants in such exposed
positions.

6. Chemical Conditions. Of the chemical soil conditions reaction is
the most important on the spoil banks. pH values of 2.5 and 8.0 can
sometimes be found only inches apart. Roof coal and other strata of
the carboniferous age contain pyrites and other forms of iron sulfide.
Upon oxidation in the presence of water the sulfide forms sulfuric acid:

2 FeS 2 + 7 2 + 2 H 2 - 2 FeS0 4 + 2 H 2 S0 4 .

The concentration of acid may be so high that water vapor from the at-
mosphere is attracted hygroscopically and during dry periods the acid
appears as wet spots on the spoils. The many limestone fragments and

324 HELMUT KOHNKE

the calcareous till account for the high pH areas. Some of the shales are
approximately neutral but usually they are slightly acid. Much of the
sandstone is strongly acid. The original surface soil is sometimes also
quite acid, depending upon parent material and the period of weathering.
While such great variations of reaction occur on most spoils, it is gener-
ally possible, nevertheless, to classify them according to the dominant pH.

The amount of total nitrogen in the fresh spoils is even lower than
that of poorer field soils (Stiver, 1949) and many observations show
clearly that nitrogen is the prime limiting factor of plant growth on most
spoils.

By chemical quick tests available phosphate has frequently been
shown to be higher in spoil material than in the neighboring undisturbed
field soil. Detailed greenhouse experiments by Stiver (1949), however,
clearly pointed to a deficiency in phosphate in 3 entirely different mate-
rials from spoil banks.

Available potash has practically always been found in adequate
amounts similar to those in field soils. Calcium is usually plentiful even
in spoils of low pH because of the many calcium bearing rocks present.
Sulfur has always been abundant, whenever spoils have been tested for
this element. Studies for other nutrients have not been reported. The
ready seed set of alfalfa, lespedeza and birdsfoot trefoil on calcareous
spoils might serve as an indication that no serious plant nutrient de-
ficiencies exist, except nitrogen.

The studies of the plant nutrient status of spoils have been few, but
application of our knowledge of soil fertility and plant nutrition point
to the fact that continuous cropping without fertilization will be as im-
possible on spoils as it is on field soils. It is probable though that the
fresh material on the spoils will contain some of the nutrients in rela-
tive abundance that have been depleted in many ordinary soils. At any
rate it is not a priori obvious, as some people think, that stripping the
ground for coal decreases "soil fertility. In many cases it does, but some-
times it is advantageous. It has been observed that cattle may prefer
to graze on spoils in preference to a neglected old established pasture
nearby.

The question is open whether toxic materials exist in spoil banks in
addition to the acid formed from the sulfide and the ions brought into
solution by this acid. Occasionally spoil banks remain bare of any vege-
tation for a number of years in spite of a nearly neutral reaction. Per-
haps this points to some toxic condition since seed sources for some of the
pioneer vegetation are always available. If such a condition exists, it is
not of long duration and therefore of no concern.

THE RECLAMATION OF COAL MINE SPOILS

325

c. Soil Profile Development. At the time the overburden is piled into
mounds and ridges, no differentiation into horizons exists. Within less
than a year soil formation at the surface becomes apparent. The first
evidence is the breaking down of rocks. The sudden change from the
depths of the earth, where the rocks have been in equilibrium with the
environment, to the surface, where they are exposed to the extremes of
the atmosphere causes rapid disintegration. Shales, sandstones and some
types of limestones break down to soil size within a few years. Many

Reaction

Total Nitrogen

s 10

Depth in Inches

.20

.15

* .10

.06
.00

Alluvium

Slope

5 10

Depth in Inches

Available Phosphate

Available Potash

ISO

< 120

. 90

5 10

Depth in Inches

ISO

120

*^_ X. Alluvium

~*~Sloi* *

5 10

Depth in Inches

Fig. 4. Soil characteristics of .spoil bank material as affected by slope.

imestones and igneous boulders are more resistant, but none have been
encountered lying on the spoil banks that have not suffered considerable
exfoliation during 2 or 3 years 1 exposure. Very few large rocks exist on
surfaces of spoils that are over 20 years old, indicating that they have
iisintegrated in the interim. With the growth and decay of the pioneer
Dlants organic matter is added to the soil. The visible change of soil
?olor on the slopes of spoils has not been found to exceed a depth of
>ne inch in a quarter of a century. The mechanical breakdown of the
ock extends considerably farther. But it is difficult to ascertain the

326 HELMUT KOHNKE

exact depth of this development, since the boundary is imperceptible.
While only this very shallow change of soil color, or development of an
A horizon, occurs on the slopes, soil of desirable texture, color and pro-
ductivity accumulates in the troughs. These widen out as a result of con-
tinuous additions of erosional debris and increase in fertility as the
nutrients wash down on them from the slopes.

Figure 4 shows the different types of soil profile developed on the
slope and in the troughs of a spoil area that had been planted to red
pines 20 years previously. The conclusion is inescapable that a deepen-
ing of the soil profile and an increase in soil fertility on the spoil banks
will be slow, if it occurs at all, as long as the slope remains around 60
per cent. Experience with naturally steep soils substantiates this
assumption.

One of the most important problems of soil development on the
spoils is the changing of the sulfuric acid spots into "soil." Since sulfuric
acid is water soluble it will be leached out and washed off freely by rain
water. The oxidation of sulfide is a slow process. Fortunately, fair
amounts of calcareous materials occur in most spoils and help to neu-
tralize the sulfuric acid and also serve as starting points for vegetation.
Tyner and Smith (1945) have shown that the rate of sulfur removal on
spoil banks is quite rapid.

3. Water

Most of the strip coal mines of the United States are in areas of gen-
erous rainfall. In many mines water must be pumped out of the pit to
permit mining. After operations end, water accumulates in the pit unless
the slope of the coal has been in the direction of the general slope of
the land. Within the ridge and trough area the lower depressions are
also frequently filled with water. As a matter of fact strip mining
adds a great number of ponds and lakes to regions that have had little
open water previously. *

To evaluate this newly created resource the following items require
consideration: chemical quality of the water, permanence of the pond
or stream, area and depth of the water.

Wherever pyrite is exposed to air and moisture it oxidizes and forms
sulfuric acid and ferrous sulfate. Ponds that receive drainage from such
material are strongly acid (up to 0.1N acid) and high in iron unless
the inflow of the acid is relatively small and is neutralized by a cor-
responding influx of lime. Such very acid waters permit no biotic devel-
opment and corrode machinery if an attempt is made to utilize them
industrially.

If most of the iron sulfide is below the water level, oxidation is

THE RECLAMATION OF COAL MINE SPOILS

327

prevented and the water is nearly neutral to slightly alkaline as a result
of the abundance of calcareous material frequently found. Such water
can be used as fish grounds, water supply for livestock, swimming, and
for industrial purposes. (Fig. 5).

Practically all mine waters are high in sulfate, but contain little or
no chloride. They contain much iron and aluminum if they are more
acid than pH 3.5 and much calcium and some magnesium if they are
above pH 4.5. The smaller ponds have no economic value, as they readily
dry up during the late summer.

Fig. 5. Pond in final cut of strip mine. (Courtesy Indiana Coal Producers Asso-
ciation.)

The final cuts of the mines present the best possibilities for deep,
reliable and sound water. This is of course only the case where all sur-
rounding land is higher so that the water will cover the pyritic material.
It has sometimes been observed that a mine pit first fills up with acid,
iron bearing water, but eventually, as the water covers the pyrites and as
lime washes into the pond, the water is neutralized so that it becomes
a good habitat for vegetation and fish.

While much of the pond water in the strip mine areas is nearly neu-

328 HELMUT KOHNKE

tral, most of the running water is acid. The reason for this is that the
presence of a stream presupposes entry of air into the strata above the
sources of the stream and, therefore, washing of the products of the oxi-
dation of pyrites into the creeks is a probability. The degree of acidity
depends on the relative amounts of pyrite, lime, and water. Generally
the farther away from the mine the higher is the pH of the creek water
because of the neutralization by calcareous material and by the calcium
bicarbonate present in most natural waters.

By comparison with drift mines the production of acid stream water
by strip mines is of minor importance. In areas of extensive drift min-
ing, even larger rivers have reactions near pH 3.0.

4. Volunteer Vegetation

Very soon after land has been stripped for coal, vegetation of some
sort appears on the spoil banks. If the soil reaction is very acid, and
sometimes for other reasons, this takes a number of years, but usually
a few individual plants make their appearance in the first year. De-
pending on the potential fertility of the spoil banks and the source of
seed the rate of establishment of a plant cover will vary. The multitude
of volunteer species that grow on spoil banks is truly amazing. Mc-
Dougall (1925), Croxton (1928), Maloney (1941), Riley (1947) and
Stiver (1949) have enumerated such species. They are predominantly
those with airborne seeds, or those that have been introduced into the
spoils in those portions of the original top soil that were deposited at the
surface of the spoils. Of the trees, cottonwood, Populus deltoides, Bartr.,
sycamore, Platanus occidentalis, L., and black willow, Salix nigra, L. are
the most common invaders. All three of these species normally grow in
a rather humid environment. Their occurrence and vigorous growth on
the spoils bespeaks the moist conditions found on recent strippings.
Wherever calcareous material occurs sweet clover, MelilotuA spp., invades
the banks. After a few^ plants of it are established, it spreads quickly
over large areas. Without artificial inoculation it seems never to suffer
from nitrogen deficiency. It is clearly one of the best site preparers for
grasses on the spoils because it provides nitrogen and surface mulch.

Among the other early pioneers on the spoils are the wild aster, Aster >
spp., the prickly lettuce, Lactuca scariola, and the blackberry, Rubus
allegheniensis, Porter. The rate of establishment of vegetation and of the
complete covering of the spoils is a rather clear and simple key to their
potential ability in supporting crop plants. Practical foresters use the
appearance of the first "fuzz" on the banks as an indication that the
spoils may safely be planted to trees.

The troughs are usually more quickly and more densely vegetated

THE RECLAMATION OF COAL MINE SPOILS 329

than the slopes and the ridges. This difference is particularly obvious
on the more acid spoils where generally a marked differential in pH
exists in favor of the troughs. Here the abundant water has removed
the free acid while lime fragments have been deposited and neutralize
the acid as it is formed.

The occurrence of some weeds in very acid conditions is one of the
ecologic puzzles of the spoil banks. Cattails, Typha latifolia, L. have
been found to thrive in mine water of pH 2.5 (Stiver, 1949) while vig-
orous plants of ragweed, Ambrosia elatior, L. grow on spoil material
below pH 3.0.

5. Plantings

Reclamation of coal mine spoils in the United States by planting
orchard and forest trees was started about 1918. The first plantings
were rather sporadic and many of them were done only along the outside
of spoils especially near highways. Gradually an increasing percentage
of the spoils was planted to forest trees, with pines and black locust pre-
dominating. The adapted species of pine have grown well on many of
the spoils (Fig. 6). After a promising start, black locust, where planted
alone, has, in many cases, succumbed to the ravages of the locust borer.
Nevertheless foresters believe that it has been valuable as a site condi-
tioner for other trees. Tree plantations on coal mine spoils have been
discussed by many authors (Chapman, 1944, 1947, 1948, 1949; Mather

Fig. 6. Pine plantation on coal mine spoils. (Courtesy Indiana Coal Producers
Association.)

330 HELMUT KOHNKB

and Mclntosh, 1947; Den Uyl, 1946, 1947; Limstrom, 1948; Sawyer,
1946a, 1946b, 1949; Schavilje, 1941; Stiver, 1949; Toenges, 1939; Win-
chell, 1948). Beside many other timber species, fruit trees and shrubs
have been planted with varying success. On coal spoils in Clay County,
Indiana, there is an abandoned pear orchard that was planted in 1918 and
still produces fruit. At that time strip mining was shallow, much top
soil was mixed into the surface of the banks and the height differences
between troughs and ridges are only 4 to 5 feet. Productive orchards
and vineyards have been observed on flattened spoils in Illinois. Many
of the calcareous spoils have been seeded to pasture. The establishment
of forage plants on spoils is described by Brown (1949), Croxton (1928),
Grandt (1949), Holmes (1944), Sinks (1946), Stiver (1949), Tyner
and Smith (1945), and Tyner, Smith and Galpin (1948). The seed set
of legumes on calcareous spoils is frequently better than on undisturbed
field soils. Where the areas are large enough and where non-mined land
belongs to the same holding, reasonable profits have been obtained. The
abundance of minerals in the spoils and the many ponds present unques-
tionably contribute to such encouraging results.

6. Wildlife and Fish

Where spoil banks are left to revegetate themselves or where forest
trees have been planted, not much feed for wildlife exists. Nevertheless,
the good cover and the many ponds seem to encourage some of the
species, especially near the borders of the mined areas. Increased num-
bers of rabbits and of non-game birds have been observed in the spoils
(Riley, 1947) but also foxes and raccoons seem to thrive. In certain
regions the spoils are regularly visited by deer which evidently appre-
ciate their inaccessibility and the shelter provided by the tree plantations.

Most of the larger bodies of nonacid water are stocked with fish.
The land owners or neighboring fishermen have stocked some of the
ponds in anticipation of some good sport but in many cases birds must
have been responsible for the introduction of the spawn.

III. METHODS OF TESTING SPOIL BANK MATEBIALS

In view of the tremendous variability of soil material in the spoils
quick qualitative methods of testing are of more immediate need than
quantitative and time-consuming analyses. On most spoils soil condi-
tions change so much from spot to spot that it becomes difficult to take
a representative sample.

The three outstanding tests are for carbonate, for free acid and for
sulfide.

THE RECLAMATION OF COAL MINE SPOILS 331

Dilute hydrochloric acid (HC1, 1:10) dropped on the ground reveals
the presence of carbonate (usually calcium carbonate) by gas bubble
formation.

Potassium thiosulphate (10 per cent solution of KCNS) combines
with ferric iron to form a complex ion of deep red color. This reaction
occurs only in a very acid medium because ferric iron is practically in-
soluble at reactions above pH 3. The appearance of red color indicates
the presence of hydrogen ion concentrations considerably in excess of
those produced by hydrogen clay and therefore too acid for plant growth.
This test is simple in the extreme as it only consists in dropping the solu-
tion on the ground and in watching for the color. The intensity of the
color is somewhat indicative of the concentration of the acid.

Sodium azide (3 g. sodium azide dissolved in 100 ml. 0.1 N iodine
solution) is used to detect the presence of sulfides. This test is unneces-
sary after the material has been lying on the surface of the bank for
several months. By that time enough of the sulfur has oxidized, so
that its presence, in the form of acid, can be detected by the potassium
thiocyanate test. But the azide test is invaluable to detect acid forming
substances in the "highwall" and in fresh spoils. For this test a piece
of the size of a pinhead is placed in a small glass vial with a conic bot-
tom, a drop or two of the azide solution is added. Sulfide acts as the
catalyst in the following reaction :

2NaN 3 + I 2 > 2NaI + 3N 2 (gas).

Therefore the formation of gas bubbles indicates the presence of sulfide.

Indicators of the hydrogen ion concentration within the range from
pH 3 to pH 7 are next in importance. They can be used as solutions or
in impregnated paper. Any of the customary organic dye indicators can
be used.

Other soil tests are also applicable to spoil material, for instance
those for total nitrogen, available phosphate and potash and lime re-
quirement. A more reliable picture of the plant nutrient condition
can be received by plant tissue tests (Thornton et al., 1939).

The physical conditions within spoils also have a great influence on
their land use capability. The two most important ones for this purpose
are the texture and the rate of decomposition of the rocks.

The content of the medium-sized and smaller rocks can be deter-
mined by screening through a half inch hardware cloth screen. The
texture of the finer material can be determined by the use of screens
and the pipette or hydrometer method of mechanical analysis.

The rate of decomposition of rocks can be estimated during the first
inspection of the spoil bank by observing the appearance of the medium

332 HELMUT KOHNKE

sized and larger rocks. Where doubt exists it is best to photograph a
number of rocks and to mark the locations and to return one or several
years later and to compare the condition of the rocks.

IV. LAND USE CAPABILITIES

To the casual observer of a fresh spoil with its ragged mounds, its
stony surface and its forbidding colors any land use seems out of the
question. But a closer scrutiny of its topography, its soil and its biotic
characteristics makes it clear that such land can well be made to serve
mankind. In the preceding pages an effort has been made to give a pic-
ture of these conditions in order to provide the background for the de-
cision as to which land use may be appropriate in each individual case.

The forms of potential land use of coal mine spoils include any use
possible under the given climatic and economic conditions, i.e., various
forms of agriculture and horticulture, pasture, forest, fish production, and
recreation. The most adapted land use for a given spoil will depend
on the present slope conditions and our willingness to alter them, on
the stoniness, texture and leaction of the spoil, on the distribution and
nature of the surface water, and the total area involved. In order to
classify spoils according to their land use capabilities, we have first to
decide whether to smooth down the topography to a certain desired
maximum slope or to utilize the banks as they are left after the mining
operation.

1. Land Use Capability; Classification of Ungraded Spoils

If it is assumed that no mechanical smoothing out of the ridges will
be done most forms of agriculture and horticulture are excluded and
forestry and range type pasturing remain. Obviously, fish raising and
recreation are always possible if there is sufficient good water available
and after vegetation has-covered the spoils.

It is rather simple to decide whether a certain spoil area would be
better adapted to forestry or pasture. Since pasture is normally the
higher form of use, a spoil capable of supporting forage plants over its
entire surface and readily accessible to livestock should be used for
pasture. It must be remembered that the steepness of ungraded spoils
precludes liming, fertilizing, and mowing. This means in practice that
the majority of the spoil should contain enough calcareous material
to support legumes. Legumes are necessary for a pasture on an un-
leveled spoil because they are the only major source of nitrogen. On
the other hand, an excess of large, slowly disintegrating rocks makes
a spoil unsuited for pasture. This means that spoils generously inter-

THE RECLAMATION OF COAL MINE SPOILS

333

mixed with glacial, alluvial or loessial material or with soft limestones
are the only ones valuable for pasture. Even then the type of livestock is
restricted to beef cattle and sheep. The steep topography and the rough-
ness of the forage preclude the profitable use of dairy cattle. Other fea-
tures in addition to topography and soil material determine the value of
a given spoil for pasture. Large size of the potential pasture area,
the possibility of its integration with an existing farming enterprise and
the presence of well distributed ponds enhance its usefulness.

All other spoils, having an average reaction below pH 6.5, or having
an excess of rocks, are potential forest land. Wherever sufficient free
acid exists to depress the reaction below pH 3.5 no revegetation should
be attempted until sufficient oxidation of the pyritic materials and leach-
ing of the acid has occurred. Fortunately only a very minor percentage
belongs to this latter group.

2. Land Use Capability; Classification of Graded Spoils

This classification presupposes the grading of excessively steep slopes
where considered necessary, but no resurfacing of the mined area with the

90%

70%

25%

10%

Fig. 7. Maximum slopes for various land uses on spoil banks.

334 HELMUT KOHNKE

original top soil. Where this latter elaborate improvement is under-
taken, as for instance in England (Brown, 1949), the land is fit for
practically any use after an initial 5- to 10-year period in a well-fer-
tilized grass and legume mixture to reestablish soil structure.

Grading of spoils to smoother slopes or to nearly level topography
permits the use of any implement required for the improvement of the
soil, and for planting, treating, and harvesting any crop (Fig. 7) .

a. Rotation Crops and Orchards. Spoils containing few stones and
no free acid can be developed into rotation and orchard land by grading
them to slopes of less than 10 per cent. They will require intensive im-
provement during the first years to raise the pH if necessary, to supply
the necessary plant nutrients and to develop a satisfactory structure.
Only a small fraction of the spoil observed can be included in this group.
The majority has too high a rock content. Of course, if time is allowed
for rock disintegration to occur after leveling, much more land could be
included in this group.

6. Improved Pastures. The soil material requirements for potential
pasture land are much lower if the spoils are graded than if they are left
in steep ridges. The grading for pasture land need only reduce the slopes
to an extent that fertilization, cultivation and mowing are possible by
use of power driven implements. This means grading to a* maximum
slope of 25 per cent. With this prerequisite the range of potential pas-
ture land on spoils becomes quite large. It includes all sites not ex-
cessively rocky or sandy. If a quick return from the pasture is desired
this excludes areas of excessive free acid, since these will not permit
plant growth for some time. It is very important to recognize that the
lower pH range of potential pasture land is around 6.5 if the spoils are
not smoothed out, but that it can go as low as pH 4.0 if leveling is
practiced. This is because of the possibility of liming and fertilizing.
Moreover, there is no restriction on the type of livestock used on graded
spoil pastures.

c. Forests. The growth of trees on ungraded spoils indicates that
no leveling work is required to prepare forest sites. Experiments by the
U.S. Forest Service (Chapman, 1949) showed even a decrease in height
growth of young hardwood plantations on graded spoils when compared
to ungraded spoils. Whether this site impairment is temporary or per-
manent and whether eventually the situation may be reversed is an
open question that will receive attention under VI.

All spoils that cannot be used for crops, orchards, or pastures by

THE RECLAMATION OF COAL MINE SPOILS 335

grading and that have a reaction above pH 3.5 are best used for forest.
The only grading required then is for the establishment of access roads
and fire lanes.

d. Other Uses. Spoils containing large areas affected by free acid
cannot be reasonably used for any plant growth because the amount of
lime required to neutralize the acid is excessive. Even then more acid
is formed by the further oxidation of sulfides. The only practical method
is to wait until most of the sulfides have been oxidized and the resulting
acid leached out.

The suggestion has been made to grind highly pyritic material and
to use it to neutralize alkaline mucks and greenhouse soils. So far this
has not been tried.

The use of the ponds for fishing and other forms of recreation is
rather obvious and requires no elaboration. The s*ame is true for hunting
and picnicking on the spoil areas. Where these uses are to be enjoyed
by more than a few people, grading of some areas near the water and for
roads, paths, and building and camping sites becomes necessary.

V. METHODS OF REVEGETATION
1. Forest Trees

Trees should be planted on spoil banks only after they have settled
and eroded sufficiently so that no excessive changes in topography are
to be anticipated. As trees are to be planted on the more acid spoils,
it is wise to permit sufficient time to pass until volunteer vegetation
shows up in many places. Hand labor is required for planting trees on
ungraded spoils. A mattock or planting bar is used to open the holes.
Fertilization is not practicable although it has been found that additions
of nitrogen have greatly stimulated the growth of pines on spoils. Trees
are usually planted 6, 7, or 8 feet apart in both directions. Special at-
tention is needed to see that the crews do not bend the roots and neglect
to tamp the ground around the roots. Since planting by hand is very
time-consuming, attempts at direct seeding of trees on spoils have been
made, but so far with no success.

The species of trees to be used on spoils depend on their adaptation
to the local climate and to the soil conditions. Although pine trees have
a definite place in the reforestation of spoil banks because of their low
nutrient requirements and their acid tolerance, a number of successful
plantings of hardwoods show that these should be used more extensively.
Black locust, Robinia pseudoacacia, L., sycamore, Platanus occidental,
L., cottonwood, Populus deltoides, Marsh, tulip poplar, Liriodendron

336 HELMUT KOHNKE

tulipifera, L., black walnut, Juglans nigra, L., and red oak, Quercus
borealis, var. maxima, Ashes are some of the species that show promise
in Indiana and neighboring states. Both in the case of pines and hard-
woods, seedling trees are cheaper, easier to plant and have a better
survival percentage than transplants.

For further detail on reforestation of spoils reference should be made
to Technical Paper No. 109 of the Central States Forest Experiment
Station (Limstrom, 1948).

H. Pastures

a. Unleveled Spoils. Seeding of forage crops on tinleveled spoils is
done either by hand with a whirlwind seeder or from an airplane; which
of these methods is more economical depends largely on the size of the
areas to be seeded.

Due to the nitrogen deficiency of the soil material, legumes should
be seeded previously to or simultaneously with grasses. Late winter is
generally the best time for legume seeding, and since the establishment
of legumes is the prime requirement for developing a pasture on un-
leveled spoils, the first seeding should be done at that time. Stiver (1949)
found that seeding in the middle of February gave the best stands of
legumes on Indiana spoils as compared to later seedings. The surface
crust of unleveled spoils without cover is dried out by March winds and
sunshine, and development of the young plants is impaired. He also
noticed that seeding legumes on spoils less than a year old gave better
stands than seeding on older spoils. The loose soil surface of fresh spoils
is the reason. Where no perfect catch has been obtained with the first
seeding it usually will be achieved in the third year when the few
plants that did grow have spread prolific amounts of seed and this seed
has germinated. Since seeding unleveled spoils is rather expensive it is
probably better to include grass seed with the legume seed and to sow
the mixture in late winter, than to make two separate seedings. The
species to be used should combine vigor in establishment with sufficient
palatability to be of value as pasture plants.

The following legumes and grasses have shown promise on unleveled
calcareous spoils in Illinois, Indiana, and Ohio:

White sweet clover, Melilotus alba, Desv.

Yellow sweet clover, Melilotus officinalis, (L) Lam.

Alfalfa, Medicago sativa, L.

Ladino clover, Trifolium repens, L.

Korean lespedeza, Lespedeza stipulacia, Maxim.

Birdsfoot trefoil, Lotus corniculatus, L.

THE RECLAMATION OF COAL MINE SPOILS 337

Kentucky blue grass, Poa pratensis, L.
Brome grass, Bromus inermis, Leyss.
Orchard grass, Dactylus glomerata, L.
Tall fescue, Festuca elatior, arundinacea Schreb.

Other species may be useful, especially in other regions. Where non-
calcareous spoil is to be used for pasture, more acid tolerant species
should be included but seldom much success can be expected. Since no
ground preparation is possible and a great variety of soil materials
exists, it is best to use a mixture of several, if not all, of the species
listed.

The rate of seeding depends largely on the time when the land is to
be ready for pasture. If pasturing is to begin 3 years after seeding,
rates similar to those used in ordinary field plantings arc adequate. If
the spoils are to be ready in the year after seeding, about double these
rates are required.

b. Graded Spoils. The establishment of pasture on graded spoils
presents problems other than those encountered where ungraded spoils
are used. The use of power driven equipment offers the possibility of
cooperating with nature to develop a productive soil. The first task is
to control erosion and to develop a good structure. A vigorous growth of
a grass and legume mixture is the best means to achieve both of these
ends. While only spoils with small amounts of stones are recommended
for grading for pasture establishment, there will still be difficulty in the
first years to cultivate such an area until some of the rocks have dis-
integrated. It may be difficult, therefore, to produce a conventional seed
bed. Under such conditions a good catch can be obtained by covering
the land with a thin layer of mulch. Manure, straw, chopped corn stover
or other similar material can be used. This cover also reduces the erosion
hazard.

Liming should be adequate to bring the surface soil to a pH of about
6.5. The fertilizer applied should be particularly high in phosphorus,
but should also include nitrogen and potassium. About 12 to 25 Ibs.
of nitrogen (N), 25 to 75 Ibs. of phosphate (P 2 O 5 ) and 25 to 50 Ibs. of
potash (K 2 0) per acre are recommended for initial application at the
time of seeding. As graded spoils are accessible to all agricultural im-
plements improvement treatments can be similar to those on ordinary
pasture land. The same species as those mentioned for ungraded spoil
pastures should be used. Other more acid tolerant plants can be in-
cluded on acid spoil material, even if it be limed. Whether the seeding

338 HELMUT KOHNKE

is done with a companion crop of small grain or alone is of little
importance.

Grazing should not begin until the second and preferably the third
year to permit the formation of a continuous sod. Fertilization at lower
rates should be repeated every second or third year and weeds should be
clipped whenever necessary.

In addition to establishing a dense vegetative cover soon after grading,
diversion ditches may be needed to control erosion. During the grading
operations the location of future fences should be smoothed out so that
the fences will not follow dips and hills. In all cases some of the ponds
resulting from the mining operations should be retained.

3. Rotation Crops and Orchards

One of the main considerations in using smoothed out spoils for rota-
tion crops is to create good soil tilth. A thick growth of grasses and
legumes for 5 to 10 years is the surest way to achieve this. After that
period any rotation suited for the given soil conditions, slope and climate
can be started. No experience of planting rotation crops on smoothed
spoils after such an improvement of several years exists. In Germany
(Meyer, 1940) and in England (Sisam and Whyte, 1944; Robinson, 1945;
Brown, 1949) the original soil is replaced on the graded spoils and there-
fore the establishment of rotation crops is much simpler.

Orchards can be established as soon as the erosion danger on the
freshly graded spoil has been controlled, provided the reaction of the
spoil is not too acid as it is practically impossible to lime the total po-
tential root zone of the fruit trees.

VI. GRADING

No other item in the reclamation of coal mine stoppings has been as
bitterly discussed as the -grading or "leveling" of the banks. It is obvious
that opinions clash on a subject where emotion, ethics and economics
are involved and where very few facts have been established by research.

1. Reasons for Grading Spoils

Grading spoil banks permits the use of power machinery and there-
fore widens greatly the land use capabilities. Ungraded strippings belong
to land use capability classes VII (Forestry) and VIII (Wildlife), ac-
cording to the Soil Conservation Service classification. These are the
two lowest groups. In many cases grading can bring the spoils into
groups VI or V (Permanent Pasture) or even III or II (Cultivated
Crops) .

THE RECLAMATION OF COAL MINE SPOILS 339

Grading can thus make it possible to grow more food and feed. Grad-
ing greatly speeds up the establishment of forage plants (Groves, 1949).
Grading permits the soil formed from the raw spoil materials to stay
in place and eventually to develop a deep and productive soil profile.
On ungraded spoil, with slopes of over 50 per cent, a deep profile cannot
form because erosion, even under dense cover, removes much of the
freshly formed soil. An example of this situation is shown in the dif-
ferences in pH and total nitrogen in the soils on the slope (60 per cent)
and the alluvium of a 20 year old pine plantation on spoils in West
Central Indiana (Fig. 4.)

Another reason for grading spoils frequently mentioned is the im-
provement of the scenery. Many people object to the ridge and valley
topography and prefer a smoother landscape. As soon as dense forest
covers the spoils, their surface configuration is largely hidden. Near
towns and cities grading is esthetically desirable, and may also furnish
valuable building sites.

2. How to Grade Spoils

In planning to grade spoils prime consideration should be given to
the eventual land use and the drainage system desired. As previously
stated, only land that is potential pasture, crop or orchard land need be
graded. Pasture land should have no slope steeper than 25 per cent in
order to permit safe operation of agricultural equipment, while for crop
land and orchard land the maximum slopes should be 10 per cent in
order to avoid serious erosion hazards. The slope pattern should be
designed in such a way that water does not flow far before reaching a
fairly level drainage channel. This can be accomplished frequently by
merely sloping down the ridges without completely eliminating them.
In some cases diversion ditches may be necessary. Where the spoil
material is of particularly porous nature, one can dispense with such
precautions.

A complete leveling of the land is of little value. It is costly, it com-
pacts the soil and it creates erosion hazards. Also, under most conditions,
grading to the original contour is not necessary. In very hilly territory
it is much better to smooth out the spoils more nearly level, allowing
for runoff on protected waterways than to attempt to regain the original
topography (Tyner et al, 1948), It is preferable to leave the "high-
wall" as a steep and valueless escarpment in order to gain some level
land that may be put to a higher form of use than could be practiced
before stripping.

Generally it is not desirable to fill in the larger and deeper ponds and
lakes that contain good water. They are a byproduct of strip mining

340 HELMUT KOHNKE

which can be regarded as an asset in areas deficient in natural open
water. On the other hand ponds containing water that is useless be-
cause it is too acid or too shallow or because it dries out during the
summer, should be filled in, if possible. The same is true of final cuts
that remain dry. It seldom will be practical to fill in final cuts com-
pletely since the overburden has all been deposited on one side, but by
grading down the last one or two ridges of spoil it is generally possible
to include the final cuts into the land use pattern.

One form of grading frequently employed is called "striking off the
ridges." A bulldozer or similar machine pushes the material of the
ridges aside so that a smooth surface of from 8 to 16 feet in width results.
This makes the area more accessible, gives the raw spoils a less ragged
skyline and simplifies tree planting to a certain extent. Striking off the
ridges does not permit operation of agricultural equipment since only a
small portion of the area is accessible.

As the available evidence points to the fact that leveling does not
help the growth of forest trees the only grading that should be under-
taken on potential forest land is to prepare access and fire lanes. Since
individual spoils are usually of restricted area not much consideration
need be given fire lanes. It is advantageous to locate access roads along
the ridges. They should be crowned so that water will not flow in the
lanes and cause severe gullying. Some lanes connecting the ridge lanes
with the main road should be provided. Where it is necessary to cut
across the ridges it is well to do this at relatively high points so that
water and erosional debris will not damage the lanes. In many cases
the original coal haulage road may serve as part of the system.

Frequently damage to soil structure has been observed as a result
of grading spoils. Probably the reason for this is that the lower strata
of spoils are almost continuously wet, because of the loose nature of the
spoils. Working heavy soils wet tends to puddle them. During the min-
ing process the overburden increases from a third to a half over its
original volume. Many large pores are created in the resulting over-
burden that stop the downward flow of all but the tension-free water.
In natural soils no such impediment exists. It is the obvious conclusion
that grading spoils, where desirable, should be done as soon as possible
after mining so that there will be no accumulation of percolating rain
water in them. The best method would be to do it as an integral part
of the deposition of the overburden in the banks. The rotating shovel
excavators used today in a few of the strip mines come closest to accom-
plishing this feat. The ordinary power shovels and draglines cannot do
it without special operations.

THE RECLAMATION OF COAL MINE SPOILS 341

3. Objections to Grading of Spoils

The main objections against grading spoils are the expense and the
possibility of impairing soil structure. Trees have been found to grow
slower on leveled spoils than on ungraded ones (Chapman, 1949; Sawyer,
1949). It is difficult to say how the same trees would have grown on
the same material, had the grading taken place immediately after mining
when the spoils had not had time to collect much water.

Another objection against grading is the increased erosion hazard.
If the smoothing out is done in such a way that water has a chance
to accumulate over larger sloping areas, grading can result in considerable
gullying, which makes successful revegetation difficult. The necessary
precautions have already been described.

VII. ECONOMIC ASPECTS

While ethics and esthetics play important roles in planning the
reclamation of coal mine strippings, an economically sound approach
must be the first consideration. Even under the most favorable condi-
tions spoil banks are "unimproved" land, and no quick returns from any
use can be expected. The choice exists between complete abandonment,
inexpensive improvement and intensive improvement. Where no im-
provement is made, it will be many years before desirable trees become
established and grow to merchantable size. Where calcareous spoils are
seeded to grass and legume mixtures grazing will bring returns within
very few years. While the returns per acre will be considerably less
than from improved pastures, the investment for seeding and fencing
will soon be amortized.

Planting trees on spoils has been costing around $25 per acre during
the last few years. This investment is much greater than the cost of
seeding grass and legumes but still is a rather modest figure. Occasion-
ally costs of establishing a forest cover have been a good deal higher
where trees had to be planted for a second or third time on account of
poor survival of the first plantings. Such extra expense can be avoided,
if the banks are checked for pH and planting is delayed until a scant
cover of pioneer plants is established. Obviously returns from the trees
begin only 10 or 12 years later when fence posts can be cut from black
locusts. The main income, however, will not materialize until pulpwood
or saw lumber can be harvested, usually after 30 to 40 years. As prac-
tically all tree plantings on coal mine stoppings in the United States are
less than 30 years old, little can be said about the value of the produce
and the expense of harvesting the trees. But the vigorous growth of the

342

HELMUT KOHNKE

adapted tree species on spoils and the constantly increasing demand for
lumber and pulpwood make it very probable that this form of land use
will bring a certain, if modest revenue commensurate with the expense
of planting.

Any form of spoil bank reclamation that includes grading is of
necessity quite expensive except where the grading is restricted to

For Complttt
Ltveling, 3.75%
of Totol Ovtrburden
is Moved Down

For 25% Slope,
2.5% of Total
Overburden is
Moved Down

50'

Natural Surface

50 feet

Fig. 8. Amount of earth movement needed for grading spoils.

striking off the ridges. In addition to the cost of grading are the expenses
for liming, fertilizing, seeding and fencing which may be considerable,
especially on the more acid spoils. The aggregate of these costs may
actually be higher than the purchase price of improved, high quality
farm land. Where only spoil material is graded which is adapted for
the intended land use there can be little doubt that an income is assured
that will be greatly in excess of that produced by forest plantations and
one that will start much earlier. Nevertheless in many cases the expense
of grading will make this form of reclamation an unprofitable under-

THE EECLAMATION 01 COAL MINE SPOILS 343

taking for the individual, if viewed as an independent business operation.

Grading may be regarded as part of the original earth-moving opera-
tion that is necessary to uncover the coal. Assuming an overburden
thickness of 50 feet, a volume increase of 50 per cent (personal communi-
cation L. L. Newman, U.S. Bureau of Mines) , spoil ridges 50 feet apart
and average slopes of ninety per cent, the complete leveling of the spoils
would require a second moving of less than 4 per cent of the total over-
burden material (Fig. 8). Moreover, this movement is of loose mate-
rial in generally downhill direction while the original material was solid
earth and rock and had to be lifted to considerable heights.

One very important item of the reconversion of spoils into improved
pasture land or crop land is that the adjacent land is benefited. Since
coal mine boundaries have irregular patterns, fields may be cut to small
sizes and sometimes made essentially inaccessible from the farm build-
ings. If the spoils are left ungraded and planted to trees they have
to be fenced out from pastured land. As a result much of the land
neighboring ungraded spoils is left idle. Grading increases the eventual
field size, permits fencing at more practical locations, and can help to
bring the entire original farm back into agricultural production.

Communities with strip coal lands suffer an eventual decrease in tax
income since usually the tax valuation of stripped coal lands is lower
than it was originally (Walter, 1949). Proper reclamation can minimize
or even eliminate this loss in tax revenue. The increased production from
agriculturally reclaimed coal land will also contribute to the economic
life of the area.

In the discussion of the economics of spoil bank reclamation a clear
distinction has to be made between the viewpoints of the individual and
of the country as a whole. Returns too small or too remote to pay for
the investment offer no incentive for the individual to reclaim spoils.
The nation, however, has to consider the potential productive power of
an area that may contribute materially to the furnishing of life's neces-
sities to the increased population of the future. Whether the individual
shall bear the full cost of restoring the productivity of the land is a mat-
ter of state or national policy. In view of the lower cost of producing
coal by the strip method compared to underground mining it seems pos-
sible that the reclamation cost be included in the price of coal. Only
the technical aspects of the reclamation of coal mine spoils are discussed
here. Walter (1949) has given a masterly presentation of the economic
problems facing agriculture as a result of strip mining of coal.

344 HELMUT KOHNKE

VIII. LEGISLATION

Soil destruction brought about by strip mining in the United States
is very minor compared with that due to erosion. Yet, while no laws
force a farmer to conserve his soil, legislation regulating reclamation of
coal mine spoils exists in 4 states: West Virginia, Pennsylvania, Indiana
and Ohio. In all cases revegetation of the stripped land is required, in
West Virginia and Ohio a certain amount of grading is also compulsory.
In all 4 states the legislation recognizes that different methods of reclama-
tion are appropriate for different types of spoil materials. The West
Virginia law also distinguishes between former arable and former forest
land. No grading is required in the latter case.

Illinois also had a reclamation law. It was declared unconstitutional,
however, on the ground that it was discriminatory legislation since it
regulated the reclamation of coal mine spoils while it ignored spoils of
other mining operations, such as gravel pits.

It is difficult to say what the most reasonable form of legislation
would be. One thing is certain, however, that it involves first of all the
establishment of a policy concerning the obligation of an individual
toward a piece of land to which he has full title. Since the United States
has passed the pioneer era of unlimited land resources it seems appro-
priate that restrictions be imposed on the destruction of the productive
capacity of land. If this philosophy were accepted, regulations concern-
ing the protection and reclamation of land would have to extend to all
land, and not only to mined areas.

Whatever the moral and constitutional background of coal spoil bank
legislation, it has to be based on well understood physical and economic
facts and to differentiate between the various conditions of spoils.

IX. DISCUSSION
1. Establishing a Reclamation Policy

It is impossible to view the coal mine strip bank problem as a sep-
arate entity and to arrive at an equitable solution. Reclamation of the
spoils is just a fragment of the general problem of conservation. As a
nation we have become conscious of the value of our natural resources.

Soil losses by water, wind and alkali have received widespread atten-
tion. The researches and practical experiences of the past 2 decades
have shown the tremendous possibilities of regaining and increasing the
productive capacity of abused soils by proper management. Great
strides have been made in halting the further inroads of excessive erosion
but the road ahead is long before the goal can be reached. The destruc-

THE RECLAMATION OF COAL MINE SPOILS 345

tion of the soil by strip mining is merely a faster and more spectacular
way of decreasing the food and feed producing area than erosion, excess
water, alkali or inadequate fertilization. We have established the prin-
ciple in this country that all land should be brought to its highest eco-
nomic use so that the United States as a nation can continue to enjoy
a satisfactory standard of living. Today there are about 3 acres of
arable land per inhabitant in the United States. Diminishing the crop-
producing area in the face of a rapid increase of population would even-
tually lead to disastrous results.

Smoothing out spoil banks to a more stable topography will permit
the soil that is formed to accumulate and eventually to become fertile
agricultural soil. It seems advisable, therefore, to grade down those
spoils that give promise of producing pasture or field crops. While the
material in the spoils varies greatly, most of it is of a nature that will
decompose rapidly and form soil. Besides, the majority of the coal strip
mines are in an area with excellent climate for plant growth. Half of
them lie in the Corn Belt and most of the others nearby.

The land involved in strip-mining operations represents only a small
fraction of the country and other lands deserve our watchful interest in
the same way. It has been interesting to see many old gravel pits graded
by bulldozers and put back into agricultural production. Such voluntary
reclamation work is the one best befitting a democratic nation. What-
ever the details of the eventual policy will be, everybody seems to agree
today that a temporary gain should not be paid for by the permanent
impairment of land resources. It is pertinent to inquire whether there
is a moral obligation on the part of the land owner to leave his land in
as good or better condition as when he took it over. No such viewpoint
prevailed in nineteenth century America. But viewpoints change and
today such an obligation is slowly developing. So far other miners have
not been required to smooth out their spoils and to improve and revegetate
them. The reason for this inconsistency may be found in the smaller
size of many of these undertakings, e.g., gravel and clay pits, and the
distance from centers of population of others, e.g., gold, iron, and copper
mines. Reclamation work on the spoil banks of phosphate mines in
Florida is under way.

In other more densely populated countries reclamation is obligatory. In
Europe, saving the original surface soil and spreading it over the leveled
spoils is either required or suggested. The attitude of German farmers
toward this problem is demonstrated by this quotation from Gray et al.
(1938) : "A German land owner, asked how he could justify an outlay
of about $100 an acre for reforesting stripped coal land, replied that he
considered it his duty to leave the property to his successors in at least

346 HELMUT KOHNKE

as good condition, as when he took it over. Whether or not the expendi-
ture would earn compound interest or show a profit did not enter into
his calculations. This epitomizes the prevailing attitude of enlightened
forest owners in much of central and northern Europe." Another example
is the reclamation of land from the North Sea which cost Holland $1250
per acre (Bear, 1949).

2. Future Research Needs

Much work has been done to find the best ways of reclaiming coal
mine spoil banks. The coal operators themselves have supported such
research with funds and facilities. The results of these studies together
with established facts of soil science, forestry, agronomy and ecology give
sufficient tools to reclaim the banks. Many improvements are still
needed. Future research may include some of the following items:
improved methods of removing the overburden and placing it in proper
sequence and in smoother surface on the spoils, developing better tests
and criteria to determine the potential land use capability of the spoil
material, finding the plant species best adapted for profitable growth on
the spoils, and study of the economic relations involved.

S. Contributions of Spoil Bank Research to Science

Since conditions on coal mine spoil banks are entirely different from
those on undisturbed land many interesting and valuable facts of soil
science, ecology, forestry and other fields present themselves for study.
One of the outstanding ones is the rate of soil formation. The moment
when the overburden is deposited in the spoil bank is zero time of soil
formation. Every year changes can be observed. Actual rates of soil
formation as well as criteria of soil formation can be discovered in sucli
research.

The steep slopes of soil material practically devoid of organic matter
give an opportunity of studying fundamentals of erosion. The high cal-
cium concentration of many spoils in spite of low pH values offers a
unique possibility of separating the effects of these two factors on plant
growth. Temperature effects on soil formation, erosion and plant growth
can be studied by comparing north and south facing slopes of the same
spoil bank. The differences between the effects of calcium ions and hy-
drogen ions on flocculation of clay and aggregation of soil can be inves-
tigated in the absence of organic matter. The many ponds give an
opportunity to determine the pH ranges of adaptability of various plants,
fish and plankton. Similarly interesting observations are possible on the
growth of various plant species in an unusual environment. It has been
said in connection with the ecology of the spoil banks that "you have

THE RECLAMATION OF COAL MINE SPOILS 347

to throw away the book," but more properly it seems that a new chapter
has to be added. Spoil banks offer a fertile and fascinating field of
research to the biologist.

X. SUMMARY

While the total area stripped for coal in the United States is only
about a quarter million acres, the problem of reclamation of the spoil
banks is important for various reasons. Each acre mined affects the
land use of at least one more acre of the neighboring land. The public
resentment against the despoilation of the land in the most fertile cli-
matic belt of the country calls for a clear understanding of the situation.
In the face of nationwide efforts to control the destruction of soil by
erosion, the destruction of soil by mining calls for the establishment of
a policy concerning the conservation of natural resources for future use.

The strip mining of coal leaves the overburden in steep ridges with
only little of the original top soil near the surface. The material is
extremely heterogeneous; from finest clay to large rocks, and from pH 2
to pH 8. Nitrogen and organic matter are present in only very small
amounts but otherwise the potential fertility is generally similar to that
of natural soils. The nonrocky part of the spoil is generally of desirable
texture.

Soil tests to determine the potential land use capabilities of spoils
have to be rapid because of the variability of the material. Hydrochloric
acid shows the presence of carbonates, potassium thiocyanate the pres-
ence of free acid and sodium azide the presence of sulfides. Ordinary
pH indicators and the interpretation of the occurrence of natural pioneer
vegetation as well as observation of the texture and rockiness are addi-
tional aids in planning land use.

On all spoils, except those of reactions below pH 3.5, natural vege-
tation is established in a few years. The calcareous areas are soon
covered with legumes which are followed by grasses. Practically all
native and many introduced plant species thrive on the spoil banks.

Plantings bring such areas more quickly into production and help to
camouflage the ragged topography. For more than 30 years trees have
been planted on spoil banks and many attempts at revegetation have been
successful.

The wildlife population has increased as a result of creating rather
inaccessible shelter and many ponds. These latter give an opportunity
to raise fish in regions deficient in natural lakes.

The land use capability of the spoils is dependent on the reaction, the
rockiness and the texture of the soil material as well as on our willing-
ness to grade down the ridges to suit a certain use. Ungraded spoils are

348 HELMUT KOHNKE

principally forest land and the calcareous and least rocky ones are po-
tential pasture land. Where grading is done pasture can be the dominant
use even though the reaction and fertility may not be very favorable,
and the areas with the best texture are suitable for cultivated crops and
orchards.

Trees are planted as one- or two-year old seedlings about 7 by 7 feet
apart. Legumes and grasses for pastures are seeded either by hand or
from an airplane. Once spoils are graded the possibility of using any
kind of agricultural implement permits a variety of methods of estab-
lishing vegetation. Grading spoils calls for extra precautions to avoid
erosion because of the lack of water-stable aggregates in the spoil mate-
rial and the resultant low infiltration capacity.

Grading of spoils raises the potential land use capability. The ma-
jority of graded spoils can be used for pasture and some of them after
a period of improvement for rotation crops. The greatest advantage of
grading is that the weathered soil can remain in place and eventually
develop considerable depth. Complete leveling is unnecessary and fre-
quently undesirable.

The main objection to grading is its high cost. The grading may cost
more than the purchase price of similarly productive natural land.
Legislation regulating the reclamation of coal mine spoil banks exists
in West Virginia, Pennsylvania, Indiana and Ohio. One of the most
essential features of such legislation is that it does not treat all spoils
alike but considers each case according to the physical and economic
situation. It seems that the development of an overall conservation
policy has to precede this type of legislation.

Further research on grading and revegetation of spoils is needed.

XI. ACKNOWLEDGEMENTS

The author takes pleasure in expressing his indebtedness to E. N.
Stiver, formerly graduate fellow, Purdue University, and now Assistant
Agronomist, Blackland Experiment Station, Texas, for the major con-
tribution to the spoil bank reclamation research which has provided
much of the background for this article. In the same way L. E. Sawyer,
Director of the Division of Forestry and Reclamation of the Indiana
Coal Producers Association has given valuable aid by sharing much of his
rich experience in this field. The author is also grateful for assistance
from his colleagues in practically every state concerned.

THE RECLAMATION OF COAL MINE SPOILS 349

REFERENCES

Adams, W. W. 1942. U.S. Dept. Int. Bull. 462, 4-5.

Adams, W. W., and Geyer, L. E. 1942. U. Dept. Agr. Bull. 462, 4-5.

Bear, F. E. 1949. Agron. J. 41, 497-507.

Brown, G. F. 1949. Soil Conserv. 15, 107-109.

Chapman, A. G. 1944. Central States Forest Expt. Sta. Tech. Paper 104, 1-25.

Chapman, A. G. 1947. Central States Forest Expt. Sta. Tech. Paper 108, 1-13.

Chapman, A. G. 1948. The Land 7, 41-45.

Chapman, A. G. 1949. Coal Mine Modernization Yr. Bk. 309-315.

Croxton, W. C. 1928. Ecology 9, 155-175.

Den Uyl, D. 1947. Proc. Indiana Acad. Sci. 56, 173.

Grandt, A. F. 1949. Proc. Natl. Coal Assn. Convention, to be published.

Gray, L. C., Bennett, J. B., Kraemer, E. and Sparhawk, W. N. 1938. Spoils and

Men. U.S. Dept. Agr. Yearbook of Agriculture, p. 133.
Groves, D. E. 1949. Ohio Farmer, Oct. 15, 14-15.
Holmes, L. A. 1944. Sci. Monthly 59, 414-420.

Limstrom, G. A. 1948. Central States Forest Expt. Sta. Tech. Paper 109, 1-79.
McDougall, W. B. 1925. Ecology 6, 372-380.
Maloney, M. M. 1941. Proc. Oklahoma Acad. Sci. 22, 123-129.
Meyer, L. 1940. Bodenkunde Pflanzenemdhr. 21-22, 707-722.
Riley, C. V. 1947. M. S. Thesis, Ohio State Univ., Columbus, Ohio.
Robinson, D. H. 1945. Agriculture (J. (Engl.) Ministry Agric.) 52, 178-180.
Sawyer, L. E. 1946. J. Forestry 44, 19-21.
Sawyer, L. E. 1949. J. Soil Water Conserv. 4, 161-165, 170.
Schavilje, J. P. 1941. J. Forestry 39, 714-719.
Sinks, A. H. 1946. Harper's Mag. pp. 432-438.
Sisam, J. W. B., and Whyte, R. O. 1944. Nature 154, 506-508.
Stiver, E. N. 1949. Ph.D. Dissertation, Purdue Univ., Lafayette, Ind.
Thornton, S. F., Conner, S. D., and Frazer, R. R. 1939. Purdue Univ. Agr. Expt.

Sta. Circ. 204.

Toenges, A. L. 1939. UJS. Dept. Interior, Bur. of Mines, R. I., 3440.
Tyner, E. H., and Smith, R. M. 1945. Soil Sci. Soc. Am. Proc. 10, 429-436.
Tyner, E. H., Smith, R. M., and Galpin, S. L. 1948. /. Am. Soc. Agron. 40, 313-

323.

Walter, G. H. 1949. Agr. Econ. Research 1, 24-29.
Winchell, J. H. 1948. Outdoor Indiana 15, 12-13.

Irrigated Pastures

WESLEY KELLER AND MAURICE L. PETERSON

U. *S. Department of Agriculture, Logan, Utah, and California Agricultural
Experiment Station, Davis, California

CONTENTS

Page

I. Introduction 351

II. Pasture Soils 352

III. Choosing Productive Mixtures 356

IV. Establishing Pastures 360

1. Preparation of Land for Irrigation 360

2. Seedbed Preparation and Seeding 362

3. Management of the New Stand 364

V. Management of Pastures 365

1. Grazing Management 365

2. Prevention of Bloat 367

3. Irrigation 369

4. Fertilization 371

5. Molybdenum Toxicity 373

6. Weed Control 374

VI. Economy of Pastures 375

1. Productivity 375

2. Economic Studies 376

3. Effect of Pastures on Crops that Follow 379

VII. Pastures in Relation to Other Sources of Feed 380

1. Relation to Other Forage Resources 380

2. Supplemental Feeding 381

References 382

I. INTRODUCTION

Irrigated pastures have provided feed for the livestock of the western
United States since its settlement. The acreage has expanded to an esti-
mated 2.7 million in the 17 western states, according to the 1940 Census
of Irrigation. Twelve per cent of the irrigated land of the west was in
pastures, 37 per cent in Nevada and nearly 25 per cent in Oregon. The
California acreage was estimated at about 560,000 in 1949, most of which
developed since 1930 and nearly half since 1946. Five acres planted in
the Werribee District of Victoria, Australia in 1914, was the beginning
of a development which reached approximately one-third million acres
by 1947 according to Morgan (1949).

351

352 WESLEY KELLER AND MAURICE L. PETERSON

Factors which have contributed to a large irrigated pasture acreage
are the development of new irrigated land, selection and use of species
which are highly productive on a wide variety of soils, and the need for
additional forages to supplement other sources, particularly rangeland.
The low labor requirement for this kind of irrigation farming has appealed
to many operators during recent years of scarce and costly labor.

Many of the Agricultural Experiment Stations of the West included
studies on irrigated pastures in their earliest projects. Recent research
has been directed both towards improvement of existing pastures and
development of new pastures on various types of soils including some
of the best.

The writers have placed greatest emphasis in this review on recent
experimental data but the lack of information on numerous points has
necessitated the use of some less well authenticated evidence.

II. PASTURE SOILS

The soils upon which irrigated pastures are grown vary widely in
physical and chemical characteristics. In recent years some of the most
fertile and productive soils in the West have been seeded to irrigated
pastures. A large part of the acreage, however, is on soils not well suited
for tillage because of poor drainage, excessive salts, shallowness, or the
presence of stones, steep slopes and other conditions unfavorable to culti-
vation. The many species of grasses and legumes used in pastures show
wide differences in soil adaptation and tolerance to adverse situations.

Characteristics of the soil which influence production and selection
of species are fertility, texture, depth, drainage, and salinity and alka-
linity. For a more thorough treatment of the management of irrigated
soils, the reader is referred to Thorne and Peterson's (1949) book on this
subject, and to the manual by Richards (1947) on the diagnosis and im-
provement of saline and alkaline soils. The present discussion is con-
fined to the adaptation of species to particular soil conditions.

Much of the irrigated pasture acreage is on soils which are typical
of arid conditions. Thorne (1948) characterizes these soils as being low
in organic matter and containing adequate or excessive quantities of
calcium, sodium, magnesium, potassium, carbonates and sulfates. He
further indicates that these soils, when placed under irrigation, often
contain insufficient phosphorus and nitrogen for maximum production.
With irrigation and growing of crops, organic matter is increased, mi-
crobial activity stimulated, and many mineral constituents are brought
into solution.

The irrigated land of the western United States is on valley bottoms

IRRIGATED PASTURES 353

and terraces which developed from material transported and deposited
by water. Most of the streams emerged from mountain canyons to
deposit sediment at floodtime in alluvial fans. Pastures are mostly on
heavy textured soils laid down under slow moving water or lacustrine
deposition. The latter soils are particularly heavy and may be quite
high in organic matter. Some of the older depositions have developed
profiles with heavy textured or cemented hardpan subsoils. Varying
degrees of salt accumulations are found.

Soil texture and depth are important factors in species adaptation.
Deep rooted plants such as alfalfa (Medicago saliva) are used on the
deep, coarse to medium textured soils. Hamilton et al. (1945) point out
that good- pastures are not readily established and maintained on very
sandy soils. The low water holding capacity of these soils and injury
to ladino clover (Trifolium repens latum) stolons by trampling limit the
use of this species. On fine textured soils, the shallow rooted species such
as ladino clover are quite satisfactory. Ladino clover and narrowleaf
birdsfoot trefoil (Lotus corniculatus tenuifolius) are two of the few species
which produce satisfactorily on the extremely heavy adobe soils in central
California. These same species are grown successfully on soils under-
lain with a claypan layer a few inches below the soil surface. This im-
pervious subsoil increases the efficiency of water use by preventing
downward percolation below the root zone.

Excessively wet soils in irrigated regions are caused by (1) direct
application of water, (2) seepage from canals and ditches, and (3) sub-
surface flows from areas receiving excessive precipitation or irrigation,
(Thorne and Peterson, 1949). Wet conditions arising from direct water
application are generally localized and result from over-irrigating, im-
proper leveling, failure to provide drainage, or to unsatisfactory balance
between the head of water and the size of the check or basin. These
conditions are normally avoidable and are discussed under later headings.

More extensive wet areas are found where drainage is difficult or not
practical such as in large valley bottoms and along natural waterways
which are subject to flooding or seepage. Because of wet conditions, and
often the presence of salts, these areas are difficult to manage and pro-
ductivity is low because the most productive species are not well adapted.

These valley bottoms or mountain meadows are used extensively for
spring and fall grazing and the production of wild hay for wintering range
cattle. Pittman and Bennett (1948) were able to double yields in the
second and third year after alsike clover (Trifolium hybridum) and red
clover (Trifolium pratense) were broadcast on undisturbed sod. An addi-
tional significant increase was obtained by plowing and seeding timothy
(Phleum pratense) and redtop (Agrostis alba) or bromegrass (Bromus

354 WESLEY KELLER AND MAURICE L. PETERSON

inermis) and meadow fescue (Festuca elatior) with the clovers. The
above treatment when combined with fertilization, frequent light irri-
gations and control of water to prevent prolonged flooding, resulted in
nearly 4 times the yield of native sod.

Reed canarygrass (Phalaris arundinacea) in combination with straw-
berry clover (Trifolium fragiferum) produced excellent pasture on land
too wet for alfalfa in unpublished studies conducted at the Utah Station.
In preliminary trials at the California Station, Reed canarygrass, peren-
nial ryegrass (Lolium perenne), tall fescue (Festuca elatior arundinacea),
narrowleaf trefoil, and strawberry clover ranked in the order named in
total production when grown under continuous flooding over a period of
several months.

Research is being initiated by federal agencies in cooperation with
several of the western states on the improvement of natural meadows.
This is an important source of livestock feed and the problem of improve-
ment warrants greater attention than it has received in the past.

According to Magistad and Christiansen (1944), "A large part of the
20,000,000 acres under irrigation in the 19 Western States contains enough
soluble salt to depress crop yields. A much smaller area contains so
much alkali that crop production is greatly curtailed and unprofitable.
Thousands of acres have been abandoned because of salinity." In arid
regions, salts accumulate chiefly because of irrigation and poor drainage
(Richards, 1947). Since poorly drained soils are not easily tilled, they
are used extensively for the production of forage. Thus, in irrigated
regions the problem of saline and alkali soils is one of great importance
in connection with forage production.

Richards (1947) has classified salted soils into saline, saline-alkali,
and nonsaline-alkali soils. The saline soils are defined as soil "for which
the conductivity of the saturation extract is greater than 4 millimhos
per cm. (at 25C.) and the exchangeable-sodium-percentage is less than
15. The pH of the saturated soil paste is usually less than 8.5." These
soils are characterized by white crusts on the surface or by streaks of
salt in the soil. They are reclaimed by leaching and drainage, after which
they become normal soils. The saline-alkali soils are defined as "soils
for which the conductivity of the saturation extract is greater than 4
millimhos per cm. (at 25C.) and the exchangeable-sodium-percentage
is greater than 15. The pH of the saturated soil paste may exceed 8.5."
The nonsaline-alkali soils are those "for which the exchangeable-sodium-
percentage is greater than 15 and the conductivity of the saturation ex-
tract is less than 4 millimhos per cm. (at 25C.). The pH values for
these soils generally range between 8.5 and 10." The latter two types

IRRIGATED PASTURES

355

of soil are more difficult to reclaim because of the low rate of water
penetration.

Hamilton et al. (1945) point out that the roots of salt-tolerant forage
plants increase the permeability of salty soils and speed up the rate at
which salt may be leached from them. According to Richards (1947)
alkaline soils require measures to improve the soil structure after suitable
base exchange and leaching has removed harmful amounts of sodium.
For this purpose they consider grass roots especially effective.

Bartels and Morgan (1944) consider the degree of reclamation of
salty soil to be proportional to the amount of water applied. They found
that when sufficient leaching had occurred to permit growth of barley

TABLE I

Salt Tolerance of Forage Crops According to Richards (1947). Tolerance Decreases
from Top to Bottom in Each Division. Scientific Names have been Added

GOOD SALT TOLERANCE

Alkali sacaton (Sporobolus airoides)
Salt grass (Distichlis spp.)
Nuttal alkali grass (Puccinellia nuttalliana)
Bermuda grass (Cynodon dactylon)
Rhodes grass (Chloris gay ana)

Rescue grass (Bromus catharticus)
Canada wild rye (Elymus canadensis)
Beardless wild rye (Elymus triticoides)
Western wheatgrass (Agropyron smithii)

MODERATE SALT TOLERANCE

White sweet clover (Melilotus alba)
Yellow sweet clover (Melilotus Officinalis)
Perennial ryegrass (Lolium perenne)
Mountain brome (Bromus carinatus)
Barley (hay) (Hordeum vulgar e)

Birdsfoot trefoil (Lotus corniculatus)
Strawberry clover (Trijolium fragiferum)
Dallis grass (Paspalum dilatatum)
Sudan grass (Sorghum vulgar e sudanense)
Hubam clover (Melilotus alba annua)
Alfalfa (California Common) (Medicago
saliva)

MODERATE SALT TOLERANCE
(Continued)

Tall fescue (Festuca elatior arundina-

cea)

Rye (hay) (Secale cereale)
Wheat (hay) (Triticum aestivum)
Oats (hay) (Avena saliva)
Orchardgrass (Dactylis glomerata)

Blue grama (Bouteloua gracilis)
Meadow fescue (Festiwa elatior)
Reed canary (Phalaris arundinacea)
Big trefoil (Lotus uliginosus)
Smooth brome (Bromus inermis)

Tall (meadow) oat (Arrhenatherum

elatius)

Cicer milk vetch (Astragalus cicer)
Sour clover (Melilotus indica)
Sickle milk vetch (Astragalus falcatus)

POOR SALT TOLERANCE

White (dutch) clover (Trijolium re-
pens)

Meadow foxtail (Alopecurus pratensis)
Alsike clover (Trijolium hybridum)
Red clover (Trijolium pratense)
Ladino clover (Trijolium repenslatum)
Burnet (Sanguisorba minor)

356 WESLEY KELLER AND MAURICE L. PETERSON

grass (Hordeum maritimum) , the land would support wimmera ryegrass
(Lolium rigidum). Morgan (1947) considers land leveling essential to
reclamation of salty land. Leveling makes possible the uniform applica-
tion of water, to leach salts downward. He reports that a field which
had a salt concentration in 1939 of 0.84 to 1.01 per cent in the 6- to 60-
inch zone was leveled, sown to pasture, and irrigated 12 to 19 times a year.
By 1946 the salt was reduced to 0.10 to 0.25 per cent. Light irrigations
were given the pasture, the annual average totaling approximately 2.5
acre feet. During the course of the study the productivity of the pasture
steadily increased.

Many native species possess marked tolerance of salty soils, but they
are almost without exception of relatively low forage value. Experience
gained through the years has been sufficient to permit a rough classifi-
cation of plants as to their salt tolerance. In recent years the work of
the U.S. Regional Salinity Laboratory at Riverside, California has
greatly expanded and refined our conception of the adaptation of plants
to salty soils. The salt tolerance of a number of forage species, as re-
ported by Richards (1947) is reproduced in Table I. The scientific names
have been added to aid in identification. The growth made by wheat,
oats or barley on salty land serves as a useful guide in choosing the best
species for seeding the area to pasture.

The principal effect of salt on crop production is a reduction in growth
of plants. Since the salts in the soil solution retard the movement of
water into the plants, it should be kept as diluted as possible by frequent
irrigation. The applications should be light if there is danger of raising
the water table, but heavier applications may be made if adequate drain-
age has been provided. Magistad (1945) and Hay ward and Wadleigh
(1949) have presented reviews of plant growth relations on saline and
alkali soils, and Hay ward and Magistad (1946) state the problem and
describe the w r ork of the Laboratory at Riverside.

III. CHOOSING PRODUCTIVE MIXTURES

Relatively few species are extensively used in irrigated pastures.
These are listed together with some of their characteristics in Table II.
Smooth bromegrass, reed canary grass, and tall oatgrass (Arrhenatherum
elatius) are used most widely in the cooler regions as contrasted to dallis
grass (Paspalum dilatatum) and rhodesgrass (Chloris gayana) which are
confined to the hot southern regions and areas of mild winter tempera-
tures. Similarly, bur clover (Medicago hispida) , a winter annual is used
only in areas with mild winters. Italian ryegrass (Lolium multiflorum)
and perennial ryegrass are most widely used in the Pacific Coast states.

IRRIGATED PASTURES

357

W
CQ
H

CO ^j

o "g

*^J

-D

S 5

^ CB ,i
* J- cc
QQ < C/J

d
d

n'3 'S '3 *3

d d a d

QJ CD (D O

PH P-i PH PH

Qi Cj

^.tfo"?
H H S

pq pq c/3 pq pq

<y cy

rt & cfl

^ 48 S

Q; 0? CS

13 ft 'g 1

3 .+j d d

2 2

CD o>
OH OH

Q^ D

fi fi

Bunch
Bunch
Bunch
Rhizo

Mode
Mode

Perenn

w S

d
^3

d
o

"J t3 T3
fc -22 .

13 -

d d

d 3

d fe .

<J OH

Nonspreading
Nonspreading
Nonspreading

^
fi X - 8 ^

*- 8

S o^Q
o'B 0^
S S g^ |

W W ^ 02

s-s

d a

z s

-*- d
S
_- ^ ,p

OH QQ PLi

enni

a-g-a
s s w

Stoloniferous
Nonspreading
Stoloniferous

f
e

358 WESLEY KELLER AND MAURICE L. PETERSON

Narrowleaf trefoil is used extensively in Oregon and California. Broad-
leaf trefoil (Lotus corniculatus arvensis) is coming into some use at
higher elevations where winter temperatures exclude the narrowleaf
trefoil. Strawberry clover has as yet achieved little importance except
on wet lands in widely scattered areas. The remaining species listed in
the table, show a wide range of adaptation throughout the western states.

Many writers have pointed out that if several species are used in a
mixture, the grazing animal has a more varied diet, stands may be im-
proved, and higher and more uniform yields are obtained. Reasons
given for using a number of species in a mixture are differences in sea-
sonal growth habits, depth of rooting, and soil nutrient requirements.
Studies at the Utah Experiment Station have at least partially supported
these statements but data are also available showing that simple com-
binations may be highly productive.

Pasture mixture studies are difficult to conduct because of the large
number of possible combinations which can be compared. Only 3 grasses
and 3 legumes give rise to 49 different mixtures containing one or more
grasses with one or more legumes. Eight grasses and 8 legumes provide
64 mixtures of a single grass with a single legume, 784 mixtures of 2
grasses with 2 legumes, 3,136 mixtures of 3 grasses with 3 legumes and
4,900 mixtures of 4 grasses with 4 legumes. There are a possible 65,025
different mixtures, using 1 to 8 grasses with 1 to 8 legumes, not including
differences in seeding rates. Most pasture mixture studies have included
selected species put in combinations considered of most value by the
experimenter.

Some early studies brought forth excellent recommendations although
they were not always carried over into agricultural practice. Sanborn
(1894) rated tall oatgrass, timothy and alfalfa in the order named, and
pointed out that Kentucky bluegrass (Poa pratensis) was relatively un-
productive as a pasture grass. French (1902) recommended 4 mixtures
for pasture, none of width contained Kentucky bluegrass although he
stated that it, and some other species, might be added. He pointed out
that a simple mixture of 10 pounds orchardgrass (Dactylis glomerata)
and 6 Ibs. red clover per acre was good for hay or pasture. He recog-
nized the difference between meadow fescue and tall meadow fescue,
characterizing the latter as a coarser, less desirable species. Welch
(1914) recommended a mixture of Kentucky bluegrass 8, orchardgrass 5,
smooth bromegrass 5, meadow fescue 4, timothy 4, and white clover
(Trifolium repens] 2 Ibs. per acre. It was a modification of a mixture
he had grown for 4 years with excellent results. Later, Welch (1917)
pointed out that orchardgrass and bromegrass were the more important
components, while Kentucky bluegrass, meadow fescue and timothy

IRRIGATED PASTURES 359

were of lesser importance. Hansen (1924) reported on a study of 3
mixtures, all quite similar except that one lacked legumes. On the basis
of both hay yields and grazing tests on these 3 mixtures he proposed a
fourth, which became widely known as the Huntley mixture. It is much
like Welch's (1914) mixture, differing only in seeding rate, the omission
of timothy and the inclusion of alsike clover. In a slightly modified
form the Huntley mixture became widely used in Utah and some adja-
cent areas under the name of Standard mixture No. 1.

Current recommendations of most experiment stations in western
United States omit Kentucky bluegrass from pasture mixtures. Com-
mon white clover has been largely replaced by ladino clover. Several
experiment stations include tall fescue in nearly all mixtures (Jones and
Brown, 1949; Klages et al., 1948; Rampton, 1945). Tall oatgrass is
recommended as a component of mixtures by Law et al. (1945) and
Keller et al (1947b).

Pasture mixtures for well-drained irrigated land have received in-
creased attention in Utah since 1943. Reports by Keller et al. (1945,
1947a, 1947b) and by Bateman et al. (1949) have shown that the modi-
fied Huntley mixture is a relatively low producer when utilized by dairy
cattle under rotation grazing. These studies have shown that high pro-
ducing mixtures were those dominated by smooth bromegrass, orchard-
grass, tall oatgrass, tall fescue or reed canarygrass, or combinations of
these, with 50 to 60 per cent alfalfa, red clover or ladino clover. In
contrast, mixtures dominated by Kentucky bluegrass, meadow fescue,
meadow foxtail (Alopecurus pratensis), or perennial ryegrass, with straw-
berry clover, alsike clover or any of several sources of ordinary white
clover, would be much less productive.

Because of the low palatability of tall fescue in these studies, and
the difficulty of obtaining good stands of reed canarygrass, these species
are not recommended in Utah for well-drained irrigated land. The
mixture currently recommended (Department of Agronomy, 1949) is
based on 6 years' study of 36 mixtures and 3 years' study of 32 mixtures.
It includes bromegrass 4, tall oatgrass 4, orchardgrass 3, wilt resistant
alfalfa 3, red clover 3, and ladino clover 2 Ibs per acre. In this mixture
tall oatgrass and red clover reach high production quickest, following
seeding. Bromegrass, orchardgrass, wilt resistant alfalfa and ladino
clover are the more permanent components of the mixture. They have
remained productive through 6 grazing seasons.

Many problems surrounding pasture mixtures need further investiga-
tion. Tall fescue is widely used in California, Oregon and other states.
However, Cunningham (1948) reports tall fescue is poisonous to cattle

360 WESLEY KELLEE AND MAURICE L. PETERSON

in New Zealand. There have been no reports of poisoning from areas
where tall fescue has been extensively used in the United States.

Almost no research has been carried out on seeding rates, A wealth
of experience indicates that under favorable conditions for both germina-
tion and establishment, 50 to 60 per cent legumes will result if the legumes
comprise one-fourth to one-third of the total weight of the seed. Size
and viability of seed and vigor of seedlings may considerably modify
these proportions. Excellent stands of ladino clover have been obtained
with one-half Ib. of seed, although 1 or 2 Ibs. are more commonly recom-
mended per acre. Mixture totals vary with the species used, but usually
range between 10 and 20 Ibs. per acre.

IV. ESTABLISHING PASTURES
1. Preparation of Land for Irrigation

The method used for irrigating pastures is determined by topography,
soil and subsoil texture and the amount of water available (Hamilton
et al., 1945; Jones and Brown, 1949) and in different regions is strongly
influenced by local custom (Stewart, 1945). Although numerous types
of irrigation systems are used, all may be grouped into either sprinkling
or flooding methods. Land leveling is required for most flooding systems.

The strip check or border method of flood irrigation is widely used
on relatively flat areas (Hamilton et al., 1945; Bartels and Morgan,
1944; Rayner, 1941; Jones and Brown, 1949). Land is graded to provide
0.2 to 0.5 foot fall per 100 feet, although steeper slopes are used in some
areas on soils which resist erosion. On land which has considerable side
fall, the width of the checks should be adjusted to keep elevation dif-
ferences between adjoining checks to 0.2 foot. Levees which guide the
water moving across the field are about 2 feet wide at the base and have
a settled height of about 6 inches. They are spaced at regular intervals
but these may vary in width from field to field.

Factors which influence width of levee spacing are soil texture, slope,
length of strips and rate of water delivery to each. The relationship
which exists between availability of water and the size of strip checks
for clay loam and clay soils is shown in Table III. For porous loam or
sandy-loam soils, the delivery rates should be increased from 2 to 5
times those indicated in the table or the size of checks correspondingly
decreased. Levees are often discontinued a few feet from the lower end
of the field and the excess water is carried away by a drainage ditch.
This avoids ponding of water and retards the encroachment of water
tolerant species. Advantages of the strip check method of irrigation
are low-labor requirements for irrigation and reasonably good control

IRRIGATED PASTURES 361

of water application. The cost of land preparation for irrigation may
be large because of the leveling which is required.

The contour check method is used on heavy soils where the land is
nearly flat or gently sloping. Levees are constructed on the contour
to form irregular shaped basins of varying sizes. The vertical interval
between levees is usually 0.2 foot, or less on very level land. Jones and
Brown (1949) recommend that contour levees have a base width of 30
to 36 inches and a settled height of at least 12 inches. Fields are irri-
gated from basis to basin starting at the upper side of the field. Drain-
age following irrigation is improved by construction of a broad shallow
ditch from the upper to the lower levee near the center of the basin.
The ditch also serves to carry water for irrigating each next lower basin
in the pasture.

TABLE III

Size of Strip Checks for Clay Loam and Clay Soils at Various Rales of Water

Delivery to each Strip a

Flow
to ei
cu.

delivered
ach strip
ft. sec.

Length of check

for various widths

of strip

10 ft. wide

15 ft. wide

20 ft. wide

25 ft. wide

0.2

440

0.3

660

440

0.4

880

660

440

0.5

880

660

440

0.6

1320

880

660

0.7

1320

880

660

0.8

1320

880

660

0.9

1320

880

660

1.0

1320

880

A Jones and Brown (1949).

Little land moving is required for contour check irrigation other
than to fill small depressions and remove high points with a land plane.
Therefore, the initial preparation costs are much less than for the strip
check method. Labor requirements are low although large heads of
water are required and it is difficult to control the amount applied. Some
modifications of this method are being tried in an effort to avoid diffi-
culties resulting from slow or improper drainage.

Wild flooding is used for irrigating pastures in the Sierra Nevada
foothills and throughout much of the Intermountain Region. Irrigation
ditches are built on grades of 1% to 2 inches per 100 feet and spaced at
intervals of 50 to 300 feet depending upon the steepness of the slope.
Water is distributed from the ditches at frequent intervals by raising the

362 WESLEY KELLER AND MAURICE L. PETERSON

level with a dam at the downstream edge of the section to be irrigated.
Little or no land preparation is required other than the construction
of irrigation ditches. The initial cost is therefore very low, but constant
attention and considerable skill is required by the irrigator if he is to
make efficient use of his water. Thus the cost of labor for irrigation is
large compared to other methods.

Little or no land preparation is required for sprinkler irrigation. A
sprinkler system may have an advantage on shallow soils, and especially
those underlaid with hardpan because the removal of surface soil may
be very detrimental. Veihmcyer (1948) states that other advantages
of sprinkling include effective use of a small flow of irrigation water,
uniform distribution, and case of adjusting water needs of different
soils in the same field. He lists disadvantages of sprinklers as their
high cost, more water lost by evaporation, and slow penetration of water
on some soils which results in runoff before an adequate amount is
applied. Various types of sprinklers are discussed in the abovementioned
publication.

2. Seedbed Preparation and Seeding

Methods used in the preparation of a seedbed for irrigated pastures
are similar to those used for all small seeded species. Hamilton et al.
(1945) list the requirements of a good seedbed as fine textured, firm,
moist, fertile, and free of weeds. Tillage operations to accomplish the
desired results usually involve plowing or disking, harrowing, and pack-
ing except when seeding in pea or grain stubble. A springtooth harrow is
sometimes used in place of the plow or disk on land relatively free of
weeds.

Jones and Brown (1949) in California recommend an irrigation
before seeding to settle the fills, firm the soil, and provide subsoil
moisture. The field is then planed if irregular settling has occurred, and
harrowed just before seeHing. Nitrogen fertilizers, if used, are applied
just before or at seeding time. Manure and phosphate fertilizers are
normally worked into the soil during the seedbed preparation. Irriga-
tion before seeding is seldom necessary in the Intermountain Region.

Pastures can be seeded at any time that favorable moisture condi-
tions and temperature can be maintained. In areas having cold winters,
6 to 10 weeks of growing weather are required for the young plants to
become winter hardy. Hamilton et al. (1945) point out that cool-season
grass seedlings attain winter hardiness at an earlier age than the asso-
ciated legumes. In the Intermountain Region of the Western United
States it has been customary to seed pastures in the spring, . on fall
plowed land. However, an increasing number of farmers are seeding in

IRRIGATED PASTUBES 363

August in grain stubble or on land from which canning peas have been
harvested. Post andTretsven (1939) and Hamilton et al. (1945) recom-
mend fall seeding if the land is not weedy, the grain has not shattered,
and adequate irrigation water can be applied. Bingham and Monson
(1946) consider grasshopper injury is avoided by late summer seeding
in grain stubble.

In areas having mild winters, Jones and Brown (1949) recommend
fall and early winter seeding. Matlock (1943) recommends seeding in
August and September for various sites in Arizona. Robertson et al.
(1948) recommend spring seeding but state that at high elevations where
snow cover is dependable in winter, good success has resulted from late
fall planting, germination occurring in early spring. This practice takes
full advantage of spring precipitation, and saves some irrigation water
and the labor of applying it. In California, the seeding time is adjusted
to take advantage of the natural winter rainfall. It is seldom possible
to establish a stand in dry weather by flood irrigation. However, a few
farmers have established successful stands of ladino clover in the Sacra-
mento Valley by airplane seeding in standing water. Ladino clover
germinates rapidly in water and the seedlings are established by the
time the field dries.

Drilling the seed is preferred if the seedbed is firm, but broadcasting
is satisfactory on loose soil. Double seeding in different directions in-
sures good broadcast distribution. Robertson et al. (1948) also recom-
mend double seeding in drilling. Companion crops, if used, are seeded
first. When planting a pasture in grain stubble the seed should be
drilled, otherwise it is not easily covered.

Airplane seeding of pastures is becoming increasingly important in
California on fields which are large enough to justify this method.
From 300 to 500 acres can be seeded in a day at a cost of $1.00 to $2.00
per acre for double seeding. This method costs slightly more than
ground broadcasting, but has the advantages of speed and the ability
to seed when the field is too wet for ground equipment. Winter seedings
in California are normally broadcast without covering.

Depth of seeding can be rather accurately controlled on a firm seed-
bed, but may be improved by the use of depth regulators on the drill.
Shallow seeding of not more than one-half inch is normally recommended.
Deeper seeding may be advisable on sandy soils or under conditions
where adequate surface moisture is uncertain. The legume fraction of
the mixture is usually seeded through the alfalfa hopper and the grass
fraction through the grain side of a drill. Southworth (1949) has recently
reported the use of rice hulls to improve the mechanical seeding quali-
ties of chaffy grass seeds. A drill set to seed 160 Ibs. barley per acre

364 WESLEY KELLER AND MAURICE L. PETERSON

will seed an acre of pasture grasses and legumes mixed with 2 bushels
(16 Ibs.) rice hulls. Many grasses which are otherwise impossible to
put through a drill, can be processed in a hammermill to remove awns
and appendages. Instructions on seed processing and some effects of
processing on germination of tall oatgrass are reported by Schwendiman
et al. (1940) and Schwendiman and Mullen (1944).

The cultipacker-type seeder developed at the Wisconsin College of
Agriculture and reported by Ahlgren and Graber (1940) and Ahlgren
(1945) is ideal for seeding pastures. It covers the seed lightly, and firms
the seedbed.

Irrigated pasture seedings are made either with or without com-
panion crops. Companion crops are almost never used in California.
Hamilton et al. (1945), Klages et al. (1948), and Davies and Christian
(1945) consider a companion crop desirable to prevent wind damage.
Bracken and Evans (1943) regard a companion crop as useful in helping
to establish pastures on land that crusts easily. Companion crops have
been used successfully in the establishment of experimental pastures in
Utah by Keller et al (1945, 1947b). Bateman (unpublished) at the Utah
Station has successfully established both pastures and alfalfa while
producing high yields of barley. He considers a companion crop worth-
while if in the Intermountain region proper management practices can
be followed during its growth. Barley is an ideal companion crop when
seeded at not over 50 to 60 Ibs. per acre. Further information on com-
panion crops is found under IV-3.

3. Management of the New Stand

New stands should be managed to promote rapid development of
the seedlings. Prolonged close grazing or grazing when wet are condi-
tions to be avoided. The stand should have a good top growth before
winter temperatures cause growth to cease. Frequent light irrigations
may be required until the roots become well developed. Davies and
Christian (1945) in Australia, and Levy (1945) in New Zealand report
satisfactory establishment of pastures under periodic heavy grazing
during the seeding year, whether with or without a companion crop.
Stands are grazed when the plants are 6 to 9 inches in height, usually
at 8 to 12 weeks after planting. This grazing is repeated whenever the
plants have made a 6 to 9 inch regrowth. Bartels (1947) points out that
heavy grazing of young pastures is sometimes necessary to prevent
perennial ryegrass from smothering out slower growing white clover.

Careful irrigation is required when a companion crop is seeded with
the pasture. In a study conducted at the Utah Station it was found more
profitable to harvest the companion crop for grain than to graze it. In

IRRIGATED PASTURES 365

an unpublished report, Bateman showed an advantage of 1,424 Ibs. total
digestible nutrients (T.D.N.) per acre plus 3,644 Ibs. of straw for this
method compared to grazing. During the seeding year (1946) the com-
panion crop harvested for grain yielded 2,952 Ibs. T.D.N. plus 3,644 Ibs.
straw per acre, while the grazed companion crop yielded 1,078 Ibs. T.D.N.
per acre, or a difference of 1,874 Ibs. T.D.N. plus the straw. In the
following year (1947) plots periodically grazed in 1946 yielded 450 Ibs.
more T.D.N. per acre than those taken through to grain. In 1948 and
later, yields from the 2 treatments have not differed. Excellent stands
were obtained under both treatments. The study was part of a pasture
experiment containing 32 different mixtures (Keller et al., 1947b). Bate-
man et al. (1949) point out that in 1947 this pasture yielded 5,342 Ibs.
T.D.N. per acre, or the equivalent of 5.31 tons alfalfa hay.

V. MANAGEMENT OF PASTURES
1. Grazing Management

The objectives of grazing management are (1) to maintain the de-
sired balance between species, (2) to obtain continuous high production,
and (3) to obtain utilization of the forage when it is most nutritious.
Some forage species will tolerate close or continuous grazing while others
will not. Most of the pasture mixtures now being recommended be-
cause of their high production of nutritious forage consist of species
that require periods for regrowth, provided by rotation grazing, and will
not survive if continuously closely grazed. With rotation grazing, two
and preferably three or more pastures are grazed in rotation. After
grazing, each pasture is irrigated and allowed to recover. The animals
return to the first pasture from 3 to 6 or 8 times in one season. The
system is highly flexible, and can be adjusted to fit into the other opera-
tions and requirements of each farm. Important considerations in de-
veloping a rotation grazing system are the number of subdivisions in the
pasture, the number of days grazing in each, and the interval between
grazings. Maximum yield of milk or beef will result only when the
proper regrowth interval is used. If the interval is too short (the herbage
too young) vigor of the plants will decline, and the grazing animals will
expend an unduly large amount of energy in grazing. If the interval
is too long the herbage will have lost palatability, and probably nitrogen
also, and the grazing animals will not clean it up eagerly. Likewise,
if the grazing period is too long in each subdivision the least desirable
components of the pasture will be left until last. California dairymen
have observed fluctuation in milk flow when the grazing period was as
short as 5 days.

366 WESLEY KELLER AND MAURICE L. PETERSON

Rotation grazing was advocated many years ago by Harris (1913)
and Welch (1914, 1917) and is widely used for irrigated pastures (Ham-
ilton et al, 1945; Starke, 1947; Bartels, 1944a; Semple and Hein, 1944),
even though there are no experimental data under irrigation to indicate
its value. The work of Hodgson et al. (1934) reporting a gain of 8.82 per
cent from rotation over continuous grazing, has been referred to by Ham-
ilton et al. (1945) and by Bracken and Evans (1943). According to
Semple et al. (1934) rotation grazing increased production 10 per cent
at Beltsville, Md., while studies in Missouri, Virginia, and South Dakota
have given like results. Apparently these studies were conducted on
pastures containing species that are tolerant of close grazing. Levy
(1949) reports that close continuous grazing reduces production of New
Zealand pastures by 50 per cent, and permits entry of weeds and un-
desirable grasses.

When a pasture is rather heavily stocked for a short period, the
forage can be consumed when it is most nutritious, and fuller utilization
and less selective grazing results, 't Hart (1949) reports that in Holland
dairy cows grazing continuously achieved a utilization of 50 to 75 per
cent of the forage. If the grazing period was reduced to 5 to 10 days,
utilization was increased to 60 to 80 per cent but by reducing the graz-
ing period to 1 to 2 days utilization was increased an additional 10 to
20 per cent. These data suggest a trend toward heavier stocking rates,
for shorter periods under rotation grazing. The advantage of many
subdivisions in a pasture, with short grazing periods is strikingly il-
lustrated by data from the Blaettler Dairy in Santa Clara County,
California, reported by Assistant Farm Advisor M. S. Beckley. In 1948
a 52-acre pasture was used in 3 subdivisions and grazed by 92 cows. Graz-
ing 10 days in each subdivision they obtained forage with a feed replace-
ment value of 2.5 tons alfalfa hay per acre. In 1949 the 52 acres was
divided into 30 subdivisions that were grazed one day each by 110
cows. In 1948, milk production was uniformly maintained only 4 days
out of the 10-day grazing period in each subdivision, with an average
loss of 3 cans of milk per day for the last 6 days of each grazing period.
In 1949 continuous high production was maintained. Feed replacement
in 1949 was 5.5 tons alfalfa hay per acre.

Investigations at Werribee (Australia, 1947) have shown that sheep
made similar gains when on rotations of 10 to 30 days. The more fre-
quent grazings of a field did not alter the grass-legume ratio, but at
10-day intervals orchardgrass thinned out. Bartels (1944a) found that
a 3-weeks J rest period between grazings enabled all seeded species to
remain in the mixture; this is in agreement with observations in Utah.

In South Africa, Starke (1947) lists the following 5 reasons for ro-

IRRIGATED PASTURES 367

tation grazing of sheep: (1) less selective grazing, (2) less fouling of
the forage, (3) more regular irrigation, (4) less internal parasite infection,
and (5) better quality and more palatable forage. He used pastures of ap-
proximately 4.25 acres for 150 to 200 sheep.

According to Hamilton et al. (1945) and Klages et al. (1948), a pas-
ture is considered ready to graze when about 6 inches of growth has
occurred if tall species are used, and when 3 to 4 inches high if low-grow-
ing species predominate. Schoth (1944) considers ladino clover ready to
graze at 3 to 4 inches, but recommendations of California investigators
include not grazing ladino clover closer than 3 to 4 inches in order to
permit rapid recovery.

It is generally considered good grazing management to allow the pas-
ture to go into the winter with at least 3 to 4 inches growth. Schoth
(1944) stresses the importance of avoiding close fall grazing of ladino
clover, and points out that the fleshy stolons are damaged if ladino is
pastured when the ground is frozen or wet.

Selective grazing cannot be avoided entirely, but it will be reduced
to a minimum under rotation grazing if the various species in the mixture
are of approximately equal palatability. Mowing the pasture occasion-
ally, with the cutter bar raised to about 3 inches, does much to keep
the pasture fresh and the forage palatable. Bartels (1944a) found clip-
ping especially worthwhile if the pasture contained orchardgrass or
Paspalum.

Continuous close grazing is still common in the Intermountain Region
of the United States. Under intensive use, it leads in a few years to
pastures consisting of Kentucky bluegrass and white clover. In experi-
mental plots Keller et al. (1947a) and Bateman et al. (1949) found this
combination a consistently low producer. The clover will be reduced or
even eliminated, with further reduction in yield, if use is heavy, fertility
low, and irrigation applications erratic.

2. Prevention of Bloat

The prevention of bloat has been a major problem in the management
of irrigated pastures in some areas. Although surveys show the actual
percentage loss from bloat to be small, individual stockmen have had
catastrophic losses (Cole et al., 1945). Practical experience has shown
that there is little likelihood of bloat on pastures containing 40 to 50
per cent or more grass. It is difficult, however, to maintain a proper
proportion of grasses and legumes at all seasons. The percentage of
legumes in pasture can be reduced by applying nitrogen fertilizer or
barnyard manure and withholding phosphate fertilizer (Klages et al.,
1948).

368 WESLEY KELLER AND MAURICE L. PETERSON

Another practical solution suggested by Cole et al. (1945) is to pas-
ture legumes only after they reach the early bloom stage although these
workers admit that no well-controlled experiments have been conducted
relating to this factor. Bartels (1944a) supports this idea in suggesting
3 weeks for recovery in rotation grazing which gives the forage enough
maturity to reduce bloat. This procedure may be expected to be more
effective in pasturing a legume like alfalfa than with ladino clover which
has an indeterminant habit of growth and low fiber content. Schoth
(1944) recommends continuous grazing except for animals that bloat
easily, which should be removed. Cole et al. (1945), however, discount
the value of continuous day and night pasture and grain feeding as a
means of preventing bloat. The feeding of minerals and the pasturing of
legumes only when free of dew or rain also appear to lack supporting
evidence.

The feeding of dry hay or straw before pasturing legumes is advo-
cated by Robertson et al. (1948). This method was tested experimentally
by Cole et al. (1943) who found that overnight feeding of alfalfa hay
did not always prevent bloat although coarse-stemmed hay was more
effective than fine stemmed hay. Cole and Kleiber (1945) found over-
night feeding of 17 Ibs. of sudan hay (Sorghum vulgare Sudanense) com-
pletely effective, while 4 to 7 Ibs. fed 2 hours preceding pasturing was
ineffective in preventing bloat. Mead et al. (1944) found that 5 Ibs. of
barley straw was not sufficient to prevent bloat.

Overnight pasturing of sudan was effective in preventing bloat on
alfalfa the following day in studies by Cole et al. (1943). Advantages
of sudan pasture suggested by these authors were high palatability and
a growth habit permitting rapid ingestion. It might also have been
added that this procedure is less costly than hay feeding. Starke (1947)
states that dry sheep and pregnant ewes are less likely than lactating
ewes to bloat on alfalfa.

According to Professor Glen Staten (unpublished data) alfalfa-grass
mixtures are less productive in Southern New Mexico than alfalfa alone,
while over a considerable area water economy prohibits use of ladino
clover. Here, pasturing alfalfa is very common. Danger from bloat
is somewhat reduced by grazing only relatively mature plants, but this
results in considerable waste from trampling and fouling of the forage.
Staten reports that some large operators are now obtaining full forage
utilization, and have eliminated bloat, by harvesting each day's require-
ments and feeding as green chopped material.

Birdsfoot trefoil apparently does not cause bloat in either cattle
or sheep. For this reason, some farmers are using this legume in pref-
erence to ladino clover or alfalfa.

IRRIGATED PASTURES 369

S. Irrigation

For rapid growth and high production the soil occupied by the roots
of pasture plants should have readily available moisture at all times.
Depth of rooting determines the volume of soil available for supplying
water to the plant. Although pasture plants vary in depth of rooting,
the depth of soil is often the principal factor limiting root penetration.

The capacity of different soils to hold water varies greatly. The
amount of readily available water (field capacity to permanent wilting
percentage) held by a group of California soils ranged from 0.67 to 2.66
inches per foot depth of soil. This range, which is approximately 4-fold,
emphasizes one of the reasons for the wide differences in amount and
frequency of irrigation required.

Kramer (1949) discusses factors affecting the absorption of water
and points out that poor aeration may cause a considerable reduction
in absorption of water by plants and that the accumulation of carbon
dioxide may be a more important factor in reduced water absorption
than lack of oxygen. Poorly aerated conditions are common on many
of the heavy clay soils used for irrigated pastures.

The total amount of irrigation water used during the year will depend
upon the length of the growing season, natural rainfall, temperature,
frequency and depth of wetting, and the species involved. Frequent
light irrigation increases the percentage loss from evaporation. Too
heavy irrigation on permeable soils will cause water to penetrate below
the root zone and be lost. However, uniformity of penetration is difficult
to attain.

Water used by transpiration and evaporation of 3 pasture crops under
the climatic conditions at Davis, California are shown in Table IV. In
1943, Sullivan and Winright obtained records of acre-inches of water
used per acre on 7 farm pastures in the Imperial Valley which indicated
a range from 48 to 81 acre-inches with an average of 68.2 inches. Sim-
ilar studies in San Bernardino County, also in Southern California, by
Shultis and Campbell in 1943 showed a range from 36 to 80 acre-inches
per acre, the average being 65. Jones and Brown (1949) state that 33
to 36 acre-inches of water were used on shallow clay loam soils applied
in 12 irrigations over a 6-months' period in the Sierra foothills of Nevada
County in Northern California. Robertson et al. (1948) recommend
2 to 4 acre-inches of water per irrigation at 10- to 14-day intervals during
June through September for most areas of Colorado. Total water re-
quirements for the season were 2% to 3 acre feet.

The frequency of irrigation required depends upon how soon after
irrigation the soil moisture within the root zone will again be reduced

370 WESLEY KELLER AND MAURICE L. PETERSON

TABLE IV

Water Used (Acre-Inches per Acre by Months) by Transpiration and Evaporation
of Three Pasture Crops Grown at Davis, California R

Crop Mar. Apr. May June July Aug. Sept. Oct. Nov. Total

Alfalfa

1.5

3.9

5.3

6.5

8.0

6.6

5.0

2.0

40.0

Ladino clover

1.8

4.5

6.0

7.5

9.0

7.5

5.5

2.2

44.0

Sudangrass

2.0

3.2

4.0

3.4

2.2

17.0

* Data from Irrigation Division, California Agricultural Experiment Station.

to the permanent wilting percentage. The water holding capacity of the
soil, temperature, and the crop influence the rate of water extraction.
Many of the pastures in the interior valleys of California which con-
tain ladino clover are irrigated at approximately 10-day intervals. On
soils which are porous, irrigation may be required as frequently as every
7 days during the heat of summer. Pastures composed primarily of
deeper rooted species, such as alfalfa and birdsfoot trefoil are irrigated
at 2- to 4-week intervals depending upon the soil and temperature
conditions.

Bartels et al. (1932) carried out extensive studies on frequency and
amounts of irrigation water required for pastures in Victoria, Australia.
The soil was described as a shallow clay loam which was slow to absorb
water. Production of the pasture under prevailing conditions was best
when 24 inches were applied in 6 irrigations of 4 inches each. Equal total
amounts applied in either 4 or 8 irrigations were inferior.

The rate at which water is applied during irrigation has much influ-
ence on penetration, Doneen (1948) has pointed out that 50 per cent
or more of the total water applied may be lost through deep percolation.
Loss is greatest when the head of water is insufficient to reach the end of
the check in a relatively short time. Deep percolation takes place near
the head ditch. He suggests a large head be used to force the water
through to the end of the check after which the head may be reduced
to maintain an even flow to wet the length of the check until the desired
depth of penetration has been reached.

The relationship of water delivery rate to size of the check is shown
in Table III. The large delivery rates are desirable for the contour
check method of irrigation because individual basins may contain up
to 4% or 5 acres (Jones and Brown, 1949). These require rapid flooding
to obtain uniform penetration.

Sprinkling systems often can be advantageously used where only a
small head of water is available. Frequent, light irrigations are also

IRRIGATED PASTURES 371

possible on porous soils which normally would require rather large
amounts of water if flood irrigated. Veihmeyer (1948) has pointed out
that while sprinkling can save water loss from deep penetration, more
water will be lost by evaporation.

Problems of drainage should be considered along with irrigation.
Poor drainage inhibits growth of desirable pasture plants and encourages
water tolerant weedy species. Poorly drained pastures are so slow to
dry out after irrigation that the likelihood of grazing while wet is in-
creased. Salts may accumulate under poor drainage conditions in some
of the arid regions. Careful preparation of the land for irrigation is
the best method of avoiding improper drainage. Some of the adobe
soils of lacustrine origin in the Sacramento Valley present a drainage
problem because the soil is very heavy and extremely flat so that down-
ward percolation is almost nonexistent. If good drainage is not possible,
the choice of species for the pasture should include only those capable
of growing on wet land. Bartels et al. (1932) have pointed out the im-
portance of the method of irrigation upon the balance of species in the
pasture. Ryegrass and white clover were more favored by frequent
heavy waterings than cocksfoot and subterranean clover.

Ladino clover, alfalfa and other legumes may "scald" if irrigated
during extremely hot weather. The difficulty usually occurs on imper-
vious soils on still days when temperatures exceed 100F. and water
stands for several hours. This difficulty is avoided by providing good
drainage and by irrigating at night during very hot weather.

4. Fertilization

In the United States, fertilizers have been used less extensively on
irrigated pastures than on most cultivated crops. In recent years, how-
ever, experimental studies have shown large increases in yield are ob-
tained from the fertilization of pastures under many conditions.
According to Hamilton et al. (1945) the average farm in the irrigated
west produces enough barnyard manure to provide 8 to 10 tons per
acre every 3 years. Robertson et al. (1948) in Colorado recommend top
dressing with manure in February or March to provide 2 weeks more
of early spring grazing. Boyd (1945) in Wyoming suggests using 8
to 10 loads of manure every few years, applied in fall or winter. Klages
et al. (1948) consider top dressing with manure the best fertilization for
pastures, supplemented with phosphorus (40 to 50 Ibs. P 2 5 per acre)
in southern Idaho and with sulfur (100 Ibs. gypsum per acre) in parts
of northern Idaho.

It is generally recognized that manure stimulates the growth of grasses
more than legumes (Bateman, 1940; Hamilton et al., 1945; Klages et al.,

372 WESLEY KELLER AND MAURICE L. PETERSON

1948; Schoth, 1944). To maintain a vigorous growth of legumes in pas-
tures that are manured, phosphate fertilizer should also be applied.
Schoth (1944) recommends 8 to 10 tons manure per acre, supplemented
with 300 Ibs. superphosphate or the equivalent. Bateman (1940) recom-
mends that farmers in Utah apply 10 to 15 tons manure with 200 Ibs.
"treble" superphosphate (43 per cent PsOg) to their pastures every 3
years. He reports that Pittman found that one application of 600 Ibs.
of treble superphosphate increased yields 63.6 per cent over a 5-year
period, the gain the first season amounting to 211 per cent. He found
that phosphorous fertilizer increased the nitrogen content of the herbage,
by increasing the per cent legumes in the pasture. One application of
treble superphosphate at 200 Ibs. per acre resulted in a 22.7 per cent
increase in the phosphorus content of the forage over a 3-year period.
In another study Bateman (1943) found that one application of 6.8 tons
manure with 200 Ibs. treble superphosphate, applied to the pasture plots
between March 27 and April 6, increased the first grazing 27 per cent,
the second grazing 108 per cent, the third grazing 117 per cent, the fourth
grazing 102 per cent, and a small fifth grazing 187 per cent, the year's
gain amounting to an increase of 95.7 per cent. Ewalt and Jones (1939)
report a 75 per cent increase in production over a 5-year period from
annual applications of 300 Ibs. superphosphate (16 per cent P2O r> ). This
treatment did not alter the chemical composition of ladino clover in
the pasture.

Pittman and Bennett (1948) in a study of irrigated meadow pastures
in ranching areas of northern Utah, report a 50 per cent increase in
production from either 10 tons manure or 200 Ibs. ammonium sulphate
per acre. One application of manure stimulated growth through 4 years.
Applications of 200 Ibs. treble superphosphate per acre were profitable
where legumes were abundant. Liquid manure containing some solids,
urine, and barn washings is especially valuable pasture fertilizer (Ham-
ilton et al.j 1945). Yearly applications equivalent to 10 tons of barn-
yard manure per acre may be used in light applications after each
grazing. Fertilization costs may be much reduced if the manure can
be applied with the irrigation water.

The rapid development of irrigated pastures in Victoria, Australia
is attributed by Bartels and Morgan (1944) to the universal response
of these pastures to phosphate fertilizer. At Werribee, Morgan (1949)
reports profitable responses from applications of superphosphate (22
per cent P20s) at 400 to 500 Ibs. per acre per year if moisture is not
limiting, and under similar conditions Rayner (1941) recommends 400
Ibs. Annual applications of 200 Ibs. superphosphate increased production
77 per cent with 6 irrigations totaling 2 feet of water per year (Australia,

IRRIGATED PASTURES 373

1947). Additions of 100 Ibs. ammonium sulphate, nitrate of soda, or
potassium sulfate, or 3000 Ibs. lime or gypsum, failed to increase yields
further, and in some instances lowered them. Rock phosphate was found
to be highly inefficient. In northern Victoria Morgan and Rayner (1941)
recommended annual applications of 300 Ibs. superphosphate. Accord-
ing to Bartels (1944a) many farmers apply as much as 500 Ibs. super-
phosphate per acre to their pastures each year. Applying phosphorus
at one time (in the fall) was as effective as several fractional applications
during the year (Australia, 1947; Morgan, 1949; Rayner, 1941). Rayner

(1947) found that liberal top dressings of superphosphate not only in-
creased yield but fostered development of the more valuable species to
the exclusion of weeds and other less desirable forage.

In a few instances pastures have responded to fertilizers other than
manure and phosphorus. According to Schoth (1944) in Oregon, and
Andrew (1947) in Victoria, potash has been profitably applied on some
soils. Both Schoth (1944) and Klages et al. (1948) report beneficial re-
sponses from sulfur on some soils of Oregon and northern Idaho. Al-
though the response to manure and phosphorus is widespread, use of
other fertilizers should be based on prior tests conducted on the land.

5. Molybdenum Toxicity

Britton and Goss (1946) reported an ailment in cattle and sheep
which had been prevalent for many years along the southwest edge of
San Joaquin Valley in central California. Ingestion of green feed con-
taining abnormally high quantities of molybdenum was found to be the
cause. The symptoms in cattle are excessive scouring, loss of weight
and roughening and gradual fading of the hair. Loss of hair and even-
tual death of the animal may result. Ferguson et al. (1943) and Lewis
(1943a, 1943b) have reported a similar condition in England. Barshad

(1948) has shown that, in general, abnormality in cattle occurred when
a large portion of the pasture plants contained 20 or more parts per
million of molybdenum but no difficulty was observed if less than 10
p.p.m. was found. Analysis of soils in the affected area of California
indicated the molybdenum content was only slightly higher than normal
soils although the solubility was relatively high. The alkalinity of these
soils was responsible in part for the high solubility.

Legumes growing on affected soils absorb greater quantities of
molybdenum than grasses. Lotus corniculatus and Trifolium repens la-
turn contained as much as 150 p.p.m. dry matter compared with 33
p.p.m. in Lolium perenne and 9 p.p.m. in Paspalum dilatatum. The total
molybdenum content of soils ranged from 5 to 10 p.p.m. where these

374 WESLEY KELLER AND MAURICE L. PETERSON

samples were taken, but Barshad (1948) points out that cattle may be
affected on soils containing as little as 1.5 p.p.m.

Grazing and livestock management systems for control of molyb-
denum poisoning are as yet based upon limited evidence and only par-
tially published. Barshad (1948) states that feeding of dry roughage
tended to reduce scouring. Rotation of livestock between badly affected
and less badly affected areas has been suggested. Emrick (1948) recom-
mends a special pasture mixture from which legumes are eliminated for
soils that may contain excessive amounts of molybdenum.

Use of copper in the form of copper sulfate in the drinking water has
been successful in overcoming molybdenum poisoning in some instances
according to an unpublished report of the California Agricultural Ex-
periment Station. Acidifying the soil with sulfur or other acidifying
materials has been suggested as a method of reducing the availability
of molybdenum. The use of nitrate fertilizer has reduced the uptake of
molybdenum and its value on a field scale should be determined.

Even though very low concentrations of molybdenum in the soil
may be taken up by plants in concentrations that become toxic, the
element is essential to plant growth, and particularly to thrifty develop-
ment and nodulation of legumes. Anderson (1946) has shown that some
soils of south Australia are so deficient in molybdenum that applications
of 1 or 2 ounces per acre were associated with very large increases in
forage, particularly on land not deficient in phosphorus.

6. Weed Control

Annual weeds are seldom, if ever, a serious problem in well-managed
irrigated pastures, particularly if the initial establishment is good. Oc-
casionally pastures become infested with perennial thistles or other un-
palatable perennial weeds. It is not advisable to seed pastures on land
known to be badly infested with weed seeds. Management practices
which help to keep weeds under control are fertilization, careful irrigation,
and mowing. Scattered weeds can be removed with a sharp shovel. If
the infestation is general, resulting in greatly reduced productivity of
the pasture, it is advisable to plow and return the land to row crops or
fallow until the weeds are controlled.

According to Hamilton et al. (1945) many perennial weeds that com-
monly occur in pastures will be eaten by cattle if mowed about 2 days
before the end of a grazing period. In the Intcrmountain Region -irri-
gated pastures often become infested with dandelion (Taraxacum of-
ficinale). Dandelion can be controlled in pastures by spraying with
2,4-D, without permanent damage to legumes, but whether this is a prac-

IRRIGATED PASTURES 375

tical measure is not now known. Spraying has been used with some
success on curly dock (Rumex crispus) in California.

VI. ECONOMY OF PASTURES
1. Productivity

The carrying capacity and productivity of irrigated pastures have
been measured and reported in various ways. Methods for evaluating
pasture research have been reviewed by Ahlgren (1947) who empha-
sized the advantages of using grazing animals. Comparison of results
from different experiments is often complicated by the variety of measur-
ing units used. Some effort has been made to standardize the measuring
and reporting of grazing studies (Report, 1943).

Production in terms of live weight gains of lambs and steers was
between 400 and 500 Ibs. per acre of irrigated pasture in studies con-
ducted on farms in the Sacramento Valley in California (Burlingame,
1949). Albaugh and Sullivan (1949) in Monterey County, California
compared gains per acre made by cows and calves on irrigated pasture
with gains per acre on alfalfa cut and fed green in the dry lot. Gains
on pasture alone were 458 Ibs. per acre compared with gains on alfalfa
of 581 Ibs. Over 2,000 head of cattle were used in this study. Bartels
(1944a) in Australia reports gains from young sheep of over 1,000 Ibs.
per acre per year.

Burlingame (1949) in California obtained production records of dairy
cattle, sheep and beef cattle on irrigated pasture through the use of
"Animal unit months." All classes of livestock were compared on the
basis of total digestible nutrients required per day. Morrison's feeding
standards were used with minor rounding for ease of calculation. An
Animal unit month (A.U.M.) was estimated to be 400 Ibs. T.D.N. for
a mature cow giving 200 Ibs. of butter fat per year. He found that irri-
gated pastures in the Sacramento Valley had an average production of
10 A.U.M. per acre. Over 90 per cent of the pasturage obtained was
during the eight-month period of March through October.

Bateman and Packer (1945) report pasture production of 253 stand-
ard cow days per acre, over a 3-year period. This is an equivalent of
approximately 4 tons alfalfa hay per acre. A standard cow day was
taken to equal 16 Ibs. T.D.N. as proposed in the Preliminary report on
pasture investigations technique (Report, 1943). In another study Bate-
man et al. (1949) found that an unimproved pasture, without fertiliza-
tion, produced 2,921 Ibs. T.D.N. per acre, or 183 standard cow days of
grazing (4-year average) . This pasture was fertilized with 5 tons manure
and 87 Ibs. 43 per cent phosphate annually and yielded 4,111 Ibs. T.D.N.,

376 WESLEY KELLER AND MAURICE L. PETERSON

or 257 standard cow days of grazing annually during the following 5
years. New mixtures now being studied, with fertilization, yielded
5,204 Ibs. T.D.N., or 325 standard cow days of grazing. This produc-
tion is equivalent to 5.1 tons alfalfa hay per acre, and with a herd
averaging 366 Ibs. butter fat per year amounted to 230 Ibs. butter fat
per acre.

In a 5-year study of some improved dairy pastures in Utah, Rich
et al. (1950) report an average of 303 standard cow days of grazing by
high producing cows during a grazing season averaging 156 days per
year. The alfalfa hay equivalent averaged 4.85 tons per acre and in
different seasons ranged from 4.42 to 5.43 tons.

Annual production of good irrigated pastures ranged from 20 to 25
tons green weight in Australia (Andrew, 1947; Rayner, 1947). Bartels
(1944a) reports experimental plots yielding 31 tons green weight per
acre from 12 clippings. Unpublished data from experimental plots in
California showed a range from 22.1 to 35.8 tons green weight per acre,
depending upon the mixture. In Utah, with a comparatively short
grazing season, new mixtures in experimental plots have yielded 16 to
17 tons per acre green-weight, according to Bateman et al. (1949) but
with identical treatment the mixture in general use in the area produced
only 10 tons. The data are based on 4 clippings, each preceding a
grazing period. Jones and Brown (1949) state that an irrigated pasture
should yield as large a tonnage of feed per acre as alfalfa, although
there may be a few exceptions.

2. Economic Studies

Initial costs for developing irrigated pastures for flood irrigation in-
clude surveying, leveling, land planing, construction of irrigation ditches,
drainage ditches and levees, turnout structures, seedbed preparation and
seeding. As of 1948, the total costs of preparing land for flood irriga-
tion by the strip-check* method varied from $30 to $130 per acre in
California (Jones and Brown, 1949). Total annual costs of production
of established irrigated pastures ranged from $17.43 to $58.35 per acre
for different operating units in the Sacramento Valley in 1948 (Reed and
Geiberger, 1948; Burlingame and Kolb, 1948). Cost items included
labor, materials such as irrigation power and fertilizer, taxes, general
expenses, depreciation and interest on the investment.

Burlingame (1949) found that labor for irrigation, fencing and other
purposes accounted for slightly over one-fourth of the total costs per
acre. Interest on land values, facilities, and the pasture stand approxi-
mately equaled labor costs. Total costs averaged $32.62 per acre on
24 records totaling 1,188 acres in 1947, as shown in Table V. On an

IRRIGATED PASTURES 377

average, a little over 60 per cent of total cash and labor costs was for
water and irrigation labor. Water costs ranged from about $2.00 per
acre on land in some irrigation districts to over $20.00 where pumping
from considerable depths was required. Irrigation labor varied from

TABLE V

Principal Items of Cost in the Production of 24 Irrigated Pastures Totaling 1,188
Acres in the Sacramento Valley, 1947 *

Item Cost

Labor for irrigation, fencing, etc. $ 8.67

Water costs, fertilizer, and other materials 6.90

Taxes and general expenses 3.42

Depreciation on stand, irrigation system, etc. 4.88

Interest on land values, facilities, and stand 8.75

Total cost per acre $32.62

Animal unit months of pasturage por acre 9.4

Total cost per animal unit month 3.48

'Burlingame (1949).

less than $3.00 to more than $17.00 per acre depending upon the method
of irrigation, size of head of water, and efficiency of the irrigation sys-
tem. Irrigated pastures have decided economic advantage in regions
of abundant cheap water and relatively level land.

Total costs per acre are of much significance only when considered
in relation to productivity. Gorton (1941), Burlingame and Kolb
(1948), and Hedges (1948) have shown that high producing pastures
usually have lower costs per unit of feed produced than low producing
pastures although total costs per acre may be higher. Three California
studies are reported in Table VI. Burlingame and Kolb (1948) obtained
costs ranging from $1.60 to $8.61 per Animal unit month (A.U.M.).
Differences in productivity were primarily responsible for this range.
These average costs per A.U.M. were used to calculate the equivalent
value of alfalfa hay per ton on the basis of equal amounts of total
digestible nutrients from the pasture and alfalfa. The alfalfa hay aver-
aging 50 per cent total digestible nutrients could cost only $10.73 per
ton to equal the average cost of $4.28 for an Animal unit month. The
seasonal average price of alfalfa hay in California during this same
year (1948) was $22.10 per ton. Gorton (1941) found that in Oregon,
costs per acre decreased as the size of the pasture increased. The pro-
duction of pastures also decreased with increasing size, however, result-

378 WESLEY KELLER AND MAURICE L. PETERSON

ing in the size of pastures having little or no relationship to costs per
unit of production.

TABLE VI

Irrigated Pasture Costs per Acre and per A.U.M. in the Sacramento Valley,

California in 1948

Equiv.

value

Alfalfa

Av. total

No. units

$ Cost per A

.U.M.

per ton *

cost per

Author

studied

Low

High

Av.

acre, $

Reed and Geiberger (1948)

3.00

5.55

3.58

8.95

37.59

Burlingame and Kolb (1948)

1.60

8.61

453

10.57

25.07

Hedges (1949)

2.90

7.02

5.05

12.63

30.82

Average

2.50

7.06

4.28

10.73

31.16

11 Calculated on the basis of 400 Ibs. total digestible nutrients per A.U.M. at aver-
age costs indicated in table.

Seventy-nine dairy records obtained in 1947-1948 were reported by
Shultis and Miller (1949) in the San Joaquin Valley and classified ac-
cording to high and low use of pasture for market milk dairies and
manufacturing milk dairies. The market milk dairies making high
use of pasture showed an advantage over those making low use of
pasture, amounting to $33.99 per year in feed costs per cow and $28.31
profit per cow. Corresponding comparisons for manufacturing milk
dairies was $24.62 advantage in feed costs and $42.24 greater profit
per cow. For both types of dairies the pounds of butterfat sold per
cow were greater under low use of pasture but added costs for hay, con-
centrates, silage and green feeds decreased profits.

Bateman and Packer (1945) have reported production and cost studies
on the Utah Agricultural Experiment Station Dairy Experimental Farm
herd. A summary of their 3-year study is reported in Table VII. On
the basis of butterfat at $.90 per lb., they obtained a return of $5.15 per
dollar of production cost, and a gross return, above feed production cost,
of $149.39 per acre. They conclude "that pasture, when planted on
fertile soil and well managed, is an economical feed for the production
of milk and butterfat and gives a high return per acre when grazed by
good dairy cattle."

Albaugh and Sullivan (1949) made a direct comparison of costs of
pasture feeding and feeding of green alfalfa hay to beef cattle in the
dry lot. The alfalfa was mowed daily, mechanically loaded, hauled by
truck to the feed lot and unloaded by hand. Costs per 100 Ibs. gain

IRRIGATED PASTURES 379

TABLE VII

Per Acre Returns from Pasture ; Utah Agricultural Experiment Station Dairy

Experimental Farm *

Item 3- Year average

Standard cow days of grazing b 253

Supplemental feeds fed: alfalfa (Ibs.) 1127

grain (Ibs.) 1707

Gain in body weight (Ibs.) 224

Total production: milk (Ibs.) 8704

butterfat (Ibs.) 303

Butterfat produced from supplement feeds fed (Ibs.) 98

Production from pasture: milk (Ibs.) 5909

butterfat (Ibs.) 206

Value of butterfat produced from pasture at $.90 per Ib. $185 40

Production cost per acre 36.01

Gross return above feed production cost 149.39

Dollars returned for each dollar cost of pasture production 5.15

Pasture feed production cost of 1 Ib. butterfat .18

* Bateman and Packer (1945).

h A standard cow day is defined as an animal obtaining 16 Ibs. of total digestible
nutrients from pasture per day.

were less on pasture than from dry lot feeding of green alfalfa. Al-
though the alfalfa produced more feed per acre, costs of harvesting and
feeding more than offset this advantage. After considering purchasing
and selling prices, pasture feeding was the more profitable.

Hedges (1948) studied 17 pasture units in the Sacramento Valley
ranging in size from 60 to 490 acres. In these studies gains from beef
cattle and lambs were obtained in addition to costs and carrying capaci-
ties. This method avoided the limitations of the above reported studies
of carrying capacities which provided no information on costs in relation
to gain in weight or milk production. Gains per acre ranged from 154
to 440 Ibs. per acre with an average of 275 Ibs. Animal unit months per
acre averaged 6.1. Costs per 100 Ibs. gain ranged from a low of $6.47
on a 440 acre unit to a high of $14.94 on a 310 acre unit. Average costs
per 100 Ibs. gain was $11.20.

8. Effect of Pastures on Crops that Follow

The beneficial effects of sod crops on soil structure, and fertility
have been extensively investigated by many State Experiment Stations.
There apparently have been no studies conducted specifically with irri-
gated pastures in crop rotations. General farm practice has been to

380 WESLEY KELLER AND MAURICE L. PETERSON

maintain pastures so long as they are productive. Many irrigated pas-
tures in California are still producing satisfactorily after 20 years.

However, several writers have emphasized the importance of using
irrigated pastures in a rotation and have mentioned increased yields of
crops which follow (Klages et al., 1948; Robertson et al., 1948; Starke,
1947; Bartels, 1944b; and Hamilton et al., 1945). Klages et al. further
mention that sweet clover with mountain brome leaves the soil in better
condition for the crops to follow than sweet clover alone when these
plants are used for temporary irrigated pastures. These writers also
suggest that 4 years are usually long enough for a field to remain in
pasture, although the exact length of time may be influenced by many
factors.

VII. PASTURES IN RELATION TO OTHER SOURCES OF FEED
1. Relation to Other Forage Resources

Irrigated pastures are depended upon as the sole source; of forage
in relatively few areas. Throughout the western United States they are
used most extensively for dairy cattle. Irrigated pastures are used to
supplement the range in the production of both beef and sheep. Guilbert
and Hart (1946) report that one and a half million beef cattle and
calves derived 13 per cent of their feed from 150 thousand acres of irri-
gated pasture, while also harvesting the forage from 40 million acres
of range land. About half as many dairy cattle utilized the forage
from 224 thousand acres irrigated pasture and one million acres of
range, while nearly 3 million sheep grazed 120 thousand acres of irri-
gated pasture and 18 million acres of range. The use of irrigated pas-
tures for the production of beef and sheep is increasing in the western
United States.

Throughout much of the western range country beef cattle are win-
tered on wild hay produced on irrigated mountain meadows. The com-
mon practice is to remove the cattle from the meadows as early in the
season as forage is available on the spring ranges. The meadows then
produce a crop of wild hay which is cut in midsummer. The aftermath
is grazed by the cattle after they return to the home ranch in the fall.
The cattle remain on these meadows throughout the winter, being fed
the wild hay produced there. Stewart and Clark (1944) found that
greater total yields, and higher quality forage were produced when the
meadows were grazed 20 to 35 days longer than customary in the spring,
and the hay crop cut at the bloom stage, which is earlier than most
ranchers harvest. This practice benefited spring ranges by reducing
grazing on them, and at the same time kept the animals at the home

IRRIGATED PASTURES 381

ranch during calving and immediately after, allowing the rancher to
give them better care than when they are on the range. Stoddart (1944)
reported that in northern Utah dairy bred heifers did not make as good
gains on summer range as beef, but that they did make satisfactory
development for subsequent milk production. Collins (1945) has re-
ported the successful operations of several ranchers in improving the
productivity of their irrigated mountain meadows.

Jones and Mumford (1944) in Oregon, recommend sudangrass as a
supplement to permanent pasture in late summer, and Abruzzi winter
rye for late winter and early spring on well drained land. In South
Africa, Bonsma (1947) found winter cereals, where adapted, a cheaper
source of succulent feed for dairy cows than silage. Madson and Love
(1948) report that in California silage is made from a variety of crops,
but that while pasturage is available it is a much more economical source
of feed. In Australia, Bartels (1944b) reports the effective use of a num-
ber of annual forage crops, both as a supplement to irrigated pastures
and to take livestock from nonirrigated areas through drought years.
Corn, millets, and sorghum have been used for this purpose.

2. Supplemental Feeding

Jones and Brown (1949) recommend having dry roughage available
to cattle and sheep at all times while on irrigated pasture, to reduce
bloat and increase dry-matter consumption. Jones and Mumford (1944)
point out that the capacity of a dairy cow to consume green forage is
limited, and that high producing cows on pasture will require a supple-
ment of grain for maximum production. If cows on pasture are liberally
fed such supplements as ryegrass hay or corn silage, some protein con-
centrate may be needed, in addition to grain. According to Klages et al.
(1948) dairy cows will produce one Ib. of butterfat daily and maintain
body weight on irrigated pasture alone but production is increased if
alfalfa hay is also fed. Hamilton et al. (1945) recommend feeding an
increasing amount of concentrate to high producing cows, as the produc-
tivity of the pasture declines. Ewalt and Morse (1942) and Bateman
(1945) have prepared tables to guide dairymen in feeding grain to cows
on irrigated pasture.

Beef cattle will make excellent growth on irrigated pasture without
any supplements. Carbohydrate concentrates are frequently used in the
fattening process.

382 WESLEY KELLER AND MAURICE L. PETERSON

REFERENCES

Ahlgren, H. L, 1945. Imp. Bur. Pastures & Forage Crops, Aberystwyth. Bull. 34,

139-160.

Ahlgren, H. L. 1947. J. Am. Soc. Agron. 39, 240-259.

Ahlgren, H. L., and Graber, L. F. 1940. Wisconsin Agr. Ext. Ser. Circ. 300, 8 pp.
Albaugh, R., and Sullivan, W. 1949. Monterey County Agr. Ext. Ser. (California),

Mimeo. 4 pp.

Anderson, A. J. 1946. J. Council Sci. Ind. Res. Australia 19, 1-15.
Andrew, W. D. 1947. J. Dept. Agr. Victoria Australia 45, 1-9, 55-63.
Australia. 1947. J. Dept. Agr. Victoria Australia 45, 473-479, 567-576.
Barshad, I. 1948. Soil Sci. 66, 187-195.

Bartels, L. C. 1944a. J. Dept. Agr. Victoria Australia 42, 391-397.
Bartels, L, C. 1944b. J. Dept. Agr. Victoria Australia 42, 433-436.
Bartels, L. C. 1947. J. Dept. Agr. Victoria Australia 45, 201-210.
Bartels, L. C., Beruldsen, E. T., and Morgan, A. 1932. J. Dept. Agr. Victoria

Australia 3D, 187-205.

Bartels, L. C., and Morgan, A. 1944. J. Dept. Agr. Victoria Australia 42, 291-295.
Bateman, G. Q. 1940. Farm Home Sci. (Utah) 1, No. 2, 1, 9-10.
Bateman, G. Q. 1943. Farm Home Sci. (Utah) 4, No. 1, 8-9, 15.
Bateman, G. Q. 1945. Utah Agr. Expt. Sta. Mimeo Ser. 315, 1-7.
Bateman, G. Q., Keller, W., and Packer, J. E. 1949. Farm Home Sci. (Utah) 10,

No. 1, 6-7, 17.

Bateman, G. Q., and Packer, J. E. 1945. Farm Home Sci. (Utah) 6, No. 2, 10-11.
Bingham, G. H., and Monson, O. W. 1946. Montana Agr. Ext. Ser. Bull. 237, 1-24.
Bonsma, J. C. 1947. Union So. Africa Dept. Agr. Bull. 284.
Boyd, G. W. 1945. Wyoming Agr. Ext. Ser. Circ. 90, 1-19.

Bracken, A. F., and Evans, R. J. 1943. Utah Agr. Ext. Ser. N.S. Bull. 120, 1-48.
Britton, J. W., and Goss, H. 1946. J. Am. Vet. Med. Assoc. 108, 176-178.
Burlingame, B. B. 1949. California Agr. 3, 13-14.
Burlingame, B. B., and Kolb, A. C. 1948. Calif. Agr. Ext. Ser. Fifth Ann. Irrigated

Pasture Study, Colusa County, California, Mimeo. 10 pp.
Cole, H. H., Huffman, C. F., Kleiber, M., Olsen, T. M., and Schalk, A. F. 1945.

/. Animal Sci. 4, 183-236.

Cole, H. H., and Kleiber, M. 1945. J. Vet. Res. 6, 188-193.
Cole, H. H., Mead, S. W., and Jiegan, W. M. 1943. J. Animal Sci. 2, 285-294.
Collins, W., Jr. 1945. Soil Conservation 10, 165-167, 175.
Cunningham, I. J. 1948. New Zealand J. Agr. 77, 519.
Davies, J. G., and Christian, C. S. 1945. Imp. Bur. Pastures & Forage Crops,

Aberystwyth. Bull. 34, 73-96.

Dept. of Agronomy. 1949. Utah Agr. Ext. Ser. Bull. 173, 10-13.
Doneen, L. D. 1948. Univ. California, Division of Irrigation, Mimeo.
Emrick, W. E. 1948. Cal. Agr. Ext. Ser. Kern Co., Mimeo.
Ewalt, H. P., and Jones, I. R. 1939. Oregon Agr. Expt. Sta. Bull. 366, 1-25.
Ewalt, H. P., and Morse, R. W. 1942. Oregon Agr. Ext. Ser. Bull. 592, 1-4.
Ferguson, W. S., Lewis, A. H., and Watson, S. J. 1943. J. Agr. Sci. 33, 44-51.
French, H. T. 1902. Idaho Agr. Expt. Sta. Bull. 33, 87-107.
Gorton, W. W. 1941. Oregon Agr. Expt. Sta. Bull. 392, 1-51.
Guilbert, H. R., and Hart, G. H. 1946. California Agr. Ext. Circ. 131, 1-157.

IRRIGATED PASTURES 383

Hamilton, J. G., Brown, G. F., Tower, H. E., and Collins, W., Jr. 1945. U.S. Dept.

Agr. Farmers Bull. 1973, 30 pp.

Hansen, D. 1924. Montana Agr. Expt. Sta. Bull. 166, 1-26.
Harris, F. S. 1913. Utah Agr. Expt. Sta. Circ. 15, 33-43.

Hayward, H. E., and Magistad, O. C. 1946. UJS. Dept. Agr. Misc. Pub. 607, 1-27.
Hayward, H. E., and Wadleigh, C. H. 1949. Advances in Agron. 1, 1-38.
Hedges, T. R. 1948. U. California, Giannini Foundation Agr. Econ., Mimeo. 20 pp.
Hodgson, R. E., Grunder, M. S., Knott, J. C., and Ellington, E. V. 1934. Wash-
ington Agr. Expt. Sta. Bull. 294, 1-36.
Jones, B. J., and Brown, J. B. (Rev. by Miller, M. D., and Booher, L. J.). 1949.

California Agr. Ext. Ser. Circ. 125, 1-59.

Jones, I. R., and Mumford, D. C. 1944. Oregon Agr. Expt. Sta. Circ. 165, 1-4,
Keller, W., Bateman, G. Q., and Packer, J. E. 1945. Farm Home Sci. (Utah) 6,

No. 4, 7-10, 15.
Keller, W., Bateman, G. Q., and Packer, J. E. 1947a. Farm Home Sci. (Utah) 8,

No. 1, 6-7, 14.
Keller, W., Bateman, G. Q., and Packer, J. E. 1947b. Farm Home Sci. (Utah) 8,

No. 4, 5, 16-18.
Klages, K. H., Stark, R. H., Anderson, G. C., Fourt, D. L., Whitman, E. W., and

Keith, T. B. 1948. Idaho Agr. Ext. Ser. Bull. 174, 1-14.

Kramer, P. J. 1949. Plant and Soil Water Relationships. McGraw-Hill, New York.
Law, G. A., Singleton, H. P., and Ingham, I. M. 1945. Washington Agr. Ext. Ser.

Bull. 319, 1-8.
Levy, E. B. 1945. Imp. Bur. Pastures & Forage Crops, Aberystwyth. Bull. 34,

97-121.

Levy, E. B. 1949. United Nations Sci. Conf. Conser. Util. Resources. Land Re-
sources Sec. 6 (b), Mimeo.
Lewis, A. H. 1943a. /. Agr. Sci. 33, 52-57.
Lewis, A. H. 1943b. /. Agr. Sci. 33, 58-63.

Madson, B. A., and Love, R. M. 1948. U.S. Dept. Agr. Yearbook Agr., pp. 582-586.
Magistad, O. C. 1945. Botan. Rev. 11, 181-230.

Magistad, 0. C., and Christiansen, J. E. 1944. U.S. Dept. Agr. Circ. 707, 1-32.
Matlock, R. L. 1943. Arizona Agr. Ext. Ser. f Mimeo. 15 pp.
Mead, S. W., Cole, H. H., and Regan, W. M. 1944. J. Dairy Sci. 27, 779-791.
Morgan, A. 1947. J. Dept. Agr. Victoria Australia 45, 111-115.
Morgan, A. 1949. /. Dept. Agr. Victoria Australia 47, 97-105, 199-207, 241-247.
Morgan, A., and Rayner, G. B. 1941. /. Dept. Agr. Victoria Australia 39, 15-16,

43-47.
Pittman, D. W., and Bennett, W. H. 1948. Utah Agr. Expt. Sta. Rept. on Project

160, Mimeo. 7 pp.

Post, A. H., and Tretsven, J. O. 1939. Montana Agr. Ext. Ser. Bull. 174, 1-14.
Rampton, H. H. 1945. Oregon Agr. Expt. Sta. Bull. 427, 1-22.
Rayner, G. B. 1941. /. Dept. Agr. Victoria Australia 39, 314-318.
Rayner, G. B. 1947. J. Dept. Agr. Victoria Australia 45, 123-126.
Reed, A. D., and Geiberger, R. C. 1948. California Agr. Ext. Ser., Mimeo. 3 pp.
Report. 1943. /. Dairy Sci. 26, 353-369.
Rich, L. H., Bracken, A. F., Bennett, W. H., and Baird, G. T. 1950. Utah Agr.

Ext. Ser. Bull. 188, 1-24.
Richards, L. A. (Ed.) 1947. Diagnosis and Improvement of Saline and Alkali

Soils. UJS. Regional Salinity Laboratory, Mimeo. pp. 1-157.

384 WESLEY KELLER AND MAURICE L. PETERSON

Robertson, D. W., Weihing, R. M., and Tucker, R. H. 1948. Colorado Agr. Ext.

Ser. Bull. 403-A, 1-46.

Sanborn, J. W. 1894. Utah Agr. Expt. Sla. Bull. 33, 1-8.
Schoth, H. A. 1944. Oregon Agr. Expt. Sta. Circ. 161, 1-12.
Schwendiman, J. L., and Mullen, L. A. 1944. J. Am. Soc. Agron. 36, 783-785.
Schwendiman, J. L., Sackman, R. F., and Haf enrich ter, A. L. 1940. U.S. Dept. Agr.

Circ. 558, 1-16.

Semple, A. T., and Hein, M. A. 1944. UJS. Dept. Agr. Farmers Bull. 1942, 1-22.
Semple, A. T., Vinall, H. N., Enlow, C. R., and Woodward, T. E. 1934. UJS. Dept.

Agr. Misc. Pub. 194, 1-89.
Shultis, A., and Campbell, A. L. 1943. Calif. Agr. Ext. Ser., San Bernardino Co.,

Mimeo. 7 pp.

Shultis, A., and Miller, M. D. 1949. California Agr. 3, 6.
Southworth, W. L. 1949. Soil Conservation 14, 280-282.

Starke, J. S. 1947. Union S. Africa Dept. Agr. Bull 279, (Agr. Res. Ser. 47), 1-30.
Stewart, G. 1945. Imp. Bur. Pastures cfc Forage Crops, Aberystwyth. Bull. 34,

180-195.

Stewart, G., and Clark, I. 1944. J. Am. Soc. Agron. 36, 238-248.
Stoddart, L. A. 1944. Utah Agr. Expt. Sta. Bull. 314, 1-24.
Sullivan, W., and Winright, G. L. 1943. California Agr. Ext. Ser., Imperial County,

Mimeo.
't Hart, M. L. 1949. United Nations Sci. Conf. Conser. Util. Resources. Land

Resources Sec. 6 (b), Mimeo.

Thorne, D. W. 1948. UJS. Dept. Agr. Yearbook Agr. pp. 141-143.
Thorne, D. W., and Peterson, H. B. 1949. Irrigated Soils, Their Fertility and

Management. Blakiston, Philadelphia.

Veihmeyer, F. J. 1948. California Agr. Expt. Sta. Circ. 388, 1-18.
Welch, J. S. 1914. Idaho Agr. Expt. Sta. Bull 80, 1-15.
Welch, J. S. 1917. Idaho Agr. Expt. Sta. Bull. 95, 1-17.

Author Index

Names in parentheses indicate coauthors of the references and are included to assist in
locating references where a particular name is not on a given page.

Example: Adams, J. E., 20 (see Rigler) means that Rigler et al. will be mentioned on page
20, the et al. accounting for Adams. This article can be located under Rigler in the list of
references.

Numbers in italics refer to the pages on which references are listed in bibliographies at the
pnd of each article.

Aandahl, A. R., 186, 187, 202, 208
Adams, J. E., 20 (see Rigler), 24 (see

Hooten), 27 (see Jordan), 42, 43, 48,

75, 76, 79

Adams, W. W., 318, 849
Adsuar, J., 127, 164
Ahlgren, G. H., 209, 212, 216, 230
Ahlgren, H. L., 210, 213, 214, 216, 219,

221, 222, 224, 225, 227, $80, 231, 364,

375, 382

Albada, M., 288, 311
Alban, E. K., 148, 152
Albaugh, R., 375, 378, 882
Alexander, L. T., 184, 203, 239, 269
Allaway, W. H., 177, 185 (see Larson),

$04

Allen, L. A., 283, 311
Allen, T. C., 140, 153
Allison, F. E., 88, 90, 93, 97 (see Pinck),

103 (see Pinck), 109, 111
Allison, L. E., 248 (see Christiansen),

262, $69, 270
Allison, R. V., 85, 109
Allyn, R. B., 262, $89
Alsmeyer, H. L., 47, 78
Amos, A., 280, 281, 811
Anderson, A., 295, 313
Anderson, A. B. C., 239, 240, 245, $69,

270

Anderson, A. J., 374, 882
Anderson, D. B., 25, 56, 57, 74, 75, 79
Anderson, G. C., 359 (see Klages), 364

(see Klages), 367 (see Klages), 371

(see Klages), 373 (see Klages), 380

(see Klages), 381 (see Klages), 383
Anderson, L. C., 118, 153
Anderson, M. A., 199, $08

Anderson, M. E., 123, 124, 152
Anderson, W. S., 122, 16$
Andrew, R. H., 122, 152
Andrew, W. D., 373, 376, 382
Andrews, W. B., 17, 74, 117, 152
Andrus, C. F, 123, 125, 15$
Appleman, D., 249, 271
Archibald, J. G., 283, 285, 286, 290, 294,

811
Arminger, W. H., 93 (see Allison, F. E.),

109

Arnell, J. C., 253, $69
Aronovici, V. S., 249 (see Donnan), 266

(see Donnan), $70

Asbury, C. E., 293 (see Pentzer), 313
Ashton, T., 128, 15$
Atkinson, G. F., 26, 28, 30, 76
Atwood, S. S., 212, 213, 230
Autrey, K. M., 286, 311
Ayers, A. D., 117, 168

Babcock, S. M., 282, 311

Badley, J. E., 208, $80

Bailey, T. L. W., Jr., 59 (see Richard-
son), 60 (see Richardson), 79

Bainer, R., 143, 152

Baird, G. T., 376 (see Rich), 383

Balasubrahrnanyan, R., 71, 80

Baldwin, M., 193, 195, 203

Ballard, W. W., 29, 77

Barber, L., 281, 314

Barber, T. S., 38, 77

Barducci, T. D., 29, 76

Barger, E. L., 302, 814

Barker, H. D., 24, 25, (see Berkley), 57
(see Barre; Berkley), 58 (see Berk-
ley), 62, 74, 79

385

386

AUTHOR INDEX

Barlow, G. E., Jr., 303, 311

Barnes, J. C., 60, 79

Barnes, W. C., 124, 162

Barr, G. W., 49, 78

Barr, H, T., 305, 911

Barre, H. W., 57, 79

Barrens, K. E., 121, 162

Barshad, I., 158, 166, 175, 184, 209, 373,

374, 882
Bartels, L. C., 355, 360, 364, 366, 367, 368,

370, 371, 372, 373, 375, 376, 380, 381,

382

Bartholomew, W. V., 88, 110
Bateman, G. Q., 359, 364, 365, 367, 371,

372, 375, 376, 378, 379, 882, 383
Bauer, F. C., 201 (see DeTurk), 202, SOS
Baver, L. D., 243, 245, 255, 269, 271, 272
Bay, C. E., 99 (see Hays), 110
Bear, F. E., 15 (see Wallace), 76, 100,

109, 346, 349

Beasley, J. 0., 4, 69, 70, 80
Beatie, H. G., 96 (see Collison), 109
Beaumont, A. B., 117, 162
Beavens, E. A., 277 (see Phillips), 314
Bechdel, S, I., 282, 283 (see Stone), 286

(see Stone), 314
Beckley, M. S., 366
Becnel, I. J., 35, 77

Beeson, W, M., 291 (see Peterson), 313
Bell, F. G., 98 (see Borst), 100 (see

Smith, D. D.), 109, 111
Bender, C. B., 285, 288, 290 (see Taylor),

811, 314

Bendixen, T. W., 252, 255, 269
Benne, E. J., 120, 163
Bennett, E., 294 (see Archibald), 311
Bennett, H. H., 85, 109
Bennett, J. B., 349
Bennett, W. H., 353, 372, 376 (see Rich),

383

Bentley, C. F., 172, 204
Berkley, E. E., 25, 57, 58, 74, 79
Beruldsen, E, T., 370 (see Bartels), 371

(see Bartels), 382

van Beynum, J,, 282, 284, 291, 292, 911
Bibby, F. F., 37, 78
Biddulph, 0., 17, 18, 74
Bingham, F, T., 85, 86 (see Jenny), 110
Bingham, G. H., 363, 382
Bird, H. R., 291 (see Peterson), 313

Bishop, J. C., 145, 162

Bizzell, J. A., 91, 92, 110

Black, A., 279 (see Swift), 914

Blair, A. W., 102, 109

Blaney, H. F., 266 (see Donnan), 270

Blank, L. M., 27 (see Jordan), 76

Blaser, R. E., 215, 290 9 303, 307, 309

Bledsoe, M. R., Jr., 90, 109

Bledsoe, R. P., 19, 20, 78, 96 (see Hoi-
ley), 110

Bledsoe, R. W., 19 (see Harris), 75

Bodman, G. B., 252, 289, 270

Bogess, T. S., Jr., 96 (see Holiey), 110

Bohn, G. W., 134, 152

Bohstedt, G., 287, 289 (see Johnson,
B. C.), 311, 312

Bondy, F. F., 35, 75

Bonnen, C. A., 45, 47 (see Magee), 78, 79

Bonsma, J. C., 381, 982

Booher, L. J., 212, 219, 220, 221, 222, 226,
231, 359 (see Jones, B. J.), 360 (see
Jones, B. J.), 361 (see Jones, B. J.),
362 (see Jones, B. J.), 363 (see Jones,
B. J.), 369 (see Jones, B. J.), 370
(see Jones, B. J.), 376 (see Jones,
B. J.), 381 (see Jones, B. J.), 383

Borst, H. L., 98, 109

Bosshardt, D. K., 285, 288, 311

Botelho da Costa, J. V., 239, 262, 269,
272

Bourbeau, G. A., 185 (see Jackson), 203

Bowlsby, C. C., 97, 110

Boyd, G. W., 371, 382

Boyoucos, G, J., 245, 269

Bracken, A. F., 92, 95, 96, 109, 364, 366,
376 (see Rich), 982, 383

Bradfield, R., 241, 252, 289

Bradshaw, M. A., 299, 312

Brain, S. G., 43, 79

Brandt, P. M., 212, 231

Bratzler, J. W., 279 (see Swift), 914

Bray, R. H., 183, 184, 185, 208

Breazedale, E., 261, 289

Briggs, L. J., 260, 269

Brigl, P., 293, 911

Brinkman, H. C., 256, 289

Britton, J. W., 373, 382

Broadbent, F. E., 89, 101, 109

Brooks, 0. L., 216, 218, 220, 221, 230

Brouwer, E., 285, 291, 311

AUTHOR INDEX

387

Brown, A. B., 24, 74

Brown, A. L., 94, 109

Brown, B. A., 211, 213, 214, 215, 217, 224,
225, 230, 231

Brown, G. A., 13, 74

Brown, G. F., 330, 334, 338, 349, 353 (see
Hamilton), 355 (see Hamilton), 360
(see Hamilton), 362 (see Hamilton),
363 (see Hamilton), 364 (see Hamil-
ton), 366 (see Hamilton), 367 (see
Hamilton), 371 (see Hamilton), 372
(see Hamilton), 374 (see Hamilton),
380 (see Hamilton), 381 (see Hamil-
ton), 383

Brown, H. B., 17, 74

Brown, H. D., 18, 74

Brown, J. B., 359, 360, 361, 362, 363, 369,
370, 376, 381, 383

Brown, J. G., 29, 30, 76

Brown, P. E., 90, 109, 175, 201, 804, 206

Brown, W. L., 96 (see Holley), 110

Browning, D. R., 248 (see Smith, R. M.),
249, 253, 255 (see Smith, R. M.), 272

Browning, G. M., 241, 261, 268 (see Wil-
son), 269, 270, 272

Brues, C. T., 36, 75

Bruhn, H. D., 302, 811

Bruins, J. F., 211, 231

Bryant, J. C., 245, 272

Buchanan, R. E., 291, 811

Buckingham, E., 243, 269

Buckman, H. O., 91, 93, 105, 109, 110

Buice, G. D., 216, 218, 220, 221, 230

Bullard, E. T., 147, 153

Burcalow, F. V., 210, 213, 214, 216, 219,
221, 222, 224, 225, 227, 230, 231

Burck, C. R., 244 (see Colman), 270

Burgess, P. S., 90, 109

Burgesser, F. W., 144, 145, 152

Burgis, D. S., 136, 152

Burlingame, B. B., 375, 376, 377, 378, 382

Burr, H. R., 150, 152

Bushnell, L. D., 282, 812

Bushnell, T. M., 196, 208

Byron, M. H., 49 (see Smith, H. P.), 79

Calhoun, P. W., 19 (see Harris), 75
Calhoun, S. L., 35, 77

Call, L. E., 90, 109

Camburn, O. M., 279, 290, 296, 297, 299,
305, 307, 308, 309, 811, 818

Campbell, A. L., 884

Campbell, R. B., 239, 271

Carew, H. J., 145, 152

Carew, J., 137, 146, 147, 152

Carleton, E. A., 98, 111

Carman, P. C., 253, 254, 269

Carolus, R. L., 141, 152

Carpenter, C. W., 29, 76

Carter, M., 136, 155

Gary, C. A., 297 (see Kane), 313

Cassidy, T. P., 34, 38, 77

Chamberlin, V. D., 226, 231

Chandler, R. F., Jr., 171, 208, 245, 270

Chang, S. C., 117, 1S8

Chapman, A. G., 329, 334, 341, 849

Chapman, A. J., 33, 77

Chapman, H. D., 116, 117, 152

Cheftel, H., 289 (see Voinovitch), 315

Chenoweth, O. V., 116, 153

Childs, E. C., 236, 241, 244, 245, 251, 253,
255, 257, 258, 265, 266, 267, 269

Christensen, H. R., 242, 248, 252, 270

Christian, C. S., 364, 382

Christiansen, J. E., 248, 249, 267, 270,
354, 383

Churchman, W. L., 75

Clark, I., 380, 884

Clark, J. C., 72, 80

Clarke, A. E., 132, 158

Clarke, M. F., 301 (see Kalbfleish), 302
(see Kalbfleish), 303 (see Kalb-
fleish), 313

Cline, M. C., 171, 208

Clore, W. J., 140, 152

Coad, B. R., 34, 77

Coke, J. E., 208, 209, 212, 231

Cole, H. H., 367, 368, 882 9 383

Coleman, R., 17, 19, 74

Collander, R., 15, 75

Collins, W., Jr., 353 (see Hamilton), 360
(see Hamilton), 362 (see Hamilton),
363 (see Hamilton), 364 (see Hamil-
ton), 366 (see Hamilton), 367 (see
Hamilton), 371 (see Hamilton), 372
(see Hamilton), 374 (see Hamilton),
380 (see Hamilton), 381, 882, 888

388

AUTHOR INDEX

Collis-George, N., 251, 253, 255, 257, 889.
(See also George, N. C.)

Collison, R. C., 96, 109

Colman, E. A., 244, 245, 252, 261, 869 9 270

Conner, S. D., 331 (see Thornton), 349

Conrad, C. M., 57, 59 (see Richardson),
60 (see Richardson), 79

Conybeare, A. B., 91, 92, 93, 104, 110

Cook, 0. F., 11, 66, 80

Cooper, H. P., 15, 16, 17, 75

Cooper, T. P., 286, 811

Cope, J. T., Jr., 18, 75

Corbet, A. S., 86, 109

Corkill, L., 212, 231

Cottrell-Dormer, W., 131, 152

Cox, H. R., 222, 831

Crafts, A. S., 146, 147, 162

Crampton, E. W., 278, 279, 311

Crasemann, E., 285, 287, 290, 311

Crim, R. F., 147, 153

Crooks, G. C., 279 (see Camburn), 290
(see Camburn), 296 (see Camburn),
. 297 (see Camburn), 299 (see Cam-
burn), 305 (see Camburn), 307 (see
Camburn), 308 (see Camburn), 309
(see Camburn), 311

Croston, F. E., 133, 153

Crowther, E. M., 95, 110

Crowther, F., 20, 22, 75

Croxton, W. C., 328, 330, 349

Cruess, W. V., 289, 313

Cuba, E. F., 31, 76

Cullison, A. E., 293, 311

Cummings, R. W., 106, 110, 245, 270

Cunningham, I. J., 359, 388

Currence, T. M., 130, 131, 158, 153

Curtis, L. C., 133, 152

Curtis, 0. F., 295, 304, 311

Cuthbertson, D. C., 122, 152

Dahlstrand, N. P., 196, 804
Damon, S. C., 120, 153
Dana, B, F, 124, 158
Danielson, L. L., 147, 158
Darcy, H., 246, 247, 270
Dastur, R. H., 13, 75
Davidson, A. L. C., 239, 270
Davies, E. B., 120, 158

Davies, J. G., 364, 388

Davis, G. N., 122, 127, 132, 158, 153, 154

Davis, R. B., Jr., 301, 303, 311

Davis, R. E., 296 (see Hodgson), 318

Davis, W. E., 245, 270

Dawson, J. E., 277, 293, 299, 301, 304, 311,

313

Day, P. R., 239, 270
Dean, H. A., 33 (see Gaines, J. C.), 35,

37, 78

Dean, L. A., 18 (see Hall), 75, 85, 110
Dearborn, C. H., 147, 158
van Deemter, J. J., 265, 266, 270
Denman, T. E., 122, 158
Dennett, R. K, 127, 153
Den Uyl, D., 330, 349
DeTurk, E. E., 201, 803
Dexter, S. T., 227, 228, 838, 294, 303, 318
Dhar, N. R., 90, 110
Dickerson, W. H., Jr., 300 (sec Schaller),

314

Dickey, J. B. R., 222, 831
Dijkstra, N. D., 286, 290, 291 (see Brou-

wer), 311, 312
Doak, B. W., 212, 831
Dobie, J. B., 298, 312
Dobson, S. H., 209 (see Lovvorn), 213,

214, 217, 218, 220, 221, 225, 831
Dodd, D. R., 213, 214, 216, 219, 220, 221,

222, 223, 224, 227, 229, 838
Dodge, D. A., 94, 110, 202, 203
Domingo, C. E., 268, 270
Donaldson, R. W., 212, "831
Donat, J., 241, 270
Doneen, L. D., 117, 142, 152, 370, 388
Donnan, W. W., 249, 266, 267, 270
Doty, D. M., 128, 158
Doyle, C. B., 57 (see Barre), 79
Drewes, H., 122, 158
Dudley, R. F., 295, 812
Duff, G. H., 276, 277, 318
Duffee, F. W., WO; 318
Duggar, F., 92, 110
Duley, F. L., 268, 270
Dulin, T. G., 21, 75
Dunham, R. S., 147, 153
Dunlap, A. A., 13, 23, 28 (see Lyle), 76,

AUTHOR INDEX

389

Dunnam, E. W., 34, 35, 72, 77, 80
Durocher, J., 289 (see Voinovitch), 815

Eaton, F. M., 12, 14, 15, 16, 18, 21, 22,
24, 25, 75

Eby, C., 214, 216, 217, 220, 221, 224, 225,
226, 229, 281, 232

Eddy, C. O., 33, 77

Eden, A., 275, SIS

Edlefson, N. E., 239, 240, 242, 245, 261
(see Veihmeyer), 269, 270, 272

Edler, G. C., 229, 231

Edwards, F. E., 117, 152

Ellenberger, H. B., 279 (see Camburn),
290 (see Camburn), 296 (see Cam-
burn), 297 (see Camburn), 299 (see
Camburn), 305 (see Camburn), 307
(see Camburn; Newlander), 308
(see Camburn; Newlander), 309 (see
Camburn; Newlander), 811, 313

Ellington, E. V., 366 (see Hodgson), 383

Elliott, R. F., 278, 279 (see Forbes), 303,
312

Ellis, G. E, 279, 312

Ellis, N. K., 147, 149, 16S, 155

Ellis, N. R., 280 (see Shepherd), 283 (see
Shepherd), 314

Ellison, W. D., 268, 270

Elting, J. P., 60, 79

Elvehjem, C. A., 291 (see Johnson, B.
C.), 312

Emery, F. E., 208, 231

Emrick, W. E., 374, 382

Enlow, C. R., 366 (see Semple), 384

Ensminger, L. E., 18, 75, 87, 88, 110

Epps, J. M., 123, 153

Ergle, D. R., 13, 20, (see Rigler), 24, 75,
76

Esten, W. M., 282, 312

Eto, W., 122, 153

Evans, R. E., 276, 315

Evans, R. J., 364, 366, 382

Ewalt, H. P., 226, 281 9 372, 381, 382

Ewing, E. C., 23

Ewing, K. P., 34, 35, 37, 77, 78

Ezekiel, W. N., 27, 77

Fahmy, T., 28, 76

Fair, G. M., 254, 270

Fairbank, J. P., 48, 79

Faulwetter, R. C., 31, 76

Feeder, K. E., 285 (see Kirsch), SIS

Feng, C. L., 241, 258, 270, 271

Fenton, F. A., 34, 77

Fergus, E. N., 214, 227, 231

Ferguson, W. S., 275, 278, 279, 296 (see
Watson), 305, 312, 315, 373, 382

Fife, L. C., 33 (see Chapman), 77

Fingerling, 287

Fink, D. S., 211, 213, 215, 217, 221, 223,
225, 227, 281

Finn-Kelcey, P. A., 299, 301, 312

Fireman, M., 242, 248 (see Christiansen),
249, 270, 271

Fisher, E. A., 140, 153

Fitch, J. B., 285, 814

Fleischmann, F., 277, 312

Fletcher, J. E, 245, 270

Fletcher, R. K., 33, 77

Flint, R. F., 169, 203

Flock, M. V., 226, 231

Flodkvist, 268

Fokin, V. M., 288 (see Mikhin), SIS

Forbes, E. B., 279, 812

Ford, W. H., 229, 231

Forward, D. F., 277, 312

Foster, R. E., 126 (see Walker), 155

Fourt, D. L., 359 (see Klages), 364 (see
Klages), 367 (see Klages), 371 (see
Klages), 373 (see Klages), 380 (see
Klages), 381 (see Klages), 383

Fowler, R. H., 84, 110

Fox, W. W., 249 (see Donnan), 270

Fraps, G. S., 90, 110

Frazer, R. R., 331 (see Thornton), 849

Frazier, W. A., 125, 127, 153

Freckman, 268

Fred, E. B., 282 (see Peterson), 283 (see
Peterson), 284 (see Peterson), 288
(see Peterson), 290 (see Peterson),
814

Free, G. F, 268, 271

French, H. T., 358, S82

Frudden, C. E., 300, 301, 302, 812

Fry, W. H., 184, 208

390

ATJTHOB INDEX

Fuelleman, R. F., 211, 219, 220, 221, 225,

281

Fullmer, F. S., 116, 164
Fulmer, E. I., 291, 311
Futral, J. O., 17 (see Skinner), 76

Gaddy, V. L., 93 (see Allison, F. E.), 97

(see Pinck), 103 (see Pinck), 109,

111
Gaines, J. C., 11, 19, 33, 34, 35, 37, 40,

76, 77, 78

Gaines, R. C., 33, 35, 78
Gainey, P. L., 90, 93, 110, 111
Galloway, L. D., 277, 818
Galpin, S. L., 330, 339 (see Tyner), 849
Garber, R. J., 213, 215, 881
Gardner, M. W., 126, 155
Garman, W. H., 15 (see Cooper), 16, 17,

Garwood, F., 244 (see Schaffer), 272
Gauch, H. G., 262 (see Wadleigh), 272
Geiberger, R. C., 376, 378, 383
George, L. V., 125, 153
George, N. C, 249, 251, 253, 255, 257,

269, 270. (See also Collis-George,

N.)

Gerlach, M., 287, 296, 312
Gessel, 8. P., 85, 86 (see Jenny), 110
Geyer, L. E., 318, 349
Gieseking, J. E., 87, 110
Giglioli, I., 287, 812
Gilbert, W. W., 31, 76
Godden, W. J., 290, 312
Godfrey, G. H., 31, 76
Goff, C. C., 31, 77
Golding, N. S., 280, 313
Goode, W. E., 249, 270
Goring, C. A. I., 88, 110
Gorini, C., 291, 292, 312
Gorton, W. W., 377, 382
Goes, H., 373, 888
Goss, M. J., 277 (see Phillips), 314
Gottlieb, S., 87, 110
Gouy, M. G., 242, 270
Gowda, R. N., 90, 110
Graber, L. F., 364, 382
Graham, E. R., 176, 208
Graham, J. B., 177, 804
Grandt, A, F., 330, 849

Grau, F. V., 216, 218, 227, 231

Graves, R. R., 305 (see Hodgson), 812

Gray, L. C., 849

Greaves, J. E., 95, 96, 109

Green, T. C., 95, 111

Greene, H., 260, 270

Gregg, L. E., 266, 271

Greulach, V. A., 13, 76

Grimball, E. L., Jr., 132, 155

Grimes, M. A., 49, 79

Grinnells, C. D., 302 (see Weaver), 315

Groves, D. E., 339, 849

Grunder, M. S., 366 (see Hodgson), 383

Guilbert, H. R., 278, 297, 312, 380, 382

Gull, P. W., 42, 43, 48, 79

Gunn, K. C., 27 (see Neal), 77

Gunness, C. I., 290, 311

Gunther, E., 287 (see Gerlach), 312

Gustafson, A. F., 92, 110

Gustafsson, Y., 265, 270

Haas, H. J., 83 (see Myers), 84 (see
Myers), 94 (see Myers), 95 (see
Myers), 108 (see Myers), 111

Haber, E. S., 122, 154

Haddock, J. L., 166, 184, 804, 212, 217,
220, 221, 225, 229, 881

Haddon, C. B., 102, 110

Hafenrichter, A. L., 364 (see Schwendi-
man), 384

Hagood, E. S., 13 (see Brown, G. A.), 74

Haines, W. B., 242, 270

Haise, H. A., 245, 270

Hale, G. A., 101, 103, 110

Hall, N. S., 18, 75

Hallsted, A. L., 83 (see Myers), 84 (see
Myers), 94 (see Myers), 95 (see
Myers), 108 (see Myers), 111

Hamilton, J. G., 353, 355, 360, 362, 363,
364, 366, 367, 371, 372, 374, 380, 381,
383

Hamilton, J. M., 118, 153

Hammons, J. G., 117, 152

Hamner, K. C., 293 (see Pentzer), 313

Hanawalt, W. B., 244 (see Colman), 270

Hanley, F., 290, 815

Hanna, G. C., 135, 154

Hansen, D., 359, 883

Hansford, C. G., 31, 7$

AUTHOB INDEX

391

Hanson, W. K., 37 (see Loftin), 78

Harding, S. W., 266, 270

Hardy, F,, 86, 110

Hare, J. F., 31, 76

Hargrove, B. D., 28 (see Lyle), 76

Harlan, J. D., 96 (see Collison), 109

Harland, S. C., 4, 66, 67, 69, 80

Harmer, P. M., 120, IBS

Harper, J. H., 91, 94, 108, 110

Harris, F. S., 260, 270, 366, 883

Harris, H. C., 19, 75

Harris, K., 49, 50, 79

Harrison, G. J., 29, 77

Harrison, J., 283, 311

Hart, E. B., 283, 297, 312, 313

Hart, G. H., 380, 382

Hartwell, B. L., 120, 153

Hartwig, H. B., 295, 312

Harvey, W. A., 146, 147

Haseman, J. F., 183, 203

Haskell, G., 128, 153

Hastings, E. G., 282 (see Peterson), 283
(see Peterson), 284 (see Peterson),
288 (see Peterson), 290 (see Peter-
son), 314

Hatch, L. P,, 254, 270

Havis, J. R., 147, 148, 162, 153, 155

Hawkins, R. S,, 50, 79

Haynes, J. L., 260, 270

Hays, O. E., 99, 110

Hayward, H. E., 356, 383

Hazelwood, B. P., 107, 111

Hearn, W. E., 804

Hedges, T. R., 377, 378, 379, 383

Hedlin, W. A., 149, 153

Heggeness, H. G., 147, 153

Hegnauer, L., 210, 212, 231

Hegsted, D. M., 289 (see Johnson, B. C.),
S12

Hein, M. A., 290 (see Shepherd), 291
(see Shepherd), 296 (see Shepherd),
303 (see Shepherd), 304 (see Shep-
herd), 306 (see Shepherd), 307 (see
Shepherd), 308 (see Shepherd), 309
(see Shepherd), 814, 366, 884

Heineman, P. G., 283, 812

Heinton, T. E., 290 (see Shepherd), 296
(see Hodgson; Shepherd), 303 (see
Shepherd), 304 (see Shepherd), 306
(see Shepherd), 307 (see Shepherd),

308 (see Shepherd), 309 (see Shep-
herd), 818, 314

Heinzl, O., 285, 287, 290, 311

Hemphill, D. D., 139, 158

Hendricks, S. B., 87, 110, 166, 184, MS,
204

Hendrickson, A. H., 260, 261 (see Veih-
meyer), 272

Hendrix, A. T., 301, 312

Hendrix, J. W., 125, 153

Hendrix, T. M., 245, 270

Hennebury, G. 0., 253, 969

Henson, E. R., 295, 296, 299, 312

Henson, L., 214, 227, 231

Herbert, F. W., 29, 76

Hermans, P. H., 57, 79

Hernandez, T. P., 147, 155

Hertel, K. L., 59, 60, 79, 80

Hessler, L. E., 57, 79

Hewitt, C. W., 86, 110

Hibbard, A. D., 123, 168

Hide, J. C., 200, 203

Higgins, F. L., 295, 312

Hildebrandt, H., 283, 818

Hill, H. O., 28 (see Lyle), 76

Hilton, J. H., 290 (see Monroe), 813

Hinds, W. E., 33, 34, 78

Hinkle, D. A., 47, 79

Hixson, C. R., 283, 312

Hodgson, R. E., 280 (see Shepherd), 283
(see Shepherd), 290 (see Monroe;
Shepherd), 291, 296, 303 (see Shep-
herd), 304 (see Shepherd), 305, 306,
307, 308, 309, 812, 818, 314, 366, 388

Hoffman, A. H., 289, 312

Hoffman, E. J., 298, 299, 812

Hoffman, O., 293 (see Windheuser), 815

Holben, F. J., 95 (see White), 111

Holdeman, Q. L., 13 (see Brown, G. A.),

Holley, K. T., 21, 75, 96, 110
Hollowell, E. A., 209, 212, 213, 217, 220,

222, 231

Holmes, A. D., 139, 158
Holmes, F. O., 126, 153
Holmes, L. A., 330, 849
Holt, M. E., 76, 120, 168
Holtz, H. F., 84, 111
Hooghoudt, S. B., 266, 267, 270
Hooten, D. R., 24, 27, 76, 76

392

AUTHOR INDEX

Horton, E. A., 296 (see Watson), 815

Horton, R. E., 250, 271

Hosking, H. R., 31 (see Hansford), 76

Hosking, J. S., 117, 153

Hosterman, W. H., 290 (see Shepherd),
291 (see Shepherd), 296 (see Hodg-
son; Shepherd), 303 (see Shepherd),
304 (see Shepherd), 306 (see Shep-
herd), 307 (see Shepherd), 308 (see
Hodgson; Shepherd), 309 (see Shep-
herd), 312, 314

Howard, L. 0., 33, 78

Howell, M. J., 138, 155

Howlett, F. S., 139, 153

Hubbard, J. W., 29, 76

Huelson, W. W., 153, 158

Huffman, C. F., 294 (see Dexter), 303
(see Dexter), 312, 367 (see Cole),
368 (see Cole), 382

Hull, H. H., 99 (see Hays), 110

Hunter, A. S., 244, $71

Hunter, C. A., 283, 312

Hunter, 0. W., 282, 283, 312

Hunter, W. D., 33, 34, 37, 78

Hutcheson, J. D., 214, 281

Hutchins, A. E., 133, 153

Hutchinson, J. B., 3, 66, 80

Hutton, C. E., 177, 179, 182, 184, 185,
BOS

Hutton, E. M., 125, 153

Iglinsky, W., Jr., 40, 78

Ingham, I. M., 221, 222, BS1 9 359 (see

Law), 883

Ingham, R. W., 288, 312
Isbell, C. L., 141, 153
Israelson, 0. W., 260, 271
van Itallie, Th. B., 15, 75
Ivanhoff, S. S., 127, 153
Ivy, E. E., 35, 37 (see Moreland; Paren-

cia), 78
Iyer, K. R. N., 87, 111

Jacks, H., 31, 76

Jackson, H. C., 278 (see Guilbert), 812

Jackson, I. R. C., 278, 279, 811

Jackson, M. L., 117, 153, 185, 203

Jacobson, L., 15, 75

Jacobson, W. C., 279 (see Kane), 309
(see Kane), 313

James, L. H., 277 (see Phillips), SI 4

James, W. H., 279 (see Forbes; Swift),
312, 314

Jamieson, V. C., 241, 252, 268, 269, 271

Janert, 268

Jeffries, C. D., 181, 203

Jenkins, P. M., 24 (see Hooten), 75

Jennings, 301

Jenny, H., 83, 85, 86, 95, 99, 110, 117, 15S,
171, 172, 173, 182, 192, 204, 268, 271

Joham, H. E., 15, 18, 75

Johnson, B. C., 289, 291, 312

Johnston, H. G., 35, 78

Johnstone, W. F., 212, 281

Jolivette, J. P., 126, 155

Jones, B. J., 208, 231, 359, 360, 361, 362,
363, 369, 370, 376, 381, 883

Jones, C. H., 279 (see Camburn), 290
(see Camburn), 296 (see Camburn),
297 (see Camburn), 299 (see Cam-
burn), 305 (sec Camburn), 307 (see
Camburn; Newlander), 308 (see
Camburn; Newlander), 309 (see
Camburn; Newlander), 811 9 313

Jones, D. L., 47, 48, 49 (see Smith, H.
P.), 79

Jones, H. A., 122, 132, 152, 153

Jones, H. E., 94, 110, 202, BOS

Jones, I. R., 212, 226, 231 9 372, 381, 882,
383

Jones, R. J., 96, 110

Jones, T. N., 295, 312

Jones, W. L., 48, 79

Jordan, H. V., 24 (see Hooten), 27, 75, 76

Kahane, E., 289 (see Voinovitch), 815
Kalbfleish, W., 301, 302, 303, 813
Kane, E. A., 279, 297, 309, 318
Kardos, L. T., 97, 110
Kauffman, W., 38, 78
Kay, G. F., 177, 204
Keeney, L. G., 298, 818
Keirns, V. E., 148, 152

AUTHOR INDEX

393

Keith, T. B., 359 (see Klages), 364 (see

Klages), 367 (see Klages), 371 (see

Klages), 373 (see Klages), 380 (see

Klages), 381 (see Klages), 383
Keller, W., 359, 364, 365, 367, 375 (see

Bateman), 376 (see Bateman), 382,

383

Kelley, 0. J., 244, 245, 270, 271
Kelley, W. P., 86, 110
Kellner, O., 278, 313
Kellogg, C. E., 193 (see Baldwin), 195,

203

Kemp, W. B., 214, 231
Kennard, D. C., 226, 231
Kenney, R., 214, 227, 231, 296, 313
Kenworthy, A. L., 244, 971
Kerr, T., 25, 56, 57, 58 (see Berkley), 61,

74, 75, 79

Kidder, E. A., 268, 271
Kiesselbach, T. A., 295, 313
Kikuta, K., 125, 153
Killough, D. T., 49 (see Smith, H. P.),

Kime, P. H., 63, 70, 80
Kincer, 162, 174, 804
King, A. S., 116, 153
King, C. J., 25 (see Berkley), 27, 28, 31,

57 (see Berkley), 58 (see Berkley),

74, 76, 79

King, H. E., 72 (see Parnell), 80
King, W. A., 288, 290 (see Monroe), 313
Kirkham, D., 253, 258, 266, 267, 271
Kirsch, W., 283, 285, 313
Klages, K. H., 359, 364, 367, 371, 373,

380, 381, 383

Kleiber, M., 367 (see Cole), 368, 382
Knight, R. L., 68, 70, 72, 80
Knisely, A. L., 289, 313
Knodt, C. B., 286 (see Autrey), 311
Knott, J. C., 305 (see Hodgson), 312,

366 (see Hodgson), 383
Kolb, A. C., 376, 377, 378, 388
Konstantinov, N. N., 12, 75
Kozeny, J., 254, 271
Kraebel, C. J., 261, 271
Kraemer, E., 349
Kramer, P. J., 369, 383
Krantz, B. A., 18 (see Hall), 75, 102, 103,

107, 110
Kraus, W. E., 290 (see Monroe), 313

Kreitlow, K. W., 213, 215 (see Garber),

230, 231

Krusekopf, H. H., 97, 98, 111
Kuhn, A. O., 214, 831
Kulkarni, Y. S., 28 (see Uppal), 77
Kuntz, J. E., 125, 126 (see Walker), 155
Kuska, J. B., 83 (see Myers), 84 (see

Myers), 94 (see Myers), 95 (see

Myers), 108 (see Myers), 111
Kuzmeski, J. W., 139 (see Holmes, A.

D.), 153, 294 (see Archibald), 311
Kvasnikov, E. I., 289, 313

Lachman, W. H., 139 (see Holmes, A

D.), 149, 153
Lamb, A. R., 283, 313
Langmuir, I., 242, 271
Larsinos, G. J., 117, 152
Larson, L. H., 92, 109
Larson, R. E., 130, 131, 153
Larson, R. H., 126 (see Walker), 156
Larson, W. E., 185, 804
Latshaw, W. L., 96, 111
Law, A. G., 221, 222, 231
Law, G. A., 359, 383
Lawes, 81

Lea, F. M., 254, 271
Leach, L. D., 145, 153
Learner, R. W., 241, 271
LeClerc, J. A., 282, 288, 313
Lee, S. H., 141, 158
Leonard, C. D., 192, 804
Leonard, O. A., 25, 75
Lepard, O. L., 288, 313
Levy, E. B., 364, 366, 383
Lewis, A. C., 28, 76
Lewis, A. H., 373, 382, 383
Lewis, M. R., 268, 271
Lewis, M. T., 121, 153
Lewis, W., 275, 313
Liebig, 81

Limstrom, G. A., 323, 330, 336, 349
Linn, M. B., 144, 153
Lipman, J. G., 91, 92, 93, 104, 110
Livingston, E. M., 33, 77
Loftin, U. C., 37, 78
Loomis, W. E., 168, 804
Loosli, 303

394

AUTHOR INDEX

Lorenz, 0. A., 117, 118, 15S

Lorenzen, C., 151, 164

Love, R. M., 381, 883

Lovvorn, R. L., 209, 213, 214, 217, 218,

220, 221, 222, 225, 231, 302 (see

Weaver), 815

Lukaczewicz, J., 285 (see Kirsch), 318
Lundegardh, H., 15, 75
Luthin, J. N., 267, 271
Lutz, J. F., 241, 271
Lyford, W., 804
Lyle, E. W., 25 (see Eaton), 27 (see

Jordan), 28, 75, 76
Lynn, H. D., 72, 80
Lyon, T. L., 91, 92, 93, 105, 110

MacArthur, J. W., 124, 155

McAuliffe, C., 215, 230

McAuliffe, H. D., 283 (see Stone), 286

(see Stone), 314
McCall, A. G., 98 (see Borst), 100 (see

Smith, D. D.), 109, 111
McCalla, T. M., 268, 271
McCalmont, J. R., 280 (see Shepherd),

283 (see Shepherd), 814
McCarthy, G., 208, 231
McClelland, B., 266, 271
McCollam, M. E., 116, 154
McComb, A. L., 168, 804
McCool, M. M., 117, 164
MacDonald, 276
McDougall, W. B., 328, 949
McDowell, C. H., 49, 79
McGarr, R. L., 34, 77
McGeorge, W. T., 108, 110, 261, 869
MacGillivray, J. H., 142, 146, 168, 154
McGregor, E. A., 40, 78
Mclntosh, 330

McKaig, N., 17 (see Skinner), 76
McKinney, K. B., 37 (see Loftin), 78
McLane, J. V., 260, 869
McLendon, C. A., 28, 76
McMiller, P. R., 196, 804
McNamara, H. C., 27, 76
McRuer, W. G., 84, 111
Madhok, M. R., 90, 111
Madson, B. A., 208, 209, 212, 229, 881,

381, 888

Magee, A. C., 47, 79

Magistad, O. C., 262, 871 9 354, 356, 888

Magruder, J. W., 214, 881

Magruder, R., 122, 154

Maloney, M. M., 328, 849

Malzahn, R. C., 283 (see Stone), 286
(see Stone), 314

Mann, L. K., 138, 146, 154

Manning, C. W., 68

Manning, H. L., 66, 80

Manns, M. M., 19 (see Manns, T. F.), 75

Manns, T. F., 19, 75

Marbut, C. F., 158, 164, 165, 166, 171,
182, 185, 192, 194, 804

Marchant, W. L., 130

Marcum, W. B,, 245 (see Edlefson), 270

Markle, J., 18, 74

Marsh, P. B., 57 (see Barre), 79

Marshall, C. E., 183, 185, 203, 205

Marth, P. C., 140, 166

Martin, J. A., 131, 164

Martin, J. P., 97, 110

Martos, V. F., 284, 286, 313

Maskell, 23

Mason, C. J., 282, 818

Mason, T. G., 18, 20, 23, 75, 76

Massey, R. E., 31, 76

Mather, 330

Mathews, E. D., 17, 75

Matlock, R. L., 363, 883

Matrone, G., 279 (see Ellis, G. E.), 818

Mayeux, H. S., 35 (see Becnel), 77

Maynard, L. A., 279, 288 (see Lepard),
811, 312, 813

Mead, R. M., 229, 232

Mead, S. W., 278 (see Guilbert), 818,
368, 888, 883

Medeck, C. H., 208, 881

Meek, W. E., 43

Mehring, A. L., 82, 104, 105, 110

Melin, C. G., 285 (see Shepherd), 290
(see Shepherd), 291 (see Shepherd),
296 (see Shepherd), 303 (see Shep-
herd), 304 (see Shepherd), 306 (see
Shepherd), 307 (see Shepherd), 308
(see Shepherd), 309 (see Shepherd),
314

Mendel, G., 64

Merola, G. V., 57 (see Hessler), 79

Merz. A. R., 116, 164

AUTHOR INDEX

395

Metzger, W. H., 200, 208

Meyer, L., 338, 849

Mick, A. H., 245, 269

Midgley, A. R., 215, 281

Mikhin, A. M., 288, 818

Miles, L. E., 29, 77

Miller, D. D., 280, 818

Miller, E. V., 135, 154

Miller, M. D., 212, 219, 220, 221, 222, 226,
231, 359 (see Jones, B. J.), 360 (see
Jones, B. J.), 361 (see Jones, B. J.),
362 (see Jones, B. J.), 363 (see Jones,
B. J.), 369 (see Jones, B. J.), 370
(see Jones, B. J.), 376 (see Jones,
B. J.), 378, 381 (Bee Jones, B. J.),
888, 884

Miller, M. F., 97, 98, 111

Miller, R. C., 298, 818

Mills, R. C., 297, 818

Milville, J., 212, 231

Miner, B. B., 217, 282

Minges, P. A., 138, 154

Mitchell, J. H., 16 (see Cooper), 75

Mitchell, J. W., 137, 154

Mitchell, K. J., 121, 154

Mitchell, N., 300 (see Schaller), 814

Mohr, E. C. J., 85, 111

Monroe, C. F., 290, 818

Monson, O. W., 363, 882

Mooers, C. A., 107, 111

Moore, L. A., 279 (see Kane), 290 (see
Shepherd), 291 (see Shepherd), 296
(see Shepherd), 303 (see Shepherd),
304 (see Shepherd), 306 (see Shep-
herd), 307 (see Shepherd), 308 (see
Shepherd), 309 (see Moore; Shep-
herd), 818, 814

Moore, R. E., 250, 252, 261, 271

Moore, W. D., 135 (see Miller), 136,
154, 155

Moran, J., 122

Moreland, R. W., 37, 78

Morgan, A., 351, 355, 356, 360, 370 (see
Bartels), 371 (see Bartels), 372, 373,
882, 888

Morrish, R. H., 295, 814

Morrison, F. B., 280, 281, 290, 818

Morse, R. W., 381, 882

Moser, F., 96, 111

Moskovitz, I., 306, 818

Muckenhirn, R. J., 196, 201 (see Stauf-

fer), 204, 239 (see Alexander), 269
Mullen, L. A., 364, 384
Mullison, E., 138, 154
Mullison, W. R., 138, 154
Mumford, D. C., 381, 883
Munger, H. M., 129, 134, 154
Munsell, R. I., 211, 213, 214, 215, 217,

224, 225, 231
Murdock, F. R., 283 (see Stone), 286

(see Stone), 814

Murer, H. K., 305 (see Hodgson), 312
Murneek, A. E., 138, 140, 154, 155
Musgrave, G. W., 268, 271
Musgrave, R. B., 277, 299, 301, 303, 311 9

313

Muskat, M., 247, 271
Myers, H. E., 83, 84, 88, 94, 95, 108, 111
Myers, W. M., 213 (see Garber), 215

(see Garber), 231

Narayanan, N. G., 71, 80

Neal, D. C., 27, 77

Neely, J. W., 43, 79

Neidig, R. E., 283, 288, 313

Nelson, A. H., 27 (see Jordan), 76

Nelson, M. L., 57, 79

Nelson, R. A., 184, 203

Nelson, W. H., 18 (see Hall), 75

Nelson, W. L., 24, 75

Nelson, W. R., 255, 271

Newcomb, G. T., 116, 153

Newell, R. E., 229, 232

Newell, W., 34, 78

Newhall, A. G., 144, 153, 154

Newlander, J. A., 307, 308, 309, 811, 818

Newton, J. D., 94 (see Brown, A. L.),

109, 172, 204
Nightingale, G. T., 22, 75
Nisman, B., 283, 314
Nixon, P. P., 148, 154
Noll, C. J., 132, 134, 154
Norman, A. G., 87, 89, 101, 109, 111
Norris, D. 0., 126, 154
Norton, E. A., 182, 204
Nurse, R. W., 254, 271

396

AUTHOB INDEX

Odell, R. T., 175, 197, 198, 201 (see

Stauffer), 804

Odland, M. L., 132, 134, 164
Ohlendorf, W., 37, 78
Ohlmer, E., 293 (see Windheuser), 315
Olsen, T. M., 367 (see Cole), 368 (see

Cole), 888

Olson, L. C., 19, 20, 76
O'Neal, A. M., 804
Orton, W. A., 26, 28, 77
Overbeek, J. Th. G., 242, 272
Overholser, E. L., 289, 313
Overstreet, R., 15, 75
Owen, W. L., 33 (see Fletcher), 77
Owen, W. L., Jr., 19 (see Gaines), 75
Owens, J. S., 214, 217, 220, 221, 225, 232

Packer, J. E., 359 (see Bateman; Kel-
ler), 364 (see Keller), 365 (see Bate-
man; Keller), 367 (see Bateman;
Keller), 375, 376 (see Bateman),
378, 379, 382, 383

Paddock, E. F., 139, 154

Paden, W. R., 15 (see Cooper), 75

Padilla-Saravia, B., 85, 110

Page, N. R., 15 (see Cooper), 16 (see
Cooper), 75

Page, E., 288 (see Lepard), SIS

Painter, R. H., 34, 78

Palmer, V. J., 268, 271

Palmiter, D. H., 118, 168

Pammel, L. H., 26, 27, 77

Panse, V. G., 66, 80

Parberry, N. H., 89, 111

Parencia, C. R., Jr., 35, 37, 77, 78

Park, J. B., 129, 154

Parker, E. R., 268, 871

Parker, F. W., 81, 111

Parks, R. Q., 104, 105, 110

Parnell, H. E., 72, 80

Parris, G. K., 123, 164

Paur, 8., 131, 163

Peak, A. R., 125, 158

Pearson, N. L., 57 (see Barre), 61, 79

Pearson, R. W., 164, 804

Peevy, W. J., 89, 111, 201, 804

Pennington, R. P., 185 (see Jackson),

808

Pentzer, W. T., 293, 313
Peterson, C. E., 122, 154
Peterson, H. B., 352, 353, 384
Peterson, J. B., 166, 171, 184, 804, 241,

272
Peterson, W. H., 282, 283, 284, 288, 289

(see Johnson, B. C.), 290, 291, 818 f

313, 314
PfeifYenberger, G. W., 25 (see Eaton),

60, 75, 79

Phillips, C. E., 212, 888
Phillips, M, 277, SI 4
Phillis, E., 18, 20, 75, 76
Pickett, J. E., 208, 838
Pierce, W. D., 34, 78
Pierre, W. H., 82, 107, 111, 164, 204
Pillsbury, A. F., 249, 871
Pinck, L. A., 88 (see AJlison, F. E.), 93

(see Allison, F. E.), 97, 103, 109, 111
Pinckard, J. A., 25, 75
Pittman, D. W., 353, 372, 883
Pohlman, G. G., 248 (see Smith, R. M.),

255 (see Smith, R. M.), 272
Pope, H. W, 17, 74
Pope, O. A., 62, 79
Porter, D. D., 24 (see Hooten), 27 (see

Jordan), 75, 76
Post, A. H., 363, 383
Pound, G. S, 125, 155
Powers, W. L., 129, 154, 261, 268, 871
Presley, J. T v 29, 31, 77
Pressley, E. H., 59, 79
Price, W. C., 213, 881
Prince, A. L., 102, 109
Princo, F. S., 214, 232
Procopio, M., 289, 314
Pryor, D. E., 127, 154

Quaintance, A. L., 36, 78

Raev, Z. A., 289, 813
Rahn, E. M., 119, 164
Rainwater, C. F., 35, 78
Raleigh, G. J., 120, 164

AUTHOR INDEX

397

Rampton, H. H., 359, 883

Ranadive, J. D., 28 (see Uppal), 77

Randall, T. E., 135, 154

Randhawa, G. S., 140, 164

Rasmussen, R. A., 288 (see Lepard), SIS

Rather, H. C., 295, 314

Rayner, G. B., 360, 372, 373, 376, 383

Ree, W. 0., 268, 271

Reed, A. D., 376, 378, 383

Reed, G. M., 281, 314

Reed, O. E., 285, 314

Reeve, P. A., 144, 145, 146, 155

Regan, W. M., 368 (see Cole; Mead),

882, 383

Reirnann, E. G., 262, 271
Reiner, J. M., 16, 76
Reinhard, H. J., 33, 78
Reitemeier, R. F., 240, 262, 271
Rhoades, H. F., 185 (see Larson), 204
Rich, L. H., 376, 383
Richards, L. A., 239, 240, 242, 244, 250,

252, 253, 261, 262, 271, 272, 352, 354,

355, 356, 383
Richards, S. J., 242, 271
Richardson, H. B., 59, 60, 79
Richer, A. C., 95 (see White), 111
Richey, F. D., 68, 80
Richmond, C. A., 33 (see Chapman), 77
Richmond, T. R., 65, 68, 80
Rick, C. M., 130, 131, 135, 154
Riecken, F. F, 161, 177, 178, 194, 195,

196, 204, 268 (see Wilson), 272
Rigden, P. J., 254, 271
Rigler, N. E., 12, 20, 21, 22, 75, 76
Riker, A. J., 140, 153
Riley, C. V., 36, 18, 328, 330, 349
Riner, M. E., 133, 155
Riollano, A., 127, 154
Ripley, P. 0., 301 (see Kalbfleish), 302

(see Kalbfleish), 303 (see Kalb-
fleish), SIS

Roach, J. R., 128 (see Doty), 152
Roberts, R. C., 204
Robertson, D. W., 363, 368, 369, 371, 378,

380, 384

Robinson, D. H., 338, 349
Robinson, R. R., 213 (see Garber), 215

(see Garber), 2S1
Rodenkirchen, J., 285, 291, SI 4
Rodriquez, A., 127, 164

Roethe, H. E., 298, 299, 314

Roever, W. E., 131, 154

Rogers, C. F., 290, 314

Rogers, C. H., 27 (see Jordan), 76

Rogers, H. T., 97, 111

Rogers, R. H., 49, 79

Rogers, W. S., 242, 244, 261, 271

Rolfs, F. M., 30, 31, 77

Rosenberg, A. J., 283, 314

Rosenstein, L., 116, 154

Roseveare, G. M., 286, 314

Ross, C. S., 166, 204

Rost, C. 0., 98, 111, 199, 204

Rouse, J. T., 25 (see Eaton), 49 (see

Smith, H. P.), 75, 79
Roussel, J. S., 35 (see Becnel), 77
Rowles, C. A., 199, 204
Rudolph, B. A., 29, 77
Runiler, R. H., 212, 232
Russell, E. J., Ill, 281, 282, 283, 284, 314
Russell, H. L., 282, 311
Russell, J. C., 84, 96, 111
Russell, J. L., 264, 267, 268, 271
Russell, M. B., 166, 184, 204, 239, 241,

244, 245, 272

Russell, W. C., 290 (see Taylor), 314
Ruston, D. F., 72 (see Parnell), 80
deRuyter de Wildt, J. C., 291 (see

Brouwer), 311

Sackman, R. F., 364 (see Schwendiman),

384

Salter, R. M., 29, 77, 95, 109, 111
Saltonstall, L., 279, 314
Samarani, F., 288, 314
Sanborn, J. W., 358, 384
Savage, E. S., 288 (see Lepard), 313
Sawyer, F. G., 116, 117, 164
Sawyer, L. E., 330, 341, 348, 349
Sayre, C. B., 15, 76, 119, 120, 154, 155
Schaffer, R. V., 244, 272
Schalk, A. F., 367 (see Cole), 368 (see

Cole), S82

Schaller, J. A., 300, 314
Schavilje, J. P., 330, 349
Schieblich, M., 284, 314
Schmidt, K, 288, 289, 314

398

AUTHOR INDEX

Schoenleber, L. G., 290 (see Shepherd),
291 (see Shepherd), 296 (see Shep-
herd), 303 (see Shepherd), 304 (see
Shepherd), 306 (see Shepherd), 307
(see Shepherd), 308 (see Hodgson;
Shepherd), 309 (see Shepherd), 812,
314

Schofield, R. K., 239, 240, 242, 243, 257.
270, 272

Scholl, W., 105, 111

Scholtes, W. H., 171, 804

Schomer, H. A., 135 (see Miller), 154

Schoth, H. A., 209, 211, 219, 220, 221, 222,
225, 232, 367, 368, 372, 373, 384

Schupp, A. A., 144, 145, 146, 155

Schwendiman, J. L., 364, 884

Scott, D. H., 133, 155

Seidel, C., 287 (see Gcrlach), 313

Semple, A, T., 366, 384

Serviss, G. H., 222, 232

Sewell, M. C, 90, 111

Shafer, J. T., 120, 154

Shaw, B., 245, 272

Shaw, R. S., 216, 232

Shaw, T. M., 239 (see Alexander), 869

Shedd, C. K., 302, 314

Sheldon, W. H., 227, 228, 232, 294 (see
Dexter), 303 (see Dexter), 312

Shepherd, J. B., 280, 283,, 285, 286, 290,
291, 296, 298, 299, 303, 304, 306, 307,
308, 309, 312, 814, 315

Sherbakoff, C. D., 28, 29, 77, 123, 153

Sherman, J. M., 282, 814

Sherman, M. S., 88 (see Allison, F. E.),
109

Shifriss, O., 129, 133, 134, 155

Shirlow, N. S., 121, 124, 1Z5

Shrader, W. D., 167, 196, 804

Shultis, A., 378, 384

Shutt, F. T., 94, 96, 111

Sievers, F. J., 84, 111

Silow, R. A., 3 (see Hutchinson), 80

Silver, E. A., 298 (see Miller, R. C.), 813

Simonson, R. W., 163, 188

Simpson, D. M., 31, 63, 71, 77, 80

Simpson, R., 33, 78

Sinclair, J. D., 261, 871

Singh, S., 13, 76

Singleton, H. P-, 359 (see Law), 888

Sinrieton, W. R., 128, IBS. 155

Sinks, A. H., 318, 330, 849

Sisakyan, N. M., 289, 814

Sisam, J. W. B., 338, 849

Sisson, W. A., 58, 80

Skinner, J. J., 17, 76

Skirm, G. W., 125, 155

Skovsted, A., 4, 80

Slater, C. S., 98, 111, 245, 252, 255, 268,

269, 270, 272
Smith, A. L., 29, 31, 77
Smith, C. D., 34, 78
Smith, C, T., 139 (see Holmes, A. D.),

153

Smith, D., 211, 220, 888
Smith, D. C., 227, 281
Smith, D. D., 100, 111
Smith, E. F., 30, 77
Smith, F. B, 201, 804
Smith, G. D., 160, 166, 172, 177, 178, 182,

183, 188, 194, 195, 196, 804
Smith, G. E., 101, 111, 148, 154
Smith, G. L., 33 (see Gaines, R. C.), 78
Smith, G. M., 128 (see Doty), 152
Smith, H. P., 48, 49, 79
Smith, H. V., 97, 109, 111
Smith, H. W, 91, 111
Smith, L. H., 201 (see DeTurk), 203,

217, 888

Smith, O., 141, 155
Smith, P. G., 124, 126, 127, 136, 155
Smith, R. M., 248, 249, 253, 255, 272,

321, 326, 330, 339 (see Tyner), 349
Smith, R. S., 199, 201, 205
Smith, V. F., 279 (see Forbes), 312
Smith, W. O., 242, 870
Snyder, H., 94, 96, 111, 226, 888
Snyder, H. J., 212, 232
Somner, A. L., 19, 76
Sotola, J., 276, 278, 814
Southworth, W. L., 363, 884
Sparhawk, W. N., 849
Spiegelman, S., 16, 76
Sprague, G. F., 68, 80
Sprague, H. B., 209, 230
Sprague, M. A., 216, 217, 220, 221, 224,

225, 226, 229, 232
Sprague, V. G., 213 (see Garber), 215

(see Garber), 231
Springer, M. E., 176, 804
Springer, V., 88, 111

AUTHOR INDEX

399

Spry, R., 164, 204

Stacy, S. V., 96 (see Holley), 110

Stallworth, H., 138, 155

Stark, R. H., 359 (see Klages), 364 (see
Klages), 367 (see Klages), 371 (see
Klages), 373 (see Klages), 380 (see
Klages), 381 (see Klages), 383

Starke, J. S., 366, 368, 380, 384

Staten, G., 13, 47, 76, 79, 368

Stauffer, R. E., 268 (see Kidder), 271

Stauffer, R. S., 175, 180, 193, 201, 204,
262 (see Reimann), 271

Steinberg, R. A., 121, 155

Stelly, M., 164, 204

Stephens, S. G., 3 (see Hutchinson), 68,
69, 80

Sterges, A. J., 90, 110

Sterling, L. D., 90, 109

Stevenson, W. A., 38, 78

Stewart, G., 360, 380, 384

Stiver, E. N., 321, 323, 324, 328, 329, 330,
336, 348, 349

Stoddart, L. A., 381, 884

Stone, R. W., 283, 286, 314

Storgaard, L. H., 208, 232

Stoughton, R. H., 31, 76, 77

Strait, J., 302, 314

Streets, R. B., 27, 30, 77

Strong, D. G., 262 (see Wadleigh), 272

Sturkie, D. G., 24, 76

Sullivan, J. T., 128 (see Doty), 152, 213
(see Garber), 215 (see Garber), 231

Sullivan, R. R., 60, 80

Sullivan, W., 375, 378, 382, 384

Swaby, R. J., 89, 111

Swfcnson, C. L. W., 241, 272

Swanson, C. 0., 96, 111

Sweet, R. D., 147, 148, 162, 155

Sweetman, W. J., 290 (see Shepherd),
291 (see Shepherd), 296 (see Shep-
herd), 303 (see Shepherd), 304 (see
Shepherd), 306 (see Shepherd), 307
(see Shepherd), 308 (see Shepherd),
309 (see Shepherd), 314

Swift, R. W., 279, 312, 314

Talbot, R. F., 217, 232
Taubenhaus, J. J., 27, 77

Taylor, C. A., 142, 155

Taylor, F. J., 146, 155

Taylor, M. W., 290, 314

Templin, E. H., 204

Tenney, F. G., 89, 111

Terry, C. W., 295, 300, 314

Terzaghi, C., 254, 272

Tesche, W. C., 208, 232

Thacker, E. J., 279 (see Swift), 314

Tharp, W. H., 25 (see Eaton), 75

't Hart, M. L., 366, 384

Thatcher, L. E., 213, 214, 216, 219, 220,
221, 222, 223, 224, 226, 227, 229, 231,
232

Thibodeaux, B. H., 45, 47 (see Magee),
78, 79

Thomas, H. R., 136, 155

Thomas, W. I., 46, 79

Thompson, H. C., 137, 140, 154

Thompson, R. C., 122, 155

Thome, C. E., 100, 111

Thorne, D. W., 352, 353, 384

Thome, M. D., 245, 272

Thornton, S. F., 331, 349

Thorp, J., 160, 171, 193 (see Baldwin),
195, 196, 203, 204

Tidmore, J. W., 96, 111

Tilley, R. H., 63, 70, 80

Toenges, A. L., 330, 349

Toth, S. J., 15 (see Wallace), 76

Tower, H. E., 353 (see Hamilton), 355
(see Hamilton), 360 (see Hamilton),
362 (see Hamilton), 363 (see Hamil-
ton), 364 (see Hamilton), 366 (see
Hamilton), 367 (see Hamilton), 371
(see Hamilton), 372 (see Hamilton),
374 (see Hamilton), 380 (see Hamil-
ton), 381 (see Hamilton), 383

Tretsven, J. O., 363, 383

Tucker, R. H., 363 (see Robertson), 368
(see Robertson), 369 (see Robert-
son), 371 (see Robertson), 380 (see
Robertson), 384

Tupikova, A. A., 288 (see Mikhin), 313

Turk, 303, 307, 309

Turner, J. H., 17, 76

Turpin, H. W., 260, 270

Twist, G. F., 150, 155

Tyler, S. A., 185 (see Jackson), 203

Tyner, E. H., 321, 326, 330, 339, 349

400

AUTHOR INDEX

Tysdal, H. M., 290 (see Shepherd), 291
(see Shepherd), 296 (see Shepherd),
303 (see Shepherd), 304 (see Shep-
herd), 306 (see Shepherd), 307 (see
Shepherd), 308 (see Hodgson; Shep-
herd), 309 (see Shepherd), 81$, 314

Tvulin, A. T., 88, 111

TThland, R. E., 97, 98, 111
Ulrich, R., 199, 204
Uppal, B. N., 28, 77

Van Alstine, E., 212, 217, 232

Vandpcavpve, S. C., 91, 111

Vandermark, J. S., 141, 152

Vander Meulen, E., 232

Van Doren, C. A., 262 (soe Roimann),

268 (see Kidder), 271
Vasil'eva, N. A., 289, 314
Vaughan, E. K., 135 (seo Miller), 154
Vavilov, N. I., 63, 80
Veihmeyer, F. J., 260, 261, $72, 362, 371,

384

Verwey, E. J. W., 242, 272
Vinall, H. N, 366 (see Semple), 384
Virtanen, A. I., 284, 287, 288, 314, 315
Vittum, M. T., 15, 76', 120, 155
Vogelsang, P., 144, 145, 146, 155
Voinovitch, I., 289, 315
Volk, N. J., 16, 75, 76, 96, 111, 120, 153

Wade, B. L., 124, 155

Wadleigh, C. H., 14, 20, 21, 22, 23, 24,
76, 261, 272, 356, 383

Wagner, R. E., 290 (see Shepherd), 291
(see Shepherd), 296 (see Shepherd),
303 (see Shepherd), 304 (see Shep-
herd), 306 (see Shepherd), 307 (see
Shepherd), 308 (see Hodgson; Shep-
herd), 309 (see Shepherd), 312, 314

Waksman, S. A., 87, 89, 90, 111, 285, 315

Walker, J. C., 125, 126, 155

Walker, R. H., 175, 205

Wallace, A., 15, 76

Wallace, H. M., 105, 111

Wallace, J., 244 (see Schaffer), 272

Wallihan, E. F., 245, 272

Wallis, G. C., 297, 315

Walter, G. H., 343, 349

Ward, A. S., 172, 204

Ward, J. W., 150

Wardle, R. A., 33, ?S

Warn, J. O., 57 (see Barre), 68, 79

Waring, K. J., 121, 155

Warington, K., 121, 155

Warren, G. F., 147, 149, 165

Wascher, H., 175, 205

Watkins, W. L, 196 (see Thorp), 204

Watson, J. R., 31, 77

Watson, S. J., 275, 278, 280, 281, 284,
285, 286, 287, 288, 289, 290, 291, 292,
295, 296, 297, 298, 299, 305, 315, 373
(SOP Ferguson), 382

Watts, J. W., 33, 35, 78

Watts, V, 127, 155

Weaver, J. W., Jr., 299, 301, 302, 304, 315

Weaver, L. R., 261, 262, 271

Webb, H. J., 288, 315

Webb, R. W., 59, 80

Wedernikov, V. V., 265, 272

Wcidinger, A., 57, 79

Weihing, R. M., 363 (see Robertson),
368 (see Robertson), 369 (see Rob-
ertson), 371 (see Robertson), 380
(see Robertson), 384

Weindling, R., 31, 77

Weir, W. W., 266, 272

Welch, C. D., 18 (see Hall), 75

Welch, J..E., 132, 166

Welch, J. S., 358, 359, 366, 384

Went, F. W., 136, 137, 138, 155

Wester, R. E., 27 (see Neal), 77, 122,
140, 164, 155

Wheeting, L. C., 84, 110, 111

Whisler, P. A., 302, S16

Whitaker, T. W., 121, 127, 134, 162, 154

White, J. W., 95, 111

Whiteside, E. P., 185, 199, 201, 205

Whiting, A. L., 90, 111

Whitman, E. W., 359 (see Klages), 364
(see Klages), 367 (see Klages), 371
(see Klages), 373 (see Klages), 380
(see Klages), 381 (see Klages), 383

Whitney, D. J., 212, 838

AUTHOR INDEX

401

Whitt, D. M., 100 (see Smith, D. D.),
Ill

Whyte, R. O., 338, 849

Wiegner, G., 278, 296, 815

Willaman, J. J., 283, 312

Willard, C. J., 213, 214, 216, 219, 220,
221, 222, 223, 224, 227, 229, 232, 298
(see Miller, R. C.), 318

Williams, B. H., 196 (see Thorp), 204

Williams, P. S., 286 (see Autrey), 311

Williams, R. D., 212, 232

Williamson, M. N., 49, 79

Willis, A. L., 185 (see Jackson), 203

Willis, L. G., 19, 76

Wilson, B. D., 93, 110

Wilson, H. A., 268, 272

Wilson, J. K., 284, 285, 286, 288, 315

Wilson, R. D., 121, 155

Winchell, J. H., 330, 849

Windhcuser, C., 293, 311, 315

WinriRht, G. L., 384

Winters, E., 175, 205

Wipprecht, R., 19 (see Gaines), 33 (see
Gaincs, J. C.), 37 (see Gaines, J. C.),
75, 78

Wiseman, H. G., 290 (see Shepherd),
291 (see Shepherd), 296 (see Shep-
herd), 297 (see Kane), 303 (see
Shepherd), 304 (see Shepherd), 306
(see Shepherd), 307 (see Shepherd),
308 (see Shepherd), 309 (see Shep-
herd), 813, 314

Wittwer, S. H., 138, 140, 155

Wood, A. D., 229, 282

Wood, J. K., 266, 270

Woodman, H. E., 276, 280, 281, 284, 290,
811, 315

Woodward, 0. C., 25 (see Berkley), 57
(see Berkley), 58 (see Berkley), 74,
79

Woodward, T. E., 276, 285 (see Shep-
herd), 286, 298, 299, 814, 815 9 366
(see Sample), 384

Work, R. A., 262, 269

Wright, N. C., 277, 297, 305, 315

Wyatt, F. A., 94 (see Brown, A. L.), 109

Wylie, C. E., 301, 304, 815

Yamell, S. H., 125, 155

Yates, F., 31 (see Hansford), 76

Yeager, A. F., 150, 155

Yernm, E. W., 277, 815

Young, 318

Young, M. T., 33 (sec Gaines, R. C.),

35, 78

Young, P. A. 124, 155
Younge, 0. R., 19, 76

Zaumeyer, W. J., 123, 155

Zervigon, M. G., 59, 79

Zingg, A. W., 100 (see Smith, D. D.), Ill

Zink, F. J., 295, 815

Zink, F. W., 147, 149, 155

Zunker, F., 254, 272

Subject Index

Agrostis alba, 218, 353

Alfalfa, 106, 336, 363, 359, 364, 368, 370,

371

digestibility, 276
Allium cepa, 122, 132
Alluvial soils, 194
Alopecurus pratensis, 359
Alsike clover, 353, 359
Alta fescue, 218, 219
Ammonia, anhydrous, 42, 116-117
Andropogan furcatm, 162, 168, 171
Anthonomus grandis, 34-36
Aphis gossypii, 33
Asparagus, 135, 148
Asparagus officinalis, 135
Austrian winter peas, 101
Azotobacter, 93
Azotogen, 93

Barn driers, 300-301
Barnyard manure, 100-101

production in U.S., 105
Beans, green, 119, 123, 140

snap, 119, 124, 140, 148, 150
Beets, 119, 120, 141, 148, 149, 150
Bentonite, 88

Benzene hexachloride, 35, 37, 38, 39
Bermuda grass, 219
Beta ctcZa, 93
Betula nigra, 323

Birdsfoot trefoil, 219, 336, 353, 368, 370
Bloat, 367-368

Brassica oleracea, 126, 134-135
Brassica oleracea acephala, 93
Brassica pekinenm, 122
Broadleaf trefoil, 358
Broccoli, 120

Bromegrass, 219, 225, 337, 353, 356, 358,
359

composition, 276
Bromus inermis, 217, 276, 337, 353

402

Brunigra soils, 203
Bur clover, 356

Butyric acid, 280, 281, 282, 283, 285, 286,
291

Cabbage, 119, 126, 134-135

Cabbage yellows, 126

Cantaloupe, 122, 127, 134

Capsicum jrutescens, 127, 131-132

Carotene, 279, 297, 309

Carrot, 132, 141, 143, 144, 146, 148, 150

Cauliflower, 120, 121, 137, 141

Celery, 115, 119, 120, 142

Cercospora zebrina, 213

Chernozems, 82, 172, 193, 194, 197

Chinese cabbage, 122

Chloris gayana, 356

p-chlorophenoxyacetic acid, 138, 139, 140

a-o-chlorophenoxypropionic acid, 141

Citrullus vulgaris, 123

Claypan formation, 183-184

Coal mine spoils, 317-349

acidity, 323-324

erosion, 321-322

establishment of pastures, 334, 336-338

graded, 333-335

grading, 334, 338-341

land-use capability, 332-335

legislation, 344

physical condition, 322-323

plant nutrient status, 324

presence of sul fides, 331

profile development, 325-326

reclamation costs, 341-343

reforestation, 329-330, 335-336

research needs, 346

sulfuric acid production, 323-326

topography, 319-322

ungraded, 332-333

volunteer vegetation, 328-329
Coal, strip mining, 318
Cotton, 2-74, 101, 106

aphis injury, 33

SUBJECT INDEX

403

bacterial blight, 30-31

boll weevil, 32, 34-36

bollworm, 36-37

boron requirements, 19

breeding systems, 66-70

competitive position among fibers,
5-10

defoliation, 44, 48-49

diseases, 26-32

drought effects, 24-25

end-uses, 6-7

fertilization, 41-42, 47

fiber development, 56-58

fiber fineness, 60-62

fiber length, 58-59

fiber properties, 56-63

fiber strength, 25, 59-60

flame cultivation, 42-43

fleahopper injury, 33-34

floral initiation, 11

Fusarium wilt, 28

ginning practices, 50-55

hormone responses, 13-14

hybrid vigor, 70-71

improvement, 63-74

injury by 2,4-D, 13

insecticide recommendations, 39, 43-44

insect pests, 32-40

irrigation, 49-50

leafworm, 38

linters, 2

mechanical pickers, 3, 44-45, 49, 73

mineral nutrition, 14-23

nitrogen metabolism, 20-23

phosphorus requirements, 17-18

photoperiodism, 12

physiology, 11-25

pink bollworm, 37-38

production in U.S., 3, 4

production practices, 40-50

research program, 9-10

root knot, 31-32

root rot, 26-28

seed treatment, 26

sodium requirements, 15-17

thrip injury, 32-33

Verticillium wilt, 29-30
}over crops, 102-104
>imson clover, 96
Cucumber, 122, 124, 134

Cucumia melo, 122, 127, 134
Cucumis sativus, 122, 134
Cucurbita maxima, 133
Cucurbita pepo, 133

D.D.T., 35, 37, 38, 39

2,4-dichlorophenoxyacetic acid (2,4-D) ,

13, 138, 147-148

Dactylis glomerata, 217, 337, 358
Dallis grass, 218-219, 356
Daucus carota, 132
Drainage, 258-259, 262-267

Egg plant, 119, 132-133
Erosion, 97, 98-100

Fayette silt loam, 170

Feldspars, 185, 191

Fertilizers, irrigated pastures, 371-373

Ladino clover, 214-217

vegetable crops, 116-120
Fertilizer placement, 118
Festuca elatior, 218, 337, 354
Festuca rubra, 219
Forage crops, artificial drying, 304-306

handling losses, 306-309

preservation and storage, 273-315

salt tolerance, 355, 356
Forages, lignin content, 279
Fusarium vasinfectum, 28-29

Glycine soja, 96
Gossypium arboreum, 3, 69
Gossypium barbadense, 3, 26; 28, 30, 64,

Gossypium hirsutum, 3, 63, 69, 70
Gossypium herbaceum, 69
Gossypium thurberi, 69, 72
Grass juice factor, 291
Green manure, 101-104
Grey-brown Podzolic soils, 168, 169, 171,

172, 190, 194, 197

404

SUBJECT INDEX

Grey-wooded soils, 172
Ground water, 258, 267

Hairy vetch, 93

Hay, barn drying, 299-304

field-cured, 294-299

field losses, 296

heating, 297-299

leaching losses, 277-278

losses in barn curing, 303-304

respiration losses, 276-277, 296

storage losses, 297-299

swath-curing, 295

time to cut, 295

windrowing, 295
Helio this armigera, 36-37
Herbicides, 146-149

oils, 148

selective, 146
Heleroda marioni, 31-32
Heterosis, 128
Hydrous mica, 88

Illite, 166, 175, 185

Infiltration, 262

Infiltration rate, 268

Interceptor drains, 263,

Irrigated pastures, 351-381
carrying capacity, 375-376
establishment, 360-364
fertilizer applications, 371-373
grazing management, 365-375
land preparation, 360-362 *
molybdenum toxicity, 373-374
production costs, 376-379
productive mixtures, 356-360
rotation grazing, 365-366
saline soite, 352, 354
seed bed preparation, 362-364
seeding operations, 363-364
soil requirements, 352-356
stand management, 364-365
suitability of saline soils, 354
U.S. acreage, 351
water management, 369-371
weed control, 374-375

nitrogen, 108-109
Irrigation, 234, 258-262
Isopropyl phenylcarbamate, 149
Italian ryegrass, 356

Juglans nigra, 336

Kale, 93

Kaolinite, 14, 88, 166, 184

Kentucky bluegrass, 219, 337, 358, 359

Korean lespedeza, 336

Lactic acid, 281, 283, 284, 285, 286, 287,

291

Lacluca sativa, 121

Ladino clover, 208-230, 310, 336, 353, 359,
360, 367, 368, 371, 372

acreage in U.S., 209, 211

adaptation, 210-212

carrying capacity, 225

cold resistance, 210-212

companion crops, 221-222

diseases, 213

establishment, 213

fall grazing, 225

fertilizer applications, 214-217

grass mixtures, 217-219

heat injury, 212

inoculation, 220-221

insect pests, 213

management, 223-225

origins, 208

pasture yields, 225

rotational grazing, 224-225

seeding methods, 222

seeding practices, 220-222

seed production, 227-229

silage, 227

soil requirements, 212

topdressing, 216-217
Lespedeza, 96, 106
Lespedeza stipulacia, 336
Lettuce, 120, 121, 122, 143, 144, 145, 148

SUBJECT INDEX

405

Lignin, 279

in soil, 87

protein complexes, 87
Lima bean, 115, 122, 139, 140, 150
Liriodendron tulipifera, 335
Lithosols, 186, 189, 190, 194
Loess, 173, 174, 175, 176, 177, 199
Lolium multiflorum, 219, 356
Lolium perenne, 218, 354, 373
Lotus corniculatus, 219, 336, 353, 358, 373
Lycopersicon esculentum, 122

Nitrojection, 117
Nitrosomonas, 86

Oatgrass, 358, 359, 364

Onion, 115, 122, 132, 143, 144, 145, 148,

149
Orchardgrass, 217, 219, 225, 358, 359, 366,

371
Organic phosphorus, 88

Meadow fescue, 218, 354, 358, 359
Meadow foxtail, 219, 359
Medicago hispida, 356
Medicago sativa, 217, 276, 336, 353
Melilotus alba, 336
Melilotus officinalis, 336
Melilotus sp, 328
Methocel, 144

Moisture characteristic, 236, 239, 240-246
Moisture equivalent, 260, 261
Molybdenum, deficiency, 121
irrigated pastures, 373-374
Montmorillonite, 14, 88, 144, 166, 185
Morrow plots, 200, 201
Muskmelon, 119, 134

a-naphthaleneacetic acid, 140
P-naphthoxyacetic acid, 138, 139, 140, 141
Narrowleaf trefoil, 358
Nitrate accumulation, 91
Nitrification, 89-91

photochemical, 90
Nitrobacter, 86
Nitrogation, 116
Nitrogen, cycle, 89

in rain water, 93

soil, see Soil nitrogen
Nitrogen fertilizers, consumption in
North Central states, 107

consumption in U.S., 105
Nitrogen fixation, 86

non-symbiotic, 92, 93-94

symbiotic, 91-93

Paspalum dilatum, 218, 356, 373
Pastures, irrigated, 351-381
Peas, 115, 119, 123, 148, 149
Pectinophorn gossyjriella, 37-38
Pedalfers, 158, 192
Podocals, 192
Pepper, 119, 127, 131-132
Perennial ryegrass, 354, 356, 364, 371
Permeability, field measurements, 267-
268

soil, 246-256

Phalaris arundinacea, 218, 354
Phaseolus lunatus, 122
Phaseolus vulgaris, 123
Phlcum pratense, 217, 278, 353
Phymatotrichum root-rot, 26-28
Pisum sativum, 123
Pisum sativum arvense, 101-102
Planosols, 185, 189, 192, 194, 197
Platanus occidcntalis, 328, 335
Poa pratensis, 97, 201, 219, 337, 358
Populus deltoides, 328, 335
Pore size distribution, 241
Potassium cyanate, 149
Prairie soils, 82, 157-203

age, 182-186

biotic factors, 167-173

calcium content, 164

changes under cultivation, 198-202

characteristics, 159

classification, 192-196

climatic range, 173-175

crop yields, 197-203

distribution, 161, 196-197

erosional effects, 202-203

families, 195-196

406

SUBJECT INDEX

geological origin, 177-178

mineralogical composition, 176-177

movement of clay, 101

organic matter distribution, 171

parent materials, 175-176

phosphorus content, 164

properties of horizons, 165

rotations, 197

series, 194*195

soil forming factors, 166-190

thickness of solum, 180-182

topography, 186-190
Pressure plates, 239
Psallus seriatus, 33-34
Pyrites, 323, 326, 327

Quartz, 185
Quercus borealis, 336

Radiophosphorus, 17, 18

Rayon production in U.S., 5

Reclamation, coal mine spoils, 335-344

Red clover, 353

Reddish Prairie soils, 192, 194, 197

Redtop, 218, 219, 353

Reed canarygrass, 218, 354, 356, 359

Rhodesgrass, 356

Rhizobium, 92

Robinia pseudoacacia, 335

Rock phosphate, 19

Roughage, 275, 280, 293, 307, 308, 309

Rye, 93

Ryegrass, 218, 219

Saline soils, 352, 354
Salix nigra, 323, 328
Sclerotinia trifoliorurn, 213
Secale cercale, 93
Seed, coated, 144

pelleted, 143-146
Silage, 279-294

acids present, 283

addition of acids, 287-288

A.I.V. process, 287

characteristics, 280

cold fermentation process, 286-287

controlled fermentation, 284-290

fermentation losses, 290-291

fermentation process, 284-290

inoculation, 291-292

Ladino clover, 227

making, 280

moldy, 281-282

sterilization, 288-289

warm fermentation process, 286-287
Soils, nitrifying capacity, 86

nitrogen content, 82
Soil moisture, constants, 243

freezing point method, 239

vapor pressure method, 239
Soil nitrogen, 81-109

adsorption by clays, 87-88

barnyard manure, 100-101

chemistry of, 87-89

climatic factors, 83-87

cropping practices, 94-98

crop removal losses, 94, 96

erosional losses, 97, 98-100

green manuring, 101-104

in irrigated areas, 108-109

in midwest soils, 107-108

in southeast soils, 106-107

in temperate soils, 83-84

in the Great Plains, 108

in tropical soils, 85-86

leaching losses, 97

trends in US., 104-109
Soil permeability, 246-258

saturated soil, 247-249

unsaturated soil, 249-253
Soil water, 234-268

diffusion, 256-258
Solanum melongena, 132-133
Sorghum vulgare, 93, 295, 368
Soybeans, 96, 101, 102
Spiriacea oleracea, 124
Spinach, 124, 147, 148, 149
Sprays, pre-emergence, 142, 146, 147
Squash, 133-134
Strawberry clover, 354
Sudangrass, 93, 101, 295, 368
Superphosphate, 19, 145

SUBJECT INDEX

407

Sweet corn, 115, 119, 120, 122, 128, 147,

148, 150
Swiss chard, 93

Tall fescue, 337, 354, 359

Tama silt loam, 159, 162, 163, 164, 202

Tedding, 295

Tensiorneter, 261

Thermistors, 239

Timothy, 353, 358

Tomato, 115, 119, 120, 122, 124, 125, 127,

129-131, 135, 136, 137-139, 142, 143,

145

disease resistance, 124-126
Toxaphene, 35, 37, 38, 39
2,4,5-trichlorophenoxyacetic acid, 141
a-2,4,5-trichlorophenoxypropionic acid,

139

Trifolium hybridum, 210, 353
Trifolium incarnatum, 96
Trifolium pratense, 210, 278, 353
Trifolium repens, 208-230, 310, 336, 358,

373

Tropical soils, 85-86
Typha latifolia, 329

future developments, 151-152
growth control techniques, 135-141
harvesting machinery, 149-151
molybdenum requirements, 120
new varieties, 121-127
nitrogen sprays, 118
plow-sole fertilization, 119
sodium requirements, 120
starter solutions, 119-120
thinning, 143
trace elements, 120-121
utilization of heterosis, 128-135

Verticillium albo-atrum, 29-30

Vetch, 96, 101, 102

Vitia villosa, 92

Vitamin A, 279

Vitamin C, 297

Vitamin D, 291, 297

Watermelon, 123
Weed control, 146-149
Whiptail, 120-121
White clover, 358, 364
White sweet clover, 336
Wiesenboden, 188, 189, 190, 194
Wilting point, 261, 262

Urea, 117, 118

Xanthomonas malvacearum, 30-31

Vegetable production, 114-152
acreage, 114

changes by states, 114-115
direct field seeding, 142-146
fertilization, 116-120
fertilizer placement, 118
fruit setting, 137-141

Yellow sweet clover, 336

Zea mays, 122, 128
Zircon, 183

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Advances In Agronomy Vol II"

See other formats