THE UNIVERSITY

-*L '*

OF ILLINOIS
LIBRARY

The person charging this material is re¬
sponsible for its return on or before the
Latest Date stamped below.

Theft, mutilation, and underlining of books
are reasons for disciplinary action and may

EXPERIMENTAL VEGETATION

THE RELATION OF CLIMAXES TO CLIMATES

Frederic E. Clements and John E. Weaver

THE UHHUr f]r thf

M 11 i£Z5

mmssn m mu

Published by the Carnegie Institution of Washington
Washington, December, 1924

CARNEGIE INSTITUTION OF WASHINGTON
Publication No. 355

Judd & Detweiler, Inc.

WASHINGTON, D. C.

^VX 2,1,'* \

CONTENTS.

5SU5

C 5^

\S)

U">

FAGE.

List of illustrations. v

1. Introduction.

Scope and plan. 3

Principles. 3

Methods..... 4

Values. 5

The grassland climax. 5

Tests of unity. 6

Further evidence. 7

Relationship of the associations...

Climatic unity. 8

Objectives of the present investigation 9

Stations. 10

Location. 10

True-prairie stations. 11

Mixed-prairie station. 12

Plains station. 13

Physical factors. 15

Rainfall and holard. 15

Temperature. 17

Wind and evaporation. 18

Detailed methods. 21

Germination of seed. 21

Depth of planting. 21

Methods of sowing and planting.. 23

. Experiments During 1920 . 27

High prairie, Lincoln. 27

Surface sowing. 27

Trench sowing. 27

Quadrat sowing. 27

Root habits of seedlings. 29

Summary of planting results. 31

Sod transplants. 31

Mixed prairie, Phillipsburg. 31

Surface sowing. 31

Trench and quadrat sowing. 33

Short-grass plains, Burlington. 33

Surface sowing. 33

Trench and quadrat sowing. 33

Sod transplants. 34

Summary of experiments. 34

Seeding in cultivated areas at all

stations. 35

True prairie. 35

Root development. 35

Mixed prairie and short-grass

plains. 36

Experiments at other climatic stat ions. 36
Subclimax prairie, Nebraska City. 36

Character. 36

Results. 38

Mixed prairie, Colorado Springs. . 39

Character. 39

Results. 39

Experiments at edaphic stations.. 40

Gravel-knoll station. 40

Character. 40

iii

PAGE.

2. Experiments During 1920.— Cont.

Gravel-knoll station— Cont.

Behavior. 41

Low-prairie station. 42

Character. 42

Behavior. 43

Salt-flat station. 44

Character. 44

Behavior. 45

Swamp station. 45

Character. 45

Behavior. 46

Summary. 47

Experiments. 47

Physical factors. 48

Experiments during 1918 and 1919.. 49

Season of 1918. 49

Season of 1919. 50

Sod transplants. 51

Gravel-knoll. 51

Low and high prairie. 52

Mixed prairie. 55

3. Experiments During 1921. 56

Physical factors. 56

Rainfall. 56

Water relations. 56

Temperature. 59

Planting results. 60

Surface sowing. 60

Trench sowing.. 60

Denuded quadrats. 61

Root development in denuded

quadrats. 63

Seedling transplants. 65

Summary. 65

Sod transplants. 66

Root development. 67

Experiments at other stations, 1921. 70

Physical factors. 70

Rainfall and holard. 70

Evaporation and temperature. . 71

Planting results. 71

Surface sowing. 71

Trench sowing. 74

Growth in cultivated soil. 75

Root development at Lincoln... 76

Sowing in denuded quadrats. . . 78

Root development at Peru,

Nebraska. 80

Seedling transplants. 80

Summary. 81

Sod transplants. 83

Gravel-knoll. 83

Low prairie. 84

Salt-flat. 85

Salt-basin. 86

Swamp. 86

Regional transplants. 88

568008

CONTENTS.

PAGE.

4. Experiments During 1922. 89

Physical factors. 89

Rainfall. 89

Holard. 89

Temperature. 89

Humidity. 91

Evaporation. 92

Planting results. 92

Surface sowing. 92

Trench sowing. 93

Denuded quadrats. 95

Summary. 96

Seedling transplants. 97

Summary. 98

Sod transplants. 98

Experiments at other stations, 1922. 99

Physical factors. 99

Water relations. 99

Planting results. 101

Surface sowing. 101

Trench sowing. 102

Denuded quadrats. 102

Seedling transplants. 103

Effect of competition. 104

Summary. 105

Sod transplants.. 105

5. Experiments During 1923. 109

Development of seedlings and trans¬
plants. 109

Physical factors. 109

Water relations. 109

Survival results. 110

Surface sowing. Ill

Trench sowing. Ill

Denuded quadrats. 112

Seedling transplants. 112

Experiments at edaphic stations.... 112

Survival results. 112

Surface sowing. 112

Trench sowing. 113

Denuded quadrats. 113

Seedling transplants. 113

Summary of survival. 113

Sod transplants. 115

Phytometric results. 116

Transpiration and growth.' 116

Plan. 116

Methods. 116

Conditions and results at Lin¬
coln. 117

Physical factors. 117

Transpiration and increase in

leaf-area. 119

Conditions and results at Phil-

lipsburg. 119

Installation. 119

Physical factors. 119

Transpiration and increase in
leaf-area. 120

page.

5. Experiments During 1923— Cont.
Phytometric results— Cont.

Conditions and results at Bur¬

lington. 119

Installation. 119

Physical factors. 121

Transpiration and increase in

area. 121

Summary and conclusions. 123

Transpiration and growth of

communities. 123

Transpiration from natural

cover and crops. 123

Objectives. 123

Methods. 124

True prairie, series 1. 125

Installation... 125

Results. 126

True prairie, series 2. 126

Installation. 126

Results. 127

Mixed prairie. 127

Installation. 127

Results. 128

Short-grass plains. 128

Installation. 128

Results. 129

Summary. 130

Growth of natural cover and crops. 131

Objectives. 131

Plan. 131

Results for 1920. 132

Results for 1921. 133

Results for 1922. 135

Results for 1923. 136

Summary.. 136

Structural reponse of dominants

and subdominants. 137

Scope and significance. 137

6. R£sum£. 138

Plan and methods. 138

Communities. 139

Physical factors for 1920. 140

Ecesis during 1920. 140

Physical factors for 1921. 141

Ecesis during 1921. 142

Physical factors for 1922. 142

Ecesis during 1922. 143

Behavior during 1923. 144

Behavior of trees and shrubs. 146

Germination and survival at the cli¬
matic stations. 147

Germination and survival at the

edaphic stations. 148

Survival of sod transplants. 151

Phytometric methods. 151

Attainment of objectives. 152

Bibliography. 154

Behavior tables. 155

LIST OF ILLUSTRATIONS.

PLATES.

Plate 1.

A. Detail of high prairie, showing the luxuriant estival society and the amount

of forage produced in June.

B. General view of high prairie at Lincoln, Nebraska.

Plate 2.

A. Mixed-prairie station at Phillipsburg, Kansas.

B. Corner of short-grass plains station at Burlington, Colorado.

Plate 3.

Root system of year-old Andropogon scoparius ( A ) and Stipa viridula (B), upland
cultivated soil at Lincoln; maximum penetration 4 and 2.7 feet respectively.

Plate 4.

Year-old Elymus canadensis (A), Bouteloua racemosa (B), and Andropogon nutans
( C ); maximum penetration about 3, 3, and 4 feet respectively.

Plate 5.

A. Gravel-knoll station, Lincoln.

B. Low-prairie station, Lincoln, showing transplanted sods.

Plate 6.

A. Low-prairie station in September showing height of the subclimax dominants.

B. Salt-flat station, Lincoln; the chief grass is Distichlis spicata.

Plate 7.

A. View of sod transplants in the swamp at Lincoln, showing water standing on the

surface.

B. Subclimax prairie at Nebraska City, showing rank growth of Stipa spartea

and Ceanothus ovatus held in check by mowing.

Plate 8.

A. A single season’s growth of Andropogon nutans from block of sod (left) and from

seed (right) on low prairie at Lincoln.

B. Year-old Bouteloua hirsuta grown on high prairie at Lincoln.

Plate 9.

A. Year-old seedlings of Andropogon scoparius (right), A. nutans (center), and A.

furcatus (left) grown on high prairie; plants about 6-8 inches tall.

B. Root system of year-old Andropogon furcatus from cultivated lowland, Lincoln;

scale in feet.

Plate 10.

A. Bulbilis dactyloides sod in high prairie.

B. Andropogon nutans in September of first season, showing marked growth with¬

out competition.

Plate 11.

A. Calamovilfa longifolia in September of first season, grown in rich silt-loam.

B. Bouteloua gracilis one year old, grown in cultivated soil.

Plate 12.

Roots of year-old Muhlenbergia pungens (A) and Calamovilfa longifolia (B) grown
in rich silt-loam; depths of penetration 2.5 and 6 feet respectively.

C. Andropogon halli at the end of the first year in silt-loam.

Plate 13.

Andropogon nutans planted in denuded quadrats at Nebraska City, (A) in 1922,
(B) in 1920; photographed June 3, 1922.

Plate 14.

A. Two-year-old Liatris punctata (left) and L. scariosa (right) grown without com¬

petition in cultivated soil: the largest plant had 107 heads.

B. Year-old Desmodium canescens and two-year-old Onagra biennis ( C) growm in

cultivated soil at Lincoln; maximum heights 5 and 7 feet respectively.

Plate 15.

A. Sod-cores and details of installation, Burlington.

B. Clip-quadrat in Bulbilis-Bouteloua short-grass, Burlington.

LIST OF ILLUSTRATIONS.

TEXT-FIGURES. page.

1. Map showing location of stations in the four grassland communities. 10

2. Mean annual precipitation at Lincoln (solid line), Phillipsburg (long broken

lines), and Burlington (short broken lines). 15

3. Average daily temperature at Lincoln (solid line), Phillipsburg (long broken

lines) and Burlington (short broken lines), 1920; here, as elsewhere,
the numbers denote the first, second, third, or fourth week in the month,
respectively. 17

4. Average day (heavy lines) and night temperatures (light lines), at Lincoln (solid

lines), Phillipsburg (long broken lines), and Burlington (short broken
lines), 1920. 17

5. Average daily soil-temperatures at Lincoln (solid line), Phillipsburg (long

broken lines), and Burlington (short broken lines), 1920. 18

6. Hvgrothermograph records from Lincoln (upper) and Burlington (lower),

June, 1921; light lines temperature, heavy lines humidity. 19

7. Average day (light lines) and night humidity (heavy lines) at Lincoln (solid

lines), and Burlington (broken lines), 1920. 20

8. Average daily evaporation at Burlington (short broken lines), Phillipsburg

(long broken lines), and Lincoln (solid line), 1920. 20

9. Monthly precipitation at Lincoln 1920-1922; the monthly mean is shown by

the heavy cross-bars. 27

10. Plants of Bouteloua gracilis (A), Bouteloua hirsuta ( B ), and Sporobolus asper

( C) 44-days old; scale 1 foot. 28

11. Bouteloua gracilis 3 months old. 29

12. Bouteloua hirsuta 3 months old. 29

13. Stipa spartea 3 months old. 30

14. Agropyrum glaucum 3 months old. 30

15. Monthly precipitation at Phillipsburg 1920-1922; the monthly mean is shown .

by the heavy cross-bars. 32

16. Monthly precipitation at Burlington 1920-1922, the monthly mean is shown

by the heavy cross-bars. 32

17. Root of Liatris punctata at end of first season’s growth. 35

18. Monthly precipitation at Nebraska City 1920-1922; the monthly mean is

shown by the heavy cross-bars. 37

19. Average daily evaporation, high prairie (solid line), and low prairie, Lincoln

(short broken lines), and Nebraska City (long broken lines), 1920..... 48

20. Average daily evaporation on high prairie (upper line) and low prairie (lower

line), 1920. 52

21. Average day and night temperatures at Lincoln (solid lines), and Colorado

Springs (broken lines), during 1919. 53

22. Average daily evaporation at Colorado Springs (upper line), and Lincoln

(lower line), during 1919. 54

23. Average daily evaoraption at Burlington (short broken lines), Phillipsburg

(long broken lines), and Lincoln (solid line), 1921. 56

24. Average daily temperature at Lincoln (solid line), Phillipsburg (long broken

lines), and Burlington (short broken lines), 1921. 57

25. Average day (heavy lines) and night temperatures (light lines), at Lincoln

(solid lines), Phillipsburg (long broken lines), and Burlington (short
broken lines), 1921. 59

26. Average daily soil temperatures at depths of 3 and 18 inches respectively at

Lincoln (solid line), Phillipsburg (long broken lines), and Burlington
(short broken lines), 1921. 62

27. Liatris punctata from Phillipsburg, on July 1 of second year after planting. ... 63

28. Bouteloua hirsuta from Phillipsburg, on July 1 of second year after planting... 63

29. Andropogon scoparius from Phillipsburg, on July 1 of second year after planting. 64

30. Roots of Andropogon nutans after the block of sod had been transplanted for

2.5 months at Burlington. 64

31. Roots of Elymus canadensis at the end of June of the second year after the

block of sod had been transplanted and watered at Burlington. 64

LIST OF ILLUSTRATIONS.

VII

TEXT-FIGURES— Continued. page.

32. Roots of Panicum virgatum at the end of June of second year after the block

of sod had been transplanted and watered at Burlington. 69

33. Average daily evaporation on high prairie (solid line), gravel-knoll, Lincoln

(short broken lines), and at Nebraska City (long broken lines), 1921.. 71

34. Pinus ponderosa 2.5 months old. 76

35. Liatris punctata 2.5 months old. 76

36. Robinia pseudacacia 2.5 months old. 77

37. Root system of Gleditsia triacanthus less than 3 months old. 78

38. Average daily temperatures at Lincoln (solid line), Phillipsburg (long broken

lines), and Burlington (short broken lines), 1922. 91

39. Average day (heavy lines) and night temperatures (light lines) at Lincoln

(solid line), Phillipsburg (long broken lines), and Burlington (short
broken lines), 1922. 92

40. Average daily soil temperatures at depths of 3 and 12 inches respectively at

Lincoln (solid line), Phillipsburg (long broken lines), and Burlington
(short broken lines), 1922. 93

41. Average daily evaporation at Lincoln (solid lines), Phillipsburg (long broken

lines), and Burlington (short broken lines), 1922. 94

EXPERIMENTAL VEGETATION

THE RELATION OF CLIMAXES TO CLIMATES

Frederic E. Clements
and

John E. Weaver

EXPERIMENTAL VEGETATION.

1. INTRODUCTION.

SCOPE AND PLAN.

The investigations that have led to the development of the phytometer
and the experimental method in vegetation are companion efforts in the en¬
deavor to organize ecology as a quantitative science. No subject has suf¬
fered more from the lack of training and experience or from the absence of
basic procedure and adequate perspective. It has become obvious that this
condition can be remedied and actual progress insured only by instrumental
and experimental methods and by a combination of intensive and extensive
studies that will check superficiality on the one hand and broaden limited
horizons on the other. It is certain that much work of purely superficial or
local character will continue to be done under the name of ecology, but the
touchstones of instrument, quadrat, and experiment afford a ready means of
eliminating such papers from consideration. No study deserves to be called
ecological that does not deal with the cause-and-effect relation of habitat and
organism in a quantitative and objective manner.

The task of focusing the efforts of the ecologist upon the habitat as cause
and the plant and community as effect and reaction furnishes the theme for
four related studies, namely, phytometers, experimental vegetation, climatic
cycles, and changes of climate and climax. The first two constitute an
endeavor to employ plant and community as the best measures of cause and
reaction, while the last two emphasize the fundamental significance of change
as the proper approach to this complex of problems.

PRINCIPLES.

The basic guides in this task are the cause-and-effect relation on the one
hand and the cyclic nature of change on the other. The former is paramount
in every ecological problem, since investigation that does not deal directly
and constantly with causes lacks the very essence of science. However, in
dealing with processes in a way that is at once complete and synthetic, ecology
must not overlook the fact that both cause and effect must be inseparably
associated and that effect sometimes affords the readier approach to the prob¬
lem. It is equally essential that ecology be treated as inherently dynamic,
that change in some degree or at some rate be regarded as universal, and that
such changes be recognized as recurrent or cyclic, both as cause and in effect.
The fact has elsewhere been emphasized that effects in turn produce reactions,
which consequently become causal to the extent that they modify the original
causes (Plant Succession, 79). This constitutes the basic cycle which finds
its expression in the larger cycle of the sere. Thus, succession is an intrinsic
function of all vegetation, though the shifting of climaxes is a matter of thou¬
sands or tens of thousands of years rather than of scores in the subsere or
hundreds in the prisere.

INTRODUCTION.

The cyclic changes of the habitat are likewise reflected in the individual and
the species. Adjustment and adaptation are clearly cyclic processes, and
this is even more evident in the major evolutions of groups and floras. As
the consequence of reaction and competition in the sere and of climatic shift
in the major seres, invasion is necessarily cyclic and finds its typical expression
in terms of dominant and relict phases. Finally, cyclic changes serve to
link the present with the past in such a broadly continuous fashion that the
gap between them disappears and such terms as ecology and paleo-ecology
are seen to be mere conveniences.

METHODS.

The methods of experimental vegetation are based on the processes inher¬
ent in the natural development of vegetation, but segregated and controlled
in such a way as to reveal the value and significance of each individual step
or process. This is even true of migration, regarded merely as transport,
the difference in the role of man as an unintentional or purposeful agent being
immaterial, except as to the number and fate of the migrules. The most
exact evidence as to the nature and success of invasion and ecesis is doubtless
that to be obtained in the form of natural experiments, but except in the case
of propagules these are difficult to discover outside of ecotones and are alto¬
gether absent from the typical areas of great associations. Hence, the only
feasible procedure is to approximate the method of natural experiment as
closely as possible, bearing in mind that the numbers must be large enough
to insure against complete loss. Moreover, it must be recognized that arti¬
ficial aids to ecesis, such as watering, screening, etc., fall well within the vari¬
ation of factors in nature and are in consequence more or less natural.

The basic methods of experimental vegetation are as follows: (1) sowing
seeds; (2) planting seeds or propagules; (3) planting seedlings; (4) transplant¬
ing adult plants of various ages; (5) transplanting small communities or por¬
tions of communities. The converse of this is to be found in the method of
denuded quadrats and transects, but as this is primarily a question of ecesis
within a community, it has not been extensively employed in the present
study. The accessory methods involve the use or manipulation of (1) com¬
petition; (2) physical factors; (3) protection against animals; (4) instruments;
(5) phytometers; (6) seasons and cycles. Time also enters the equation, inas¬
much as germination and the ecesis of seedlings are periods of the greatest
stress. Various ways are available for reducing or eliminating competition
and modifying the physical factors, but the following have proved the sim¬
plest and most satisfactory, viz: (1) sowing on the surface in the midst of
natural vegetation; (2) sowing or planting in trenches, by which competition
is prevented for a short time; (3) sowing or planting in denuded quadrats,
which eliminates competition for a much longer period, but renders the water
relations usually less favorable at the same time that it improves light con¬
ditions; (4) transplanting adult plants into living cover or denuded areas,
with similar consequences as to competition, water, and light; (5) improving
the conditions for germination, establishment, or survival by watering, shad¬
ing, thinning, draining, etc. A full battery of recording instruments is indis¬
pensable and must be supplemented by determinations of chresard, evapora¬
tion, and light intensity in all cases, and of salts and air-content whenever

GRASSLAND CLIMAX.

these are controlling. Equally essential are phytometers of various types
to serve as interpreters between the physical instruments and the natural
cover on the one hand, and the transplants on the other. Furthermore, the
differences between seasons and the wet and dry phases of the climatic cycle
must be taken into definite consideration, as they are often greater than the
climatic differences between the several grassland associations or even between
formations. Finally, a clear distinction must be drawn between persistence
and dominance, the former indicating the potential and the latter the actual
role of a species in the climax.

VALUES.

The basic value of the experimental method in vegetation is evident, but
its far-reaching significance warrants detailed consideration. While its first
importance inheres in its being indispensable in the search for causes, the
problems of the origin, unity, and relationship of formations and associations
and their correlation with climate can be solved objectively and permanently
in no other way. In fact, the objectivity afforded by comprehensive and
repeated experiment is the paramount reason for its constant and universal
use. The opinions and hypotheses arising from observation are often inter¬
esting and suggestive and may even have permanent value, but ecology can
be built upon a lasting foundation solely by means of experiment. This does
not mean that observation and reconnaissance have no value, but such results
are to be regarded as provisional, pending adequate experimental study. Still
less does it mean that hypothesis and theory are to have no part in an exact
and quantitative ecology. Such a view is possible only to those who ignore
the decisive role that theory has played in the development of modern chem¬
istry and physics and are unfamiliar with its essential stimulus in the ad¬
vance of biology.

The study of vegetation demands the use of experiment to even a greater
degree than that of the plant, owing to the complexity and extent of the great
communities and the consequent opportunity for making general observations
and forming unchecked opinions. Moreover, it must be fully recognized
that the intensive-extensive method of investigation applies with peculiar
force to the plant community, since it was developed for the express purpose
of turning complexity and extent to the advantage of thorough and detailed
analysis. As a procedure, it has been fully justified by the experience of
years in the vegetation of western North America, in which the extensive
results of constant field studies have given perspective and balance as well
as new points of departure to the results of the laboratory and station, and
the latter have served to refine and definitize the natural experiments and
observations over a vast territory.

THE GRASSLAND CLIMAX.

The development of the views as to the nature and structure of the grass¬
lands of North America illustrates the need of objective methods of determin¬
ing vegetation units and their relationships. This is all the more convincing,
since the ecological investigation of the prairie and plains has been the work
of a group with the same general training and outlook. In the first analysis
of the grassland, Pound and Clements (1898:243, 1900:347) recognized two

INTRODUCTION.

prairie formations, viz, the prairie-grass and the buffalo-grass formations, a
bunch-grass formation of the sandhills, and a meadow formation. In the
light of successional studies, the last two are to be regarded as subclimaxes.
In a few years (Clements, 1902) it had become clear that the prairie-grass
or Stipa-Agropyrum formation and the buffalo-grass or Bulbilis-Bouteloua
formation were the two great communities of the prairie-plains region.
This was essentially the view also of Shantz (1906, 1911) and of Pool (1914).
This conception was maintained in Plant Succession (180, cf. note) after
many additional years of successional research. However, the developmental
concept of the formation had broadened its scope and afforded a clearer view
of its structure. As a consequence of the special study of these relations, it
became necessary to abandon the view of two separate grassland formations
and to recognize a single formation composed of several associations. Mean¬
while, it had become increasingly evident that the Agropyrum consociation
of the Northwest was closely related to the Stipa-Agropyrum prairie. This
was first suggested by finding the three dominants associated from Washing¬
ton to Montana during the field work of 1914, a view adopted by Weaver,
who had worked intensively in this community (1917:40). This idea was
confirmed by further studies in 1917, but the true relationship was obscure
until it became certain in 1918 that Stipa setigera and S. eminens were the
original bunch-grasses of California. As a consequence, it proved possible
to recognize a fourth grassland association, composed of bunch-grasses and
characteristic of the Pacific region of winter precipitation (Plant Indicators,
115). By means of extensive field work and the comparative study of com¬
munities, the actual relationship of the grassland units was determined, but
the experimental method might well have accomplished in 5 years what obser¬
vation required 20 years to attain.

TESTS OF UNITY.

In reaching the conclusion that the grasslands constitute a single climax
formation and not several more or less related but distinct units, recourse was
had to a number of tests, all as objective as possible.

“The conclusion that the grassland is a single great climax formation is based in
the first place on the fact that the three most important dominants, Stipa, Bouteloua,
and Agropyrum , extend over most of the area, and one or the other is present in
practically every association of it. This would seem the most conclusive evidence
possible, short of actual vegetation experiments, that the grassland is a climatic vege¬
tation unit. Equally cogent is the fact that these dominants, together with Carex,
Bulbilis, and Koeleria, mix and alternate in various groupings throughout the Stipa-
Bouteloua association. Indeed, this association appears so conclusive as to the gen¬
eral formational equivalence of these seven dominants that it is regarded as the
typical or base association. In addition, the characteristic societies either extend
through several of the associations or are represented by corresponding communities
belonging to the same genus. The relation of the associations to such subclimax
species as Andropogon scoparius, Calamovilfa longifolia, Aristida purpurea, and Elymus
sitanion further confirms the relationship of the dominants. The most obvious differ¬
ence between the various associations is exhibited by the tail-grass prairies, Stipa-
Koeleria poium, and the short-grass plains, Bulbilis-Bouteloua poium. Yet these
are closely related, as shown not only by the criteria given above, but also by their
geographical contact. Still more eloquent is the fact that grazing favors Bouteloua
and Bulbilis at the expense of Stipa and Agropyrum, and thus frequently converts the

GRASSLAND CLIMAX.

base association of Stipa-Bouteloua into a pure short-grass cover. Concrete evidence
of this has been obtained in widely separated areas and has led to the working hypoth¬
esis that a pure short-grass community is partly if not largely a response to over-
grazing.” (Plant Indicators, 115.)

FURTHER EVIDENCE.

In the six years since the above was written, a large amount of additional
evidence has been accumulated as to the unity of the grassland and the
close phylogenetic relationship of its associations, quite apart from the experi¬
mental results brought forward later. In fact, the detailed field study of
the ranges of the dominants makes it clear that the associations are even
more closely related than was at first supposed, and the problem becomes one
of maintaining several of these units as distinct rather than one of justifying
their inclusion in the same formation. The number of dominants common to
contiguous units has steadily increased, as well as the areas concerned. The
most unique of the associations, the short-grass plains, has been shown to be
a recent modification of mixed prairie due to overgrazing and has been more
closely connected with the desert plains. The widespread dominants that
occur in all the climatic associations have increased in number and now com¬
prise Stipa comata, Agropyrum glaucum, Bouteloua gracilis, B. racemosa,
Sporobolus cryptandrus, and Koeleria cristata.

Even more striking has been the evidence drawn from relicts and from sea¬
sonal variations arising out of the climatic cycle (Clements, 1921, 1922, 1923).
These are the results of natural experiments on a scale and over an area that
can not even be approximated by the experimenter, and their value is of the
greatest, especially when supplemented by control experiments. In addition,
the repetition and checking that are so indispensable are not only provided
by recurrence in thousands of localities, but particularly also by the wide
variation in behavior from season to season at the wet and dry phases of the
climatic cycle. Of the first and the most direct importance are the relicts
produced by overgrazing and by rodents, since these factors have operated
directly upon the normal climax. Such relicts permit the complete recon¬
struction of the original community, both as to structure and extent. Indeed,
this is usually done in the most graphic fashion by nature herself during sea¬
sons or phases of excessive rainfall, and when a drought period of several
years is followed immediately by a year of exceptional rainfall, as happened
from 1916-1919, a revealing picture is obtained. The expansion and con¬
traction of relict areas at such a time reproduces in miniature what occurs
during the wet and dry phase of a major climatic shift extending over a period
of a thousand years or more. In the light of grazing relicts and cyclic changes
it has proved possible to reconstruct all the grassland associations in detail
as they were before the historical period in the West. Furthermore, recon¬
struction has been carried much farther back to a time probably 10,000 to
20,000 years ago, when the differentiation of the grassland mass into the
various associations had barely begun. In short, the relict method confirms
the view of the unity of the grassland climax by disclosing several stages in
the differentiation of the original mass and affording a measure of the extent to
which this has progressed. This is the theme of the volume on the changes
of climate and grassland, and hence it will not be further elaborated here.

INTRODUCTION.

RELATIONSHIP OF THE ASSOCIATIONS.

The above suffices to indicate what appears to be clearly proven by the
detailed evidence, namely, that the relationship of the six communities of
the grassland is a phylogenetic one. This seems almost obvious in the case
of the subclimax prairie and the short-grass plains. The latter is demon¬
strably a grazing subclimax of the mixed prairie, though such a hard-pan area
as that about Burlington may be an actual relict of a short-grass climax
developed during the last major dry phase. On the other hand, the sub¬
climax prairie is just as clearly an eastward extension of the taller meadow-
grasses from the true prairie, made in response to the same climatic shift
and profiting by fire, especially during the prehistoric period. These changes
are still so recent that the detailed steps can be readily followed, and similar
adjustments can be traced between the other associstiona. However, the
latter have been largely differentiated at an earlier period, and the phylo¬
genetic and geographic connections are less broad.

The clue to the relationships of the four actual associations is to be found
in the mixed prairie, which most nearly represents the original formation, and
consequently affords a measure of the divergence of the others. The true
prairie stands closest to the mixed prairie, Agropyrum glaucum and Koeleria
cristata being common dominants, and the closely related Stipa spartea and
S. comata, and Sporobolus asper and S. cryptandrus being reciprocal dominants
in the two respectively. The chief difference lies in the fact that the short-
grasses of the Stipa-Bouteloua association are usually absent, or form a frag¬
mentary layer in the true prairie, except where grazing has favored their
spread. The bunch-grass association resembles the true prairie in the absence
of the short-grasses, but its endemic dominants are naturally different, e. g.,
Agropyrum spicatum, Stipa setigera, and Festuca ovina in place of Stipa
spartea and Sporobolus asper. Its cognate relation to the mixed prairie is
shown by such common dominants as Stipa comata, Agropyrum glaucum, and
Koeleria cristata, and by the persistence to-day at the higher altitudes in the
mixed prairie of such bunch-grasses as Stipa setigera, S. eminens, and Festuca
ovina. The desert plains bear a close resemblance to the short-grass condition
of the mixed prairie, the reduction of the upper layer of tail-grasses being more
a matter of drought than of grazing. The close relationship of the two is
shown by the dominance of Bouteloua in both, as well as by the fact that
such tail-grasses as Sporobolus cryptandrus, Bouteloua racemosa, Stipa pen -
nata, and S. comata persist as an upper layer in the desert plains wherever
protection permits. Moreover, the relict areas in and about the Mohave
Desert and Death Valley prove beyond doubt that the desert plains and bunch-
grass associations were once in contact and that at an antecedent period they
were commingled and probably identical (Clements, 1922, 1923).

CLIMATIC UNITY.

The fundamental fact in the relation of climaxes to climate is that this
must be determined by plant judgments rather than by human ones (Plant
Indicators, 116). This is necessarily truer of grasses than of trees, since their
life-habits enable them to minimize or escape the rigors of winter as well as
the dangers of the dry season of the year. Consequently, it should be regarded
as neither strange nor perplexing to find such great grassland dominants as
Bouteloua gracilis, Stipa comata, and Koeleria cristata ranging through a num-

OBJECTIVES.

ber of climates in the human sense. The human conception of a climate is
loose enough at best, and it regularly ascribes much greater importance to
temperature than to water relations. On the contrary, grasses are more
responsive to differences of rainfall and evaporation than to those of tempera¬
ture, Bouteloua gracilis in particular ranging much more extensively in lati¬
tude than in longitude as a climatic dominant. This is confirmed by the
construction of dominance maps of the chief grasses, which disclose the fact
that the interruption or disappearance of a consociation is primarily a matter
of water, a conclusion entirely supported by the present investigation.

In addition, it must be recognized that obvious differences in the amount
and distribution of rainfall may have slight significance for grasses. It is
the amount and distribution during the growth period of the grass dominants
that controls, and this is equally true of the effect of evaporation. The mixed
prairie may range from 25 to 10 inches without undergoing any essential
change in composition, owing to the fact that the period of development is
shortened to fall within the limits of optimum water conditions. The exist¬
ence of two rainy seasons, as in southern Arizona, seems to indicate an impor¬
tant difference in climate, but the growth of the grassland is still an effect of
the much more adequate summer rains. The rainfall of the Pacific Coast is
of the winter type and at first thought appears entirely different. However,
this is only apparently true in so far as the native bunch-grasses are con¬
cerned. These bear the same general relation to the spring rains as that pre¬
vailing in the mixed prairie, namely, development begins during the rains,
followed by maturity and the drying of the aerial parts in late spring or early
summer respectively. The differentiating effect of the peculiar distribution
of rainfall has been recorded in the characteristic life-form, though the
progressive assumption of the bunch-grass habit is revealed by the fact that
it is found in practically all the dominants of the desert plains and in nearly
all the tail-grasses of the mixed prairie.

OBJECTIVES OF THE PRESENT INVESTIGATION.

As has been already indicated, the primary purpose of these researches is
to develop an experimental method that will combine the maximum of demon-
strability and objectivity. Because of its size and apparent complexity,
vegetation affords an open field for prepossession and interpretation, and its
study can be made scientific in the proper degree only by the resolute employ¬
ment of experiment and measurement. The special ends of this investigation
are many, chief among which are the demonstration of the unity of the grass¬
land climax and climate, and the growing differentiation of the associations,
partly in response to the subclimates and partly to the changes wrought directly
or indirectly by man. It has also proved possible to cast further light upon
the relationship of climatic and edaphic areas in the same region, as well as
upon the significant role of climatic cycles and communi'y shiftings in this. A
special endeavor has been made to analyze competition in minuter detail and
to evaluate more fully the effect of light intensity in a grassland cover. Par¬
ticular attention has been directed to germination, seasonal ecesis, and final
survival, especially in relation to annual variations in climate. Finally, the
entire study has been designed to exemplify a new aspect of the phytometer
method, in which native species and communities have received the major
emphasis.

INTRODUCTION.

STATIONS.

Location.

The stations were chosen to represent typical conditions in each of the
four grassland communities occurring in the vast area between the Missouri
River and the Rocky Mountains. Lincoln, Nebraska, was selected as repre¬
sentative of true-prairie conditions. Aside from the matter of convenience
in the use of the facilities of the Department of Botany of the University of
Nebraska, the vegetation and environment of this region are better known
than anywhere else in the grassland formation, due especially to continued
study by Pound and Clements (1897, 1898, 1900), Clements (1905, 1907,
1916, 1920), Pool (1914), Weaver (1917, 1919, 1920, 1921, 1922, 1923), and
others. Phillipsburg, in north-central Kansas, was selected as representative
of mixed-prairie conditions, while Burlington, in eastern Colorado, was chosen
in the short-grass plains. Phillipsburg lies about 190 miles southwest of
Lincoln, and Burlington about 180 miles further west and south, 10 miles

Fig. 1 . —Map showing location of stations in the four grassland communities.

west of the Kansas-Colorado State line (fig. 1). The altitude at the several
stations rises from 1,100 feet at Lincoln to 1,900 feet at Phillipsburg and
4,160 feet at Burlington. Precipitation, the chief factor in determining the
type of vegetation, ranges from 28 to 23 and 17 inches at the respective sta¬
tions, decreasing westward. A fourth station was maintained at Nebraska
City, Nebraska, about 50 miles southeast of Lincoln, near the Missouri River.
Although the rainfall of 33 inches is sufficiently ample to produce scrubland
or woodland, except for grazing, mowing, fires, etc., this station lies just
within the border of the subclimax grassland and was chosen largely because
of its ready accessibility.

In addition to the four stations mentioned, considerable experimental work
was carried on at the mixed-prairie station at the foot of Pike’s Peak near
Colorado Springs, the Alpine Laboratory being used as a base. At Lincoln

STATIONS.

a whole series of stations, ranging through gravel-knoll, high prairie, low
prairie, salt-flat, swamp, and cultivated fields, was maintained. Moreover,
reciprocal transplants from Tucson, Arizona, and Berkeley, California, have
been made at many of the stations.

True-prairie Stations.

The high-prairie station at Lincoln (plate 1) was located on a rather flat
hilltop about 60 feet above the general level of the flood-plain of Salt Creek
and 2 miles north of the city. The fertile soil is of the type commonly called
loess, but is really a glacial drift. It is a silt-loam belonging to the Marshall
series, with a moisture equivalent of about 28 per cent and a maximum water
capacity of approximately 60 per cent. The mechanical and chemical analyses
of soils taken from an adjacent field in the same area (Weaver, 1920:140)
indicate its general characteristics (tables 1 and 2).

Table 1 —Mechanical analysis of high-prairie soil.

Depth of sample.

Coarse

gravel.

Fine

gravel.

Coarse

sand.

Me¬

dium

sand.

Fine

sand.

Very

fine

sand.

Silt.

Clay.

0.0 to 0.5 foot.

p. ct.
0.0

p. ct.
2.0

p. ct.
1.9

p. ct.
6.0

p. ct.

26 1

p. ct.

39 3

p. ct.

24 7

0.5 to 1.0 foot.

0.0

1.2

2.2

3.8

21.8

38.4

32.6

1 to 2 feet.

0.0

2.7

2.9

4.8

19 6

45 6

24 4

2 to 3 feet.

0.0

5.6

6.9

10.1

23.1

32.8

21 5

3 to 4 feet.

0.0

7.0

8.8

12.8

23.5

28.3

19.6

Table 2. —Chemical analysis of high-prairie soil.
[Digestion with HCL (sp. gr. 1.115) for 120 hours.]

Depth of sample.

Insolu¬

ble

resi¬

due.

Solu¬

ble

salts.

Vola¬

tile

mat¬

ter.

Iron
and alu¬
minium
oxid.

Cal¬

cium

oxid.

Mag¬

nesium

oxid.

Phos¬

phorus

pent-

oxid.

Nitro¬

gen.

0.0 to 0.5 foot.

0.5 to 1 foot.

1 to 2 feet.

2 to 3 feet.

p. ct.

76.87

75.70

76.17

77.80

p. ct.

17.12

18.58

19.08

18.46

p. ct.
6.01
5.72
4.75
3.68

p. ct.

13.20

14.25

14.72

14.03

p. ct.
0.68
.70
.75
.86

p. ct.
1.19
1.32
1.68
1.69

p. ct.
0.13
.12
.12
.15

p. ct.
0.159
.134
.079
.045

The soil is fine in texture, being composed mostly of silt and clay, and is
sufficiently supplied with calcium to lack acidity. Determinations through¬
out a number of years show that the subsoil is usually moist to great depths,
although at infrequent intervals during drought periods the holard may be
reduced below the wilting-coefficient of Briggs and Shantz (1912) to depths
of 4 to 5 feet, leaving only a small chresard for vegetation. Root excavations
and bisects show that the plants are not only rooted deeply, but also that
the root-systems of different species form layers in the soil, the shallowest one
ending at about 3 feet and an intermediate one at 5 feet, while a third layer
extends far below this level (Weaver, 1920:28, 40).

INTRODUCTION.

The vegetation is distinctly of the tail-grass sod type. Andropogon sco-
parius, Stipa spartea, Koeleria cristata, and Bouteloua racemosa are the chief
grasses, although Andropogon fmeatus and nutans occur more or less along
with Poa pratensis. The interstitial Panicum scribnerianum and the relict
Bouteloua gracilis are of much less importance. Prevernal societies are repre¬
sented by Antennaria campestris and Carex pennsylvanica. Such vernal
bloomers as Astragalus crassicarpus, Baptisia bracteata, Senecio plattensis, and
Nothocalais cuspidata are abundant, while the variety and abundance of esti¬
val herbs indicate favorable growth conditions throughout the early summer.
Chief among these are Psoralea floribunda, Erigeron ramosus, Brauneria pal¬
lida, Meriolix serrulata, and Achillea millefolium, although a host of others
occur (Pound and Clements, 1900; Weaver and Thiel, 1917). Species of
Solidago, Aster, Liatris, Helianthus, Kuhnia, etc., constitute the most impor¬
tant autumnal societies, the subdominants during summer and autumn giving
a truly kaleidoscopic appearance. An average grass-level of 6 to 8 inches
and an upper story of forbs at 15 to 22 inches is attained by June 1, although
the flowerstalks of Stipa and the later grasses and herbs are 2.5 to 3.5 feet
tall (plate 1).

Mixed-prairie Station.

This occupies an area just south of Phillipsburg, quite typical of the gently
rolling topography, on a hillside which slopes gently southward (plate 2a).
The fertile soil is a mellow, dark-brown, very fine sandy loam of the Colby
series. At a depth of 12 to 15 inches it is slightly lighter in color and con¬
tains enough clay to be quite sticky, although when wet it is dark in color to
a depth of 2 feet. Below this level it is light yellow and shows its loess origin
throughout. The first 4 feet have a water-holding capacity of about 66 per
cent. As is true of most soils of semiarid regions, it shows no acidity at
any depth, and the mellow subsoil is very deep.

Table 3. —Mechanical analysis of soil at Phillipsburg, Kansas.

Depth of sample.

Coarse

gravel.

Fine

gravel.

Coarse

sand.

Me¬

dium

sand.

Fine

sand.

Very-

fine

sand.

Silt.

Clay.

p. ct.

0.0 to 0.5 foot.

0.0

0.3

0.2

1.2

43.5

35.8

19.0

0.5 to 1 foot.

0.0

0.2

0.5

44.4

32.8

22.1

1 to 2 feet.

0.0

0.2

0.3

39.7

34.0

25.8

2 to 3 feet.

0.0

0.3

0.6

41.2

31.9

26.0

3 to 4 feet.

0.0

0.1

0.2

37.5

31.4

30.8

—

Repeated excavations for the examination of the roots of native and crop
plants from 1919 to 1921 showed that the soil was moist to 8 feet. This
condition is assumed to be abnormal for the region, but it is directly to be
correlated with the excessive precipitation of 1919 (Weaver, Jean, and Crist,
1922:77). However, this station is somewhat subject to drought, the holard
to a depth of 4 feet being sometimes reduced approximately to the hygroscopic
coefficient, about 10.6 per cent. Under these conditions the native vegetation
is rooted almost or quite as deeply as in the true prairie (Weaver, 1920:93).
The vegetation is essentially mixed prairie, the tail-grasses alternating with

STATIONS.

or forming a layer above the short ones (plate 2a). Andropogon scoparius,
nutans, and furcatus often form more or less continuous irregular sods varying
from 6 inches to 7 feet in diameter, from which short-grasses may be almost
entirely excluded, while Agropyrum glaucum likewise frequently occupies large
areas rather exclusively. Bouteloua racemosa and Elymus canadensis are
other important tail-grasses. Alternating with these are similar or on drier
slopes even larger areas of Bulbilis dactyloides and Bouteloua gracilis, the
intervals being nearly devoid of vegetation, often to the extent of one-fourth
of the surface. More usually, however, the two kinds of grasses are inti¬
mately mixed, the short-grasses often showing a strong tendency toward the
bunch habit. Car ex filifolia and stenophylla supplement the understory of
grasses, the leaves of which reach an average height of about 4 inches (before
flower-stalk production), as contrasted with the midsummer tail-grass level
4 to 10 inches above. A taller open layer of Psoralea tenuiflora at 2 feet
characterizes much of the area in late June, when societies of Ratibida colum-
naris and Morongia uncinata are also conspicuous. However, as emphasized
by Clements (1920:138), the mixed prairie shows its xerophytic tendency
by less numerous and less extensive societies. Antennaria campestris, Astrag¬
alus crassicarpus, Nothocalais cuspidata, Anemone caroliniana, Senecio plat-
tensis , Vida americana, etc., are all represented, even if sparingly, in the
spring and early summer, but the absence of Viola, Stipa, Koeleria, and
Brauneria is at once noted, and the presence of Astragalus mollissimus, Oxy-
tropis lamberti, Malvastrum cocdneum, Opuntia fragilis, 0. camanchica, Aristida
purpurea, and Plantago purshi indicates a more xerophytic type of vegetation.
The sub dominants of midsummer are usually smaller and less abundant than
eastward, while the autumnal aspect likewise lacks many species common to
the true prairies. The part of the mixed prairie used for experimental planting
was carefully selected with due regard to a proper balance between tall and
short grasses.

Plains Station.

This is located just north of Burlington, on a vast level tract (plate 2b) .
The soil is a rich, brown, fine sandy loam, very compact and hard when dry.
It has a water-holding capacity of 65 to 70 per cent to a depth of 4 feet. At
a depth of 2 to 2.5 feet it is underlaid with a so-called hard-pan. Soil analyses
show that the concentration of colloidal clay and carbonates in the subsoil is
sufficient to give rise to a hard-pan, i. e., a much more compact intercalated
stratum of soil, upon its becoming completely dried out (Weaver and Crist,
1922). Silt constitutes about one-third of the soil at all depths, while the sand
decreases and the clay increases in amount to 4 feet (table 4).

Table 4. —Mechanical analysis of soil at Burlington, Colorado.

Depth of sample.

Coarse

gravel.

Fine

gravel.

Coarse

sand.

Me¬

dium

sand.

Fine

sand.

Very

fine

sand.

Silt.

Clay.

Hygro¬
scopic co¬
efficient.

p. ct.

0.0 to 0.5 foot..

0.0

0.1

2.6

48.6

33.4

15.3

10.9

0.5 to 1.0 foot..

0.0

0.1

2.2

49.1

32.5

16.1

10.9

1 to 2 feet.

0.0

0.2

1.9

46.7

32.0

19.3

12.2

2 to 3 feet.

0.0

0.1

1.5

45.5

31.0

21.9

12.0

3 to 4 feet.

0.0

0.1

0.9

42.2

34.2

22.6

11.4

INTRODUCTION.

Chemical analyses show that carbonates are practically absent in the
surface soil, but increase rapidly with depth, and often reach concentrations
of 5 or 6 per cent in the hardpan layer, which appears somewhat chalky in
color. The soil is not at all acid; the carbon dioxid increases very rapidly
with depth and is high at 2 to 4 feet. This soil is rich in phosphorus
and potassium and has a sufficient supply of nitrogen, and hence all the
essential raw materials are present in abundance.

Table 5. —Chemical analysis 1 of soil at Burlington , Colorado.

Depth of
sample.

Acidity.

Carbon

dioxid.

V olatile
matter.

Phosphorus

pentoxid.

Sulphur

trioxid.

Potassium

oxid.

Nitro¬

gen.

0.0 to 0.5 foot..
0.5 to 1.0 foot..

1 to 2 feet.

2 to 3 feet.

3 to 4 feet.

p. ct.
None.
None.
None.
None.
None.

p. ct.
0.03
0.30
1.71
2.10
2.60

p. ct.
4.67
3.13
3.11
3.34
2.84

p. ct.

0.189

0.504

0.428

0.406

0.525

p. ct.
0.007
0.017
0.006
0.006
0.005

p. ct.

2.32

2.39

2.45

2.51

2.22

p. ct.
0.184
0.130
0.101
0.086
0.084

1 Phosphorus determinations were made by digestion with HN0 3 and HF, sulphur by fusion
with Na 2 C> 2 , potassium by fusion with calcium carbonate, and nitrogen by the modified Gun¬
ning method.

Water penetrates very slowly and run-off is usually high in this fine-
textured soil. Shantz (1911) has shown that the average run-off from the
short-grass sod at five stations in this region was 37 per cent of the total
rainfall, while the maximum run-off reached 55 per cent. Consequently, the
actual precipitation of 17 inches is a poor index of the efficiency of the rainfall.
After heavy rains, 3 days were required for water to penetrate to a depth
greater than 6 inches. The excellent root development of native plants in
the surface 1.5 to 2.5 feet of soil enables them to absorb water readily, and
further prevents a deeper penetration of rain. Determinations of the holard
through a series of years (1920 to 1923) show that the soil is seldom moist
below 2 feet, while by midsummer the amount above this level is frequently
reduced to the hygroscopic coefficient. When this happens, the short-grass
cover dries out and “cures” on the ground.

Closed mats of Bulbilis dadyloides, usually mixed with Bouteloua gracilis ,
make up fully 90 per cent of the vegetation, forming a carpet seldom over
inches deep, exclusive of the flower-stalks (plate 2b). Agropyrum glaucum is
frequently associated with the short-grasses, but overgrazing and drought
have resulted in a dwarf habit and it forms flower-stalks sparingly, except in
years of more than normal rainfall. These grasses, with their widely spreading
roots, occupy the soil so thoroughly that relatively few important sub¬
dominants are present. Most conspicuous among these, and increasing in
abundance where overgrazing has occurred, are Aristida purpurea, Opurdia
camanchica, O.fragilis, 0. polyacantha, Festuca octoflora, Plantago purshi, and
Schedonnardus paniculalus. Erysimum asperum and Psoralea tenuiflora also
are often abundant, while small, poorly developed societies of Astragalus
crassicarpus, Malvastrum coccineum, and Ratibida columnaris are infrequent.
Decreased size, vigor, and number are characteristic of most of the species
when compared with their growth in more favorable conditions.

CLIMATES.

While some of the species of the meager flora are rooted entirely in the
surface soil, others reach depths of 4 feet or more, especially the dominant
grasses and the legumes. However, even these are relatively shallow when
compared with the great depth attained by species in the moist subsoil of the
true prairie.

PHYSICAL FACTORS.

Rainfall and Holard.

The grassland associations concerned lie in the region of summer rainfall;
by far the greater part of the rain falls during the growing-season and only
about one-tenth during the three winter months (fig. 2). Such a seasonal
distribution of the precipitation is very favorable to the growth of grasses.
The normal decrease of 5 inches at Phillipsburg in comparison with Lincoln
and the further decrease of 6 inches at Burlington are quite evenly dis¬
tributed through the season. Moreover, the type of rainfall is similar through¬
out, consisting usually of heavy showers, often of rather short duration,
though this is more marked upon the high plains than eastward, where che
rains are more general. At all stations it results in much run-off, but this is
particularly heavy on the compact soil at Burlington. However, in the plains
region more of the precipitation falls in light showers of 0.2 inch or less, which
are of practically no value in increasing the holard. At each station the holard
has in general been very similar at any particular time during the three seasons
(1920 to 1922), and the comparison of a single season’s data will suffice here
(table 6).

Fig. 2. —Mean annual precipitation at Lincoln (solid line), Phillipsburg (long
broken lines), and Burlington (short broken lines).

At Lincoln a sufficient amount of water to promote good growth was avail¬
able at all depths and at all times. At the mixed-prairie station, July and early
August were periods of drought and at times of actual deficiency. The holard
at Burlington was favorable until June, but after this time marked defi¬
ciencies were of frequent occurrence. However, the value of water-content to
the plant is not determined entirely by its quantity, but largely also by the

INTRODUCTION

Table 6 .—Holard in excess of the hygroscopic coefficient at the several

stations, 1920.

Lincoln, Nebraska.

Date.

0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
feet.

3 to 4
feet.

Apr. 10.

17.6

16.1

14.1

10.0

8.6

Apr. 21.

17.5

22.4

20.3

May 5.

20.5

19.1

19.7

June 9.

18.7

20.3

June 16.

5.9

12.5

14.8

15.6

June 23.

4.9

7.1

9.6

July 14.

22.0

16.8

9.1

13.2

14.7

July 28.

7.2

3.9

6.3

Aug. 5.

2.5

4.0

5.4

9.7

11.6

Aug. 12.

8.7

5.7

4.8

Aug. 19.

26.3

7.7

3.2

Aug. 31.

Continued heav 1

v rains: no sanmles taken.

Wilting coef.

14.0

12.9

12.6

10.5

9.1

Hygroscopic coef.

9.5

8.7

8.6

7.1

6.2

Phillipsburg, Kansas.

Date.

0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
feet.

3 to 4
feet.

May 7.

14.7

15.3

12.5

14.7

14.0

June 2.

4.8

6.6

11.6

11.8

13.0

June 10.

7.6

9.7

9.0

June 24.

6.6

3.1

5.0

9.0

July 1.

1.9

4.1

2.8

July 9.

1.5

2.8

1.6

July 21.

-1.4

0.1

-0.2

2.3

5.3

Aug. 4.

0.7

7.7

-0.3

0.5

3.5

Aug. 18.

12.6

12.3

5.4

Aug. 26.

4.0

4.1

0.4

2.0

5.5

Wilting coef.

13.3

13.4

13.5

13.1

Hygroscopic coef.

10.6

10.9

10.6

10.7

Burlington, Colorado.

Date.

0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
feet.

3 to 4
feet.

Apr. 15.

16.7

13.7

11.1

4 9

1 8

June 3.

2.3

5.2

7.3

6.9

2 9

June 12.

-1.8

-0 1

2 8

June 25.

7.4

2.5

1 8

1 4

-2 5

July 2.

-1.6

-0.5

0 0

July 8.

-2.9

-1.3

-2.2

July 20.

-0.7

-2.7

-1.7

2.5

-0.6

Aug. 5.

4.6

2.7

-2.7

-0.6

-3 1

Aug. 19.

-3.1

-0.5

Aug. 24.

-0.8

-2.1

-1 9

-0.6

14 5

-1 4

Wilting coef.

13.3

14 0

Hygroscopic coef.

10.9

12.2

12.0

11.4

CLIMATES.

rate of loss both through the plant and by surface evaporation. These in
turn are controlled by humidity as affected by temperature, wind, etc., all of
which are more or less integrated in the evaporation.

June July August

the numbers denote the first, second, third, or fourth week in the month,
respectively.

Fig. 4. —Average day (heavy lines) and night (fight fines) temperatures, at Lin¬
coln (solid fines), Phillipsburg (long broken fines), and Burlington (short
broken fines), 1920.

Temperature.

Because of differences in elevation, which more than offset those of latitude,
spring usually opens about 7 to 10 days later at Phillipsburg, and 18 to 23

INTRODUCTION.

days later at Burlington, than at Lincoln. These stations are 800 and 3,000
feet higher respectively than Lincoln. Air temperature was usually 5° to
7° less at Burlington than at Phillipsburg, and also lower at the latter
than at Lincoln. Likewise, the average day temperatures were highest at
Lincoln and lowest at Burlington, and this same general relation held for
average night temperatures (figs. 3 and 4). The night temperatures at
Burlington varied from 45° to 67° F. These were more or less unfavorable
to growth, but their chief effect was probably through humidity, differ¬
ences of temperatures as such probably having little effect upon the type of
grassland.

Fig. 5. —Average daily soil-temperatures at Lincoln (solid line), Phillipsburg
(long broken lines), and Burlington (short broken lines), 1920.

Soil temperatures were highest at Lincoln (70° to 77° F.) and lowest at
Burlington (64° to 72° F.) during the first half of June, but by the last week
this relation was reversed, the soil at Burlington remaining warmest through¬
out the season (fig. 5). The average weekly differences were often as great
as 6° to 8° F., the soil at Lincoln being coldest, at Phillipsburg intermediate,
and the dry soils at Burlington having the highest temperatures.

The much greater daily range of both temperature and humidity at the
Burlington station as compared with that at Lincoln is shown in figure 6,
and the average day and night humidities for 1920 are given in figure 7.
Conditions at Phillipsburg were more or less intermediate. This combination
of high temperature and low humidity, which occurs rather regularly in the
afternoons at Burlington, when coupled with dry soil, obviously promotes
transpiration and depresses growth.

Wind and Evaporation.

The wind movement is much greater at Burlington than at either of the
other stations and is an important factor in desiccating both plants and soil.
An average day velocity of 8 or 10 miles per hour at a height of 0.5 meter is
quite usual, while periods of several days with a velocity of 20 to 30 miles per
hour are not uncommon. The amount of wind is less at Phillipsburg and
much less at Lincoln (for example, a daily average of 4 miles per hour from

CLIMATES.

July 13 to September 19, 1916). The evaporation, which in a measure inte¬
grates the factors of radiant energy, humidity, and wind movement (fig. 8),
was greatest throughout the season of 1920 (23 to 60 c. c. average daily
evaporation from white cylindrical non-absorbing atmometers) at Burlington,

THURSDAY FRIDAY. SATURDAY. SUNDAY.

Fig. 6.—Hygrothermograph records from Lincoln (upper) and Burlington (lower), June 1921; light

lines temperature, heavy lines humidity.

intermediate at Phillipsburg (11 to 32 c. c.), and least at Lincoln (9 to 25
c. c.). Similar constant differences of evaporation rates were obtained the
following seasons.

INTRODUCTION.

In consequence, the conditions for plant-growth as regards rainfall, holard,
temperature, humidity, wind, and evaporation are normally most favorable
at Lincoln, intermediate at Phillipsburg, and least favorable at Burlington.
These conditions are indicated by the native vegetation and borne out by
the growth of crop plants (Weaver, Jean, and Crist, 1922).

June July August

Fig. 7. —Average day (light lines) and night humidity (heavy lines) at Lincoln
(solid lines), and Burlington (broken lines), 1920.

Fio. 8.—Average daily evaporation at Burlington (short broken lines), Phillips¬
burg (long broken lines), and Lincoln (solid line), 1920.

METHODS.

DETAILED METHODS.

Germination of Seed.

In these experiments the fruits and seeds of a large variety of grasses’
forbs, 1 shrubs, and trees from a wide range of habitats were employed.
These were collected in quantity, when ripe, chiefly at Lincoln, Nebraska
City and Halsey, Nebraska; Phillipsburg, Kansas; and Burlington and
Colorado Springs, Colorado. In addition, others were secured from the
vicinity of Berkeley, California; Tucson, Arizona, and in lesser quantities
from other sources. After drying, the seeds were placed in large cloth bags
and kept over winter in a dry, well-ventilated, unheated building at Lincoln.
A new crop of seeds was secured each season. After threshing and cleaning
the seed, germination tests were conducted in pots and flats n the greenhouse
of the University of Nebraska, and to a less extent in moist chambers in the
laboratory. This was primarily for the purpose of eliminating the species
that germinated rarely or not at all, for it proved possible to sow seeds even
with a low viability so thickly as to secure a good stand. Seeds from plants
growing in sand, viz, Calamovilfa longifolia, Muhlenbergia pungens, Redfieldia
exuosa, etc., were germinated in sandy soil.

The vitality of the seeds differed greatly from season to season, as would be
expected. Among various causes for this, climatic conditions at the time of
anthesis may have been a potent factor. In some cases, early frosts on low¬
lands proved harmful to the unripe seeds of late species of grasses. While
the 1920 crop of Kuhnia glutinosa gave no germinable seeds, that of 1921
yielded 84 per cent. This is the highest record for any species, values of 20
to 25 per cent ( Psoralea tenuiflora, Ratibida columnaris, Stipa viridula,
Amorpha canescens, Desmodium canescens, and Liatris scariosa ) being very
good, while a germination of 10 to 15 per cent was quite usual. Some of the
legumes germinated very slowly. A few species that did fairly well in the
greenhouse {Amorpha canescens and Aster multiflorus) gave no germination
under field conditions (1922).

Depth of Planting.

In dealing with seeds so variable in size and natural depth of planting
(viz, Sporobolus asper, Aster, and Solidago as compared with Stipa spartea and
Gleditsia triacanthus), great care was exercised to place the seed at such a
depth as to favor germination and establishment. The results of a single
greenhouse experiment upon this phase of the problem are of interest (table 8).
In this experiment 50 seeds of each species were planted at various depths in
loam soil of good tilth, except where the seeds, because of small size or low
viability, were used in larger but equal (measured) numbers. Sandhill species
were grown in sandy loam.

These results show that depth of planting has a pronounced effect upon
germination and establishment. This is due in most cases to the actual
failure of the seed to germinate at increased depth of soil and in others to its
inability to grow through the greater thickness of soil and still have sufficient

1 The term “forb” is here used for herbs other than grasses in order to afford a clear cut dis¬
tinction and at the same time avoid an awkward phrase. It is derived from the Greek-Latin

root which appears in 4°pP4. an d herba (*ferb—).

INTRODUCTION

Table 7. —Germination of seeds . 1

Species.

1919.

1920.

1921.

1922.

1923.

Ar>pr neonmdo^.

Fair.

Acer saccharinum.

Fair.

Seeds fro-

Good....

A ornnvrnin erlaiieiiTn . . .

zen; no
crop

Poor.

Good....

None.

Arrmmlia canescens.

Good....

Andropogon con tortus 5 .

Poor.

fnrnftt.ns .

Poor . _ .

Poor.

halli .

Poor.

nnt.a.ns .

Poor ....

Good. . . .

Fair.

ReonariiiR.

Poor . . .

Fair.

None. . . .

None....

Poor.

Arcremono nlatvceras 1 * 3 .

Poor.

Ariatida. rmrnnrpa 1 .

Good....

Poor.

Excellent.

Good....

Excellent.

Aater mnltiflorus.

None....

Fair.

None.

.aalieifoliiiR.

None....

Astrn.palns crassicarpus.

None. . . .

Fair.

Roiiteloiia. bromoidea 6 .

Good....

erionoda, 3 .

None. . . .

gracilis.

Good....

Excellent.

Poor.

Excellent.

hiranta.

Poor . . .

Excellent.

Fair.

Good....

None.

racemosa.

Fair.

Poor.

Excellent.

Poor.

racemosa 5 .

Good....

Fair.

rnthroeki 5 .

None....

Rrannpria pallida.

None. . . .

Poor.

None.

Bulbilis dactyloides.

Fair.

Calamovilfa longifolia®.

Fair.

Very poor

Poor.

GaRRia chamaecrista.

Fair.

Cornus canadensis 2 .

Very poor
Very poor

Gorylna n.mericana 2 .

Desmodium canescens.

Excellent.

Elymus canadensis.

Good. . . .

Excellent.

Fraxinus lanceolata 2 .

Poor.

Fair.

Gleditsia triacanthus.

Fair.

Good....

Glycyrrhiza lepidota.

None. . ..

None....

Helianthus petiolaris 3 .

None. . ..

rigidus.

None....

None. . . .

None.

Hilaria rigida 5 .

None. . . .

Koeleria cristata.

Excellent.

Kuhnia glutinosa.

None....

Excellent.

Lespedeza capitata.

Poor.

Fair.

Liatris punctata.. .

Excellent.

Good....

Good.

scariosa.

Excellent

Excellent.

Good.

Muhlenbergia pungens 6 .

Fair.

Onagra biennis.

Fair.

Excellent.

Panicum virgatum.

Poor

Fair.

None. . . .

None....

Norte.

Petalostemon candidus.

Fair.

Excellent.

Pdnus ponderosa 7 .

Good. . . .

Good....

Psoralea tenuifiora 3 .

None. . . .

Good....

Ratibida columnaris 3 .

Good....

Redfieldia flexuosa®.

None....

Fair.

Rhus glabra.

None....

Robinia pseudacacia 2 .

Good. . . .

Good....

Salvia pitcheri.

None.

Solidago missouriensis.

None....

Fair.

serotina.

None. . . .

None....

rigida.

Fair.

Sporobolus cryptandrus 3 .

None....

Good....

asper.

Excellent

Good . . ,

Excellent.

Stipa comata 3 .

Poor.

Good....

Poor.

spartea.

None....

Poor.

Poor ....

None.

viridula 3 .

Good. . . .

None. . . .

Fair

Excellent.
None....

Good.

Symphoricarpus occidentalis.

vulgaris.

None.

Ulmus americana.

Good....

No crop,
frozen

Good....

1 All seeds were collected at Lincoln, unless otherwise indicated.

1 Seed from Pennsylvania. * Seed from Arizona.

1 Seed from Colorado Springs. « Seed from sandhills, Halsey, Nebraska.

4 Seed from Burlington. 1 Seed from Wyoming.

METHODS.

reserve material to establish itself. The decrease in the number of seedlings
that appeared above ground was due in part to this cause, but especially to
damping-off and other abnormal conditions arising under control. With the
grasses a depth of planting exceeding 0.5 to 1 inch was distinctly detrimental,
and most of them, like the composites, did best at 0.12 to 0.25 inch. Un¬
doubtedly, aeration and frequency and amount of precipitation exerted a
profound effect upon ecesis in the field. Even under uniform conditions,
some seeds, e. g., Elymus canadensis , germinate and grow rapidly, while others
are much slower, as w r ould be expected.

Methods of Sowing and Planting.

Seeds and fruits of plants were sown under three different sets of conditions
at the several field stations, namely, surface-seeding, trench, and denuded
quadrat. Surface-seeding consisted of selecting areas with a typical cover of
vegetation in which wooden stakes bearing the names of the species were
driven. Near the stakes and without disturbing the surface, a small quantity
of seed was scattered. A little debris was added to simulate conditions
during fall and wdnter after the seeds have fallen naturally from the plant.

Table 8. —Germination and depth of planting.

Species planted
Mar. 24.

Depth

inches.

No. of plants.

Apr.

14.

Apr.

25.

Remarks.

Aristida purpurea....

0.25

3 lvs. Apr. 25.

0.5

Same size as those at 0.25 inch

depth.

Agropyrum glaucum 1 .

0.5

4 to 5 in. tall with 3 lvs., Apr.

40 •

Same size as preceding.

All smaller than preceding.

Amorpha canescens...

0.25

Smaller than preceding.

Andropogon halli 1 ....

0.25

2 in. tall, 3 lvs., Apr. 25.

0.5

Same as above.

Do.

Nearly as large.

Andropogon nutans...

0.25

Tillering.

0.5

As large as preceding.

Bouteloua racemosa 1 ..

0.12

Progressively smaller to the

0.5

maximum depth.

Bouteloua hirsuta 1 ....

0.25

3 or 4 leaves; 1.5 in. tall.

0.5

Desmodium canescens.

0.25

Largest, with 5 leaves.

Smaller than above.

As large as 1 inch.

'Number of seeds not counted.

INTRODUCTION

Table 8. —Germination and depth of planting —Continued.

Species

planted

Mar. 17.

Depth

inches.

No. of plants.

Mar.

27.

Mar.

29.

Apr.

14.

Apr.

25.

Remarks.

Elymus cana-

0.12

Poorest.

densis. 1

0.5

About same size.

Smallest.

Kuhnia gluti-

0 12

Thick.

nosa.

0 5

Thick.

Smaller than

above.

A few.

Thick.

Do.

Liatris scariosa.

0.25

4 to 5 in. tall.

Liatris punc-

0.25

2 to 3 in. tall.

tata.

Muhlenbergia

0.12

2 tillers; 3.5 in.

pungens. 1

tall.

0.5

Smaller.

Smallest.

Onagra bien-

0.12

Very

Excessive com-

nis. 1

thick.

petition.

0.5

Very

Much competi-

thick.

tion.

Largest rosettes

of all.

Very small.

Petalostemon

0.25

2 in. tall.

candidus.

0.5

0.5 in. tall.

Smaller, very del-

icate.

Ratibida col-

0.25

Rosettes with 4

umnaris.

to 5 lvs.

0.5

Smaller.

Sporobolus as-

0.12

3 to 4 in. 4 lvs.

per. 1

0.5

Large as above.

Pot waterlogged.

Smaller than

above.

Stipa viridula..

0.25

4 to 5 in. 3 lvs.

Do.

Much smaller.

1 Number of seeds not counted.

This was sufficient to keep the seeds from being blown away. Upon germina¬
tion, seeds sown under such conditions meet with keen competition from the
established vegetation for water and nutrients below ground and for light
above. The degree to which light intensity is reduced under the grassland
cover has never been adequately emphasized, but frequently it is as low as
5 to 10 per cent, even early in the summer.

METHODS.

The trench method of seeding was used in order to remove competition
both below and above ground for longer or shorter periods. Long, narrow
trenches, approximately 4 inches wide and deep, were cut in the native sod.
The soil from the trench was then thoroughly pulverized and all of the larger
roots and rhizomes carefully removed, after which it was again replaced in
the trench Enough additional surface-soil was added so that when firmly
compacted the level of the trench was the same as that of the surrounding
surface This detail is of fundamental importance, since the amount of run-off
or the amount of water running into the disturbed area is thus determined.
Stakes bearing the species name were then driven in place at about 12-inch
intervals and the seeds planted, the soil firmed, and a little loose surface-soil
added as a mulch. The rapidity with which rhizomes and neighboring roots
invade this newly prepared area varies considerably with the type of vegeta¬
tion and climate, as also the rate at which the grasses arch over the trench
and reduce the light intensity

In the case of denuded quadrats, the native sod was removed to a depth of
4 or 5 inches over an area about 2.5 feet long and 16 inches wide. The soil was
pulverized, the plant parts removed, and the surface brought to such a height
as to be even with the surrounding sod after settling. One and sometimes two
species were planted in each quadrat thus prepared. Rhizomes of Corylus
americana and Symphoricarpus vulgaris and occidentalis were transplanted in
1921. Obviously, root and rhizome invasion from the sides would affect the
area in the quadrat much more slowly than that in the trench, and this was
also true of shading. Consequently, the quadrat method obviates both root
and especially shoot competition for a considerable period, but certain other
factors more or less unfavorable to growth are introduced.

Since the most critical period in the life of the plant is that of ecesis, the
method of seedling transplants was also used. Seeds were germinated and the
seedlings grown in an unheated greenhouse in loam in 2.5 to 4 inch flower¬
pots. The larger pots were used for those species which grew most vigorously
and whose roots penetrated deepest. Planting was timed in such a manner
that the seedlings were well established and 3 or 4 weeks old before trans¬
planting them into the grassland, usually early in May. At this time trenches
were prepared in the way already described, except that the subsoil in the
trench was well watered before replacing the soil. The trench was half filled
with soil, and the seedlings were then transplanted without injury by simply
removing the entire contents of the pot as a core, placing this at the proper
level in the trench, and firming the soil about it. A sufficient number of
plants were available so that any injured in removing the pot were discarded.
After watering again, the trench was brought to the proper level and a thin
layer of loose soil applied. If the weather required it, the trenches were
watered for a 10-day period during establishment. As with the other methods
of planting, the plants were thinned from time to time to reduce competition
when necessary. All the seedlings were grown at Lincoln and transported to
the other stations as needed.

Finally, a fourth method was used, which consisted of transplanting mature
plants or sods. Blocks of sod about 10 inches square were selected and cut
with almost vertical sides to a depth of 8 or 10 inches and removed with the
minimum disturbance of the underground parts above the level. The size

INTRODUCTION.

varied somewhat, depending upon the root-habit and nature of the soil.
In transplanting Bulbilis dactyloides, Bouteloua gracilis , Muhlenbergia gracil-
lima, and other species with fine, dense surface roots, the blocks were often
larger but not so deep. Conversely, transplants from sandy or gravelly soil
were necessarily smaller. Many forbs were also transplanted, care being
taken to wrap each block securely in burlap for transportation. By this
method of sod transplants, species were planted reciprocally among the
different stations; for example at Lincoln, Spartina cynosuroides was trans¬
planted from swamp to salt-flat, low prairie, high prairie, and gravel-knoll,
care being taken to replant a control block in the swamp for a check. In
transferring the blocks of sod, great care was taken to make a hole of the
exact size, so that the transplant was in good contact at both the bottom and
sides, just enough loose soil being a*dded and firmly tamped around the block
to make the contact complete. The sods were placed 2 or 3 feet apart.
Ordinarily, root and rhizome growth had more or less obscured the lines of
disturbance after the first season, unless there were marked differences in
soil color or texture.

As far as possible, transplants were made preceding or coincident with the
renewed growth and at a time when the holard was distinctly favorable.
Moreover, green parts were clipped back, often level with the soil, in order to
reduce water-loss. The surplus food materials of the older roots, rhizomes,
etc., rendered vigorous growth possible, and in the main this was the most
successful method employed, early-blooming species like Koeleria cristata and
Stipa spartea alone sometimes failing to seed the first season. The others,
such as Agropyrum glaucum and Bulbilis dactyloides, not only fruited, but often
also extended their area by vigorous propagation.

2. EXPERIMENTS DURING 1920.

HIGH PRAIRIE, LINCOLN.

Surface Sowing. •

Surface sowings of 6 species were made on the high prairie, April 16. The
seed was obtained at Lincoln, as in all cases, except as otherwise stated,
Stipa viridula from Colorado Springs being the only exception in this instance.
Koeleria cristata and Stipa alone had germinated by May 15, and Andropogon
nutans and A. scoparius by June 2, while the seeds of Elymus canadensis and
Aristida purpurea (seeds from Burlington) did not germinate. By June 15,
Stipa and A. scoparius were mostly dead or dying, and from this time there
was a steady decline until August 30, when all had succumbed.

It is interesting to note that, although heavy rains occurred immediately
after sowing and at two other intervals, only two species had germinated by
May 15. Holard determinations show no indication of failing moisture in the
surface 6 inches until the middle of June (table 6), but this is no criterion of
the alternate periods of wetting and drying undergone by the seeds in the
surface inch of soil. The drought during June was marked, no efficient
precipitation falling from the 7th until the 25th; in fact, the rainfall was less

Fig. 9.—Monthly precipitation at Lincoln, 1920-1922; the monthly mean is shown

by heavy cross-bars.

than half of the normal (fig. 9). Drought weakened the seedlings, but not to
such an extent that they were unable to survive during the favorable growing
weather of July and August, and competition for light undoubtedly played the
decisive role. By late in May the light intensity at the soil surface (where the

EXPERIMENTS DURING 1920.

prairie was burned the preceding spring) was reduced to 12.5 per cent, with a
range of 2.5 to 26 per cent as determined by the stop-watch photometer, and
later in the season this was much further decreased.

Trench Sowing.

The various species sown by this method all germinated, except Koeleria.
The June drought resulted in the loss of 6 of the 12 species, nearly all of which
were in fine condition early in the month. Five of the six remaining species
not only survived the summer, but became permanently established. The
failure of the four introduced species is of interest, especially since the holard
was quite favorable for growth throughout the season (table 6). The attenu¬
ated condition of certain species noted was due to competition for light; by
midsummer (July 25) the light values were only 20 per cent, even at a height
of 4 inches above the surface, under a moderate cover.

Quadrat Sowing.

With three exceptions, all of the species sown on the surface or in the
trench were also planted in denuded quadrats. Two plantings were made, one
on April 16 and another on May 4 (table 58). Although the June drought
made its impress upon the seedlings, all but three species, Koeleria cristata
Stipa comata, and S. coronata, survived the summer. In fact, the following
season showed that all of these were permanently established, except the
California stipas, which were winterkilled. It is interesting to note that
grasshoppers, while not attacking the native seedlings, greedily ate the intro¬
duced ones (except Stipa viridula). This alone might constitute a biological
barrier. Little difference was noted in the success of the seedlings planted
at different dates. At both plantings the soil was quite moist and good rains
soon followed. Even a short period of drought after germination is often
disastrous, as can be understood from a study of the life-history of the seed¬
lings. Since it is extremely difficult to recover with certainty the delicate

Fig. 10.—Plants of Bouteloua gracilis (A), Bouteloua hirsuta (B),
and Sporobolus asper (C), 44-days old; scale 1 foot. >

HIGH PRAIRIE.

roots among the dense tangle of the native sod, plants for this purpose were
grown in a similar soil free from vegetation. In early growth these were
comparable to those in the quadrats, though differences occurred later.

Root Habits of Seedlings.

Plants for root study were planted in fertile cultivated soil on April 20.
The roots were first examined on June 3 and 4, when the plants were about 44

Fig. 11 .—Bouteloua gracilis 3 months old.
Fig. 12 .—Bouteloua hirsuta 3 months old.

days old. The plants of Bouteloua gracilis, Sporobolus asper, and Bouteloua
hirsuta were 3 or 4 inches tall, each had 3 to 5 leaves, and most of them were
beginning to tiller (figs. 10, a, b, c). The position of the seed and the pro¬
nounced branching of the single primary root are clearly evident, and the
depth of 7 to 11 inches so early attained is significant. Tillering occurred
simultaneously with the development of a secondary root system, a phenome¬
non which is common also to the cultivated cereals (Weaver, Kramer, and

EXPERIMENTS DURING 1920.

Reed, 1923). The period of tillering, before the new roots have become well
established in the moist layers of soil, is a very critical one for the plant,
drought at this time often causing great mortality. Where a little erosion
occurs, it may be plainly seen that the seedling is literally hanging on to life
by a single thread.

By July 11 to 14 the root development of Bouteloua gracilis, under the favor¬
able growth conditions and lack of competition, had a height of foliage of about
8 inches, while flower-stalks 12 to 18 inches tall bore spikes of flowers almost
in bloom. However, the plants in the quadrats were only 4 to 6 inches tall.
The primary root was still plainly distinguishable from the abundant widely

MIXED PRAIRIE.

spreading or deeply penetrating ones of the secondary system. The working-
level was about 20 inches; some roots reached depths of 33 inches (fig. 11).
Bouteloua hirsuta was even more deeply rooted than the preceding, although
its above-ground parts were scarcely so well developed (fig. 12). The root
penetration of Stipa spartea was similar to that of the gramas, though the
two or three roots of the primary system alone had penetrated deeply, the
larger fleshy branches of the secondary system mostly ending at a depth of
about 8 inches. The plants were 8 to 10 inches tall; the parent plants had
about 5 leaves and some were well tillered (fig. 13).

The development of Agropyrum glaucum is shown in figure 14. Aristida
purpurea at this time had formed heavily tillered clumps 6 to 8 inches tall,
which were only slightly better developed than those in the prairie quadrats.
The roots had a working depth of 20 inches and a maximum penetration of 32
inches. Plants of Elymus with a height of 12 to 15 inches (compared with 5
inches in the quadrats) showed a similar root depth. Stipa viridula, which
also grew more vigorously in these plats, with a height of 8 to 10 inches, had a
working level of 16 inches, while the longer roots penetrated 33 inches into the
moist soil. These data show clearly why a species, after the hazards of the first
season are past, has much greater chances of survival.

Summary of Planting Results.

With respect to results on high prairie, the germination on the surface, in
the trench and in the quadrats was 66, 92, and 100 per cent respectively, and
the survival, 0, 45, and 80 per cent. Most of the introduced species were
badly damaged by grasshoppers, although some in the quadrats survived the
season but were winterkilled. Aside from these, all the species that persisted
to the end of the first season became permanently established. Andropogon
scoparius, A. nutans , and Bouteloua gracilis grew well in both trench and
quadrat.

Sod Transplants.

Sods of the following species were planted on April 23, mostly in triplicate.
All showed some drought effects during June, such as rolling of the leaves, but
most species grew normally and all survived. Among those from the high
prairie, Agropyrum glaucum, Elymus canadensis, Koeleria cristata, and Stipa
spartea produced seed, while Andropogon scoparius and Bouteloua racemosa
failed to form flower-stalks. Bouteloua gracilis from the gravel-knoll did not
blossom, but Distichlis spicata from the salt-flat seeded profusely. Andro¬
pogon nutans was the only species from the low prairie which did not fruit,
A. furcatus, Poa pratensis , Panicum virgaium, and Spartina cynosuroides all
producing seed.

MIXED PRAIRIE, PHILLIPSBURG.

Surface Sowing.

Seeding was done on May 7 with the same six species as those used at
Lincoln. When examined on June 10, three species only had germinated,
but these were quite abundant. By July 1, two, Koeleria cristata and Stipa
viridula, had died, while Elymus was represented by a few slender narrow¬
leaved plants with dead tips, which succumbed by July 10. Good rains

EXPERIMENTS DURING 1920.

occurred soon after planting and at two later periods in May. However, an
examination of the rainfall data shows that the precipitation for June was much
less than half of the normal, as at Lincoln (fig. 15). No rains sufficient even
to wet the surface-soil fell after June 18 until the 14th of July, when drought
again prevailed until late July (table 6). The attenuated condition of the
longer-lived species was due to the competition for light. An average value
of 28 per cent (range 2.8 to 80 per cent) was determined for this station on
June 11.

Fig. 15. —Monthly precipitation at Phillipsburg, 1920-1922; the monthly mean is

shown by heavy cross-bars.

Fig. 16.—Monthly precipitation at Burlington, 1920-1922; the monthly mean is

shown by heavy cross-bars.

SHORT-GRASS PLAINS.

Trench and Quadrat Sowing.

The behavior of the same species, sown in the trench, is summarized in
table 9. While all germinated, two died by July 1, another by August 4, and
only two, Andropogon scoparius and A. nutans , grew throughout the season
and became permanently established. These plants, with others, did much
better in the quadrats. While all but Arislida germinated by July 1, several
showed a high mortality due to drought, although only two species had died.
Conditions became drier during July and only six species were represented by
the end of the summer. All of them became permanently established, Bou-
teloua gracilis forming flower-stalks the first season.

Thus, there was a mortality of 100 per cent among plants where the seed
was surface-sown, while 33 per cent survived in the trench, and 67 per cent
in the quadrats. Andropogon scoparius and A. nutans did well in both trench
and quadrats, but Bouteloua gracilis alone flowered the first year. No sods
were transplanted at this station.

SHORT-GRASS PLAINS, BURLINGTON.

Surface Sowing.

The account of the growth of surface-planted species at Burlington is brief.
Of the 13 species sown on April 15, evidences of germination were found for
only 6 on June 11, and the plants of two of these species were dead. By
July 2, three species alone remained, and one of these, Liatris punctata, died
before the end of the summer. Bouteloua gracilis merged into and became
scarcely distinguishable from the native sod; Stipa viridula tillered heavily
and was over 6 inches tall.

Precipitation records show that showers (0.25 to 0.63 inch) fell in April after
seeding, but this was followed by no efficient rainfall (0.15 inch or more)
until the middle of May. Another drought period intervened until the 27th,
when 0.25 inch of rain fell, after which drought ensued until June 18, all of this
serving to emphasize the unfavorable conditions to which the plants were
subjected (fig. 16). The almost constant lack of rainfall throughout the
summer was enhanced by the frequent high winds, low humidity, and high
evaporation (figs. 7 and 8).

Trench and Quadrat Sowing.

Thirteen species were sown in the trench on April 15. Evidence that
Andropogon nutans and Elymus canadensis had germinated was found, but
both species had died and none germinated later. With one or two exceptions
the same species that were sown on the surface and in the trench were also
planted in denuded quadrats. Their growth throughout the season is shown
in table 60. Seven of the 13 species germinated; all but four were dead by
July 2, and only three species, Andropogon nutans, Stipa viridula, and Liatris
punctata, became established.

A summary shows that 33 per cent of the surface-sown species that germi¬
nated survived; all of those in the trench succumbed, while in the denuded
quadrats 43 per cent lived through the season and as usual became established.
Stipa viridula survived both on the surface and in the quadrats, Bouteloua
gracilis on the surface, and Andropogon nutans and Liatris punctata in the
quadrats.

EXPERIMENTS DURING 1920.

Sod Transplants.

Blocks of sod of the following species, mostly in quadruplicate, were secured
from the several stations at Lincoln and planted at Burlington: Agropyrum
glaucum, Andropogon scoparius, A. furcatus, Bouteloua racemosa, Elymus
canadensis, Koeleria cristata, Panicum virgatum, Poa pratensis, and Stipa
sparlea. Before planting the sods in the holes made in the short-grass mats,
these were filled with water, which was allowed to sink into the deeper layers
before the block was tamped in place. The grasses were planted in two long
rows, each having two blocks of each species, and several feet of unbroken
sod thus intervened between the blocks. After planting, all were thoroughly
watered, and this was repeated on one row at intervals of 2 or 3 weeks through¬
out the season, as proved necessary. It was found later that the watered
row, which had been planted only 2 or 3 feet from a broken area, had received
some run-off water from the furrow separating the broken field from the
grassland.

Although none of the transplants died, the drought affected the imwatered
row in a number of ways, while even the watered plants did not attain normal
development. Nearly all of the former exhibited a marked rolling of the
leaves and even severe wilting at times, and the tips on several species as well
as whole plants died back considerably. Elymus, Koeleria, Panicum, and
Stipa flowered under both conditions, but the flower-stalks were usually fewer
and decidedly shorter in the unwatered row, as was true of the foliage also.
Thus, the maximum leaf-height of Panicutn under the two conditions was 13
and 20 inches respectively, while the height of the panicles ranged in the one
case from 7 to 13 and in the other from 19 to 20 inches. Stipa, Koeleria , and
Elymus seeded in both habitats, but Agropyrum and Andropogon furcatus in
neither, while A. scoparius and Boutelaua racemosa each had a single flower-
stalk in the watered area. The ultimate fate of these transplants is indicated
on pages 68-70.

Summary of Experiments.

The results of the 1920 experiments at the three stations are summarized in
table 9.

Table 9. —Summary of experimental seeding, 1920.

Method.

Per cent of germination. 1

Per cent of germinated species
established.

Lincoln.

Phillips-

burg.

Burling¬

ton.

Lincoln.

Phillips-

burg.

Burling¬

ton.

Surface-sowing.

Trench.

100

Denuded quadrat ....

100

Average.

1 As indicated elswhere, it is entirely possible that some seeds germinated, but did not appear
above ground, or died later and disappeared between visits to the stations, though it seems prob¬
able that few remnants were overlooked.

Of the plants surviving, only three, Andropogon nutans, Bouteloua gracilis,
and Stipa viridula, succeeded under at least one condition at all three stations.

CULTIVATED AREAS.

.4 ndrcrpogon scoparius and Bouteloua hirsuta grew at all stations but Burlington,
and Liatris punctata at all but Lincoln. Elymus, Bouteloua racemosa , and
Aristida purpurea lived throughout the season at Lincoln only. In most cases
a somewhat better growth occurred at Lincoln than westward. For example,
Andropogon nutans was 7 to 10 inches tall at Lincoln, 6 to 9 inches at Phillips-
burg, and 4 to 5 inches at Burlington late in August. Among the transplanted,
about 70 per cent flowered at both Lincoln and Burlington, those at the latter
station having been watered. Of the unwatered sods at Burlington, 53 per
cent had flower-stalks, but some had only one each. "While the transplants at
Lincoln developed quite normally, most of those at Burlington were much
dwarfed, even when watered.

SEEDING IN CULTIVATED AREAS AT ALL STATIONS.

True Prairie.

The germination and growth of the preceding species in cultivated soil at
the several stations is instructive. In every case a good seed-bed was prepared
by plowing, disking, harrowing, etc., and a part of the fields used in growing
experimental crop-plants (Weaver, 1920,1922) was set aside for native vegeta¬
tion. "Weeds were removed from time to time by hoeing. In these experi¬
ments the trench method alone was used. The upland station at Lincoln was
located on a broad, level hilltop of silt loam adjoining the upland prairie.

.All but two species, Aristida purpurea and Bouteloua
hirsuta, germinated, and all of the others except
Koeleria came through the first season successfully.

Because of the lack of the root competition which is
regularly encountered by plants in the native prairie,
these plants suffered less from midsummer drought.

The tillering was much heavier and the plants were
in all respects more vigorous. For example, Andro¬
pogon nutans was 7 to 10 inches tall in the prairie
quadrats and 12 to 17 inches in the field on August
30. Elymus was 4 to 8 inches and 10 to 14 inches
tall and Bouteloua gracilis 4 and 9 inches respectively
in the two habitats. Like Bouteloua racemosa , both
were putting forth flower-stalks in the cultivated field.

They all seeded by October, together with Andropogon
scoparius, although the flower-stalks of A ndropogon
and Elymus were dwarfed.

Root Development.

The striking growth made by these native species
in a single season emphasizes the effect of the removal
of competition, as shown especially by the root devel¬
opment, e. g., Andropogon scoparius, with a maximum
penetration of 4 feet (plate 3a). Fully mature plants
excavated in the prairie in similar soil differed mainly
in their greater depth of penetration, about 5.5 feet
("Weaver, 1919:4). Siipa viridula (plate 3b) had a well-developed root system,
a few of the longer roots extending to the 33-inch level, but compared with the
roots of mature plants in the native habit at the root system had scarcely

Fig. 17.—Root of Liatris
punctata at end of first
season’s gr owth.

EXPERIMENTS DURING 1920.

begun. At Colorado Springs they had a lateral spread of over 1.5 feet and
a working level of nearly 11 feet. The roots of Elymus canadensis under¬
went a remarkable development, as shown in plate 4a, and as with mature
specimens, most of the roots were in the first 1.5 feet of soil. However, this
year-old plant exceeded in depth (34 inches) any excavated in grassland,
and especially in the large number and profuse branching of its roots.
Liatris punctata , though only 3 inches tall and with but 2 leaves, possessed
a tap-root that penetrated to a depth of 3.3 feet (fig 17). A considerable
supply of reserve food had been stored in the enlarged portion of the tap,
which had a maximum diameter of 5 mm. Mature plants reached depths
of 6 to 15 feet.

Both tops and roots of Bouteloua racemosa made exceptional growth; one
plant was photographed and excavated on August 25, 1919, only 124 days
after the seed was planted in the greenhouse on April 23, the seedlings having
been transplanted to the field on May 10 (plate 4b). Mature plants reach¬
ed depths of 5.5 feet. Stipa comata and Andropogon nutans attained working-
levels of 18 and 28 inches respectively and maximum depths of 34 and 48
inches. The root habit and extent in both cases were similar to those of
mature plants (plate 4c).

Mixed Prairie and Short-grass Plains.

At Phillipsburg, six species were planted on May 7 in a field adjoining the
prairie station. Andropogon nutans, A. scoparius, and Koeleria cristata did
not germinate, Aristida purpurea grew until midsummer only, while Elymus
canadensis and Stipa viridula both prospered. By July 10 the two tillers on
each plant of Elymus were practically as tall as the parent, but, unlike the
plants at Lincoln, none produced flower-stalks.

At Burlington the plantings were made on the same day as in the grassland,
viz, April 15. The following either did not germinate or died soon after
germinating: Andropogon nutans, A. scoparius, Aristida purpurea, Bouteloua
gracilis, B. racemosa, Koeleria cristata, Liatris punctata, Panicum virgatum,
Stipa comata, and S. setigera. Stipa viridula made excellent growth throughout
the summer, reaching a height of 8 to 12 inches. A thick stand of Elymus
canadensis was obtained, but the plants died during the summer.

To summarize, germination of 83, 50, and 18 per cent respectively was
obtained at the several stations going westward. Among those which germi¬
nated, Koeleria died at Lincoln and Aristida at Phillipsburg. Elymus was
successful at these two stations, but flowered only at Lincoln. Stipa viridula
succeeded at all three stations, doing quite as well or even better westward.
Seven other species grew at Lincoln but not elsewhere. The holard of the
cultivated fields decreased in nearly the same proportion as in the grassland
at the several stations. The response of the seedlings to the different climates
was in close agreement with that of the seedlings and transplants in the
grassland.

EXPERIMENTS AT OTHER CLIMATIC STATIONS.

Subclimax Prairie, Nebraska City.

Character.

The subclimax-prairie station was located near Nebraska City, on a gentle
east slope somewhat over halfway down one of the large rolling hills char-

SUBCLIMAX PRAIRIE.

acteristic of the topography of the region. It was situated in an area of
native grassland of over 125 acres in extent. The soil is a dark-brown fine silt
loam, consisting of glacial materials intermixed with loess. Below 1.5 feet it is
quite yellowish in color and shows more plainly its loess origin. The high per¬
centage of clay and silt is indicated by its hygroscopic coefficient, which is
about 12 per cent; hence it is very retentive of moisture. As already indi¬
cated, the precipitation at this station is approximately 5 inches greater (33
inches) than at Lincoln. This additional moisture promotes the growth of
a more luxuriant grassland vegetation and permits the presence of a large
number of herbs and shrubs found rarely or at least much less abundantly
in true prairie (plate 7b).

ill

Nov Dec

Fig. 18. —Monthly precipitation at Nebraska City, 1920-1922; the monthly mean

is shown by heavy cross-bars.

As shown in figure 18, April had about twice the normal amount of precipita¬
tion. Moreover, this was well distributed and probably accounts for the
high percentage (80) of germination. However, the rainfall of May was far
below normal, while that of June was less than one-third the usual amount,
only two rains falling (0.35 inch on the 15th and 0.70 on the 25th). July was
likewise a dry month.

The dominant grasses are Andropogon nutans, A. Jurcaius, A. scoparius, and
Stipa spartea. Koeleria cristata, Poa pratensis, Bouteloua racemosa, Panicum
virgatum, and Spartina cynosuroides, which runs far up the ravines, are of
lesser importance. The regular occurrence of A. furcatus and Panicum on
highland and their excellent development indicate more favorable conditions
for growth than in true prairie. The seasonal aspects and herbaceous societies
resemble those described for Lincoln (Pound and Clements, 1900; Thornber,
1901). Throughout the grassland, even on the ridges, postclimax com¬
munities of Ceanothus ovatus , Amorpha canescens, and Rosa arkansana occur,
being held in check by annual mowing. Symphoricarpus vulgaris and occi-

EXPERIMENTS DURING 1920.

dentalis, Corylus americana, Cornus asperifolia, Rhus glabra , and Xanthoyxlum
americanum invade the grassland from the vantage-ground of fences, hedge¬
rows, and ravines. In unmown and unpastured areas these with others,
including many trees, form thickets or groves overrun with Vitis, Rubus,
Ampelopsis, Celastrus, Smilax, etc., indicating the approach to climax condi¬
tions.

Results.

On April 2, ten species were sown on the surface of an area mown
the preceding fall. Aristida purpurea and Bouteloua gracilis failed to germi¬
nate. Stipa viridida, S. comata (like Aristida , from seed collected at Colorado
Springs), Koeleria cristata , and Liatris punctata died by midsummer, while
A. scoparius, A. nutans, Bouteloua hirsuta, and Elymus, after suffering some¬
what severe mortality, came through in good condition and became perma¬
nently established.

Table 10. —Holard in excess of hygroscopic coefficient,
Nebraska City, 1920.

Date.

0.0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
feet.

3 to 4
feet.

Apr. 2.

24.2

21.3

18.3

14.2

12.6

June 6.

14.2

16.9

17.9

....

June 14.

16.7

10.6

....

June 21.

8.4

6.7

12.8

....

June 30.

8.9

7.9

....

July 4.

9.8

4.7

7.1

. . . .

July 19.

4.5

5.3

7.3

8.9

9.9

July 28.

2.3

3.4

4.9

. . . .

Aug. 4.

0.3

1.2

3.4

6.1

7.6

Aug. 12.

9.7

2.6

3.1

....

. . , .

Aug. 19.

28.5

3.9

2.4

....

. . . .

Aug. 28.

Hygroscopic

12.4

9.5

5.2

3.3

5.5

coefficient.

12.1

11.7*

12.3

13.7

12.9

The attenuation of the leaves of the seedlings became noticeable early in
June. By this time the average level had reached 11 inches, above which
Stipa (3 feet), Euphorbia corollata, Silphium integrifolium, and various other
herbs occurred. Light values ranged from 2.5 to 40 per cent.

The behavior of the species planted in the trench was exceptionally good.
All germinated and all but four lived through the season, owing to the rather
favorable chresard of early summer, which promoted rapid root growth
(table 10).

The same species planted in the quadrats did much better. All germinated,
became established, and lived throughout the summer. As usual, a fairly
high mortality occurred among most species at the critical period, when the
primary root system was supplying a maximum of transpiring area before the
secondary one became established. As a consequence, the seedlings thrived
best or held out longest about the edges of the quadrats. In solving this
problem a comparison of conditions within the disturbed area, as compared
with that in the undisturbed grassland, is illuminating.

MIXED PRAIRIE. 39

Table 11 .—Comparison of chresard and soil temperatures in denuded quadrats and in adjacent

undisturbed grassland.

Station.

Date.

Depth.

Grass¬

land.

Chresard.

Denuded

quadrat.

Differ¬

ence.

Remarks.

inches.

p. ct.

July 14

0 to 6

22.0

14.0

8.0

Lincoln, high prairie.

..Do. . .

6 to 12

16.8

12.1

4.7

June 14

0 to 6

6.8

11.3

-4.5

r Do...

0 to 6

10.7

19.9

-9.2

Lincoln, low prairie..

. .Do. . .

6 to 12

13.8

20.2

-6.4

June 22

0 to 3

6.9

-0.6

• • • •

,• Do. . .

3 to 6

• • . .

13.3

• ■ • «

June 21

0 to 3

2.0

-4.2

• • * •

Nebraska City.

. . Do. . .

3 to 4

• • • •

17.9

, , . ,

. .Do. . .

0 to 3

• • • •

-2.6

• • • •

Phillipsburg.

May 19

0 to 3

• • . •

5.8

• • • .

. .Do. . .

0 to 6

5.4

12.9

7.5

Temperature.

Station.

Date.

Depth.

Remarks.

Grass-

Denuded

Differ-

land.

quadrat.

ence.

inches.

°C.

Lincoln, high prairie.

May 5

Under surface

27.6

16.2

11.4

1 p. m., clear,

hot.

Lincoln, low prairie. .

. .. Do. . .

. . .Do.

22.5

19.5

3.0

12 noon, clear,

hot.

Lincoln, gravel-knoll.

June 14

26.5

21.2

5.3

2 p. m., clear,

hot.

Lincoln, high prairie.

.. . Do. . .

Under surface

37.0

41.0

4.0

Do.

Lincoln, low prairie. .

. . . Do. . .

34.0

47.0

13.0

2 p. m., burned

area.

Do.

. . .Do. . .

30.0

38.0

8.0

Do.

Lincoln, gravel-knoll.

. . .Do. . .

41.5

46.0

4.5

2 p. m., clear,

hot.

Do.

. . . Do. . .

37.5

41.0

3.5

Do.

Nebraska City.

June 21

Under surface

58.0

38.2

19.8

2 h 30 m p. m.,

clear, hot.

Do.

. . .Do. . .

37.6

32.4

5.2

Do.

It is evident that it is often much drier in the quadrats, the soil hotter, and
the humidity much lower. The average daily evaporation is from 15 to 40
per cent higher in the bare area. These conditions are obviously unfavorable
for seedlings. The partial shade about the edges of the quadrats lowers the
temperature, lessens wind movement, increases the humidity, and undoubt¬
edly accounts for the seedlings growing here when drought kills them else¬
where.

Mixed Prairie, Colorado Springs.

Character.

The station at Colorado Springs was located on a hillside sloping gently
southward and in the midst of a great expanse of mixed prairie, about 2 miles

EXPERIMENTS DURING 1920.

east of the city. As typical for much of the region, the soil consists of a light-
colored loam intermixed with some sand. It is very compact and run-off is
high; at a depth of 6 to 10 feet it is underlaid with sand. The distribution of
the rather meager precipitation (15 inches) throughout the year is very
similar to that for other stations. The holard of the surface foot is about 45
per cent and the echard 8.2 per cent. Studies of the water-content throughout
a number of years show that it is frequently reduced to the hygroscopic co¬
efficient in the surface layers, and the soil may become very dry to a depth
of at least 4 feet. The high evaporation, which is often twice as great as in
true prairie, accentuates drought. The extremes of day and night tem¬
perature are similar to those at Burlington. The high day temperatures
cause low humidity (frequently 5 to 15 per cent) in the afternoon and cor¬
responding water-loss from both plants and soil.

The climate is reflected in the composition of the grassland, both in the
absence of many true-prairie species and the relative dwarfing of the others
that grow in this habitat. As a result of close grazing before fencing the
station, Bouteloua gracilis was quite dominant. Other shorfc-grasses or sedges
were Muhlenbergia gracillima and Carex pennsylvanica. Stipa comala, Andro-
pogon scoparius, Agropyrum glaucum, Aristida purpurea, Koeleria cristata,
Bouteloua racemosa, and Andropogon furcatus all reappeared after cattle were
excluded, the latter only sparsely. Artemisia frigida, Aragalus lamberti, Lupi-
nus argenteus, Senecio aureus, Abronia fragrans, Chrysopsis villosa, Psoralea
tenuiflora, and Argemone platyceras all form extensive societies. Ratibida
columnaris, Opuntia camanchica, 0. fragilis, and Thelesperma trifidum were
somewhat abundant. Most of the grasses are rooted at a depth of 3 to 5 feet,
while many forbs penetrate much more deeply, some to 12 feet. This grass¬
land never reaches the luxuriance of true prairie and consequently light played
a ver}^ minor role in the following experiments.

Results.

The seeds of 9 species were planted in quadrats, but not until June 17.
All were from Lincoln except Aristida purpurea and Stipa viridula. Bouteloua
gracilis, B. hirsuta, and Koeleria failed to germinate, while Liatris punctata
died soon after germination. Andropogon scoparius, A. nutans, Elymus
canadensis, and Stipa viridula became established, but like the others were
watered from time to time.

On June 9 the following sods from Lincoln were transplanted, being abun¬
dantly watered at the time of planting and at several intervals afterward:
Agropyrum glaucum, Andropogon furcatus, A. scoparius, Bouteloua racemosa,
Bulbilis dactyloides, Elymus canadensis, Koeleria cristata, Panicum virgatum,
and Stipa spartea. Although they made a rather poor growth, due partly to
the late period of transplanting, all survived the first season.

EXPERIMENTS AT EDAPHIC STATIONS.

Gravel-Knoll Station.

Character.

A comprehensive series of edaphic stations was located at Lincoln, all within
a distance of a half-mile, ranging from gravel-knoll through high prairie, low
prairie, lowland cultivated field, and salt-flat to swamp.

GRAVEL-KNOLL.

The gravel-knoll station, which was only 300 feet south of the high prairie,
occupied the crest of a steep hill about 60 feet above the valley of Salt Creek
(plate 5a). The soil is a drift and consists of very porous coarse sandy to
gravelly loam with a water-holding capacity of only 40 per cent of its dry
weight. At a variable depth, about 4 feet, it is underlaid with clay. The
hygroscopic coefficients at the several depths, together with the chresard
during 1920, are given in table 12.

Table 12 .—Holard in excess of hygroscopic coefficient.

Date.

0.0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
feet.

3 to 4
feet.

Apr. 21.

13.3

7.3

13.3

May 5.

14.2

5.5

15.1

. . . .

June 9.

4.6

5.1

....

June 16.

-0.2

0.0

4.1

6.6

....

June 23.

-1.1

-0.2

1.8

....

July 15.

-0.9

1.7

2.6

1.1

5.8

July 29.

-1.8

-0.6

8.1

. . . .

Aug. 5.

-1.9

-1.7

0.5

1.4

7.2

Aug. 12.

3.3

-1.3

0.4

. . , .

. . . .

Aug. 19.

8.6

2.3

0.3

....

Aug. 31.

Continued heavy rains; no samples taken.

Hygroscopic

coefficient.

5.2

4.9

3.1

2.8

7.4

The low water-content is due not alone to the nature of the soil, but also to
the steep and generally south slope, which produces a high run-off and greatly
increases the temperature. These factors, together with greater wind move¬
ment and higher evaporation than at the high-prairie station, subject both
soil and vegetation to frequent drought. The area was dominated by rather
open mats of Bouteloua gracilis intermixed with B. hirsuta, which Stipa
spartea, Andropogon, and other tail-grasses were invading with difficulty from
their foothold in the silt loam farther down the slope. However, the transition
from one soil type to the other is quite abrupt and the ecotone correspondingly
sharp, a narrow belt of what is essentially mixed-prairie occupying this area.

Behavior.

Of the seven native species sown in the trench on April 16, Aristida purpurea
and Liatris punctata failed to germinate. Elymus canadensis and Koeleria
cristata were all dead by July 15, together with Stipa coronata and S. setigera,
dominants introduced from California. The single remaining clump of Stipa
viridula was disturbed by rodents and died later, leaving only Andropogon
nutans and A. scoparius as survivors of the first season. Among these the
mortality was very high; the leaf-tips of many plants were dead, while the
best plants were only 4 to 6 inches tall. In the quadrats the same species
did even more poorly, only five germinating. One of these died by June 2,
while all of the others, after a period of leaf-rolling, wilting, and progressive
drying of the leaves, succumbed by the middle of July or not long afterwards.

Sods of Agropyrum glaucum, Andropogon scoparius, Bouteloua racemosa,
Elymus canadensis, Koeleria cristata, and Stipa spartea were transplanted
from high prairie to gravel-knoll on April 20. Andropogon furcatus, A. nutans,
Boa pratensis, Panicum virgatum, and Spartina cynosuroides were also trans-

EXPERIMENTS DURING 1920.

planted here from the low prairie. Duplicate and often triplicate blocks of
sod were used. One lot of each was watered thoroughly at two or three of the
most critical periods during the summer. On June 15, none of the sods having
been watered since transplanting, Andropogon furcatus and Panicum were
badly wilted and the leaves on Andropogon were rolled, while both Koeleria
and Stipa had blossomed earlier than elsewhere and the plants were drying at
the base. Two weeks later, no rain having occurred until the 25th, all of the
unwatered plants had either ripened seeds and were drying {Koeleria, Poa,
Stipa), or had revived as a result of the rains, but were again in a wilted state.
During August, one sod of Poa and another of Andropogon scoparius died,
while the unwatered Agropyrum, Andropogon furcatus and another A. sco¬
parius were nearly dead. The plants with a rather shallow root habit, viz,
Koeleria, Stipa, and Elymus, suffered less, probably because they could com¬
pete more successfully with the gramas for water in the surface soil. However,
Bouteloua racemosa and even Spartina survived in fair condition, although
dwarfed. In addition to the three early bloomers already mentioned, Elymus
and Agropyrum were the only ones that flowered. The ultimate fate of these
species is indicated later (p. 83). The low chresard, or its entire absence during
certain periods of drought, is shown in table 12.

Low-prairie Station.

Character.

This station was located on a level tract at the foot of the hill about 60 feet
below the high-prairie and gravel-knoll stations and only a quarter of a mile
southward (plate 5b). The soil is a fertile, dark-colored silt loam of the
Wabash series. It is very fine in texture, being composed mostly of silt and
clay (tables 13 and 14), and with the Truog test showed no trace of acidity.
The greater amount of volatile matter and the greater nitrogen-content at all
depths than in soil from the upland indicate more favorable conditions for
plant growth.

Table 13. —Mechanical analysis of soil from lowland cultivated area.

Depth of
sample (feet).

Coarse

gravel.

Fine

gravel.

Coarse

sand.

Me¬

dium

sand.

Fine

sand.

Very

fine

sand.

Silt.

Clay.

Moisture

equivalent.

0.0 to 0.5....

0.1

0.4

2.2

1.8

5.0

25.0

41.3

24.3

27.7

0.5 to 1.0 ft. .

0.3

0.7

2.1

2.2

5.0

25.4

38.8

25.8

27.9

1 to 2.

0.2

0.3

1.3

1.5

3.7

21.4

40.8

31.0

30.6

2 to 3.

0.0

0.1

0.4

0.5

1.7

19.2

43.4

34.7

32.9

Table 14. —Chemical analysis of soil from lowland cultivated area.

Depth of
sample
(feet).

Insoluble

residue.

Soluble

salts.

Volatile

matter.

Iron and
aluminium
oxids.

Calcium

oxid.

Magne¬

sium

oxid.

Phospho¬
rus pent-
oxid.

Nitrogen.

0.0 to 0.5..

79.34

12.96

7.70

9.57

0.68

0.75

0.13

0.218

0.5 to 1. . .

79.63

13.66

6.71

10.27

0.63

0.77

0.10

0.187

1 to 2.

78.11

15.83

6.06

12.11

0.64

1.01

0.08

0.135

2 to 3.

74.78

19.82

5.40

15.20

0.76

1.27

0.09

0.082

LOW PRAIRIE.

Moreover, the chresard is constantly greater than on the high prairie,
though the area is well-drained and the soil is never waterlogged.

Table 15. —Holard in excess of hygroscopic coefficient.

Date.

0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
feet.

3 to 4
feet.

Apr. 10.

28.7

22.3

21.8

18.7

18.8

Apr. 21.

26.1

• • • •

22.1

• • • •

May 5.

33.1

25.6

23.9

• • • •

a a • •

June 9.

25.7

24.0

• • • •

June 16.

13.6

16.6

19.5

26.4

June 23.

9.3

11.5

20.4

• • • •

July 14.

27.4

21.1

17.0

19.5

22.7

July 28.

17.6

12.5

15.7

• • • •

a a a a

Aug. 5.

2.3

8.2

11.5

12.0

23.6

Aug. 12.

15.4

12.4

11.8

• • • •

a a a a

Aug. 19.

21.8

8.8

10.1

a • • ■

a a a a

Aug. 31.

Continued heavy rains; no samples taken.

Hygroscopic

coefficient.

11.8

11.1

10.3

10.2

11.9

The low-prairie area is dominated by a few species less xerophytic than
those occurring on the high prairie. Many of the species of the high prairie
are absent, but are replaced in part by others of a more mesophytic kind.
The dominant grasses are Andropogon furcatus, Panicum virgatum, Andropogon
nutans , and Spartina cynosuroides, each of which often covers small areas
with a pure or nearly pure growth. Poa pratensis is also very important, but is
overtopped by the taller grasses, which reach a height of 5 or 6 feet (plate 6a).
Characteristic forbs are Solidago canadensis , S. rigida, Glycyrrhiza lepidota,
Aster multiflorus, A. salicifolius, Physatis heterophylla, Polygonum muhlenbergi,
Artemisia gnaphalodes, Achillea millefolium, and Callirrhoe alcaeoides. The rank
growth forms a dense cover and makes ecesis in the area very difficult. Bisects
show that the plants are rooted deeply, most of them reaching depths of 5
to 12 feet. Root-layers at 3 and 5 feet are quite as distinct as those on the
high prairie.

Behavior.

Surface sowings of eight species were made on April 10. Two failed to
germinate, while Bouteloua gracilis germinated but soon died. By June 15
all were very delicate and much attenuated, owing to the low light intensity.
A month later Liatris punctata had been shaded out and the others were
slender and pale in color. Stipa viridula and Andropogon scoparius died
later, leaving attenuated individuals of A. nutans and Koeleria surviving
the first season. Koeleria had disappeared by the following spring, but A.
nutans made a good growth and merged into the native sod. The light
intensity under the cover of tail-grasses was so reduced as to constitute the
major factor in competition, in view of the high holard. On June 14, 1922,
in a portion of the prairie burned over the preceding spring, the light intensity
under vegetation of average density and with an average height of 12 inches,
was only 6 per cent. Some seedlings were growing in even deeper shade. By
July 25 the average had dropped to 3.5 per cent; under a thin cover it was
12 per cent, but where the vegetation was very dense, only 1.5.

EXPERIMENTS DURING 1920.

Seeds planted in the trench at the same time did somewhat better.
Although three species failed of germination, all of the others, viz, Andropogon
nutans, Bouteloua hirsuta, Elymus canadensis , and Stipa viridula, came
through the season, although somewhat blanched and attenuated. Thus, A.
nutans had reached a height of 7 to 8 inches as compared with a growth of 5 or
6 inches on the high prairie. Light in the trench in June varied from 5 to 10
per cent near the edge to 50 or 60 per cent in the center. By late July this
was reduced to 4.5 to 16 per cent, even at 4 inches above the surface. Of the
eleven species sown in denuded quadrats, all germinated, and all but one,
Stipa coronata , came through the season in excellent condition. In fact,
except for Stipa setigera, which was winterkilled, all became established.
The marked development shows the possibilities of growth in this habitat
when sufficient light is available. Little difference was apparent between
the earlier and later sowings.

A comparison of growth in the denuded quadrats of the prairie and in the
adjoining cultivated area is of interest. Both were seeded on the same dates,
weeds being kept out of both areas. The native vegetation in the field was
consequently free from any competition. Aristida purpurea, Stipa comata,
S. eminens, and S. setigera failed to germinate, as did also the sowing of
Andropogon nutans, A. scoparius, and Koeleria cristata made on April 9. All
others made an excellent growth, even exceeding that in the quadrats. For
example, the height in inches attained by August 30 by typical representatives
in field and grassland respectively were: Stipa viridula, 20 and 5 to 7; Koeleria
cristata, 3 to 4 and 2 to 3; Elymus canadensis, 14 to 16 and 7 to 11. Moreover,
the species in the cultivated area were more heavily tillered and three,
Andropogon nutans, A. scoparius, and Elymus, came into blossom in Sep¬
tember. The root systems were well developed; the roots of Elymus reached
a maximum depth of 4 feet and those of Andropogon scoparius were abundant
to 3 feet, while some extended to 4.5 feet. The roots appeared finer than on
older plants and were not so well-branched near the tips. The ample water-
content of the fertile soil in the cultivated area, which was similar to that on
low prairie, together with the lack of competition, accounts for the excellent
growth.

Blocks of sods of the same species used on the gravel-knoll (except Andro¬
pogon nutans) and secured from the same sources were transplanted into low
prairie from March 22 to April 24. Bulbilis dactyloides from overgrazed low
prairie, Distichlis from salt-flat, and Bouteloua gracilis from gravel-knoll were
also included. The prairie had been mown the preceding year. During the
first season all of the plants flourished, including the short-grasses, Bulbilis
and Bouteloua gracilis. The former produced stolons abundantly, but none
were able to become rooted. Because of light conditions, the height growth
of the shorter grasses was somewhat emphasized, while that of the taller
species was quite normal for lowland. All the transplants flowered the first
season, except Spartina. The effect of a reserve food-supply in the roots and
rhizomes is shown by comparing a single season’s growth of Andropogon nutans
from a block of sod in the low prairie and from seed sown in the lower culti¬
vated plats (plate 8a).

Salt-flat Station.

Character.

This station is located on a level tract of land just below and only 200 feet
south of the low prairie. Owing to a small percentage of sodium chloride which

SALT-FLAT AND SWAMP.

has deflocculated the clay (table 16), it is of a very different consistency
from that of the low prairie (Hall, 1920: 283). Moreover, the adsorption
of the Na ion in part has left the soil in a slightly acid condition, according to
Truog's test.

Table 16 .—Acidity and salt-content at salt-flat station.

Depth of
sample.

Acidity. 1

Total soluble
salts. 2

Chlorides.

p. ct.

p. p. m.

p. ct.

p. p. m.

0.0 to 0.5 foot.

Slight.

0.17

1,700

0.06

600

0.5 to 1.0 foot.

Very slight .

1.14

11,400

.18

1,800

1 to 2 feet....

... Do.

.46

4,600

.10

1,000

2 to 3 feet....

. . .Do.

.50

5*000

.12

1*200

3 to 4 feet....

.. .Do.

.64

6,400

.15

1,500

1 Neutral to litmus in water-extract. 2 No carbonates were present.

The above conditions, as the holard is very similar to that in the low
prairie, have greatly affected the type of vegetation. This consists of an open
growth of salt-grass, Distichlis spicata, which is frequently mixed with a
sparse growth of Agropyrum pseudorepens (plate 6b). In places occur scat¬
tered plants of Poa pratensis; other species, especially of forbs, are practically
excluded. Agropyrum is much dwarfed, and with Distichlis forms a layer
seldom exceeding 5 to 8 inches in height. The sparse flower-stalks are only
8 or 10 inches high, and consequently light plays a minor role in deciding the
fate of transplants.

Behavior.

The same species as in the case of the low prairie and from the same
sources as before, were transplanted into the salt-flats, with the exception of
Bulbilis. Andropogon furcatus, A. scoparius, and Panicum virgatum had been
transplanted during the spring of 1919. The first season they made a poor
growth and none put forth flower-stalks; from the outset in 1920, Stipa
appeared yellow and dwarfed. All of the others suffered more or less severely
during the summer drought. Late in June nearly all showed more or less
wilting and some yellowing. In July the ground was cracked on one or more
sides of the blocks of sods, a phenomenon due to local soil texture and one
which did not occur elsewhere, even in the dry soils at Burlington. However,
none of the transplants died. Compared with their growth in the adjoining
low pi airie, nearly all were greatly dwarfed and lacked vigor and but 6 of the
13 species blossomed. These were Agropyrum, Distichlis, Bouteloua gracilis,
Elymus, Koeleria, and Stipa. The flower-stalks were usually short and the
inflorescence smaller than normal.

Swamp Station.

Character.

The swamp station is located in the valley at the foot of a hill 0.25 mile
north of the high-prairie station. During 1920 water stood above the soil-

EXPERIMENTS DURING 1920.

level throughout April, May, and most of June. For the remainder of the
season the soil was saturated at nearly all times to near the surface. The vege¬
tation consisted chiefly of a rank growth of Spartina cynosuroides, with Scirpus
atrovirens and Heleocharis palustris. Phalaris arundinacea and relict Typha
latifolia, with several species of Juncus and Carex , played a minor role.
Adjoining the wettest area and making the transition to the grassland was a
zone of almost pure Poa pratensis. Sods were transplanted into both areas.
In the latter the water stood above the surface during April and May, while *
throughout much of the remainder of the season the soil was saturated or
nearly so almost to the surface. Vegetation in both areas grew rank, but
aeration was obviously a factor of equal or greater importance than light.

Behavior.

Sods of the same species used on the low prairie were placed in the wettest
portion of the swamp and in the adjoining Poa zone on April 24. Water stood
on the surface of the swamp until after June 15. By June 2, Andropogon
scoparius and Bouteloua gracilis had died, and the leaf-tips were dead on
A. furcatus, Elymus, and Stipa, while Koeleria , which was heading out at 4
to 8 inches high, had most of its leaves dead. Spartina and Panicum were
thriving. By the middle of June the whole area was badly overgrown.
Heleocharis had reached a height of 20 and Scirpus atrovirens 30 inches, while
both Poa and Spartina overtopped all the transplants. Light values ranged
from 2 to 10 per cent. Bouteloua racemosa had died and Stipa was nearly
dead. The basal leaves on Agropyrum and Elymus had died, on the latter to
a height of 9 inches, and Distichlis was much attenuated. The flower-stalks of
Koeleria had rotted off at the base and only a few green leaves remained.
Koeleria, Stipa, A. furcatus, and A. nutans completed the mortality list for the
first season. Spartina had flower-stalks 5 feet tall. A few poor, slender plants
of Agropyrum, Distichlis, Poa, and better ones of Panicum survived. The
stems of Elymus rotted off near the ground, but new shoots came out from the
base. Panicum did not seed. The other survivors, except Spartina, had
gone through a period of anthesis earlier in the season.

Aside from a deficit of oxygen, low light intensities must be considered in
accounting for these results. During July and August a dense growth of Spar¬
tina, Scirpus, Phalaris, etc., to a height of 4 feet or more so completely over¬
shadowed the transplants that they were found with difficulty, and light
values ranged between 1.8 and 7.5 per cent. Transplants in the Poa zone
grew under slightly less unfavorable conditions in regard to both deficient
aeration and reduced light. However, Andropogon scoparius and A. nutans
both died in May and Koeleria and Stipa in July. Some of the other species
had dead basal leaves and were more or less attenuated. All of the survivors
except A. scoparius, A. furcatus, Bouteloua gracilis, and B. racemosa produced
seed.

Sods of Stipa setigera and S. eminens from California were transplanted to
both high and low prairie on April 27. They were watered from time to time
as necessary. During May and early June all showed slight to fair growth, but
by July 15 three had died at each station and the others were growing poorly.
All had died on the high prairie by August 9, while only one of each species
survived on the lowland. The leaves on these did not exceed 2 or 3 inches in

SUMMARY.

height. They remained alive until fall, when they were winterkilled. Sods
of several species from Arizona were transplanted into high prairie on June 2
and were also watered from time to time. By June 15 some plants showed
a new growth, but a month later half of them were dead and the others were
doing poorly. These died during August, with the exception of a single Bou-
teloua, which produced a flower-stalk 20 inches high; however, this plant was
winterkilled.

Summary.

Experiments.

A summary of the planting experiments is given in table 17. Omitting
the data from Colorado Springs, where the plants were watered, the per¬
centage of germination was greatest at Nebraska City and in the low prairie
at Lincoln, as was also that of establishment, if the cultivated area is omitted.
Establishment was lowest, 15 per cent, on the gravel-knoll. Of the species
which frequently failed to germinate, Aristida failed at 6 places of planting,
Bouteloua hirsuta, B. gracilis, Koeleria , and Liatris at 2 or 3 each, and Andro-
pogon scoparius at 1 . Andropogon nutans grew in 9 of the 10 places sown;
A. scoparius, Elymus, and B. hirsuta in 6, while Koeleria and Stipa viridula
grew in 4 of 5. In general, the percentage of germination, as well as that of
establishment, increased from surface to trench to quadrat, and inversely
with competition for water and light.

Table 17 .—Summary of planting experiments, 1920.

ct. of species germinating.

ct. of germinated species
established.

Method of
planting.

Gravel-

knoll.

Low

prairie.

Cultivated

land.

Nebraska

City.

Colorado

Springs.

Gravel-

knoll.

Low

prairie.

Cultivated

land.

Nebraska

City.

Colorado

Springs.

Surface.

•

Trench.

100

Denuded quadrat..

100

Average.

100

Including with these data those from the high prairie at Lincoln, Phillips-
burg, and Burlington, and arranging the stations in order of the average per¬
centage of germination under the different methods of sowing, the sequence is
that shown in column 1 of table 18. Column 2 gives the arrangement based
on the percentage of the germinated species that became established, the
low-prairie station ranking ahead of Nebraska City. The difference between
the percentage of germination in high and low prairie is much less than the
difference in that of establishment. The use of the most favorable method of
planting, i. e., the denuded quadrat, gives practically the same sequence of
stations.

EXPERIMENTS DURING 1920.

Table 18 .—Comparison of germination and growth at all stations } 1920.

Average per cent of
germination.

Average per cent of
establishment.

Per cent of establish¬
ment in denuded quadrats.

Nebraska City. 93

High prairie, Lincoln.... 86

Phillipsburg. 80

Low prairie, Lincoln.... 77
Gravel-knoll, Lincoln.... 71
Burlington. 38

Low prairie. 75

Nebraska City. 70

High prairie. 42

Phillipsburg. 33

Burlington. 25

Gravel-knoll. 15

Nebraska City.100

Low prairie. 91

High prairie. 80

Phillipsburg. 67

Burlington. 43

Gravel-knoll. 0

In accounting for the above results, it remains only to compare the habitats
at Nebraska City, low prairie, and gravel-knoll a little more fully. Those of
the high prairie (Lincoln), Phillipsburg, and Burlington have already been
shown to decrease in all conditions favorable for plant growth in the order
mentioned.

Physical Factors.

A comparison of the holard at the gravel-knoll and Burlington stations shows
that it is not greatly different, both very frequently falling to the hygroscopic
coefficient (tables 12 and 6). In the first 6 inches especially it is often reduced
by surface evaporation to even a lower point. The erosion of the soil about
the seedlings was often pronounced on the gravel-knoll and in this respect
growth conditions were less favorable than at Burlington. The cover of short-
grasses on both areas indicates xerophytic conditions, the vegetation often
drying and turning brown late in July or in August. The chresard was con¬
tinuously higher on low than on high prairie, a difference of 5 or 10 per cent
in favor of the former not being unusual. The low prairie was also better
supplied with water than the Nebraska City station, though the hygroscopic
coefficient w r as at no time reached in either habitat, a margin of 3 to 10 per
cent or more at all depths usually being present. The average daily evapora¬
tion was also slightly less in the low prairie, Nebraska City being intermediate
between the low and high prairie stations at Lincoln (fig. 19).

Fig. 19.—Average daily evaporation, high prairie (solid line),
and low prairie, Lincoln (short broken lines), and Ne¬
braska City (long broken lines), 1920.

However, a comparison of the average weekly humidity at these three
stations shows this to be much higher at Nebraska City (68 to 82 per cent)
than on the low prairie (57 to 79 per cent), the high prairie being lowest of all
(58 to 73 per cent). The average weekly humidity at the three stations was

i EXPERIMENTS.

73.S. 71.9. and 65 .4 respectively from June 3 to August 26. This factor seems
suffi cient to account for Nebraska City exceeding the low--prairie station both
in per cent of germination and of establishment in denuded quadrats.

In summary, two sod transplants died from drought on the gravel-knoll
All showed repeated rotting and dying back of the leaves, including those
Altered from time to time, and the hve which blossomed did so earlier than
elsewhere. Drying at the base was often quite as pronounced as at Burlington.
AH of the transplants nourished on the low prairie and all but Spcrtina blos¬
somed. On the salt-fats the transplants suffered more or less severely during
drought, as shown by the wilting and yellowing of leaves. None died, but all
were considerably dwarfed. The dower-stalks were short on the species that
blossomed and the inflorescence smaller than normal. The sods in the swamp
suffered, from deficient aeration especially and also from deficient light. Five
high-prairie species and two from the lew prairie died. All but Spcmbu and
Paver--. .. umr..- . shewed the erects of deficient aeration, the basal leaves of

several clams dying and the rower-stalks rotting off rear the base. Sir of the
13 produced seed, some or much abbreviated rower-stalks. Sods in the Pckl
core cd the swamp fid somewhat better, only three high-prairie and one low-
prairie steedes succumbing two other high-prairie species failed to dower.
N earlv ah sh;wed the erects of deficient aeration and shade, but to a less

degree than those in the swamp. The transplants at Colorado Springs were
made late in the season -June 9 and watered at intervals: ah survived, but

nine made a good growth.

Exseeimests Dueixg 1918 no 1919.

*> <~v Of 19 IS

Feme preliminary experiments were conducted during 1918 ani 1919. and
an account of the results seems desirable for the sake of completeness. On
April 6. seeds of the following species were sown in 4-inch pots in rich loam
soil in the greenhouse: .irrm^ rkucmn. Aodr?poeeri /urcatu.?. A. rerun?.
A. sc&parius. Aristide oLi.;nr.tha . Baiticloua h:mru . and 8yer>:o. ae osper.
When the roc-ts had reached a depth of 3 to 5 inches ani the plums were wed
established, the latter were transferred to trenches S inches wide and 6 inches
deep. This was done on April 26 in both gravd-knoll and high prairie. 4 to
14 pets cf each species being transplanted at each station. They were wmered
from time to time until June 1. when they were f.rally weeded and given no
further attention until late in September. On October 25 to 27 the plants
were counted, the roots excavated, and measurements taken. Tne summer
was exceedingly dry and only one plant, eye uve . survived on tne gravei-
knoiL In the high prairie all of the An skim died, ad but one of Amcymivu,
and over half of fkrceiovu and ep-sr;eol.-?. while only tne trree species of
Ardmpceon came through with small mortality.

Fci.reTu.u t ecu had reached an average height of 6.5 inches and most of
the plums had seeded. The bulk of the root system plate was in tne
drst S inches of soil, although seme roots penetrated considerably deeper.
A frycc T■verne attained a height of 5 or 6 Irenes and the coarse roots
were traced well below the 18-inch level: A. scope:rim was c to 10 inches tall;
the central plant in plate 9 a shews the major portion of me root system, tne
roots being much utter and mere abundant than in the preceding species.

EXPERIMENTS DURING 1920.

A. nutans grew to a height of 7 or 8 inches (plate 9a). The largest of the three
survivors of Sporobolus asper reached a height of 10 inches, the plants were
in seed, and the roots penetrated more than 15 inches, the lateral spread
reaching distances of 8 to 10.5 inches. The difficulty of extracting the entire
root system in grassland has already been mentioned. Because of competition
the shoot growth was less than in cultivated soil, and it seems probable that
the root system was also shortened. These root systems, which were only
partially recovered, afford a striking contrast to those of similar species of a
single year’s growth in cultivated soil (p. 35 and 36).

Season of 1919.

During 1919, Andropogon furcatus was sown in quadrats on May 10 and
again on June 15 in high prairie, low prairie, gravel-knoll, and in the lower
cultivated plats. Those on the high prairie which were watered on May 28 at
the same time as the others, did very well, and little difference was noted be¬
tween the two plantings. In June 1920, both quadrats contained plants 6 to 8
inches tall, which reached 7 to 10 inches by fall, but no flower-stalks appeared.
A good sod, almost merging into the native prairie, was developed during 1921,
but no species flowered. By June 1922, some invasion had occurred, and
later in the season, as well as in 1923, flower-stalks were produced, although
rather sparsely. Those on the low prairie made a better growth the first
season, but, because of the plowing of the area, were transplanted to salt-flat
and the new low-prairie station the following spring. Both lots continued to
thrive (pp. 44 and 45). Plantings of this same species on the gravel-knoll
died the first season or the following winter. Those made in the lower
cultivated plats did best of all. The roots of the plants were excavated on
August 26, when they had a working depth of 2 feet and a maximum depth of
over 3 feet (plate 9b). A few stalks flowered and set seed. The vegetative
growth by August of the second season was 2 feet high. Flower-stalks, some
of which were 44 inches tall, occurred in abundance and good seed was
developed.

When about 3 weeks old, duplicate seedlings of Aristida purpurea, Stipa
viridula, and Elymus canadensis were transplanted in the usual way on
May 5, into high and low prairie, gravel-knoll, and upland and lowland
cultivated areas. Bouteloua racemosa was also included in the last two sta¬
tions. All of those on the gravel-knoll died before the end of August, after
repeated wilting and drying throughout the summer. Those on the low
prairie grew well until fall, but were plowed up the following spring. On the
high prairie all lived through the first season, but did not survive the winter.

Aristida and Stipa both died on the cultivated upland, while Bouteloua and
Elymus made an excellent growth and were in full bloom August 1. By August
15, Bouteloua had reached a height of 23 inches (plate 4b). Clumps of Elymus
1.5 to 2 inches in diameter at the base were common, and as many as 66 stalks
were counted in a single clump. The dense root system (plate 4a) reached a
depth of 2 feet and a maximum lateral spread of 22 inches.

All the plants grew very well in the lower cultivated plots, except Aristida,
wdiich did rather poorly. Bouteloua came into full bloom late in August, after
making an excellent vegetative growth, and it matured seed abundantly.
Elymus also blossomed, and by August 9 of the second season had flower-

SOD TRANSPLANTS.

stalks 58 inches tall with very large spikes. Stipa viridula made an excellent
growth, but did not seed until the next year. By August 9, 1920, the seed was
ripening on flower-stalks over 5 feet tall and in spikes 14 inches in length.
The root system at this time had reached a depth of over 5 feet and had a
lateral spread of 18 to 20 inches near the surface. Roots were very abundant
to 2 feet. In fact, the fine growth and great size attained by native species
under favorable conditions emphasize the controlling effects of competition in
the native grassland.

Sod Transplants.

Gravel-knoll.

The following sods were transplanted to the gravel-knoll during the spring
of 1919: Agropyrum glaucum, Andropogon furcatus, Bulbilis dactyloides,
Ely?nus canadensis, Koeleria cristata, and Panicum virgatum. They were
watered a few times until established. Agropyrum, Bulbilis, Elymus, and
Koeleria all blossomed somewhat earlier than elsewhere, the flower-stalks
being somewhat dwarfed. During July all dried out badly, A. furcatus and
Koeleria dying by the end of the season. Four sandhill dominants, Andropogon
halli, Calamovilfa longifolia, Muhlenbergia pungens, and Sporobolus cryptan-
drus, were secured from the extra-regional sandhills near Central City,
Nebraska, and transplanted into the high prairie on June 9. Great difficulty
was encountered in holding the sand in place about the root systems. The
new shoot growth was clipped back as usual to lessen transpiration, but in
spite of repeated watering, none survived.

Table 19 .—Holard in excess of hygroscopic coefficient on the gravel-knoll, 1919.

Date.

0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
feet.

3 to 4
feet.

Apr. 28.

24.0

25.3

20.5

19.0

12.0

May 10.

7.3

5.6

7.1

7.9

1.1

May 27.

1.0

1.7

3.6

• • • •

« • • •

June 16.

7.0

6.6

....

• • « •

....

July 1.

0.2

0.1

• • • •

....

. . . .

July 8.

-1.1

-0.4

2.4

3.5

• • • •

July 14.

-1.7

-0.1

• • • •

....

July 28.

1.6

1.7

6.7

11.4

7.1

Aug. 26.

-0.7

6.6

7.2

6.7

7.1

Sept. 6.

Hygroscopic

-0.9

-0.5

0.4

4.1

8.3

coefficient.

5.2

4.9

3.1

2.8

7.4

An examination of the rainfall records shows that while May and June had
ample rainfall, July was extremely dry. The total rainfall was only 0.38 inch
(normal 3.83 inches), 0.30 inch of which fell on the last day of the month.
The rainfall for August was also an inch below normal. The low holard of the
gravelly soil in which the plants were growing is shown in table 19. On July 14
the native grama grasses ( Bouteloua gracilis and B. hirsuta) began to dry
badly and two weeks later all were brown and cured on the ground.

During 1920, which was a more favorable season, Agropyrum produced a
thin crop of flower-stalks 34 inches high, while Bulbilis flowered profusely and
put forth numerous stolons 8 to 12 inches long, some of which rooted in the

EXPERIMENTS DURING 1920.

grama sod. Only 2 stalks of Elymus about a foot tall were left; Panicum
did poorly, reaching a height of but 12 to 14 inches. Undoubtedly by this
time most of these species were rooted rather deeply in the clay subsoil at
about 4 feet depth. In 1921, Agropyrum again blossomed, but not profusely,
while Bulbilis seemed quite at home, blooming luxuriantly and extending its
territory nearly a foot into bare areas. One of the two stems of Elymus put
forth a flower-stalk at 18 inches height; Panicum remained dwarfed (13 inches)
and did not blossom. During 1922 all made an excellent growth, exeept
Elymus, which was winterkilled. As usual, Agropyrum blossomed early;
Bulbilis had spread until it occupied nearly twice its original area, and
Panicum flowered at 18 to 24 inches. All dried earlier in the late summer
than elsewhere, owing to the severe drought in August. The next year was
an excellent one for growth, and Agropyrum spread over 30 inches into the
adjacent grassland, flowering profusely at 2.5 to 3 feet. Bulbilis spread
even more widely than before, but the stolons rooted with difficulty in the
heavy growth of grasses; it also flowered profusely. Panicum likewise
increased its area slightly, but blossomed at a height of only 27 inches.

Low and High Prairie.

The same species, except for Panicum and Andropogon furcatus, but in¬
cluding A . scoparius and Stipa spartea, were transplanted to the low prairie
during April and early May 1919. All flourished and bloomed profusely
except Stipa. No further data were obtained, since this part of the prairie
was broken the following spring. A similar lot of sods, with the addition of
Bouteloua gracilis from the gravel-knoll and Panicum, Poa, Andropogon

Fig. 20.—Average daily evaporation on high prairie (upper line) and low

prairie (lower line), 1920.

furcatus, and Spartina from the low prairie, were transplanted to high
prairie. Because of the severe drought during July, the transplants wilted
somewhat and both the production and size of flower-stalks and inflorescence
were reduced. The chresard on both high and low prairie is given in table 20.

The high evaporation rate (daily average 30 to 50 c. c.) on the high prairie
during the drought period, and the relatively lower one on low prairie (15 to
28 c. c.), together with the evaporation throughout the season, are compared

SOD TRANSPLANTS.

in figure 20. Greater wind movement at the upper station was an important
factor in causing the difference in evaporation, although humidity played a
large share also. From July 7 to August 18 the average daily evaporation
was 64 per cent greater at the upland station.

Table 20 .—Holard in excess of hygroscopic coefficient on high and low prairie at Lincoln , 1919.

High prairie.

Low prairie.

Date.

0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
feet.

3 to 4
feet.

0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
feet.

3 to 4
feet.

Apr. 28.

May 5.

18.5

19.8

22.2

18.2

19.1

15.6

13.8

9.8

11.4

14.2

24.4

26.1

26.5

17.3

16.1

17.5

16.1

19.5

May 27.

8.2

13.7

11.2

....

8.8

17.8

16.4

....

June 16.

16.6

....

24.5

18.6

....

July 1.

11.5

10.0

9.5

9.9

10.0

11.7

11.1

17.1

18.5

July 8.

2.3

5.3

9.2

10.7

9.2

5.4

5.7

16.8

17.1

20.6

July 14.

0.0

3.5

5.2

....

4.4

5.6

9.6

....

July 28.

-0.6

1.2

1.6

7.1

11.1

0.1

1.9

5.1

13.3

18.5

Aug. 11.

4.4

2.0

3.3

7.0

6.8

3.6

5.9

8.1

18.1

Aug. 26.

10.4

3.2

2.2

5.4

17.2

6.1

7.3

8.9

17.9

Sept. 6.

1.2

3.3

3.4

5.6

8.3

-0.2

4.4

2.6

9.3

18.9

Sept. 26.

14.5

15.0

5.2

2.6

5.7

17.2

15.2

5.8

10.0

16.2

Hygroscopic coef-

ficient.

9.5

8.7

8.6

7.1

6.2

11.8

11.1

10.3

10.2

11.9

June

.23 30

July

August

11 18

Sept

1 8

Fig. 21.—Average day and night temperature at Lincoln (solid lines), and Colorado

Springs (broken lines), during 1919.

Thermograph records of soil temperature at the two stations at a depth of 6
inches show no marked differences. The daily range at the lowland station
was usually 4° to 7° F.; in the drier soil of the upland station 4° to 12° F.
The minimum soil temperatures were 2° or 3° F. lower at the former station and
the maximum 3° to 9° F. higher on the upland. The soil temperatures varied
from 53° F. (average 60° to 65° F.) late in May, to 85° F. (average 70° to 78° F.)
in August. Except in early spring, when the low temperatures on the lower

EXPERIMENTS DURING 1920.

wetter soil may retard growth, and the higher temperatures at all depths on the
upland may facilitate absorption, it is not probable that the effects of tempera¬
ture on the development of vegetation at the two stations are very different.
A comparison of the two sets of thermograph records shows that differences of
air-temperature greater than 5° F. are rare at the two stations. It seems cer¬
tain that such small differences in temperature variations in the range of
growth conditions (minimum 65° F. late in May, maximum 105° F. in July and
August) would be almost negligible in the development of natural grassland.
The average day and night temperatures throughout the season, together with
those at the Colorado Springs station, are compared in figure 21 and the
average daily evaporation rates at the same station in figure 22. Notwith¬
standing the adverse conditions, all the transplants, including Spartina ,
bloomed more or less profusely, with the exception of Andropogon furcalus,
Koeleria, and Stipa.

Fig. 22. —Average daily evaporation at Colorado Springs (upper line),
and Lincoln (lower line), during 1919.

During 1920, all of the transplants did quite well, except two of the Stipas
which were not well developed. Bulbilis made an excellent growth, blossoming
profusely by the middle of June (plate 10a), and both Panicum and Spartina
flowered abundantly at heights of 30 to 44 inches. In fact, Andropogon
scoparius and A. furcatus were the only species that did not blossom, but they
made a vegetative growth of 8 to 12 inches. By the following season, Andro¬
pogon furcatus had almost merged into the common sod. Koeleria was
nearly all dead; two of the Stipas were behaving normally, while the other
two did not blossom and were in very poor condition. Elymus was repre¬
sented by only one or two stalks, but all of the other species were developing
quite normally.

On August 23, when Panicum virgatum was blooming profusely with flower-
stalks 3 to 4 feet tall, a trench was dug to a depth of 11 feet and the roots of
this 3-year-old transplant examined. Below 3 feet the silt-loam soil became
more sandy and below 6 feet it changed to a sandy gravel with many small
rocks. It was quite moist at all depths. The roots penetrated to a maximum
depth of 9.8 feet, and were fairly abundant to 7 feet. The general root habit
and branching were about normal (Weaver, 1920), but the degree and extent
of branching were considerably more pronounced than in the low prairie.

SOD TRANSPLANTS.

However, it did not approximate the remarkable network of branches dis¬
played by this species in the drier soils at Burlington (fig. 32).

Spartina cynosuroides was also excavated from the same trench at this time.
It had made a good growth, notwithstanding the rolling of leaves from time to
time as a result of drought. Many leaves reached a height of 40 inches. It
had increased its area about one-third by rhizome propagation, but the root¬
stocks had not developed many new shoots. Roots were abundant to 9 or 10
feet and were densely branched throughout, perhaps having a third more
branches than those in the lowland, and branches were longer and rebranched
to a greater degree than those growing in wetter soil. Koeleria and Poa died
during the winter of 1921-22. The following summer all the rest, except
Andropogon scoparius, made a good growth, and flowered rather profusely,
except the very late bloomers, which were somewhat retarded by the late
summer drought. Panicum extended its territory nearly 3 feet, but few of the
others gained in area. During 1923 all did very well, owing to good growth
conditions, Andropogon furcatus merging completely with the native sod. All
flowered and set seed, the height of the inflorescence being about normal.

Mixed Prairie.

On June 19, 1919, the following sods from Lincoln were transplanted to
the mixed-prairie station at Colorado Springs: Agropyrum glaucum, Andro¬
pogon furcatus, Bulbilis dactyloides, Elymus canadensis (2 sods), Koeleria cris-
tata (3 sods), Panicum virgatum, Poa pratensis, and Stipa spartea (3 sods).
Before planting, the hollow made for each sod was filled with water that was
allowed to sink away before the sod was planted, tamped in place, and again
watered. Owing to watering at 2-week intervals, all did fairly well the first
season. In June 1920, cattle broke in and closely grazed the entire fenced
area. However, most of the sods lived throughout the Season, but made
a poor growth, Bidbilis and Agropyrum alone coming into blossom. One
Elymus and one Koeleria died late in the summer. During 1921, perhaps
as a result of repeated grazing, the remaining Elymus and the two remain¬
ing koeierias died, as did also one of the stipas; Bulbilis alone came into blos¬
som. The area was grazed again in 1922. This repeated grazing, combined
with drought, resulted in a very poor growth, Poa succumbing and Bulbilis
alone blossoming. Because of further grazing in 1923, it was clearly evident
that most of the plants (except Bulbilis) were rapidly losing ground, Agropyrum,
Panicum, and Andropogon furcatus being represented by remnants only.

3. EXPERIMENTS DURING 1921.

PHYSICAL FACTORS.

Rainfall.

The season of 1921 at Lincoln was quite favorable as to precipitation.
Rainfall for April, June, and July was slightly above normal, while August
was slightly below normal, as was May also (fig. 9). An examination of the
records shows that the showers were exceptionally well distributed, no marked
drought periods occurring, and at no time did the vegetation become dry or
brown, even on the gravel-knoll.

At Phillipsburg an excess of 0.14, 0.98, and 1.30 inches fell during April,
May, and June respectively. The July precipitation was slightly above
normal and that of August and September slightly below (fig. 15). A drought
period occurred between June 9 and 28, during which time no efficient mois¬
ture fell. At Burlington the precipitation for April was nearly twice normal
(and 0.86 inch more than at Lincoln). May fell 0.91 inch below normal;
June was somewhat above (0.58 inch); July had less than half the usual amount
(only 1.15 inches); August had an excess of 1.19 inches, but September a
deficit of more than one-half the normal (fig. 16). No efficient rainfall oc¬
curred (i. e., showers of over 0.15 inch) between March 7 and April 15, April
17 and May 19, June 22 and July 4, and July 8 and August 7. By July 24

Fig. 23.—Average daily evaporation at Burlington (short broken lines), Phil¬
lipsburg (long broken lines), and Lincoln (solid line), 1921.

the short-grasses were about half-dried and brown, while two weeks later
not only the grama and buffalo grasses were curled and brown, but the wheat-
grass heads were dry as well.

Water Relations.

The above conditions were reflected in the holard at the several stations
(table 22). At Lincoln, a margin of 5 per cent or more (usually 8 or 10)

PHYSICAL FACTORS.

existed at all times and at all depths to 4 feet. At Phillipsburg the echard
was approached once in July and twice in August, actually reaching the
danger-mark to a depth of 4 feet late in that month. Conditions at Burling¬
ton were, as usual, much worse. At no time was water available in the third

Table 21. — Humidity at Lincoln and Burlington , 1921.

For week ending—

Average day.

Average night.

Average daily.

Lincoln.

Burlington.

Lincoln.

Burlington.

Lincoln.

Burlington.

Apr. 21.

53.9

64.7

59.3

Apr. 28.

53.7

• • • •

81.0

• • • •

67.3

....

May 5.

57.4

37.0

86.4

71.5

71.9

54.2

May 12.

57.6

• • • •

73.5

....

65.5

• • • •

May 19.

56.1

....

73.9

....

65.0

• • • •

May 26.

43.7

, , . .

64.2

....

53.9

• • • •

June 2.

57.2

39.4

79.4

78.7

67.9

59.0

June 9.

63.2

....

84.4

....

73.8

• • • •

June 16.

61.4

50.4

81.7

87.0

71.5

68.7

June 23.

61.7

....

80.2

....

70.9

• • • •

June 30.

60.4

30.0

81.7

72.1

71.0

51.0

J uly 7.

66.0

50.9

83.6

81.0

73.5

65.8

July 14.

58.2

....

84.3

, . . .

71.2

....

July 21.

57.7

42.5

80.4

75.5

69.0

59.0

July 28.

63.0

40.0

78.4

78.3

70.7

59.1

Aug. 4.

66.7

38.4

85.1

72.6

75.9

55.5

Aug. 11.

62.0

42.8

81.9

79.8

71.8

61.3

Aug. 18.

65.3

49.3

83.4

83.6

74.3

66.4

Aug. 25.

58.0

....

83.4

....

70.7

....

Sept. 1.

62.0

....

86.4

....

74.2

....

or fourth foot, while after June 30 it was depleted repeatedly above the hard-
pan (at about 2 feet). A complete record of humidity was not obtained, but
a comparison of the average daily humidity at Lincoln and Burlington at
the several intervals (table 21) show T s that without exception the air during the
day at Burlington was 10 and often 30 per cent less humid than at Lincoln.

Fig. 24. —Average daily temperature at Lincoln (solid line), Phillipsburg (long broken lines), and Bur¬
lington (short broken lines), 1921.

EXPERIMENTS DURING 1921.

Owing to the lower night temperatures at Burlington (altitude 4,160 feet),
the humidity at this less critical period averaged only slightly lower than at
the true-prairie station.

Table 22. —Holard in excess of the hygroscopic coefficient at the several

stations, 1921.

Lincoln, Nebraska.

Date.

0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
feet.

3 to 4
feet.

Apr. 9.

20.2

17.9

13.7

13.3

12.5

May 18.

18.0

19.7

16.7

15.0

14.5

May 25.

6.2

13.2

• • • •

....

• • • •

June 1.

7.9

10.3

....

• • • •

June 7.

23.0

18.6

14.9

14.5

17.2

June 22.

8.8

14.1

15.1

....

June 29.

7.0

8.8

11.8

15.2

■ • • •

July 13.

13.4

20.1

10.4

....

July 20.

21.0

18.5

12.7

13.1

16.8

July 28.

7.5

7.3

. . , .

• • • •

Aug. 3.

10.2

11.0

11.3

....

Aug. 10.

16.4

17.9

12.0

12.1

26.8

Aug. 23.

10.8

12.9

8.5

11.4

8.8

Aug. 31.

8.5

7.0

5.1

....

• • • •

Hygro. coeff.

9.5

8.7

8.6

7.1

6.2

Phillipsburg, Kansas.

Apr. 28.

15.8

14.5

11.4

5.9

2.3

May 19.

16.8

13.0

9.8

6.4

0.7

May 25.

12.2

10.9

....

June 9.

19.8

15.8

13.9

14.5

13.0

June 22.

5.3

8.6

10.8

11.7

9.3

July 19.

1.6

2.2

• • • •

....

July 21.

8.2

3.3

3.9

4.6

• • • •

July 25.

1.6

2.0

0.9

2.7

4.9

Aug. 4.

3.5

4.4

....

• • • •

. • • •

Aug. 9.

0.1

0.8

0.6

2.1

Aug. 17.

20.9

7.5

1.5

1.3

....

Aug. 30.

-1.6

-0.4

-0.6

0.0

0.3

Hygro. coeff.

10.6

10.9

10.6

10.7

Burlington, Colorado.

Apr. 30.

9.6

10.7

8.6

0.0

1.3

May 20.

6.4

5.6

4.7

-0.7

-2.2

June 10.

10.3

0.0

0.7

-0.5

-0.6

June 30.

-2.6

-1.6

— 1.5

-1.0

-1.3

July 17.

-1.0

-3.6

....

July 24.

-3.5

-2.2

-3.4

-1.8

-0.4

Aug. 8.

2.0

-0.7

-2.1

-0.3

-1.0

Aug. 17.

16.9

11.7

-2.4

-2.0

....

Aug. 29.

-0.3

0.5

-2.1

-0.4

-1.0

Hygro. coeff.

10.9

12.2

12.0

11.4

The average daily evaporation at the several stations is shown in figure 23.
The rather uniform rate at Lincoln (daily average 8 to 27 c. c.), as compared
with the erratic and excessively high one at Burlington (18 to 62 c. c.), empha-

PHYSICAL FACTORS.

sizes the precarious conditions for plant growth at the latter station. Con¬
ditions at Phillipsburg were intermediate (8 to 43 c. c.).

Temperature. .

The average daily air-temperature was in general lowest at Burlington,
while that at Phillipsburg after the third week in June was somewhat higher
than at Lincoln (fig. 24). During May and June, day temperatures were
highest at Lincoln and lowest at Burlington, but after July 1 this was
usually reversed. Night temperatures throughout the season were distinctly

Fig. 25.—Average day (heavy lines) and night temperatures (light lines), at Lincoln (solid lines),
Phillipsburg (long broken lines), and Burlington (short broken lines), 1921.

lower at Burlington by 5° to 10° F., those at the other stations being very
similar. The soil-temperature at a depth of 3 inches was invariably lower
at Burlington than at Phillipsburg. Later in the season soil-temperature was
correlated with the holard, the soils becoming progressively drier and warmer.
A series of readings in cropped areas (oat field) to a depth of 4 feet is
instructive in this connection (table 23).

Table 23. —Soil temperatures , 1921.

Depth

feet.

April 28-

30.

May 19-21.

June 9-10.

June 22.

June 30.

Lin¬

coln.

Phil¬

lips¬

burg.

Bur¬

ling

ton.

Lin¬

coln.

Phil¬

lips¬

burg.

Bur¬

ling¬

ton.

Lin¬

coln.

Phil¬

lips¬

burg.

Bur¬

ling¬

ton.

Lin¬

coln.

Phil¬

lips¬

burg.

Bur¬

ling¬

ton.

°C.

0 to 0.5. . .

15.4

24.0

12.0

21.0

20.0

22.2

22.5

21.4

19.8

21.2

23.2

29.2

0 5 to 1. . .

13.0

23.1

11.0

19.0

17.8

15.0

21.8

20.2

18.8

21.1

22.4

26.1

1 to 2.

12.0

15.0

10.5

13.4

16.4

14.0

20.2

18.8

16.8

20.1

21.5

23.9

2 to 3.

12.0

14.0

10.5

13.1

15.0

12.4

18.6

17.2

15.8

18.8

21.2

21.8

3 to 4.

10.8

12.0

10.0

12.0

13.5

11.8

17.8

16.0

15.0

17.4

20.5

20.0

EXPERIMENTS DURING 1921.

Isolated readings at similar depths showed the temperature of the grassland
soil (below the first foot) to be very similar to that in the cultivated areas.

To summarize; as during 1920, conditions for plant growth were much more
favorable at Lincoln and least favorable at Burlington, while those at Phil-
lipsburg were intermediate, holard and humidity being the controlling factors.

PLANTING RESULTS.

Surface Sowing.

Seeds of 14 species were sown on the surface of the high prairie at Lincoln
on April 20. Aristida purpurea, Elymus canadensis, and Lespedeza capitata
failed to germinate. Bouteloua gracilis, Koeleria cristata, Liatris punctata,
and Sporobolus asper germinated rather abundantly, but died before the
end of June, the seedlings being much attenuated. The light values were low;
by May 5 they fallen to 16 to 27 per cent, as the prairie had not been mown
the previous autumn. There was not only considerable dead grass, but the
new growth was rank and flower-stalk production greater than usual. Before
the end of July, Bouteloua hirsuta, Bromus inermis (which was badly eaten
by grasshoppers), Liatris scariosa, and Pinus ponderosa were added to the
mortality list. Robinia pseudacacia held out until August, as did also Stipa
viridula. Andropogon nutans alone survived the season, making a fair growth
6 to 9 inches high and becoming indistinguishable from the native sod.

At Phillipsburg the surface plantings made on April 28 did very much
better, 83 per cent living throughout the summer. After planting, showers
were opportune at both stations, all 12 species germinating at Phillipsburg.
Unquestionably the light relation was much more favorable at the latter
station. Reduced light and surface drought were the chief factors to which
plants gaining a foothold in climax grassland were subjected. The same
species as those used at Phillipsburg, with the addition of Sporobolus, were
sown on the surface at Burlington, April 30. None had germinated by May
20, and only 5 of the 13 germinated at all. These were Bouteloua racemosa,
B. gracilis, Bromus inermis, and the two species of Liatris. All were found
in greater or less abundance on June 10, but had disappeared by the 30th.

The fate of surface-sown plants of the preceding year is of interest. At
Lincoln and Phillipsburg none survived the first season. By May of the second
year at Burlington, Bouteloua gracilis had merged into the native sod. Stipa
viridula, the other survivor, made a good growth (8 to 10 inches) and tillered
heavily, standing out quite plainly in the short-grass turf. Its leaves were
tightly rolled and somewhat wilted from time to time, and it did not produce
flower-stalks. During 1922 it grew vigorously, but was much smaller than
similar plants in denuded quadrats. By the end of the season it was clear that
competition with the native short-grasses was too severe, Stipa rapidly
losing ground. However, it held out through the very favorable season of
1923, being represented by only a few culms 6 to 8 inches high, while the leaves
of grama were 6 inches tall all around it. It was clearly evident that it would
soon lose in the unequal struggle.

Trench Sowing.

Although poor at all stations, the results from trench planting April 20 to
30 were best at Lincoln. Of 12 species used, all but Pinus ponderosa germi-

TRENCH AND DENUDED QUADRATS.

nated; Elymus, Koeleria, Bromus, and Stipa viridula succumbed before
the end of June, the last two being badly eaten back by grasshoppers.
The slender plants of Lespedeza and Liatris punctata, the latter reaching a
height of 5 or 6 inches, died in August. Andropogon nutans, B. hirsula, and
Sporobolus came through in excellent shape, tillering heavily and forming a
rather dense sod from 2 to 6 inches tall. Bouteloua gracilis was represented by
remnants only, and Liatris scariosa by a dozen one-leaved slender plants.

In the trench at Phillipsburg, Bouteloua racemosa, B. gracilis, B. hirsuta,
Koeleria, Lespedeza, Liatris scariosa and punctata, and Pinus failed to ger¬
minate, while Bromus and Elymus died in June and Stipa viridula in July. A
single plant of Robinia pseudacacia with 5 leaves and a height of 4 inches lived
through the summer but was winterkilled. The only species that became
permanently established were Andropogon nutans and Aristida purpurea;
the former was represented in late summer by 4 or 5 good clumps, 3 tolO inches
tall. Of the same species sown in the trench at Burlington, Bromus and Elymus
alone germinated, and very sparsely, but both died during the June drought.

The development of the plants grown in the 1920 trench at Lincoln was
above the average. During 1921, Andropogon nutans, A. scoparius, Bouteloua
hirsuta, and B. racemosa all made excellent growth. The boutelouas reached
heights of 3 to 8 inches, and the andropogons of 5 to 10 inches, the latter
forming a fairly dense sod. Bouteloua gracilis was represented by a single
remnant, which was winterkilled. During 1922 the four remaining species
made a fair growth, the boutelouas alone flowering and producing a single
flower-stalk each. These grew well in 1923, though Andropogon scoparius and
Bouteloua racemosa were quite sparse. The height growth ranged from 4 to 14
inches, but none had developed flower-stalks late in August.

At Phillipsburg, Andropogon scoparius and A. nutans alone survived the
first season. During 1921 they flourished, forming dense bunches and reaching
heights of 14 and 24 inches respectively, but they were favored by some water
running into the trench. A. nutans developed few flower-stalks. Both made
an excellent growth in 1922, extending the area of the bunches, but Andro¬
pogon scoparius alone flowered. In 1923 they again grew rank, reaching
heights of 2 feet, and both flowered abundantly. At Burlington there were
no survivors of trench sowing.

Denuded Quadrats.

Twenty-eight species were planted in the quadrats at Lincoln (high prairie)
on April 14. Eight failed to germinate. Elymus and Stipa viridula died
in June, Agropyrum and Bromus inermis in July, and Acer negundo, A sac -
charinum, and Calamovilfa in August; of these, Bromus and Stipa were badly
damaged by grasshoppers; 13 species survived. Corylus americana and
Symphoricarpus spp., all of which survived, were propagated at the several
stations from rhizomes obtained from Weeping Water, Peru, and Lincoln,
Nebraska rather than from seed.

Among the 28 species planted in quadrats at Phillipsburg, 9 failed to
germinate. Among these were Koeleria cristata, Liatris scariosa, Muhlenbergia
pungens, and Onagra biennis, none of which germinated at Lincoln and all but
Muhlenbergia failed in the quadrats at Burlington. Calamovilfa, Sporobolus
asper, and S. cryptandrus died in June, Corylus and Pinus in July, while Stipa

EXPERIMENTS DURING 1921.

viridula succumbed in August. Except for Calamovilfa, the losses here and at
Lincoln were of different species. Thirteen species survived.

Of 25 species planted on April 30 at Burlington, only 6 did not germinate.
Ten species disappeared in June, another in July, and 4 others in August,
Andropogon nutans, Calamovilfa longijolia, Gleditsia triacanthus, and Stipa
viridula being the only ones to survive the season. A comparison of the
growth of Gleditsia and Andropogon, the only species that lived at all three
stations, is instructive. By the end of the summer the former was 4 to

Fig. 26.—Average daily soil-temperature at depths of 3 and 18 inches respectively at Lincoln (solid
line), Phillipsburg (long broken lines), and Burlington (short broken lines), 1921.

8 inches tall at Lincoln with 12 to 13 leaves 5 to 10 inches high, with the same
number of leaves at Phillipsburg, and only 2 to 3 inches high with 4 to 6
leaves at Burlington. Andropogon had formed a dense sod 4 or 5 inches high
at Lincoln, one 12 to 18 inches high at Phillipsburg, while at Burlington the
scattered growth was only 2 to 7 inches in height.

Of the survivors in the 1920 quadrats at Lincoln, Stipa setigera and S.
eminens were winterkilled (table 58). Andropogon nutans and A. scoparius
flourished, forming good sods 10 to 13 inches tall, the former having flower-
stalks at 2 feet. Stipa viridula and Aristida purpurea developed good bunches
6 or 7 inches in height. Bouteloua gracilis and Elymus blossomed, the Boute-
loua tufts being 4 to 6 inches tall and those of Elymus 12 to 15 inches. All
made a good growth during 1922, but Andropogon scoparius and Bouteloua
gracilis alone came into blossom, probably owing to the severe drought in late
summer. None died during the very dry winter following (table 58).

At Burlington, only 3 species, Andropogon nutans, Lialris punctata, and
Stipa viridula, survived the 1920 season in the denuded quadrats. The
vicissitudes of these plants are of interest (table 60); wilting and rolling of the
leaves, dying back, and discoloration were of usual occurrence. However,
they were well rooted and all survived the summer of 1921, notwithstanding

DENUDED QUADRATS.

the ravages of grasshoppers and drought. In 1922, all but Stipa viridula
blossomed. The flower-stalks of Andropogon nutans were 1 to 2.5 feet high
late in August, while Liairis flowered profusely. Stipa exhibited many fine
clumps 12 to 18 inches tall in 1923. As in the other quadrats, its area had been
scarcely invaded, but it did not blossom. All the species prospered and

seemed permanently established, at least as long
as competition with the short-grasses did not
occur (table 60).

Root Development in Denuded Quadrats.

Fig. 27 .—Liatris punctata from
Phillipsburg, on July 1 of
second year after planting.

From July 1 to 3 several species from the 1920
quadrats at Phillipsburg were excavated and the
root development studied. Liatris punctata had
5 leaves and a height of 5.5 inches. The portion
of the root which began 2 or 3 inches below the
surface was 5 mm. in diameter for a distance
of about an inch, beyond which it tapered
rapidly and descended more or less vertically
downward to a depth of nearly 4 feet (fig. 27). Nine branches came
off from the tap in the first 8 inches of soil, none of which exceeded 17
inches in length. Like the tap-root, they were fairly well supplied with
short branches.

Bouteloua hirsuta formed fine, well-tillered clumps 6 to 9 inches tall. The
root system was characterized by its wide-spreading habit in the surface soil.
The delicate, thread-like, exceedingly well-branched roots reached a distance
of 16 to 18 inches on all sides of the plant, the longest ones often lying in the
surface inch or two of mellow soil (fig. 28). The deeper roots, which ran

Fig. 28. —Bouteloua hirsuta from Phillipsburg, on July 1
of second year after planting.

EXPERIMENTS DURING 1921.

obliquely downward, extended to depths of 12 and a few to 24 inches. Many
new thick, w r hite roots only 2 to 4 inches long came off from the new tillers.
Bouteloua gracilis had well-developed clumps 4 to 7 inches high. The fine
fibrous root system was very similar to that of the preceding, especially in
the number and extent of fine lateral branches. However, it differed in
having a smaller lateral spread (about 12 inches) and greater depth of pene¬
tration. Many of the roots reached 2 and some nearly 3 feet in depth.

Andropogon scoparius produced densely rooted plants 8 to 12 inches tall
and so w T ell tillered that they formed a continuous sod. The dense network
of roots and branches reached a working level of 30 inches and a maximum

29 30 31

Fig. 29 .—Andropogon scoparius from Phillipsburg, on July 1 of second year after planting.

Fig. 30.—Roots of Andropogon nutans after the block of sod had been transplanted for 2.5 months
at Burlington.

Fig. 31.—Roots of Elymus canadensis at end of June of second year after the block of sod had been
transplanted and watered at Burlington.

depth of 37 inches. Many new roots of the current season’s growth were only
6 to 12 inches long. The last 3 to 8 inches of these, as well as those of many
of the rapidly growing ones of 1920, were 1 to 1.5 mm. thick, glistening white,
and destitute of branches (fig. 29). Stipa viridula was represented by bunches
4 to 9 inches in height. The roots were abundant to a depth of 2 feet, some
extending to the 4-foot level. They had a lateral spread of 12 to 14 inches
on all sides of the plant. These data are sufficient to show the fairly deep
root penetration of the species by midsummer of the second year in denuded
quadrats. It also explains how they persist so tenaciously during periods of
drought after they have had a year’s growth.

The removal of the preceding species left Andropogon nutans alone in the
1920 quadrats. It grew very well, reaching a height of 15 to 18 inches by fall
and producing a few flower-stalks 40 inches tall. During 1922 it made ex-

SEEDLING TRANSPLANTS.

cellent growth but did not blossom, while in 1923 it reached a height of 2
feet and flowered abundantly.

Seedling Transplants.

Seedlings of 13 species were transplanted into high prairie on May 11,
when 3 or 4 weeks old (table 61). As at all stations, these seedlings were
thoroughly watered when planted and during a period of 10 days thereafter
whenever necessary. Notwithstanding this care, two weeks later two of the
four lots of Andropogon nutans, Aristida purpurea, Bouteloua racemosa, and
Koeleria cristata had died. Agropyrum glaucum, Elymus canadensis, Stipa
comata, and S. viridula were eaten back by grasshoppers and many individuals
had died. One lot of Bouteloua gracilis had also died, and Andropogon fur-
catus, Stipa spartea, and Liatris punctata were the only ones which had not
suffered a high mortality. The soil at this station is of such a type that it
becomes quite compact following heavy rains or watering. This is detri¬
mental to seedlings, as well as germinating seeds, since the surface crust
hinders aeration and offers more or less of a mechanical barrier to tender
growing parts. When dry it cracks and in this way injures the plants by
mechanical tearing of roots, rhizomes, etc., near the surface. At Phillips-
burg and Burlington the soil is of such texture that it is mellow at all times.
However, all species were represented until the middle of June, after which
Stipa comata and S. viridula died, and Agropyrum, Aristida, Elymus, and
Liatris also succumbed during July. The other 7 species came through the
season in fair to excellent condition, although much reduced in number.
Bouteloua hirsuta flowered.

A similar lot of seedlings of the same age was transplanted at Phillipsburg
on May 19. These were in groups varying from 2 to 6 per species (table 62).
By June 9 all were flourishing and most of them continued to do so throughout
the month. By July 25, Agropyrum, Elymus, and Stipa spartea had suffered
heavy losses; Koeleria, Stipa comata, and S. viridula were nearly all dead,
and by the end of the summer they had disappeared. Agropyrum, Andro¬
pogon furcatus, Bouteloua gracilis, Elymus, and Liatris scariosa were nearly
all dead or dying, the last having been eaten to the ground. However, A.
nutans, B. hirsuta, Liatris punctata, and Stipa spartea were in fair to excellent
condition.

Seedlings were transplanted at Burlington on May 20 (table 63). By June
10, Koeleria had died, while Agropyrum, Liatris scariosa, and Stipa viridula
had suffered a high mortality. However, the other species were in fairly
good condition. By the end of the month, Liatris scariosa and Stipa viridula
had succumbed and most of the other species had lost heavily. Many had
their leaves rolled or were badly wilted. By the end of July, A. nutans, B.
hirsuta, and B. gracilis alone were barely alive and these died during August.

Summary.

The sequence of germination was the same as in preceding years (table
24), i. e., Lincoln being highest (81 per cent), Phillipsburg second (68 per
cent), and Burlington last (43 per cent). On the basis of establishment,
however, Phillipsburg, unlike the preceding year, ranked first with 60 per
cent, Lincoln second (40 per cent), and Burlington third (7 per cent) The

EXPERIMENTS DURING 1921.

unusually favorable rainfall at Phillipsburg during spring and early summer
has already been pointed out, as well as the relatively favorable light con¬
ditions for surface-sown plants (p. 60) .

On the basis of survival of seedlings, Phillipsburg also ranked first, none
surviving at Burlington. As to the surface-sown plants of 1920, none were
left at either Lincoln or Phillipsburg for growth during 1921, but Stipa

Table 24 .—Summary of planting experiments at the three stations, 1921.

Method of
planting.

P. ct. of species germinating.

P. ct. of germination established.

Lincoln.

Phillips¬

burg.

Burling¬

ton.

Lincoln.

Phillips¬

burg.

Burling¬

ton.

Surface sowing.

100

Trench.

Denuded quadrat....

Average.

Seedlings.

viridula survived at Burlington. Among the 5 survivors of the 1920 trench
planting at Lincoln, none died during 1921; 1 was winterkilled in 1921-22,
but the others were growing fairly well at the end of the summer of 1923.
At Phillipsburg only 2 of the 6 species planted lived, but both of these grew
well, flowered from time to time, and persisted to the end of 1923. None
survived the 1920 summer at Burlington. Among the 8 survivors in the
1920 quadrats at Lincoln, all lived throughout the next summer, but 2 were
winterkilled. The 6 remaining species did well until the fall of 1923. At
Phillipsburg all but 1 were excavated for root study. This plant made an
excellent growth and flowered from time to time until the end of 1923. The
3 species at Burlington lived not only through the 1921 season, but also during
1922 and 1923, all but Stipa viridula coming into blossom.

Sod Transplants.

Between March 20 and 31, sods of 16 species were planted on high prairie
(table 64). Distichlis was secured from the salt-flats, Bulbilis from overgrazed
low prairie, Bouteloua gracilis and B. kirsuta from gravel-knoll, Andropogon
furcatus, A. nutans, Panicum virgalum, and Poa pralensis from low prairie,
and Spartina cynosuroides from the swamp. The rest were high-prairie
species. All made a good growth during the summer, except Distichlis, which
was nearly all dead by August 31, and all blossomed except the three Andro-
pogons, Bouteloua kirsuta, B. gracilis, Koeleria, and Spartina. During 1922
all not only survived, but Bulbilis, Panicum, and Spartina increased their
area. However, by the end of August, Agropyrum, Distichlis, and Elymus
were doing very poorly, owing in part to the severe drought. Twelve of the
16 species blossomed (table 64). Koeleria died during the following winter.
The rest survived the following summer, but by August Bouteloua gracilis,
Agropyrum, Distichlis, and both lots of Elymus were in straits. Bulbilis,
Stipa, and Panicum alone produced flower-stalks before the final check was

SOD TRANSPLANTS.

made in August. This lack of vigor was probably a result of the extremely
dry late summer, fall, and winter preceding (p. 109). The 1920 sods
made fair to good growth, except for one block of Andropogon furcatus
and two of Distichlis, which were nearly dead by the last of August. The
following did not blossom: Agropyrum, A. furcatus, Bouteloua gracilis, B.
racemosa, Distichlis, Spdrtina, and Stipa. During 1922 a severe drought
occurred in late summer and both plantings of Distichlis and Elymus and one
of Koeleria succumbed. Agropryum, one Andropogon scoparius, and Poa
pratensis were in very poor condition at the end of the summer, while Andro¬
pogon nutans, A. scoparius, Poa, and Stipa alone had seeded. The following
winter one Stipa died, but all of the others made a fair to good growth during
this fourth summer. On August 25 only the following had blossomed or
showed signs of developing flower-stalks: Poa, Bouteloua gracilis, B. racemosa,
Koeleria, and Panicum.

Sods of the same species as those used at Lincoln during 1921, except
Bouteloua gracilis, Bulbilis, Distichlis, and Spartina, were secured from the
various stations at Lincoln and transplanted into the short-grass sod at Bur¬
lington on April 15. As in 1920, they were placed in duplicate rows, one
being thoroughly watered 5 times during the season (table 65). By May
20 the effects of drought were apparent. The leaves on many species were
rolled and the leaf-tips dead. Others had been frayed by whipping in the
wind, a very common phenomenon in this region among both native and
crop plants. Koeleria and Stipa were almost dead. The plants in the
watered row were in better condition.

On June 29, the root development was examined in two species, which had
been growing for a period of 75 days. Andropogon nutans was in fairly good
shape and had made a growth of 4 to 6 inches. The block of sod was found
to be in close contact with the soil on all sides. The old roots had practically
all died, though in a few instances laterals from them had continued growth.
Many new white roots had originated from the rhizomes and ended at 6 to 12
inches depth. They were evidently growing rapidly, since 3 to 6 inches of the
tips were unbranched; nearer their origin they were thickly beset with short
branches. Several longer roots penetrated from 18 to 32 inches, but none
spread much laterally, and all were well branched (fig. 30). The soil was
moister under the block of sod than elsewhere to a depth of 20 inches. Panicum
virgatum was growing nicely and had reached a height of 10 inches. As
before, the old roots were nearly all dead, but many new ones descended
vertically and were fairly abundant to 24 to 30 inches. At the last depth they
came in contact with the dry hard-pan. They were densely beset with
laterals to near the tip.

By June 29, Stipa had died in the unwatered row and Elymus and Koeleria
were nearly dead. Even some of the species in the watered row were doing
poorly, especially Poa and the three andropogons. By late July Andropogon
nutans and Poa had died in the watered row and several other species were in
bad condition. In the unwatered row, Elymus and Poa had succumbed,while
Andropogon furcatus, Bouteloua hirsuta, and Koeleria died soon after. Autumn
found only 7 of the 11 species alive in the watered row, and these were repre¬
sented by mere remnants of the original fine blocks of sod. Of the survivors,
Agropyrum, A. scoparius, Bouteloua hirsuta, Elymus, and Stipa were high-

EXPERIMENTS DURING 1921.

prairie species, and A. furcatus was from the lowland. Agropyrum alone had
blossomed. Of the 11 duplicate unwatered species (2 having been dug up)
only Agropyrum, Bouteloua racemosa, and Andropogon scoparius remained,
and Agropyrum alone had a single flower-stalk.

During 1922, both Agropyrum and Bouteloua racemosa in the unwatered row
survived, although both suffered severely from drought, and Andropogon
scoparius was winterkilled. Agropyrum put forth flower-stalks mostly less
than a foot high, and the spikes were very much dwarfed. In the lot that
were formerly watered, Andropogon scoparius died in July. The other 6
species made a fair to good growth, often showing the effects of drought in
the rolled leaves or dead leaf-tips, as well as by the dwarfing of the whole
plant. Bouteloua racemosa, B. hirsuta, Agropyrum, and Elymus alone blos¬
somed. Andropogon furcatus, Bouteloua hirsuta, and Bouteloua racemosa died
before the next spring. The single survivor of the unwatered lot, Agropyrum,
reached a height of only 11 inches the following summer, and did not blossom
or extend its territory into the surrounding short-grasses. Agropyrum in the
other lot grew only slightly better, Elymus was represented by only 2 shoots,
and Stipa by a small c'ump, while Bouteloua racemosa alone made a good
growth and seeded. Although the season was unusually favorable for growth,
the results indicate that one more year would probably have been sufficient to
eliminate most if not all of the transplants. As to the growth of the 1920
transplants at this station, it may be recalled that none died during the first
season, even in the unwatered row, but all suffered from drought and were
dwarfed. During 1921 none were watered.

Root Development.

On June 28-29 the root development of several species was examined.
Elymus, which had blossomed the preceding season, had made a fair growth
in the unwatered row, reaching a height of 21 inches with flower-stalks appear¬
ing. The old roots had not renewed their growth, but laterals on them were
functioning (fig. 31), and in one case a large lateral had replaced the cut main
root. It was traced to a depth of 20 inches. Many new roots reached a
depth of 18 inches, while a few penetrated to the hard-pan, about 30 inches
deep. All were quite profusely branched to near the tips, and the lateral
spread was about 12 inches. Some new roots, evidently of the current year’s
growth, extended just through the old block of sod.

Andropogon furcatus was excavated in the unwatered row. The original
sod, which was in complete contact with the soil, was a foot square and 8
inches deep. The new growth of leaves was about 8 inches high. Most of
the old roots had died, but new ones had grown out thickly from the rhizomes.
Many of these penetrated to the hard-pan layer at 28 to 30 inches, but only
slightly into it. All of the roots were well branched, but were very dry, like
the soil in which they grew. The dry, mellow loess below the hard-pan layer
(which was 8 to 12 inches thick) contained roots of the short-grasses and
Psoralea. There was very little lateral spread of the roots beyond the area
occupied by the original block of sod.

Koeleria, Stipa spartea, and Panicum virgatum were excavated in the
watered row, which had also received some run-off water because of its loca¬
tion near a furrow separating an adjoining plowed area. The clump of

SOD TRANSPLANTS.

Koeleria was 3.5 inches in diameter and had flower-stalks 12 inches tall, with
rather normal spike development at a height of 6 to 15 inches. The roots,
although very fine, were easily distinguished from those of the short-grasses
by their yellowish color and slightly larger size. A great mass of profusely
branched roots filled the soil to a depth of about 18 inches, a few penetrating
8 inches deeper. A lateral spread of 14 inches
in the surface inch of soil was determined, the
occupation of a considerable area on all sides
of the plant being characteristic of the species.

Stipa spartea had made a growth of 2 feet,
with 5 flower-stalks, some of which were 34
inches tall, and formed a representative clump.

The roots reached a maximum depth of 34
inches in the fairly moist soil, many ending at
or above 2 feet. The lateral spread was about
normal (10 inches), but the roots as a whole
were more profusely branched than is usual in
its native region. Panicum virgatum, which
flowered the preceding season, had made a good
growth, reaching a height of 18 inches. The
root development was remarkable. The coarse
roots, 2 to 4 mm. in diameter, pursued a nearly
vertically downward course to a maximum
depth of 7.3 feet, spreading only a little near
the surface. The profound branching to the
extreme depth of penetration is shown in figure
32. Branches at the rate of 25 per inch were
not uncommon, and many of the roots were
branched to their tips. In fact, the profuse
branching, which undoubtedly was a response
to the dry soil, can scarcely be overemphasized.

It stood out in marked contrast to the rather
poorly developed lateral roots found on plants
growing in lowland soil near Lincoln (Weaver,

1919:4).

Notwithstanding the root growth of these
transplants, the severe drought conditions
caused the death of several species. One sod
of the watered Koeleria and one in the un¬
watered row succumbed. In the unwatered
row, 1 Andropogon scoparius was winter-killed,

2 individuals of Elymus , and all 3 of Poa
died. Agropyrum was the only species that
increased its territory perceptibly. It invaded
the short-grass sod on all sides for a distance of about 18 inches, the
new shoots from the rhizomes reaching a height of 8 to 13 inches. How¬
ever, none of these or those of the original block bore flower-stalks.
Bouteloua racemosa, Elymus , Stipa, and Panicum (in the watered row only)
bore a few flower-stalks and seeded, as did also Poa and Koeleria , the

Fig. 32. —Roots of Panicum vir¬
gatum at end of June of second
year after the block of sod had
been transplanted and watered
at Burlington.

EXPERIMENTS DURING 1921.

last in the unwatered row also. All suffered severely from spring and summer
drought, and in general growth was poor. Most of the species, including such
late bloomers as A. scoparius, A. furcalus , and Panicum, took on their late
autumn reddish color, dried out, and went into winter condition by August 1.
As a whole, those that had not been watered the preceding season, and hence
were less well established, suffered most. During 1922 none of the plants
died in the lot which had been watered, but Koeleria of the unwatered
lot succumbed in July. Of the former, Agropyrum had extended its area 3
or more feet on all sides, even into the densest sod. Although it flowered
profusely in its original area at a height of about 2 feet, only a few dwarfed
flower-stalks appeared in the invaded area. This was the only species which
extended its area, though, on the other hand, the short-grasses were unable to
invade markedly. Because of the severe drought of July and August, the
grasses at this time mostly took on their late autumn or winter color and
habit. Aside from Agropyrum , Stipa, Elymus, Koeleria, and Bouteloua race-
mosa blossomed; the last, unlike the others, had flower-stalks and inflorescences
of about normal size. In the row that was not watered the preceding year,
conditions were much more severe. Agropyrum possessed only a few flower-
stalks, while Panicum put forth a few small panicles at 8 inches height.
Elymus and Koeleria also blossomed.

The very dry fall and winter following took severe toll. Of the 15 lots of
plants in the area formerly watered, 4 died, while among those less well
established 8 succumbed, leaving Agropyrum as the sole survivor. Notwith¬
standing the excellent growing conditions of the following summer (1923), it
had no flower-stalks, but had sent out a few rhizomes to distances of 2 to 3
feet. The rest of the plants received extra water from a furrow which was
only 2 feet distant and separated the unbroken and broken land. Hence,
their growth was very good. Agropyrum had spread nearly 4 feet, but
no flower-stalks occurred in this sodded area. Elymus had also extended
its area over a foot, but another block of the same species died. Panicum ,
Bouteloua racemosa, and Stipa blossomed. Among the forms introduced with
the sods, Erigeron ramosus blossomed the second year (1921), as did also
Sporobolus asper, though both died later. Brauneria pallida in one instance
made a good vegetative development only, while in another case it blossomed
normally the second year after transplanting before succumbing. Aster
multiflorus occurred in several blocks of sod and usually spread a foot or more
by rhizomes; it blossomed profusely. In some cases it persisted 4 years
near the furrow, but in others it soon succumbed.

EXPERIMENTS AT OTHER STATIONS, 1921.

Physical Factors.

Rainfall and Holard.

Studies on experimental vegetation were continued in the series of edaphic
stations at Lincoln, at Nebraska City, and Colorado Springs during 1921.
The general conditions of precipitation at Lincoln have already been given
(p. 56). The season at Nebraska City was one of drought. The precipitation
from February to July was decidedly below normal, that of April, May, and
June being 0.86, 1.56, and 0.52 inches respectively below the mean. However,

EDAPHIC STATIONS.

July had an excess of 5.17 inches, but August was slightly below normal.
No efficient rain fell between May 9 and 26 or from June 18 to July 2. The
holard at the Nebraska City, low-prairie, and gravel-knoll stations is given
in table 25. Notwithstanding the decreased rainfall, the soil at Nebraska
City usually had a margin of 10 per cent above the hygroscopic coefficient.
In fact, a more constant supply was maintained here than on the low prairie
at Lincoln, where the chresard in the surface 6 inches was practically exhausted
during the last half of June. An abundant supply was at all times available
below the 6-inch level. Even on the gravel-knoll conditions were quite
favorable as compared with the preceding year. However, the xerophytism
of this habitat is shown by the small chresard, usually only 1 to 6 per cent,
even at depths of 3 or 4 feet.

Evaporation and Temperature.

The average daily evaporation at Nebraska City (fig. 33) was considerably
lower (often 33 per cent) than on the high prairie. In general, evaporation
on the gravel-knoll was greater than on high prairie. This was especially
true during the earlier and later parts of the season. The average daily
air-temperature at Nebraska City was usually higher than that at Lin¬
coln, and during the first 3 weeks of June it ranged from 2° to 11° F. higher.
This was largely due to the much higher average night temperatures (often
6° to 10° F.), since the day temperatures nearly always averaged lower than
those at Lincoln. The temperature of the soil at 18 inches was constantly
1° to 3° warmer.

Fig. 33.—Average daily evaporation on high prairie (solid line), gravel-knoll,
Lincoln (short broken lines), and at Nebraska City (long broken lines), 1921.

Planting Results.

Surface Sowing.

During 1921, surface sowings were made on the low prairie at Lincoln and
at Nebraska City. Of the 9 species planted at Lincoln on April 23, Aristida
purpurea and Bouteloua gracilis did not germinate. Elymus canadensis had
died by June 27, as the following did also during July and August: Lespedeza
capitata , Liatris punctata , L. scariosa, Sporobolus asper, and Stipa viridula.
Koeleria cristata was still alive in September, but was missing the following

EXPERIMENTS DURING 1921.

May. Owing to the low light intensities, all of the seedlings were extremely
attenuated as compared with those growing in the quadrats, and were unable
to make any real growth. The ability of Koeleria to endure shading was noted

Table 25 .—Holard in excess of the hygroscopic coefficient at Nebraska
City, Lincoln low prairie , and gravel-knoll , 1921.

Nebraska City.

Date.

0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
* feet.

3 to 4
feet.

Apr. 9.

24.4

21.4

18.5

11.7

8.7

Apr. 23.

19.4

20.1

• • • •

May 6.

10.7

16.5

15.6

....

, , , ,

May 20.

10.1

16.1

16.8

12.6

12.8

June 3.

23.8

21.6

19.5

....

June 24.

12.0

9.9

15.0

13.6

13.8

July 15.

9.2

14.9

14.3

11.5

9.6

Aug. 13.

14.8

10.5

16.0

15.9

10.1

Aug. 26.

12.8

10.2

10.9

12.5

9.1

Hygro. coeff.

12.1

11.7

12.3

13.7

12.9

Low Prairie.

Apr. 9.

20.8

25.2

Apr. 23.

28.4

23.3

21.5

20.5

29.7

May 18.

10.5

24.8

21.6

22.3

23.0

May 25.

9.7

20.0

....

• • • •

June 1.

6.8

16.4

....

• • • •

....

June 7.

9.7

21.7

21.3

23.9

22.4

June 22.

1.4

16.0

19.2

....

June 29.

0.3

11.1

14.2

16.8

....

July 20.

25.4

23.0

15.3

15.2

18.9

July 28.

18.2

17.9

....

• • • •

. . . .

Aug. 3.

15.0

12.1

14.5

....

Aug. 23.

12.3

14.9

11.8

13.7

19.3

Hygro. coeff.

11.8

11.1

10.3

10.2

11.9

Gravel-Knoll.

Apr. 9.

6.5

3.9

5.0

1.0

May 18.

6.3

5.3

6.6

9.2

1.7

May 25.

0.1

-0.2

• • • •

....

June 1.

3.3

-0.3

....

June 7.

7.8

6.4

11.8

5.7

1.2

June 22.

-0.1

0.4

3.4

4.5

3.9

June 29.

1.9

3.3

3.1

3.3

....

July 13.

4.5

2.6

3.7

• • . .

....

July 20.

4.8

3.1

2.1

4.7

9.6

July 28.

0.4

1.3

• • . •

....

Aug. 3.

4.8

4.0

2.7

....

Aug. 10.

3.1

2.3

2.0

6.7

10.8

Hygro. coeff.

5.2

4.9

3.1

2.8

7.4

in many instances, this adaptation probably resulting from its smaller stature.
Light values at the soil-surface in the several stations are summarized in
table 26.

EDAPHIC STATIONS*. LIGHT VALUES.

Table 26 .—Light values at the several stations.

Cover.

High prairie, Lincoln.

May 28, 1922 :*

Andropogon scoparius.

Psoralea tenuiflora floribunda
Psoralea, Kuhnia glutinosa..

Brauneria pallida.

Astragalus crassicarpus.

July 25, 1921:

Mixed grasses.

Average density.

Light cover.

Medium, 4 inches above

surface.

Light.

Trench, medium cover.

Swamp, Lincoln.

May 28, 1922:

Spartina cynosuroides.
Phalaris arundinacea. .
Mixed grasses, etc....
Mixed grasses, average.
August 24, 1922:

Spartina cynosuroides

Densest places.

Medium density....
Most open.

Nebraska City.

June 3, 1921: 3

Stipa spartea.

Andropogon furcatus.

Andropogon scoparius.

Liatris scariosa.

Grasses, Amorpha canescens.

Rosa arkansana..

Brauneria pallida..

A. scoparius, Koeleria cristata

A. scoparius, A. nutans..

Stipa spartea.

Very open grass mixture.

Ceanothus ovatus.

A. furcatus at side of quadrat.
June 3, 1922:

Very open grass mixture.

Euphorbia corollata, grasses..

Rosa arkansana.

Baptisia bracteata.

Vernonia fasciculata.

A. furcatus, Desmodium.

Stipa spartea, Andropogon. . .
Koeleria cristata, Poa praten-

sis.

Ceanothus ovatus.

Per cent.

20 to 26
6

10 to 15
2.5 to 3.5

8.3

5.0

9.5

15.0
17.0
20 to 23

7.5

6.5
4.5 to 20

11.2

1.5 to 2.6
4 to 5.7
11.4 to 16.6

18
11.5
17.5 to 26.5

7.5
8 to 11.5

2.5
7

8 to 10
60

1.5
15

40 to 45
20

6.5
6.5

. 8
9
9

22 to 26
2 to 2.5

Cover.

Low prairie, Lincoln.

May 28, 1922:

Mixed grasses:

Above 1921 growth.

Average.

Under old growth.

June 14, 1922: 2

Surface seeding, average den¬
sity.

Side of 1922 trench:

Spartina cynosuroides.

Verbena stricta.

Light, mixed.

Andropogon furcatus..

Grasses, Art. ludoviciana.. .

Lighter mixed cover.

South edge of 1922 quadrats. . .
West edge of 1922 quadrats... .
July 25, 1922:

Mixed grasses:

Average density.

Lighter.

Very dense.

4 inches above trench surface:

Medium.

Lighter.

Dense.

Phillipsburg.

June 10, 1922:

Bouteloua gracilis and Bul-

bilis dactyloides.

Psoralea tenuiflora.

Elymus canadensis.

Andropogon furcatus.

A. scoparius.

Bouteloua, Gaura, Lygodes-

mia.

Verbena stricta.

Grasses:

Edge of 1922 trench.

Edge of 1921 quadrat. . . .

Burlington.

June 11, 1922:

Bulbilis dactyloides 4 .

Bulbilis, B. gracilis, open 4 .

Bulbilis, very dense 4 ..

Agropyrum glaucum 6

Stipa spartea.

Short-grasses:

Malvastrum coccineum 4 .
Erysimum asperum 4

Edge of 1922 trench.

Edge of 1920 quadrat.

Edge of 1922 quadrat.

Per cent.

7.5 to 28
16

3.4 to 14

5 to 7

12.5

7 to 12

8 to 13

15 to 27
10

2.5 to 4
12

1.5

4.5

60 to 80
35
24

3 to 8

4 to 5

11.5

37.5

1 Burned over in late winter.

2 This portion of the area burned over in late winter.

3 Mowed the preceding September.

4 Photometer sunken so as to be
level with soil surface.

6 Transplanted sods.

EXPERIMENTS DURING 1921.

Of the 9 species sown on the surface at Nebraska City, Liatris scariosa
and Pinus ponderosa failed to germinate, squirrels having eaten the seeds
of the latter. A single plant of Elymus was found, but it died before June.
By June 24, Liatris punctata and Bromus inermis were also dead, the latter
having been eaten to the ground by grasshoppers. Sporobolus asper and
Stipa viridula (also somewhat eaten) died in later summer. Bouteloua gracilis
and B. hirsuta survived the season, reaching a height of 1 or 2 inches, tiller¬
ing somewhat, but persisting as very delicate seedlings.

Of the 1920 surface sowing (p. 43), Koeleria alone survived the first sum¬
mer on the low prairie, but was then winterkilled. At Nebraska City, 4
species, Andropogon scoparius, A. nutans, Bouteloua hirsuta, and Elymus cana¬
densis survived. During 1921 a single small weak survivor of B. hirsuta
remained alive until July 15, when it succumbed. The other species had
become indistinguishable from the general plant cover by midsummer.

Trench Sowing.

Plants in the trench at Nebraska City did even more poorly than those on
the surface. Aristida purpurea, Bouteloua gracilis, B. hirsuta, Liatris scari¬
osa, Pinus ponder osa, and Stipa spartea did not germinate, squirrels having
eaten the seeds of Pinus. By June 24, Elymus canadensis, Liatris punctata,
and Sporobolus asper, all represented by a few plants only, had died. Calamo-
vilfa longifolia, which had germinated abundantly, also succumbed in July.
Bouteloua racemosa survived the first season, tillered rather freely, and reached
a height of 4 to 7 inches; however, it was winterkilled. Andropogon nutans
alone survived, and it was represented by a good stand that had reached a
height of 10 to 15 inches by fall.

Out of 11 species planted in a trench on the gravel-knoll April 20, Aristida,
B. gracilis, Koeleria, Lespedeza, and Liatris scariosa failed to germinate. A
few of each of the following species germinated, the two last dying almost at
once and the rest during June: Elymus, Liatris punctata, Sporobolus, Stipa
viridula. Pinus died in July and Andropogon nutans alone survived, being
represented by a few plants 4 to 5 inches tall in September. These lived
over the winter, but were doing poorly in May 1922, and all died in June.
Six of the 8 species planted in the low-prairie trench germinated, but Bouteloua
gracilis, Lespedeza capitata, and Liatris punctata died soon after. Elymus
canadensis succumbed to dense shading in June, and Stipa viridula, which
was represented by only 3 delicate plants, in July. Aristida purpurea, simi¬
larly undeveloped and few in number, consisted of a few remnants 4 to 6
inches tall in late summer, but unfortunately these were dug up in trans¬
planting sods.

Of the seedings in the trench on the gravel-knoll during 1920, Andropogon
scoparius and A. nutans were the only survivors, and these were in a very
poor condition. Both made a feeble growth and died late in June 1921.
In the low-prairie trench, 4 species survived the first season (1920), but all
showed the effects of shading. Of these, Stipa viridula and Bouteloua hir¬
suta made a poor growth the next spring and died in July. Andropogon
nutans and Elymus canadensis, although much attenuated, made a good
growth, reaching heights of 12 to 15 inches. In 1922 a small clump of each
reached a height of about 32 inches, but neither blossomed, owing to the

EDAPHIC STATIONS: CULTIVATED AREAS.

dense shade. However, the following summer both blossomed at a height
of 2.5 feet. At Nebraska City all but 4 of the 1920 species planted in the
trench survived, but Bouteloua gracilis died the following winter. Liatris
punctata and Elymus were both quite abundant in May 1921, but succumbed
during June and July. Andropogon nutans and A. scoparius came through
the second season in a thriving condition, reaching a height of 5 to 9 inches.
A sparse growth of Bouteloua hirsuta also survived the season, but the plants
were quite delicate. During 1922, Bouteloua succumbed during August
drought, after making a good growth earlier. The two Andropogons formed
good sods 10 to 13 inches tall; both grew well in 1923, but neither blossomed.

Growth in Cultivated Soil.

Native species made remarkable development when grown in narrow rows
(trench method) in fertile cultivated soil, kept at a favorable holard at all
times, and free from competing species (table 27). These were planted on
April 22 and 23 on a level area of soil, the physical and chemical composition
of which was almost identical with that of the cultivated lowland area already
described (p. 41). Potatoes had been grown on the area the preceding year.
A month before planting a good seed-bed was formed by plowing and repeated

Table 27 .—Growth of species in cultivated soil , 1921.

Species.

Aug. 2.

Oct. 11.

Agropyrum glaucum. . .

Excellent growth, 14 in.
tall.

Excellent growth, IS in., no flower-
stalks.

Andropogon nutans. . . .

Excellent, 24 to 32 in.

Lvs. 2 ft., flower-stalks 3 to 5 ft.,
seed ripe.

Aristida purpurea.

Fine, 12 in. tall.

Fine bunches 1 ft., flower-stalks 18
to 20 in., abundant, seed ripe.

Bouteloua gracilis.

Excellent, 1 ft., flower-
stalks to 2 ft., abundant
blooming.

Lvs. 16 to 18 in., flower-stalks 23 to
33 in., seed ripe.

Bromus inermis.

Good 10 to 15 in.

Lvs. 9 in., dense growth, flower-stalks
18 in.

Calamovilfa longifolia..

Excellent 20 to 24 in.

Fine development, 24 to 27 in.;
flower-stalks 3 to 5 ft.; seed ripe.

Elymus canadensis....

Fine, flower-stalks 2 to 2.5
ft., heads 5 to 5.5 in.
long.

Fine 15 to 18 in., heads abundant at
30 in.

Gleditsia triacanthus...

Good, 7 to 14 in.

Good, 13 to 20 in., stems 7 to 10 mm.
diameter.

Lespedeza capitata....

Good, 4 to 5 in.

Good, 5 to 6 in.

Liatris punctata.

Good, none over 5 in.

Good, 4 to 6 in., max. of 7 lvs.

Muhlenbergia pungens.

Excellent, 6 in.

Flower-stalks 15 to 18 in. seeded;
bunches 12 to 16 in. diameter, 8-
in. lvs.

Onagra biennis.

Rosettes abundant, 2 to 3
in. tall, 6 to 7 Ivs.

Rosettes 5 to 8 in. diameter, 2 to 3
in. tall.

Robinia pseudacacia. ..

2 to 3.5 ft., excellent.

5 to 5.5 ft. stems over 0.5 in. diam¬
eter.

Sporobolus asper.

Excellent, 20 in.

Bunches 3 to 3.5 in., flower-stalks
abundant, 3 to 4 ft., seed ripe.

Stipa spartea.

Good growth.

Lvs. 15 to 22 in., very broad; no
flower-stalks.

Stipa viridula.

20 to 22 in., fine.

Excellent; no flower-stalks.

EXPERIMENTS DURING 1921.

harrowing (Weaver, Jean, and Crist, 1920, p. 80). The growth made during
the first season by Andropogon nutans and Calamovilfa longifolia is shown in
plate 10b and 11a, and that of Bouteloua gracilis in plate 11b.

Root Development at Lincoln.

Pinus ponderosa reached a height of 1.5 inches when 45 days old (June 6).
The first whorl of leaves was as long as the cotyledons, while the strong,
vertically descending tap-root had penetrated to a depth of 13 inches. The
root was 1 mm. in diameter, brown in color, and had no laterals over 4 cm.
long. These were usually about 2 cm. in length, white near the end, and
entirely unbranched (fig. 34). Liatris punctata of similar age had a single
leaf, 2 to 4 inches high, m addition to the cotyledons. The glistening white
tap-roots, which had already stored some food in their enlarged upper parts,
tapered rapidly and reached depths of 10 to 16 inches. They descended almost
vertically and gave off but few short, unbranched laterals (fig. 35).

Robinia pseudacacia seedlings 45 days old had about 6 compound leaves and
a height of 5 inches. The tap-roots, which were about 1 mm. in diameter,
tapered rapidly, but regained their maximum diameter again in the vigorously
growing unbranched portion of the tip. The much branched tap-root de¬
scended irregularly to a depth of 12 to 15 inches; branching began about an
inch below the surface and continued at the rate of 10 to 20 per inch, some of
the horizontal laterals having a spread of 5 inches (fig. 36).

The root development of Gleditsia triacanthus was even more remarkable.
On July 12, when 81 days old, the trees had a height of 7 to 9 inches and
15 to 17 compound leaves. The tap-roots reached depths of 30 to 40 inches.
The course was rather directly downward, long branches being given off
profusely to a depth of 20 inches. Below this for 9 to 12 inches many shorter

EDAPHIC STATIONS: ROOT DEVELOPMENT.

ones arose, while for the rest the tap-root was unbranched. The great spread
of the longer horizontal laterals (12 to 18 inches), their large number and
their characteristic branches, which as often pursued an upward as a down¬
ward course, are shown for one of the largest and best developed root sys¬
tems in figure 37.

The rosettes of Onagra biennis
were 5 to 8 inches in diameter and
2 to 3 inches tall on October 5.

The glistening white tap-roots, 5
or 6 mm. in diameter, penetrated
downward in a somewhat devious
course, tapering to 1 mm. in width
at 1.5 feet, but reaching a depth of
39 to 44 inches in the stiff wet clay.

Branches to the number of 100
were produced just below the soil
surface and extended to a depth
of 1 foot. These varied from 1 cm.
to 6 inches in length in a few; they
ran off rather horizontally and
were poorly rebranched. Below a
foot the branching was much less
pronounced (3 to 6 per inch of
tap-root) and the laterals short.

In the deeper soil the tap-root took
on the appearance of a delicate
white thread.

Near the end of the growing-
season (October 5 and 6), the root Fig. 36 .—Robinia pseudacacia 2.5 months old.

development was examined in

Muhlenbergia pungens and Calamovilfa longifolia, both characteristic sand¬
hill species but growing here in rich silt-loam. The former was repre¬
sented by bunches 12 to 16 inches in diameter and about 8 inches in
average height. Flower-stalks 15 to 18 inches tall were abundant (plate
12a). The root development was marked, roots being traced in the clay sub¬
soil to a maximum depth of 2.5 feet; they were very abundant in the first 22
inches. Many new roots, with thick, white, rapidly growing tips ended in the
first foot, some of these originating from the prostrate stems. All branched
profusely upon entering the soil. However, there were not so many surface
roots as described for sandhill specimens (Weaver, 1920:89), but this may have
been due to the age of the plants. Fine laterals were exceedingly numerous,
as many as 30 to 50 per linear inch. While most of these were only 1 cm. or
less in length, others were 8 to 9 cm. long All were profusely and minutely
branched, the larger ones to the third and fourth order, many even to the
tips, although some that were growing rapidly were wide and free from
branches.

Calamovilfa longifolia had made an excellent growth, having leaves 24 to
27 inches high and an abundance of flower-stalks at 24 to 27 inches. Many
tough, wiry rhizomes, thickly covered with long scales and tipped with buds
an inch long with very sharp, hard points, extended out on all sides of the

EXPERIMENTS DURING 1921.

clumps for distances of 6 to 8 inches. Multitudes of tough, wiry roots pene¬
trated the soil vertically or obliquely downward to depths of 5 feet and some
extended beyond a depth of 6 feet. Beginning just below the soil-surface and
extending to near the tips of the larger roots, laterals well provided with

Fig. 37.—Root system of Gleditsia triacanthus less than 3 months old.

branches to the third and fourth order were abundant. Most of the major
laterals were only an inch or two long, but with the main roots they thoroughly
occupied the soil beneath and for several inches on all sides of the clumps
(plate 12, a and b).

The excellent growth of all of these native species under favorable condi¬
tions, 73 per cent of the grasses bearing seed the first season, emphasizes the
keen competition prevailing in stabilized grassland.

Sowing in Denuded Quadrats.

Fourteen species were planted in denuded quadrats on the gravel-knoll,
April 14. Koeleria, Liatris 'punctata, L. scariosa, and Muhlenbergia pun -

EDAPHIC STATIONS: DENUDED QUADRATS.

gens did not germinate, while Lespedeza capitata, Pinus ponderosa, Sporo-
bolus asper, and Stipa viridula germinated in small numbers (except the last,
which was abundant), but all died before the end of June. At this time the
mortality among the remaining species was also pronounced. Bouteloua
gracilis and Cala7novilfa longifolia died in August. Symphoricarpus vulgaris,
which was propagated from transplanted rhizomes, was nearly dead; Elymus
canadensis was represented by a single plant 9 inches tall, Aristida purpurea by
4 small clumps, and Andropogon nutans, which had formed a thin sod, was in
poor condition. Thus, while 71 per cent of the species germinated or grew
from rhizomes, only 40 per cent of these survived the first summer. Of the 16
species sown in the low prairie, Koeleria, Liatris punctata, L. scariosa, and
Symphoricarpus vulgaris failed to germinate or grow. Stipa viridula died
during June, but the other species (except Robinia pseudacacia, which died
in August) lived throughout the summer. Nearly all made a fair growth,
but showed the effects of shading, the trees having thin, broad, pale leaves,
while those of several of the grasses were much attenutaed.

At Nebraska City, 7 of the 22 species failed to germinate. The
seeds of Pinus ponderosa had been dug up and eaten by squirrels. Bouteloua
gracilis, Lespedeza, Liatris scariosa, and Muhlenbergia died in June, as did also
Bromus inermis and Elymus canadensis, both of which were eaten to the
ground by grasshoppers. All of the others made good growth, shading at this
station being less pronounced than on the low prairie. Andropogon nutans
formed a dense sod (plate 13). Calamovilfa was all dead but a few plants, and
Robinia was represented by a single seedling, grasshoppers having damaged
the good stand of early summer.

On the low prairie all the species germinated and all but one came through
the season in excellent condition. Stipa setigera (seed from California)
was winterkilled; the rest made a fine growth until the end of July, Aris¬
tida and both species of Bouteloua blossoming. Andropogon scoparius, A.
nutans, A. furcatus, Koeleria, and Elymus formed good sods and, like Stipa
viridula, had tillered heavily. A. scoparius, A. nutans, A. furcatus, and
Elymus continued to thrive, reaching heights of 18 to 24 inches, Koeleria grew
to a height of 7 to 11 inches, and Bouteloua gracilis and B. hirsuta put forth
flower-stalks 18 to 20 inches tall. Stipa viridula was so badly shaded that half
of the plants died, as did nearly all the tips of the leaves on the remaining
plants. Autumn found Aristida purpurea also nearly dead or dying. In 1922,
one lot of Aristida died in May and one of Stipa in June, while Bouteloua
hirsuta succumbed in August. Although the others made a good growth for the
most part, Andropogon nutans, A. scoparius, Aristida, Bouteloua gracilis, and
Elymus alone flowered. During the summer of 1923, Aristida, Bouteloua
gracilis, Stipa viridula, and one lot of Koeleria died as a result of the dense
shade. While all three andropogons showed a normal development and
formed dense sods, Koeleria and Elymus were clearly losing ground.

At Nebraska City, all of the species in the 1920 quadrats germinated,
became established, and lived throughout the summer. However, the
mortality among the individuals was high and in general the stand was
not as good as on the low prairie. All but Koeleria survived the second
season. Grasshoppers kept this species eaten back and also made consider¬
able ravages on Stipa viridula, Liatris punctata, and Elymus. Most of the

EXPERIMENTS DURING 1921 .

grasses tillered heavily, but no sod was formed comparable to that on the
low prairie. Neither did they make an equal height-growth, Andropogon
reaching heights of only 8 to 15 inches and Bouteloua 4 to 10 inches. Andro¬
pogon scoparius and Bouteloua hirsuta were the only species that seeded.
Stipa viridula was winterkilled in 1921-22, but all the rest came through the
following summer. Aristida and Elymus did poorly, but the others made
fair to excellent growth, although flower production was not common, prob¬
ably owing in part to the dry fall. These 3-year-old quadrats were mostly
invaded by Poa pratensis, species of Car ex, and Helianthus rigidus, but none
to the extent of causing serious competition. In 1922-23, Elymus was winter-
killed, while all the others not only made a very good growth, but some also
produced seed.

At Colorado Springs, 7 of the 9 species planted in denuded quadrats in
1920 germinated, and of these 4 survived the first season. During 1921,
Andropogon nutans and Elymus canadensis died, but Andropogon scoparius
formed an open tufted sod 2 inches high, which covered nearly the whole
quadrat and reached a height of 4 inches by 1923. Stipa viridula also did
well, tillering freely and reaching heights of 6 and 12 inches in 1922 and
1923 respectively.

Root Development at Peru, Nebraska.

A number of species were grown in a well-prepared seed-bed in a cultivated
field near Peru, Nebraska, a station about 20 miles south of Nebraska City.
The mellow silt-loam soil, underlaid at a depth of 1 to 1.5 feet with a loess
of very loose texture, not only absorbs water readily, but has a high water¬
holding capacity. This ranges from 57 to 64 per cent and is rather uniform
to a depth of at least 4 feet, the same type of subsoil extending to depths of
many feet. The mechanical analysis of this soil shows that it is approximately
one-half silt, while the remainder is composed almost entirely of very fine sand
and clay. Weaver has shown that in this soil-type roots of species common
to both true and subclimax prairie penetrate more deeply than in the stiffer
silt loam at Lincoln (1919:15). Normally, the rainfall is about 5 inches
greater than at Lincoln; however, during 1921 it was not only deficient, but
also poorly distributed. During April, the rainfall was nearly 3 inches below
normal, 70 per cent of it falling at one time. For May it was 1.6 inches below
normal, 72 per cent of it falling between May 7 and 10. June, with 3.4
inches of precipitation, practically all of which occurred before the middle of
the month, had a deficiency of 1.4 inches. The holard was much below nor¬
mal and the poor growth of crops marked the season as one of distinct drought
(Weaver, Jean, and Crist, 1922:78), though the native species made a fairly
good growth. When they had reached an age of 90 days (July 18-19), the
development of the plants both above and below ground was recorded (table
28). The development below ground was marked and agreed well with pre¬
vious findings as to the deep-seated nature of mature root systems in this
mellow loess soil.

Seedling Transplants.

Seedlings of 14 species were transplanted to the gravel-knoll on May 9.
Agropyrum glaucum, Aristida purpurea, and Stipa viridula had died by May
25, Elymus canadensis and Koeleria cristata succumbed early in June, and

EDAPHIC STATIONS! SUMMARY.

Bouteloua racemosa and Stipa comata later in the month. Liatris punctata
and Stipa spartea died in July and Andropogon furcatus in August. By the
end of the summer Bouteloua gracilis and Liatris scariosa were represented
by remnants only, and both species died later in the year. Bouteloua hirsuta
came through the summer with only one small clump about 3 inches tall and
Andropogon nutans with two, 6 or 7 inches high, the survival for the first
season being 14 per cent.

Table 28. —Development of seedlings at Peru, Nebraska, July 19, 1921.

Species planted
Apr. 19.

Development
of shoots.

Work¬
ing
depth
of roots.

Maxi¬
mum
depth
of roots.

Lateral

spread

roots.

in.

Andropogon nutans..

Good stand; luxuriant growth; ave.
height 11 in.; max. height 16 in.

Aristida purpurea. . .

Good growth; height 4 to 6 in.; 12 to 30
tillers per plant.

• •

Bouteloua gracilis. . .

Good stand; lvs. 4 to 8 in. high; exceed¬
ingly well tillered; flower-stalks about

9 in. tall.

Elymus canadensis. .

Vigorous plants; height 10 to 16 in.; 3
to 4 tillers per plant; heads appearing.

Liatris punctata.

Good growth; 3 to 5 leaves per plant;

3 to 4 in. tall.

Stipa viridula.

Thin stand, vigorous growth, ave. height

7 to 9 in.; 6 to 10 tillers per plant.

Seedlings that were transplanted to low prairie at the same time made a
much better growth. All grew well, but by the end of June Agropyrum, Boute¬
loua racemosa, and Elymus particularly, showed the effects of shading. In
July, through error, a part of the prairie into which the transplant area ex¬
tended was mown. Those in the area of better illumination made a fine
growth, the shorter boutelouas reaching heights of 4 to 8 inches and the
andropogons 6 to 12 inches. In the shaded area Agropyrum died and the
other species made but a poor to fair growth, the leaves being thin and atten¬
uated; the loss of only 1 of the 13 species is remarkable.

All of the seedling transplants at Nebraska City did very well until about
the middle of June, when Agropyrum, Koeleria, and Liatris punctata died
and by July 15, Bouteloua racemosa had likewise succumbed. The other
boutelouas, Aristida, and the andropogons, however, were in a flourish¬
ing condition. Elymus, Stipa comata, and S. viridula had been badly eaten
back by grasshoppers and were represented by remnants only, as were also
Liatris scariosa and S. spartea. Elymus and Stipa viridula died in August,
and S. comata was found to be in very poor condition in the autumn, but the
other 7 species were doing quite well.

Summary.

A summary of the experiments for this year is given in table 29. The
results from the Colorado Springs station are omitted because the area was
broken into by cattle and closely grazed. Owing to the drought at Nebraska
City, the percentage of germination is lower here than in low prairie at Lin-

EXPERIMENTS DURING 1921.

coin. However, the percentage of establishment averaged slightly higher
at the Nebraska City station, although growth was somewhat poorer. The
germination averaged slightly lower on the gravel-knoll (where surface sow¬
ing was omitted) than at Nebraska City, while the establishment was decidedly

Table 29. —Summary of sowing experiments, 1921.

Method of
planting.

Percentage of germination.

Percentage of germinated species
established.

Gravel-

knoll.

Low

prairie.

Nebraska

City.

Gravel-

knoll.

Low

prairie.

Nebraska

City.

Surface.

Trench.

Denuded quadrat....

Average.

Seedlings.

• •

lower. The survival of transplanted seedlings was highest on low prairie
(92 per cent), intermediate at Nebraska City (57 per cent), and least on the
gravel-knoll (14 per cent).

Of the survivors of the surface-sown species of the preceding year (1920) on
low prairie, Koeleria was winterkilled. One of the 4 survivors at Nebraska
City died. Of those sown in the trench, both survivors of 1920 on the gravel-
knoll died, 2 of the 4 on low prairie, and 3 of the 6 at Nebraska City.
Among the species planted in the denuded quadrats during 1920, none
persisted on the gravel-knoll. One of the 10 survivors on the low prairie
died during 1921, and also 1 of the 10 at Nebraska City.

Table 30. —Comparison of germination and growth at the several stations, 1921.

Average per cent of germination.

Average per cent of establishment of
germinated species.

Lincoln high prairie. 81

Lincoln low prairie. 76

Phillipsburg. 68

Nebraska City. 65

Lincoln gravel-knoll. 63

Burlington. 43

Phillipsburg. 65

Nebraska City. 41

Lincoln high prairie. 40

Lincoln low prairie. 38

Gravel-knoll. 29

Burlington. 7

Per cent of establishment in denuded
quadrats:

Per cent of establishment of seedlings:

Lincoln low prairie. 83

Lincoln high prairie. 65

Phillipsburg. 63

Nebraska City. 60

Gravel-knoll. 40

Burlington. 21

Lincoln, low prairie. 92

Phillipsburg.,. 75

Nebraska City. 57

Lincoln, high prairie. 54

Gravel-knoll. 14

Burlington. 0

If the data from the other stations are included and arranged in the order
of the average percentage of germination (table 30), they prove to be in
general agreement with those of the preceding year (p. 48), with two excep¬
tions. Owing to the drought, Nebraska City ranks after Phillipsburg, in¬
stead of heading the list, while Phillipsburg falls behind low prairie in rank.

When the stations are arranged according to the average percentage of
establishment, Phillipsburg ranks highest with 65 per cent (largely because

EDAPHIC STATIONS! SOD TRANSPLANTS.

of the excellent establishment of surface-sown plants). The Lincoln stations
and Nebraska City are about the same (38 to 41 per cent), while the gravel-
knoll and Burlington have changed in relative position, the latter giving the
lowest percentage of establishment. On the basis of establishment in de¬
nuded quadrats, the same sequence occurs as during 1920, except that Ne¬
braska City then ranked first and the gravel-knoll last. If an average is
obtained from all four criteria for growth, the stations aline them¬
selves as follows: Low prairie 72, Phillipsburg 67, high prairie 60, Nebraska
City 56, gravel-knoll 37, and Burlington 18.

Sod Transplants.

Gravel-knoll.

Between March 20 and 31, 18 blocks of sod, including 14 species from
the various Lincoln stations, were transplanted to the gravel-knoll.
Distichlis spicata died in June, while Agropyrum glaucum, Andropogon sco-
parius, and two lots of Koeleria cristata died during August, notwithstanding
the relatively favorable season for growth. With the exception of Poa
pratensis, all of the other species made a fair growth, but the scarcity of water
was indicated by the rolling of leaves and frequent wilting (table 25). Andro¬
pogon furcatus and A. nutans reached a height of a foot and Panicum virgatum
one of 1.5 feet, but did not put forth flower-stalks. Bouteloua hirsuta and
racemosa had a few flower-stalks each at 1 and 2 feet respectively, while
Bulbilis blossomed profusely in June, but like the other species did not
increase its area. Elymus bore several fine heads at a height of 2 to 2.5 feet
and Spartina cynosuroides from the swamp flowered at 27 to 34 inches; Poa
and Stipa also blossomed.

Poa and both clumps of B. racemosa were winterkilled. Elymus died during
the August drought, but all of the others made a fair to good growth, although
adversely affected by the late summer drought. Seven of the 8 species bloomed,
Andropogon furcatus from the low prairie alone failing to put forth flower-
stalks. All survived the drought of the following fall and winter. Panicum
alone made a poor growth the following season, while Andropogon furcatus,
Bulbilis , and Spartina spread a foot or more beyond their original area. By
August 25, Stipa and Bulbilis were the only species that gave signs of having
flowered or preparing to flower, a striking contrast to the preceding year,
when drought promoted reproduction. Even during the very favorable year
of 1923, but more especially during preceding seasons, sod transplants on the
gravel-knoll were all much shorter than elsewhere. They also began renewed
development earlier and produced flower-stalks sooner.

Of the species planted in the spring of 1920, although suffering severely
from drought, only one sod of Poa and another of Andropogon scoparius in
the unwatered row died (table 66). One of Koeleria was winterkilled, while
a sod of Elymus and another of Koeleria died during August of the second
season. Practically all of the species, including Panicum virgatum and
Spartina cynosuroides, made a fair growth, rolling of leaves and wilting
occurring much more rarely than in transplants of the current year, owing to
the well-established root systems. Exceptions to this occurred in the case of
Agropyrum, both lots of which did very poorly, as did also one sod of Elymus
and the remaining block of Koeleria. After the middle of August, when the

EXPERIMENTS DURING 1921.

short-grass cover of the knoll began to dry, most of the tail-grass sods began
to change color and enter the winter condition also. One lot each of A.
furcatus, A. scoparius, B. racemosa, Elymus canadensis, and Poa pratensis and
all three lots of Panicum virgatum had flowered and set seed, although the
flower-stalks were fewer in number and smaller than normal. Of the species
that seeded the first season after transplanting, viz, Koeleria, Stipa, Agro-
pyrum, Elymus, Panicum, and Poa, only the last three seeded the second
year. The single remaining lot of Koeleria and Poa were both winterkilled
and one Agropyrum died in June 1922. Both lots of Elymus died later in the
summer, as did also one of Andropogon furcatus. The effect of drought upon
the other species is shown in table 66. So well were these species rooted,
however, that the dry fall and winter of 1922-23 took but a single Andropogon
scoparius. The growth during 1923 was marked, Panicum doubling and
Andropogon nutans tripling its area.

Low Prairie.

Fifteen species of grasses were transplanted into the low prairie late in
March 1921. The rank growth of the native tail-grasses produced a very
dense shade (plate 6a). Bouteloua hirsuta, Distichlis spicata, and Koeleria
cristata had died by late summer and Agropyrum had made but poor growth.
Bouteloua gracilis and Bulbilis dactyloides both reached a height of 12
inches, the leaves being much attenuated, but neither was able to increase
its territory. Andropogon furcatus, A. nutans, Bouteloua racemosa, Spartina
cynosuroides, and Stipa spartea all developed about normally, but none blos¬
somed. Bulbilis, Poa, Andropogon scoparius, Elymus, and Panicum bore
well-developed flower-stalks and seeded rather profusely. During the follow¬
ing May and June, owing to the dense cover and resulting competition,
Andropogon scoparius, Bouteloua racemosa, B. gracilis, and Stipa spartea, all
high-prairie species, died. Bulbilis died in July, while by the end of the
summer Agropyrum, one Elymus , and Poa were represented by remnants only.
Elymus, Panicum, and Spartina alone blossomed. All these survived the dry
winter, but by August 1923, Agropyrum was dead and Poa nearly shaded out.
Both lots of Elymus were heading at 40 to 42 inches high, while the other
subclimax species, viz, Spartina, Panicum, Andropogon nutans, and A.
furcatus, were all flourishing.

As to the 1920 transplants at this station, all made a good growth, and all
except Spartina cynosuroides bloomed (table 67). By May 18, 1921, Bouteloua
racemosa, one lot of Koeleria, and two of Stipa spartea were dead. Many
others, particularly Agropyrum, Bulbilis, Distichlis, and Stipa, were suffer¬
ing from shading and invasion, but all other species made a good growth
during June and July. By the end of the summer, one Agropyrum had
died, and the other was much attenuated. Bulbilis, Distichlis, and Koeleria
were nearly dead, being represented by remnants only, and Bulbilis alone had
blossomed. However, all of the other species had made an excellent growth.
The three andropogons exhibited abundant flower-stalks 3 to 5 feet high, and
Bouteloua gracilis had leaves a foot tall, but flowered sparingly. Elymus and
Panicum possessed flower-stalks in profusion 4 to 5 feet tall, and Stipa seeded
at 33 to 36 inches, while Spartina made a good growth but did not blossom.

By May of the following spring (1923), one Andropogon scoparius and
Koeleria had died in addition to Distichlis. Panicum virgatum and Andro-

EDAPHIC STATIONS: SOD TRANSPLANTS.

pogon nutans and one lot of A.furcaius were indistinguishable from the general
cover. Poa pratensis and another Andropogon scoparius died in July, and
Agropyrum glaucum and Bulbilis by August. At this time the following were
in rather poor condition: one Andropogon scoparius, both plants of Bouteloua
gracilis, and Elymus and Koeleria, but the rest had made a good growth.
By the spring of 1923, Bouteloua gracilis and one Andropogon scoparius
were dead. By the end of this, the fourth season, sufficient time had elapsed
for a fair adjustment of the species to the low-prairie habitat. Although
the remaining high-prairie species (two lots of Stipa and one of Koeleria)
were still in fair condition, the last Andropogon scoparius was in straits.
The other species of the subclimax prairie, viz, Andropogon furcatus, Elymus
canadensis, and Spartina cynosuroides, were flourishing, while Andropogon
nutans and Panicum virgatum had already become indistinguishable from
the native sod. The following had entirely lost out in the struggle for light:
Agropyrum, Bouteloua gracilis, Bulbilis, Poa, Bouteloua racemosa, and Dis-
tichlis spicata. As a consequence, these experiments clearly reveal the effect
of competition in sorting out species and stabilizing vegetation.

Salt-flat.

Sods of each of the preceding species were also transplanted into the salt-flat
late in March, 1921. During June most of the plants suffered from
drought, great cracks appearing on the sides of the transplanted blocks, owing
to the peculiar nature of the soil (p. 44). Growth was very poor, wilting and
even dying of parts being not uncommon. Throughout July, growth was
slow, most of the sods making only a sparse growth and Spartina alone
flourishing. Koeleria and Stipa failed to blossom or bore only a few dwarfed
flower-stalks, and the inflorescence of Bulbilis was also shorter than normal.
By the end of the summer, the leaves of Andropogon furcatus, A. nutans, and
Bouteloua racemosa were dying back, while Panicum virgatum and B. hirsuta
were nearly dead. These, like Agropyrum, B. gracilis, and Spartina, had
failed to blossom. Andropogon scoparius, Elymus, and Poa had dwarfed
flower-stalks and Distichlis alone developed normally.

As a result of the drought of the following June, all the sods were in poor
condition, large cracks occurring around most of the transplanted blocks,
even those that had been transplanted 2 or 3 years, and many of them were
very dry. One Koeleria died in June and the other two did very poorly.
Poa and Panicum were represented only by remnants at the end of the season
and the growth in many of the other blocks was very sparse. Practically
all of the plants were dwarfed and only 4 species blossomed, viz, Andropogon
scoparius, Bulbilis, Elymus, and Koeleria; on these the inflorescences were
much dwarfed.

During the very favorable growing-season of 1923, another Koeleria and
Stipa died and Andropogon furcatus, A. scoparius, Bouteloua racemosa, Panicum
virgatum, and Spartina cynosuroides were represented by remnants only.
Seven species seeded, although not abundantly.

The sods transplanted into the salt-flat during 1920 made a poor growth
the first season, though 8 of the 13 species blossomed (table 68). Compared
with the growth of similar species in the adjoining low prairie, nearly all were
greatly dwarfed. However, they survived the first season. During May
1921, Andropogon scoparius, A. nutans, Stipa spartea, one lot of Elymus

EXPERIMENTS DURING 1921.

canadensis, and one Koeleria cristata were in poor condition or dying. How¬
ever, many of the others, including Panicum and Spartina, were doing quite
well. Throughout June and July they grew fairly well, except A. scoparius,
which was mostly dead and badly invaded by Distichlis. All lived through¬
out the second season but Bouteloua racemosa, though none grew normally
except Distichlis. Agropyrum, A. scoparius, A. nutans, B. racemosa, Spar¬
tina, and Stipa failed to flower. The dwarfed condition of the remaining
species, all of which blossomed, was well represented by A. furcatus, which
bore flower-stalks only about 2 feet high. Panicum was an exception, with
a few stalks reaching a height of 40 inches, but its vegetative growth, as for
all of the other species, was much below normal. Elymus canadensis failed
to grow in 1922, while one Panicum virgatum died in June. Although the
plants in general made a fair growth, they were dwarfed and several
species did poorly. Andropogon furcatus, Distichlis, Koeleria, Panicum, and
Poa blossomed. Owing to an excellent holard, all the species except
Elymus, Stipa, and Andropogon nutans made a good growth during 1923.
Consequently, it seems evident that several species can tolerate the saline
areas dominated by Distichlis and Agropyrum, as was clearly shown during
the wet year of 1923. As a result, the salt-basin is being gradually invaded
by various prairie species.

Salt-basin.

A similar lot of sods was transplanted to an area in the salt-basin just west
of Lincoln, which is dominated by Dondia depressa. This halophyte occu¬
pies an intermediate zone between the less saline Distichlis area and the wetter
and more saline one of Salicornia herbacea. Repeated analyses of the soil
in the Dondia area show that it has a salt-content of 1.8 to 2.0 per cent, prac¬
tically all of which is sodium chloride. The holard in spring and early summer
is more or less favorable for growth. Dondia, Salicornia, and Distichlis
were also transplanted early in April, together with the sods of the grasses.
Within 20 days the transplants from non-alkaline soil began to lose their
green color and the tips of the leaves to dry, and after another 20-day period
all were dead. The three halophytes did very well throughout the season,
except Salicornia, which suffered apparently from drought late in the
summer.

Swamp.

Sods of the same species were transplanted into the swamp in March
1921. Owing to the building of a dam higher up the ravine, conditions were
markedly different from the preceding year. The soil in the wettest
area was not saturated to the surface throughout the spring, although water
stood on the top during the first week in May. For much of the remainder
of the season the area was relatively dry, the holard being so low that water
could not be pressed out by hand from the surface 6 inches. The Poa zone
was also drier than the preceding year. This area, which was nearly pure
bluegrass at the beginning of the experiments (1920), was now invaded by a
rank growth of Spartina, and light conditions were less favorable than for¬
merly.

In the swamp all of the grasses made a very good growth until the end of
June, Bulbilis flowering profusely, but Stipa and Koeleria, other early bloom-

EDAPHIC STATIONS! SOD TRANSPLANTS.

ers, failed to bloom. At this time Spartina was 3 feet high all around the
sods and the transplants began to show the effects of shading, the light values
falling as low as 2 to 5 per cent. By the end of July the surrounding vegeta¬
tion was 4 feet tall and some of the sods were found with difficulty. The
soil was wet, and hence many of the plants were slender, with narrow leaves.
Late in August, Agropyrum, Bouteloua hirsuta, B. gracilis, Distichlis, and
Koeleria were dying back or mostly dead, some being represented by rem¬
nants only. They were all attenuated and none had produced flower-stalks.
Andropogon nutans, A. scoparius, B. racemosa, Spartina, and Stipa had all
made excellent growth, but none produced flowers. Bulbilis, Elymus, Poa,
and Panicum also grew well and blossomed about normally.

The sods in the Poa zone, which were planted at the same time, made a
good growth early in the season. By the end of June, Spartina, Cyperus,
Mentha, etc., had reached a level of about 3 feet and the shade was dense.
Bulbilis and Bouteloua gracilis both had delicate slender leaves 10 to 14 inches
tall, while those of Koeleria reached 15 inches. None had blossomed, but
Stipa had a few fruits which were smaller than normal. On August 31, one-
half of Bouteloua gracilis was dead and Distichlis was nearly so. However,
most of the other species, though slender, had made in almost every case a
much better growth than in the wetter part of the swamp. Six of the 13
species, viz, Andropogon scoparius, A. nutans, Stipa, Distichlis, Elymus, and
Panicum, blossomed, in contrast to 4 of the 14 species in the swamp.

In 1922, the swamp area was much drier than during preceding years; in
fact there was nowhere excess holard or deficient aeration, light being the
controlling factor. Spartina was over 2.5 feet tall by the middle of June
and 4 feet in August, covering the area with a dense growth. Conditions
were very similar in the Poa zone. The leaves on practically all of the plants
were very narrow, long, and delicate, Andropogon scoparius did not appear
in the spring, and Koeleria died in August. The following early bloomers
produced flowers: Bulbilis, Koeleria, Stipa, and Poa, while the subclimax
dominants Elymus and Panicum also blossomed. By June of 1923, Boute¬
loua hirsuta, Bulbilis, and Distichlis were dead, while Bouteloua gracilis, B.
racemosa, Elymus, and Stipa vanished later in the summer. This left a few
remnants of Agropyrum and well-developed plants of Andropogon nutans,
Panicum virgatum, and Spartina cynosuroides at the end of the third season.
In the Poa zone all survived the 1922 season, but the following species died
the next summer: Andropogon nutans, A. scoparius, Bulbilis, Bouteloua
gracilis, Distichlis, and Elymus. Agropyrum,, Bouteloua racemosa, and Stipa
spartea were represented by remnants only, while Andropogon furcatus, Pani¬
cum, and Spartina were the only species that had developed normally.

The fate of the sods planted in the swamp during 1920 is instructive. A.
scoparius, A. fur catus, A. nutans, B. gracilis, B. racemosa, Koeleria, and Stipa
died the first season, while many of the others came through in very poor
condition (table 69). In May of the following spring Agropyrum and Elymus,
and especially Poa, showed signs of distress. Spartina developed nor¬
mally, while Panicum and Distichlis were in good health. Both Agropyrum
and Poa had died by the end of July; autumn found remnants only of Distich¬
lis and Elymus, but Panicum and Spartina had made a good growth and
seeded. During 1920, Koeleria, Stipa, Andropogon scoparius, and A. nutans

EXPERIMENTS DURING 1921.

died in the Poa zone. The following spring Agropyrum glaucum and Bouieloua
racemosa failed to appear, while B. gracilis died in June after a feeble
growth. Distichlis succumbed in August, but Elymus, Poa, Panicum , and
Sparlina made a fairly normal growth, all ripening seed.

Since the swamp was much drier and the air-content higher, the chief
factor inhibiting growth was not aeration but light (p. 73). Spartina be¬
came indistinguishable from the surrounding vegetation. Panicum, blos¬
somed normally, Elymus bore a single flower-stalk 2 feet high, while Distichlis
did very poorly. In the Poa zone, Poa and Elymus made an excellent growth,
the former merging into the common sod and the latter flowering at a height
of over 4 feet. Panicum and Spartina developed normally in every way.
In 1923 Distichlis died, but the three low-prairie species developed very well,
the 4 species in the Poa zone also prospering. The results are in accord with
the expectation, since all the species except those of the mesophytic subclimax
prairie, viz, Andropogon furcatus, Elymus, Panicum, and Spartina, had lost
in the struggle during the period of 3 or 4 years.

Regional Transplants.

As to the fate of the sods transplanted to Colorado Springs in 1920, both
blocks of Stipa spartea had died by the end of 1921, as well as one each of the
following species: Andropogon furcatus, A. scoparius, Bouieloua racemosa ,
Elymus, Koeleria, and Panicum. This was chiefly due to drought, but partly
also to overgrazing, stock having broken into the inclosure at several times.
In 1922, Bulbilis alone blossomed, repeated grazing again weakening the other
plants. By the autumn of 1923, grazing having been continued, the last lot
of Elymus had died, several of the other species were represented by rem¬
nants only, and all were in rather poor condition.

During 1921, sods of Stipa setigera, S. eminens, Poa tenuifolia, and Melica
imperfecta from California were transplanted into high and low prairie re¬
spectively on June 23. These were received at Lincoln on May 18, but were
kept in a garden and watered frequently before the large blocks of soil con¬
taining the transplant were removed to the native grassland. Here they were
again watered from time to time as necessary. Only a few made a feeble growth
and all on the high prairie were dead by July 30. On the low prairie four lots
put out new shoots 3 to 5 inches long, but these died by the end of the summer.

On July 20 the following species from Arizona were transplanted in quad¬
ruplicate into a garden at Lincoln and kept well watered: Bouteloua eriopoda,
B. rothrocki, B. bromoides, and Aristida divaricata. All made some growth. In
September one clump of B. eriopoda bore flower-stalks about 15 inches tall,
and one B. bromoides possessed leaves 9 inches high and flowered abundantly
at 15 to 22 inches.

4. EXPERIMENTS DURING 1922.

PHYSICAL FACTORS.

Rainfall.

The season of 1922 was fairly favorable for growth, except the latter part,
when severe drought occurred at all the stations, being relatively less marked
at Burlington. At Lincoln the combined precipitation for March and April
averaged about normal; that of May and June was approximately 1 and 2
inches below the mean respectively, July had an excess of 2.4 inches, but only
0.7 of the 3.7 normal rainfall of August occurred (fig. 9). At Phillipsburg the
April precipitation was nearly 1.5 inches above normal (no report for May),
June, July, August, and September showing deficiencies of 2.4, 0.6, 1.8, and
1.6 inches respectively (fig. 15). The Burlington precipitation was about
half the normal during March (April report missing), 0.4 inch in excess during
May, while June gave an excess of 1.4 inches over the mean. July had a
deficiency of 0.8, August an excess of 0.5, and September was also far below
the normal (fig. 16). Spring and early summer drought periods, i. e., rainfall
less than 0.21 inch, occurred at Lincoln from April 11 to May 8 and May 26
to June 24. At Phillipsburg a 10-day drought occurred in April (May report
missing) and no rain fell from June 1 to 26. The severity of the late-summer
drought at both Lincoln and Phillipsburg is well illustrated by the fact that
by the end of August the late-developing Andropogons were already half dried
and appeared to be entering the winter condition. No efficient precipitation
fell at Burlington from May 2 to 19 and 20 to 29 (April report missing) or
June 2 to 15.

Holard.

A study of the water-content (table 31) shows that a margin of 5 per cent
(more usually 7 to 11 per cent) above the hygroscopic coefficient existed at all
times to a depth of 4 feet in the high prairie at Lincoln (except in the surface
6 inches on June 22).

At the mixed-prairie station a good water-content prevailed to 3 feet in the
spring, but the margin of safety was much less at nearly every determination
than at Lincoln. Late in June, and again throughout July and August,
drought prevailed, no water being available often to depths of 4 feet. Condi¬
tions at Burlington were not very different, the soil in spring and early summer
being as usual quite moist. However, the deficiencies were very marked and
practically continuous after June 18.

Temperature.

Air-temperatures in general were lowest at Burlington, 7° F. higher at
Lincoln during early May, and usually 3 to 5° higher during the rest of the
season; those at Phillipsburg exceeded those at Lincoln by 2 to 5°, especially
after June 7 (fig. 38). The night temperatures at Burlington were decidedly
lowest (by 2 to 10°), those at Lincoln intermediate, and at Phillipsburg
highest (fig. 39). A similar general relation, but one less marked and with
overlappings, holds for the day temperatures. Soil-temperatures at a depth
of 3 inches were obtained until the middle of June, after which the thermo¬
graph bulbs were buried at a depth of 12 inches and left for the remainder of

EXPERIMENTS DURING 1922.

the season. In figure 40 it may be seen that the moister soil at Lincoln was
also that with the lowest temperature, although the differences, as in the case

Table 31 .—Holard in excess of the hygroscopic coefficient at the several

stations, 1922.

Lincoln, Nebraska.

Date.

0 to 0.5

0.5 to 1

1 to 2

2 to 3

3 to 4

foot.

feet.

Apr. 30.

15.3

15.8

14.7

10.9

9.1

May 10.

10.1

13.7

15.3

. . . .

....

May 17.

12.0

13.1

14.1

12.1

7.8

June 7.

6.4

11.2

10.9

. . . .

June 14.

6.8

11.3

10.5

8.4

6.6

June 22.

3.5

6.3

11.1

....

July 6.

5.7

4.7

8.9

8.0

5.4

July 13.

17.8

11.7

8.1

. . , .

, . . .

July 20.

16.4

10.7

6.2

....

July 27.

21.2

14.7

7.5

. , . .

. • • •

Aug. 3.

8.2

10.1

7.4

8.0

6.0

Aug. 10.

5.6

6.2

7.4

....

. . . .

Aug. 17.

0.9

1.8

0.8

. . . ,

....

Aug. 24.

3.2

5.4

5.1

7.1

5.3

Hygroscopic coeff....

9.8

10.9

10.1

10.0

10.3

Phillipsburg, Kansas.

Apr. 28.

18.8

19.4

14.1

1.5

1.3

May 19.

5.4

10.2

9.3

6.2

0.4

June 11.

1.7

9.5

7.8

8.2

0.5

June 17.

1.4

4.4

....

. . . .

June 24.

-1.6

0.9

....

. . . .

June 30.

10.8

1.1

3.7

5.1

4.7

July 17.

-1.7

0.2

-0.5

-0.2

0.9

July 28.

-0.1

1.6

, . . .

. . . .

. . , .

Aug. 3 .

0.5

0.7

-0.5

-0.1

0.0

Aug. 11.

-1.5

-1.8

-1.3

....

. . . .

Aug. 17.

-3.7

-1.5

....

. , . .

. . , .

Aug. 27.

-2.6

-1.4

0.7

-1.7

0.2

Hygroscopic coeff....

10.6

10.9

10.6

10.7

Burlington, Colorado.

Apr. 30 .

17.6

16.9

8.5

-1.0

May 20.

19.8

9.6

8.7

6.8

-0.4

June 4.

12.2

10.3

8.3

6.8

. . . .

June 11.

-0.1

1.2

4.5

4.9

-1.1

June 18.

0.4

1.1

3.8

....

June 25 .

0.2

0.0

1.2

2.1

-0.2

July 2.

11.1

-1.8

-2.6

-0.6

-1.3

July 16.

-2.5

-1.9

-0.5

0.0

-1.0

July 23.

6.7

-1.1

-0.1

1.1

. . . .

July 30.

-0.7

-0.2

. . . .

Aug. 4.

5.4

-1.5

-1.9

-0.4

0.6

Aug. 20.

-1.7

-1.9

-1.1

-0.2

. . . .

Aug. 26.

-0.3

-2.1

-2.7

-1.9

-0.4

Hygroscopic coeff....

10.9

12.2

12.0

11.4

of air-temperatures, are not great enough to be of much importance in the
establishment and growth of native vegetation.

PHYSICAL FACTORS.

Humidity.

Continuous records of humidity were not obtained during 1922, but a
comparison of the hygrograph records from Burlington and Lincoln, from
which the average day and average night humidities have been obtained,
shows striking differences. With few exceptions, the air at the former was

Fig. 38.—Average daily temperatures at Lincoln (solid line), Phillipsburg (long broken lines), and

Burlington (short broken lines), 1922.

13 to 25 per cent drier by day (table 32). Owing to the high altitude (4,160
feet) and cool nights at Burlington, differences in night humidites were less
marked. Frequent isolated readings at Phillipsburg, when compared with the
records at the other stations, showed that they were usually intermediate.

Table 32 .—Average day and night humidites at Lincoln and Burlington, 1922.

Date

Average day humidities.

Average night humidities.

Lincoln.

Burlington.

Lincoln.

Burlington.

p. ct.

May 1 to 7.

44.7

45.5

73.5

77.3

8 14.

50.5

26.2

77.5

58.2

15 21.

44.5

39.7

70.9

74.5

22 28.

77.4

64.5

89.0

87.9

29 to June 4.

64.2

• • • •

90.2

• . » •

June 5 to 11.

58.3

45.8

87.5

75.4

12 18.

60.6

40.5

86.6

74.2

19 25.

48.8

37.2

73.2

73.0

26 to July 2.

61.4

• • • •

85.3

• • • •

July 3 to 9.

60.8

51.0

78.5

82.2

10 16.

71.0

• • • •

93.0

• • • •

17 23.

70.6

44.5

89.4

78.5

24 30.

72.2

46.1

92.3

78.3

31 to Aug. 6.

66.0

....

90.2

• . . •

Aug. 7 to 13.

55.3

61.0

85.3

90.6

14 20.

54.0

42.8

81.0

62.1

21 27.

50.2

• • • •

80.0

• • • •

28 30.

56.3

....

87.8

....

EXPERIMENTS DURING 1922.

Evaporation.

The average daily evaporation (fig. 41) shows again the relative xerophytism
of the several communities, as well as the periods of stress in June and August,
although that at the two western stations remained rather high throughout
the growing-season. The evaporation at Lincoln ranged from 9 to 33 c. c., at
Phillipsburg 8 to 41 c. c., and at Burlington from 15 to 47 c. c. In general,
the losses were higher than for the preceding year (fig. 23).

Fig. 39.—Average day (heavy lines) and night temperatures (light lines), at Lincoln (solid line), Phil¬
lipsburg (long broken lines) and Burlington (short broken lines), 1922.

To summarize, the season of 1922 was less favorable for growth than 1921,
and deficiencies in rainfall and soil moisture were marked, being pronounced
even at Lincoln, the least xerophytic of the stations, during late summer.
Humidity and holard were again the controlling factors in plant growth,
conditions as regards both being progressively more severe westward.

PLANTING RESULTS.

Surface Sowing.

Twenty-two species of grasses, forbs, and trees were planted on the sur¬
face of the high prairie at Lincoln on April 18; 82 per cent of the species ger¬
minated, only Aster muitiflorus, Bouteloua gracilis , Solidago missouriensis, and
Sporobolus asper failing to do so. However, germination was considerably
delayed because of dry weather. All of the species did at least fairly well
until July, Lespedeza, Liatris spp., Onagra, and Pinus having succumbed by
the 14th, while 4 other species were suffering badly. The drought period,

SURFACE AND TRENCH SOWING.

May 26 to June 24, was quite severe. During the latter part the long-establish¬
ed native grasses on the high prairie had rolled leaves and Erigeron ramosus
was drying, many without flowering. It seems probable that most, if not all, of
the surface-sown species would have succumbed except for regular watering.
Late August found Bouteloua hirsuta, Kuhnia, Ratibida, and Robinia added to
the mortality list. Thus, the mortality during the first summer was 50 per
cent, notwithstanding regular watering at critical periods throughout the
season. Liatris scariosa died in August 1922, and Bouteloua hirsuta did
very poorly, but the 5 remaining species flourished, except during the late-
summer drought, Aristida purpurea, Bouteloua racemosa, and Elymus cana¬
densis producing seed. Andropogon nutans, Bouteloua hirsuta, and Bouteloua
racemosa died during the dry fall and winter. During the wet summer of 1923
the shade was very dense. At the end of the season Elymus canadensis was
represented by a single weak plant; Liatris punctata, which was growing in a
dense sod, was over a foot tall, but failed to blossom; but Aristida purpurea,

Fig. 40.—Average daily soil temperatures at depths of 3 and 12 inches respectively
at Lincoln (solid line), Phillipsburg (long broken lines), and Burlington (short
broken lines), 1922.

the other survivor, developed normally and produced an abundance of seed.
It seemed only a question of a year or two more, however, until all three would
disappear from the stabilized grassland.

Trench Sowing.

Twenty-three of the same species sown on the surface were also planted in a
trench on the high prairie at Lincoln on April 18. Seven, viz, Amorpha can-
escens, Aster multifiorus, Bouteloua hirsuta, B. gracilis, Onagra biennis, Rati¬
bida columnaris, and Solidago missouriensis, failed to germinate. Of the 70
per cent germinating, Agropyrum glaucum and Liatris punctata died before the
end of the June drought and many others wilted badly. Lespedeza capitata,
Muhlenbergia pungens, Petalostemon candidus, and Sporobolus asper succumbed
by the middle of July, while the remaining ten species (63 per cent) lived
throughout the summer. Most of them, however, were in poor condition,
being badly dried by August 21.

EXPERIMENTS DURING 1922.

Plants in the trench at Phillipsburg did much more poorly. Of the 23
species planted on April 29 only 8 (35 per cent) germinated. One-half of the
trench was rather densely shaded by Bouteloua racemosa and Andropogon
furcatus, and germination in it was very low. Only about hah of those that
germinated did so abundantly. Liatris scariosa and Sporobolus asper died
by June 30, and Agropyrum glaucum in July, while Elymus canadensis, Des-
modium canescens, and Kuhnia glulinosa succumbed in August. Andropogon
nutans and Bouteloua racemosa were the only ones that survived the first
summer. Although the prairie had been burned over the preceding winter
through accident, by May 27 the general level of the grasses had reached 5
or 6 inches, while an upper story of Psoralea, Helianthus, Erigeron, and flower-
stalks of Boa and Siipa at 15 to 22 inches added to the shade. Consequently,
light early became an important factor.

c.c

Fig. 41. —Average daily evaporation at Lincoln (solid line), Phillipsburg (long
broken lines), and Burlington (short broken lines), 1922.

Twenty-two species were sown on the surface of the mixed prairie at Phil¬
lipsburg on April 29, the spring being late at all stations. Of these, 14 per
cent, including Aster multiflorus, Lespedeza capitata, and Solidago missouri -
ensis, failed to germinate. By June 30 (no efficient rain having fallen dur¬
ing June 1 to 26), 4 forbs, Pinus, and Sporobolus had succumbed. On
August 3, Agropyrum was found dead, and by the last of the month
Amorpha, Aristida, Petalostemon, and Ratibida had also succumbed. This
left 8 species (42 per cent) surviving, of which at least 4 were in very poor
condition. Shading was an important factor. As early as June 30 the grass-
level was 8 to 14 inches high, with subdominant herbs such as Psoralea at a
level of 24 inches.

Twenty of the same species grown at Lincoln and Phillipsburg were also
sown on the surface of the short-grass sod at Burlington on April 30. Ten
species (50 per cent) failed of germination. In fact, only Andropogon nutans,
Robinia pseudacacia, Desmodium canescens, and Liatris scariosa germinated
at all abundantly. Some species lay dormant until late in June, being
stimulated to grow by a heavy shower. By the middle of July only one
plant of Liatris scariosa remained alive, and it succumbed later. The follow¬
ing species alone survived in both the true and mixed prairies: Andropogon
furcatus, A. nutans, Bouteloua racemosa, Desmodium canescens, and Elymus

DENUDED QUADRATS.

canadensis. Shading played a minor role at Burlington, except in protecting
the plants from too rapid water-loss, since the short-grass sod was only 3
inches tall by July 1. The fate of the surface sowing of 1921 is of interest
here. At Lincoln 9 per cent survived, but none at Burlington, while, owing
to opportune showers and to a much more favorable light relation at Phillips-
burg than at Lincoln, 83 per cent survived the first season. Establishment
and growth in the mixed prairie, however, were better than on the short-
grass plains. Among 22 species planted in the trench at Burlington on April
30, 68 per cent failed of germination. However, 5 of the 6 species that grew
germinated rather abundantly. Robinia pseudacacia , Desmodium canescens ,
and Kuhnia glutinosa died by the middle of July, and Agropyrum glaucum
in August, while 5 plants of Andropogon nutans alone survived the summer.
Andropogon nutans was the only species that survived at all three stations,
reaching a height of 1 to 3 inches at Burlington, 5 to 8 inches at Phillipsburg,
and 5 to 9 inches at Lincoln. Bouteloua racemosa survived at Phillipsburg
and Lincoln, and 8 other species at Lincoln alone.

As to the 1921 trench plantings, the best results were obtained at Lincoln,
where 5 of the 11 species that germinated survived the first season.
Bouteloua gracilis died the following June, but the rest made fair to good
growth. During the favorable season of 1923 growth was excellent,
Andropogon nutans and Sporobolus asper reaching heights of 8 to 10 inches,
while Bouteloua hirsuta and Liairis scariosa were 5 to 7 inches tall, but none
flowered.

Of the 6 species that started growth at Phillipsburg, only Andropogon
nutans and Aristida purpurea became permanently established. Both did
quite well during 1922, but Andropogon , which was densely shaded, died in
late summer. Aristida was represented in 1923 by a small clump which
blossomed profusely at a height of 22 inches. No species planted in the
trench in 1921 survived at Burlington.

Denuded Quadkats.

Thirty-five species of grasses, forbs, shrubs, and trees were planted in
denuded quadrats in the true prairie on April 18. Among these, ten (29
per cent) failed to germinate. Brauneria pallida died early in June and
Acer negundo before the June drought was broken. During this period
many other species were wilted or drying. By the middle of July, Agropyrum,
Elymus, and Ulmus had also died, while Andropogon halli and Calamovilfa
longifolia were added to the mortality list by August 22. However, 18 spe¬
cies or 72 per cent survived the first summer, the list including trees and
forbs as well as grasses. Many of these were seriously affected by drought.

Of 32 species planted on April 29 in denuded quadrats in the mixed prairie,
10 (31 per cent) failed of germination. By June 10, Lespedeza and
Liatris scariosa had died, and Agropyrum , Muhlenbergia, and Ulmus
disappeared by the end of the month. Robinia dried out during July and
Acer negundo early in August. The end of August found Desmodium , Elymus ,
Onagra, and Pinus added to the list of non-survivors. This left 11 species,
50 per cent of those which germinated, that survived the first summer. As
usual, most of these showed by dead leaf-tips, rolled leaf-blades, or browning,
the effects of intermittent periods of drought.

EXPERIMENTS DURING 1922.

Twenty-eight species were sown in denuded quadrats at Burlington
on April 30; 50 per cent of them failed of germination. Among these

were Amorpha, Aster, Liatris, Ratibida, Redfieldia, and Solidago, all of
which had also failed at Lincoln and Phillipsburg, notwithstanding they
showed fair to good germination in the greenhouse as well as under other
methods of planting. Seeds of certain species kept on germinating until
July. By this time Acer saccharinum and Andropogon furcatus had died, but
no further losses were recorded until August, when Agropyrum and Robinia
also succumbed. In fact, except for the death of Aristida, no further loss
of an entire species occurred, although many plants were eaten off by grass¬
hoppers, badly wilted, or entirely destroyed by drought. Thus, 64 per cent
survived, showing at once the better conditions furnished by this method as
compared with that of surface or trench sowing. The surface of the soil in
these quadrats, as was true of those 2 or 3 years older, was always mellow;
Bulbilis had started to invade the older ones from all sides, but the invasion
was very incomplete, even on areas denuded for over 2 years. While the
grama in the buffalo-grass sod was flowering rather sparsely at 6 to 12 inches,
that around the edges of the quadrats, owing to an increased holard due to
lessened competition, was 6 to 20 inches tall.

Summary.

Andropogon nutans, Bouteloua gracilis, Bouteloua racemosa, Gleditsia tria¬
canthus, Kuhnia glutinosa, and Panicum virgatum grew throughout the sum¬
mer at all three stations. The following survived the summer at two: Des-
modium canescens, Muhlenbergia pungens (Burlington and Lincoln), and Acer
saccharinum, Andropogon furcatus, Aristida purpurea, Petalostemon candidus,
and Sporobolus asper (Phillipsburg and Lincoln). Andropogon scoparius,
Liatris punctata, Onagra biennis, Pinus ponderosa, and Robinia pseudacacia
survived only at Lincoln.

A comparison of the results shows that Bouteloua gracilis made the best
growth at Burlington (5 to 13 inches) and the poorest at Lincoln (2
to 5 inches). Gleditsia grew poorest at Burlington (2 to 5 inches)
and about the same at the other stations (3 to 7 inches). However,
all of the others, Andropogon nutans, Bouteloua racemosa, Kuhnia glutinosa,
and Panicum virgatum, showed the best growth at Lincoln (4 to 12 inches),
intermediate at Phillipsburg (3 to 9 inches), and least at Burlington (1 to
5 inches).

The fate of the species planted in quadrats in 1921 is of interest here. At
Lincoln, 13 of the 20 species that germinated survived the first season, but
the following spring 7 of these failed to come up. The 6 remaining
species did very well throughout the next summer, but none came into blos¬
som. In 1923, Andropogon nutans formed a dense sod; Gleditsia was repre¬
sented by a single remnant, and Stipa spartea by a scattered growth. The
other species did fairly well, but all were much shaded and none produced
seed. At Phillipsburg, 19 species germinated and 13 survived the first sum¬
mer; 6 of these, including 4 species of trees, were winterkilled, and
Stipa setigera died by the end of 1922. Agropyrum and both species of Sym-
phoricarpus did but poorly; Andropogon nutans formed a dense sod, while
Aristida and Elymus both ripened seed. Of the 19 species which germinated

SEEDLING TRANSPLANTS.

at Burlington in 1921, only 4 survived the first summer. Gleditsia was winter-
killed and Calamovilfa was accidentally destroyed. Andropogon nutans grew
well during both 1922 and 1923 and reached a height of 6 to 9 inches, but Bul-
bilis had invaded rather extensively. Stipa viridula died during the severe
winter of 1922-23.

Seedling Transplants.

Seedlings of 20 species were transplanted into the high prairie on May
17 and watered freely from time to time, as weather conditions demanded,
until August 23. Notwithstanding this aid to establishment, Aristida
purpurea and Petalostemon candidus died early in June, Agropyrum
glaucum and Solidago missouriensis in July, and Psoralea tenuiflora flori-
bunda and Stipa viridula following the drought in August. Several other
species were represented by mere remnants, 14 (70 per cent) surviving the
summer. As sometimes happens under cultivation, Onagra biennis grew a
flower-stalk the first season, but failed to blossom and died in the fall.

On May 19, seedlings of 19 species were transplanted into the mixed prairie
at Phillipsburg. As at the other stations, they were about 15 days old. In
planting, water was placed in the bottom of the trench and after it had settled
down, moist soil was added around each root-mass until the trench was two-
thirds filled, w T hen a second watering was given. The trench was then filled
with drier soil and finally covered with a dry mulch. At all stations the
plants were watered for about a week thereafter. Agropyrum, Bouteloua
racemosa , and Stipa viridula died during the June drought, and Psoralea by
the first of August, but all the others made a good growth, the grasses tillering
freely and Bouteloua gracilis and B. hirsuta blossoming. The mortality was
only 22 per cent.

Seedlings at Burlington did very poorly. Of the 19 species planted here on
May 20, only 5 survived, 74 per cent succumbing to the unfavorable condi¬
tions of the short-grass plains; 1 died early in June, 5 more by July 2, 7 others
before the first week in August, and another later in the same month. On
August 26, wilted or half-dead clumps of Bouteloua racemosa, Stipa viridula,
and Sporobolus asper, and better lots of Bouteloua hirsuta and B. gracilis,
2 to 3 inches tall, alone remained. The history was one of repeated wilting
and revival only to wilt again, the conditions causing transpiration at this
station often being so severe that vegetation wilted even in soil of good
water-content. Under such conditions growth was poor. Only three spe¬
cies survived at all the stations, viz, Bouteloua gracilis, B. hirsuta, and Sporo¬
bolus asper, even under this favorable method of transplanting.

As to the 1921 seedlings, 6 of the 13 species transplanted into high prairie
at Lincoln died the first summer, while 2 more were winterkilled (table 61).
The 5 remaining species grew throughout 1922, all surviving the drought.
However, Andropogon nutans and Stipa spartea died the following winter,
leaving only the three species of Bouteloua. All these made a good growth
in 1923, forming five clumps 5 to 8 inches tall.

At Phillipsburg, 9 of the 12 species survived the first season and 1 died the
following winter (table 62). At the end of 1922 Liatris scariosa had died
and Agropyrum and Andropogon nutans were doing very poorly. The rest
made a good growth, Bouteloua hirsuta alone producing flower-stalks. Andro-

EXPERIMENTS DURING 1922.

yog on nutans and Agropyrum died the following winter. The two shorter
gramas both made a good growth in 1923 and produced seed. Andropogon
furcatus was densely shaded and did only fairly well; Stipa spartea was repre¬
sented by a few narrow, densely shaded leaves only, while Liatris punctata
had 5 fine plants about a foot high that blossomed profusely. None of the
1921 seedlings at Burlington survived the first season.

Summary.

The average germination under all methods of planting was similar to
that of the two preceding years, Lincoln being highest (74 per cent), the mixed-
prairie station second (63 per cent), and the short-grass plains last (44 per
cent). However, as regards surface sowing, Phillipsburg ranks slightly ahead
of Lincoln, and this in spite of frequent watering at the latter station. The
explanation for this seems to lie in soil structure, that at Phillipsburg being
much mellower, and hence it does not crust when alternately wet and dry.

Table 33. —Summary of planting experiments, 1922.

Method of
planting.

Per cent of germination.

Per cent of establishment
of germinated species.

Lincoln.

Phillips¬

burg.

Burling¬

ton.

Lincoln.

Phillips¬

burg.

Burling¬

ton.

Surface sowing.

Trench.

Denuded quadrats.. .

Average.

Seedlings.

This probably accounts for the 8 per cent margin in favor of the seedling trans¬
plants, since in all cases the percentage of germination falls off to the west¬
ward.

Sixty-two per cent of the species that germinated at Lincoln became estab¬
lished, 39 per cent at Phillipsburg, and only 26 per cent at Burlington. The
same sequence held for all methods of planting, except that establishment
in the quadrats at Burlington exceeded by 14 per cent that at Phillipsburg.
Both the lowest percentage of germination and establishment occurred in
the trench, the denuded quadrats ranking highest in percentage of estab¬
lishment.

Sod Transplants.

Large blocks of soil containing 26 species of grasses and forbs were
transplanted into high prairie at Lincoln March 22 to April 5. All the
transplants made a good growth except Psoralea tenuiflora floribunda, which
died before the middle of June, probably because of the cutting of its
deep tap-root, Silphium integrifolium, which succumbed by the end of August,
and Andropogon scoparius which grew poorly. In addition to 10 of the
grasses, the following forbs blossomed and set seed: Anemone cylindrical
Brauneria pallida, Grindelia squarrosa, Liatris scariosa, Solidago missouriensis,
S. rigida, and Vernonia fasciculata; only 7 per cent of the entire lot died.

EDAPHIC STATIONS.

Eighteen blocks of sod, representing 13 species of grasses, were transplanted
from Lincoln into the short-grass sod at Burlington on April 15. Although
none died, the growth was with few exceptions rather poor, being
characterized by scattered shoots, frequently with rolled leaves and dead
leaf-tips. Flowers were present in 61 per cent as compared with 67 per cent
of the grasses transplanted at Lincoln. In all cases, however, the flower-
stalks were 3 to 10 inches shorter and usually few in number, while the
inflorescence itself was dwarfed. For example, Agropyrum headed at Bur¬
lington at 11 inches, while at Lincoln the flower-stalks were 21 to 27 inches
tall; Elymus headed at 12 to 18 inches in the Great Plains, but in the true
prairie at 28 inches. The vegetative growth was proportionately dwarfed.

EXPERIMENTS AT OTHER STATIONS, 1922.

Physical Factors.

Water Relations.

Studies were continued during 1922 at the series of edaphic stations at
Lincoln, and at Nebraska City and Colorado Springs. The general conditions
of precipitation at Lincoln have already been given. The season at Nebraska
City was one marked by June and August drought. April precipitation was
nearly normal, May showed a deficiency of 1 inch, June of 2.6 inches, and,
although July had an excess of 3 inches, August had a rainfall of only 0.7
inch instead of the normal 3 inches (fig. 18). Drought periods occurred on
April 11 to 30, May 6 to 20, May 25 to June 10, and June 11 to 25. The
water-content in excess of the hygroscopic coefficient at Nebraska City and
Lincoln low prairie is given in table 34, where that of the high prairie is added
for purposes of comparison. A study of the table shows that notwithstanding
the decreased rainfall, the soil at Nebraska City usually had a margin of at
least 8 per cent chresard below the first foot of soil. Exceptions to this
occurred after the first week in June and again in August. However, the
chresard in the surface foot, which most critically affects seedlings, was
practically exhausted on June 17, and was very low on August 17, as well as
at certain other periods. As a whole, the holard was less favorable than
during 1921.

Water-content on the low prairie was more favorable than that at Nebraska
City and considerably in excess of that on the high prairie. At no time was
there a margin of less than 7 per cent available at any level, and it was usually
10 to 12 per cent. Conditions on the gravel-knoll were much less favorable
than during the preceding year, as was evidenced by the drying of the grama
grasses during a drought when no water was available in the first 2 feet
of soil.

The average daily evaporation at Nebraska City was considerably lower
than on the high prairie at Lincoln. During the 4 months (last half of May,
June, July, and August) the average daily rates were as follows: 12.4 and 15.1,
18.4 and 29, 12.7 and 13.5, 14.4 and 20.2 c. c. respectively. Isolated humidity
readings taken from time to time and compared with the hygrograph record
at Lincoln for the same hour showed conditions not greatly different from
from those of the preceding year.

100

EXPERIMENTS DURING 1922

Table 34. —Holard in excess of hygroscopic coefficient in 1922.

High Prairie.

Date.

0 to 0.5
foot.

0.5 to 1
foot.

1 to 2
feet.

2 to 3
feet.

3 to 4
feet.

Apr. 30.

15.3

15.8

14.7

10.9

9.1

May 10. .

10.1

13.7

15.3

May 17.

20.0

13.1

14.1

12.1

7.8

June 7.

6.4

11.2

10.9

June 14.

6.8

11.3

10.5

8.4

6.6

June 22.

3.5

6.3

11.1

July 6.

5.7

4.7

8.9

8.0

5.4

July 13.

17.8

11.7

8.1

July 20.

16.4

10.7

6.2

July 27.

21.2

14.7

7.5

Aug. 3.

8.2

10.1

7.4

8.0

6.0

Aug. 10.

5.6

6.2

7.4

Aug. 17.

0.9

1.8

0.8

Aug. 24.

3.2

5.4

5.1

7.1

5.3

Hygroscopic coeff....

9.8

10.9

10.1

10.0

10.3

Low Prairie.

Apr. 30.

18.4

10.4

20.6

12.5

18.2

May 17.

13.4

16.1

17.0

17.6

May 31.

26.7

21.3

18.7

June 7.

13.2

14.6

15.2

June 14.

10.7

13.8

15.1

13.8

19.0

June 22.

6.9

9.4

15.4

June 29.

21.9

10.2

11.0

July 6.

13.4

8.0

6.8

12.8

16.6

July 13.

28.1

16.8

11.5

July 20.

26.3

21.1

11.2

July 27.

30.6

28.7

15.3

Aug. 3.

17.5

14.6

11.8

11.1

15.9

Aug. 10.

10.2

15.8

13.7

Aug. 17.

7.9

9.8

11.0

Aug. 24.

7.8

6.8

11.1

11.8

15.9

Hygroscopic coeff....

10.0

9.6

9.2

11.1

10.8

Nebraska City.

Apr. 22.

17.9

19.3

18.3

15.0

13.3

May 13.

8.7

16.2

17.4

14.4

13.2

May 20.

4.4

11.0

16.1

June 3.

18.4

18.8

17.0

15.0

14.3

June 10.

4.3

9.4

12.3

June 17.

-0.3

1.5

7.9

11.2

June 21.

2.0

4.2

10.4

10.3

10.2

June 28.

4.4

0.8

5.0

12.2

11.8

July 5.

8.5

0.9

5.9

July 12.

25.6

21.6

17.7

10.2

10.3

July 19.

7.6

13.4

9.9

July 26.

5.5

10.3

9.1

Aug. 1.

18.2

16.3

13.1

9.9

9.5

Aug. 8.

11.9

12.7

10.4

Aug. 15.

1.5

6.1

7.8

Aug. 22.

11.2

7.1

4.1

Aug. 31.

3.7

4.8

6.8

8.0

3.4

Hygroscopic coeff....

12.1

11.7

12.3

13.6

12.9

EDAPHIC STATIONS! SURFACE SOWING.

101

Planting Results.

Surface Sowing.

Of typical grasses and forbs, 19 species were sown on the surface of the
low prairie on April 20, 1922; unlike the high prairie and gravel-knoll, this
area had not been burned. A dense tangle of grass-leaves and flower-stalks
covered the ground and formed a loose mulch to a depth varying from 2 to 4
inches. This not only greatly affected the light relations, but also the air was
of much higher humidity and lower temperature below the mulch, and the
soil was cooler. For example, at 2 p. m. on a clear day (May 28) an average
of four thermometer readings at a depth of an inch under the mulch gave a
temperature of 68° F., while that at a similar depth in denuded quadrats was
95° F.

Under these conditions the plants germinated slowly, Desmodium canescens,
Liatris punctata and scariosa, and Petalostemon candidus alone appearing above
ground by May 17. Moreover, by May 26 the new growth of the dominants
was 10 to 15 inches tall, so that light relations played an exceedingly important
role from the first. However, only 3 species, Aster multiflorus, Ratibida colum-
naris, and Solidago missouriensis, failed to germinate. Weak plants of
Amorpha canescens, Onagra biennis, and Sporobolus asper lasted only until the
middle of June, and Kuhnia glutinosa, Liatris scariosa and punctata, and
Lespedeza capitata died within the next 30 days, all showing the effects of
shading. Bouteloua hirsuta succumbed by August and B. gracilis, B. racemosa,
Desmodium canescens, and Elymus canadensis during this month. This left
remnants of Aristida purpurea and Andropogon nutans, as well as rather good
growths of Andropogon furcatus and Petalostemon candidus . The total mor¬
tality was 75 per cent during the first summer. The severe competition for
light on the low prairie was chiefly responsible for this; by June 7 the surface
and trench plantings were already badly shaded, while by June 22 the general
grass level was 18 inches; by September 1, that of the foliage reached over 2
feet and the flower-stalks 43 inches. In the part of the area unmown since
1919, growth was even greater.

On the surface at Nebraska City, 20 species were sown on April 22. Spring
opened late and the soil was in excellent condition as regards holard. Nine
species were above ground by May 13. Aster multiflorus, Bouteloua gracilis,
Lespedeza capitata, Onagra biennis, Solidago missouriensis, and Sporobolus
cryptandrus failed to germinate. By June 21, most of the plants were badly
wilted and Liatris scariosa and Pinus ponderosa had succumbed. Robinia
pseudacacia died in July, Agropyrum glaucum and Petalostemon candidus by
August, while Amorpha canescens and Ratibida columnaris were found dead
on August 31. Only 50 per cent survived. These included slender, delicate
specimens of Andropogon nutans, A. furcatus, Bouteloua hirsuta , B. racemosa,
Elymus canadensis, Kuhnia glutinosa, and Liatris punctata. A similar lot of
seeds was sown on the surface at Colorado Springs on May 5, but these, like
those planted in the trench and denuded quadrats, showed no germination
when examined on May 25 and again on July 4, owing to unfavorable holard.

As regards the 1921 surface sowings, it may be recalled that none survived,
owing to the dense shade. At Nebraska City, Bouteloua gracilis and B.
hirsuta alone were represented by delicate seedlings at the end of the first

102

EXPERIMENTS DURING 1922.

season. However, they held out during the summer of 1922, although very
much attenuated. During 1922 they were still very slender and delicate, but
they survived this third season of growth, although densely shaded, indicating
an unexpectedly high tolerance of shade for the short-grasses.

Trench Sowing.

Seeds of 20 species were planted in a trench on the low prairie on April 20.
Amorpha canescens, Aster multiflorus, Onagra biennis, Ratibida columnaris,
Sporobolus asper, and Solidago missouriensis did not germinate. Nearly all
of the others grew in considerable abundance, except Bouteloua hirsuta and
Liatris punctata, which were represented by a few plants each, all dying before
the end of June. The others did quite well during the rainy month of July,
Liatris scariosa, Lespedeza capitata, and Muhlenbergia pungens disappearing
by August. Nine species (64 per cent of those that germinated) became
permanently established, and although slender were in fairly good condition
at the end of the summer.

A similar lot of seeds, representing 21 species, was planted in a trench in
subclimax prairie on April 22. Eight species, including 3 which grew at
Lincoln, failed of germination. Agropyrum glaucum, Muhlenbergia pungens,
Petalostemon candidus, and Pinus ponderosa died before June 20, the last
having been dug up by rodents. Andropogon nutans, A. furcatus, Liatris
scariosa and punctata died before the middle of July, and Elymus canadensis
in August. This left a few weak plants each of Bouteloua racemosa, Kuhnia
glutinosa, and Robinia pseudacacia, but a good sod of Andropogon nutans 7 to
9 inches high.

Of the 1921 trench plantings on the low prairie, remnants of Aristida
purpurea alone survived the first season, and these were accidentally destroyed
the next spring in transplanting. At Nebraska City, Andropogon nutans
alone was alive the following spring. In 1922 it reached a height of 13 inches,
and did quite as well the next season, but it failed to blossom. On the gravel-
. knoll, Andropogon nutans alone survived, but although it grew well early the
next spring, it died the following June.

Denuded Quadrats.

On April 20, 27 species were sown in denuded quadrats on the low prairie.
Five did not germinate. These belonged to different species from those
failing of germination in the trench, and included Brauneria pallida, Corylus
americana, Fraxinus lanceolata, Gleditsia triacanthus, and Ratibida columnaris.
During the June drought, Agropyrum, Aristida, Aster, Lespedeza, Liatris
punctata, and Solidago succumbed; Elymus, Muhlenbergia, and Ulmus also
died before August 1. This gave a mortality of 40 per cent, though most of
the survivors were in good condition at the end of the summer.

Among 32 species planted at Nebraska City, 10 did not germinate. Muhl¬
enbergia and Petalostemon died before June and the following as a result of
the June drought: Acer saccharinum, Aristida, Liatris scariosa and punctata,
Ratibida, and Ulmus. Later, Agropyrum and Lespedeza were lost, giving a
total mortality of 46 per cent. Among the survivors were 4 species of trees
(two of which, Acer negundo and Gleditsia triacanthus, were in fine condition),
2 forbs, and 6 grasses. In general, they were in good condition. By the

EDAPHIC STATIONS! SEEDLING TRANSPLANTS.

103

end of August the average height-level of the grassy vegetation was 16
inches, shade being dense over most of the quadrats. However, very little
invasion had occurred.

As to the denuded quadrats planted in 1921, 10 of the 12 species that grew
on the low prairie survived the first summer. Acer negundo, A. saccharinum,
and Lespedeza capitata failed to grow the following spring, and Elymus died
in June, as did also Bouteloua gracilis, perhaps as a result of the surface
soil washing away from the roots. All of the others except Corylus ameri-
cana made a good growth, the latter being badly eaten by grasshoppers. All
were densely shaded and none flowered. Corylus did not survive the dry
fall and winter following, while Aristida died during the summer of
1923, evidently being unable to longer endure the dense shade. At the end of
the summer Sporobolus was represented by remnants only. Symphoricarpus
made only a fair growth, while Andropogon nutans alone developed normally,
reached a height of 25 to 30 inches and merged into the native sod.

At Nebraska City only 8 of the 15 species that grew survived the first
summer. Gleditsia and Robinia failed to appear in the spring of 1922 and
Calamovilfa died early in June. All of the other species made a good growth,
Andropogon merging into the native sod. None of the quadrats were much
invaded. No flower-stalks were produced by any of the plants. In 1923 the
6 remaining species all did quite well, several of them seeding rather abun¬
dantly.

On the gravel-knoll, 40 per cent of the species that grew survived the first
season. During 1922 some of the Lespedeza seeds, sown the preceding
spring, germinated, but these and the few feeble shoots from Symphori¬
carpus rhizomes were all dead by midsummer. Aristida and Elymus both
seeded, but, like Andropogon, suffered severely from the August drought.
During the following dry fall and winter Elymus died, but in 1923 both Aris¬
tida and Andropogon made a good growth and seeded.

Seedling Transplants.

Seedlings were transplanted both on the low prairie and at Nebraska City,
22 pots containing 18 species being used in growing the plants for each station.
The transplanting was done at Lincoln on May 17. One lot of Agropyrum
died in June and another in July, together with Elymus, Koeleria, Liatris, and
Stipa spartea; Ratibida, Kuhnia, Argemone, Onagra, and Psoralea succumbed
in August, making a total loss of 55 per cent. Most of the survivors came
through the season in good condition, owing to the adequate water-content
even during the August drought.

At Nebraska City, Agropyrum, Koeleria, Lespedeza, Solidago, and Stipa viri-
dula died before the end of June. Three others failed in July, viz, Aristida,
Glycyrrhiza, and Liatris, and August added Psoralea to the list, making a
total of 41 per cent. Most of the survivors grew well, Bouteloua hirsuia
putting forth flower-stalks.

To complete the history of the 1921 seedling transplants, only 14 per cent
survived on the gravel-knoll, constituted by Bouteloua hirsuta and Andro¬
pogon nutans. During 1922 these made only a poor vegetative growth, fre¬
quently wilting, and did not put forth flower-stalks. By August 1923 they
had reached a height of 8 inches, neither coming into blossom.

104

EXPERIMENTS DURING 1922.

On the low prairie a single species died the first season, but by the next
spring Aristida purpurea and Liatris scariosa were dead and Bouteloua race-
mosa was represented by a single plant which died in June. The rest
made a fair to excellent growth, except Stipa comata. The shade was
dense and Elymus canadensis alone grew above the general level and
headed at 35 inches. No others blossomed.

During the summer of 1923 the following species were so deeply shaded
that they died: Bouteloua hirsuta, B. gracilis, Liatris punctata, Stipa viridula,
and S. comata. Elymus headed at about 3 feet and Stipa spartea seeded at
2 feet, but neither Andropogon nutans, which did poorly, nor A. furcatus
showed signs of flower-stalks by the last of August.

Table 35. —Growth of species in cultivated soil.

Species.

Degree of development during first season.

Agropyrum glaucum.

Good growth of foliage; maximum 2 feet; no flower-stalks.

Andropogon furcatus.

Fine bunches; foliage 18 in.; a few flower-stalks 36 to 38 in.

halli.

Foliage 16 in.; flower-stalks 30 in. (cf. plate 12 c).

nutans.

Excellent growth; 20 in., seeding abundantly at 3 to 4.5 feet.

Aristida purpurea.

Large bunches; foliage 11 in.; flower-stalks 18 in.

Bouteloua gracilis.

Dense growth; leaves 14 in.; flowering profusely at 15 to 22 in.

hirsuta.

Leaves 10 in.; flower-stalks appearing Aug. 8; later 12 in. tall;
abundant.

racemosa.

Good growth; ave. height 13 in.; flower-stalks 3 to 3.5 feet.

Calamovilfa longifolia. . ..

Foliage 16 in.; no flower-stalks.

Desmodium canescens....

Flower-stalk appearing July 10; fruited abundantly at 5 feet
(cf. plate 14 b).

Elymus canadensis.

Foliage 2 feet; flower-stalks 4 to 5 feet; very large heads.

Kuhnia glutinosa.

34 in. tall; fruiting abundantly.

Lespedeza capitata.

Fine plants, 20 to 24 in.; seeded abundantly.

Liatris scariosa.

3 to 6 in.; many leaves per plant.

Muhlenbergia pungens....

9 in.; flowered and seeded profusely at 14 to 16 in.

Onagra biennis.

One rosette 28 in. in diameter; many smaller ones; many had
flower-stalks 5 to 5.5 feet tall; blossomed.

Panicum virgatum.

Average height foliage 29 in.; seeded abundantly at 3 to 3.5 feet.

Petalostemon candidus. . .

Excellent growth; very much bunched stems, 8 to 12 in. tall.

Pinus ponderosa.

Five plants, 4 to 5 in. tall.

Robinia pseudacacia.

Stems nearly an inch in diameter and 7.5 feet tall.

Sporobolus asper.

9 in. tall; flower-stalks 14 to 16 in.; seeded profusely.

Stipa viridula.

Foliage about 16 in. tail; maximum 26 in.; no flower-stalks.

At Nebraska City, 6 of the 13 species of seedlings died the first summer;
Regardless of this heavy mortality, all the species survived the summer of
1922, notwithstanding the drought. However, all suffered more or less
severely in late summer, Stipa comata being represented by remnants only at
the end of the season. Bouteloua hirsuta and B. gracilis alone blossomed.
The seedling area, although quite shaded by the adjacent vegetation, was not
much invaded.

In 1923, Stipa comata became badly invaded and died. All the others made
a good growth.

Effect of Competition.

In order to determine the effects of competition upon the development of
the species investigated, many of them were grown in a cultivated area at
Lincoln, kept free from weeds by hoeing, and watered from time to time
(table 35).

EDAPHIC STATIONS! SOD TRANSPLANTS.

105

The fine growth made in a single season by these species is in striking con¬
trast to that under even the most favorable method of planting in the grass¬
land and emphasizes the striking effect that competition plays in limiting
plant development under natural conditions. Plants growing for a second
year under cultivation made no less remarkable growth. Bouteloua gracilis
had an average height of foliage of 19 inches, spread 15 inches on either side
of the trench, and some of the flower-stalks were 32 inches tall. Agropyrum
glaucum was 29 inches tall, with a maximum height of 44 inches, and had
spread 6 feet by rhizome propagation. A single plant of Onagra biennis had
3 stems which reached a height of 7 feet and 12 others of lesser height, all
from the same rosette (plate 14).

Summary.

A summary of planting experiments is given in table 36, where for purposes
of comparison the high prairie is included. The germination of surface-
sown seed was not greatly different at the three stations (70 to 84 per cent),

Table 36. —Summary of planting experiments , 1922.

Method of
seeding.

Per cent of germination.

Per cent of establishment of
germinated species.

High

prairie.

Low

prairie.

Nebraska

City.

High

prairie.

Low

prairie.

Nebraska

City.

Surface sowing.

Trench.

Denuded quadrat....

Average.

Seedlings.

r 45

that on the low prairie being highest. Germination in denuded quadrats was
also about 10 per cent greater on the low prairie than elsewhere, Nebraska
City ranking third. Averaging the results of the three methods, 75 per cent
germination occurred on low prairie, 1 per cent less on high prairie, and 8 per
cent less at Nebraska City.

Because of the effect of the luxuriant vegetation of the low prairie in reduc¬
ing the light intensity, a factor which quite outweighed the higher water-
content, establishment averaged lower here (50 per cent) than on high prairie
(62 per cent) and was least at Nebraska City (46 per cent). The last fact can be
directly correlated with drought, especially when it is recalled that the surface
sowing at Lincoln was watered. Similar conditions hold for the seedling trans¬
plants. A survey of the data shows clearly that while surface sowing gave the
greatest per cent of germination, the best establishment occurred in the
denuded quadrats. Compared with the western stations, both the per cent
of germination and establishment averaged higher.

Sod Transplants.

Twenty-eight blocks of soil, comprising 25 species, were transplanted from
the several habitats about Lincoln into the low prairie on March 22. These
included 11 subdominant forbs, as well as all of the most important grasses.

106

EXPERIMENTS DURING 1922.

Agropyrum glaucum grew poorly, not attaining a height greater than 11
inches and failing to flower. Astragalus crassicarpus, Brauneria pallida,
and Psoralea tenuiflora floribunda, all provided with strong tap-roots neces¬
sarily cut off in transplanting, languished and died within a few weeks. All
of the rest, however, came through the season in good condition, 75 per
cent of the dicotyls and 50 per cent of the grasses flowering.

A lot of sods, including 11 species from Arizona, were transplanted into
low prairie early in May. The blocks were very small, being only 3 or 4 inches
in diameter, and it seems probable some were dead when transplanted. In
spite of sufficient watering the following did not grow: Aristida divaricata, A.
purpurea, Bouteloua eriopoda, B. hirsuta, B. racemosa, Hilaria mutica, H. cen¬
chroides, Sporobolus wrighti, and Valota saccharata. Andropogon saccha-
roides put forth 3 green shoots, which reached a height of 11 inches before
they died in August. Bouteloua bromoides and Hilaria cenchroides formed
good clumps 5 to 8 inches high, the former bearing flower-stalks 8 to 13
inches tall. However, both were winterkilled, notwithstanding the fact that
they were well covered with a mulch of dead grasses. The following species
from Arizona were also transplanted into high prairie at the same time:
Aristida divaricata, A. purpurea, Bouteloua eriopoda , B. bromoides, B. hirsuta
(all of which died promptly), and Hilaria cenchroides . The last made a
sparse growth, reaching a height of 8 to 10 inches in August, but died under
a mulch during the following winter.

Sod of 14 species of grasses from both high and low prairie were trans¬
planted into the salt-flat near the low-prairie station late in March. Andro¬
pogon nutans, A, furcatus, Panicum virgalum, and Sporobolus asper did very
poorly, making only a sparse growth, yellowing and wilting from time to time.
By June 22 not only this lot of transplants, but also those of previous years
were in very poor condition. Large cracks occurred around most of the blocks
of sods and some were almost dried out. Shading played rather an unim¬
portant role, since the height of the Distichlis was only 18 inches. Sporo¬
bolus died late in July, and the others made a fair to poor growth only.
Nearly all were more or less dwarfed and of a light-green color, and the
inflorescences were often reduced. In fact, only 5 species (36 per cent)
seeded; these were Bidbilis dadyloides, Bouteloua gracilis, Elymus canaden¬
sis, Koeleria cristata, and Stipa spartea.

On April 5, 14 species were transplanted into the former swamp area.
Silting up since the building of a dam, as well as the enormous losses by
transpiration from the rank growth of vegetation, had decreased the water-
content. On June 7 the holard at any depth did not exceed 22 per cent, which
was the amount in the surface 6 inches; by June 22 it had fallen to 15 per
cent or less in the surface 2 feet. It seems evident that there was no excess
moisture or deficient aeration and that light was the controlling factor. By
May 6, Spartina was 8 to 10 inches tall, and it had reached a level of 32 inches
by the middle of June. In August a rank growth of Spartina, Phalaris, etc.,
4 feet tall, covered the area; light readings at this time gave values of 1.5 to
11 per cent. Andropogon furcatus, A. scoparius, and Panicum virgatum, al¬
though starting out well, were badly invaded by the swamp dominants and
died by the middle of July, as did Andropogon nutans later in the summer.
All suffered severely from shading, as was evidenced by the attenuated leaves

EDAPHIC STATIONS: SOD TRANSPLANTS.

107

and slender flower-stalks of the 5 species that blossomed, viz, Agropyrum,
Bulbilis, Elymus, Koeleria, and Stipa. The species of the following genera
came through the season in fair to good condition: Agropyrum, Bulbilis,
Elymus, Koeleria, Spartina, and Stipa. Bouteloua gracilis, B. hirsuta, B.
racemosa, and Sporobolus were in poor condition by autumn.

Duplicate blocks of sods of the following species were shipped from Lincoln
and transplanted at Colorado Springs on March 28: Andropogon scoparius,
A. furcatus, A. nutans, Elymus canadensis, Sporobolus asper, Stipa spartea,
Koeleria cristata, Bouteloua racemosa, B. gracilis, B. hirsuta, Bulbilis dacty-
loides, Panicum virgatum, Spartina cynosuroides, and Agropyrum glaucum.
All were growing vigorously on May 21, but were considerably eaten by
grasshoppers and doing rather poorly by July 4. A final check on August 24
showed that Bulbilis dactyloides alone had blossomed, and no flower-stalks
were appearing on any of the other species. Sporobolus asper had died, and
Elymus, Andropogon furcatus, and B. racemosa were eaten back to mere rem¬
nants. Triplicate sods of Stipa viridula, Muhlenbergia gracilis, M. gracillima,
and Sporobolus cryptandrus were shipped from Colorado Springs and trans¬
planted in the cultivated plats at Lincoln on June 22. The sods were very
small and the plants died in spite of ample watering.

Large duplicate blocks of sods were shipped from Lincoln to Berkelej^,
California, which has an average rainfall of 24 inches, occurring chiefly in
the winter months, and a long dry season in summer and autumn. These
were transplanted on December 4 (1921), the species employed being Andro¬
pogon scoparius, A. furcatus, A. nutans, Elymus canadensis, Stipa spartea,
Koeleria cristata, Bouteloua racemosa, and Bulbilis dactyloides. The sods were
set on a dry west exposure in a heavy clay soil covered chiefly with introduced
weeds, such as Arena fatua, Brassica campestris, and Raphanus saiivus.
Owing to an abundance of rain, conditions at the time of transplanting were
very favorable for growth. The original vegetation was bunch-grass prairie
composed mostly of Stipa setigera, as indicated by the relicts of this. Some
attention was given the sods in 1921 and 1922, at first a little water, and then
occasional removal of the competing weeds. Most of the grasses did not
survive the first dry season, but Elymus canadensis made a good growth.
Bulbilis dactyloides persisted for a time, and Poa pratensis spread at first,
but succumbed when water was withheld. Even Elymus was unable to sur¬
vive the exceptionally dry spring and summer of 1923, and the plot was
finally destroyed by the great fire of September 1923. The failure of the
grasses to establish themselves was due chiefly to inability to pass the long
dry seasons and to competition as a contributing factor, but the experiment
was on too small a scale to justify the conclusion that none of the species
could be established.

A number of grasses from the Santa Rita Reserve near Tucson, Arizona,
were planted under the same conditions at Berkeley. The sods were set
April 3, 1921, and comprised Andropogon saccharoides, A. coniortus, Aristida
divaricata, A. purpurea, Bouteloua bromoides, B. hirsuta, B. racemosa, B. roth-
rocki, B. eriopoda, Hilaria cenchroides, H. mutica, and Sporobolus wrighti.
All of these made some growth in 1921, but by the autumn of 1923 all had
succumbed except the following: Andropogon saccharoides grew vigorously
in 1922, reaching a height of 4 feet, flowered, and set seed abundantly; in 1923

108

EXPERIMENTS DURING 1922.

the 3 plants were growing and flowering, with a slight tendency to increase
but not spreading over the ground, and no seedlings were noticed; Hilaria
mutica made an attempt to establish itself and 1 of the 3 tufts was still living;
Aristida divaricata was in the same condition as the Hilaria; Bouteloua bro-
moides made a good start and flowered in 1922, but disappeared during the
long dry summer of 1923; Sporobolus wrighti grew each season, and although
it did not spread, was holding its own, even against the weedy grasses with
which it had to compete during the spring and early summer. It is instruc¬
tive to note that the grasses from the desert plains grew better than those
from the true prairies, a fact readily explained by the greater similarity in
climate and climax between Arizona and California.

5. EXPERIMENTS DURING 1923.

DEVELOPMENT OF SEEDLINGS AND TRANSPLANTS.

It remains only to trace the history of the 1922 plantings through the period
from September 1922 to the end of the summer of 1923.

Physical Factors.

Water Relations.

The fall and winter of 1922 constituted a period of great drought. Some
idea of its intensity may be gained by an examination of the rainfall data in
table 37.

Table 37 .—Precipitation from August 1922 to August 1923.

Burlington.

Phillipsburg.

Lincoln.

Total.

Departure
from mean.

Total.

Departure
from mean.

Total.

Departure
from mean.

August.

3.04

+ 0.50

1.13

-1.77

0.69

-3.02

September....

0.16

-1.23

0.94

-1.58

2.12

-0.52

October.

0.18

-0.75

0.44

-1.04

1.42

-0.40

November....

1.44

+ 1.01

0.40

-0.28

2.62

+ 1.77

December....

0.00

-0.63

0.00

-0.82

0.06

-0.61

January.

0.00

-0.26

0.00

-0.44

0.14

-0.48

February.

0.10

-0.36

0.16

-0.65

0.66

-0.04

March.

0.36

-0.47

1.31

+0.52

1.77

+0.44

April.

1.66

-0.40

3.40

+ 1.07

2.21

-0.56

May.

No report.

....

7.86

+4.85

3.31

-0.94

June.

3.03

+0.26

4.94

+ 1.03

5.25

+0.93

Julv.

4.33

+ 1.56

2.34

-0.62

1.88

-1.95

August.

4.89

+2.30

1.85

-0.95

.5.50

+ 1.79

Although Burlington had an excess of 0.5 inch in August, drought at the
other stations began during this month. Precipitation was below normal
at Burlington (except during November) until May, and the same was true
at Lincoln until June, except a slight excess in March, while at Phillipsburg
it was exceedingly low until March. However, it should be noted that the
spring and summer rainfall was high at the western stations, although below
normal at Lincoln during April and May and again in July. The season was
one which promoted excellent growth. Not only was the rainfall generally
above normal, but the rains were also well distributed, drought periods west¬
ward being fewer and shorter than usual. This was indicated by the excellent
growth of Bouteloua gracilis , which reached a height of 12 to 18 inches at
Burlington and also by the fact that the short-grasses remained green or dried
only slightly throughout the entire summer. Conditions were slightly less
favorable in August at Lincoln and at Phillipsburg, where the three grama
grasses ripened seed. On the whole, vegetation grew well at Lincoln, the
grasses on the gravel-knoll remaining green all summer. An examination of
the chresard taken at intervals in the various stations (table 38) is illuminating
in this connection.

The very low water-content on the high prairie in October is remarkable;
on the gravel-knoll practically no water was available to 4 feet, and it was not

109

110

EXPERIMENTS DURING 1923.

high even on low prairie. By March 1 a good water-content was found at
both prairie stations and available moisture to the amount of 3 to 7 per cent
was present throughout the season. The adequate holard in the surface layers
at Burlington during July and August was very unusual. At Phillipsburg, as
already indicated, it was drier in August. Because of the drought of autumn
and late winter, many species under all methods of planting and transplanting
were killed (table 39).

Table 3S .—Holard in excess of hygroscopic coefficient at the several stations, 1923 (unless

otherwise indicated ).

Station.

Date.

Depth in feet.

0 to 0.5

0.5 to 1

1 to 2

2 to 3

3 to 4

Lincoln:

High prairie.

Oct. 27,1922

1.8

3.7

3.0

2.8

3.8

Do.

Mar. 1

18.1

18.3

12.1

10.7

7.3

Do.

June 2

10.4

16.6

16.7

11.5

8.-2

Do.

July 23

3.0

4.3

5.4

7.1

7.5

Do.

Aug. 3

4.2

5.4

5.0

6.4

8.2

Do.

Aug. 23

12.5

16.4

6.3

7.1

8.7

Gravel-knoll.

Oct. 27

-1.5

1.0

-0.8

1.5

3.1

Do.

June 2

-0.6

1.0

3.1

9.8

7.6

Burlington.

July 6

0.4

-0.7

0.2

0.8

-0.6

Do.

July 19

10.0

0.0

-0.3

-0.2

Do.

Aug. 21

5.8

4.5

-0.2

-0.7

-0.6

Phillipsburg.

June 20

11.3

13.3

11.1

12.6

11.9

Do.

Aug. 22

-0.6

1.5

0.7

1.6

5.6

Nebraska.

May 27

20.9

20.0

17.7

13.9

12.7

Do.

Aug. 28

22.7

20.0

9.5

6.7

7.7

Lincoln:

Low prairie.

Oct. 27

3.5

5.8

6.2

8.3

13.2

Do.

Mar. 1

22.4

21.1

11.8

11.3

16.4

Do.

Aug. 23

11.4

18.4

13.4

8.0

14.0

Survival Results.

An examination of the data in table 39 reveals a number of interesting facts,
the two most important of which are as follows: Losses at the three major
climatic stations were greatest at Burlington, intermediate at Phillipsburg,
and least at Lincoln. The losses at all stations averaged highest for the
plants of a single year’s establishment, next for those 2 years old, and least
for those established for 3 seasons. With a single exception, survival was
greatest at all 3 stations under the method of denuded quadrats, this even
exceeding that of transplanting blocks of sods. Surface sowing gave the
lowest survival.

Survival on the low prairie and at Nebraska City was very similar, following
the sequence stated above as regards length of establishment. Except for
1920 (when all averages were very high) it exceeded that on the high prairie at
Lincoln, and naturally that at the western stations. Survival on the gravel-
knoll of species grown for at least two summers was remarkably high, though
few species were concerned, with the exception of sod transplants.

SURVIVAL RESULTS.

Ill

Surface Sowing.

Nine of the 18 species that germinated on the surface of the high prairie
at Lincoln died by the end of August. Because of the exceedingly dry
autumn and winter, 6 species were winterkilled, leaving only 33 per cent of
the survivors of the first summer. Petalostemon candidus had only two very
delicate plants 3 to 5 inches tall by autumn, but Andropogon furcatus and A.
nutans had both made a good growth, reaching a height of 10 or 12 inches.

Table 39 .—Comparative survival at the different stations during fall and winter of 1922-23.

Method of planting.

Lincoln,

high

prairie.

Phillips¬

burg.

Burling¬

ton.

Method of planting.

Lincoln,

low

prairie.

Nebraska

City.

Gravel-

knoll.

1920.

Surface sowing....

, , ,

. . .

100

Surface sowing....

...

Trench planting. . .

100

• . •

Trench planting . . .

100

...

Denuded quadrats.

100

Denuded quadrats.

100

...

Seedlings.

Sod transplants... .

...

Sod transplants... .

Average.

100

Average.

1921.

Surface sowing. . . .

...

Surface sowing....

...

100

...

Trench planting. . .

100

. . .

Trench planting. . .

. . .

100

...

Denuded quadrats.

100

Denuded quadrats.

100

Seedlings.

100

Sod transplants....

100

Average.

1922.

Surface sowing....

...

Surface sowing....

Trench planting. . .

Denuded quadrats.

Seedlings.

Sod transplants....

Average.

, 66

...

In the mixed prairie at Phillipsburg, 8 species, 4 of which were in very poor
condition, survived the first season. Andropogon nutans, A. furcatus, Des-
modium canescens, Elymus canadensis, and Kuhnia glutinosa were winter-
killed, leaving only the 3 species of Bouteloua. None of these had seeded
by the following August, but all were in fair condition, having reached
heights of 4 to 8 inches. None of the surface-sown plants survived the
summer at Burlington.

Trench Sowing.

Of the species sown in a trench on the high prairie, 63 per cent lived through
the first season, most of them being badly dried out late in August; 5 of the 10
did not grow the next spring, viz, Andropogon furcatus, Aristida purpurea,
Elymus canadensis, Liatris scariosa, and Rohinia pseudacacia. Andropogon
nutans and Pinus ponderosa had succumbed by August of 1923, leaving only
several fine clumps of Bouteloua racemosa and 4 good plants of Desmodium
canescens, each about 8 inches tall. Plants in the trench at Phillipsburg were
represented in the fall of 1922 by Andropogon nutans and Bouteloua racemosa
only. The last was winterkilled, while Andropogon was represented by very
much attenuated plants 8 inches tall, the shade being very dense again during
1923. At Burlington, Andropogon nutans alone survived the summer, but it
died the following fall or winter.

112

EXPERIMENTS DURING 1923.

Denuded Quadrats.

Of the 25 species that grew in the denuded quadrats on the high prairie,
18 survived the first summer. Of these, Aristida purpurea, Desmodium
canescens, Onagra biennis, and Pinus ponderosa succumbed by spring. Andro-
pogon scoparius and Muhlenbergia pungens, which were represented by a single
clump each, died before September 1923. Acer saccharinum was killed back
to within 6 inches of the soil, 16 of the trees dying; however, like all the other
species, they made a good growth during the summer. Exceptions to this
were Bouteloua gracilis, Sporobolus asper, and Robinia pseudacacia, which did
poorly. None of the species blossomed. The quadrats were only slightly
invaded. It is interesting that not only grasses and forbs, but also certain
trees, survived the second summer.

At Phillipsburg only half of the species that germinated in the denuded
quadrats lived to the end of the first summer. Andropogon furcatus,
Gleditsia triacanthus, and Acer saccharinum were winterkilled. Although all
quadrats were densely shaded by a rank growth of Bouteloua gracilis 18 to 24
inches high, the 8 remaining species all did well. Aristida purpurea failed to
blossom, as did also Andropogon nutans and Panicum virgatum. The rest,
viz, Bouteloua racemosa, B. gracilis, Sporobolus asper, and Petalostemon can-
didus, seeded, the last at a height of 2 to 2.5 feet. At Burlington, 64 per cent
survived in the denuded quadrats. Because of the very dry fall and winter
that ensued, only 2 of the 9 species showed renewed growth the following
spring. Bouteloua gracilis formed excellent clumps and flowered profusely
in 1923. Elymus canadensis, the other survivor, was represented by a
single stalk which headed at 16 inches.

Seedling Transplants.

Of the seedlings transplanted into high prairie in the spring of 1922,
only 70 per cent survived, notwithstanding that they had been regularly
watered throughout the summer. Of the 14 survivors, Elymus, Petalostemon,
and Argemone were winterkilled. Onagra died during the following summer,
but all of the others made a good growth, though none came into blossom.
Of the 19 species of seedlings grown in the mixed prairie, 15 survived,
but 8 of these were winterkilled. Although the rest made a good growth
during the following summer, Bouteloua gracilis and Ratibida columnaris
alone came into bloom. At Burlington only 5 of the 19 species trans¬
planted survived. All but one were winterkilled, viz, Sporobolus, Bouteloua
racemosa, B. hirsuta, and Stipa viridula, Bouteloua gracilis alone surviving.
This did fairly well in 1923, but produced no flower-stalks.

EXPERIMENTS AT EDAPHIC STATIONS.

Survival Results.

Surface Sowing.

On the low prairie, owing to dense shading, only 25 per cent of the 1922
surface-sown plants survived the first summer. Aristida, Andropogon
furcatus, A. nutans, Bouteloua gracilis, B. racemosa, and Petalostemon candidus
did not grow the following spring. Andropogon nutans alone survived, and at
the end of the following summer the slender plants were a foot high.

EDAPHIC STATIONS! SURVIVAL.

113

At Nebraska City, 50 per cent of the species survived. Bouteloua hirsuta,
Elymns canadensis, and Kuhnia glutinosa were winterkilled and Bouteloua
racemosa died by the end of the next summer. Andropogon nutans, A. fur-
catus, and Liatris punctata alone survived. In all cases the plants were very
much attenuated and it seemed doubtful if they would ecize.

Trench Sowing.

As regards the trench plantings in the low prairie, 9 species, i. e., 64 per
cent of those that germinated, grew throughout the first summer. Aristida
and Kuhnia were winterkilled and Andropogon furcatus died the following
summer. All the remaining species were very slender, Elymus and Petalo-
stemon being represented in August by remnants only. Andropogon nutans,
Bouteloua gracilis, B. racemosa, and Desmodium canescens, which made up
the remainder, reached heights of 6 to 8 inches, but none seeded. Of the 21
species planted in the Nebraska City trench, only 4 survived, and of these
Kuhnia glutinosa and Robinia pseudacacia were winterkilled. Andropogon
nutans and Bouteloua racemosa were each represented by clumps of rather
slender plants at the end of the following summer.

Denuded Quadrats.

On the low prairie at Lincoln, 60 per cent of the 1922 plantings in quadrats
survived, Amorpha canescens and Onagra biennis alone succumbing dur¬
ing the following winter. At the end of August 1923, the following were in
very poor condition, being represented by remnants only: Bouteloua gracilis,
B. hirsuta, Sporobolus asper, and Acer saccharinum. The 7 remaining species,
although somewhat slender, had made fair to good growth, but none seeded.
At Nebraska City, 12 species survived the first summer. A fourth of these,
including Andropogon furcatus, Desmodium, and Kuhnia were winterkilled,
and Robinia pseudacacia died later. All of the remaining species made good
growth, Andropogon nutans forming a dense sod.

Seedling Transplants.

Of the species transplanted as seedlings into the low prairie, 45 per cent
survived the first summer, but among these Stipa viridula, Petalostemon
candidus, and Kuhnia were winterkilled. All the rest, including the 3
gramas, Sporobolus, Desmodium, and Andropogon nutans, prospered until the
end of the growing-season.

At the Nebraska City station, 13 species (59 per cent) lived to the end of
the first summer. Petalostemon and Ratibida were winterkilled. All
of the others made a good growth because of the favorable season, with the
exception of Onagra, which died in midsummer.

Summary of Survival.

The fate of the 1922 plantings by the end of 1923 is shown in table 40.

Under all methods of planting or transplanting, the survival at the end of
the second season of growth was greatest at Lincoln, intermediate at Phillips-
burg, and least at Burlington. Of the several methods of planting, survival
in the trench was lowest and surface sowing next, while the denuded quadrats
were only slightly more successful than the seedling transplants. The average
survival on the low prairie at Lincoln was greater than at Nebraska City,
but the latter was nearly equal to that on high prairie at Lincoln. On low
prairie and at Nebraska City the success of the several methods of seeding
was in the same order as at the other stations, except that trench planting

114

EXPERIMENTS DURING 1923.

outranked surface sowing. The average survival of the seedling transplants
at these stations was less than on high prairie (where they were watered the
first season), but exceeded that in mixed prairie.

An examination of the species which gave the greatest percentage of sur¬
vival is of interest. As to surface sowing, Petalostemon, Andropogon nutans,
and A. furcatus alone survived at Lincoln, and the three boutelouas at Phillips-
burg, so far as the three major stations are concerned. However, Andropogon
nutans survived in the two stations at Lincoln as well as at Nebraska City,
and A. furcatus on high prairie and at Nebraska City. Liatris punctata was
the only other survivor and was found at Nebraska City alone.

Table 40. —Summary of survival of 1922 plantings in August 1923.

Method of seeding.

Lincoln.

Phillips¬

burg.

Burling¬

ton.

Lincoln

(low

prairie).

Nebraska

City.

Surface sowing.

Trench.

Denuded quadrats.

Average.

Seedlings.

In the trench, Bouteloua racemosa and Desmodium were the sole survivors at
Lincoln, and Andropogon nutans at Phillipsburg, while, as before, none were
found at Burlington. But Bouteloua racemosa also survived at Nebraska City
and on the low prairie and Andropogon nutans at both of these stations also.
Desmodium persisted on low prairie as well as on high prairie. The only
other survivors occurred on the low prairie alone, viz, Elymus, Petalostemon ,
and Bouteloua gracilis .

With respect to denuded quadrats, Bouteloua gracilis persisted at all five
stations, B. racemosa and Andropogon nutans at all but Burlington, Petalo¬
stemon, Panicum , and Sporobolus at three, and Andropogon furcatus, Kuhnia,
Liatris scariosa, Gleditsia, Acer saccharinum, and Elymus at two, while the
following were found at only one station each: Robinia, Bouteloua hirsuta,
Desmodium (at Lincoln), Pinus, and Acer negundo (at Nebraska City).

Among the seedlings, Bouteloua gracilis occurred at all the stations, and
B. hirsuta and Sporobolus at all but Burlington. Liatris scariosa, Andropogon
nutans, Bouteloua racemosa, Stipa spartea, and Desmodium were found at
three stations, while Ratibida columnaris occurred at two. The following
were represented at one station only: Liatris punctata, Koeleria (at Lincoln),
Petalostemon (at Phillipsburg), Kuhnia, and Elymus (at Nebraska City).

Arranging the species in order of their survival under all methods of planting
at the several stations gives Andropogon nutans 14, Bouteloua gracilis 12,
B. racemosa 11, Sporobolus and Desmodium each 7, Petalostemon and Bouteloua
hirsuta each 6 , Andropogon furcatus and Liatris scariosa each 5, Elymus
canadensis 4, Panicum virgatum, Kuhnia glutinosa, and Stipa spartea each 3,
Liatris punctata, Gleditsia triacanthus, Acer saccharinum, and Ratibida colum¬
naris each 2, while the following occurred only once: Robinia pseudacacia,
Pinus ponderosa, Acer negundo, and Koeleria cristata.

EDAPHIC STATIONS! SURVIVAL.

115

Bouteloua gracilis was the only species which survived under some method
of planting at all 5 stations. The following persisted at both Lincoln and
Phillipsburg, but not at Burlington: Petaloslemon, Andropogon nutans,
Bouteloua racemosa, B. hirsuta, Panicum, Sporobolus, Stipa spartea, Ralibida,
and Liatris scariosa.

Sod Transplants.

As to the 1922 sod transplants on the high prairie, the following died during
the fall or winter: Anemone , Grindelia, Helianthus, and one lot of Bouteloua
gracilis, while Astragalus crassicarpus died before autumn. Aside from the
grasses, nearly all of which made an excellent growth, it is interesting to
note that several forbs, viz, Liatris scariosa, Solidago missouriensis, and S.
rigida, blossomed profusely.

At Burlington the fall and winter were so extremely dry that 83 per cent
were winterkilled, including 9 species. This left a single sod of each of the
following: Agropyrum, Bouteloua gracilis, Panicum virgatum, and Andropogon
scoparius. The last died before the end of the summer. Agropyrum was
represented by only 3 shoots and Panicum by 2, neither species blossoming,
but Bouteloua gracilis made a good growth and put forth flower-stalks abun¬
dantly. The contrast between prairie and plains is consequently very sharp
and the disaster met by planted or transplanted species in the latter area
needs no emphasis. The indications in 1923 were that Bouteloua gracilis and
possibly Agropyrum glaucum were the only species that would survive per¬
manently.

Sods transplanted to the low prairie in 1922 were winterkilled to the extent
of 5 species, while Agropyrum, Bouteloua gracilis, and Bouteloua hirsuta died
the following summer. By the end of the season Koeleria and Anemone
were nearly dead, due to the severe competition and dense shade. Although
all of the other species were still prospering, the condition of the transplants
made in 1920 and 1921 indicated that a year or two would suffice to eliminate
the upland species and leave the taller subclimax grasses and forbs in pos¬
session of the area.

None of the 1922 sods in the salt-flat were winterkilled, but Bouteloua
hirsuta died the following summer. In August 1923, Koeleria was represented
by mere remnants, while several of the other species evidenced by their paler
color and dwarfed habit the uncongenial nature of the habitat, notwith¬
standing the rather wet season. However, all did fairly well, but several
failed to bloom.

As to the species transplanted into the former swamp area in 1922, Sporo¬
bolus and two lots each of Bouteloua hirsuta and B. racemosa failed to grow the
following spring. The slough-grass cast a very dense shade, and by the end of
1923 Bulbilis, Elymus, and Bouteloua gracilis had died, while Agropyrum was
represented by mere remnants. This left Spartina, which had made a normal
growth, Stipa spartea, the flowers of which were still in the leaf-sheath in late
August, and Koeleria, which failed to blossom. In another season it is
probable that these two high-prairie species would also have been eliminated.

The sods planted in 1922 at Colorado Springs did poorly because of grazing.
By the end of the second summer, Andropogon furcatus, Panicum virgatum,
and Bouteloua racemosa had died, and Elymus and Spartina were represented
by mere remnants. None but Bulbilis had blossomed.

116

EXPERIMENTS DURIN& 1923.

PHYTOMETRIC RESULTS.

Transpiration and Growth.

Plan.

In order to obtain further light as to the climatic differences of the major
stations in terms of functional response, a special investigation was made of
transpiration and growth during the summer of 1923. Plants of Helianthus
annuus, Avena sativa, Elymus canadensis , and Acer negundo were grown from
seed or transplanted as seedlings into sheet-metal containers of appropriate
shape and sufficient size to accommodate the root systems throughout the
duration of the experiment. After the plants were well established, the leaf
area and the weight of plant and container were determined. They were
then installed at the several stations, together with a complete battery of
instruments, for a period of 14 days, and measurements made of transpiration
and growth in terms of increased area. The considerable differences in
altitude and hence in seasonal development made simultaneous studies unde¬
sirable, and consequently the periods of observation were successive. This
was also imperative because of the time and effort involved.

Methods.

The individual plants of sunflower and box-elder were placed in cylindrical
galvanized-iron containers 5 to 6 inches in diameter and 9 to 10 inches deep,
filled with rich loam soil tamped firmly in place. A layer of one-half inch of
coarse gravel in the bottom of the container covered an exit-tube, consisting
of an automobile tire valve-stem with the inner end cut off and covered with a
fine copper gauze soldered in place. The core of each tube had been removed.
The tube was soldered in place with the threads projecting through the wall
of the container, so that an exhaust pump could be attached for aerating the
soil. The tube also assisted materially in watering the plants, the usual cap
preventing loss during the intervals. The soil was well screened, brought to an
optimum holard, and weighed at the time of filling the containers. By
restoring the containers to their original weight from time to time, the holard
was maintained at the desired level. To prevent loss other than by transpira¬
tion, the containers were furnished with a sloping metal top provided with a
circular opening with the edges reamed upward, and large enough to receive a
cork 2.5 inches in diameter. An effective seal was formed by boring a hole
large enough for the plant stem, splitting the cork and fitting it into place
after padding the sides of the opening with a little cotton. The seal was tested
by a check container without a plant, but fitted with a wooden peg to simulate
the plant stem. During the period of 14 days this did not lose water in an
amount sufficient to be detected by a balance sensitive to 2 grams under a
load of 7 kg.

The containers for wild rye and oats were similar to those already described,
except for the tops, which were furnished with a slit 5 inches long and 1 inch
wide. The edges of the metal cut in making the slit were turned down into
the container about a quarter of an inch on each side and a half-inch at the
ends of the slit. These furnished supports for narrow strips of shellacked oak,
which were held in place by thin wedges of similar material at each end.
These were put in place after the containers had been filled and the soil pressed

TRANSPIRATION AND GROWTH.

117

firmly under the metal tops. They were then coated with shellac, so that no
openings remained except between the wooden strips, which narrowed the
slit to about 10 mm. The seeds were planted through this opening, which was
nearly filled with soil kept moist by frequent watering. After the plants came
up, the remainder of the opening was filled with sand level with the top of the
wooden strips. The containers were provided with felt tops, cut to fit around
the slits, and sunken level with the soil. The efficiency of the sand-mulch in
preventing water-loss was found to be high, only 6 grams escaping in a period
of 2 weeks. Twelve plants of average size were selected from the several
containers and the leaf-area determined. From this, and the number of plants
in each container, the initial area of the group of plants in any container was
calculated. The final area was determined in a similar manner. This method
was used because it was quite impossible to determine the total area of the
plants in place, as could be done with the dicotyls.

Owing to cool, wet weather, the plants grew slowly, except during the last
week of May. They were kept covered during rains and at night, watered
from a burette, and aerated from time to time. To keep the metal containers
from heating the soil, they were surrounded by sand and the top covered by
a collar of felt about a centimeter thick, held in place by strips of adhesive
tape. On June 1 the leaf-area was determined by means of solio prints and
the planimeter, corks were inserted, and the containers (with collars removed)
brought back to their initial weight after they had been transported to the
station in the high prairie at Lincoln. Here they were placed in the soil and
thoroughly covered after the collars had been replaced, the exposed plants
being sheltered during rains.

After the 14-day period of the experiment, following final weighings,the
number of parent plants and number of tillers in each container was ascer¬
tained. Eighteen specimens of each group were then selected and their areas
determined. From these data the final areas were calculated.

Conditions and Results at Lincoln.

Physical Factors.

The period (May 31 to June 15) was exceptionally cool and wet. Only one
entire day was sunny, and the sun shone for over half of the time on two others
alone. During a 4-day interval the sun did not appear, while the total sun¬
shine over the whole period was approximately 5.5 days. A total of 4.9
inches of rain fell during 8 days, and rains occurred in the daytime on 6 days,
during one of which it was necessary to keep the plants sheltered for the
entire day. In fact, they were under cover for a total period of 30 hours of
daylight in which rain was falling.

The air-temperature during the first 7 days ranged from 55° to 91° F., while
the humidity was never lower than 55 per cent. This was followed by a period
of 4 days when the temperature remained at 58° to 65° F. and the humidity
above 80 per cent. The remaining time was similar to that of the first 7 days.
The average day temperature for the entire period was 70.1° F. and the day
humidity 75 per cent. The soil temperature at a depth of 6 inches among the
containers varied from 62° to 71° F. The total wind movement at a height of
39 cm. during the period was 1,282 miles, an average of only 3.6 miles per
hour. The exceptionally mesophytic conditions are shown by the average

118

PHYTOMETRIC RESULTS

Table 41. — Water-loss and growth at Lincoln.
Helianthus annuus.

Water added

/*N

-tT c3

(grams).

Final weight,

June 15 (gran

-d

• M

Con¬

tainer.

Initial weigl
June 1 (gi

June

12.

June

14.

Total water

(grams).

Initial area

(sq.cm.).

Final area.

(sq. cm.)

Average are

1 (sq.cm.).

Loss per sq.

(grams).

P. ct. increa

area.

6,748

300

250

100

6,464

934

229.1

960.5

594.8

15.7

319.2

7,053

300

250

100

6,700

1,003

266.9

1,135.5

701.2

14.3

325.4

7,035

300

250

100

6,690

995

237.9

1,142.4

690.1

14.4

380.2

6,793

300

250

100

6,587

856

167.0

1,029.1

598.1

14.3

516.2

6,942

300

250

100

6,769

823

156.7

1,075.9

616.3

133

586.6

6,966

300

250

100

6,660

956

258.5

1,099.4

678.9

141

325.3

6,898

300

250

100

6,646

902

201.0

1,127.9

664.5

135

461.1

6,700

300

250

100

6,452

898

202.7

921.1

561.9

159

354.4

6,974

300

250

100

6,629

995

177.3

1,124.4

650.8

153

534.2

6,800

300

250

100

6,516

934

244.2

928.8

586.5

159

280.3

6,980

300

250

100

6,655

975

275.1

1,205.7

740.4

132

338.2

6,962

300

250

100

6,672

940

187.5

1,106.1

646.8

145

489.9

Aver....

145

409.2

A VENA SATIVA.

Initial weight,

June 1 (grams).

Water added
(grams).

Final weight,

June 1 (grams).

<73

• ri

Con¬

tainer.

June

13.

June

15.

Total water
(grams).

Initial area
(sq. cm.).

Final area
(sq. cm.).

Average are
(sq. cm.).

Loss per sq.
(grams).

P. ct. increa

area.

8,918

600

300

8,442

1,376

414.0

1,601.7

,007.8

136

286.9

8,360

500

180

8,081

959

414.0

1,620.8

,017.4

291.5

8,233

500

120

7,885

968

414.0

1,614.4

1,014.4

290.0

8,221

500

7,860

861

414.0

1,544.0

979.0

272.9

8,237

500

7,927

810

414.0

1,436.1

925.0

246.8

8,316

500

138

8,004

950

414.0

1,596.7

,005.4

285.6

278.9

ElYMUS CANADENSIS.

1-31

5,273

100

5,186

187

145.4

492.6

319.0

238.8

2-40

5,010

100

4,896

314

187.6

635.7

411.6

238.8

3-41

5,010

100

4,916

294

192.3

651.6

421.9

238.8

4-28

5,206

100

5,070

236

131.3

445.0

288.1

238.8

5-43

5,180

100

5,044

336

201.7

683.3

442.5

238.8

Control.

5,093

5,087

238.8

Aver....

238.8

Acer negundo.

4,355

4,306

38.6

67.6

53.1

75.1

6,104

. . •

6,049

42.2

107.6

74.9

155.5

5,910

, , ,

5,872

36.2

45.6

40.9

25.9

6,363

6,300

143

99.2

241.2

170.2

143.1

5,933

5,878

58.8

72.6

65.7

23.4

Aver. . . .

84.6

TRANSPIRATION AND GROWTH.

119

daily evaporation, which was only 8.4 c. c. per day, in contrast to a normal
daily loss of 20 to 30 c. c. during this portion of the growing-season (table 41).

Transpiration and Increase in Leaf-are a.

The behavior of plants of the four species with respect to water-loss and
increase in area during the control period of two weeks at Lincoln is exhibited
in table 41.

Conditions and Results at Phillipsburg.

Installation.

Plants of the same species as those used in the preceding experiment were
grown at Lincoln and shipped to the mixed-prairie station at Phillipsburg,
where they were used experimentally for the 14-day period from June 18 to
July 3. Some differences in the age and size of the various species from those
at Lincoln were brought about by more or less favorable growing conditions
during the few weeks preceding the beginning of the experiments, but the
several species were as nearly as possible comparable with those previously
used. The same methods were employed in both cases.

Physical Factors.

The period June 19 to July 3 was one of fairly typical weather for early
summer at this station, when compared with that of the three preceding years
for which factor data were obtained. Eight days were clear, except for a few
floating clouds, and the sun shone on four others for over half of the time.
Two days alone were sufficiently cloudy so that the sun shone for less than one-
third of the day. Heavy showers occurred at night on June 18, 21, 26, and 29,
as well as during the day on June 21, when it was necessary to cover the
plants for 2 hours. Three days were so extremely hot, windy, and dry that
it was desirable to shade the plants and protect them from the wind for
periods of 2 to 5 hours. Notwithstanding an adequate holard, partial wilting
occurred at these times, crop and ruderal plants suffering likewise. During
one of the most severe days (June 24) the temperature was 98° F. and the
humidity 42 per cent, while a south wind of 20 to 30 miles per hour was blow¬
ing. The air-temperature during the first five days of the period ranged from
65° to 98° F. and the humidity from 41 to 97 per cent. For the remainder
of the time the variations were from 42° to 90° F. and 34 to 100 per cent humid¬
ity. The average day temperature for the period was 80.6° F. and the
average day humidity 60 per cent. Soil temperatures at a depth of 6 inches
varied from 74° to 82° F.

The total wind movement at a height of 39 cm. during the 14-day period
was 1,100 miles, an average of 3.3 miles per hour; moreover, most of this
occurred during the day. The somewhat xerophytic conditions are indicated
by the average daily evaporation of 25.5 c. c. Under these conditions of
prevailingly clear, warm weather, growth was marked and the losses by
transpiration were high (table 42).

Conditions and Results at Burlington.

Installation.

The plants used at Burlington were also grown in appropriate containers at
Lincoln for the usual period and shipped to Burlington, where they were used

120

PHYTOMETRIC RESULTS

experimentally for the 14-day period from July 6 to 20. Methods similar to
those utilized for the other stations were employed throughout. Owing to the
abundance of grasshoppers, it was necessary to screen the inclosure, a fence of
hardware cloth about 4 feet high, the top edge of which was turned outward
and downward, being employed. This modified the conditions of wind and
evaporation within to a considerable extent and shaded the plants for about 2
hours each day.

Transpiration and Increase in Leaf-are a.

Table 42 . — Water-loss and growth at Phillipsburg.
Helianthus annuus.

^ 3
3

Water added (grams).

u. i

<D •

• H

Con-

-g 2

c3 X

0) A

c3 G"

Average area

(sq. cm.).

'■o

cr .

tainer.

is ^

ca 2

June

23.

June

26.

June

30.

July

£ M

* >>
i i 1 l—-H

c3 3

• S

Total wa
(grams

Initial ar
(sq. cr

Final are

(sq. cn

Loss per

(grams

P. ct. inc

area.

6,862

200

300

200

6,627

•

985

65.89

860.86

463.38

212

1,206.5

6,974

200

300

200

6,639

1,085

52.84

925.73

489.29

222

1,651.9

7,030

200

300

200

6,724

1,056

57.11

934.08

495.59

213

1,535.6

7,036

200

300

200

6,628

1,158

53.75

727.39

390.57

296

1,253.3

7,061

200

300

200

6,797

1,014

54.91

870.29

462.60

219

1,484.9

7,058

200

300

200

6,718

1,090

51.29

960.98

506.14

215

1,773.6

7,045

200

300

200

6,760

1,035

50.39

945.36

497.88

208

1,776.1

7,074

300

400

200

6,717

1,307

91.86

1,133.60

612.73

213

1,134.0

Aver. . .

225

1,476.9

Avena sativa.

8,3S8

200

350

400

200

8,138

1,400

316.37

1,254.24

785.31

178

296.4

8,385

150

350

300

200

8,210

1,175

260.54

905.84

583.19

201

247.6

8,442

200

350

400

200

8,200

1,392

353.59

1,297.50

825.55

169

266.9

8,121

100

350

300

200

8,053

1,018

353.59

836.16

594.88

171

136.4

8,369

200

350

400

200

8,129

1,390

353.59

1,239.82

796.71

174

250.6

8,211

100

350

400

200

7,958

1,303

316.37

1,196.56

756.47

172

278.2

Aver.

178

246.0

Elymus

CANADENSIS.

5,048

100

200

4,900

648

214.08

720.90

467.49

139

236.7

5,289

200

400

300

4,939

1,250

214.08

780.54

497.31

251

264.6

5,422

100

300

100

5,172

750

214.08

763.50

488.79

153

256.6

Check....

5,291

4,741

200

400

300

4,913

4,733

1,278

214.OS

789.06

501.57

255

268.6

Aver.

199

256.6

Acer

NEGUNDO.

6,190

200

100

6,068

422

128.94

336.70

232.82

181

161.1

6,079

...

100

5,979

250

77.52

214.99

146.26

171

177.3

4,337

100

4,208

379

139.28

393.41

266.35

142

182.5

4,349

100

4,260

339

124.81

285.53

205.17

165

128.8

6,022

. . .

100

, . ,

5,956

166

59.17

157.24

108.21

153

165.7

5,955

100

5,878

177

69.38

179.46

124.42

142

158.6

Aver.

159

162.3

TRANSPIRATION AND GROWTH.

121

Physical Factors.

The weather conditions for the first four days of the experiment were very
characteristic of the high plains during July, with day temperatures reaching
90° to 96° F. and falling to 60° or 65° F. at night, while the humidity ranged
from 25 to 30 per cent in the afternoon to 80 or 85 per cent at night. However,
this was followed by 10 days during which the atmosphere was more humid
and cooler, due largely to an unusual amount of cloudy weather and rain.
Rains fell in the evening or at night as follows: July 11, 0.02 inch.; 14th,
1.48 inches; 16th, 0.51 inch; 17th, 0.58 inch. Moreover, the heavy rains
were general and greatly increased the humidity. One day was entirely
cloudy, and four were cloudy half or more of the time, while on 3 others the
sky was overcast more than one-fourth of the time, only 6 days being clear.
With the exception of the first 4 days, clouds invariably obscured the sun after
5 or 6 p. m.

The temperature ranged rather uniformly between 55° to 60° and 80° to 85°
F., and the humidity usually reached 100 per cent at night (the vegetation
being covered with rain or dew) and fell to 50 to 55 per cent during the after¬
noon (on one day to 36 per cent). The average day temperature and humidity
for the period from 8 a. m. to 6 p. m. were 80° and 57 per cent respectively.
While the temperature was nearly the same as the average day temperature
for the same period during the three preceding years, as determined by
hygrothermographs in the field, the humidity was 11 per cent higher. The
soil temperature at a depth of 6 inches among the containers was 73° to 86° F.

The wind movement at a height of 39 cm. during the first 4 days averaged
over 7 miles per hour, but for the rest of the time it was much less (3.4 miles
per hour). The average daily evaporation was 35.4 c. c., which was 5 c. c.
lower than for the same period of the preceding years. It was necessary to
shade the plants for an hour or so to prevent wilting during the first day
or two.

Transpiration and Increase in Area.

Table 43 . — Water-loss and growth at Burlington.
Helianthus anntjus.

Con¬

tain¬

er.

Initial weight,

July 6 (grams).

Water added (grams).

Final weight,

July 20 (grams).

Total water used
(grams).

Initial area
(sq. cm.).

Final area
(sq. cm.).

Average area
(sq. cm.).

Loss per sq. dm.
(grams).

P. ct. increase in
area.

July

11.

July

13.

July

16.

July

18.

July

19.

6,526

100

300

6,444

1,182

30.62

1,235.16

632.89

187

3,933.9

7,572

100

300

7,637

1,035

22.22

775.20

398.71

259

3,388.7

6,591

100

300

6,409

1,182

29.59

1,159.44

594.52

199

3,818.4

7,538

100

300

7,617

921

21.58

650.91

336.25

274

2,916.2

7,599

100

300

7,612

987

20.20

802.33

411.27

240

3,871.7

6,600

100

300

6,517

1,183

26.10

1,314.35

670.23

176

4,935.8

6,692

100

300

6,522

1,270

21.91

1,354.01

687.96

185

6,079.9

7,545

100

300

7,500

1,145

27.78

868.16

447.97

256

3,023.7

6,477

100

300

6,446

1,031

28.17

1,114.18

571.18

181

3,855.2

6,636

100

300

6,754

882

18.22

955.05

486.64

181

5,141.7

Aver

214

4,096.5

122

PHYTOMETRIC RESULTS

Table 43. — Water-loss and growth at Burlington —Continued

Avena sativa.

Container.

Initial weight,

July 6 (grams).

Water added (grams).

Final weight,

July 19 (grams).

Total water used

(grams).

Initial area

(sq. cm.)

Final area

(sq. cm.).

Average area

(sq. cm.).

Loss per sq. dm.

(grams).

P. ct. increase in

area.

July

11.

July

13.

July

16.

July

18.

8,850

200

250

200

8,722

778

92.92

566.94

329.93

236

510.1

8,732

150

100

200

150

8,593

739

100.33

721.32

410.83

180

618.9

8,702

150

100

200

8,583

769

85.48

642.81

364.15

211

652.0

8,803

100

150

100

8,734

519

38.93

506.17

272.55

190

1,200.0

8,915

150

100

200

8,816

749

89.21

523.91

306.56

244

487.3

Check.

8,742

8,734

Average.

212

693.7

Elymus canadensis.

Initial weight,

July 6 (grams).

Water added (grams).

/•—N

Container.

July

11.

July

13.

July

16.

July

18.

Final weight,
July 20 (gra

Total water us
(grams).

Initial area
(sq. cm.).

Final area
(sq. cm.).

Average area
(sq. cm.).

Loss per sq. di
(grams).

P. ct. increase
area.

5,253

100

5,211

192

47.10

176.87

111.98

172

275.5

5,277

• . •

100

5,226

301

88.81

212.08

150.45

200

138.8

5,197

• . •

100

5,159

238

64.98

155.25

110.12

216

138.9

5,158

...

100

5,110

298

74.18

175.53

124.86

239

136.6

5,336

100

5,284

252

59.15

132.02

95.58

264

123.2

Average.

218

162 6

Acer negundo.

Container.

Initial weight,

July 5 (grams).

Water added (grams).

Final weight,

July 19 (grams).

Total water used
(grams).

Initial area
(sq. cm.).

Final area
(sq. cm.).

Average area
(sq. cm.).

Loss per sq. dm.
(grams).

P. ct. increase in
area.

July

13.

July

16.

July

18.

6,112

100

150

6,055

357

134.50

375.84

255.17

139

179.4

5,995

100

5,986

209

70.54

234.11

157.33

137

231.8

4,384

150

100

4,300

484

163.83

584.06

373.95

129

256.5

4,339

200

100

150

4,252

687

283.08

737.99

510.54

134

160.7

4,355

150

100

150

4,290

615

218.84

685.40

452.12

136

213.2

4,194

100

4,151

443

125.58

475.71

300.65

147

278.8

Average.

137

220.1

TRANSPIRATION AND GROWTH.

123

Table 44. —Transpiration and growth at the stations.

Species.

Water-loss per
sq. dm. (grams).

Increase in area
(p. ct.).

Rate of growth, based
on actual increase in
area (p. ct.).

Lin¬

coln.

Phil¬

lips¬

burg.

Bur¬

ling¬

ton.

Lin¬

coln.

Phil¬

lips¬

burg.

Bur¬

ling¬

ton.

Lin¬

coln.

Phil¬

lips¬

burg.

Bur¬

ling¬

ton.

Sunflower.

145

225

214

409

1,477

4,097

854

860

998

Wild rye.

199

218

239

257

163

410

549

140

Oats.

178

212

279

246

694

1,155

796

510

Box-elder.

159

137

162

220

161

370

Average. . .

100

190

195

253

535

1,294

618

591

504

Table 45. —Environmental conditions at the stations.

Station.

Approxi¬

mate

hours

sunshine.

Average

day

temp.

(° F.).

Soil

temp.

(° F.)

Average

day

humidity
(p. ct.).

Average
daily evap¬
oration
(c. c.).

Wind
per hour
(miles).

Lincoln.

70.1

62 to 71

8.4

3.6

Phillipsburg.

80.6

74 82

25.5

3.3

Burlington.

80.0

73 86

35.4

3.4!

Summary and Conclusions.

This onerous series proved a disappointment in so far as normal climatic
relations were concerned, owing to the wholly exceptional weather. This is
revealed by table 45, and is brought out even more graphically by comparing
the data here with those obtained during other years. The amount of sun¬
shine at Lincoln was little more than half that at the other stations, while the
air-temperature averaged 10° lower and the soil temperature ranged from 11°
to 15° lower. The average humidity was 15 to 18 per cent higher and the
evaporation but a third or a fourth of that at Phillipsburg or Burlington.
Hence, it is easy to understand why the increase in area was twice as
great at Phillipsburg and more than four times as great at Burlington
ihough the normal relation is suggested by the rate of growth based on the
actual increase in area and by the order of water-loss at the three stations.
Thus, while the plant responses are in agreement with the physical factors for
the respective fortnights concerned, it is obvious that entirely comparable
results could be insured only by dealing with the growth season for each species.
An adequate record of transpiration and growth for such a period at stations
widely separated demands at least one resident investigator for each station,
and such studies must await the future.

TRANSPIRATION AND GROWTH OF COMMUNITIES.

Transpiration from Natural Cover and Crops.

Objectives.

One of the major tasks of quantitative ecology is to determine the functional
responses of plants when grouped in communities. While much light can be
obtained by the use of individual plants under control in the field, in the form

124

PHYTOMETKIC RESULTS.

of standard phytometers, these differ essentially in their soil and competition
relations from plants growing together in the actual cover. Hence, the task
is to maintain these natural relations of the community and at the same time to
secure a degree of control that modifies the efficient factors little or not at all.
In the case of transpiration, for example, these requisites can be met
only by weighing, as all other methods modify the physical factors to an
undesirable degree. The method that maintains the soil and community
relations with the minimum disturbance is the soil-block, which was first
employed for determining the chresard in the field (Clements, 1904, 1905).
This requires only such slight modifications as those of size and form to
become applicable to all problems in which an undisturbed soil-root core
is indispensable.

In consequence, the first objective was to perfect the soil-block method so
that it could be used in the field with both convenience and accuracy. Because
of its importance in the grassland climate, the chief function to be measured
was transpiration, though chresard and aeration can be studied with something
of the same readiness. In the present case the transpiration from representa¬
tive cores was followed in the proper climate of each association, but it is
evident that the containers can be moved or exchanged between different
edaphic or climatic stations and thus serve as reciprocal phytometers. This
permits the determination of the transpiration behavior of each climax in its
own climate in terms of adjustment and adaptation and at the same time
affords a basis for comparing adjacent climaxes. A further use of funda¬
mental value arises out of the rainfall relation. The method of the soil-core
not only makes it possible to trace the complete water-cycle of rainfall, holard,
evaporation, and transpiration, but also to estimate the extent to which the
vegetation of each region may furnish the water-vapor for its own rainfall
(Clements, 1923). Finally, it also opens up a new field in the functional rela¬
tion of roots to the soil as an actual structure, which shows striking differences
from climate to climate, as well as from one local habitat to another.

Methods.

A steel cylinder 12 inches tall and with an inner area of 1 square foot, the
lower edge of which was sharpened, was driven into the grassland soil to a
depth of 4 inches. Care was taken to cut off none of the leaves belonging to
the plants in the square foot selected, which was chosen with a special regard
to its representative structure. The cylinder was then carefully removed,
leaving the column of soil intact, and replaced by one of heavy galvanized
iron 3 feet long and reinforced at both ends by a heavy wire over which the
metal was turned back smoothly. After starting a row of these cylinders at
distances of 8 inches, a trench 2 feet wide was dug around them to a depth of
over 3 feet. In this process no soil was removed within 3 or 4 inches of the
cylinder. As the trench was deepened, the columns of soil were carefully
pared away with large knives in such a manner that the cylinders could be
forced down under considerable pressure from above. By shaping the column
for a few inches in front of the descending cylinder, it was possible to force
the latter into place over a tightly fitting soil-core to a depth of 3 feet. The
columns were then undercut and smoothed off level with the lower end of the
cylinder. A loose-fitting metal bottom with the edges 2 inches deep was

TRANSPIRATION FROM NATURAL COVER AND CROPS. 125

placed over the end and the entire container was then weighed on a portable
Fairbanks scale sensitive to one-fourth pound.

In the meantime, a trench sufficiently wide and deep to receive the cylinders
in an upright position had been dug in a nearby area, care being taken not to
cover the grass with soil. The containers were lowered in the new trench and
slid into place on a plank in the bottom, after which the bottoms were made
water-tight by means of a measured amount of hot wax of the usual composi¬
tion. The trench was then filled with soil and pieces of sod were fitted around
the tops so that the surface conditions would be essentially normal. The
trenches were selected so that the surface water would readily drain away
from them and in addition the plants were covered by wooden roofs whenever
rain was actually falling. This was imperative because of the varying inter¬
ception of rainfall by the different vegetation in the several containers. In
the case of the cultivated crops, oats and millet, the usual type of bottom was
replaced by one 3 feet deep, owing to the difficulty of selecting a proper slope
for drainage. In order to determine the amount of water evaporated from
the cultivated soil, the plants were removed from one container in each field.
In another check, the natural grasses were left in place after having been
killed by the addition of a measured amount of boiling water.

From time to time, depending upon the weather and the needs of the plants,
water in measured amounts was slowly added to all the containers, and as a
result there was little shrinkage of the core from the sides of the container.
None of the plants died, and even those near the edges gave no signs of wilting,
demonstrating that the roots in the core supplied abundant water for trans¬
piration. Much care was exercised in watering, so that there was little or no
run-off down the sides of the core. This was accomplished by pouring the
water on slowly and pressing the moist soil firmly against the cylinder wherever
the contact was not complete. At the end of the period the containers were
again weighed and the losses calculated.

At the end of the experiment the vegetation was carefully removed at the
soil surface by means of a hand grass-clipper. The dense foliage of former
years was carefully separated from the living plants, the latter oven-dried at
60° C., and weighed.

True Prairie, Series 1.

Installation. —The first series consisted of 6 sod-cores which were installed
on the high prairie during the period of May 31 to June 15. The group in
container 2 consisted chiefly of a sod of Andropogon nutans and Bouteloua
racemosa, in which were found 4 bunches of Koeleria cristata. Container 1
was very similar, having been taken within 18 inches from the former; the
plants in it were killed with hot water and it was then employed as a control.
The third group consisted mainly of Andropogon nutans, A. scoparius, and
Koeleria, and container 4 resembled it closely. The group in container 5
comprised several bunches of Stipa spartea, 4 small clumps of Bouteloua
racemosa, and 3 of Andropogon scoparius; No. 6 was occupied chiefly by
Stipa spartea and Andropogon scoparius.

The entire period was exceptionally cool and wet, only one day being wholly
sunny and the total sunshine being approximately only a third of the total
daytime period. The average day temperature was 70° F. and the day

126

PHYTOMETRIC RESULTS.

humidity 75 per cent. The unusual nature of the weather is perhaps best
revealed by the fact that the average daily evaporation was only 8.4 c. c. in
contrast with a normal loss of 20 to 30 c. c. per day during this period.

Results. —A striking character of the soil columns is their rather uniform
weight. Cores 1 to 4 were obtained from a single trench, the extreme distance
between the containers not exceeding 6 feet. The weight of the 3 cubic feet of
soil (exclusive of container) varied only from 289 to 295 pounds. Soil columns
5 and 6 were obtained about 50 feet distant from the first trench.

Table 46. — Water-losses from sods in high prairie, Lincoln, Nebraska.

Container.

Original
weight,
May 31.

Weight

seal.

Water
added
May 31.

Final
weight,
June 15.

Loss.

Dry

weight of
foliage.

Loss per
gram of dry
weight.

lbs.

gm.

1 (control).

309 50

0.60

5.5

312.5

3.1

2..

307.00

0.60

4.0

303.0

8.6

19.18

203.4

305.25

1.20

4.0

302.0

8.45

18.95

202.3

303.75

1.00

4.0

302.5

6.25

13.88

204.3

313.00

0.80

4.0

311.0

6.8

18.95

162.8

308.00

1.20

4.0

308.0

5.2

15.18

155.4

The water-losses, though low, are fairly consistent, ranging from 2.1 to 5.5
pounds per square foot in excess of that of the area covered with dead grasses.
Moreover, the variations are directly proportional to the extent of the grassy
cover, as expressed in dry weight. The mixture of grasses (chiefly andropogons
and Koeleria ) lost about 203 grams of water per gram of dry matter, which
was somewhat more than that from Stipa spartea (average 159 grams).

True Prairie, Series 2.

Installation. —Because of the unfavorable conditions during the first series,
the experiment was repeated from July 24 to August 8, 5 containers being
installed on the high prairie and 4 in the adjoining low prairie. In container
1 a mixture of Andropogon nutans, Koeleria cristata, and Poa pratensis, with
a single plant of Amorpha canescens (.Andropogon being most abundant),
covered half of the area. These had an average height of 12 inches. The
surface of three-fourths of the second container was covered principally with
Andropogon nutans , but also with a mixture of Koeleria cristata, Bouteloua
racemosa, and a little Poa pratensis, with an average height of 15 inches.
The grasses in container 3 were chiefly Andropogon scoparius, with a third as
much A. nutans and a very little Bouteloua racemosa; the height was 14 inches.
In container 4, one-fifth only of the area was bare, the rest being occupied
with a mixture of the grasses named above, Andropogon scoparius dominating.
Amorpha canescens and Antennaria campestris were each represented by a
single small clump. The control container, No. 5, in which the grasses were
killed as usual, was very similar to No. 3.

In the low prairie two clumps of Andropogon furcatus, 22 inches in average
height, occupied slightly less than half of the area in container 6, the rest being
bare. Nearly two-thirds of the area in container 7 was bare, the rest being
occupied by a large clump of A. furcatus 16 inches in height. Spartina cyno-

TRANSPIRATION FROM NATURAL COVER AND CROPS. 127

suroides and a little Panicum virgatum, with an average height of 38 inches,
filled about one-third of container 8, the remainder being destitute of plants.
The remaining cylinder was sunk and a soil-core obtained in a field of alfalfa
where the plants were 21 inches high and beginning to blossom freely. A
single clump in the center of the core was obtained which occupied only
one-fourth of the area.

Because of a prolonged period of hot, dry weather, the grasses at both sta¬
tions had rolled leaves and some of them wilted during the afternoons, a
result rarely seen in the true prairie. However, the weather conditions for
the following 15 days were below normal in the amount of sunshine and heat.
Only 2 days were entirely clear and rain fell on 8 different days. The average
day temperature was 79° F. and the average day humidity was 80 per cent.
With respect to transpiration, conditions were much more nearly normal
than in the preceding series, as is shown by the fact that the average daily
evaporation was 22 c. c. in contrast to 8 c. c.

Results. —The losses per square foot in the upland prairie ranged from 9.4 to
13.7 pounds, as compared with 4.8 pounds from the control. In the low prairie
the losses varied from 12.3 to 17.5 pounds, owing to the more luxuriant growth
of the vegetation, while that from alfalfa was 21.5 pounds. Based on the
quantity of water lost per dry weight, the upland prairie grasses show a
relatively higher loss (maximum 332.5 gm.) than the coarser, larger-stemmed
lowland species, where a maximum loss of only 165 gm. of water per gram of
dry matter was found. On this basis the woody-stemmed Spartina transpired
less even than the upland grasses. Alfalfa lost 442 grams of water per gram
of dry matter.

Table 47. — Water-losses from sods and from alfalfa, Lincoln, Nebraska.

Container.

Original
weight
July 24.

Weight

seal.

Water added.

Final
weight
Aug. 8.

Loss.

Dry

weight

plants.

Water
lost per
gram dry
weight of
plants.

July

25.

July

26.

July

28.

July

30.

Aug.

1 .

Aug.

lbs.

lbs.'

lbs.

gm.

1 .

294.75

0.61

300

9.36

12.77

332.5

294.50

1.22

296.0

13.72

32.07

194.1

291.25

1.08

295.0

11.33

23.12

222.3

4 .

286.25

0.81

291.25

9.81

22.70

196.0

fo.hp.AkT

293 75

0.81

300.75

4.81

308.75

2.03

313.25

17.53

66.53

119.5

291.00

0.81

296.5

15.31

42.17

164.7

322.00

2.03

331.75

12.28

56.07

99.4

307.50

2.03

312.0

21.53

22.09

442.1

Mixed Prairie.

Installation. —At Phillipsburg 6 cylinders were used in the mixed prairie
and 3 in an adjoining field of oats. Container 2 was two-thirds occupied by a
dense sod of Bouteloua gracilis with a very little pistillate Bulbilis dactyloides;
these had an average height of 9 inches. The third container was fully two-
thirds covered by a practically pure growth of Bouteloua gracilis with an
average height of 9 inches, in which were found a few stalks of Sporobolus
cryptandrus and one small Callirrhoe involucrata. The control, No. 1, bore a
vegetative cover of a kind and density almost exactly like that in No. 3.

128

PHYTOMETRIC RESULTS.

About one-fourth of the area in container 4 was bare; the rest was covered with
a dense sod of pistillate Bulbilis dactyloides 4 to 8 in. tall. There were also
four small clumps of Car ex filifolia and a very small amount of Bouteloua
gracilis. A fine clump of Andropogon furcatus, 18 inches in average height,
occupied 96 square inches of the area in container 5, the rest being devoid of
vegetation. Container 6 also bore a clump of the same grass 91 square inches
in extent at the base and with an average height of 20 inches.

The period from June 18 to July 3 was typical of early summer at this
station, as shown by the factor data for the preceding years. Except for a
few clouds, 9 days were clear and 4 others were sunny for more than half of
the time. The average day temperature was 80.6° F. and the average day
humidity was 60 per cent. The effect of the factors concerned in transpiration
is indicated by the average daily evaporation, which was 25.5 c. c.

Results. —Among the short-grasses the amount of water added slightly
exceeded the amount used; it was approximately the same in amount in the
oats, but the soil lost 3 to 7 pounds of its original weight in the case of the
bluestems. The losses from the control with a dead short-grass cover and
from bare soil were 4.4 and 7.2 pounds respectively. Those from containers
2 and 3, where the dominants were grama grass, were 14.7 and 16.2 pounds
per square foot respectively, or approximately 1 pound (pint) of water per
day. The buffalo-grass sod lost nearly as much, 14.2 pounds, while loss
from the big bluestem was nearly twice as great, 27 and 28 pounds respectively.
The oats, even in the late stage of development, lost practically the same
amount of water as the short-grasses (14.7 and 16.4 pounds).

Table 48. — Water-losses from sods and oats , Phillipsburg, Kansas.

Con¬

tain¬

er.

Dominant

species.

Origi¬

nal

weight,
June 18.

Weight

seal.

Water added.

Final
weight,
July 3.

Loss.

Dry

weight

plants.

Loss

per

gram
of dry
weight.

.Tune

18.

June

19.

June

22.

June

26.

June

30.

Control, dead

Bouteloua gra-

lbs.

am.

Qm.

cilis.

276.25

0.9

286.75

4.40

B. gracilis....

278.75

0.9

281.00

14.65

19.30

344.3

Do.

274.00

1.2

275.00

16.20

20.42

359.9

Bulbilis dacty-

loides.

275.50

0.75

278.00

14.25

21.76

297.0

Andropogon

furcatus.

282.50

0.9

271.25

28.15

53.90

236.9

Do.

282.30

0.75

275.00

27.05

53.94

227.5

Oats.

269.25

2.43

271.00

14.68

34.30

194.1

Do . . ..

267.50

2.43

267.5

16.43

36.43

204.6

Control, bare

soil.

285.75

2.43

288.00

7.18

.. ..

• . •

Short-grass Plains.

Installation. —The experiments in the short-grass plains were carried out
at Burlington from July 5 to 20. Six containers were used in the grassland
and three in an adjoining field of millet. Container 2 was covered with a
dense closed mat of nearly pure pistillate Bulbilis dactyloides, except for about
one-sixth of Bouteloua gracilis. This formed a fine green carpet 4 inches in
height. Only one-fifth of the area inclosed by the cylinder was bare. The
check container, No. 1, in which the vegetation was killed, was similar, except

TRANSPIRATION FROM NATURAL COVER AND CROPS. 129

that one-fourth of the area was bare. Container 3 was very much like No. 1,
except for a mixture of about one-third grama, which was flowering at a
height of 8 inches. The average height of the short-grass foliage was 4.5
inches. Container 4 consisted of nearly pure pistillate Bulbilis about 4 inches
tall, with just a little grama, some of which had flower-stalks; only one-sixth
of the area was bare. Containers 5 and 6 had a rather dense growth of Agro~
pyrum glaucum. The foliage averaged 20 inches in height, most of the stems
being dry and yellowish to a height of 6 or 8 inches. There were no flower-
stalks. The two containers were very similar, No. 6 having slightly less
vegetation. The millet was about 4 weeks old, of thick stand, and had an
average height of 18 inches. It had not begun to head at the beginning of the
experiment (July 5), but by the end of the 15-day period it was well headed
at an average height of 20 inches. Two containers with millet were used, and
one from which the crop had been removed by pulling out the plants.

The weather during the period was rather exceptional for July on the high
plains, only 7 of the 15 days being entirely clear. The average day tem¬
perature was 80° F. and the average day humidity was 57 per cent. While the
temperature was much the same as that for this period during the 3 preceding
years, the humidity was 11 per cent higher. The average daily evaporation
was 35.4 c. c. as compared with 40.5 c. c. for the same period in previous
years.

Table 49. — Water-losses from sods and millet, Burlington, Colorado.

Con¬

tain¬

er.

Dominant

species.

Origi¬

nal

weight,
July 5.

Weight

seal.

Water added, July—

Final
weight,
July 20.

Loss.

Dry

weight

plant.

Loss
per
gram
of dry
weight

12.

17.

18.

19.

Control,

dead

Buffalo

lbs.

gm.

grass....

254.75

2 03

265.75

5.03

Same, living

260.5

1.22

264.25

12.47

15.4

367.3

Do...

254

1.22

259.00

11.22

17.75

286.7

Do...

265.25

1.22

268.75

12.72

15.75

366.3

Wheat

grass....

251.5

1.22

254.00

18.72

45.31

187.4

Do...

254.25

1.08

258.50

16.83

40.32

189.3

Millet.

274.78

1.22

274.50

21.50

46.95

207.7

Do. . .

278.75

1.22

279.00

18.97

40.20

214.2

Control

bare soil.

267.25

1.22

271.50

6.97

• • * *

Results .—The final weights of the containers were only slightly greater than
their original weights, which means that approximately all of the water added
was again lost by transpiration or direct evaporation from the soil and dead
plant surfaces. The latter amount was only 5.03 pounds (container 1). The
square foot of Bulbilis and Bouteloua in each of the three containers lost 11 to
13 pounds. The wheat-grasses lost 17 to 19 pounds, varying directly with the
amount of foliage concerned, while the losses from the millet were even greater,
19 to 21.5 pounds. It is interesting to note that the bare area lost approxi¬
mately 2 pounds more water than a similar area covered with dead grasses.
The loss of water in grams per gram of dry weight of the vegetation is almost
identical (about 366 grams) in containers 2 and 4, the smaller ratio in No. 3

130

PHYTOMETRIC RESULTS.

probably being due to the presence of flower-stalks of grama grass, which
increased the dry weight but transpired little. The relatively low ratio of the
wheat-grasses (188 grams) may readily be explained when their large, very
fibrous stems are considered. Losses from the millet were intermediate,
probably being relatively less than earlier because of the heads.

Considered solely on the basis of loss per square foot, the millet ranks first,
wheat-grass second, and the short-grasses last. Thus, as is shown by the dis¬
tribution of the native species, wheat-grass is less adapted to this semiarid
region than are the short-grasses, more than one-third more water escaping
from the area occupied by the former. These data, however, hold only for
the conditions of the experiment, viz, with the plants growing in very dry soil
and under unusual humidity for the Great Plains. This is clearly brought out
by a comparison of the physical factors w r ith those of preceding seasons.

Summary.

Transpiration from natural cover .—As in the case of water-loss from the
phytometers, the exceptional weather of the summer obscured the normal
climatic response of the sod-cores. When the physical factors and the type of
vegetation are taken into account (table 50), it is clear why the low short-grass
cover at Burlington in the driest climate transpired less than the mixed prairie
at Phillipsburg in a moister atmosphere, and more than the luxuriant true
prairie at Lincoln in a much more humid climate. Thus again, while the use of
sod-cores contributed no clear-cut evidence as to the relation of the three
climaxes and their climates, it does demonstrate the value of the method and
what can be expected of it when employed through a series of years. The
losses from the three crops, while of interest, have no comparative value, since
a different species was used at each station; they are distinctly helpful, how¬
ever, in showing the similarity in the behavior of the native cover and repre¬
sentative crops for each station and in potential importance for the rainfall of
each region. However, in the future development of the method it is obvious
that the same dominant and the same crop should be employed throughout
the series of stations, and this should involve the reciprocal transfer of sod-
cores and crop-cores between the three stations.

Table 50. —Comparison of factors and average water-loss.

Station.

Dominant

grasses.

Date of
experi¬
ment.

Approx¬

imate

sun¬

shine.

Average

day

temper-

ture.

Average

day

humid¬

ity.

Average

daily

evapo¬

ration.

Average
daily loss
from
sq. ft.
of cover.

Bulbilis.

July 5
to

July 20

] p. ct.

\ 71

0 F.

p. ct.

c. c.

lbs.

Burlington....

•j Bouteloua. .. >
w Agropyrum. . (

35.4

0.96

Phillipsburg.. .

( Bouteloua. .. .

June 18

\Andropogon. .

July 3

80.6

25.5

1.33

Lincoln.

Andropogon .)
Stipa.1

July 24

22.0

0.8.5

| Koeleria.

Bouteloua. ..

Au&. 8

CLIP-QUADRATS.

131

Growth of Natural Cover and Crops.

Objectives.

The growth of a representative area of a community may be used as a
climatic index as well as a measure of response in much the same way as
transpiration. It possesses three distinct advantages over the latter, though
the two are complementary and hence one can not replace the other. Growth
demands no laborious installation, as it can be determined directly from the
native or culture community in position. While it can be measured at any
time, it yields the major values at the end of the growing-season, and thus
does not require the services of a resident investigator. Moreover, it inte¬
grates the response for the whole season, though as a complex it permits less
ready analysis than transpiration and is also less satisfactory for short
periods. In short, it is the simplest and most convenient of all community
phytometers when employed in the form of the clip-quadrat. This is the
only practicable method, as the measurement of the individuals in a com¬
munity group is too time-consuming to be desirable. There are certain cases
in which it is profitable to pull the individuals out with their roots, but this is
hardly feasible in a close cover or a compact soil. The clip-quadrat is merely
the usual one of a square meter in extent, from which the shoots are cut at any
desired time. It may be either smaller or larger in order to meet special con¬
ditions, as in the case of crops planted in rows. The growth is regularly
expressed in terms of dry matter, but in the case of grazing ranges or forage
crops, the green weight should likewise be found. Finally, the clip-quadrat
facilitates the analysis of community response to climatic factors by making it
possible to measure growth during different portions of the season or its
variations from season to season or from the wet to the dry phase of a climatic
cycle, and to determine the part played by the various species in the total
production.

In the case of the grains, it is often preferable to select the individuals to be
cut in accordance with the results desired, instead of taking all those in a
particular area (Weaver, Jean, and Crist, 1922). This may be regarded as an
aggregate clip-quadrat; it has the further advantage of permitting measure¬
ment of particular individuals throughout the season.

Plan.

Clip-quadrats were first installed at the three stations in 1920; these repre¬
sented high and low prairie in the true-prairie association, mixed prairie, and
short-grass plains. They were again cut in 1921 and 1922 to determine the
fluctuation from year to year. The first step in the simple procedure was to
select a considerable number of quadrats in typical areas of each climax.
The height and density of the cover, the abundance of dominant and sub¬
dominant species, the presence of layers, etc., were recorded and photographs
made of certain of the quadrats. In some instances it has proved desirable
to make a chart of these as well. The cover was then removed by cutting it
near the surface and at a uniform level with a hand-clipper. It was collected,
sorted, and then shipped into the laboratory to be thoroughly air-dried, after
which the actual production was determined on the basis of the dry weight.
This gave an expression of the growth of the community as a unit, as well as
the role taken by each dominant or subdominant in this. As complete factor

132

PHYTOMETRIC RESULTS.

records were obtained at each station, this made it possible to correlate the
yield of community or species with the climate and the season at each. The
physical factors for the several years have been given in the preceding chapters
and hence are not repeated here.

Since even the most uniform cover shows some variation in density, the
clip-quadrats were selected with much care and in sufficient number to insure
dependable results. As in all ecological studies that are adequate, i. e., causal
and quantitative, rather than merely mathematical, this demands considerable
knowledge of the community and can not be met by random selection. The
best plan is to locate a proper proportion of quadrats in pure or nearly pure
stands of each dominant whenever these are present and to distribute the
others among the various mixtures. It is often desirable to take the sub¬
dominants into account in doing this, as their yield may be much greater than
that of the grasses.

Results for 1920.

The growth response for this year was obtained from about 400 clip-
quadrats, of which 50 were in the true prairie at Lincoln, 50 from mixed prairie
at Phillipsburg, 50 from the short-grass plains at Burlington, and 30 from each
of the cultivated fields at the various stations. The short-grasses at Burling¬
ton averaged 103 gm., those at Phillipsburg 290 gm. Wheat-grass ( Agropyrum
glaucum) at Burlington yielded 398 gm. per square meter, while at Phillipsburg
the yield was 480 gm. A mixture of short-grass and tail-grasses gave 244 gm.
per square meter at Burlington and 470 gm. at Phillipsburg, and the mixed
prairies at Phillipsburg produced a total yield of 439 gm. per square meter,
while those at Lincoln averaged 452 gm. The average yield at Phillipsburg
(439 gm.) exceeded that from the hilltops at Lincoln (361 gm.) or even the
slopes (429 gm.) at this time (July 9), but did not equal that of the lowland,
which was 564 gm. Moreover, an examination of the factor data shows that
late-summer drought usually prevails at the western stations and good plant
growth is not maintained. This is well illustrated by a second series of
quadrats cut August 16 to 24. The average yield at Burlington was 196 gm.
(more mixed short and tall grass quadrats being included than before), at
Phillipsburg 310 gm., and at Lincoln 465 gm. On the basis of the 50 quadrats
taken at each station during the season, the average total production, pro¬
ceeding from the drier western station eastward, was 183, 378, and 458 gm.
respectively.

Crops of oats ( Avena sativa), spring wheat ( Triticum aestivum), barley
{Hordeum vulgare), alfalfa ( Medicago sativa), and white sweet clover ( Melilotus
alba) were grown in plats adjoining the several grassland stations (Weaver,
Jean, and Crist, 1922:76). Each kind of crop was planted from the same lot
of seed and grown under conditions of farm practice common to the several
localities respectively. The crop plants were thoroughly air-dried before
weighing, and the relative height of the mature oats and wheat at the several
stations, as well as the comparative yield from an average square meter, were
determined. The yield is the average of 25 to 30 square-meter quadrats taken
from each of the several plats at the three stations when the grain was ripe.
400 plants of alfalfa and 300 of sweet clover of average size were carefully
selected at each of the stations, cut at the ground-line, thoroughly air-dried,
and the dry weight ascertained. This was done during July and again in

GROWTH OF NATURAL COVER AND CROPS.

133

August for the sake of obtaining comparative values between clip-quadrats
and a group of selected plants (table 51).

The height-growth of all crops was greatest at Lincoln, less at Phillipsburg,
and least at Burlington, the height of the former averaging more than twice
that of the latter. The maximum penetration of roots was usually greatest at
Phillipsburg, next at Lincoln, and least at Burlington, a divergence readily
explained by the greater holard at the first and the dry subsoil and hard-pan
at the last.

Table 51. —Growth and yield of crop-quadrats and plants, 1920.

Crop and station.

Date of
harvest.

Average

height.

Maximum
depth of
roots.

Average
yield in
grams per
sq. met^r.

Weight
of 1,000
kernels.

Oats:

Lincoln.

July

feet.

3.0

feet.

4.8

706

gm.

20.1

Phillipsburg.

July

2.6

6.0

379

16.6

Burlington.

July

1.5

2.7

175

16.2

Wheat:

Lincoln.

July

3.2

4.8

740

29.8

Phillipsburg.

July

2.3

5.8

322

9.1

Burlington.

July

1.7

2.7

205

20.1

Barley:

Lincoln.

July

2.7

5.4

607

32.7

Phillipsburg.

July

2.4

6.7

407

14.7

Burlington.

July

1.7

2.9

176

23.9

Alfalfa:

Lincoln.

July

1.5

5.7

Dry weight
of plants.
528

Phillipsburg.

July

0.7

5.0

292

Burlington.

July

0.4

2.3

122

Sweet clover:

Lincoln.

July

2.0

5.5

840

Phillipsburg.

July

1.3

5.7

461

Burlington.

July

0.4

2.8

213

Alfalfa:

Lincoln.

Aug.

1.8

5.9

739

Phillipsburg.

Aug.

1.2

• • •

601

Burlington.

Aug.

0.6

2.0

214

Sweet clover:

Lincoln.

Aug.

2.5

• • •

1,103

Phillipsburg.

Aug.

1.7

• • •

869

Burlington.

Aug.

0.8

2.7

323

The average yield of grain for the three cereals was about twice as much at
Lincoln as at Phillipsburg, and nearly twice as much at the latter as for
Burlington. A similar relation obtained for the first cutting of alfalfa and
sweet-clover, but at the second cutting Phillipsburg more nearly approached
Lincoln. Thus, the five crops decreased decisively in both height and yield
from Lincoln westward in correspondence with the rainfall, holard, and
humidity.

Results for 1921.

Two series of cuttings were made during 1921, one about July 1 and a second
during the last half of August. Ten clip-quadrats each of Bulbilis dactyloides
and Agropyrum glaucum were used at each of the stations early in the summer;
at all of these both grasses had headed or blossomed at the time of cutting.

134

PHYTOMETRIC RESULTS.

The buffalo-grass averaged about 4.5 inches high at Burlington, while the
staminate spikes were 5.5 inches tall. In the mixed-prairie the foliage of this
grass averaged 5 inches in height and the flower-stalks 6 inches. Bulbilis
occurs in the prairies at Lincoln only where the tail-grasses are held in check by
grazing; hence it was impossible to secure an ungrazed sod comparable to that
at the other stations. The foliage was only 4.5 inches high and the effects of
heavy overgrazing the preceding year were apparent. On July 1 the average
yield of 10 quadrats at each of the respective stations was 207, 266, and 235
gm. per square meter. At Burlington the leaves of wheat-grass had reached an
average height of 16 inches and the flower-stalks of 26 inches at this time,
at Phillipsburg the respective values were 22 and 30 inches, and at Lincoln 24
and 32 inches. The average yield of 10 quadrats from each station was 400,
457, and 606 gm. respectively.

Table 52 .—Growth and yield of crop-quadrats in 1921.

Crop and station.

Aver¬

age

height.

Average
number
stalks
per sq.
meter.

Average

number

stalks

per

plant.

Average
length
heads or
panicles
in inches.

Average
number
heads per
sq.

meter.

Average

total

weight dry
matter per
sq. meter.

Maxi¬

mum

depth.

Oats:

feet.

gm.

feet.

Lincoln.

3.2

375

3.3

10.5

283

792

4.8

Phillipsburg. . .

2.8

353

2.9

269

366

5.3

Burlington....

1.5

414

2.5

171

180

2.5

Wheat:

Lincoln.

3.2

648

2.8

365

557

4.3

Phillipsburg. . .

2.6

475

1.8

3.5

211

314

4.5

Burlington....

1.6

419

1.6

2.5

277

172

2.5

Barley:

Lincoln.

3.1

384

3.7

3.5

306

622

4.6

Phillipsburg. . .

2.8

253

2.2

3.25

201

369

6.0

Burlington....

1.3

255

1.6

197

122

2.5

On August 17, 12 quadrats cut in the high prairie at Lincoln yielded an
average of 581 gm., and a similar number from the low prairie 929 gm., giving
an average yield of 755 gm. This considerably exceeded the average weight
of 11 tail-grass quadrats cut on August 30 at Phillipsburg, which was 477 gm.
The average of the 44 quadrats from both cuttings at each station was as
follows: Burlington 353 gm., Phillipsburg 402 gm., and Lincoln 603 gm. per
square meter.

Crops of the smaller cereals were again grown in 1921 under conditions
similar to those already described for 1920.

There was a regular decrease in height of the crop from the more humid to
the more arid stations. The same general relation holds for the average
number of stalks per square meter, except at Burlington, where many tiny
stalks only 2 to 4 inches tall started growth relatively early and soon dried out,
but remained until harvest. During 1920 the average number of stalks per
square meter at Burlington was from one-third to one-half less than at the
other stations, although the number at Phillipsburg often exceeded that at
Lincoln. The average number of stalks per plant (1921) was in direct relation
to the water-content of soil and other factors favorable or unfavorable to plant
growth. In general, this relation held also during 1920. The average number

GROWTH OF NATURAL COVER AND CROPS.

135

of heads per square meter and the average length of heads or panicles de¬
creased from Lincoln to Phillipsburg to Burlington respectively. An exception
to this occurred in the case of the number of heads of wheat at Burlington
when compared with Phillipsburg, while the difference in this respect in the
case of barley was small. However, a clear gradation in the reduction of total
dry weight from east to west is apparent, giving a direct correlation with
differences in water-content and humidity.

Results for 1922.

A single series of cut quadrats was obtained in 1922; these were taken on
August 4 to 15, beginning with the short-grass plains. A series of 7 quadrats
of Bulbilis at Burlington yielded an average of 179 gm., while the average of a
similar series at Phillipsburg was 260 gm. A series of mixed short and tall
grass cuttings at these two stations yielded 263 and 365 gm. respectively. The
differences in the height-growth of foliage and flower-stalks for both kinds of
grasses were again like those recorded for the preceding years. A series of 8
representative quadrats of wheat-grass was also obtained at both Phillips¬
burg and Lincoln. The average height of the foliage and flower-stalks re¬
spectively was 15 and 26 inches at the former station, and 17 and 30 inches at
the latter. The respective yields were 334 and 541 gm. per square meter. A
series of 27 quadrats containing mixed tail-grasses, largely Andropogon
scoparius, A. furcatus, and A. nutans, was secured from these two stations.
Repeated measurements showed that the average height of the general level
of the foliage was 3 to 8 inches greater at Lincoln (13 to 24 inches tall) than
in the mixed prairie. The yield in grams per square meter was 287 and 413
at the two stations respectively. The average yield of all the quadrats for
this year in the three grassland communities was 224, 311, and 447 gm. at
Burlington, Phillipsburg, and Lincoln respectively.

Table 53. —Growth and yield of crop-quadrats, 1923.

Crop and station.

Date of
harvest.

Height.

Dry'
weight. 1

Oats:

in.

gm.

Lincoln.

July

2,637.0

Phillipsburg.. .

July

905.0

Burlington....

July

806.5

Wheat:

Lincoln.

July

2,395.0

Phillipsburg. . .

July

1,296.0

Burlington....

July

801.5

Barley:

Lincoln.

July

1,671.5

Phillipsburg.. .

July

1,292.5

Burlington....

July

1,038.5

1 Total of straw and grain.

Five fields of maize were selected as representative of conditions about each
of the three stations respectively, all of which were within a radius of 2 miles of
the stations at which holard and other factor determinations were made. 10
stalks were then selected from various places in the field and the measure¬
ments recorded. The seventh leaf was selected for measurements of length and
width. After completing the measurements on each stalk, the ear was husked

136

PHYTOMETRIC RESULTS.

and placed in a sack, 10 ears being secured from each field. Finally, a typical
stalk, as determined by the measurements of the 10 preceding, was selected,
cut off at the ground line, cut into pieces, and after removing the ear but not
the husks, was placed in another sack. These sacks were kept well ventilated,
and after all had become thoroughly air-dried the weight of their contents
was ascertained. Consistent and marked increases in height and diameter of
stalk, height of ear, number and length of leaves, length and diameter of ear,
as well as dry weight of stalks and ears, were found proceeding from west to
east. The respective values for average weight of stalk and weight of ear were
320 and 82 gm. for Burlington, 413 and 110 gm. for Phillipsburg, and 496 and
183 gm. for Lincoln.

Results for 1923.

No clip-quadrats were taken from native vegetation in 1923, but three were
made in plats of each of the grains at the three stations.

As in previous years, there was a pronounced decrease in the height of all
three species from Lincoln westward, the plants at Burlington averaging
about half as tall as at Lincoln. With respect to weight, the differences were
even greater, oats and wheat yielding a third as much as at Lincoln. The
average production for the three crops was 2,234 gm. at Lincoln, 1,164 gm.
at Phillipsburg, and 882 gm. at Burlington, in close correspondence with rain¬
fall, evaporation, and chresard.

Table 54 .—Average yield of clip-quadrats in grams.

1920.

1921.

1922.

Dominant type
of vegetation.

Lin¬

coln.

Phil¬

lips¬

burg.

Bur¬

ling¬

ton.

Lin¬

coln.

Phil¬

lips¬

burg.

Bur¬

ling¬

ton.

Lin¬

coln.

Phil¬

lips¬

burg.

Bur¬

ling¬

ton.

Buffalo-grass.

290

235

266

207

260

179

Wheat-grass.

• • •

541

500

606

457

400

541

334

• • •

Mixed short and tall

grasses.

• • •

313

197

• . •

• ■ •

• • •

365

263

Mixed tail-grasses.

458

410

755

477

413

287

Average (based on
number of quad-

rats).

458

378

183

603

402

353

447

311

224

Summary.

In every case each grass or mixture of grasses yielded progressively less as
the rainfall and holard decreased to the westward. The only exception is the
case of buffalo-grass at Lincoln in 1921, and this has already been explained as
an effect of grazing. Moreover, the averages for each year at the several
stations show a graduated series, plant production increasing with increased
efficiency of rainfall. However, it may be readily seen that the total yield
at all of the stations was greater in 1921 than during the preceding or following
year, an increase particularly noticeable in the case of the late-maturing tail-
grasses.

It was early determined that the water relations of soil and air were con¬
trolling, other factors being merely contributory. The yield of pure stands of

GROWTH OF NATURAL COVER AND CROPS.

137

short-grasses (Bulbilis dadyloides and Bouteloua gracilis), wheat-grass
(Agropyrum glaucum), mixed short and tall grasses, and mixed tail-grasses
was found to decrease from the true prairie to short-grass plains directly with
the chresard and inversely with evaporation. The same relation was de¬
termined not only for the smaller cereals (oats, wheat, and barley), but also
for alfalfa and sweet clover, as well as for maize. The plant yield at each
station during different seasons was closely correlated with the variations in
rainfall and holard. Deficiencies in water-content were naturally most marked
late in the summer, and their effect was especially to be found in the late-
maturing tail-grasses at the eastern stations. The results obtained with
crop-plants were comparable in practically every respect, demonstrating that
the method was equally applicable to both native and cultivated species.

Structural Response of Dominants and Subdominants.

Scope and Significance.

Comprehensive studies have been made of the leaf-structure of the great
majority of the climax dominants of the grassland, as well as of a considerable
number of serai dominants, subdominants, and trees. The material was
obtained from both native and experimental plants, and in the case of the most
important species represented the whole range of adjustment and adaptation,
from salt-flat and swamp through low and high prairie to gravel-knoll at
Lincoln, and from Nebraska City to Lincoln, Phillipsburg, Burlington, and
Colorado Springs. As was to be expected, the grasses in general were very
stable as to leaf-structure, the adjustment to the several climates occurring
chiefly in function and growth, and hence in form. As a rule, distinct ecads
were produced only in the series of edaphic stations with their more efficient
differences, and the various species and genera naturally responded in very
different degree even to these. Moreover, the effect of a particular habitat
tended to be cumulative, and greater changes will probably appear in the
course of the next few years. This is suggested by the existing adaptations
shown by the leaves to the respective climates and recorded in the growth-
forms and in the genera. With occasional exceptions, the leaf-structure is
essentially the same for the dominants of each association, but differs markedly
from one association to another. This seems to be a more recent adaptation
that has been impressed upon the earlier one characteristic of each genus,
with the consequence that it is necessary to distinguish three successive
changes of varying degree in many dominants. Furthermore, this adapta-
tional series affords a new approach to climatic shifts in the past and to the
basic correlation of ungulate teeth and hoofs with changing vegetation.
For these reasons, as well as because of the large amount of material available,
it has seemed desirable to treat this phase of the investigation in a separate
paper (Clements, 1924).

6. RESUME.

PLAN AND METHODS.

An endeavor has been made to determine experimentally the factors
operating in the composition and sequence of the climax grassland associations
lying between the Missouri River and the Rocky Mountains. Lincoln,
Nebraska, was selected as representative of true prairie, Phillipsburg, in
north-central Kansas, of mixed prairie, and Burlington, in eastern Colorado,
of short-grass plains. The altitude rises from 1,100 feet at Lincoln to 1,900
feet at Phillipsburg and 4,160 feet at Burlington. Since the associations are
zoned from east to west, precipitation is the chief factor in determining the
type of grassland; it decreases through this series of stations from 28 through
23 to 17 inches annual mean. A fourth station was maintained in the edge of
the subclimax prairie at Nebraska City, 50 miles southeast of Lincoln, where
the precipitation is 33 inches. In addition to these climatic stations, a series
of edaphic ones ranging through gravel-knoll, high prairie, low prairie, salt-flat,
swamp, and cultivated field was maintained at Lincoln. Moreover, some
reciprocal transplants from Colorado Springs, Colorado; Tucson, Arizona;
and Berkeley, California, were made at several of these stations.

Continuous records of the most important ecological factors, viz, precipita¬
tion, temperature of air and soil, humidity, and evaporation, were secured and
frequent measurements were made of the holard to a depth of 4 feet, of light,
and wind velocity, in addition to physical and chemical analyses of the soils.
In the endeavor to interpret the various climatic and edaphic complexes in
terms of plant activities, seeding and transplanting were carried out in
different ways in the several communities. Seeds were sown in the undis¬
turbed grass cover, others were placed in especially prepared trenches, while
still others were sown in denuded quadrats. Seedlings were transplanted
and reestablished under favorable conditions, and finally, large blocks of sod
of mature plants were also employed. Some individuals were favored by
watering, while others were exposed unaided to the test of the new habitat.
Frequent checks were applied, resulting in complete records of the plant’s
activities. The investigation was begun in 1919 and continued until the close
of the growing-season in 1923.

Fruits and seeds of a large variety of native grasses, forbs, shrubs, and trees
from a wide range of habitats were employed. These were kept dry, but
subjected to winter temperatures at Lincoln, Nebraska, and the germinability
of the seeds was determined previous to planting, more than 60 species being
tested. Great variability in the vitality of the seeds from year to year was
found, and this could be traced to differences in climate. A germination of
20 to 25 per cent was exceptional, while one of 10 to 15 per cent was quite
usual. Experiments were also made to disclose the effect of depth of planting
upon germination and establishment. With most grasses a depth exceeding
0.5 to 1 inch was detrimental, and most species, like the composites, did best
at a depth of 0.12 to 0.25 inch.

Surface sowing consisted of scattering the seeds in selected marked areas
with a typical cover of vegetation, without disturbing the surface, and adding
only enough debris to prevent the seeds from blowing away. Upon germina¬
tion the seedlings at once met the keen competition from the existing vegeta-

138

PLAN AND METHODS.

139

tion for water and nutrients below ground and for light above. Frequently
the latter was as low as 5 to 10 per cent, even early in the summer. By
planting the seeds in a trench 4 inches deep, for which the sod had been broken,
the soil pulverized, and a good seed-bed prepared, competition both above and
below ground was removed for a short time. Denuded quadrats 0.5 meter
square were also employed, the native vegetation likewise being removed and
a good seed-bed prepared to a depth of 4 inches. This method eliminated
competition for light and also for water and nutrients for a time, but other
unfavorable factors were introduced, such as higher temperatures and greater
evaporation. To tide the plant over the most critical period in its life, that of
ecesis, the method of seedling transplants was also employed. These were
grown in 2.5 to 4 inch flower pots until they were 3 or 4 weeks old, when they
were transplanted into a specially prepared trench without disturbing the
root system and watered for a period of about 10 days. A fourth method, that
of transplanting mature perennials, was also utilized. Blocks of sod 10 inches
square and 8 inches deep were transplanted reciprocally among most of the
stations, a control block being replanted in the area where the species was
secured. As far as possible, sowing and transplanting were done preceding or
coincident with the inception of new growth and at a time when the holard
was distinctly favorable.

COMMUNITIES.

The vegetation at each station has been studied and described in detail, in
connection with the measurement of physical factors. Ninety per cent of the
cover at Burlington ( Bouteloua-Bulbilis association) consists of closed mats of
short-grasses (mostly Bulbilis) forming a carpet seldom over 4 inches deep.
At Phillipsburg ( Bouteloua-Stipa association) tail-grasses, chiefly Andropogon,
alternate with or form a layer above the shorter buffalo and grama grasses,
the former reaching a height-level of 14 inches by midsummer. In the true
prairie ( Stipa-Koeleria association) the vegetation is distinctly of the tail-
grass sod type. The chief genera are Stipa, Koeleria, Andropogon , and
Sporobolus, with which are associated subdominants to form extensive
societies. By June 1 the grasses have a height-level of 6 or 8 inches and the
upper story of forbs attains a height of 15 to 22 inches. The subclimax prairie
at Nebraska City (. Andropogon associes), in addition to the true-prairie
species which grow somewhat more rank, is characterized by the regular
occurrence of Andropogon furcaius and Panicum virgatum on highland, while
the various potential scrub societies usually held in check by mowing and fire,
indicate its t rue relationship to forest. The low prairie at Lincoln is dominated,
almost to the exclusion of other species, by Andropogon furcatus and nutans,
Panicum virgatum , and Spartina cynosuroides, and is actually an edaphic
postclimax (Plant Succession, 109), closely resembling the subclimax grass¬
land. The swamp area is covered by the Spartina consocies bordered by a
zone of Poa pratensis , but the bluegrass has been replaced by Spartina during
the course of the experiment. The vegetation both here and in low prairie
reaches a height of 4 or 5 feet. The salt-flat, on an area near the low prairie,
had just enough sodium chloride, combined with an unfavorable soil structure,
to exclude most prairie species, and is occupied by an open growth of Dis-
tichlis spicata and dwarf Agropyrum glaucum , seldom exceeding 5 to 8 inches

140 RESUME.

in height. The gravel-knoll is covered with mats of Bouteloua gracilis inter¬
mixed with Bouteloua hirsuta and often also with sparse tail-grasses in the
intervals. These grow in porous sandy to gravelly drift soil, which forms the
crest of a steep slope to the south. It is a miniature short-grass area of edaphic
nature surrounded by true prairie.

PHYSICAL FACTORS FOR 1920.

A comparison of the physical factors during 1920 at the three major climatic
stations shows that conditions for plant-growth in respect to rainfall, holard,
temperature, humidity, wind, and evaporation were most favorable at Lincoln,
intermediate at Phillipsburg, and least favorable at Burlington. These condi¬
tions are indicated by the native vegetation and are borne out by results of
these experiments. Holard and humidity were found to be the controlling
factors in plant growth, all others being secondary. The precipitation is not
only 5 inches less at Phillipsburg and 11 inches less at Burlington than at
Lincoln, but, owing to lighter showers, greater run-off, and increased evapora¬
tion, it is also progressively less efficient westward. While at Lincoln sufficient
chresard was found at all times and at all depths to 4 feet for good growth,
a period of midsummer drought occurred at Phillipsburg, and at Burlington
the holard was favorable only until June (to a depth of 2 feet only), after
which marked deficiencies were of frequent occurrence. Evaporation was
lowest at Lincoln (9 to 25 c. c. daily), intermediate at Phillipsburg (11 to 32
c. c.), and highest at Burlington (23 to 60 c. c.).

ECESIS DURING 1920.

Both germination and establishment at the three stations were in the
sequence of increasing water-content. Germination by all methods of planting
averaged 86, 80, and 38 per cent, and establishment 42, 33, and 25 per cent
at Lincoln, Phillipsburg, and Burlington respectively. Surface sowing gave
both the poorest germination and poorest establishment, and seeding in
denuded quadrats the best. In most cases better growth occurred at Lincoln
than at the other two stations. Germination and growth in cultivated areas
adjoining the several grassland stations were much better than in the sodded
areas, but the sequence of the stations was the same.

In order to more fully understand the causes for the success or failure of
seedlings, the root habit was studied both in native grassland and in cultivated
soil. During 1918, when seedling Bouteloua hirsuta, Andropogon scoparius,
furcatus, and nutans, and Sporoholus asper had reached heights of 6 to 10
inches in the grassland at Lincoln, the bulk of the root systems was found to
occur at no greater depths than 8 to 18 inches. In cultivated areas seedlings of
Bouteloua, Sporoholus, and Liatris, for example, reached heights of 3 or 4
inches in 44 days and the roots reached depths of 7 to 11 inches. By mid¬
summer they were 8 to 18 inches tall, and the root depth 20 to 33 inches.
The excellent growth made by all the species when properly spaced in fertile,
cultivated soil with adequate holard emphasizes the keen competition pre¬
vailing in stabilized grassland. In the natural cover, as well as in the trench
and denuded quadrat, most of the grasses require 2 or more years to produce
seed, but under the former environment 73 per cent bore seed the first season.
Because of the lack of root competition encountered by plants in the prairie,

RESULTS FOR 1920 AND 1921.

141

the species in cultivated soil not only suffered less from drought, but also made
a greater growth and tillered more heavily. For example, Andropogon nutans
was 7 to 10 inches tall at the end of the first season in the true-prairie quadrats
and 12 to 17 inches in the adjacent cultivated field. The tap-root of Gleditsia
reached a depth of 40 inches by midsummer and that of Onagra 44 inches at
the end of the first season, while Andropogon scoparius had a maximum lateral
spread of nearly 1.5 feet, was rooted abundantly to 26 inches depth, and had a
maximum penetration of 4 feet. These are typical examples of many findings.
In the deep, mellow loess soil at Peru, Nebraska, the root penetration of
seedlings was even greater. Of the 40 blocks of sod (9 species) transplanted at
Burlington no species died, but only 53 per cent (unwatered sods) had flower-
stalks, as compared with 70 per cent at Lincoln, where none of the 13 trans¬
planted species died (23 blocks).

In the edaphic series, the low prairie exhibited a chresard 5 to 10 per cent
greater than that of high prairie, and it also exceeded that at Nebraska City,
where unusually dry weather prevailed. However, the humidity was higher
at Nebraska City and the evaporation correspondingly less than on low prairie,
which was more mesophytic than high prairie. The holard on the gravel-
knoll and at Burlington was not very different, frequently falling to the
hygroscopic coefficient, the short-grasses on both areas drying and turning
brown late in July or in August. The average germination was highest at
Nebraska City (93 per cent), intermediate on low prairie (77; 86 per cent
on high prairie), and least on the gravel-knoll (71 per cent). The average
of establishment was greatest in the cultivated area at Lincoln (100 per
cent), 75 per cent on low prairie, 70 per cent at Nebraska City, and only 15
per cent on the gravel-knoll. Thus, all stations except the last exceeded high
prairie (42), while Burlington with 25 per cent exceeded the gravel-knoll.

As to sod transplants on the gravel-knoll, all showed repeated rolling and
drying back of the leaves; only a few flowered, and these did so earlier than
elsewhere. On low prairie all grew well and blossomed. The transplants
suffered from drought in the salt-flats, exhibiting wilting and yellowing of the
leaves and dwarfed stature. Seven species, mostly from upland, succumbed to
insufficient aeration in the swamp, and four in the Poa zone. All survived at
Colorado Springs, where they were transplanted late, but none made a good
growth. In the case of Spartina cynosuroides transplanted from swamp to
high prairie two years earlier, many leaves reached a height of 40 inches and
the roots were abundant to 9 or 10 feet depth, being much more branched than
in lowland. Panicum virgatum showed a similar growth.

PHYSICAL FACTORS FOR 1921.

The season of 1921 was very favorable for growth, no marked drought
periods occurring at Lincoln, where at least 5 and usually 8 or 10 per cent of
chresard existed at all times and at all depths to 4 feet. At Phillipsburg the
echard was approached once in July and twice in August, no water being
available to a depth of 4 feet in late summer. As usual conditions were much
worse at Burlington. At no time was water available in the third or fourth
foot of soil, while after June 30 it was depleted repeatedly above the hard-pan
at a depth of* 2 feet. The humidity was highest at Lincoln and lowest at
Burlington, where it often dropped to 10 or 20 per cent in the late afternoon.

142 RESUME.

Evaporation increased inversely with humidity and soil moisture, ranging
from 8 to 27 c. c. daily at Lincoln, 8 to 43 c. c. at Phillipsburg, and 18 to 62
c. c. at Burlington. The remaining conditions for plant growth were also most
favorable at Lincoln, intermediate at Phillipsburg, and poorest at Burlington.

ECESIS DURING 1921.

The germination results agreed with those of preceding years, Lincoln being
highest (81 per cent), Phillipsburg second (68 per cent), and Burlington last
(43 per cent). However, owing to the unusually favorable rainfall following
germination at Phillipsburg, establishment was highest here (65 per cent),
intermediate at Lincoln (40 per cent), and least at Burlington (7 per cent).
No seedling transplants survived at Burlington, but 54 per cent lived at
Lincoln, and 75 per cent at Phillipsburg. Although the surface-sown seeds
germinated better than those in the trench, establishment, as in 1920, was
least on the surface, next in the trench, and best in denuded quadrats, the
latter exceeding that of seedling transplants. As to the 1920 plantings, the
average survival during 1921 was greatest at Lincoln and least at Burlington.
An examination of the grasses and forbs during the second year in the quadrats
at Phillipsburg showed that they were well established, the roots penetrating
to depths of 2.5 to 4 feet.

Sods on the high prairie all did well, but 4 species died at Burlington, and
the rest, including those that had been repeatedly watered, were represented
by mere remnants of the original fine blocks. Of those transplanted to high
prairie in 1920, all survived and some increased their area, but at Burlington 1
died, while all suffered from drought and were considerably dwarfed. By mid¬
summer of the second season's growth in the short-grass plains, the roots were
confined to the surface 2 to 2.5 feet, except in the watered area, where they
penetrated 5 to 7 feet.

As to the other stations, the season at Nebraska City was one of drought,
though the soil usually had a 10 per cent chresard at all depths to 4 feet.
This exceeded the water-content on low prairie, where the chresard in the
surface 6 inches was nearly exhausted in June. Conditions on the gravel-knoll
were more favorable than during 1920. Evaporation at Nebraska City and
low prairie was lower than on the high prairie and was highest on the gravel-
knoll. The average germination at the three stations was very similar (63
to 76 per cent) and in all cases less than on high prairie (81 per cent). The
rate of establishment was in the following order: Nebraska City 41 per cent,
high prairie 40 per cent, low prairie 38 per cent, and gravel-knoll 29 per cent,
the last considerably exceeding Burlington (7 per cent). Seedling transplants
gave the best growth on low prairie (92 per cent), next at Nebraska City (57
per cent), then high prairie (54 per cent), and least in gravel-knoll (14 per
cent). Four species of sod transplants died on the gravel-knoll, and three
were shaded out in low prairie, while all survived on the salt-flat and in the
swamp, the latter being much drier than in 1920. Of the 1920 transplants,
no species died on the gravel-knoll, one succumbed to the shade in low prairie,
one died on the salt-flat, and 6 in the swamp and bluegrass zone.

PHYSICAL FACTORS FOR 1922.

The season of 1922 was fairly favorable for growth, except for the latter
part, when severe drought occurred at all the stations, though it was relatively

RESULTS FOR 1922.

143

less marked on the Great Plains. A margin of at least 5 per cent and more,
usually 7 to 11 per cent, chresard existed at all times to a depth of 4 feet in the
high prairie at Lincoln, while at the mixed-prairie station drought began late
in June and continued throughout the season. There was often no chresard
to a depth of 4 feet. The soil in spring and early summer was as usual quite
moist at Burlington, but deficiencies were marked and practically continuous
after the middle of June. As usual, the factors were most favorable in true
prairie and least so on the Great Plains.

ECESIS DURING 1922.

As for the two years preceding, the average germination under all methods
of planting was highest at Lincoln (74 per cent), intermediate at Phillipsburg
(63 per cent), and least at Burlington (44 per cent). In surface sowing Phil¬
lipsburg slightly outranked Lincoln, due to differences in soil structure. This
also accounts for the greater survival of seedling transplants at the former,
the percentages being 78, 70, and 26 respectively. Establishment was
decidedly highest in true prairie, viz, 62 per cent; it was but 39 per cent in
mixed prairie, and 26 per cent in the short-grass plains. Germination aver¬
aged best on the surface, next in denuded quadrats, and least in the trench,
but establishment was in the following order: denuded quadrats, trench, and
surface, the last giving the lowest percentage. Bouteloua gracilis made the
best growth in the quadrats at Burlington and the least at Lincoln, but for all
of the other species the growth-rate was in the sequence of increasing water-
content, i. e., eastward. Andropogon nutans, Kuhnia glutinosa, Bouteloua
racemosa , and Panicum virgatum were 4 to 12 inches tall at Lincoln, 3 to 9
inches at Phillipsburg, and 1 to 5 inches at Burlington.

The survival for the 1920 surface-sowings was 33 per cent at Burlington
only; for the trench it was 33 per cent at Lincoln and Phillipsburg, and
for the quadrats 55 per cent at Lincoln and 43 per cent at Burlington, the
plants at Phillipsburg having been mostly excavated for root study. The
average survival of 1921 plantings and seedling transplants was greatest at
Phillipsburg and lowest at Burlington. Two species died out of the 26 trans¬
planted to high prairie in 1922; these were both dicotyls with strong tap¬
roots. All of the 13 species of grasses transplanted to Burlington survived,
although they flowered less than at Lincoln and nearly all were dwarfed. Of
the 1920 sod transplants, 2 died on high prairie during 1922, but only 1 at
Burlington, although many more individuals out of the 40 transplanted
blocks of sod succumbed here. None of the 1921 transplants died on the
high prairie; several increased their territory, and 12 of the 16 species blos¬
somed. At Burlington several individuals died, but only 1 species vanished
entirely. Agropyrum spicatum and Bouteloua racemosa were the sole survivors
of the unwatered lot.

As to the other stations, the season at Nebraska City was one marked by
drought in June and August. However, the soil usually had a chresard of 8
per cent below the first foot, but that in the surface foot, which most critically
affects seedlings, was practically exhausted two or three times during the
summer, being less favorable than in 1921. Conditions as to the holard were
more favorable on low prairie, where a considerable excess prevailed over that
on high prairie. On the gravel-knoll the grama grasses dried during the

144 RESUME.

drought, as no moisture was available in the first 2 feet of soil. The germina¬
tion of surface-sown seed was not greatly different at the three stations
(70 to 84 per cent), this method of sowing giving the best germination. The
average germination was 67 per cent at Nebraska City, 75 on low and 74 on
high prairie. Establishment gave the usual sequence with respect to method
of planting. It averaged 46 per cent at Nebraska City, 62 per cent on high
prairie, and only 50 per cent on the more densely shaded low prairie, where
light values ranged from 2 to 10 per cent. Seedlings did best on high prairie
(70 per cent) where they were watered, next at Nebraska City (59 per cent),
and poorest on low prairie (45 per cent). Three of the sod transplants on low
prairie succumbed to the dense shade, as did 4 in the swamp, while one died on
the salt-flat. Of the 1920 sods, 3 died on the gravel-knoll, 4 were shaded out
in low prairie, and 1 died in the salt-flat, but no further mortality occurred in
the rapidly drying swamp. Of the 1921 transplants, 3 more died on the
gravel-knoll, 5 in low prairie, 1 in the salt-flat, and 2 in the swamp.

BEHAVIOR DURING 1923.

In 1923, a season with high spring and summer rainfall at the western
stations, the plants made an excellent growth. Not only was the rainfall above
the normal, but the rains were also well distributed, drought periods being
fewer and shorter than usual. At Lincoln the precipitation for spring and
summer averaged below normal. The preceding fall and winter had been one
of unusually severe drought, and high losses were sustained at all the stations.
In the major climatic stations these were greatest at Burlington, intermediate
at Phillipsburg, and least at Lincoln. For example, losses of the 1921 plant¬
ings and transplants were 43, 28, and 13 per cent in the above sequence of
stations. The mortality at all stations averaged greatest for the plants of a
single year’s establishment, next for those 2 years old, and least for those
established for 3 seasons, e. g., the respective losses at Burlington being 85,
43, and 16 per cent. The survival was greatest in denuded quadrats, this
method even outranking that of transplanting blocks of sod.

In the survival of surface-sown plants for all years Phillipsburg ranked first
(14 per cent), Burlington second (11 per cent), and Lincoln last (9 per cent).
The average germination was respectively 79, 45, and 76 per cent, Lincoln
taking second and Burlington third place. The average germination in
trenches decreased from Lincoln westward, the percentages being 85, 59, and
21 respectively. The average survival for all years at the end of 1923 was in
the same order, the percentages being 27, 21, and 0 respectively. Germination
in denuded quadrats was greatest at Lincoln for every year except one, and
least at Burlington, with one exception, the percentages for the three stations
being 81, 73, and 60. With a single exception, establishment was likewise in
the same order, viz, 72 per cent in true prairie, 60 per cent in mixed prairie,
and 43 per cent in the short-grass plains, and this was also true of survival at
the end of 1923, the percentages being 44, 22, and 21 respectively. Hence,
while surface-sowing ranked highest in germination, it resulted in the lowest
estabishment.

Transplanted seedlings usually did best at Phillipsburg, the average survival
at the end of 1923 being 44 per cent as compared with 37 per cent at Lincoln
and 3 per cent at Burlington.

RESULTS FOR 1923.

145

Permanent establishment by the end of 1923 under the three methods of
planting was 44, 27, and 9 per cent respectively at Lincoln, 22, 21, and 14 per
cent at Phillipsburg, and 21,0, and 11 per cent at Burlington. This was always
greatest in the denuded quadrats and next greatest in the trench.

The results of seeding and transplanting at Nebraska City and on low prairie
are not greatly different, although often somewhat better than on high
prairie, while ecesis on the gravel-knoll was often only slightly higher than that
at Burlington. At the latter, as on the gravel-knoll, drought was the deciding
factor, while on low prairie, as well as at Nebraska City, light played the
dominant role.

Certain species were found to survive more frequently than others. Out
of the total plantings at Nebraska City, Lincoln, Phillipsburg, and Burlington,
beginning with 1920, Andropogon nutans survived in 35 different places by
the end of 1923. The survival of other species was in the following order, the
figures indicating the different number of places in which they survived :
Bouteloua gracilis 18, B. racemosa 15, B. hirsuta 14, Sporobolus asper 11,
Aristida purpurea, Elymus canadensis, and Liatris punctata 10 each, Andro-
pogon scoparius 8, A. furcaius 7, Petalostemon candidus 6, Stipa spartea,
Desmodium canescens, and Liatris scariosa 5 each, Kuhnia glutinosa 4, Panicum
virgatum, Stipa viridula, Symphoricarpus occidentalis 3 each, Gleditsia
triacanthus, Symphoricarpus vulgaris, Koeleria cristata, Ratibida columnaris,
Robinia pseudacacia, and Acer saccharinum 2 each, and Stipa comata,
Agropyrum glaucum, Pinus ponderosa, and Acer negundo each only 1.

As regards sod transplants, by the fall of 1923 over half of the 1920 plantings
at Burlington had died. Agropyrym glaucum alone remained in the un¬
watered area. The 1921 transplants, half of which were watered, lost 15 out
of 22 individuals, Agropyrum and Bouteloua racemosa alone surviving in the
unwatered area, while practically all of the rest showed clearly that they
would not last much longer. 83 per cent of the 1922 transplants were winter-
killed, and at the end of the second season Bouteloua gracilis and remnants of
2 other species alone remained. These results clearly reveal the uncongenial
nature of the Great Plains, in so far as prairie species are concerned.

Relatively few species died on the high prairie, even Spartina cynosuroides
and Panicum virgatum forming seed. Of the 1920 plantings only 6 individuals
had died, including 2 plants of Distichlis spicata, which never flourished out
of its owtl saline habitat. Of the 1921 plantings a single one succumbed. Six
forbs among the 1922 plantings died, and 1 lot of Bouteloua gracilis. Several
of the species had considerably extended their area.

On the gravel-knoll the mortality was much greater, although here likewise
several low-prairie species became permanently established, their roots ex¬
tending well into the clay subsoil below 4 feet. Three species of the 1920 and
7 of the 1921 lot succumbed, but others planted in 1919 still survived, Bulbilis
and Agropyrum being quite at home. Nearly all were somewhat dwarfed, and
they began growth and matured earlier than at the other Lincoln stations.

Exceedingly interesting results were obtained on low prairie. Of the 14
species and 28 blocks of sods transplanted in 1920, all flourished, including
Bulbilis and Bouteloua gracilis. By the second season, however, 1 species was
shaded out, 4 the next, and 1 in 1923, so that by the end of the fourth summer
the dominants of the low prairie alone, viz, Andropogon nutans and furcatus,

146 RESUME.

Panicum virgatum, Elymus canadensis, and Spartina cynosuroides were
flourishing, Stipa, a single Koeleria, and Andropogon scoparius surviving from
the high prairie. Of the 1921 transplants, 5 subclimax dominants developed
normally, but of the 10 species remaining all had succumbed but 3 represented
by mere remnants. Two or three years were sufficient for the tail-grasses to
overshade and cause Bulbilis and Bouteloua gracilis and hirsuta to disappear.
This applies also to several true-prairie dominants. Of the 1922 transplants a
total of 11 species had succumbed by the end of 1923, the process of elimina¬
tion being well under way.

On the salt-flats, while the plants were able in most cases to tolerate condi¬
tions (only 6 species among the three years’ plantings succumbing), they
made a very poor growth, did not extend their territory, and seeded very
poorly. For the most part they had a paler color and in every way reflected
the unfavorable environment, except the indigenous Distichlis alone.

Transplants in the swamp lost heavily, at first-from poor aeration and later
from lack of light, after a dam was built above the experimental area and the
swamp dried out. High-prairie species were most affected and died in greatest
numbers, the flower-stalk often rotting off at the surface of the saturated soil.
By 1923, Spartina, Panicum virgatum, and Elymus canadensis, all of which
grew well, were the only survivors among the 13 species planted in the swamp,
while these and Poa pratensis were the 4 survivors of a similar lot in the Poa
zone, which was now overgrown with Spartina.

During the course of the experiment it was found that species of widely
different families exhibit the phenomenon of dormancy, the seeds lying
in the soil for one or more years before germinating. Among the most con¬
spicuous cases the following may be enumerated: Robinia pseudacacia, Gledit-
sia triacanthus, Petalostemon candidus, and Lespedeza capitata among legumes;
Sporobolus asper, Andropogon nutans, Panicum virgatum, and Andropogon
furcatus among grasses; and likewise Kuhnia glutinosa, Corylus americana, and
Acer negundo. While the volunteer growth of these plants during the second
or third year has not been recorded in detail, because of the risk of confusion,
it may be stated that in general they suffered the same fate as the original
plantings.

BEHAVIOR OF TREES AND SHRUBS.

The fate of trees and shrubs planted or transplanted by rhizomes into grass¬
land is of especial interest. The trees employed were Acer negundo, A.
saccharinum, Gleditsia triacanthus, Robinia pseudacacia, and Pinus ponderosa,
together with the following shrubs: Corylus americana, Symphoricarpus occi-
dentalis, and S. vulgaris. At Burlington none survived longer than the first
season, and at Phillipsburg the only survivors at the end of 1923 were the 2
species of Symphoricarpus planted in 1921. Corylus (delayed growth),
Robinia, Gleditsia, and Acer saccharinum of the 1922 planting, and Robinia
of 1921 survived on high prairie. On low prairie Symphoricarpus occidentalis
of the 1921 planting and Acer saccharinum of 1922 alone survived. None lived
on the gravel-knoll, but at Nebraska City both species of Symphoricarpus of
1921 and Acer negundo and Pinus ponderosa of 1922 were alive at the end of the
summer of 1923. However, as shown by a series of earlier experiments with
trees at Lincoln, the results of which are as yet unpublished, the possibility of

CLIMATIC STATIONS.

147

tree-growth in stabilized grassland is almost nil, since tree seedlings are
killed by drought on upland and are shaded out on lowland.

GERMINATION AND SURVIVAL AT THE CLIMATIC STATIONS.

The percentage of germination for each year and for each method and the
percentage of survival for the first year of planting as well as succeeding years
are given in table 55. The percentage of germination under surface sowing
was greatest at Lincoln during 1920, and at Phillipsburg during 1921 and
1922, while it was least at Burlington for all three years. The averages for the
three years are Phillipsburg 79 per cent, Lincoln 76 per cent, and Burlington
45 per cent (table 56). In regard to establishment, Burlington ranked first in
1920 (clearly an erratic case), Phillipsburg in 1921, and Lincoln in 1922. As
to the survival of plants of all years, at the end of 1923 Phillipsburg was first
with 14 per cent, Burlington second, 11 per cent, and Lincoln third, 9 per cent.
The greater germination and establishment at Phillipsburg were probably due
to the structure of the soil (p. 12).

Germination in the trenches was highest at Phillipsburg in 1920, but at
Lincoln for the other years, Burlington coming last. The average germination
for the three seasons was greatest at Lincoln (85 per cent), next at Phillips¬
burg (59 per cent), and lowest at Burlington (21 per cent). Survival in 1920
and for 1923 was greatest at Lincoln, but in 1921 greatest at Phillipsburg,
Burlington regularly ranking third. However, in the survival for all plantings
at the end of 1923, the sequence is Lincoln, Phillipsburg, and Burlington, with
respective percentages of 27, 21, and 0.

Germination in the denuded quadrats was greatest at Lincoln during 1920
and 1922, and least at Burlington, except in 1921, when it exceeded Phillips¬
burg and Lincoln. The average germination for the three years at the several
stations was 81, 73, and 60 per cent, in the order of decreasing rainfall.
Establishment was also greatest at Lincoln every year and least at Burlington,
except during 1922, when the latter exceeded that at Phillipsburg. The
average establishment for the three years was 72 per cent in the true prairie,
60 per cent in mixed prairie, and 43 per cent in the short-grass plains. The
percentage of survival for all plantings in quadrats for the three years was, in
the above order, 44, 22, and 21.

Transplanted seedlings did best at Phillipsburg both during 1921 and 1922
and poorest at Burlington. Moreover, the average survival at the end of 1923
was greatest at Phillipsburg (44 per cent), next at Lincoln (36 per cent), and
least at Burlington (3 per cent). The explanation of the better survival in the
mixed prairie seems clearly to be sought in the more favorable soil structure.

With reference to the success of the various methods, surface sowing gave
the poorest results as to germination in 1920, but the best the other two
seasons, the quadrat ranking second (first in 1920), and the trench third. On
the other hand, establishment was always greatest in the denuded quadrat,
second in the trench, and least on the surface. The survival for all years at the
end of 1923 was greatest in denuded quadrats, and least on the surface.
At Lincoln under the three methods it was 44, 27, and 9 per cent, re¬
spectively, at Phillipsburg 22, 21, and 14 per cent, and at Burlington 21, 0,
and 11 per cent. The seedling-transplant method was somewhat less suc¬
cessful than sowing in quadrats, but more satisfactory than sowing in trenches.

148

RESUME.

The persistence of plants is high after surviving a single growing-season, but
the mortality is also usually greater during the second winter and summer
than for any year afterward.

During the three years the average germination for all methods was greatest
in true prairie (81 per cent), second in mixed prairie (70 per cent), and least
at Burlington (42 per cent). The average establishment was, in the same
sequence of stations, 48, 46, and 20 per cent, as was also the average survival
at the end of 1923, viz, 27, 19, and 11 per cent. Thus, the experimental
evidence clearly indicates that the best conditions for growth are to be found
in true prairie and the least favorable ones in the short-grass plains, conditions
in the mixed prairie being intermediate.

GERMINATION AND SURVIVAL AT THE EDAPHIC STATIONS.

Results from the series of edaphic stations show that the germination of
surface-sown seeds was very similar in the low prairie and at Nebraska City
during the three years, where it slightly exceeded that on the high prairie at
Lincoln. The survival at the end of the first season averaged greater at both

Table 55. —Per cent of germination and survival at all stations, 1920-1923.

Surface sowing.

Trench sowing.

Station.

P. ct.
ger¬
mina¬
tion.

P. ct. survival.

P. ct.

ger¬

mina¬

tion.

P. ct. survival.

1920.

1921.

1922.

1923.

1920.

1921.

1922.

1923.

1920.

Lincoln (high prai¬
rie) .

Phillipsburg.

. .

. ,

100

Burlington.

. .

1921.

Lincoln (high prai¬
rie) .

(9)

Phillipsburg.

100

• .

. ,

Burlington.

• .

• •

• .

• •

1922.

Lincoln (high prai¬
rie) .

Phillipsburg.

• .

• •

. .

• .

Burlington.

• •

. .

• •

. ,

. .

1920.

Gravel-knoll (Lin¬
coln) .

Low prairie (Lin¬
coln) .

(17)

100

Nebraska City. ...

(38)

100

1921.

Gravel-knoll (Lin¬
coln).

Low prairie (Lin¬
coln) .

Excavated
for roots.

Nebraska City. ...

, ,

1922.

Low prairie (Lin¬
coln).

Nebraska City. ...

• •

EDAPHIC STATIONS.

149

Table 55. —Per cent of germination and survival at all stations, 1920-1923 —Continued,

Denuded quadrats, sowing.

Seedling trans-

Station.

P. ct.
ger¬
mina¬
tion.

P. ct. survival.

plants, p. ct.
survival.

1920.

1921.

1922.

1923.

1921.

1922.

1923.

1920.

Lincoln (high prairie).

100

Phillipsburg.

Exca

vated

Burlington.

for r
43

oots.

1921.

Lincoln (high prairie).

Phillipsburg.

Burlington.

1922.

Lincoln (high prairie).

Phillipsburg.

Burlington.

1920.

Gravel-knoll (Lincoln).

Low prairie (Lincoln).

100

Nebraska City.

100

1921.

Gravel-knoll (Lincoln).

Low prairie (Lincoln).

Nebraska City.

1922.

Low prairie (Lincoln).

Nebraska City.

Note: Numbers in parenthesis indicate that the plants had merged into the native sod and
were not traced further.

stations (low prairie 24, Nebraska City 43 per cent) than on high prairie (20
per cent), while the average survival at the close of 1923 was respectively 8,
29, and 9 per cent.

The average germination in the trenches for all years was least on the gravel-
knoll, 63 per cent; it reached 64 per cent in low prairie and 71 per cent at
Nebraska City, as compared with 85 per cent on high prairie. However,
ecesis on the gravel-knoll was less than half that on high prairie (23 in contrast
to 51 per cent), while in low prairie and at Nebraska City it was 60 and 42 per
cent respectively. The survival at the end of 1923 was nil on the gravel-
knoll, 31 per cent in low prairie, and 17 per cent at Nebraska City, as com¬
pared with 27 per cent on high prairie.

Germination in denuded quadrats averaged 71, 85, and 79 per cent on gravel
knoll, low prairie and Nebraska City respectively, in contrast to 81 per cent on
high prairie. The corresponding percentages were 20, 78, and 71, as against
72 on high prairie, the survival at the end of 1923 being 10, 40, 49, and 44 per
cent respectively.

The ecesis of seedling transplants was 14 per cent on the gravel-knoll (1921
only), 68 per cent in low prairie, and 58 per cent at Nebraska City, as compared
with 62 per cent on high prairie. The respective survival at the end of 1923
was 14, 35, and 48 per cent, as against 36 per cent on high prairie. Thus,

150

RESUME.

Table 56.— Summary of percentages of germination and survival.

Surface Sowing.

Station.

Lincoln (high prairie).

Phillipsburg.

Burlington.

Low prairie (Lincoln).

Nebraska City.

Year of
plant¬
ing.

1920

1921

1922

1920

1921

1922

1920

1921

1922

1920

1921

1922

1920

1921

1922

P. ct.
germina¬
tion.

79
82
50

100

80
78
70

Aver¬

age.

Survival
at end of
first year.

Aver¬

age.

Survival
at end of
1923.

Aver¬

age.

Trench Sowing.

Lincoln (high prairie).

Phillipsburg.

Burlington.

Gravel-knoll (Lincoln),

Low prairie (Lincoln).

Nebraska City.

1920

1921

1922

1920

1921

1922

1920

1921

1922

1920

1921

1920

1921

1922

1920

1921

1922

70
100

71
55
57
75
60

100

Excav.

Denuded Quadrats.

Lincoln (high prairie).

Phillipsburg.

Burlington.

Gravel-knoll (Lincoln).
Low prairie (Lincoln).

Nebraska City.

1920

1921

1922

1920

1921

1922

1920

1921

1922

1920

1921

1920

1921

1922

1920

1921

1922

100

63
50
43
21

64
0

100

s 48 <

! Excav.

31
36
43

5
14
0

20
45
25
50
70
40
36

TRANSPLANTS.

151

Table 56 .—Summary of 'percentages of germination and survival —Continued.

Seedling Transplants.

Station.

Lincoln (high prairie)

Phillipsburg.

Burlington.

Gravel-knoll (Lincoln).. .
Low prairie (Lincoln)

Nebraska City.

Year of
plant¬
ing.

P. ct.
germina¬
tion.

Aver¬

age.

Survival
at end of
first year.

Aver¬

age.

14
68

Survival
at end of
1923.

Aver¬

age.

no consistent differences are to be found between the low-prairie station
at Lincoln, which is really a postclimax stage in the true-prairie climate, and
the subclimax prairie at Nebraska City, both averaging somewhat better than
the high-prairie station at Lincoln. The gravel-knoll station, climatically a
preclimax to the true prairie, excelled Burlington in percentage of germina¬
tion and establishment, but barely equaled it in survival at the end of 1923.

SURVIVAL OF SOD TRANSPLANTS.

While the number of species lost at Burlington from the 1920 transplanting
was the same as that at Lincoln, the individual losses were far greater, owing
to the transplanting of a larger number of blocks of each species at the plains
station. The relative losses for 1921 were representative. Losses in individ¬
uals on the gravel-knoll were consistently high, while those on low prairie
were similarly high, though the cause for the latter was not drought, but
shade. The losses in the swamp were greatest during 1920, because of poor
aeration due to an excessive amount of water, while later losses were caused
almost entirely by the density of shade. While in general the high-prairie
species were the ones to disappear from swamp and low prairie, no consistent
loss of individual species occurred at the other stations. Excluding the losses
of 1923, owing to the very dry fall and winter, the following species suffered
total loss in as many instances as are indicated by the number: Koeleria
cristata 14; Andropogon scoparius 7; Elymus canadensis, Poa pratensis,
Agropyrum glaucum, and Bouteloua racemosa 6 each; Distichlis spicata 5;
Stipa spartea and Andropogon nutans 4 each; Bouteloua gracilis, Panicum
virgatum, Andropogon furcatus 3 each; and Bouteloua hir^uta, Bulbilis dacty-
loides and Sporobolus asper 1 or 2 each.

PHYTOMETRIC METHODS.

In spite of the abnormal weather during 1923, the investigation of the
transpiration of standard plants and sod-cores of native and cultivated spe¬
cies, and of the growth of natural cover and crops, as shown by cut-quadrats,
yielded results of value in confirmation of those obtained by the methods of
experimental vegetation. This was especially true of the cut-quadrats, as

152 RESUME.

these also extended through a series of nearly normal years. They not only
demonstrated the close dependence of grassland and crop communities upon
rainfall and water-content, but also the intimate relationship of the various
prairie climaxes and climates. The evidence drawn from germination, ecesis,
function, and growth was corroborated in large measure by the adaptation
shown in leaf-structure, though this was more or less obscured by the fixity of
type in certain species.

Table 57 .—Summary of sod transplants.

Station.

No. of

Loss of

species (not individuals).

species.

trans¬

plants.

1920.

1921.

1922.

1923.

1920.

Lincoln (high prairie).

Burlington.

Gravel-knoll.

Low prairie.

Salt-flat.

Swamp.

Poa zone in swamp.

1921.

Lincoln (high prairie).

Burlington.

Gravel-knoll.

Low prairie.

Salt-flat.

Swamp.

Poa zone in swamp.

1922.

Lincoln (high prairie).

JmJ

Burlington.

Low prairie.

Salt-flat.

Swamp.

....

ATTAINMENT OF OBJECTIVES.

The conclusion seems to be warranted that the general objective of de¬
veloping the basic method of experimental vegetation so that it combines the
maximum of demonstrability and objectivity has been achieved. While the
methods involved require further refinement and testing, there can be little
question that they are prerequisite to fundamental and permanent progress
in the field of vegetation. The special objectives set have also been attained,
though necessarily in varying degree, since such problems as those of the cycle
and of competition will constantly unfold with advancing research. Both
planting and factor results have confirmed the essential variety within unity
that characterizes the grassland formation with its associations, and harmonize
completely with the evidence from cyclic changes and relicts. They have
further illuminated the relation of serai habitats and communities to those of
the climax, and have likewise justified the concepts of subclimax, postclimax,
and preclimax. The essentially subclimax nature of the low prairie has been
corroborated, as have also the climatic and phylogenetic relationships of the
subclimax prairie. The evaluation of competition as a controlling process in

OBJECTIVES.

153

grassland has been attended by unexpected success, especially in connection
with the persistence of subclimax prairie in regions of forest climate. With
respect to natural migration, the experimental evidence lends no support to
the assumption that single or scattered individuals regularly invade climaxes
with success, but on the contrary indicates that all such invasion is a mass
movement in response to wet and dry phases of climatic cycles. The
demonstration of the phyletic relationship between subclimax prairie, true
prairie, mixed prairie, and short-grass areas and between the corresponding
climates in the first three cases appears conclusive, and in conjunction with
relict and cyclic results will constitute the basis for future research in the
grassland formation.

154

EXPERIMENTAL VEGETATION.

BIBLIOGRAPHY.

Clements, F. E.

1897. Peculiar zonal formations of the Great Plains. Am. Nat. 31:968.

1904. Development and structure of vegetation. Rep. Bot. Surv. Nebr. 7.

1904 2 . Formation and succession herbaria. Univ. Nebr. Studies 4:329.

1905. Research methods in ecology.

1907. Plant physiology and ecology.

1916. Plant succession. Carnegie Inst. Wash. Pub. 242.

1920. Plant indicators. Ib., Pub. 290.

1921. Aeration and air-content. Ib., Pub. 315.

1922. The original grassland of Mohave and Colorado Deserts. Carnegie Inst.

Wash. Year Book 21:350.

The method and principle of relicts. Ib., 351.

Rainfall and climatic cycles. Ib . x 354.

1923 The original vegetation of Death Valley. Ib., 22:

Application of the relict method. Ib., 317.

Rainfall and climatic cycles. Ib., 318.

-and E. S. Clements. 1921. Changes in vegetation. Ib., 20:405. 1922

Changes in grassland. Ib., 21:348, 22:316.

-and G. W. Goldsmith. 1921-23. The phytometer method. Ib., 20-22.

-. 1924. The phytometer method in ecology.

-and H. M. Hall. 1919-23. Experimental taxonomy. Ib., 18-22.

-and D. C. Lutjeharms. 1921-22. Slope-exposure studies. Ib., 20-21.

-and J. E. Weaver. 1919-23. Transplant quadrats and areas. Ib., 18-23.

Hall, A. D. 1920. The soil. An introduction to the scientific study of the growth of
crops.

Lutjeharms, D. C. 1924. Environmental differences on north and south slopes of a
canyon.

Pool, R. J. 1914. A study of the vegetation of the sandhills of Nebraska. Minn. Bot.
Studies 4:189.

Pound, R., and F. E. Clements.

1898. The phytogeography of Nebraska.

1898 2 . The vegetation regions of the prairie province. Bot. Gaz. 25:381.

1900. The phytogeography of Nebraska. Second edition.

Shantz, H. L.

1906. A study of the mesa region east of Pike’s Peak: the Bouteloua formation.

Bot. Gaz. 42:16.

1911. Natural vegetation as an indicator of the capabilities of land for crop produc¬
tion. U. S. Dept. Agr., Bur. PI. Ind. Bull. 201.

Weaver, J. E.

1917. A study of the vegetation of southeastern Washington and adjacent Idaho.

Univ. Nebr. Studies 17:1.

1919. The ecological relations of roots. Carnegie Inst. Wash. Pub. 286.

1920. Root development in the grassland formation. Ib., 292.

1924. Plant-production as a measure of environment. Jour. Ecol. 12:205; Carne¬
gie Inst. Wash. Year Book 22:312.

-and J. W. Crist. 1922. Relation of hard-pan to root penetration in the Great

Plains. Ecology 3:237.

-. 1924. Direct measurement of water-loss from vegetation without

disturbing the normal structure of the soil. Ecology 5:153; Carnegie
Inst. Wash. Year Book 22:312.

-, F. C. Jean, and J. W. Crist. 1922. Development and activities of roots of crop

plants. Carnegie Inst. Wash. Pub. 316.

- , Joseph Kramer, and Maud Reed. 1924. Development of root and shoot

of winter wheat under field conditions. Ecology 5:26-20.

-and A. F. Thiel. 1917. Ecological studies in the tension zone between prairie

and woodland. Rep. Bot. Surv. Nebr. 1917.

BEHAVIOR TABLES

155

t-H

Table 58 .—Sowing in denuded quadrats, high prairie, Lincoln, Apr. 16 and May 1920. —Continued.

156

EXPERIMENTAL VEGETATION

Table 59. — Sowing in denuded quadrats, Phillipsburg, May 7, 1920.

BEHAVIOR TABLES

157

1 During 1922 Andropogon nutans grew well but did not blossom, while in 1923 it made an excellent growth and flowered abundantly.

158 EXPERIMENTAL VEGETATION.

1. Andropogon nutans. 2 green shoots only, 15', little invasion. 2 small clumps, 8-15', will blossom.

9. Liatris punctata. 2 fine stalks, 12', thriving. Now 16', will flower abundantly.

13. Stipa viridula. Good, many well-tillered clumps, 12-15'. 12-18', thriving, very little invasion.

Table 61. —Seedlings transplanted to high prairie, Lincoln, May 11, 1921.

BEHAVIOR TABLES

159

Table 62. — Seedlings transplanted at Phillipsburg, May 19, 1921.

160

EXPERIMENTAL VEGETATION

Table 63 .—Seedlings transplanted at Burlington, May 20 , 1921 .

BEHAVIOR TABLES

tab

•tt

o 3

T 3

-T 3

-O

T 3

. >5

t3t:

2 oi
0 ) o)

P£

bll

05~

-O *

• •£>

o3 o3 c3 c3 c

0 > 0 <V <D >

^QQ^P

■Ofl

e3 cj
05 05

>->

J 3

»o

<N to I

CO'

f—«

bfl

• r*

T3 TS
cJ oj
05 05 _

T3 r O~'

-2 S.'S

o • "

O.rH

$ 8 n-

O ri

bfl qj

<Nr

I .s*

<N u _

.-q _

OPTS OP3
—■ oj<S be cj
■TJ - . - -

o 00 CO CO 03

^ -U - 4 J ,+J

M° °,2.2

C/1 1

A’S

, -T3 -rt

»H O

* >

. CO
“?H I
* ro <*

i N '

:as

. O
. c3
.pO

0 )

'■d

‘ c3

0 >

-6

bfl

**CM

CD CD CD
P> '

o o c>

CO CM

J 5

oo
co 3
0 '

cJ
co oo

cs 2 «.s « 2*8

° a e.fS o?

y S 33 ® ■*■{ -S os

Oj ” C3 'O j§ Oj CO

c a & g §

§) m a 5 S 2 2
o o P P p
a a_S2 o o o oo
o 0 3131313

(h >H -*_5 .+_} -+_5 -*_5

XJd .co 3 2 3

O o3

O •;-* -P>

£3 <-> eg
5j — S3 r.

. 2 « a d

eg a w o a

C 03 CO w CD

B^’E'E ej oj

M fli 4 J -p O ^

2 fl-c O O O-b o 03 cj.&.S* .»

<!<!<!WWWPWpI3mM oq

-O

• ri

•P

C<IC0-^‘0<0tN.00 05OrHC5je0

162

EXPERIMENTAL VEGETATION

Table 65 .—Sods transplanted at Burlington , Apr. 15, 1921.

BEHAVIOR TABLES

163

Table 65 .—Sods transplanted at Burlington, Apr. 15, 1921 —Continued.

164

EXPERIMENTAL VEGETATION

Table 66. —Sods transplanted to gravel-knoll, Lincoln, Apr. 20, 1920.

BEHAVIOR TABLES

165

Table 66. —Sods transplanted to gravel-knoll , Lincoln, Apr. 20, 1920 —Continued.

166

EXPERIMENTAL VEGETATION

Table 67 .—Sods transplanted to low prairie, Lincoln, Mar. 22 to Apr. 24, 1920.

BEHAVIOR TABLES

167

Table 67 .—Sods transplanted to low prairie , Lincoln , Mar. 22 to Apr. 24, 1920 —Continued.

168

EXPERIMENTAL VEGETATION

Table 68. —Sods transplanted to salt-flat, Lincoln, Apr. 24-26, 1920.

BEHAVIOR TABLES

1 Planted here in 1919.

Table 68. —Sods transplanted to salt-flat , Lincoln , Apr. 24~26, 1920 —Continued.

170

EXPERIMENTAL VEGETATION

BEHAVIOR TABLES

171

1 Transplanted into the Poa zone.

69 .—Sods trnnsplanted to swamp and Poa zone, 1 Lincoln, Apr. 24, 1920 —Continued.

172

EXPERIMENTAL VEGETATION

-d

■S

CLEMENTS AND WEAVER

PLATE 1

A. —Detail of high prairie, showing the luxuriant estival society and

amount of forage produced in June.

B. —General view of high prairie at Lincoln, Nebraska.

CLEMENTS AND WEAVER

PLATE 2

A. —Mixed-prairie station at Phillipsburg, Kansas.

B. —Corner of short-grass plains station at Burlington, Colorado.

«¥B»smr of ium ihirihv

CLEMENTS AND WEAVER

PLATE 3

penetration 4 and 2.7 feet respectively.

iifpvpfisrry of iujnois tuuxinv

CLEMENTS AND WEAVER

PLATE 4

Year-old Elymus canadensis (A), Bouteloua racemosa (B), and Andropogon nutans (C); maximum penetration about 3, c

and 4 feet respectively.

CLEMENTS AND WEAVER

PLATE 5

A. —Gravel-knoll station, Lincoln.

B. —Low-prairie station, Lincoln, showing transplanted sods

CLEMENTS AND WEAVER

PLATE 6

A. —Low-prairie station in September showing height of subclimax dominants.

B. —Salt-flat station. Lincoln; the chief grass is Distichlis sp:cnta.

[ CLEMENTS AND WEAVER

PLATE 7

A. —View of sod transplants in swamp at Lincoln, showing water standing on surface.

B. —Subclimax prairie at Nebraska City, showing rank growth of Stipci spar tea and

Ceanoth.us ovatus held in check by mowing.

CLEMENTS AND WEAVER

PLATE 8

A. —A single season’s growth of Andropoqon nutans from block of sod (left) and from seed fright) on low prairie at Lincoln.

B. —Year-old Bouteloua hirsula grown on high prairie at Lincoln.

CLEMENTS AND WEAVER

PLATE 9

Year-olcl seedlings of Andropogon scoparius (right), A. nutans (center), and A. furcalus (left) «
high prairie; plants about 6 to 8 inches tall.

Root system of year-old Andropogon furcatus from cultivated lowland, Lincoln; scale in feet.

CLEMENTS AND WEAVER

PLATE 10

A. —Bulbilis dactyloides sod in high prairie.

B. —Andropogon nutans in September of first season, showing marked growth

without competition.

CLEMENTS AND WEAVER

PLATE 11

A. —Calamovilfa longifolia in September of first season, grown in rich silt-loam.

B. —Bouteloua gracilis one year old, grown in cultivated soil.

aimsnr of Illinois unw

CLEMENTS AND WEAVER

PLATE 12

Roots of year-old Muhlenbergia pungens (A) and Calamovilja longifolia (B) grown
in rich silt-loam; depths of penetration 2.5 and 6 feet respectively. (C)
Andropogon halli at end of first year in silt-loam.

CLEMENTS AND WEAVER

PLATE 13

Andro pogon nutans planted in denuded quadrats at Nebraska City (A) in 1922, (B) in 1920.

Photographed June 3, 1922.

mUHl SIQHITI! JO AliSttJAlli

-1-3

- 1-3

•+3

-4-3

-M

S—i

bJO

•<s>

£
b
* c*i
rO

bJO

'O'

-P

b-C

* <s>

P XD
P p
P
03

T—I

-P

o >?

L "o

P >

<u •-

^ D

gs,

-+3 D
5-i

Q d

<4-1

P 1>
P P
P P
CO 03

g >0

O “
eo

“’d
dP

•<S>

5-.

*<s>

+=>
m
o
hC

_ Oj

° D

^>P-T

• rp

$ o

> 03

*CO

O o

i £

£ hJ

<$ PQ

CLEMENTS AND WtAVER

PLATE 15

A. —Sod-cores and details of installation, Burlington.

B. —Clip-quadrat in Bulbilis-Bouieloua short-grass, Burlington.

✓

••

■ • flu '

• *