EFFECTS OF SEEDBED MODIFICATION, SOWING DEPTH AND by Von Kenneth Winkel

EFFECTS OF SEEDBED MODIFICATION, SOWING DEPTH AND by Von Kenneth Winkel

EFFECTS OF SEEDBED MODIFICATION, SOWING DEPTH AND

SOIL WATER ON EMERGENCE OF WARM-SEASON GRASSES by

Von Kenneth Winkel

A Dissertation Submitted to the Faculty of the

SCHOOL OF RENEWABLE NATURAL RESOURCES

In Partial Fulfillment of the Requirements

For the Degree of

DOCTOR OF PHILOSOPHY

WITH A MAJOR IN RANGE MANAGEMENT

In the Graduate College

THE UNIVERSITY OF ARIZONA

1990

THE UNIVERSITY OF ARIZONA

GRADUATE COLLEGE

2

As members of the Final Examination Committee, we certify that we have read the dissertation prepared by

Von Kenneth Winkel entitled

Ufects of Seedbed Modification, Sowing Depth and

Soil Water

On

Emergpmce

of Warm-Season Grasses and recommend that it be accepted as fulfilling the dissertation requirement

Doctor of Philosophy for the De ree of

/

A(2

, ye4:2-7

1

/9fe7

Date

A l

At!.14

2

-

77

//fiO

r e.

,

Date

Date

\IL/LtIN. )

I

AULL,LZ__7

5

Date

/Fe;

Date

Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copy of the dissertation to the Graduate

College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.

,7

\

-

7N

,.._ .

Drsl'ertatIon''Director

Date

Of\,

3

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at the

University of Arizona and is deposited in the University

Library to be made available to borrowers under the rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgement of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate

College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

)

S I GNE

4

ACKNOWLEDGEMENTS

This research was funded by a grant from the U.S.

Department of Agriculture, Rangeland Research program.

I wish to thank my research advisor, Dr.

Bruce A. Roundy for his accessibility, expert advice, support and encouragement. Appreciation is expressed to Drs. Phil R.

Ogden, George B.

Ruyle, Albert K. Dobrenz,

Fredric R.

Lehle and E. Lamar Smith for their suggestions and help.

Thanks goes to Dr. David K. Blough for assistance with statistical analyses. I especially thank my wife Ann and our children Ryan and Jamee for their faithful patience and support throughout this study.

TABLE OF CONTENTS

Page

LIST OF FIGURES

LIST OF TABLES

ABSTRACT

INTRODUCTION AND OBJECTIVES

LITERATURE REVIEW

Effects of Soil Water On Emergence,

Survival and Morphology of Grass Seedlings

Effects of Sowing Depth On Emergence,

Survival and Morphology of Grass Seedlings

Effects of Seedbed

Microsites

On Grass

Seedling Emergence

Effects of Seedbed Preparation On Grass

Seedling Emergence

DETERMINATION OF SEED LOCATION IN RELATION

TO SEEDBED PREPARATION

25

34

36

41

Introduction

New Technique

Technique Test

EFFECTS OF CATTLE TRAMPLING AND MECHANICAL

SEEDBED PREPARATION ON BURIAL OF GRASS SEEDS

44

44

45

49

52

Introduction

Methods

Results and Discussion

52

54

56

EFFECTS OF CATTLE TRAMPLING AND MECHANICAL

SEEDBED PREPARATION ON GRASS SEEDLING EMERGENCE ... 68

Introduction

Methods

Study Area

Treatment Application

68

72

72

72

8

11

13

15

25

5

TABLE OF CONTENTS

(continued)

Data Collection

Experimental Design

Results

Discussion

INFLUENCE OF SEEDBED MICROSITE CHARACTERISTICS

ON GRASS SEEDLING EMERGENCE

Introduction

Methods

Results and Discussion

Seedling Emergence

Soil Water

EFFECTS OF SOWING DEPTH AND SOIL WATER ON

EMERGENCE AND SURVIVAL OF GRASS SEEDLINGS

Introduction

Methods

Results and Discussion

Maximum Emergence

Seedling Survival

EFFECTS OF SOWING DEPTH AND SOIL WATER ON

GRASS SEEDLING MORPHOLOGY

Introduction

Methods

Results

Primary Seminal Root Lengths

Shoot and Root Weights

Root:Shoot Ratios

Discussion

SUMMARY AND CONCLUSIONS

Seed Location

Effects of Seedbed Preparation On

Seedling Emergence

Seedbed

Microsites

Effects of Sowing Depth and Soil Water

On Emergence and Survival

Page

75

76

78

86

92

92

96

102

102

105

111

111

113

117

117

121

128

128

130

135

138

143

145

149

152

152

153

155

155

6

TABLE OP CONTENTS

(continued)

Effects of

Sowing Depth and Soil Water

On Morphology

Species Characteristics

Sideoats Graina

Blue Panic

Lehmann and Cochise Lovegrass

Seedbed Preparation Treatments

Cattle Trampling

Land

Imprinting

Root plowing and

Ripping

Relationships Between Greenhouse and Field

Experiments

Microsites

Emergence and

Survival

Morphological Characteristics

APPENDIX; ANALYSIS

OF VARIANCE TABLES

LITERATURE CITED

Page

156

158

158

162

164

167

167

168

169

170

170

171

171

174

177

7

LIST OF FIGURES

Figure

Page

1.

View of split soil plug showing blue panic seeds

2.

Linear regressions of known seed percentages on seed percentages estimated at different soil depths for

3 grass species using a vial sampling technique

48

51

3.

Distribution of sideoats grama, blue panic, and

Cochise and Lehmann lovegrass seeds immediately after treatment in relation to treatment and depth in the seedbed 57

4.

Distribution of sideoats grama, blue panic, and

Cochise and Lehmann lovegrass seeds after summer thunderstorms in relation to treatment and depth in the seedbed

60

5.

Distribution of sideoats grama, blue panic, and

Cochise and Lehmann lovegrass seeds that produced emergent seedlings in relation to depth in the seedbed

61

6. Precipitation

1987, 1988 and

Arizona; and periods in the in relation to data for July and August

1989 at the Anvil Ranch, lengths of wet and dry upper

3 cm of the seedbed year and treatment 79

7.

Seedling density of sideoats grama, blue panic, and Lehmann and

Cochise lovegrass in relation to year and seedbed treatment

80

8.

Seedling emergence of sideoats grama, blue panic and

Cochise lovegrass in various seedbed microsites in relation to

3 soil water treatments

103

9.

Percent soil water in relation to microsites across time

107

8

LIST OF FIGURES (continued)

Figure

10.

Seedling density of sideoats

graina, blue panic and Cochise lovegrass across time in relation to sowing depth and watering treatment

11.

Soil water content

(wt/wt) in relation to depth in the seedbed across time

12.

Primary root depths of sideoats

graina, blue panic and

Cochise lovegrass in relation to sowing depth, water treatment and -1.5 MPa drying fronts

13.

Primary root depths of sideoats graina, blue panic and Cochise lovegrass in relation to sowing depth, water treatment and

-1.5 MPa drying fronts

14.

Examples of sideoats grama, blue panic and Cochise lovegrass seedlings

15.

Morphology of 15-day old sideoats graina, blue panic and Cochise lovegrass seedlings in relation to sowing depth and soil water treatment

16.

Soil water content

(wt/wt) in relation to depth in the seedbed across time

17.

Below-ground morphology of 5-day old sideoats

graina, blue panic and Cochise lovegrass seedlings in relation to sowing depth

18.

Leaf and root weights of sideoats grama, blue panic and

Cochise lovegrass seedlings in relation to soil water treatment and time

19.

Root:shoot weight ratios of sideoats graina, blue panic and Cochise lovegrass in relation to soil water treatment

126

134

Page

118

122

125

137

139

141

144

146

9

10

LIST OF FIGURES (continued)

Figure

Page

20.

Root:shoot weight ratios of sideoats graina, blue panic and Cochise lovegrass across time

148

LIST

OF

TABLES

Table Page

1.

Repeated measures analysis of variance of depth at which cumulative percentages of broadcast seeds were found for 4 grass species immediately after seedbed treatment and after at least

20 mm of rain

175

2.

Repeated measures analysis of variance of depth at which cumulative percentages of seedlings of

4 grasses emerged in relation to seedbed treatments 176

3.

Percent of seeds found on the soil surface, within the biological limit, and below the biological limit in relation to year, sampling time, species and seedbed treatment 64

4.

Sample

ANOVA table for determining effects of seedbed preparation treatments on grass seedling emergence

77

5.

Effects of seedbed disturbance on length of wet and dry periods following summer thunderstorms 82

6.

Effects of seedbed treatment on percent cover of bareground, litter and 4 indigenous grasses in relation to seedbed treatments

84

7.

Percent cover of bareground, litter and

4 indigenous grasses in

1987, 1988, and

1989 on a loamy upland range site in southern Arizona

84

8.

Percent frequency of sideoats grama, blue panic, and Lehmann and

Cochise lovegrass in relation to year and seedbed treatment 87

9.

Sample

ANOVA table for determining the influence of seedbed microsite characteristics on grass seedling emergence

100

11

LIST OF TABLES

(continued)

Table

Page

10. Maximum seedling emergence of sideoats graina, blue panic and

Cochise lovegrass in relation to sowing depth and water treatment....... ..... .......... .............. . 120

12

13

ABSTRACT

The ability of different seedbed preparation treatments to enhance seedling establishment of 'Vaughn' sideoats graina (Bouteloua curtipendula (Michx.) Torr.),

'A-130' blue panic (Panicum antidotale Retz.), 'A-68'

Lehmann lovegrass (Eragrostis lehmanniana

Nees)

and

'Cochise' Atherstone lovegrass (Eragrostis lehmanniana

Nees X E. tricophera Coss and Dur.) was determined in field and greenhouse experiments. Field experiments compared seed burial and seedling emergence on undisturbed plots with that of plots lightly or heavily trampled by cattle, furrowed with a land imprinter or plowed. Seed burial was nearly always greatest in plots disturbed by these 3 latter treatments compared to undisturbed or light-trampled plots. Summer thunderstorms increased burial on treated andundisturbed plots. These seedbed treatments likewise increased seedling emergence of all species during a moderately-wet summer but not during a dry summer when emergence was limited. Sideoats graina emergence was low all 3 years. Seed burial and emergence of the small-seeded lovegrasses was high in a wet year even on undisturbed plots. Greenhouse experiments were designed to determine effects of seedbed microsites, sowing depth and soil water on emergence,

14 survival and morphology of sideoats graina, blue panic and

Cochise lovegrass.

Emergence of all species was highest from seeds under gravel, followed by seeds under litter, seeds in cracks and finally seeds on the bare soil surface. Cochise lovegrass had high emergence under gravel for both continuously-wet and drying seedbeds.

Cochise lovegrass had greater survival, followed by blue panic and then sideoats graina.

All 3 species initiated permanent roots from nodes near the soil surface regardless of sowing depth. Seedlings from deeper-sown seeds had shallower primary roots and less survival than seedlings from shallow-sown seeds. Although seedbed treatments may increase the number of seeds buried and thereby increase seedling emergence when summer storms are frequent, treatments which bury seeds too deep may decrease seedling emergence. Seedbed treatments do not ensure successful emergence in a dry year.

15

INTRODUCTION AND OBJECTIVES

Perennial grass densities of many southern Arizona rangelands have been decreased by overgrazing, drought and flooding (Cox et al.

1982) and encroachment of woody shrubs (Humphrey

1958,

Cox et al.

1983).

Griffiths

(1901) surveyed ranchers for descriptions of southern Arizona rangelands prior to degradation.

Describing the San Pedro Valley, H. C. Hooker, owner of the Sierra Bonita ranch, said:

...the San Pedro Valley in

1870 had an abundance of willow, cottonwood, sycamore, and mesquite timber, also large beds of sacaton and graina grasses, sagebrush and underbrush of many kinds. The river bed was shallow and grassy and its banks were beautiful with a luxuriant growth of vegetation. Now the river is deep and its banks are washed out, the trees and underbrush are gone, the sacaton has been cut out by the plow and grub hoe, the mesa has been grazed by thousands of horses and cattle, and the valley has been farmed.

Suggesting a cause for the destruction, Hooker further states that "there were fully 50

,

000 head of stock at the head of Sulphur Spring Valley and the valley of the

Aravaipa in

1890.

In

1900 there were not more than one-half that number and they were doing poorly".

C. H. Bayless of Oracle, Arizona replied that in the

1880's:

...ten animals were kept in good condition where one can now barely exist. However, those

16 ten animals were then rapidly destroying the vegetation, not making proper use of it. Of the rich graina grasses that originally covered the country so little now remains that no account can be taken of them.

W. C. Barnes (1936) described the San Simon Valley of southeastern Arizona in

1882 this way:

The meadows were covered with soft lush grasses, almost untouched by animals except for the horses and mules of an occasional traveler.

...Everywhere on the more open areas those fine stock grasses, black, blue and hairy grainas, grew luxuriantly. Here and there along the wash were tracts of alkali land on which sacaton touched my stirrups.

Barnes estimated that in

1885, 50,000 head of cattle were grazing on the San Simon ranges. Upon returning to the San Simon Valley in

1934,

Barnes discovered that,

"many of the old valuable grasses and forage plants were gone. The green meadows were replaced by wide expanses of drifting sand".

Attempts to revegetate southern Arizona rangelands began in

1900

(Griffiths 1901).

Several range reserve tracts near Tucson and at area ranches were fenced and sown with a variety of native and introduced grasses and shrubs. These seeding trials were generally unsuccessful due to non-adaptability of the species and several years of below-average rainfall (Griffiths 1904; Thornber 1905,

1907, 1908, 1909, 1910, and

1911).

17

Little revegetation work was attempted between 1912 and the early 1930's. According to Cox et al. (1982), the passage of the National Industrial Recovery Act (NIRA),

Work Progress Administration (WPA), and the Civilian

Conservation Corps (CCC) in the 1930's made funds available to scientists for reseeding research, ecology and development of range methodology.

Reseedings during the 1930's and 40's were often successful during years of above-average rainfall (Barnes

1958) and unsuccessful when rainfall was below average

(Anderson and Swanson 1949). In Arizona during the 1940's and 50's, revegetation research was abundant. Much of the research effort was directed towards determining species adaptability. Consequently, hundreds of grasses, forbs and shrubs were seeded in exclosures (Judd and Judd

1976).

During this period, mechanical seeding and soil water conserving equipment was developed including: pitting and ripping implements, contour furrowers, eccentric disks, cultipackers, and the rangeland drill

(Cox et al 1982). Glendening (1942) used straw mulches to conserve soil water and increase seedling emergence.

In the 1960's and 70's Jordan (1970) prepared seedbeds in the San Simon Valley by root plowing and

18 pitting and then sowed a variety of introduced grasses including Lehmann lovegrass

(Eragrostis lehmanniana

Nees), Boer lovegrass (E. chloromelas

Steud), and Cochise lovegrass (E. lehmanniana

Nees

X E. tricophera Coss and

Dur.); and blue panic (Panicum antidotale Retz.).

Recently the land imprinter was developed by Dixon and Simanton (1980). It was developed for use on

Southwest rangelands to concentrate limited precipitation. It has since been shown to help cover broadcast seeds, firm seedbeds, reduce wind erosion and increase seed-soil contact (Haferkamp et al. 1987, Clary

1989).

In the past 100 years, revegetation research in southern Arizona has made important advances.

Revegetation equipment has been developed, seedbed preparation techniques have been refined and adapted plant species have been selected. According to Cox et al.

(1982), since the 1890's rangeland seedings have included more than 300 forb, grass and shrub species, 40 types of mechanically-prepared seedbeds and about 400 planting sites in the southwestern United States and northern

Mexico. Of the 300 species, only 14 are recommended by

Cox et al. (1982) for seeding on rangelands in various parts of the Sonoran and Chihuahuan deserts.

19 parts of the Sonoran and

Chihuahuan deserts.

In spite of the advances in revegetation technology, seeding failures still occur. Reasons for some seeding failures on semiarid rangelands may include insufficient soil water and soil coverage of seeds, and unfavorable temperature

(Vallentine 1989); seed and seedling desiccation (Evans and Young

1970,

Nelson et al.

1970,

Evans and Young

1972a); lack of radicle entry

(Campbell and Swain

1973,

Sheldon

1974, and Cox and

Martin

1984); ant, bird and rodent predation (Howard

1950,

Nelson et al.

1970); excessive seed burial

(Mutz and

Scifres 1975) and slow root growth and lack of permanent root development (Hyder et al.

1971,

Van Der

Sluijs and

Hyder 1974).

Cox et al.

(1982) suggest that problems associated with interpreting success or failure of seeding trials include: lack of weather data collection, lack of comparisons between new and standard seedbed methods, nonreplication of trials in time and space and lack of subjectivity in data collections.

Most seeding trials are practical in nature and often only emphasize seedling emergence. In order to determine causes for seeding failures, interactions between the seed and seedling, the seedbed and climate

20 must be understood. Evans

(1987) suggested that:

...scientists and resource managers must understand the processes that support the recruitment of new seedlings into a plant community either artificially induced or naturally induced. This will require a clear description and understanding of interactions involving the soil system with the seed and seedling... What are the quantifiable effects at the soil/seedling interface of varying soil textures, varying moisture regimes and varying levels of disturbance?

Much information is lacking regarding these interactions. Specifically information is needed concerning the following.

1. Seed Burial

Seed burial is recommended for successful reseeding.

Researchers have assumed that mechanical seedbed preparation and livestock trampling bury seeds, however no information is available on where these treatments place seeds in the seedbed. Also lacking is information regarding whether summer thunderstorms bury seeds and the location of seeds that produce emergent seedlings.

2. Effects of Mechanical Seedbed Preparation and Cattle Trampling On Seedling Emergence

Researchers have suggested the use of livestock trampling to help bury seeds and increase seedling emergence. However, information is lacking regarding reasons for increased seedling emergence from trampling.

21

If cattle trampling is to be considered as a viable seedbed preparation technique, comparisons are necessary between trampling and other standard techniques. Little information is available regarding the relationship between seedling emergence and water availability in the field. Also lacking is information regarding whether seedbed preparation treatments differ in their ability to retain soil water for germination and seedling emergence.

3. Seedbed Microsites

Surface-sown seeds fall into a variety of microenvironments some of which are associated with favorable conditions for germination and seedling establishment. Favorable microsites may include cracks and depressions in the soil surface or sites associated with gravel and plant litter. Little information is available regarding types of microsites that are adequate for germination and establishment of grasses on Southwest rangelands.

4. Effects of Sowing Depth and Soil Water On

Emergence, Survival and Morphology of Seedlings

Deep sowing has been suggested as a means of placing seeds in better soil water conditions for germination and seedling emergence. Deep sowing may also increase rooting depth since primary roots will initiate deeper in the seedbed. Seedlings with deeper roots may also survive

22 longer than seedlings with shorter roots. However, deep sowing may cause adverse effects because of limited seed reserves.

In the Southwest, the soil-drying front that forms after a precipitation event may proceed downward through the seedbed so rapidly that the available water period is insufficient for germination and radicle extension.

Therefore it may or may not be advantageous to deep sow seeds on Southwest rangelands.

Seed burial is often recommended for successful seedling establishment. However, seedling emergence generally decreases with sowing depth. Information is needed regarding the biological limit of emergence of adapted and desirable revegetation species.

Seedling establishment of most grass species is dependent upon initiation of permanent roots. In general, grass seedings exhibit 2 types of below-ground root morphology. Seedlings that exhibit the "festucoid" type initiate permanent roots at seed level and those that have the "panicoid" type initiate permanent roots near the surface. If deep sowing is shown to be successful in the Southwest, then the festucoid type root morphology may be advantageous. Information is needed regarding the

23 root morphologies of revegetation species.

Initiation and growth of primary and adventitious roots, and leaves of grasses is reduced by lack of soil water availability. In addition, some species appear to exhibit morphological survival mechanisms such as large root:shoot ratios. Information is needed regarding the effects of decreased water availability on revegetation species. Determination of plant morphological characterisitics which are associated with successful establishment in water-limited environments is necessary in future plant improvement and species selection programs.

Cox et al.

(1982) recommended 14 grass species for revegetation in the Chihuahuan and Sonoran deserts. Of those recommended,

4 were selected for use in this study. They include 'Vaughn' sideoats

graina (Bouteloua curtipendula

(Michx) Torr.),

'A-130' blue panic, 'A-68'

Lehmann lovegrass and

'Cochise' Atherstone lovegrass.

All but sideoats graina are introduced to the Southwest and all have proven to establish in areas similar to our study site. These species were selected not only for their adaptability, but because

1) information was needed to help explain sideoats

graina seeding failures,

2) the seeds represent different sizes and shapes, and

3) a

24 desire to determine the germination, emergence and morphological characteristics of Cochise lovegrass that are associated with its success as a revegetation species.

The overall objective of this study was to determine the ability of different seedbed preparation treatments to enhance seedling establishment of four warm-season grasses.

Main objectives included:

1. To develop a seed location technique to identify the location of grass seeds after seedbed disturbance.

2. To determine effects of cattle trampling and mechanical seedbed preparation on burial of grass seeds.

3. To determine effects of cattle trampling and mechanical seedbed preparation on emergence of grasses.

4. To determine the influence of seedbed microsite characteristics on emergence of grasses.

5. To determine effects of sowing depth and soil water on emergence and survival of grasses.

6. To determine effects of sowing depth and soil water on morphology of grasses.

25

LITERATURE REVIEW

Effects of Soil Water On Emergence,

Survival and Morphology of Grass Seedlings

Seed germination requires adequate soil water, temperature and oxygen among other species-specific requirements. Soil water is most accessible to a seed in a seedbed where the seed is surrounded by soil particles that are firmly packed around the seed to ensure hydraulic conductivity of water from soil to seed

(Collis-George and Sands

1959).

Seed germination relative to soil water intake is described in 3 phases (Bewley and Black 1985). The first phase, referred to as imbibition, is the absorption of water by the seed. The water potential of a dry seed is much lower than the surrounding wet soil and can exceed

-100 MPa.

Phase

2 is a lag period after most of the imbibition has occurred, during which metabolic events occur in preparation for radicle protrusion. The final phase begins at the onset of radicle protrusion.

Water uptake resumes during this phase and is initially related to cell production in the radicle and later to decreases in osmotic potential due to production of osmotically active substances produced by hydrolysis of seeds reserves (Bewley and Black

1985).

26

Germination and seedling growth of range grasses occurs at a variety of water potentials depending upon the species. Knipe (1968) germinated seeds of alkali sacaton, galleta grass and blue graina in aqueous mannitol representing water potentials of

0, -0.1, -0.4,

-0.7, -1, -1.3 and

-1.6 MPa.

Alkali sacaton germination was most severely affected by low water potentials, and galleta grass and blue graina were less affected. Total germination of the latter grasses was near

50% at

-1.6

MPa.

McGinnies (1960) studied the effects of soil water and temperature on germination of

6 cool-season grasses.

As water potential decreased, germination rate was reduced and total germination after 28 days decreased.

A major cause of seeding failures on arid rangelands is inadequate soil water (Vallentine 1989). Precipitation often comes in short bursts during which time a seed must respond quickly by penetrating the soil surface with its radicle and extending its seminal primary root deep into the soil (Whalley et al.,

1966).

Limited availability of soil water decreased establishment of blue grama, crested wheatgrass and

Russian wildrye on the Central Great Plains

(Hyder 1971,

Van Der Sliujs and

Hyder 1974). Attempts to revegetate

27 the arid shadscale vegetation zone of the Great Basin have been largely unsuccessful due to inadequate precipitation (Bleak et al.

1965). Seedings in southeastern Arizona were often unsuccessful due to inadequate rain (Jordan 1970, Roundy and Jordan 1988).

Seeding failures on arid rangelands of New Mexico were also blamed on inadequate precipitation

(Herbel et al.

1973).

Seeds often germinate, but do not survive because inadequate moisture prevents the root system from becoming established and being able to support the plant through later periods of low soil water potentials

(Frasier et al.

1984).

Most perennial grass species have

2 types of root systems, seminal and adventitious. Seminal primary roots emerge from the radicle which is initiated at the scutellar or first node-.

One to several weeks after germination, depending upon the species, soil temperature and soil water conditions, adventitious roots emerge from the crown or coleoptilar node, and in some species, from other upper nodes.

In many species, seminal roots are temporary and may only live a few weeks

.

(Hyder et al.

1971; Briske and Wilson 1978). Their purpose is to provide water to the plant until the more permanent adventitious roots can develop.

28

There are generally 2 types of below ground seedling growth (Hyder et al. 1971).

Grasses of the panicoid type have an elongated subcoleoptile internode.

Grasses with the festucoid type, do not have an elongated subcoleoptile internode, but a long coleoptile (Hyder et al.

1971).

The disadvantage of the panicoid type in arid and semiarid climates is that the subcoleoptile internode elevates the coleoptilar node (the site of most adventitious root initiation) to or near the soil surface. Adventitious roots require high soil water to initiate (Wilson and Briske 1979), and if they are near or above the surface, the chance of initiation is lessened. Since festucoid seedlings do not have an elongated subcoleoptile internode, the site of adventitious root initiation is at seed level, which is usually in wetter soil (Hyder et al. 1971).

In general, grass seedlings require 2 periods of soil water to become established; one period for germination and emergence, which includes seminal root initiation; and another period for initiation of adventitious roots.

Germination and seminal root initiation of blue graina seedlings requires a moist soil surface for 2-4

29 days (Wilson and

Briske 1979).

Requirements for sideoats graina,

Cochise lovegrass, blue panic, and alkali sacaton include a minimum of

24, 30, 48, and

48 hours of soil water potential greater than -0.03 MPa (Simanton and

Jordan

1986). Heteropogon contortus

seedlings began producing seminal roots within

2 days of sowing under ideal conditions (Glendening 1941).

Adventitious roots do not begin growth until several weeks after germination depending upon the species and timing of favorable conditions. Blue grama seedlings require a wet period of 2-4 days approximately

2 to

8 weeks after germination (Wilson and

Briske 1979).

• In a greenhouse study, blue graina seedlings began producing adventitious roots

11-20 days after sowing under favorable conditions (Van Der

Sluijs and

Hyder 1974).

Adventitious wheat roots formed within

4 weeks after planting under favorable conditions (Ferguson and

Boatwright 1968).

Species differ in their rate of seminal primary root extension. Under ideal conditions in the greenhouse, sideoats

graina seeds responded very quickly to even moderate precipitation and seminal primary roots extended nearly

70 mm in only

7 days. In contrast, blue panic and alkali sacaton reached only

30 mm in

7 days, and

Cochise

30 lovegrass extended nearly 20 mm in the same time period

(Simanton and Jordan 1986).

Seminal primary root growth rates of blue graina averaged 1.1 cm/day when soil water potential was near

-0.03 MPa, and only 0.6 cm/day when soil water potential was near

-1.5 MPa (Wilson and Briske 1979). Fulbright et al.

(1984) reported decreased seedling emergence and total roots per seedling with increased dehydration.

They imposed dehydration treatments of 0, -4, and -10

MPa osmotic potential on germinating seeds of green needlegrass when seminal primary roots were 2-5 mm in length and then placed them in soil under high matric potential (-0.03 MPa). The number of seminal roots per seedling did not differ between the 0 and -4 MPa treatments, but was reduced by the -10 MPa treatment.

Length of the longest seminal root was also decreased with the -10 MPa treatment. If the seminal primary root was killed by dehydration or excised, seedlings developed up to

3 seminal lateral roots.

No additional seminal roots were developed if the seminal primary root was undamaged.

Hassanyar and Wilson (1978) reported decreased development of seminal lateral roots with increased water

31 stress. Germinating seeds of crested wheatgrass and

Russian wildrye with

2-5 mm seminal primary roots were exposed to solutions with water potentials of -10, -22,

-37, -91, or -158 MPa, after which the roots were excised and the seeds placed in a growth chamber under ideal conditions for 20 days. Water potentials of -91 and

-158 MPa for crested wheatgrass, and -32, -91, and

-158 MPa for Russian wildrye significantly reduced the percentage of seedlings that developed seminal lateral roots. After a temporary drought of -32, -91, and -158

MPa; 75, 58, and 24 % of crested wheatgrass, and 69, 20, and 6 % of

Russian wildrye seedlings developed seminal lateral roots.

Seedling survival could be dependent upon the ability of seminal roots to extend though a dry soil layer. This could be important if deeper soil layers had higher matric potentials.

In a study of root development of wheat, oats and barley,

Salim et al. (1965) discovered that most seminal roots stopped growing soon after reaching dry soil. They then performed an experiment where they planted cereals in boxes containing

2 layers of soil, soil water potential in the top layer was maintained at

-0.03 MPa, and the bottom layers were maintained at

-.045, -.10,

32

-.60, or -1.0 MPa.

After 20 days, root penetration of the bottom soil layer of both barley and oats for the -.045

and

-.10 MPa treatments was significantly greater

(p<0.01) than the dryer treatments of

-.60 and

-1.0 MPa.

Therefore, the amount of penetration of the bottom layer was dependent upon its soil water potential.

As part of the above study,

2 range grasses were also tested for their ability to penetrate dry soil.

Both sideoats grama and sand lovegrass appeared to have greater ability to penetrate then did cereals. Seminal root growth rates in top layers (-.03 MPa) and bottom layers (< -1.5 MPa) were 11 and

2 mm/day for sideoats grama, and

9 and 1 mm/day for sand lovegrass, respectively.

Although some grasses exist for a while on only a seminal root system, most must eventually rely on adventitious roots. Hyder, et al. (1971) reported that plantings of blue grama nearly always died at 6-8 weeks of age. Upon examination, they found that since blue grama has an elongated subcoleoptile internode, the coleoptilar node was nearly always within 6 mm of the soil surface. At this depth the soil was nearly always dry in this study. Most adventitious roots were dead stubs.

33 stubs.

Blue graina seedlings restricted from forming adventitious roots by low soil water, initiated them in just a few hours when soil water conditions were improved

(Van Der

Sluijs and

Hyder 1974).

Wilson and

Briske (1979) observed that adventitious roots of blue graina seedlings formed and survived after a

3.8-cm rain, but not after a lighter rain. Adventitious roots grew

2.3 cm/day in high soil matric potentials

(-0.03 MPa), and only

0.70 cm/day under lower potentials

(<-1.5 MPa).

If the site of adventitious root initiation is above the soil surface, high relative humidity must be present for initiation to occur. Briske and Wilson

(1978) varied the relative humidity around the coleoptilar nodes of blue graina seedlings and kept the seminal roots in soil at

-0.03 MPa.

In general, adventitious root lengths and numbers decreased with a decrease in relative humidity.

Numbers of adventitious roots per seedling at

100, 96,

86, 81, and

76% humidity were

7.9, 8.0, 7.0, 4.4, 3.6, and

3.2, respectively.

Olmsted

(1941) showed that sideoats

graina adventitious root initiation required

3 days of surface

34 soil wetness. He watered

4 sets of sideoats graina seedlings every

3, 6, 12, or 20 days. After

42 days, numbers of adventitious roots for the 3, 6, 12, and 20day treatments were

8.4, 4.9, 3.9,and

O. Thus, numbers of adventitious roots decreased with decreases in soil water.

Lack of soil water availability also adversely affects leaf growth of grass seedlings.

Herbel and

Sosebee (1969) compared leaf growth of Boer lovegrass and black graina in relation to soil water and temperature. At lower temperatures, average shoot height decreased with decreased soil water availability.

Glendening (1941) showed that shoot lengths and weights of Heteropogon contortus decreased with reduced irrigation frequency.

Leaf lengths of sideoats graina were also adversely affected by reduced water availability (Olmsted

1941).

Effects of Sowing Depth On Emergence,

Survival and Morphology of Grass Seedlings

Following precipitation on arid rangelands, a soil drying front is initiated at the soil surface and moves down through the seedbed. Under these conditions, deepsown seeds may be in favorable conditions for germination for a longer period than surface or shallow-sown seeds

(Tadmor and Cohen 1968).

35

Deep sowing has both increased and decreased emergence of range grass seedlings. In the majority of cases, under favorable conditions for germination, emergence decreases with depth of sowing (Mutz and

Scifres 1975,

Cox and Martin

1984, Fulbright et al.

1985, Carren et al.

1987,

Newman and Moser

1988).

Emergence of blue grama seedlings was greater from

2 than

1 cm under marginal soil water conditions

(Carren et al.

1987). Apparently soil water conditions at 2 cm were more favorable for germination than at 1 cm. Under low soil matric potential (-1.5 MPa) in a greenhouse, emergence of crested wheatgrass seedlings occurred first from seeds sown at 1.25 cm and on the surface, and from seeds sown at 2.5 cm 1 day later (Schmutz and Al-Rabbat 1969).

Deep sowing may allow seedlings to initiate primary roots deeper in the seedbed and thus give them a survival advantage over seedlings from shallow-sown seeds.

However, effects of sowing depth on primary root development have been variable and are apparently species-specific

(Tadmor and Cohen

1968,

Cornish

1982,

Fulbright et al. 1985). Primary root depth of Wilman lovegrass is independent of sowing depth, primary root depth of Klein grass increases with sowing depth and deep sowing decreases primary root penetration of weeping lovegrass (Tischler and

Voigt 1983).

36

With the exception of low survival of seedlings from surface-sown seeds due to limited radicle penetration, depth of sowing had no effect on survival or development of ryegrass and red clover (Campbell

1985) . In a greenhouse study, deep sowing did not decrease survival of crested wheatgrass seedlings (Schmutz and Al-Rabbat

1969).

Effects of Seedbed

Microsites

On Grass Seedling Emergence

Each plant species has specific requirements that must be met for successful germination and establishment.

These requirements may include certain soil water, temperature, light and oxygen conditions.

Whether these requirements are met or not is often determined by the microenvironment surrounding the seed during germination and establishment. The microsite where these requirements are met has been termed by Harper (1961,1977) as a usafesiteu. Harper

(1977) stated that:

...a safe site is envisaged as that zone in which a seed may find itself which provides (a) the stimuli required for breakage of seed dormancy, (b) the conditions required for the germination processes to proceed, and (c) the resources (water and oxygen) which are consumed in the course of germination. In addition a

'safesite' is one from which specific hazards are absent such as predators, competitors, toxic soil constituents and pre-emergence

37 pathogens.

Harper further stated that, "the degree of hetergeneity of the soil surface is dependent upon seed size". This may be why small seeds appear to need less soil tilth or surface roughness to meet their safesite requirements (Evans and Young

1972a ,

Cox and Martin

1984).

In striving to understand the germination and establishment requirements of plants, researchers have exposed seeds and seedlings to a variety of microtopographical situations, and studied the emergence of seedlings in natural microsites. These have included cracks, holes and depressions of various sizes, shapes and depths on the surface of mineral soil; various types and degrees of scarified soil; dead plant litter and a large array of artificial situations.

Microtopography modifications appear to improve moisture conditions and minimize temperature extremes.

In a study where 100 times as many downy brome seedlings emerged from pits than on the soil surface Evans and

Young

(1972a) reported that soil water potential during the spring growing season on the surface varied from

-0.2

to -1.5 MPa, while water potential in 9-cm pits stayed at

-0.03 MPa all season. They also showed that while the

38 relative humidity of both the soil surface and pit were near saturation at night, day time humidities in the pits were never less than

40% as compared to those on the surface which were often

10%.

In the same study, temperatures in the pits fluctuated much less than those on the unprotected soil surface.

In a similar study,

Oomes and

Elberse (1976) reported greater emergence of

6 grassland herbs in

10-mm deep grooves than on the surface. However emergence was even less for

3 of the herbs in 20-mm grooves. Limited emergence at

20 mm may have been due to lack of oxygen.

In a green house study where the soil water content was the same across a series of pits of different depths and a flat soil surface

(Oomes and

Elberse, 1976), greater emergence of herbs was recorded in the pits.

Since the soil water content was the same on the surface, the pits must have met another unknown germination requirement.

Harper et al.

(1965) demonstrated that Plantago lanceolata but not P. media or P. major

responded positively to

1.25-cm and

2.50-cm soil depressions.

Harper et al.

(1965) found greater emergence of B.

rigidus and B. madritensis on soil surfaces with higher

39 proportions of smaller and larger aggregates, respectively. Differences were interpreted relative to the contact the seeds made with the aggregates. Sheldon

(1974) showed that sowing in a hole produced greater germination and establishment of various forb seedlings then sowing on the soil surface. He suggested that this response was a result of higher humidity in the hole.

In an experiment designed to determine plant responses to various microsites in the Great basin,

Eckert et al.

(1986) found greater emergence of Wyoming big sagebrush, perennial grasses and various annual forbs in cracks and trenches than on the unprotected soil surface.

The presence of litter on mineral soil apparently increases water availability by decreasing evaporation and the amount of radiation reaching the soil surface

(Kay 1987) .

There is also an indication that litter may provide leverage for increased radicle entry.

Evans and Young

(1970) reported

47 times more emergence of medusahead rye and much greater seed production

(1000 seedhead/m as opposed to

30 seedhead/m) under litter than on the unprotected soil surface.

Environmental monitoring showed that litter greatly moderated maximum and minimum temperatures, and shading

40 from litter reduced light reaching the soil from

30 to

97.7%.

Relative humidities under litter varied from

60 to

95%, while those on the surface varied from

15 to

95%.

Soil water depletion was also much less rapid under litter. In a similar study, Evans and Young

(1972b) showed much greater emergence of Russian thistle under litter, and although no environmental measurements were taken, they observed much greater radicle desiccation on the unprotected soil. They also suggested that the litter provided points to hold the seeds while the radicle entered the soil.

Fowler

(1986) discovered no differences in germination of perennial grasses from under

2 quantities of litter and from the soil surface. However, fewer seedlings on the surface survived.

In a classic study, Harper et al.

(1965) set up a simple experiment where he sowed seeds of

Plantago lanceolata,

P. media

and P. ma

.

or around various objects including glass sheets and wooden boxes placed in different positions on soil. He then recorded the origin of emergence of each seedling. Approximately

90% of P.

media and P. lanceolata seedlings emerged close to objects or depressions on the surface. In contrast, seedlings of P. ma or were not generally associated with

41 any objects. Harper gives no explanation for the seedling response, but it could involve several factors including relative humidity, soil water potential, temperature, and seed leverage.

In a similar study,

Sheldon

(1974) used various artificial objects and showed that treatments imposing a high humidity over the treatment area resulted in higher germination and emergence than those that did not.

Blom

(1978) used different sizes of glass beads to simulate sand grains and showed decreased establishment of seedlings on fine beads. He observed that seedling mortality was largely a result of the inability of radicles to penetrate a dense substrate surface.

Effects of Seedbed Preparation

On Grass Seedling Emergence

Cox et al.

(1982), Jordan (1981) and Vallentine

(1989) have provided extensive reviews of seedbed preparation techniques.

Seedbed preparation is often necessary to reduce competition from resident vegetation, loosen the soil surface and bury seeds. Mechanical seedbed preparation techniques include: plowing, land imprinting, chaining, cabling, harrowing, railing, ripping, bulldozing and pitting (Jordan 1981, Vallentine 1989).

4 2

Root plowing, disk plowing, chaining and pitting were compared in a series of seeding trials in southeastern Arizona by

Jordan and

Maynard (1970). They reported similar forage yields between root-plowed and disked plots, and both treatments produced about twice as much as the chained and pitted plots.

Root-plowed plots in the semiarid Trans-Pecos region of Texas produced good initial plant densities of Lehmann and Cochise lovegrass, and sideoats graina (Nelson and

Gabel 1987).

Herbel et al. (1973) developed a machine to root plow, seed, create basin pits and deposit up-rooted shrubs on the seedbed. Of 23 seedings in southern New

Mexico, 10 had good to excellent stands and 4 had fair stands.

Land imprinting was developed for use on arid

Southwest rangelands to concentrate limited and erratic precipitation (Dixon and Simanton 1980).

Imprinted plots on a semidesert grassland in southern

Arizona produced fewer plants than railed or disked plots (Cox 1986).

Seedling density of crested wheatgrass in the northern Great Basin was nearly twice as high from imprinted plots compared to loose-disked plots, but

43 seedling density was 2 to 4 times greater from drilling than imprinting on firm unprepared seedbeds

(Haferkamp et al. 1987).

Broadcast seeding and imprinting produced greater seeded species density, cover and production than seeding with a rangeland drill on a burned sagebrush community in central Utah (Clary 1989).

Livestock trampling as been recommended as a means of increasing seed burial and seedling emergence (Plummer et al.

1955, Hormay 1970, Vallentine 1989).

Eckert et al.

(1986) reported that moderate cattle trampling increased emergence of

perennial grasses, but decreased emergence of perennial forbs on specific seedbed microsites in

Nevada. They also showed that heavy trampling decreased perennial grass emergence, but increased emergence of sagebrush and annual forbs on certain microsites.

44

DETERMINATION OP SEED LOCATION IN RELATION

TO SEEDBED PREPARATION

Introduction

Artificial seeding of semiarid rangelands is often unsuccessful partly because of limited understanding of the response of seeds to different seedbeds.

Of particular interest is the effect of seedbed preparation treatments on seed placement, and the location of seeds that produce emergent seedlings. Seeds buried too deeply may not produce emergent seedlings. In contrast, seeds buried too shallow may also fail to produce seedlings due to limited soil water.

Several seed location techniques are used.

Most are used to determine total populations of viable seeds in the seedbank (Malone 1967, Jerling 1983, Staaf et al.

1987), or numbers of seeds at various depths in the seedbed

(Moore and

Wein 1977, Fay and Olson 1978,

Granstrom 1982,

Pareja et al. 1985). Other methods are used to determine numbers of seeds that survive

(Archibold 1979) or are stimulated to germinate or emerge by various cultural treatments (Wesson and

Wareing

1969). Interest has recently developed in assessing the accuracy of seed placement by drills

(Choudhary et al.

1985, Kaviani et al. 1985).

45

Most seed location techniques involve: 1) germinating seeds from seedbed samples, 2) sieving to isolate seeds, 3) tracing seedlings to their seeds, and

4) X-raying seedbed samples.

Determining the effects of seedbed preparation techniques on seed placement and seedling emergence requires a seed location technique that determines seed depths with minimal disturbance in fragile seedbeds. The method should be rapid to permit analysis of many samples in highly variable seedbeds.

The objective of this chapter is to describe a new seed location technique used to help determine the location of seeds after various seedbed preparation treatments on a sandy loam soil in southern

Arizona, and show evidence of its accuracy as an estimator of seed location.

New Technique

Plots 1 m

2

in area were broadcast seeded to cover the surface with a single layer of seeds of either

'Vaughn' sideoats graina (Bouteloua curtipendula

(Michx.) Torr), 'A-130' blue panic

(Panicum antidotale

Retz.), 'A-68' Lehmann lovegrass (Eragrostis lehmanniana Nees) or

'Cochise' Atherstone lovegrass

(Eragrostis lehmanniana Nees X E. tricophera Coss and

46

Dur.). Plots were then treated with heavy and light cattle trampling, land imprinting, root plowing or ripping or left undisturbed. Plots that were imprinted or lightly or heavily trampled by cattle were seeded before treatment, while plowed or ripped plots were seeded after plowing or ripping. The soil was a sandy loam (fine, mixed, thermic

Ustollic Haplargid). Plots were sampled after treatment when the soil was dry by protecting the seedbed with a layer of cotton cloth and then sprinkling the sample area with water until the soil was saturated to 3-5 cm. Plots were also sampled after rain when the soil water content was near field capacity. Soil plugs were collected by inserting a

3.5 cm diameter by

6 cm plastic vial into the seedbed. The vial was extracted from the soil, capped, immersed in liquid nitrogen until the soil was frozen (10-20 seconds) and placed in an ice chest lined with dry ice for transport to a freezer.

Freezing the soil plugs has the potential to move seeds slightly and therefore is recommended only to minimize disturbance during transport.

In the laboratory, each frozen soil plug was placed upright in a shallow water bath for about

2 minutes until it slid easily from the vial. After thawing, each plug was carefully split in half with a micro spatula. Each half was then placed in a cradle made from a plastic vial

47 cut in half, and the cradle was placed in a soil plug holder. The holder was placed under a dissecting scope, the outline of each soil plug was drawn on a plastic transparency and each visible seed was located and marked on the transparency (Fig. 1).

Seed depth was determined by measuring the distance from the seed to the soil surface. The data were recorded as percent of seeds found at particular depths in the seedbed.

Ease in locating seeds in the plugs was dependent upon seed size. Sideoats grama and blue panic seeds

(5 X 1 mm, and 2 X 1 mm) were easily seen under a lighted magnifying glass or dissecting scope set at

10 power. Lehmann and Cochise lovegrass seeds

(0.75 X 0.5

mm) required a dissecting scope set at

20-30 power.

Because buried seeds were often obscured by soil, numbers of buried seeds in relation to surface seeds could have been slightly underestimated.

The technique was used during the summers of 1987 and 1988.

Approximately

1440 soil plugs were collected during each year. Plugs were collected immediately after seedbed preparation, after summer thunderstorms, and after seedling emergence.

A crew of 3-4 people collected approximately

500 plugs in about 4 hours.

In the

4 8

Fig.

1.

View of split soil plug showing blue panic seeds.

49 laboratory, each plug was analyzed in about 10 minutes.

Approximate cost per plug including labor was

$1.00.

The technique permitted quantification of depth of seed burial and seedling emergence.

This information was helpful in explaining differences in seedling emergence associated with the different seedbed preparation treatments.

Technique Test

An experiment was conducted to determine the accuracy with which the method estimated the percent of seeds at different depths. Three wooden boxes 100 X 20 by

10 cm deep were filled with sandy loam soil (fine, mixed, thermic

Ustollic Haplargid) passed through a 2-mm sieve. While filling the boxes, seeds of Lehmann lovegrass, blue panic, and sideoats graina, (one species per box) were spread evenly at

5-mm intervals from

20-mm deep to the soil surface with the following percentages of total seeds:

20-mm - 5%, 15-mm - 10%, 10-mm - 15%,

5nun

- 20 % and the soil surface - 50%. After sowing, the soil was sprinkled with water until saturated and then 20 soil plugs per box from randomly preselected positions were extracted and examined as described above.

Data from the 40 plug halves per box were pooled, and estimated seed percentages from the plugs were

50 compared to the known percentages with correlation analysis.

Coefficients of determination (r

2

) between percentages of known and estimated seeds at different depths of sideoats grama, blue panic and Lehmann lovegrass were 0.80, 0.72 and 0.92 respectively (Fig. 2).

Comparisons of regression lines with one to one lines indicated that the technique overestimated the percentage of all three species on the surface, and underestimated the percentage of all species and particularly

Lehmann lovegrass at all other depths

(Fig.

2). The small size of Lehmann lovegrass seeds may limit their visual detection and result in underestimation.

The seed location technique permitted estimation of the location of ungerminated and germinated seeds of different species in various seedbeds.

51

— — SIDEOATS CRAMA

- - - - BLUE PANIC

- — — LEHMANN LOVEGRASS

A

11 UNE

10 20 30 40 50

ESTIMATED SEED PERCENTAGE

60 70

Fig.

2.

Linear regressions of known seed percentages on seed percentages estimated at different soil depths for

3 grass species using a vial sampling technique.

Each symbol is the mean of

20 samples.

52

EFFECTS OF CATTLE TRAMPLING AND MECHANICAL SEEDBED

PREPARATION ON BURIAL OF GRASS SEEDS

Introduction

Reasons for failure of some seedings on semiarid rangelands may include insufficient soil water and soil coverage of seeds, and unfavorable temperature,

(Vallentine 1989); lack of radicle entry (Campbell and

Swain 1973,

Sheldon

1974, and Cox and Martin 1984); ant, bird and rodent predation (Howard 1950, Nelson et al.

1970); and excessive seed burial

(Mutz and Scifres 1975).

Drilling is the preferred method for sowing seeds on most rangelands, since it distributes and covers seeds more uniformly,

(Vallentine 1963, Jordan 1981); and places seeds at a desired depth more accurately than broadcasting. However, drilling is not always practical.

Broadcast seeding must be used on steep slopes and rough terrain.

Broadcast seeding of small-seeded species on prepared seedbeds prior to summer rains has produced similar seedling emergence as drilling in the

Southwestern United States (Cox et al. 1986).

Chaining, ripping, plowing, imprinting and livestock trampling have all been used to help prepare seedbeds

(Allison and Rechenthin 1956, Tiedemann and Schmutz 1966,

53

Haferkamp et al.

1987, and Clary

1988, 1989).

It has been assumed that these treatments help to bury seeds.

However, the effects of mechanical seedbed preparation on seed burial have not been quantified. Although livestock trampling has also been suggested as helping to bury seeds (Plummer

1955, Hormay 1970, Vallentine 1989,

Pearson and

Ison 1987), no information is available on where trampling places seeds in the seedbed. Also lacking is knowledge regarding the numbers of seeds buried below the biological limit of emergence, and the location of seeds that produce emergent seedlings.

In the semidesert grasslands of the Southwest, seedbed treatment and sowing are usually done just before summer thunderstorms. Information on location of seeds after preparation and after precipitation may help explain differences in seedling establishment and aid in evaluation of the effectiveness of different treatments for different species and soils.

The objectives of this study were to determine the effects of cattle trampling, land imprinting, root plowing and summer thunderstorms on placement of broadcast grass seeds; and to determine the depth of seeds that produce emergent seedlings.

54

Methods

The study was conducted on the Anvil ranch, approximately

65 km southwest of Tucson, Arizona, in conjunction with a seedling establishment study. The soil at the site is a sandy loam (fine, mixed, thermic

Ustollic Haplargid; Sasabe series).

The experimental design of the study was a randomized block, split plot with

4 grass species X 5 seedbed treatments applied to each of 3 blocks on each of

2 years (1987 and 1988). Whole plots were years and split plots were seedbed treatments and species.

Seedbed treatments applied to 6 X 6 m

2

plots included light cattle trampling (approximately

10 hoofprints per m

2

), heavy cattle trampling (5 cattle herded inside the plot for

20 minutes), land imprinting, root plowing

(1987) or ripping (1988), and no disturbance.

Two 1-m

2 subplots within each 6 X 6 m treatment plot were intensively broadcast seeded with 'Vaughn' sideoats grama (Bouteloua curtipendula

(Michx.) Torr.), 'A-130' blue panic (Panicum

antidotale Retz.), 'A-68' Lehmann lovegrass (Eragrostis lehmanniana Nees), or

'Cochise'

Atherstone lovegrass

(Eragrostis lehamanniana Nees

X E.

tricophera Coss and Dur.). Plots were seeded after root

55 plowing or ripping, but were seeded prior to all other seedbed treatments.

Seeds were sown at a high enough rate to completely cover the soil surface with a layer of seeds. The soil surface was then protected with cotton cloth, the seedbed wetted and 3.5 X 5.0 cm soil plugs extracted with plastic vials. Samples were taken to the laboratory, split in half and seeds were located with a dissecting microscope.

Four samples were collected from each subplot for a total of

480 samples per sample period and 24 samples per treatment-species combination per sample period. Data were recorded as percent of seeds found at particular depths in the seedbed.

Seedbeds were sampled immediately after treatment application

(1987 and 1988), a few weeks later after the first major thunderstorm

(1987 and

1988), and after seedling emergence

(1987 only for blue panic,

Cochise and

Lehmann lovegrass, and

1988 only for sideoats grama).

Root-plowed plots were not sampled after treatment due to the heterogeneous nature of the seedbed, but were sampled after rain and seedling emergence.

Repeated measures analysis of variance was performed on the soil depth at which cumulative

56 percentages of seeds were found, and p-values were adjusted with

Greenhouse-Geisser statistics (Morrison

1976).

This analysis was used to determine differences in seed distributions in the seedbed after treatment and after summer thunderstorms among years, species and treatments, as well as differences in depth of emergence among species and treatments. The biological limit of each species was determined by plotting and regressing the cumulative percentage of seeds that produced emergent seedlings on seed depth.

Analysis of variance was used to determine differences in percentages of seeds found on the soil surface, within the biological limit and below the biological limit in relation to the treatments. Means were separated with the LSD test

(p<0.05).

Results and Discussion

Interactions of year, species and seedbed treatment were significant

(p<0.01) for depth at which seeds were buried immediately after treatment and after rainfall

(Table

1, Appendix)

.

Seed distribution of all

4 species immediately after treatment showed similar trends between the 2 years, with the exception that seeds from all species were buried deeper by heavy trampling in

1988

(Fig.

3).

Fig. 3. Distribution of sideoats grama, blue panic, and

Cochise and Lehmann lovegrass seeds immediately after treatment in relation to treatment and depth in the seedbed. Columns having the same letters are not significantly different at the 0.05 level. Undist

= undisturbed, light = light trampling, heavy

= heavy trampling, imprint = imprinting and plow = root plowing.

o o

Pd.

o

DEPTH (mm)

o

57 o

8

CD W

o o

Fig.

3

58

Heavy trampling was the most effective treatment in burying sideoats grama seeds, followed by imprinting, light trampling and control plots. Heavy trampling buried seeds to a depth of

27 mm, compared to 17 mm for imprinting, 16 mm for light trampling and only 6 mm for undisturbed controls. Blue panic seed burial followed trends similar to sideoats graina burial. However, the smooth, more spherical blue panic seeds were placed deeper than sideoats graina seeds in the seedbed. Heavy trampling buried seeds to 33 mm, imprinting to

22 mm, light trampling to

11 mm and non-disturbance to

7 mm.

Seeds of the

2 lovegrasses showed similar distribution and depth of seed burial, with the exception that imprinting buried more Lehmann than

Cochise lovegrass seeds. Again, heavy trampling buried most seeds, followed by imprinting, light trampling and nondisturbance. Heavy trampling buried seeds of both species to depths of more than

30 mm, followed by imprinting near

20 mm, light trampling near

10 mm and nondisturbance to

7 mm.

Seed sampling after rain was done after a

20 mm rain in

1987 and after a

35 mm rain in

1988. Seed location after summer thunderstorms was highly variable between years and among species, but generally, thunderstorms

59 increased seed burial (Fig.

4).

The trend of increased burial with increased seedbed disturbance seen in seed placement after treatment, continued after rainfall.

However, in many cases, more seeds in root-plowed plots were buried than seeds in other treatments. Root plowing produces a highly-fragmented seedbed and natural soil sloughing may have helped bury seeds.

There was a significant

(p<0.01) interaction between depth at which seedlings emerged and species, but not seedbed treatment (Table 2,

Appendix). This indicates that different species emerged from different depth intervals, but that each species had similar depth intervals of emergence for all seedbed treatments.

Blue panic seedlings emerged from greater depths than the other 3 species, followed by sideoats grama, then

Cochise lovegrass and then Lehmann lovegrass (Fig.

5). All blue panic, sideoats

graina, and

Cochise and

Lehmann lovegrass seedlings emerged from above 18, 16, 10 and 9 mm, respectively. Nearly all seedlings from all species emerged from buried seeds, even though after rain, many seeds remained on the soil surface. Either seeds on the surface did not germinate, or many were buried between sampling after rain and seedling emergence. Data from an associated greenhouse study

Fig. 4. Distribution of sideoats graina, blue panic, and

Cochise and Lehmann lovegrass seeds after summer thunderstorms in relation to treatment and depth in the seedbed. Columns having the same letters are not significantly different at the 0.05 level. Undist

= undisturbed, light = light trampling, heavy = heavy trampling, imprint = imprinting and plow = root plowing.

5

10

1987

60

a.

La

5

35 -

40

- UNMT

A AA

AAA

ABA

AB AB

A

A

20

AA

AA

BB

BC BC so

AA A

ABA A

B.BB

C C C

80

100

SEEDS

(%)

• UGHT

0

HEAVY

Fig.

4

IMPRINT

x

PLOW

61 o

5 -

15 -

.0

n

1 n

1

20

o

O BLUE PANIC

O SIDEOATS GRANA x

LEHMANN

LOVEGRASS

• COCHISE LOVEGRASS

20

1

40

1

60

CUMULATIVE

SEEDLING EMERGENCE (%)

80

100

Fig. 5. Distribution of sideoats grama, blue panic, and

Cochise and Lehmann lovegrass seeds that produced emergent seedlings in relation to depth in the seedbed.

62

(Fig. 10, page 118) showed that some seedlings of sideoats graina, blue panic and Cochise lovegrass emerged from 30 mm, 30 mm and 20 mm, respectively. Differences between these data and the field data may have been due to higher soil bulk density in the field or differences in germination conditions.

In another greenhouse study (Fig. 8, page 103),

Cochise lovegrass emergence was high from the bare surface of sandy loam soil. For this reason, it is reasonable to assume that many of the seeds found on the surface after the first significant rain storm in the field were buried by the time they germinated and emerged.

Seeds of blue panic and sideoats graina emerged from greater depths in the seedbed than the small-seeded lovegrasses possibly due to greater seed size and more seed reserves. Blue panic seeds are approximately 1 X 2 min and sideoats graina 1 X 5 mm, compared to seeds of

Cochise and Lehmann lovegrass which are approximately 0.5

X 0.75 mm.

The recommended sowing depth for blue panic and sideoats graina is approximately 12 mm (Jordan 1981). The recommended sowing depth for the lovegrasses is approximately 7 mm (Jordan 1981). The results of this

63 study confirm these recommendations.

This study suggests that on coarse-textured soils in southern

Arizona, it may be practical to sow lovegrass seeds on undisturbed seedbeds. Seeds will then either establish on the surface or be buried by summer thunderstorms.

Approximately 90% of the seedlings emerged from seeds that were within 6, 7 , 12, and 14 mm from the soil surface for Lehmann lovegrass, Cochise lovegrass, sideoats graina and blue panic, respectively

(Fig.

5).

These depths were considered the biological limit of emergence for these species in this soil.

Analyses of variance on percent of seeds found after treatment and after rain on the soil surface, and within and below the biological limit produced 2 and 3 factor interactions of year, species and treatment.

In general, greater soil surface disturbance (heavy trampling, land imprinting, and root plowing) buried more seeds, but also placed more seeds below the biological limit

(table

3).

All species generally responded similarly to the treatments with the exception that less sideoats grama seeds were buried than seeds of the other species. This

tr

▪ ▪ r n spzEz

^ 4

13 ...

4 g ;

+

^

0

I:

10

414

IiIIK

SIP

" 4 1 ;* 47,

a

.3

a

n

•• N CO .0

••• tod —•

• •

CO

7 7 • •

N .I. +

. . . .

IA N V N

• • C.

W

N N

. .

0 N Vs

• • CT Cf

42W6W au. •

.

WO ••• IV

••• CO —. 0

. . . .

VI .0 NI

O.

7 • rn

N

0 t's 0 0

+

• • 7 Cf

+ V 0

N 0 V

• ft 7 •

'

42aTIN

7 • 7 fs

4Nag anon

NN

IV0000 nn

' cZoo

6456;v

• •

7 7

,ogoo wo.W6

ma 7.7

CO

N VI

V ' CO C

7a 17 • •

.

CitôgN

.....

f

.

t.

wrfl

V

O.

N

a

.0

• tr n

Fr

wt

.

AZ:;4

.

-46L46%4

• cr a

,14Poog

666k11

,

.

• • •

0

.

Cf

u,QO-..

• a.

• a

tart

NNANN

N

43

.

0..V;A:0

W OO'

• C.1 N

N —• 0 V f • k •

KtraN.

IV 0 C. 5.

er

Fr

• g• ft t7.

N

N

N

666..6

s • cr

W

'

S E

o

3

"

g

0 VI

IN 0

170 • •

tiVR

ara •

N+++

C CO 4.1

C

• •

N a • n n

• -•••

Cf n0 a

N

O CO 12 O.

6666

1

n

100

.

0'

• . .

6

.

la

a- a-

• 7o a

+

V Os O.

rf a

g

N 0

Cf • CI' 7

.1••

N 0

• a

• g n

IV IV IN N

CO Vt

—•

7 7 or o •

Iv WI 14 N

.

N

•••

• .

N N

CO —• + N 0

66:06k, a • • • 7

••

a

N VI

0 V V) 4' vs

• • • •

VS Iod VI 0.

N

O. VI IV CO NJ

• • • •

0 0 —• N

IA N N to1 V CO 0. N io666;..) wgro'gg

7.6:o:46

77170

NI WI VI N

N

•••

n

vi

:•4

' °

°'

V .0. f.

—• Lod 0 0 0

06666

O 01777

N VI 0 WI

.

OC•17000

Cf 0'

NJ

Co CO 0 0 vt

:o:o666

IlIg.C4.77

CO

O

64

65 was probably because these seeds were larger, and most were still contained in spikelets. This prevented seeds from being easily sifted through the seedbed. Heavy trampling was the most successful in burying sideoats grama seeds.

Summer thunderstorms

(20 mm of rain in 1987, 35 mm in 1988) effectively buried more seeds for all treatments, including no disturbance.

In fact, after rain, similar percentages of lovegrass and blue panic seeds were buried within and below the biological limit regardless of the treatment. Rain was less effective in burying sideoats

graina seeds.

Once again, more sideoats grama seeds were buried in the heavily-trampled plots.

Results from the associated field study

(Fig. 7

I

page 80) showed greater seedling emergence on a moderately-wet year

(1988) from plots treated with heavy trampling, imprinting and root plowing than plots not disturbed or treated with light trampling.

On a wet year

(1987) there was similar emergence of small-seeded lovegrasses among all treatments.

On a dry year

(1989) there was limited and similar emergence of all species among all treatments. Greater emergence in heavilydisturbed plots in some cases may have been due to greater seed burial by these treatments. Since many seeds

66 from less-disturbed plots were still on the surface after rain, the fact that nearly all seedlings sampled after emergence came from buried seeds may suggest that seeds remaining on the surface did not germinate or were washed away by rain.

All

4 species tested in this study have shallow biological limits. They also have root morphologies characteristic of the panicoid type. This means that the permanent

(adventitious) roots are initiated from the coleoptilar node which is nearly always within 1 to 2 mm of the soil surface. Adventitious root initiation requires 2-4 days of optimal soil water conditions

(Olmsted 1942, Wilson and Briske 1979. Soil water from the top 3 cm of soil in this study is depleted from -0.03

to

-0.1 MPa matric potential within

1 to

4 days after a rainstorm

(Roundy, unpublished data, Univ. of Arizona).

Seeds buried deeper within the shallow biological limits of these species probably do not have a significantly longer period of available water than those at a more shallow depth under the rapid drying conditions of the

Southwest. Data from an associated greenhouse study

(Figs. 10, 12 and 13; pages 118, 125, and 126) indicated that seedling emergence and primary root lengths of sideoats graina, blue panic, and Cochise lovegrass can

67 decrease with increased planting depth.

This study shows that soil disturbance by cattle trampling or mechanical treatments definitely can bury seeds at a desirable depth for emergence. However, summer rains can bury many small seeds even on unprepared seedbeds. Seed burial increases seedling emergence probably by increasing seed-soil contact and water flow to the seed (Collis-George and

Sands

1959).

Seedling emergence of species with shallow biological limits is dependent on successive storms to maintain available water in the seedbed. Treatments used to help bury seeds would not necessarily be expected to increase seedling emergence of these species in a dry year due to a lack of available water above the biological limit. In a wet year, rain itself will bury many small seeds. However, seedbed treatments increase the numbers of seeds buried and increased seed burial prior to summer rains may help reduce seed predation and thereby increase seedling emergence. Seedbed treatments are probably more essential to help bury large than small-seeded species broadcast on the soil surface.

68

EFFECTS OF CATTLE TRAMPLING AND MECHANICAL SEEDBED

PREPARATION

ON GRASS

SEEDLING EMERGENCE

Introduction

Revegetation of semiarid rangelands is often recommended to restore herbaceous forage plants. However, establishment of seeded species on southwestern rangelands is difficult due to erratic summer precipitation and rapid evaporation of soil water (Cox et al. 1982).

Root plowing and imprinting have been used to control competing shrubs and prepare seedbeds for several years. The imprinter was developed for use on Southwest rangelands to concentrate limited precipitation (Dixon and Simanton 1980).

It has since been shown to help cover broadcast seeds, firm seedbeds, reduce wind erosion and increase seed-soil contact (Haferkamp et al. 1987, Clary

1989).

Effects of imprinting on seedling emergence have been variable.

Imprinted plots on a semidesert grassland in southern Arizona produced fewer grass plants and less production than railed or disk-plowed plots

(Cox et al.

1986). This was true regardless of whether plots were broadcast or drill seeded.

The suggested reason for poor establishment and persistence on imprinted plots was

69 ineffective reduction of creosotebush competition.

Haferkamp et al. (1987) reported that seedling emergence of crested wheatgrass in the northern

Great

Basin was nearly twice as high with imprinting compared to drilling on loose-disked seedbeds, but 2 to 4 times more seedlings emerged from drilling than imprinting on firm unprepared seedbeds.

Clary (1989) reported greater seeded-species density, cover and production from imprinting a burned sagebrush community in the Great Basin as compared to seeding with a rangeland drill. He suggested that the difference was due to increased bulk densities of imprinted soil, decreased wind erosion, and greater seedsoil contact.

Root plowing was initially employed to remove unwanted shrubs prior to cultivation, and later used to control mesquite and other shrubs on rangeland

(Fisher et al. 1959).

Root plowing alone has been shown to decrease productivity of rangeland because of heavy damage to existing grass cover (Fisher 1959,

Ball

1964), although there are exceptions (Allison and Rechenthin 1956, Drawe

1977).

Seeding adapted species following root plowing has

70 generally been shown to increase production over undisturbed seedbeds and other mechanical seedbed preparation treatments. Seeding after root plowing in mesic areas of Texas has produced variable results. In a survey of

1312 root plowed seedings,

Ball

(1964) reported that

50% were considered successful. Allison and

Rechenthin (1956) reported a

10 fold increase in production over nonseeded areas after seeding with buffelgrass.

Carrying capacity was increased

2 to

4 times after rootplowing and seeding with blue panic and buffelgrass on the King Ranch in south Texas (Fisher et al. 1959).

In contrast, Mathis et al. (1971) reported significantly less forage production after root plowing and seeding with native grasses.

Less root plowing has been attempted in the semiarid

Southwest. Jordan and Maynard (1970) compared root plowing with disk plowing and a combination of chaining and pitting in the San Simon valley of southern Arizona.

Results showed similar forage yields between root plowed and disked plots, and both produced approximately twice as much as the chained-pitted plots.

Herbel et al.

(1973) reported the success of a machine developed to root plow, seed, create basin pits and deposit up-rooted shrubs on the seedbed. Of

23

71 seedings in southern New Mexico,

10 had good to excellent stands and 4 had fair stands. A study in the semiarid

Trans-Pecos region of Texas showed good initial plant density of Lehmann and

Cochise lovegrass, and sideoats graina after root plowing and broadcast seeding (Nelson and Gabel

1987).

Recently interest has developed concerning intensive management of livestock and its effect on the soil surface and seedling emergence (Savory

1978). For many years, range managers have suggested the use of livestock trampling to help bury seeds and increase seedling emergence (Plummer et al.

1955, Hormay 1970, Pearson and

Ison 1987, Vallentine 1989). In Nevada, moderate trampling increased emergence of perennial grasses, but decreased emergence of perennial forbs on specific seedbed microsites (Eckert et al. 1986). Heavy trampling decreased perennial grass emergence, but increased emergence of sagebrush and annual forbs on certain microsites. If cattle trampling is to be considered as a viable seedbed preparation technique, comparisons are necessary between trampling and other standard techniques.

The objective of this study was to determine the effects of cattle trampling, land imprinting and root

72 plowing on seedling emergence of Southwest grasses.

Methods

Study Area

The study was conducted on the Anvil ranch, approximately

65 km southwest of Tucson, Arizona. The study site is on the east slopes of the

Baboquivari mountains at an elevation of

1027 m. The soil is a sandy loam (fine, mixed, thermic

Ustollic Haplargid; Sasabe series) and the range site is a loamy upland.

The site is on the edge of the

Chihuahuan semidesert grassland subresource area (Jordan 1981), with major plant species consisting of mesquite

(Prosopis iuliflora Swartz), snakeweed (Gutierrezia sarothrae

Pursh), various cacti, and native grasses including

Arizona cottontop (Digitaria californica (Benth.) Henr.) and purple 3-awn

(Aristida purpurea Nutt.).

The climate is characterized by hot summers and cool winters. Average annual precipitation for the past

30 years, obtained

17.7 km from the site is

328 mm, of which 60-70% falls between July and

November (U.S. Dept.

of Commerce

- Weather Bureau

1957-87).

Treatment Application

The study area was fenced to exclude livestock. All

73 plots to be used in a particular year were treated with a combination of picloram (4-amino-3,5,6-trichloropicolinic acid) and 2,4-D (2,4-dinitrophenyl) to kill snakeweed, and glyphosate (N-(phosphonomethyl)glycine) to kill native grasses. Scattered mesquite trees were cut and removed.

The following 5 seedbed preparation treatments were applied in June or early July (before summer rains) on

6 X 6-m plots: 1) undisturbed control,

2) light cattle trampling,

3) heavy cattle trampling,

4) land imprinting, and

5) root plowing or ripping. Lightlytrampled plots were treated by leading a

300 to

500 kg steer or heifer around each plot until a density of approximately

10 hoof prints per m

2

was obtained.

Heavily-trampled plots were enclosed with an electric fence, and 5, 300 to 500 kg steers or heifers were herded around each plot for approximately

20 minutes. The result was a light, fluffy soil layer approximately 2 to 4-cm deep.

Plots were land imprinted with a Dixon land imprinter composed of

1 directional and 1 nondirectional geometric angle iron forms welded on separate 1 by 1 m cylinder capsules. The capsules were linked together with

74 a common axle and filled with water (Dixon and Simanton

1980).

Total weight of the imprinter was approximately 5 metric tons. The imprinter produced imprints approximately

10-cm deep.

Root-plowed plots were treated with a root plow in

1987 that was pulled through each plot at a depth of

15 to

20 cm. Due to root plow breakage, 1988 and

1989 plots were treated with a 3-tined ripper. The composition of the seedbeds within root plowed and ripped plots was similar. Seedbeds were prepared between 1 and

8

July in

1987, 14 and

17

June in

1988, and

29

May and

5

June in

1989.

Seeds of 'Vaughn' sideoats grama (Bouteloua curtipendula (Michx.) Torr.),

'A-130' blue panic

(Panicum antidotale

Retz.), 'A-68' Lehmann lovegrass

(Eragrostis lehmanniana

Nees), and

'Cochise' Atherstone lovegrass (Eragrostis lehmanniana

Nees X

E. tricophera

Coss and Dur.) were either sown with a handheld

"Cyclone" seeder, or mixed with No. 60 blasting sand and hand broadcast onto all plots at a rate of

1.2 million pure live seed

(PLS) per hectare. All but sideoats grama are introduced to the Southwest, and all have proven to establish well in areas similar to the study site (Jordan

1981, Cox et al. 1982).

All plots with the exception of

75 plowed plots were seeded prior to treatment application.

Plowed plots were seeded immediately after treatment to allow natural soil sloughing to bury seeds (Jordan 1981).

Data Collection

Seedling density data collection was begun when seedlings were approximately

2 to

4 cm tall. This corresponded with 18

August

1987, 1

August

1988 and

22

August

1989.

Each plot was sampled with 1/16 m

2

quadrats uniformly spaced across the plot in 2 transects of several quadrats each. Number of quadrats varied from 20 to

40 per plot for the lovegrasses, and 40 to 80 quadrats for sideoats grama and blue panic. Quadrat numbers varied because of seedling density differences.

To determine effects of seedbed preparation on indigenous forbs, annual and perennial grasses and, bare ground and litter, percent cover data were collected beginning on

15

September

1987, 26

September

1988 and

8

September

1989.

Each plot was sampled with a 10 point pin frame uniformly spaced across the plot in

2 transects of

5 placements each, for a total of 100 points per plot.

Frequency data were derived from density data to determine differences in distribution of the seeded species.

Soil water in the top

1-3 cm was measured in

76 selected undisturbed, imprinted and heavily-trampled plots with Colman soil moisture cells (Colman and

Hendrix

1949) in

1988 and

1989. Precipitation was measured on site with a tipping bucket rain gauge, and soil water and precipitation data were recorded with a

Campbell Scientific Instruments datalogger.

Soil from the top

20 cm of several plots was analyzed for percent water content at

-0.03, -0.10, -0.30 and

-1.50 MPa with a pressure plate.

Experimental Design

The experimental design was a randomized block, split plot. Whole plots were years and seedbed treatments and species were split plots (Table

4).

All factors were assumed fixed. Years were randomized within blocks and each block per year included

20 plots, randomized by treatment-species combination. Because numbers of quadrats or subsamples counted varied among years for some species, seedling density data were analyzed initially for significance of main effects and interactions by calculating approximate

F ratios according to the procedure of Steel and

Torrie (1980).

Significant (P<0.01) year-species-treatment interaction justified analysis of variance and LSD mean separation

(P<0.05) of treatments separately for each species for

(D

/•<

I

1-3

5

(D

0 rt

II

(D

D) rt

Cf)

1

1:$

I-'.

a)

(I) ci

1

1:1

(D

(1)

) 1.4 a)

, ti) ct

(1) sl)

(1)

0 0 11 0

H- a) *

Ha)

(1)

P.<

(D

P)

11

(D si)

11

*

* *

•-3

II

*

(D

P)

rh

5

(D

1-3

1

1 a)

a)

Z ch

rt 5

0 rt

ri)

I

(-I

I

H H

.-...

a)

11

11

0

II

$1 )

II

I

H H

Q.

1-1)

1-3

0)

V 17

11H

P a) (D ti

D) g.

Il.

P)

Il' ci)

H- al

O 5

O Pri

&

I-j

Il als

(D

Di rt Z

• 0

a)

C

rh

(n (1-

())

O b

O H

(D kg1

I

i Ft

P)0 cn

(i)

hi

y) fl a)

(

D a)

rt

a

M

1-

1

/

-

1

H-

5

Z I

-1-

La

Z

Ha)

5 4

(D

Il

(1)

4 H)

(D

H

)

O (D

M 0

a) c t

. 0 .

1

0

l-b

I t r tr V t

Y' tr tr

I su

I

77

78 each year. In those analyses, conventional

F ratios were calculated since the total number of subsamples were equal for a given year and species. Analysis of variance was performed on the arcsin square root of percentage data both for cover and frequency.

Results

1987

Consistent summer rainfall began on

21 July and continued with high frequency and intensity until

25

August (Fig.

6). Although no soil water data were collected during

1987, the surface soil was observed to remain wet from approximately

4 to

28

August. All Species had high emergence except sideoats grama (Fig.

7).

Blue panic had little emergence on undisturbed and lightlytrampled plots, moderate emergence on root-plowed plots and high emergence on heavyily-trampled plots. The lovegrasses had high emergence for all treatments but responded to them differently. Lehmann lovegrass had a trend toward lower emergence on root-plowed plots than on the other treatments.

Cochise lovegrass had greatest emergence on imprinted plots, lower emergence on undisturbed, heavily-trampled and root-plowed plots and least emergence on lightly-trampled plots.

Fig.

6.

Precipitation data for July and August

1987, 1988 and

1989 at the Anvil Ranch, Arizona; and lengths of wet and dry periods in the upper

3 cm of the seedbed in relation to year and treatment. Periods of soil water content

(vol/vol) equal to or greater than

0.09 (>-0.1 MPa matric potential) are indicated by symbols and dry periods are represented by blank spaces between symbols. Heavy

= heavy trampling, control

= undisturbed and imprint

= land imprinting.

2

0

E.

et

12.

1987

60

50 —

40 —

30 —

20 —

10 —

1988

•• •••

III

NM

A AA A111161116A•Ah,

PI

•••

AniddhfigAilliWkidiA

50

40

30

20

10

0

30

20

50

40

10

o

JULY 1 13

1989

• •• ••

• • *****

*4.

ISMII•11

AlA

AIL A11616

RAIN

pit

25

P

AUGUST 6

I

18 30

HEAVY

Fig.

6

CONTROL

IMPRINT

79

Fig. 7. Seedling density of sideoats grama, blue panic, and Lehmann and Cochise lovegrass in relation to year and seedbed treatment. Means within a treatment year with the same letter are not significantly different (p<0.05).

0 o

\

/

\N:-

A.

at tri

N

0

/

..!•:•!

•!•!•!•!•!...!...!•:•!..!...!•!

n

!..!•!

n

wk-mcc

81926

-o

001151,23 m

0 C C rn

0 o o rn

(1)

Ln

N

/

4 t • W...N.M...t•!.•:.

0

...1% OM* o

0 0 o o

as

o o

/

/

./. •

:;:onie•e•

-

•••••••:.**;,

4.!..!...!•!•!..!•!•!.'•.".

n

...- •

Fig.

7

80

81

1988

The

1988 summer rains began on

7

July with a

8.1 mm storm. Thunderstorm activity continued with good frequency until

28

August. Although number of days with precipitation events during July and August was higher in

1988 (15 in

1987, 25 in 1988), storms were generally smaller, therefore the seedbed did not stay wet as long as it did in

1987.

Starting on

17

July, soil water was available in the soil surface for about

6 days, unavailable for a day or less and then available for another

6-9 days (Fig. 6).

Most seedlings emerged during these wet periods. Heavily-trampled and imprinted plots had available water for 1 to

2 days longer than undisturbed plots (Table 5).

Seedling densities were generally higher on heavilytrampled, imprinted, or ripped plots than on undisturbed or lightly-trampled plots (Fig. 7). Sideoats grama density was highest on ripped plots. Blue panic density was highest on ripped and heavily-trampled plots followed by imprinted plots. Lehmann lovegrass densities were similar on ripped, imprinted and heavily-trampled plots.

Cochise lovegrass densities were highest on heavilytrampled plots, followed by imprinted and ripped plots.

The lovegrasses had much higher emergence than blue panic and sideoats grama for all disturbance treatments.

Table 5.

Effect of seedbed disturbance on length of wet and dry periods following summer thunderstorms. Table values are days that soil water content (vol./vol.) was greater than or equal to 0.09 (-0.1 MPa matric potential) in the top

1-3 cm of the seedbed (wet), and days of drying immediately following the wet period (dry).

1989

July

8

10

21

26

29

Aug. 4

6

7

10

14

16

22

Undisturbed

Seedbed treatment

Land imprinted

Wet Dry

Wet Dry

1988

days

July 7 1 1.2

10

1.7

12

17

24

0.3

5.6

1.8

0.1

4.8

1.4

6.6 11.4

0.8

2.1

2.3

4.8

6.7 0.3

6.7 11.3

Aug.

11

14

19

0.7

17.3

2.1

1.2

4.8

11.6

1.7

0.4

1.7

0.2

0.9 10.0

2.7

2.6

2.4 6.1

2.2

0.1

0.3

10.9

0.2

0.1

3.3

0.5

0.2 9.7

1.4

0.6

1.1 10.0

3.2

2.3

2.1

6.3

4.4

1.9

10.3 11.0

1

Dates are day that wet period began.

Heavy

Wet Dry

1.7

2.7

6.8

8.8

1.6

2.9

11.5

1.3

4.2

0.3

9.2

1.2

2.4

1.8

1.4

3.0

2.4

0.1

3.1

0.1

9.6

2.6

0.1

5.7

3.2

4.0 0.2

1.7

0.1

3.8

11.6

82

83

1989

Precipitation events in 1989 were less frequent and intense than in either

1987 or

1988

(Fig. 6).

Summer rains began on

8

July with a

9.1 mm storm and total precipitation for July and August was only

107 mm.

Periods of available water were shorter in 1989 than in

1988.

Wet periods during emergence were approximately 6 to

9 days and

2 to

4 days long in 1988 and

1989, respectively (Table

5).

Heavily-trampled and imprinted plots had similar periods of available water in the surface soil that were up to 12 hours longer than in undisturbed plots. The short periods of available soil water in 1989 probably caused the limited seedling emergence of all 4 seeded species. Emergence was low regardless of the seedbed treatment (Fig. 7). Ripping slightly increased emergence of sideoats grama and

Cochise lovegrass over that of the other treatments.

Analysis of variance for both bare ground and litter cover data produced significant treatment and year effects but the treatment-year interaction was not significant (P<0.05).

Root plowing or ripping and heavy trampling resulted in more bare ground than the other treatments (Table

6).

More bare ground was exposed in

1989 than in

1988

(Table

7).

Heavily trampled, imprinted and root-plowed or ripped plots had significantly less

84

Table 6.

Effects of seedbed treatment on percent cover of bare ground, litter and 4 indigenous grasses in relation to seedbed treatments.

Undisturbed Light trampling

Treatment

Heavy trampling

Imprint

Root plow

Bare ground

Litter

Annual lovegrass

Feather fingergrass

Six weeks grama

Rockroth grama

(seedlings)

36 b

1

27 a

5a

3a

3a l a

36 b

27 a

7a

4a

2a

O a

52 a

16 b

8

5

2 a la a a

41 b

19 b

9

5

2 a

Oa a a

1

Means in a row followed by the same letter are not significantly different (p<0.05).

55 a

15 b

7 a

6

2 a a la

Table 7. Percent cover of bare ground, litter and 4 indigenous grasses in 1987, 1988 and 1989 on a loamy upland range site in southern Arizona.

Year

1987 1988

1989

Bareground

Litter

Annual lovegrass

Feather fingergrass

Six weeks grama

Rockroth grama (seedlings)

45 ab l

12 b

10 a

5a

2a la

33 b

16 b

11 a

9 a

4 a

O b

54 a

33 a

O b

Ob l a la

1

Means in a row with the same letter are not significantly different

(p<0.05).

85

(p<0.05) litter cover than lightly trampled and undisturbed plots. There was significantly more litter in

1989 than in the other 2 years, probably as a result of a gradual build-up of litter associated with cattle exclusion.

Several indigenous perennial and annual grasses, forbs and shrubs were identified during cover data collection, but results from only

4 were analyzed due to rarity of the other species. The 4 species included an annual lovegrass (Eragrostis arida Hitchc.), feather fingergrass (Chions virgata Swartz), sixweeks needle grama (Bouteloua aristidoides (H.B.K.) Griseb.), and

Rockroth grama (Bouteloua rothrockii Vasey). Indigenous species cover analysis produced no significant treatment effects

(Table 6), however, there was a year effect for the annual lovegrass and feather fingergrass (Table 7).

There was less annual lovegrass and feather fingergrass in 1989 than in 1987 and 1988. The lack of difference among treatments probably reflects the ability of these species to germinate and establish in a variety of seedbed micros

ites.

There was a significant interaction between year and treatment for frequency of seeded species. Seedlings were most evenly distributed on seedbeds prepared by heavy

86 trampling, imprinting and root plowing or ripping (Table

8). Distribution was moderate on undisturbed and light trampled plots in 1987, and limited in 1988 and

1989.

Distribution across treatments was greatest for the lovegrasses, followed by blue panic and then sideoats grama.

Discussion

Differences in seedling emergence among years and treatments were highly related to precipitation patterns and periods of available water. Seedling emergence was highest for all species except sideoats grama in

1987 when surface-soil water was estimated to be available for about 24 consecutive days. The lovegrasses had high emergence from all treatments on that year while blue panic had greater emergence on the more disturbed seedbeds. The low density of sideoats grama in that year may be related to its rapid germination

(24-48 hours --

Simanton and Jordan

1986) and possible desiccation during the 12-day drying period after initial rainstorms (Fig.

6).

The other

3 species require a longer period of available water to germinate and apparently only germinated in early August during consistent rainfall.

The greater emergence of the small-seeded lovegrasses compared to the larger-seeded blue panic on

Table

8.

Percent frequency of sideoats grama, blue panic, and Lehmann

Cochise lovegrass in relation to year and seedbed treatment.

Sideoats grama

Species

Blue panic Lehmann lovegrass

Cochise lovegrass

Treatment

1987

Undisturbed

Light trampling

Heavy trampling

Imprint

Root plow

1988

Undisturbed

Light trampling

Heavy trampling

Imprint

Root plow

1989

Undisturbed

Light trampling

Heavy trampling

Imprint

Root plow

0

0

2

1

1 a a a

1 a a

2 bc

3 b

13 b

11 b

30 a

3

3

1 b

1 b

6 ab ab a

11 c

12 c

74 a

66 a

37 b

13 b

9

4

16

20

18

10 b

48

34 a a

53 a a a a a a

64

61

71

81

11 b

21 b

64 a

51

61 a a ab ab ab a

50 b

5 b

9 b

39

15

36 a ab a

1

Means in a column within a year with the same letter are not significantly different (p<0.05).

63 b

63 b

61

90

75 ab a ab

12 c

36 bc

82

56

55 a ab ab

3 b

5 b

21 a

8 b

45 a

87

88 the undisturbed plots and the greater

emergence

of blue panic on the more disturbed seedbeds in 1987 may have been due to greater seed-soil contact. Heavy trampling, land imprinting and root plowing buried more seeds than non-disturbance or light trampling (Fig.

3, page

57).

However, many seeds were buried by summer rains, even on undisturbed and lightly-trampled plots. Seedling emergence from surface-sown seeds on bare soil in a greenhouse study was much greater for Cochise lovegrass than for sideoats grama and blue panic (Fig.

8, page

103). A lower trend in emergence of Lehmann lovegrass on root-plowed plots in 1987 may have been related to excessive seed burial. Root plowing buried at least 60% of the Lehmann lovegrass seeds found below the biological limit of emergence (Table 3, page

64).

Greater seedbed disturbance by heavy trampling, land imprinting or ripping produced greater seedling emergence than no disturbance or light trampling on a moderatelywet year

(1988).

That was the only year that sideoats grama produced an acceptable stand of seedlings, and this occurred on the ripped plots. Ripping buried twice as many sideoats grama seeds within the biological limit of emergence as did the other treatments (Table 3, page 64).

Successful emergence of sideoats grama on this year was probably related to more consistent initial rainfall and

89 soil water availability at the start of the rainy season than occurred in 1987 or

1989.

The greater seedling emergence of all species on the more disturbed seedbeds in

1988 was probably related to greater seed burial on these seedbeds, and a slightly longer period of available water, at least for the imprinted and heavily-trampled plots. In 1988, there were significant correlations between percentage of seeds buried immediately after treatment and seedling emergence for blue panic (r

2

=

0.99, P = 0.003),

Lehmann lovegrass (r 2 = 0.99,

p = =

0.006) and

Cochise lovegrass (r 2

= 0.97, P = 0.017).

There were also significant correlations between percentage of seeds buried after a

35-mm rain and seedling emergence for sideoats grama (r 2 = 0.99, P =

0.001) and blue panic (r

2

= 0.80, P = 0.04).

None of these correlations were signficiant (P < 0.05) in

1987 when soil water was available for a larger period of time than in

1988.

Low emergence on all treatments in 1989 was related to inconsistent rainfall and only

2-4 days of available water in the surface soil in July. More consistent rainfall and longer periods of available water occurred later in August, but these did not produce a cohort of seedlings. Seeds may have initiated germination after

90 initial rains and desiccated during the subsequent dry periods in July.

Emergence of indigenous annual and short-lived perennial grasses was apparently affected much more by precipitation patterns than seedbed treatments. These grasses were not observed to emerge until after a period of consistent rainfall. Evidently they are adapted to emerge from a variety of seedbed microsites as long as soil water is available.

Analysis of past seedings is difficult because daily precipitation, temperature, humidity (Cox et al.

1982) and soil water availability are seldom measured. This study documents the fact that seedling emergence is highly related to the pattern of soil water availability.

With the exception of sideoats grama, seedling density of all species was highest during a wet year

(1987), intermediate during a moderately-wet year

(1988) and limited during a dry year

(1989).

_Greater seedling emergence from more disturbed seedbeds was most evident on a moderately-wet year. A slightly longer period of soil water availability and increased seed burial associated with greater seedbed disturbance by heavy trampling, imprinting, root plowing or ripping help explain their higher seedling emergence compared to non-

91 disturbance and light trampling. In this study, increased seedling emergence was more highly associated with increased percentage of seeds buried on a moderately-wet year than on a wet or dry year. Seed burial may increase emergence by increasing seed-soil contact and water flow to the seed

(Collis-George and

Sands

1959), increasing radical penetration (Dowling et al.

1971,

Cox and Martin 1984) and reducing predation

(Nelson et al.

1970,

Campbell and Swain 1973).

Seedling establishment of small-seeded lovegrasses may not require seedbed preparation on coarse-textured soils during wet years in southern Arizona. Sideoats grama and blue panic appear to require seedbed preparation for seedling establishment. Heavy trampling, imprinting and root plowing or ripping all provided adequate seedbed preparation for burial and emergence during wet or moderately-wet years. During dry years these seedbed treatments would not be expected to increase seedling emergence because the seedbed dries out rapidly without recurrent rainfall.

92

INFLUENCE OF SEEDBED MICROSITE

CHARACTERISTICS

ON GRASS SEEDLING EMERGENCE

Introduction

Perennial grass densities have decreased on many rangelands due to lack of grazing management, drought

(Cox et al. 1982), and encroachment of woody shrubs

(Humphrey 1958, Cox et al. 1983). Loss of herbaceous cover often results in increased soil erosion, formation of soil crusts and loss of favorable microsites necessary for germination and establishment of grass seedlings on some sites.

Natural revegetation of perennial grasses is limited in many areas (Hyder et al. 1971, Stoddart et al. 1975,

Roundy and Jordan 1988), therefore seeding adapted species is often desirable. Because of steep or rocky terrain, conventional methods of tilling and seeding are often impractical. On these sites the only alternative is broadcast seeding (Vallentine 1989). However, broadcast seeding is often unsuccessful on many sites (Cook 1958,

Bleak and Hull 1958, McWilliam et al. 1970, Campbell and

Swain 1973A, Stoddart et al. 1975).

Depending upon the species, most germination requirements are more easily met within the seedbed rather than upon it. Surface-sown seeds are exposed to

93 many hazards including predation (Howard

1950, Trevis

1958,

Nelson et al.

1970, and Campbell and Swain

1973A), seed and seedling desiccation (Evans and Young

1970,

Nelson et al.

1970,

Dowling et al.

1971,

Evans and Young

1972a, and Campbell and Swain

1973a), and lack of radical entry (Dowling et al.

1971,

Evans and Young

1972b,

Campbell and Swain

1973a,

Campbell and Swain

1973b,

Sheldon

1974, and Cox and Martin

1984).

Surface-sown seeds fall into a variety of microenvironments, some of which are associated with favorable water and temperature conditions for germination and establishment. Favorable microsites for seedling establishment have been described by Harper et al.

(1961) and Harper

(1977) as

"safesites".

Natural safesites may include cracks and depressions in the soil surface, or sites associated with gravel, and plant litter.

In a classic study, Harper et al.

(1965) reported increased emergence from surface-sown seeds of

Plantago spp.

placed in soil depressions, and near various objects including glass sheets and wooden boxes. They also reported increased emergence from seeds of

Bromus rigidus

(Roth.) when sown in large soil aggregates, and B.

madritensis (L.) when sown in small soil aggregates. They

94 explained the differential emergence in terms of the different types of contact the seeds made with the aggregates. In a similar study, also using various artificial objects, Sheldon

(1974) showed that treatments imposing a high humidity over the treatment area had better emergence than those that did not.

Apparently one of the most important factors that identifies a microsite as a safesite is its ability to provide adequate soil water for germination and growth.

Harper et al.

(1965) and Sheldon

(1974) both worked under mesic conditions in Great Britain. On arid and semiarid rangelands where storms are less frequent and seedbeds may rapidly dry out, microsites may still be able to increase emergence. Evans and Young

(1972a) reported 100 times greater emergence of downy brome (Bromus tectorum

L.) seedlings from seeds sown in

9-cm soil pits as compared to seeds sown on the bare soil surface in

Nevada. Soil water and humidity in the pits were much more favorable for germination than the bare soil surface. In a survey of plant responses to various microsites in the Great Basin, Eckert et al.

(1986) found more seedlings of Wyoming big sagebrush (Artemesia tridentata wyomingensis Beetle), perennial grasses and annual forbs emerging from cracks and trenches than on

95 the unprotected soil surface.

Evans and Young

(1970, 1972b) reported much greater emergence of downy brome, medusahead rye

(Taeniatherum asperum (Simonkai)

Nevski), and Russian thistle

(Salsola iberica

L.) from seeds under litter than from seeds on the soil surface. Environmental monitoring showed that litter greatly moderated maximum and minimum temperatures, decreased the amount of light reaching the soil surface and greatly increased relative humidity.

In the semidesert grasslands of the Southwest,

60-

70% of annual precipitation falls between July and

October. At certain times during this period, conditions may be mesic enough for natural seedbed microsites such as gravel, litter and cracks; and artificial microsites produced by livestock trampling and mechanical seedbed preparation to be more favorable for germination than the bare soil surface. This may help to explain why smallseeded lovegrasses broadcast on disturbed seedbeds have had equal or greater emergence as those drilled in the

Southwest (Cox et al.

1986).

Other then the few studies above, there has been little determination of the types of microsites that are adequate for germination and establishment of range plants.

96

The objective of this study was to determine the effects of various natural microsites typical of the semidesert grasslands in Arizona on seedling emergence in relation to soil water conditions for

3 warm-season grasses.

Methods

A field survey was undertaken at a typical semidesert grassland site on the Anvil Ranch,

65 km southwest of Tucson, Arizona to determine what types of microsites may be available to seeds on rangelands in southern Arizona. The soil is a sandy loam (fine, mixed, thermic

Ustollic Haplargid) and supports mesquite

(Prosopis luliflora Swartz.), snakeweed (Gutierrezia sarothrae Pursh), and remnant native grasses such as

Arizona cottontop (Digitaria californica (Benth.) Henr.) and purple three awn

(Aristida purpurea

Nutt.).

Three obvious microsites were gravel, cracks in the soil crust and litter. To determine the percent composition of the gravel microsites, 2 m

2

of the top

1 cm of the soil surface was collected, sieved and each fraction weighed. Of the sample,

45% was gravel, 39% was sand and

15% was silt and clay. Of the gravel fraction,

66% was between 2 and

5 mm, 21% was between

5 and 10 mm

97 and

13% was greater than 10 mm.

Percent litter cover was determined with a point frame. Of

1600 points,

281 hit litter; thus litter cover equaled

17.6%.

Litter was composed of dead stems of half shrubs and grasses and located mainly beneath living plants and in litter dams. Cracks were common but most were less than 0.5 mm in width.

All

3 microsites described above plus a control consisting of bare uncracked soil were selected for use in the study, with the following modifications. The gravel treatment was a mixture of both the 2-5 and the

5-

10 mm fractions and each was present in the same percentages found in the field. Due to the lack of an uncontaminated litter source and lack of homogeneity of litter at the survey site, fine wood excelsior was used.

Crack size was increased to

2-5 mm wide by 15-20 mm deep.

The study was conducted in a temperature-controlled greenhouse with average minimum and maximum temperatures of

17 and

46 degrees

C, respectively

.

Percent relative humidity ranged from

4 to

72%.

Soil from the A horizon (top 20 cm) of the microsite survey area was sieved to

5 mm and placed in

10x10x10-cm square pots. All soil in the pots was then subjected to 2

98 wet-dry cycles to simulate field soil conditions. The microsites were then applied as follows. Three cracks

2-5 mm wide, 15-20 mm deep and 80-100 mm long were formed in crack treatment pots with a knife in moist soil. Twentyfive cm

3

of gravel from the survey site was added to the soil surface of the gravel pots. This amounted to a layer approximately

10-mm deep. Wood excelsior was cut into 20-

50 mm lengths and placed on the litter treatment pots to a depth of 10-20 mm. Bare surface pots were left untouched.

Twenty-five seeds each of 'Vaughn' sideoats grama

(Bouteloua curtipendula (Michx.) Torr.), 'A-130' blue panic (Panicum

antidotale

Retz.) or

'Cochise' Atherstone lovegrass (Eragrostis lehmanniana Nees

X

E. tricophera

Coss and Dur.) were sprinkled onto the soil surface of pots with gravel, litter or control microsites.-Seeds were sown beneath rather than on gravel and litter to simulate field conditions. Pots with cracks were sown using a thin funnel to carefully place seeds near the bottom of cracks.

Three watering treatments were implemented, including

1) daily watering to maintain field capacity,

2) watering on day

1 and day

5 (when seedlings showed signs of stress), and

3) watering on day

1 only. These

99 conditions were selected to simulate the wide range of soil water conditions that may occur during the rainy season in southern

Arizona. Watering was accomplished by subirrigation to minimize microsite disturbance.

The experimental design was a split plot with 4 blocks. Whole plots were water treatments, and microsites, species and days were split plots (Table 9).

All factors were assumed fixed. Each block contained

72 pots;

3 species,

4 microsites and 3 water treatments with

2 pots per species-microsite-water treatment combination.

To determine why seeds and seedlings responded as they did to the microsites, gravimetric soil water content was measured for samples from additional unseeded pots. These pots were filled with soil, conditioned with

2 wet-dry cycles, irrigated and treated with the microsites as were the seeded pots.

Beginning on day 1 and continuing daily throughout the study, approximately 10 cm

3

of

soil from the top 5 mm of mineral soil from the bare surface, gravel and litter microsites was collected from the unseeded pots.

Samples from the pots with the crack microsite were collected by removing the soil core from the pot, carefully splitting the core lengthwise down

1 crack and then scraping approximately

10-15 cm

3

of soil from the bottom of the

Table

9. Sample ANOVA table for determining the influence of seedbed microsite characteristics on grass seedling emergence.

Source df Error term used

Block (r)

Water treatment

(w)

Error a

Species (s)

Microsite (m)

Species * Microsite

Water * Species

Water * Microsite r-1 w-1

(r-1) (w-1) s-1 m-1

(s-1)(m-1)

(w-1) (s-1)

(w-1) (m-1)

Water * Species * Microsite

Error b

Day

(d)

Error c

Day * Water

(w-1)(s-1)(m-1) by subtraction d-1

(d-1)(b-1)

(d-1)(w-1)

Day * Species

Day * Microsite

Error d

Day * Species * Microsite

Day * Water * Species

Day * Water * Microsite

(d-1) (s-1)

(d-1)(m-1)

(d-1)(r-1)(w-1)

(d-1)(s-1)(m-1)

(d-1)(w-1)(s-1)

(d-1)(w-1)(m-1)

Day * Water * Species * Microsite (d-1)(w-1)(s-1)(m-1)

Error a

by subtraction

Total rwsmd-1

e e d d d e e a b b

b b

c

b

b

-

100

101 crack (between 10 and 20 mm from the soil surface).

Gravimetric soil water content of the samples was determined by weighing before and after oven-drying

(Hillel 1982). This method was selected because of the necessity of sampling the small interface between the soil and seed. Soil used in the study was analyzed for percent water content at

-0.03 and -1.50 MPa with a pressure plate.

Seedling emergence data were collected daily beginning on day

4. Seedling emergence criteria per microsite were as follows:

1) bare surface, seedlings >15 mm and radicles penetrating the soil surface;

2) gravel and litter, seedlings above gravel or litter; and 3) cracks, seedlings even with soil surface.

Temperatures within microsites were measured with a probe-type thermocouple thermometer from day

1 until day

4.

The temperature probe was placed between gravel or litter and the soil surface, within cracks, or directly on the soil surface. Air temperature was also measured

1 cm above the soil surface.

Seedling density data were converted to percent emergence of germinable seed.

Germinability was determined in the greenhouse with 4 replications of each species using petri dish techniques. For each

102 replication, 100 seeds were placed on

Whatman

No.

2 filter paper in a petri dish. Seeds were then watered and the dish was placed in its respective block. Germination counts were made every 24 hours (+ or

- 1 hr) after initial watering. A seed was considered germinated when its radicle had emerged. Analysis of variance was performed on the arcsin square root of percentage data as suggested by Sokal and

Rohlf (1981) for seedling emergence and soil water content.

Seedling Emergence

Results and Discussion

Analysis of variance of seedling density data showed significant

(P < 0.05) 2, 3 and 4 factor interactions involving days, species, water treatments and microsites, therefore, results are based on comparisons of dayspecies-water treatment-micros ite means.

In general, under conditions of daily watering

(matric potential > -0.03 MPa), all microsites produced significantly more seedlings (P< 0.05) than the bare surface (Fig.

8). Emergence of all species was highest from gravel, followed by litter, then crack microsites. Cochise lovegrass was the only species to emerge from the bare surface.

Fig. 8. Seedling emergence of sideoats grama (row 1), blue panic (row 2) and Cochise lovegrass (row 3) in various seedbed microsites in relation to 3 soil water treatments. Column 1 = daily watering, column

2 = watering on days 1 and 5, and column 3 = watering on day 1 only. Lines within a species-water treatment graph having the same letter are not significantly different (P<0.05).

+

A

21)

>

CI

X

0

0 xi

>

<

I

nnsa.

MIIIIPX

, one)

,

1111111

n

10 flOUIS.

newma.

velars

, enter onces

PERCENT SEEDLING EMERGENCE

}

103 anger caws.

onsir oflOIN

CltIMP oasis.

11•111•3.

ff

OWN

(11113.2.

flui

MO*

i

Fig. 8

104

Watering on days 1 and 5 resulted in lower seedling emergence than under conditions of daily watering. Only

Cochise lovegrass emerged from the bare surface.

Emergence of the different species varied relative to gravel, litter and crack microsites. Sideoats grama emergence was similar for gravel and crack microsites during the first few days, but thereafter, seedlings in cracks began to die. Seedling emergence was generally highest from gravel, less from cracks, and least from litter microsites. Blue panic seedling emergence was similar from gravel and litter microsites.

Cracks produced the most blue panic seedlings, but after day 9, seedlings began to die. Cochise lovegrass emerged best from gravel. Litter and cracks produced similar seedling emergence, but seedlings in cracks began dying after day 9.

Premature mortality of seedlings in cracks may have been due to the method of crack preparation. The action of the knife in moist soil may have sealed the crack surfaces. An examination of dead seedlings from cracks showed that radicles of many seedlings had not penetrated crack walls. Natural cracks appear to have more large pores and rough edges for easier radicle penetration. In a more natural situation, seedlings may have survived longer in cracks.

105

Sideoats grama seedlings had the highest seedling emergence from gravel when watering occurred on day

1 only. Cracks also produced nearly as many seedlings as gravel, until seedlings began dying prematurely. All microsites produced more seedlings than the bare surface.

Results from the petri dish trial showed that blue panic requires

3-4 days of wet conditions for germination as opposed to 1 day for sideoats grama and 1-2 days for

Cochise lovegrass.

Watering on day

1 only did not appear to meet the germination requirements for most blue panic seeds within any of the microsites, although emergence was significantly higher (P<0.05) in gravel sites than in the other sites for

3 days.

Cochise lovegrass had high seedling emergence from gravel even with watering only on day 1. Litter and cracks had significantly more

(P<0.05) seedling emergence, than the bare surface. Once again, there was premature mortality among seedlings in cracks.

Soil Water

The analysis of variance for soil water data showed significant

2

and

3 factor interactions involving days, water treatments and microsites; therefore discussion of results is based on day-water treatment-microsite means.

106

Water content of soil watered daily was greater than a corresponding matric potential of -0.03 MPa throughout the study for all microsites

(Fig.

9).

Soil water content of soil watered on days 1 and 5 stayed above a corresponding matric potential of -1.5 MPa in the gravel and litter sites until day

9, but only until days

4 and

5 for the bare surface and crack sites, respectively. Although soil water content increased above a corresponding matric potential of -1.5 MPa in cracks and on the surface after watering on day 5, the

1 or 2 days below

-1.5 MPa may have caused seed and seedling desiccation. This may help explain the significantly greater seedling emergence within the gravel and litter sites. Water content within the gravel and litter sites stayed above -1.5 MPa for 1 or 2 days longer than in the crack and bare surface sites in soil watered on day

1 only.

The soil water data do not explain the limited seedling emergence of sideoats grama and blue panic under high matric potential on the bare surface site.

Observations showed that not only was there low emergence, but very little germination occurred on these sites. The greater emergence of

Cochise lovegrass (smallseeded) compared with sideoats grama and blue panic

Fig. 9. Percent soil water in relation to microsites across time (days). Lines indicating soil water content at soil water potentials of -0.03 MPa and

-1.50 MPa have been added for reference. Water 1 = daily watering, water 2 = watering on days 1 and 5, and water 3 = watering on day 1 only. Lines within a water condition graph having the same letter are not significantly different (P<0.05).

20

15

10 o

20

Fig.

9

Water 1

-0.03 PIPs

-1.50 1CPIL

107

20

0 suHace gravel

Days o

-0.03 MPs

-1.50 MP'

9 10

¶2

A crock

108

(larger seeded) on the bare soil may be due to greater seed-soil contact. Evaporative demand may have resulted in greater loss of water from the seeds than flow of water from the soil to the seeds.

The ideal seedbed is one where the seed is surrounded by soil particles that are firmly packed around the seed to ensure conductivity of water from soil to seed

(Collis-George and Sands

1959).

Seeds on the surface have varying amounts of contact with soil. Harper

(1977) stated that the degree of heterogeneity of the soil surface is dependent upon seed size. This may explain why small seeds appear to require less soil tilth or surface roughness to meet their safesite requirements

(Nelson et al. 1970,

Evans and Young

1972a,

Cox and

Martin

1984).

In the associated field study, (Fig. 7, page

80), seedling emergence of Lehmann and Cochise lovegrass in undisturbed plots with an abundance of bare soil surface was as high as in plots disturbed by cattle trampling, land imprinting and root plowing following frequent summer rains. This was probably due to slight burial of seeds after summer rains.

The soil water results do not explain the greater

Cochise lovegrass seedling emergence from gravel, since

109 soil water content under litter was similar to or higher than under gravel (Fig.

9).

In actuality, soil surface water and relative humidity in the immediate vicinity of the seeds may have been greater under gravel than litter.

The technology to measure these conditions at the scale of small seeds is not yet available (Harper et al. 1965).

Another explanation may be found in the contact of seeds to gravel. Gravel appeared to lay on or next to seeds, as opposed to litter which seldom seemed to be in contact with the small seeds. The position of the gravel may have improved emergence by providing points of contact for the seed to push against during radicle penetration.

The range of temperatures between microsites on any one day was never greater than 4.2

C. Differences in seedling emergence are probably not related to temperature differences.

This study showed that under high soil matric potentials (> 0.03 MPa), gravel, litter and cracks produced significantly more seedlings of sideoats grama, blue panic and Cochise lovegrass than the bare soil surface. However, results from watering on days 1 and

5 are probably more indicative of Southwest rangeland conditions during the summer rainy season. Under these conditions, the microsites also produced greater seedling

110 emergence than the bare soil surface. Of course, the emergence of seedlings from different microsites is highly dependent on the actual period of available water in a site for a given set of meteorological conditions.

Cochise lovegrass emerged well in gravel microsites under a variety of soil water conditions. Thus it may be practical to broadcast seed small-seeded lovegrasses onto gravelly rangeland soils with little or no seedbed preparation.

The ability of small-seeded lovegrass seeds to emerge from undisturbed seedbeds with gravel surfaces may help to explain their success in spreading and dominating coarse-textured soils in southern Arizona (Cable 1971,

Cox and Ruyle 1986, Cox et al. 1988).

111

EFFECTS OF SOWING DEPTH AND SOIL WATER ON EMERGENCE

AND SURVIVAL OF GRASS SEEDLINGS

Introduction

Reseeding efforts on semiarid rangelands may be unsuccessful due to seed predation (Nelson et al. 1970,

Campbell and Swain 1973), competition (Dowling et al.

1971), prevention of radicle penetration and seedling emergence by soil crusts (Dowling et al.

1971,

Taylor

1971), short periods of available soil moisture (Herbal et al.

1973,

Hauser

1986), and excessive sowing depths

(Mutz and

Scifres 1975, Vallentine 1989).

Seedbed preparation is often recommended on rangelands to increase seed burial in order to increase soil water contact and availability to seeds (Jordan

1981, Vallentine 1989).

Depending on the nature of seedbed disturbance, seeds may end up at a variety of depths. Seeds of small-seeded grasses broadcast on undisturbed seedbeds fell into cracks or were buried at shallow depths by natural movement associated with summer rainstorms in southern Arizona (Figs.

3 and 4, pages

57 and

60).

Mechanical seedbed preparation or cattle trampling after broadcast seeding increased the depth of seed burial and the percentage of seeds buried above and below the biological limit of emergence (that depth

112 beyond which seeds can emerge due to lack of energy reserves), (Table 3, page

64).

Following a summer rain, the soil drying front proceeds from the surface downward at a rate dependent on cloud cover and incident radiant energy available for evaporation. It could be hypothesized that for certain drying rates, seeds buried near, but not below the biological limit may have a longer period of available water for germination and seedling growth than seeds at more shallow depths.

Deep planting has both increased and decreased emergence of grass seedlings. Under optimal water conditions, emergence of most grass species is decreased with increasing sowing depth (McKenzie

1946, Mutz and

Scifres 1975,

Cox and Martin

1984, Fulbright et al.

1985,

Carren et al.

1987; and Newman and Moser 1988).

Blue graina seedling emergence was greater from

2 cm than

1 cm under marginal water conditions

(Carren et al. 1987).

Apparently soil water potential at 2 cm was high enough for germination and growth for a longer period than at 1

CM.

Effects of sowing depth on primary root development have been variable

(Tadmor and Cohen

1968,

Cornish 1982,

Fulbright et al.

1985). Tischler and

Voigt (1983)

113 reported that primary root depth of

Wilman lovegrass

(Eragrostis superba

Peyr.) is independent of sowing depth, primary root depth of Klein grass (Panicum colortum L.) increases with sowing depth, and deep sowing decreases primary root penetration of weeping lovegrass

(Eragrostis curvula

(Schrad.) Nees.). If primary root lengths increase with or are unaffected by sowing depth, then primary root depth will increase with sowing depth.

Seedlings with deeper roots may survive longer during drying cycles than seedlings with shallower roots.

The objective of this study was to determine the effects of sowing depth and soil water on emergence and survival of southwest grass seedlings.

Methods

The study was conducted in a temperature-controlled greenhouse with average minimum and maximum temperatures of 31 and 45 degrees

C, respectively. Relative humidity ranged from

5 to

78 %.

Relative humidity may range from

7 to 100 % and air temperature from 16 to 40 degrees

C in the semiarid grassland in southern Arizona during the summer rainy season (Roundy, unpublished data, University of Arizona).

The grass species used were 'Vaughn' sideoats graina

114

(Bouteloua curtipendula (Michx.) Torr.),

'A-130' blue panic (Panicum

antidotale Retz.), and 'Cochise'

Atherstone lovegrass (Eragrostis lehmanniana Nees

X

E.

tricophera

Coss and Dur.).

The experimental design was a randomized block, split plot with

4 blocks. Whole plots were soil water treatments, and split plots were time of sampling, sowing depth and species. Each block contained

72 pots:

3 species,

4 planting depths and

3 soil water treatments with

2 pots per species, depth, water combination.

Soil was obtained from the Anvil ranch,

65 km southwest of Tucson, Arizona. The soil is a sandy loam

(fine, mixed, thermic

Ustollic Haplargid). Soil from the

A horizon (top

20 cm) was sieved to 5 mm and placed in

10-cm

3 pots to a depth of 0, 10, 20, and

30 mm from the pot surface in pots for sideoats graina and blue panic, and

0, 5, 10, and

15 mm in pots for Cochise lovegrass.

Twenty-five seeds were placed on the soil surface for each depth and then covered with sieved soil to the pot surface. Primary root lengths of seedlings were measured in relation to water treatment and sowing depth on seedlings of all

3 species grown in tapered plastic cones

3.8 cm in diameter by

20-cm long. Sowing depth and water treatments were identical for pots and cones. Each block

115 contained 216 cones, 3 species,

4 sowing depths,

3 water treatments and 6 sampling days. Each species-sowing depth-water treatment combination included

4 cones with

3 seedlings sampled per cone. Seedlings were excavated by removing the soil core and washing the soil from roots with a fine water spray. Primary root lengths were measured from the seed to root tips.

Three soil water treatments included: 1) watering every

3 days to maintain high matric potentials

( > -0.03 MPa) throughout the experiment,

2) watering on days

1 and

7, and

3) watering on day

1 only. These conditions were selected to simulate the wide range of moisture conditions that may occur during the rainy season in southern Arizona. Watering of seedlings under the first watering treatment was discontinued on day 17 in order to determine seedling survival. Watering was accomplished by subirrigation to minimize disturbance of surface-sown seeds.

Seedling density data were collected daily beginning on day

3.

Seedlings of surface-sown seeds were considered emerged if they were

1-cm tall and radicles penetrated the soil surface. Seedlings from buried seeds were considered emerged when they were

1-cm tall.

116

Seedlings were judged to be dead if leaves were rolled, and brittle. Survival data were calculated as a percentage of maximum seedling density for a pot.

Soil water content was determined gravimetrically from samples in additional unseeded pots. Beginning on day 1 and then daily throughout the experiment, 1 pot per water treatment per block was randomly selected. The soil core from each pot was carefully removed and approximately 3 cm

3

of soil was sampled from 5-mm depth intervals to a depth of 30 mm. Gravimetric soil water content of the samples was determined by weighing before and after oven-drying

(Hillel 1982).

Beginning on day

9 of the experiment, soil sampling intervals were widened to

10 mm in order to sample the deeper soil in contact with extending roots. Soil used in the study was analyzed for gravimetric water content at matric potentials of

-0.03 and

-1.50 MPa with a pressure plate.

Relative humidity and air temperature were monitored using a Phys-Chemical Research PCRC-11 relative humidity sensor and a Fenwal Electronics UUT51J1 thermistor, respectively. Data were collected each minute and 30-minute averages were recorded with a Campbell

Scientific Instruments

CR10 datalogger.

Analyses of variance were performed on maximum

117 seedling density, percent survival of seedlings and soil water content.

Results and Discussion

Maximum

Emergence

There were significant

(p<0.05)

2-factor interactions for maximum seedling density between species, sowing depth and water treatment.

Seedling emergence for all

3 species was generally greatest from surface-sown seeds and decreased with increasing planting depths (Fig.

10,

Table

10).

Seedlings of all species emerged from all depths, except for sideoats grama which had low emergence at

30 mm.

Under continuously-wet soil (watering every 3 days), all species generally produced many more seedlings from surface-sown than buried seeds. Very few sideoats grama seedlings emerged from buried seeds under continuouslywet soil. In contrast, many more sideoats grama seedlings emerged from buried seeds when watering occurred on days

1 and

7 or day

1 only (Fig.

10).

This difference may have been due to inadequate gas exchange for seeds watered every

3 days.

Watering on days

1 and

7 resulted in slightly less emergence than watering every

3 days (Fig.

10,

Table

Fig. 10. Seedling density of sideoats grama (row 1), blue panic (row 2) and Cochise lovegrass (row 3) across time in relation to sowing depth and watering treatment. Column 1 = watering every 3 days, column

2 = watering on days 1 and 7, and column 3 = watering on day 1 only. Watering days = "*".

Standard error ranges are: sideoats grama and blue panic, 0.84 to 0.97 and Cochise lovegrass, 1.34.

Fig.

10

118

119

10).

More sideoats grama and blue panic seedlings emerged from 10 mm than from the surface, as opposed to Cochise lovegrass seedlings, which established best on the surface. Seeds of sideoats grama and blue panic are

1 X 5 mm and

1 X 2 mm in size, respectively, and larger than Cochise lovegrass seeds which are 0.5 X 0.75 mm. The smaller lovegrass seeds may have greater contact with soil particles on the surface, and therefore better hydraulic conductivity of water from soil to seed. In an associated greenhouse experiment (Fig. 8) seedling emergence of Cochise lovegrass was nearly as high from seeds on bare soil surfaces as on surfaces covered with litter or gravel. Germination and emergence requirements of the larger-seeded species were apparently best met at a burial depth of 10 mm. Germination occurred at deeper sowing depths but emergence was decreased probably as a result of a lack of seed reserves.

Watering on day 1 only generally produced the fewest seedlings (Fig. 10,

Table

10).

Surface-sown seeds produced the most seedlings. Few sideoats

graina or blue panic seedlings emerged from 30 mm. Surface-sown Cochise lovegrass seeds produced many more seedlings than did buried seeds. Below the surface, sideoats graina and blue panic had decreased emergence with depth while Cochise had increased emergence with depth.

Table

10.

Maximum seedling emergence of sideoats

graina, blue panic and

Cochise lovegrass in relation to sowing depth and watering treatment.

Depth

(mm) 1

Water Treatment

1

2 seedling density

3

Sideoats

graina

0

10

20

30

Blue panic

0

10

20

30

Cochise lovegrass

0

10

20

30

9.3

2

2.3

0.4

0.3

9.5

8.0

4.0

3.0

19.8

8.6

9.1

9.9

5.5

9.0

2.0

0

6.5

8.3

5.0

2.3

14.9

11.3

10.0

8.5

8.1

6.1

1.1

0.1

4.0

4.4

2.0

0.1

17.9

2.9

4.3

5.1

1

Water

1 = watering every

3 days, water

2 = watering on days

1 and

7 and water

3 = watering on day

1 only.

2

Standard errors range from

1.7 to

2.0.

120

121

Seedling Survival

Seedling survival percentage had a significant

(p<0.001) day-depth interaction for sideoats

graina; significant

(p<0.05)

2-factor interactions involving day, depth and water treatment for blue panic; and a significant

(p<0.001)

3-factor interaction for

Cochise lovegrass.

Mortality for most sideoats

graina seedlings occurred within the first

24 hours after the final day of maximum seedling density for all depths (Fig. 10).

In general, mortality of sideoats

graina seedlings increased with sowing depth. This may be because seedlings from surfacesown seeds are able to photosynthesize sooner than seedlings from buried seeds, and thus may produce better root systems. All seedlings emerging from 30 mm died the first day. All seedlings emerging from

0, 10, and 20 mm were dead by days

10, 8 and

8, respectively.

Soil water content of soil watered on days 1 and

7 stayed above a corresponding matric potential of

-0.03

MPa for all depths until day 10, and then soil water content began to differentiate according to depths (Fig.

11).

Soil layers closer to the surface were depleted first and deeper depths last. Soil water content in lower depths was above a corresponding matric potential of

Fig.

11.

Soil water content

(wt/wt) in relation to depth in the seedbed across time (days). Water

1 = watered every

3 days, water

2 = watered on days

1 and

7, water

3 = watered on day

1 only. Watering days

=

The standard error for all means

= 0.09.

20

15

I

0

5 o

20

I 5

10

5

Fig.

WATER 1

WATER 2

20

-1.5o wa,

122

1

O

DEPTH

+

DEPTH 2

DEPTH 3

DAY

DEPTH 4

DEPTHS

Dom.'

6

123

-1.50 MPa for up to a day longer than in the upper depths.

Soil water depletion from soil watered on day 1 only showed results similar to soil watered on days 1 and

7

(Fig.

11).

The extended period of available soil water at deeper depths in the seedbed may have aided sideoats graina seedlings from shallow-sown seeds with deeper roots to survive longer than seedlings from deeper-sown seeds.

Mortality of most blue panic seedlings occurred within the first

3 days after the final day of maximum emergence regardless of sowing depth or water treatment

(Fig.

10), and all seedlings were dead by day 9.

Blue panic seedling survival was similar regardless of sowing depth and water treatment. There was a trend towards decreased survival with increased sowing depth.

Complete mortality of Cochise lovegrass seedlings occurred by days 5, 7 and

10 after the final day of maximum seedling density for witering every 3 days (until day

17), watering on days 1 and

7 and watering on day 1 only, respectively (Fig.

10).

There was no obvious sowing depth response across all 3 water treatments.

Root depth data were taken for the first few days following the final day of maximum seedling density (days

124

11, 13 and

15 for watering on days

1 and

7

(Fig. 12), and days

7 and 9 for watering on day

1

(Fig.

13)).

Root depth of sideoats grama generally decreased with increased sowing depth. By day

15, for watering on days

1 and

7, the drying front had extended to near or beyond the bottom of roots from deeper-sown seeds. This may help to explain why mortality among sideoats grama seedlings from deeper-sown seeds generally occurred sooner than for seedlings from more shallow seeds. By day

9, for watering on day

1, all roots were exposed to soil with a water content corresponding with a matric potential of less than

-1.5 MPa.

Blue panic root depths were generally similar regardless of sowing depth (Figs.

12 and

13).

By days

9 and

15 for watering on day

1 and days 1 and

7, respectively, the drying front was near to or beyond the depth of all roots.

Cochise lovegrass root lengths were similar regardless of sowing depth (Figs.

12 and

13).

This meant that rooting depth increased with sowing depth. Root depths in relation to the

-1.5 MPa drying front were similar to that of blue panic.

Root depth data (Fig.

13) indicate that the more rapid drying front and shorter root depths for watering

Fig. 12. Primary root depths of sideoats grama, blue panic and Cochise lovegrass in relation to sowing depth, water treatment and -1.5 MPa matric potential drying fronts. Depths below the slanted line had a matric potential less than -1.5 MPa on that day. The dotted portion of the drying front indicates estimated data. Roots for a given day are from seeds sown at 0, 10, 20 and 30 mm for sideoats graina and blue panic, and 0, 5, 10, and 15 mm for Cochise lovegrass. Standard error ranges are as follows: sideoats graina = 9.7 to 16.7, blue panic = 11.3 to

18.7 and Cochise lovegrass = 6.3 to 7.2 mm.

M

ON

M du

O o

M

N o

M

0 o

Depth (me) co o ch o o o

0

>

.3

CO m m

0 m

0

1

125

/

/

/

/

/

I

/

/

/

/

/

/ e

,

0

0

0

=

1 ..4

M

4

M

‹ m

0

.

.

.

.

/

/

/

/

../ 4

Fig. 12

Fig. 13. Primary root depths of sideoats grama, blue panic and Cochise lovegrass in relation to sowing depth, water treatment and -1.5 MPa matric potential drying fronts. Depths below the slanted line had a matric potential less than -1.5 MPa on that day.

The dotted portion of the drying front indicates estimated data. Roots for a given day are from seeds sown at 0, 10, 20 and 30 mm for sideoats grama and blue panic, and 0, 5, 10, and 15 mm for Cochise lovegrass. Standard error ranges are as follows: sideoats grama = 9.7 to 33.4, blue panic = 14.1 to

16.7 and Cochise lovegrass = 6.3 mm.

n

J

s.

o

Depth (ma)

126

Fig. 13

127 only on day

1, caused roots from this treatment to be exposed to dry soil more quickly than roots for watering on days

1 and

7.

However, there was little difference in survival for any of the species between these

2 water treatments. This inconsistency is not easily explained.

Perhaps the method of mortality detection was not sensitive enough to detect differences.

This study suggests that deep sowing to place seeds in better conditions for germination and emergence may not be beneficial for these warm-season grasses on

Southwest rangelands due to the fact that primary root lengths decrease with increased sowing depth. Shallowsown seeds may produce seedlings that become autotrophic quicker allowing them to produce longer roots that may increase their chances for survival during a drying period between summer rain storms.

128

EFFECTS OF SOWING DEPTH AND SOIL WATER ON

GRASS SEEDLING MORPHOLOGY

Introduction

Revegetation of Southwest rangelands is often risky due to erratic precipitation and high evaporation rates.

Seedling establishment should be successful when viable seeds are in contact with moist mineral soil and primary seminal roots extend quickly downward into the seedbed at a rate that exceeds that of the soil drying front.

The primary root system is often short lived so grass seedlings must develop adventitious roots for longterm survival

(Hyder et al.

1971, Briske and Wilson

1978).

Initiation and growth of both seminal primary and adventitious roots of range grasses are adversely affected by water stress (Van Der Sluijs and

Hyder 1974,

Hassanyar and Wilson

1978,

Wilson and

Briske 1979,

Fulbright et al. 1984).

Water stress also limits leaf growth of grasses (Olmsted 1941, Glendening 1941, Herbel and

Sosebee 1969).

Seedbed preparation is often recommended to increase seed burial (Jordan

1981, Vallentine 1989).

In an associated field study, (Figs. 3 and 4, pages 57 and 60) we found that seeds of small-seeded grasses were buried at shallow depths by natural soil movement or summer

129 thunderstorms. Mechanical seedbed preparation and cattle trampling increased the depth of seed burial and in some cases buried seeds at excessive depths (Table 3, page

64).

Deep sowing has been suggested as a means of placing seeds in better conditions for germination and seedling establishment (Tadmor and Cohen 1968). However, seeds at excessive depths may fail to produce emergent seedlings or may produce seedlings that are weak due to exhaustion of seed reserves.

Conflicts exist in the literature regarding effects of sowing depth on seedling emergence and morphology.

Some of these conflicts may be explained by speciesspecific morphology and varying research methods. In general, deeper burial appears to produce adverse effects. Specifically, researchers have observed weak seedlings (Moore 1943), shorter roots and shoots (Mutz and

Scifres 1975), reduced tiller production

(Arnott

1969, Hadjichristodoulou 1977), reduced leaf numbers, seedling weight and rate of leaf and tiller production

(Arnott 1969, Carren et al. 1987, and decreased primary seminal and adventitious root lengths

(Fulbright et al.

1985). Carren et al. (1987) found that subcoleoptile internode (SCI) weights decreased per unit length in

130 relation to depth of burial. They suggested that this may be a result of smaller

SCI diameters, which could restrict water flow.

In contrast,

Tadmor and Cohen

(1968) found that root lengths of range grasses stayed constant, while those of barley and wheat increased with sowing depth. This put roots of the deepest-sown seedlings in a better position to take in soil water.

Kinsinger (1962) reported increased seedling growth for Indian ricegrass (Orvsopsis hvmenoides (Roem. and

Schult.)

Ricker) with increased sowing depth under limited soil water conditions in the field, and the same response in the greenhouse under optimal water conditions. This led him to believe that a factor other than better water conditions at greater depths was increasing seedling growth.

The objective of this study was to determine effects of sowing depth and soil water on morphology of

3 warmseason grasses.

Methods

The study was conducted in a temperature-controlled greenhouse with average minimum and maximum temperatures of

31 and

45 degrees

C, respectively. Relative humidity ranged from

5 to

78 %.

Relative humidity may range from

7

131 to

100 % and air temperature from 16 to 40 degrees C in the semiarid grassland in southern Arizona during the summer rainy season (Roundy, unpublished data, University of Arizona).

The grass species used were 'Vaughn' sideoats graina

(Bouteloua curtipendula (Michx.) Torr.),

'A-130' blue panic (Panicum antidotale

Retz.), and 'Cochise'

Atherstone lovegrass (Eragrostis lehmanniana

Nees

X

E.

tricophera

Coss and Dur.).

The experimental design was a randomized block, split plot with 4 blocks. Whole plots were soil water treatments, and split plots were time of sampling, sowing depth and species. Each block contained

216 cones; 3 species,

4 sowing depths, 3 soil water treatments, and 6 sampling periods. Each combination included

4 cones with

3 seedlings sampled per cone.

Soil was obtained from a typical loamy upland range site on the Anvil ranch,

65 km southwest of Tucson,

Arizona. The soil is a sandy loam (fine, mixed, thermic

Ustollic Haplargid).

Soil from the A horizon (top 20 cm) was sieved to 5 mm and placed in 3.8-cm diameter by 20-cm long tapered plastic cones to a depth of 0, 10, 20, and

30 mm from the cone surface in cones for sideoats grama

132 and blue panic, and

0, 5, 10, and

15 in cones for

Cochise lovegrass.

Approximately

10 seeds of each species were sprinkled on the soil surface, and each cone was then filled to the top with soil.

Soil water treatments included: 1) watering approximately every 3 days to maintain the soil matric potential above

-0.03 MPa throughout the experiment, 2) watering on days

1 and 7, and

3) watering on day

1 only.

These treatments were selected to simulate the wide range of soil water conditions that may occur during the rainy season in southern Arizona. Watering was by subirrigation to minimize disturbance of surface-sown seeds.

To determine when to begin seedling morphology sampling,

4 replications of

100 seeds of each species were placed on Whatman

No.

2 filter paper in a petri dish. Seeds were watered and each dish was placed in its respective block. Germination counts were made every 24 hours

(+ or

- 1 hour) after initial watering. A seed was considered germinated when its radicle had emerged.

Sampling began on day

5 when seedlings from surface-sown seeds were approximately 1-cm tall and continued every

2 days until day

15 for a total of

6 sampling periods.

Prior to the experiment, all cones were randomly assigned sampling days. During sampling, all emerged

133 seedlings were marked at the soil surface. The cone was placed in water until the soil core was moist. The core was carefully removed from the cone and onto a

0.5 mm wire sieve. The soil was carefully washed from the seedling roots with a fine water spray. Emerged seedlings were removed from the sieve, 3 representative seedlings were selected, and the following measurements were taken:

1) total leaf length (measured from the coleoptilar node to leaf tips), 2) length of coleoptile, 3) length of subcoleoptile internode, 4) length of primary root, 5) position of coleoptilar node in relation to soil surface and 6) number and lengths of adventitious roots on the coleoptilar node (Fig. 14).

Each seedling was oven-dried at 80 degrees

C for 48 hours, separated into above and below ground parts at the coleoptilar node, and weighed.

Soil water content was determined gravimetrically from samples in additional unseeded cones. Beginning on day

1 and then daily throughout the experiment,

1 cone per water treatment per block was randomly selected. The soil core from each cone was carefully removed and approximately 3 cm

3

of soil was sampled from

5-mm depth intervals to a depth of 30 mm.

Beginning on day 9 of the experiment, soil sampling intervals were widened to

10 mm in order to sample the deeper soil in contact with

134

135 extending roots. Gravimetric soil water content of the samples was determined by weighing before and after ovendrying (Hillel 1982). This method was selected because of small soil samples and the importance of sampling narrow soil increments. The gravemetric water content of the study soil was determined at

-0.03 and -1.50 MPa matric potentials created with a pressure plate.

Relative humidity and air temperature were monitored using a Phys-Chemical

Research PCRC-11 relative humidity sensor and a Fenwal

Electronics UUT51J1 thermistor respectively. Data were collected each minute and

30minute averages were recorded with a Campbell Scientific

Instruments CR10 datalogger.

Analyses of variance were performed on individual plant components for each species and on soil water data.

Results

Two or 3-factor interactions involving day, sowing depth and soil water were generally significant

(p<0.05) for all plant parts for all species. The exception was blue panic which had a significant

(p < 0.01) day-sowing depth interaction for subcoleoptile internode length and distance between the soil surface and coleoptilar node.

136

Lengths of subcoleoptile internodes increased with sowing depth but were unaffected by water treatment (Fig.

15). Lengths of coleoptiles were independent of sowing depth and water treatment (Fig. 15). Average coleoptile lengths for sideoats

graina, blue panic and Cochise lovegrass were

6.5, 4.2 and

1.9 mm, respectively.

According to Hyder et al.

(1971) the extension of subcoleoptile internodes with increased sowing depth characterizes these

3 species as panicoid type seedlings.

Elongation of the subcoleoptile internode elevates the coleoptilar node (the major site of adventitious root initiation) to near the soil surface. Adventitious root initiation requires 2 to 4 days of high soil water availability from 2 to several weeks after germination

(Olmsted 1942, Van Der Sluijs and

Hyder 1974,

Wilson and

Briske 1979).

Adventitious roots may fail to develop if they are initiated near the soil surface where soil water availability may be limited without successive rains. In general, poleoptilar nodes of all 3 species were slightly below the soil surface regardless of sowing depth or water treatment (Fig. 15).

Even under high soil water availability, adventitious root initiation may be decreased if the coleoptilar node is not in contact with mineral soil, thus allowing for hydraulic conductivity from soil to node. However, Briske and Wilson

(1978)

Fig. 15. Morphology of 15-day old sideoats grama, blue panic and Cochise lovegrass seedlings in relation to sowing depth and soil water treatment. Seedlings within a water treatment are from seeds sown at 0,

10, 20 and 30-mm depths for sideoats grama and blue panic; and 0, 5, 10 and 15 mm for Cochise lovegrass.

Leaf = total leaf length, cole = coleoptile length, soil = soil surface, CN = coleoptilar node, SCI = subcoleoptle internode and root = length of primary seminal root. Water 1 = watering every 3 days, water

2 = watering on days 1 and 7 and water 3 = watering on day 1 only. Standard error ranges are as follows: sideoats grama leaf, coleoptile, subcoleoptile internode, total primary seminal root and distance from the soil surface to coleoptilar node are 12.9

to 31.7, 0.6 to 1.4, .8 to 2.0, 9.7 to 23.6 and 0.7

to 1.7 mm, respectively. Those respectively for blue panic are 12.2 to 17.3, 0.4 to 0.5, .77 to 1.09,

10.79 to 15.27 and 0.3 to 0.5. Those respectively for

Cochise lovegrass are 6.6 to 7.6, 0.3 to 0.7, 0.6 to

0.7, 6.3 to 7.2 and 0.1 to 0.2 mm.

Fig. 15 ri

0

0

5

4

0.0

00

137

138 showed that adventitious roots could be initiated above the soil surface if relative humidities were near

100 %.

Adventitious root development began on day

9 for all species when watering occurred every

3 days (water treatment

1), and on days

11, 11 and

9 for sideoats grama, blue panic and Cochise lovegrass, respectively when watering occurred on days

1 and 7 (water treatment

2).

No adventitious roots developed when watering occurred on day

1 only (water treatment

3).

Effects of sowing depth on numbers of adventitious roots was highly variable.

Soil water content at the coleoptilar node was above a corresponding matric potential of

-0.03 MPa for at least

9 days prior to adventitious root development under water treatments

1 and

2, and was below

-0.03 MPa from day

6 to the end of the experiment under water treatment

3

(Fig.

16).

Lack of adventitious root development for water treatment

3 was probably the result of this short period of available water.

Primary Seminal Root Lengths

The first few days following germination are critical to seedling survival in the Southwest. Seedlings must rapidly extend primary seminal roots into the soil ahead of the drying front that often occurs following a

Fig.

16.

Soil water content

(wt/wt) in relation to depth in the seedbed across time (days). Water

1 = watering every

3 days, water

2 = watering on days

1 and

7, water

3 = watering on day

1 only.

"*" = watering days. The standard error for all means

= 0.9.

20

15

-1

1 0

5

Fig. 16

WATER

1

-0.03

MPAL

-

1.50 MPa

139

20

15

10

0

DEPTH

1 o

+

DEPTH

2

2 4

6

B

DAY

10

O DEPTH

3 0

DEPTH

4

12

14

16

O

DEPTH

5 V

DEPTH 6

140 rain storm. Primary seminal root length data from day 5 indicate the depth of roots after an initial growth period when soil water was not limiting. Data from the 3 water treatments were pooled since the soil water content of all treatments was greater than a corresponding matric potential of -0.03 MPa until day 5 (Fig. 16).

Primary seminal root lengths of sideoats grama and Cochise lovegrass decreased with sowing depth

(Fig. 17) so total root depth of these species (approximately

55 and 20 mm for sideoats grama and Cochise lovegrass, respectively) was similar regardless of sowing depth. In contrast, primary seminal root lengths of blue panic were independent of sowing depth so that root depth increased with sowing depth. Root depths of seedlings from blue panic seeds sown at 0, 10, 20 and 30 mm were 19, 29.6,

36.2 and 50.7, respectively. Although primary seminal roots of Cochise lovegrass were much shorter than those of the other species on day 5, by day

15 root lengths

(particularly under water treatments

2 and 3) were similar for all species (Fig. 15).

Primary seminal root data from day 5 indicate that deep sowing may be beneficial for increased survival during early growth of blue panic, but not sideoats graina or Cochise lovegrass seedlings. The shallow rooting depth of Cochise lovegrass should tend to decrease its survival

Fig. 17. Below-ground morphology of 5-day old sideoats grama, blue panic and Cochise lovegrass seedlings in relation to sowing depth. Root systems are from seeds sown at 0, 10, 20 and 30 mm for sideoats grama and blue panic, and 0, 5, 10 and 15 mm for Cochise lovegrass.

Soil = soil surface, CN = coleoptilar node, SCI = subcoleoptle internode and root = primary seminal root length. Standard error ranges are as follows: sideoats grama primary seminal root length, subcoleoptile internode, and distance from the soil surface to coleoptilar node are 8.1 to 12.6, 0.7 to

1.1 and 0.6 to 0.9 mm, respectively. Those respectively for blue panic are 11.8 to 21.6, 0.9 to

1.6 and 0.4 to 0.6 mm. Those respectively for Cochise lovegrass are 5.1 to 5.4, 0.5 and 0.1 mm.

40-

50-

60-

10-

20-

30-

Sideoats Grana

Blue Panic Cochise Lovegrass

Sou

CN

Sc'

Seed

Root

Fig. 17

141

142 capabilities. However, in the associated field study

(Fig.

7, page 80), we found that Cochise lovegrass had greater seedling establishment than sideoats

graina.

In a companion greenhouse study (Fig. 10, page

118) greater than

50% mortality occurred by days

1, 3, and

5 after the final day of maximum emergence for sideoats grama, blue panic and

Cochise lovegrass, respectively.

Factors other than the measured morphological characteristics are evidently responsible for the establishment capabilities of Cochise lovegrass.

Primary seminal root lengths of sideoats

graina and blue panic on day 15 were generally unaffected, while those of Cochise lovegrass increased with decreased water availability. Effects of sowing depth on primary seminal root lengths were variable.

Total leaf lengths of all 3 species on day 15 decreased with decreased water availability and in some cases with sowing depth (Fig.

15).

Total leaf lengths under water treatment 3 were only half that of leaves under water treatment

1 and they were intermediate for water treatment

2.

Olmsted

(1941), Glendening (1941) and

Herbel and Sosebee (1969) all reported decreased leaf lengths of

3 to 4-week old sideoats

graina, black graina

(Bouteloua eriopoda (Torr.) Torr.), tanglehead

143

(Heteropogon contortus

(L.) Beauv.) and Boer Lovegrass

(Eragrostis chloromelas Steud) with decreased water availability. Reduced leaf length and hence leaf area could be an advantage. Glendening (1941) suggested that grass seedlings with large absorbing root systems and reduced leaf area with a small transpiring surface may have a survival advantage over seedlings with small root systems and large leaf areas.

Shoot and Root Weights

Analysis of variance of shoot and root weights indicated a significant 3-factor interaction involving days, sowing depth and water treatments for sideoats grama and 2-factor interactions involving days, sowing depths and water treatments for

Cochise lovegrass.

Analysis of blue panic shoot and root weights produced no significant interactions and only a significant day main effect. Analysis of the effects contributed by sowing depth in the interactions showed no trends. The effects of water treatments were more clear, therefore day-water treatment means were used for comparison (Fig.

18).

Rates of shoot growth were similar among the water treatments for all 3 species until day 9

(Fig. 18). Mean rates were .2 mg/day for sideoats grama and blue panic and .06 mg/day for Cochise lovegrass.

Beginning on day

Fig. 18. Leaf and root weights of sideoats grama, blue panic and Cochise lovegrass seedlings in relation to soil water treatment and time (days). Water 1 = watering every 3 days, water 2 = watering on days 1 and 7 and water 3 = watering on day 1 only. Standard error ranges are as follows: sideoats grama leaf and root weights, 0.2 to 0.3 and 0.2 to 0.3 mg; blue panic leaf and root weights, 0.3 to 0.6 and 0.4 to

0.7 mg; and Cochise lovegrass leaf and root weights,

0.1 and 0.1 mg.

3

2

3

2

1

3

2

9 13 '5

0 WATER

1

DAY

+

WATER

2

Fig.

18 o WATER 3

13

144

145

11, rates increased rapidly under water treatment

1, moderately under water treatment 2 and slightly under water treatment

3.

Mean rates for all species across the last

3 sampling days were

.7, .3 and

.2 mg/day, respectively.

Shoot growth rates of Cochise lovegrass under water treatment

3 were only

.06 mg/day as opposed to .19 mg/day for sideoats grama and

.32 mg/day for blue panic. Root weights were heaviest under water 3 and decreased with increased water availability (Fig. 18).

Root rates differed among species. Mean rates for blue panic, sideoats

graina and

Cochise lovegrass were

.56, .34

and

.18 mg/day, respectively.

Root:Shoot Ratios

It has been suggested that the larger the root:shoot ratio, the more drought resistant the seedling

(Oppenheimer

1960,

Wright and

Streetman 1960 and

Simanton and Jordan

1986).

Both root:shoot length and weight ratios could be important. High root:shoot length ratios could indicate deep rooting depths and small transpiring leaf surfaces. Root:shoot weight ratios may indirectly indicate a ratio of total soil water absorbing surfaces to total transpiring surfaces.

Species-water treatment and species-day interactions

146

771

WATER 1 WATER 2 MD WATER 3

Fig.

19. Root:shoot weight ratios of sideoats grama, blue panic and

Cochise lovegrass in relation to soil water treatment. Water

1 = watering every

3 days, water

2 = watering on days

1 and

7, and water

3 = watering on day

1 only. Bars within a species with the same letter are not significantly different

(p<0.05).

147 were significant for root:shoot ratio which generally increased with decreased available water (Fig. 19).

This corresponds with root and shoot weight, total leaf and primary seminal root data (Figs. 15 and

17).

Blue panic had greater root:shoot ratios than the other species during the entire study (Fig. 20). Ratios were always above 1 indicating more carbon allocation for root than leaf production. Ratios of sideoats grama and Cochise lovegrass were near 1 throughout the study. Ratios for all species decreased during the first few days (Fig. 20) and then increased during the final days of the study.

Seedlings evidently allocated seed reserves into roots during the early stages of seedling growth.

Total length of below-ground plant parts for all 3 species was similar on day

15, but root weights were different. Root weights (which included the primary seminal root, seed, and subcoleoptile internode) of blue panic, sideoats

graina, and Cochise lovegrass on day 15 were 3.5, 2.0 and 1.0 mg, respectively. The greater weight of blue panic roots can be explained by the greater amount of seminal lateral and subcoleoptile internode roots produced by blue panic seedlings (Fig

14).

This also accounts for the greater root:shoot ratio of blue panic seedlings.

148

2.0

1.8

1.6

1.4

1.2

1.0

3.0

2.8

2.6

2.4

2.2

0.8

0.6

0.4

0.2

0.0

5 7 9

DAY

11 13 15

Fig. 20. Root:shoot weight ratios of sideoats grama, blue panic and Cochise lovegrass across time (days).

Standard errors for sideoats graina, blue panic and

Cochise lovegrass are 0.1, 0.1, and 0.1, respectively.

149

Discussion

This study shows that deep sowing for increased seedling establishment of sideoats grama, blue panic and

Cochise lovegrass may not be advantageous since the site of adventitious root initiation is elevated to near the often-dry soil surface regardless of sowing depth.

Although early rooting depth of blue panic seedlings increases with sowing depth, deep sowing still may not be beneficial since adventitious roots will later initiate near the soil surface.

With the exception of subcoleoptile internodes, seedling morphology of all

3 species was generally unaffected by sowing depth. Lack of available water adversely affected leaf length and weight, but root lengths and weights were unaffected or actually increased with stress. This response may be a survival strategy.

The more rapid primary seminal root extension of sideoats grama and blue panic, and greater root:shoot ratios of blue panic in relation to Cochise lovegrass should equate to greater establishment capabilities for these species. However, this is not the case.

Simanton and Jordan (1986) reported findings similar to ours and suggested that successful Cochise

150 lovegrass establishment was due to physiological rather than morphological adaptations.

Many researchers have examined the relationship between morphological characteristics and drought resistance. However, according to Johnson (1980), their results often conflict due to interactions of these characteristics with the environment. Wright (1971) stated that none of these characteristics has been useful as an indicator of drought tolerance.

In performance tests under natural and artificial conditions of water stress, Wright (1964) showed that out of 6 species of grasses, Lehmann lovegrass (Eragrostis lehmanniana

Nees) was most drought tolerant and blue panic was least. No evaluation was made on sideoats graina.

Cochise lovegrass is very similar to Lehmann lovegrass genetically, and according to Jordan

(1981) has similar drought tolerance. Examinations of characteristics that may be associated with drought tolerance of blue panic revealed that only stomatal density was positively associated (Dobrenz et al.

1969).

Similar examinations of Lehmann lovegrass showed positive associations with water use efficiency and leaf waxes

(Wright and

Dobrenz 1973,

Hull et al.

1978).

151

Although morphological characteristics such as rapid root extension and reduction of leaf area may be associated with drought tolerance and establishment of grass seedlings, apparently other factors such as water use efficiency may be more important in determining the ability of the species in this study to establish.

152

SUMMARY AND CONCLUSIONS

Seed Location

A new technique was developed to determine the distribution of seeds with respect to depth as influenced by seedbed preparation. The technique successfully determined seed location after cattle trampling, land imprinting, root plowing and ripping on a sandy loam soil in southern Arizona. It successfully determined seed location after summer thunderstorms and depth of seeds that produce emergent seedlings.

In a laboratory experiment, coefficient of determination (r

2

) between percentages of actual and observed seeds at different depths of sideoats grama, blue panic, and Lehmann lovegrass were

0.80, 0.72, and

0.92, respectively.

The technique tends to underestimate the percentage of small buried seeds, such as those of Lehmann lovegrass, but has moderately high accuracy and permits analysis of large numbers of samples. This technique should be useful in determining the effects of different seedbed preparation techniques on seed placement in rangeland soils and help explain differences in seedling emergence in revegetation studies.

153

The technique was used during the summers of

1987 and 1988 to determine the effects of various seedbed preparation techniques and summer thunderstorms on seed placement, and the location of seeds that produce emergent seedlings.

An average of 75, 42, 17, and

7 % of seeds found were buried immediately after heavy trampling, land imprinting, light trampling and non-disturbance, respectively. After summer thunderstorms an average of

78, 72, 63, 40 and 29 % of seeds found were buried by root plowing, heavy trampling, imprinting, light trampling and non-disturbance, respectively. Blue panic, sideoats grama, Cochise lovegrass and Lehmann lovegrass seedlings emerged from above 30, 27, 18 and

11 mm, respectively. Although heavy trampling and root plowing buried more seeds than the other treatments, they also placed more seeds below the biological limit of emergence. Blue panic and the lovegrasses were buried by the treatments and rain better than sideoats grama.

Effects of Seedbed Preparation on Seedling Emergence

Effects of cattle trampling, land imprinting and root plowing or ripping on seedling emergence of sideoats grama, blue panic, and Lehmann and Cochise lovegrass were compared in a field study on a sandy loam soil in

154 southern Arizona. Plots were broadcast seeded and treated prior to summer rains in

1987, 1988 and 1989. Seedling density of seeded grasses, and cover of indigenous species was measured after emergence. Differences in amount and frequency of precipitation and resultant differences in surface soil water availability among years and treatments were strongly associated with differences in seedling emergence among seeded species.

In a wet year

(1987) when surface soil water was estimated to be available for at least

24 consecutive days, heavy trampling and land imprinting increased emergence of blue panic and land imprinting increased emergence of Cochise lovegrass.

In that year, lovegrass emergence was high even on undisturbed plots. In a moderately wet year

(1988), measured surface soil water was available for 6 to

9 days during seedling emergence and heavy trampling, land imprinting and root plowing increased emergence of all species compared to light trampling and non-disturbance. In a dry year (1989), measured surface soil water was available for

2 to

4 days and seedling emergence was low and similar for all treatments.

Sideoats

graina emergence was low all 3 years, but was highest in

1988 when initial thunderstorms were followed closely by subsequent storms. Seedling emergence of indigenous annual and perennial grasses was generally

155 similar for all treatments and years and was closely associated with the time of most consistent rainfall.

Seedbed preparation to enhance revegetation in the

Southwest may be most useful in years of moderate precipitation and may be unnecessary in wet years or futile in dry years depending on species and soils.

Seedbed Microsites

A greenhouse study was implemented to determine the response of sideoats grama, blue panic and

Cochise lovegrass to naturally-occurring rangeland seedbed microsites in relation to 3 soil water treatments.

Although there were several interactions, in general, emergence of all 3 species was highest from gravel, followed by litter, cracks and finally the bare soil surface. Bare surface sites decreased in water content more quickly than the other sites. Cochise lovegrass performed well in gravel under all water conditions. Small seeded species such as Cochise lovegrass broadcast on coarse-textured surface soils may establish with minimal seedbed preparation, provided summer precipitation is adequate.

Effects of Sowing Depth and Soil Water

On Emergence and Survival

A greenhouse study was implemented to determine

156 effects of sowing depth and soil water on seedling emergence and survival of sideoats graina, blue panic and

Cochise lovegrass.

Under high soil water availability

(matric potential above

-0.03 MPa during entire

15 days of the experiment), maximum emergence of all species was greatest from surface-sown seeds. Maximum emergence under drier conditions

(matric potential above

-0.03 MPa for

6 and

10 days) was greatest from

10 mm for sideoats

graina and blue panic, and from surface-sown seeds for Cochise lovegrass. Survival of sideoats graina and blue panic generally decreased with increased sowing depth, probably as a result of decreased primary root lengths from deepsown seeds. Soil water content increased with depth in the seedbed. Soil water content in lower depths was above a corresponding matric potential of

-1.50 MPa for this soil for up to a day longer than in shallower depths and corresponded with greater survival of seedlings with deeper roots than those with shorter roots. Greater sowing depths may result in decreased seedling survival due to shallower primary roots.

Effects of Sowing Depth and Soil Water on Morphology

A greenhouse study was implemented to study effects of sowing depth and soil water on morphology of sideoats graina, blue panic and

Cochise lovegrass. Lengths of

157 subcoleoptile internodes of all species increased with sowing depth but were unaffected by water treatment.

Coleoptiles of all species were independent of sowing depth and water treatment. The position of the coleoptilar node (site of adventitious root initiation) was at or within 4 mm of the soil surface for all species. Adventitious roots began emerging on day

9 for all species when watering occurred every 3 days and on days

11, 11 and 9 for sideoats grama, blue panic and

Cochise lovegrass, respectively when watering occurred on days

1 and 7. No adventitious roots developed when watering occurred on day 1 only. Primary seminal roots of sideoats grama and blue panic were unaffected by a lack of available water.

Primary seminal root lengths of

Cochise lovegrass increased with decreased water availability. Total leaf lengths and rates of leaf weights of all species decreased with decreased water availability. Root weights of all species were generally heaviest when watering occurred on day 1 only and decreased with an increase in available water. Root:shoot ratios of all species increased with decreased available water. Blue panic seedlings had greater root:shoot ratios than sideoats grama and Cochise lovegrass seedlings throughout the study. Deep sowing for long-term survival of sideoats grama, blue panic and Cochise lovegrass may

158 not be beneficial due to elevated coleoptilar nodes and therefore possible reduced adventitious root development.

Drought tolerance of Cochise lovegrass may be associated with physiological rather than morphological characteristics.

Species Characteristics

Sideoats

Graina

Sideoats grama is difficult to establish, and little work has been done to determine reasons for seeding failures. Results from the above studies may help discover some reasons for these failures.

Sideoats grama seeds were difficult to bury by either seedbed preparation treatments or summer thunderstorms. Heavy cattle trampling buried the most seeds, followed by root plowing or ripping and then imprinting. Seeds of sideoats grama were sown as caryopses within intact florets and spikelets, and therefore were not easily sifted down into the seedbed.

In contrast, seeds of blue panic, and Lehmann and Cochise lovegrass are smaller and were sown only as naked caryopses. Most seedlings emerged from buried seeds, indicating the importance of burial for this species.

Sideoats grama seedlings emerged from

0 to

27 mm in

159 the field and from 0 to

30 mm in the greenhouse, however, we determined that the majority (90%) of seeds emerged from a depth of 12 mm or less in the field. The recommended sowing depth for sideoats

graina is approximately 12 mm (Jordan 1981).

The results of the above studies confirm these recommendations.

Sideoats

graina emergence in a sandy loam soil in the field was low during a wet year (1987), moderately wet year

(1988) and dry year

(1989).

Petri dish studies in the greenhouse showed that most germination of sideoats graina seeds occurred within

24 hours after the beginning of imbibition. According to Jordan (1981) rapid germination may increase the chances of establishment in areas with sporadic summer rainfall. However, Simanton and Jordan

(1986) suggest that the rapid germination of sideoats grama may be either a positive or negative factor for establishment depending upon the amount and temporal distribution of rainfall. Under periods of extended water availability, sideoats

graina could establish well. However, if a rain did not wet subsurface soil, the fast growing roots of sideoats

graina would extend into dry soil. In each of the 3 years of the field study, initial rains of 10 mm or less were followed by periods of unavailable soil water. It could be

160 hypothesized that sideoats graina seeds germinated after these initial rains and that the subsequent drying periods caused seedling desiccation.

Even though sideoats grama can emerge from up to 30 mm, under wet soil conditions in the greenhouse, emergence was greatest from seeds on the surface.

However, radicle penetration was limited for many of these seedlings. In some cases, radicle penetration had occurred, but as much as 4 cm of primary seminal root was exposed. This phenomenon could have a negative impact on the long-term survival of these seedlings. Low emergence from buried seeds under wet soil conditions may have been caused in part by inadequate gas exchange.

Surface-sown sideoats graina seeds responded best to gravel microsites, followed by litter and crack microsites, and then the bare soil surface. There is an incongruity between the above 2 greenhouse studies regarding emergence of surface-sown seeds. In the first study, emergence was greatest from surface-sown seeds as compared to buried seeds, and in the second study emergence was least from surface-sown seeds on the bare soil surface as compared to gravel, litter and crack microsites. In both cases the seedbed was the same. The probable cause was the greenhouse environment. The

161 environment during the period of germination of the first study was relatively cool and humid and skies were overcast for much of the time. Conditions during the first few days of the second study were warm and dry.

Although soil water content data from the 2 studies do not reflect the differences in the greenhouse environment

(Figs. 9 and

11), it is possible that the soil surface in the second study dried out enough to prevent adequate hydraulic conductivity from soil to seed.

Sideoats

graina seedling survival was the least of the 3 species studied. Most seedlings died within the first 24 hours after the final day of maximum emergence, as opposed to 3 days for blue panic and 5 days for

Cochise lovegrass.

Mortality of sideoats graina seedlings increased with sowing depth, probably as a result of shallow rooting depth from deep-sown seeds.

Sideoats

graina is characterized as a panicoid species due to elevated coleoptilar nodes caused by elongating subcoleoptile internodes.

Because the coleoptilar node is the major site of adventitious root initiation it would be inadvisable to deep-sow sideoats graina seeds to take advantage of better germinating conditions. Leaf lengths, weights and weight rates decreased; and primary root lengths, weights and weight

162 rates were generally unaffected by a decrease in availability of soil water. Leaf and root rates were both intermediate compared to those of blue panic and Cochise lovegrass.

Root:shoot weight ratios were near 1 regardless of sowing depth or water treatment.

Blue Panic

Blue panic seeds were buried best by heavy cattle trampling, followed by land imprinting, light cattle trampling and then non-disturbance. Compared to sideoats grama seeds, more blue panic seeds were buried by seedbed treatment and rain storms.

Nearly all seedlings in the field emerged from buried seeds and all seedlings emerged from

0 to 30 mm in the field and greenhouse. The biological limit of emergence in the field was 14 mm.

The recommended sowing depth for blue panic is approximately

12 mm

(Jordan

1981).

The results from these studies confirm those recommendations.

Blue panic emergence in a sandy loam soil in the field was best from plots treated with land imprinting and heavy cattle trampling during a wet year (1987); ripping, heavy trampling and imprinting during a moderately wet year

(1988); and was low regardless of seedbed treatment during a dry year

(1989).

This study

163 indicates that blue panic seeds germinate and emerge best from seedbeds prepared by treatments that bury seeds.

However emergence in the greenhouse was best from the bare soil surface as compared to buried seeds in 1 study, and least from the soil surface as compared to gravel, litter and crack microsites in a second study. This incongruity is addressed above in the sideoats grama section. The response from the second study is probably more indicative of conditions in the field during the growing season in southern Arizona.

Blue panic seedling survival was intermediate between sideoats graina and Cochise lovegrass. Most seedlings died within the first

3 days after the final day of maximum emergence, as opposed to 1 day for sideoats grama and 5 days for Cochise lovegrass. Blue panic seedling survival was similar regardless of sowing depth or water treatment.

Blue panic is also characterized as a panicoid species. For this reason it would not be advisable to deep sow blue panic seeds to take advantage of better soil water conditions. As with sideoats grama seedlings, leaf lengths, weights and weight rates of blue panic seedlings decreased; and primary root lengths, weights and weight rates were generally independent of a decrease

164 in available soil water. Blue panic root weight rates were higher than either sideoats graina or Cochise lovegrass, however primary seminal root lengths of 2-week old seedlings were similar for all species. The probable reason for the difference is that blue panic seedlings have many more primary seminal lateral roots than either sideoats graina or

Cochise lovegrass. This should allow blue panic seedlings to exploit more soil area and consequently more soil water. The greater root weight is also the reason that blue panic had greater root:shoot ratios than the other species throughout the entire study.

Lehmann and Cochise Lovegrass

Lehmann and

Cochise lovegrass are discussed together due to their similar genetic make-up and size. The studies showed that both species responded in a similar manner to seedbed treatments in regards to seed burial and emergence. It was partly for this reason that only

3 species were used in the greenhouse experiments.

As with sideoats graina and blue panic, the lovegrasses were buried best by heavy trampling, followed by imprinting, light trampling and then non-disturbance.

Although more lovegrass seeds were buried than sideoats graina and blue panic seeds, in general more lovegrass

165 seeds were buried beyond the biological limit compared to the other species. Nearly all seedlings emerged from buried seeds even in undisturbed plots.

Lovegrass seedling emergence was very high even in undisturbed plots in a wet year (1987), highest from disturbed plots in a moderately wet year (1988) and low regardless of seedbed treatment in a dry year (1989).

In the greenhouse,

Cochise lovegrass responded best to gravel microsites, followed by litter, cracks and then the bare soil surface. Cochise did well in gravel under all water treatments. In contrast to sideoats grama and blue panic,

Cochise lovegrass emerged reasonably well from the bare soil surface.

In another greenhouse study, Cochise seedling emergence was twice as high from surface-sown than from buried seeds. In contrast to sideoats grama and blue panic seedlings, radicals of Cochise lovegrass seedlings were not impeded by the soil surface. In no case were exposed radicles observed. Apparently, Cochise lovegrass radicles are smaller in diameter than radicles of the other species and can therefore penetrate the soil surface easier than the other species. In a laboratory pilot study, designed to observe radicle penetration of

166 lovegrass seedlings from surface-sown seeds, I observed a network of fine seminal roots that extended from the seminal node to the soil surface.

These roots may have aided radicle penetration by securing the seeds during radicle extension.

These studies indicate that it may be practical to sow lovegrass seeds onto coarse-textured soil surfaces with little or no seedbed preparation. Seedlings would then become established on the soil surface or be buried by summer thunderstorms.

Cochise lovegrass was more drought tolerant than either blue panic or sideoats grama, even though the initial rate of primary seminal root extension was much less than that of the other 2 species. It was suggested that the reasons for this phenomenon were probably physiological rather than morphological.

Cochise lovegrass exhibits panicoid-type belowground morphology similar to the other 2 species.

In contrast to those species, primary seminal root lengths of Cochise lovegrass increased with decreased water availability.

Rates of leaf and root weights as well as root:shoot weight ratios were less than either sideoats grama or blue panic.

167

Seedbed Preparation Treatments

Cattle Trampling

This study showed that cattle trampling can bury seeds at a desirable depth for grass seedling emergence.

Heavy trampling and root plowing or ripping buried similar amounts of seeds and both buried more seeds of all species than land imprinting, light trampling and non-disturbance. However, heavy trampling also buried more seeds below the biological limit than all treatments except root plowing or ripping. Heavy trampling was the best treatment to bury seeds of all species regardless of seed size.

In a wet year

(1987), heavily-trampled and imprinted plots produced similar blue panic seedling emergence, and both produced more seedlings than the other seedbed treatments. However, heavy trampled plots had similar or less emergence than the other seedbed treatments for the other species. In a moderately-wet year

(1988), heavy trampling, imprinting and ripping increased emergence of all species compared to light trampling and nondisturbance. In a dry year

(1989), all seedbed treatments produced similar low seedling emergence.

Light cattle trampling was generally ineffective as a method for increasing seed burial and seedling

168 emergence. Light trampling and non-disturbance generally produced similar seed burial and seedling emergence.

Heavy cattle trampling as used in this study (herding

5 cattle around a 6 X 6 in paddock for 20 minutes) is probably rare in typical rangeland conditions, except around stock ponds, corrals and portions of intensivelygrazed pastures. If deliberate cattle trampling was practiced to help bury seeds, the resulting seedbed would probably be between the seedbeds created by heavy and light trampling as defined in this study. If this were the case, we would expect reasonably good seed burial and seedling emergence depending upon other variables including soil water and texture, and amount of residual vegetation.

Land

Imprinting

Land imprinting generally buried less seeds than heavy cattle trampling, root plowing or ripping; and more than light cattle trampling or non-disturbance.

Imprinting buried seeds of Lehmann and

Cochise lovegrass best, followed by blue panic and then sideoats grama.

Imprinting buried less seeds below the biological limit than heavy trampling and root plowing or ripping.

During a wet year

(1987), imprinting increased blue panic and

Cochise lovegrass emergence over the other

169 seedbed treatments. In a moderately-wet year

(1988), imprinting, heavy trampling and ripping increased emergence of all species over light trampling and nondisturbance. In a dry year

(1989), emergence was low and similar for all species and seedbed treatments.

Imprinting buried less seeds than heavy trampling and root plowing or ripping, but produced similar seedling emergence. Imprinting both presses seeds into the seedbed and covers them with soil. The pressing action may increase emergence by providing a firm seedbed and increasing seed-soil contact.

Root Plowing and Ripping

Root plowing and ripping produced the most fractured and heterogeneous seedbed. These treatments also buried the most seeds and placed more seeds below the biological limit than all other seedbed treatments with the exception of heavy trampling.

Blue panic and

Cochise lovegrass emergence in rootplowed plots was less than in all other plots in a wet year

(1987).

In a moderately-wet year

(1988), sideoats grama emergence was much higher after ripping than after all other treatments. Emergence of the other species in ripped plots was similar to heavy trampling and

170 imprinting.

Heavy cattle trampling, land imprinting and root plowing or ripping all provided adequate seedbed preparation for seed burial and seedling emergence during wet and moderately wet years. In dry years seedbed preparation by these treatments will not ensure successful seedling establishment due to rapid drying of the seedbed and lack of recurrent rains to maintain sufficient available water for germination and seedling growth.

Relationships Between Greenhouse and

Field Experiments

Microsites

Although the seed burial study indicated that most seedlings emerged from buried seeds, some remained on the soil surface.

These seeds occupied various microsites, some of which provided favorable conditions for germination and seedling emergence while others did not.

A field survey indicated that 3 obvious microsites at the study site were gravel, cracks in the soil surface and plant litter. The greenhouse study showed that all 3 microsites produced greater seedling emergence than the bare uncracked soil surface. Of the 3 microsites, gravel generally produced the most seedlings, followed by litter and cracks.

These results indicate that it may be

171 practical to surface-sow seeds on coarse-textured soils on southern Arizona rangelands.

Emergence and Survival

The seed burial study showed that seeds of all species used in these studies were buried by seedbed preparation at a variety of depths including those below the biological limit. It was hypothesized that seeds buried near but not beyond the biological limit may be in better soil water conditions for germination, emergence and survival. The emergence and survival greenhouse study showed that deep sowing may not be beneficial for these grasses on Southwest rangelands due to rapid drying of the seedbed and decreased primary root lengths with increased sowing depth.

Morphological Characteristics

Grass seedling morphological characteristics such as adventitious roots and their depth of initiation in the seedbed, lengths and rates of growth of primary seminal roots, and root:shoot ratios may affect seedling establishment. This greenhouse study showed that regardless of sowing depth, the site of adventitious root initiation for all species was at or near the soil surface where harsh conditions decrease the probability of adventitious root initiation.

172

Root weights and lengths of all species were unaffected or actually increased with a decrease in available soil water, while leaf lengths and weights decreased with soil water availability. Root:shoot ratios of all species also increased as available soil water decreased. These results indicate that these species may exhibit deep rooting and decreased transpiring leaf surfaces as a survival strategy. The field study showed that Cochise lovegrass was the easiest of all the species to establish.

However, the greenhouse studies indicated that sideoats graina and blue panic were better equipped morphologically for establishment than was Cochise lovegrass. Successful Cochise lovegrass establishment may be due to physiological rather than morphological characteristics.

In conclusion, the seedbed preparation treatments in this study are helpful in burying seeds within the biological limit of emergence. Seed burial was necessary for seedling emergence in the field.

There is apparently no advantage to using seedbed preparation treatments which bury seeds deeper within the biological limit.

Although these seedbed preparation techniques help to bury seeds, which increases seed-soil contact and seedling emergence on wet or moderately-wet years, they

173 do not ensure seedling establishment on dry years.

Seedlings apparently need 6 or more consecutive days of available water to emerge and subsequent precipitation to permit seedling survival. Sideoats grama which germinates rapidly, apparently requires consistent available water after initial rains to establish. The lovegrasses are slower to germinate and are able to establish under a greater range of precipitation and soil water conditions than sideoats grama. The small seed size of the lovegrasses increases their chance of burial even without seedbed preparation or soil disturbance. These factors help explain the successful natural and artificial revegetation of lovegrasses in the Southwest.

APPENDIX

ANALYSIS OF VARIANCE TABLES

174

(D (D (D (D (D (D (D rt ri ri ri ri

1

1

O 000000

M

(D (D (D (D CD

O 000000 rr rr

A) A) Cu rl• rt rr rr

0) (a) ta)

41 4 4 4 4 4 fr) fD

(D (D (D

XX>CXXXX

(D 'D (D

(I)

.3

c1 MI (D

CD il CU QI (D (D Cu ri

Cn o< cf)

0 tt 0) 0 rt ris'-

X(

5

(D ca

(DUC

Co

1-3 cil

• a)

<pI 'D

(D (D

O -3 CD s.•

N

rh

Cri

(1 5

(D

0)

(D

Cn rr

5 rr

(D

. 3

PI nt rr ni rt

5 fr)

CO 03 NJ NJ NJ NJ i-

,

0

I-, IV (...)

CO

NJ Co

h

,

h

,

CM

VD

Ch h

,

h

,

CD

Ln ni

Co

I-.

h

,

CD CD lJ NJ LI NJ

CO

Co CT J

CM

VD h

,

h

,

NJ CD cn

Ln

CD

• • • • • • •

4..

h

,

NJ VD -4 l0 NJ

VD

CO CD h

,

CO -.1

0,

TI

'13

P.

CO

(D ri- N

(DO

W

rr

▪ M

CD

0 0.

(D

Q.

(D 5

0(n (0

(1)

a.

M

0 0

ri-

C

t

N

N

(D

ID

CO

• tr rt

(no

(D A) Al rt 0

ID niC-

10)

• ri- DC

Q.

DI

.n

D) ri- CD

(D 0.0

Il rn ni £ ri- CD

<

(D

(1) ri

CA ch

(-h Ai

0

0 0

N.)

00.

(D

• mi

5 0 0

HI

O f:11

4 rr

• rt

ni

rn ni ri-

175

Table

2.

Repeated measures analysis of variance of depth at which cumulative percentages of seedlings of

4 grasses emerged in relation to seedbed treatments.

Source

Percentage

X

Species

Percentage X Treatment

Percentage X Treatment X

Species

**

Significant at the 0.01 level.

df

27

36

108

F -value

6.56 **

1.94

0.66

176

177

LITERATURE CITED

Allison, D. V. and C. A.

Rechenthin. 1956.

Root plowing proved best method of brush control in south Texas.

J. Range Manage. 9:130-133.

Anderson, D. and A. R. Swanson.

1949.

Machinery for seedbed preparation and seeding on southwestern ranges. J. Range Manage.

2:64-66.

Archibold,

O. W.

1979.

Buried viable propagules as a factor in postfire regeneration in northern

Saskatchewan. Can. J. Bot.

57:54-58.

Arnott,

R. A.

1969.

The effect of seed weight and depth of sowing on the emergence and early seedling growth of perennial ryegrass

(Lolium perenne). J.

Brit. Grassland Soc.

24:104-110.

Ball, D. E. 1964.

Range seeding introduced grasses on root plowed land in the northwest Rio Grande plain.

J. Range Manage.

17:217-220.

Barnes, O. K., D. Anderson and A.

Heerwagen. 1958.

Pitting for range improvement in the Great Plains and southwest desert regions.

U.S.D.A.,

Production

Research Report No.

23, 17 p.

178

Barnes, W. C. 1936. Herds in San Simon Valley. Amer.

Forests 42:456-457.

Bewley, J. D. and M. Black.

1985.

Seeds, Physiology of

Development and Germination. Plenum Press, New York.

367 p.

Bleak, A. T., N. C. Frischknecht,

A. P. Plummer and R. E.

Eckert, Jr.

1965.

Problems in artificial and natural revegetation of the arid shadscale vegetation zone of Utah and Nevada. J. Range Manage. 18:59-65.

Bleak, A. T. and A. C. Hull. 1958. seeding pelleted and unpelleted seed on four range types. J. Range

Manage.

11:28-33.

Blom, C. W. P. M. 1978.

Germination, seedling emergence and establishment of some plantago species under laboratory and field conditions. Acta Bot. Neerl.

27:257-271.

Briske

D. D. and A. M. Wilson. 1978.

Moisture and temperature requirements for adventitious root development in blue graina seedlings. J. Range

Manage.

31:174-178.

179

Cable, D. R.

1971.

Lehmann lovegrass on the Santa Rita

Experimental Range, 1937-1968.

J. Range Manage.

24:17-21.

Campbell, B. D.

1985.

Planting depth effects on overdrilled seedling survival in summer. New Zealand

J. Exp. Agric. 13:103-109.

Campbell, M. H., and F. G. Swain. 1973a.

Factors causing losses during the establishment of surface-sown pastures. J. Range Manage. 26:355-359.

Campbell, M. H. and F. G. Swain. 1973b.

Effects of

Strength, tilth and heterogeneity of the soil surface on radicle-entry of surface-sown seeds. J.

Br.

Grassld.

Soc.

28:41-50.

Carren,

C. J., A. M. Wilson, R. L. Cuany and G. L. Thor.

1987. Caryopsis weight and planting depth of blue grama

I. morphology, emergence, and seedling growth.

J. Range Manage. 40:207-211.

Choudhary,

M. A., Guo Pei Yu, and C. J. Baker. 1985.

Seed placement effects on seedling establishment in direct-drilled fields. Soil Till. Res.

6:79-93.

180

Clary, W. P.

1988.

Plant density and cover response to several seeding techniques following wildfire. Res.

Note INT-384 USDA, Forest Service, Intermountain

Research Station.

6 p.

Clary, W. P. 1989. Revegetation by land imprinter and rangeland drill. Res. Pap. INT-397. USDA, Forest

Service, Intermountain Research Station. 6 p.

Collis-George, N. and J. E. Sands. 1959.

The control of seed germination by moisture as a soil physical property.

Aust.

J. Res. 10:628-636.

Colman, E. A. and T. M. Hendrix. 1949.

The fiberglas electrical soil-moisture instrument. Soil Sci.

67:425-438.

Cook, C. W. 1958.

Sagebrush eradication and broadcast seeding. Utah

Agr.

Exp. Sta. Bull.

404. 23 p.

Cornish, P. S.

1982.

Root development in seedlings of ryegrass

(Lolium perenne L.) and Phalaris (Phalaris aquatica L.) sown onto the soil surface. Aust.

J.

Agric. Res.

33:665-677.

Cox, J. R. and M. H. Martin.

1984.

Effects of planting depth and soil texture on the emergence of four lovegrasses.

J. Range Manage.

37:204-205.

181

Cox, J. R., M. H. Martin-R,

F. A.

Ibarra-F,

J. H.

Fourie,

N. F. G. Rethman, and D. G. Wilcox.

1988.

The influence of climate and soils on the distribution of four African grasses. J. Range Manage. 41:127-

139.

Cox, J. R., M. H. Martin-R., F. A. Ibarra-F., and H. L.

Morton.

1986.

Establishment of range grasses on various seedbeds at creosotebush (Larrea tridentata) sites in Arizona, U.S.A. and Chihuahua, Mexico. J.

Range Manage.

39:540-546.

Cox, J. R., H. L. Morton, T. N. Johnson, G. L. Jordan, S.

C. Martin, and L. C. Fierro. 1982. vegetation restoration in the

Chihuahuan and Sonoran Deserts of

North America. USDA,

Agr.

Res.

Serv., Agr.

Reviews and Manuals. ARM-W-28. Oakland, Calif. 37 p.

Cox, J. R., H. L. Morton, J. T. Lebaume, and K. G.

Renard.

1983.

Reviving Arizona's rangelands. J.

Soil and Water Cons. 38:342-345.

Cox, J. R. and G. B. Ruyle. 1986.

Influence of climatic and edaphic factors on the distribution of

Eragrostis lehmanniana

Nees in Arizona. J. Grassi.

Soc.

Sth. Afr. 3:25-29.

182

Dixon R. M. and J. R. Simonton.

1980. Land imprinting for better watershed management, p.

809-826. In: Symp.

on watershed Manage. Vol. II. 21-23

July 1980.

Boise, Ida. Amer. Soc. Civil Eng.

Dobrenz,

A. K., L. N. Wright, A. B. Humphrey, M. A.

Massengale and W. R. Kneebone. 1969.

Stomate density and its relationship to water-use efficiency of blue panicgrass

(Panicum antidotale

Retz.). Crop

Sci.

9:354-357.

Dowling, P. M., R. J. Clements, and J. R.

McWilliam.

1971.

Establishment and survival of pasture species from seeds sown on the soil surface.

Aust.

J. Agric.

Res. 22:61-74.

Drawe,

D. L. 1977.

A study of five methods of mechanical brush control in south Texas.

Rangeman's J. 4:37-39.

Eckert, R. E., Jr., F. F. Peterson, M. S.

Meurisse, and

J. L. Stephens.

1986.

Effects of soil-surface morphology on emergence and survival of seedlings in big sagebrush communities. J. Range Manage.

39:414-420.

183

Evans, G. R. 1987.

Identification of future research. p.

309-311. In: G.

W. Frasier and R.

W. Evans

(eds.)

Proceedings of Symposium,

Seed and Seedbed Ecology of Rangeland Plants. 21-23 April 1987,

Tucson,

Arizona. U.S.D.A., Agricultural Reserch Service.

Evans, R. A., and J.

A.

Young. 1970.

Plant litter and establishment of alien annual weed species in rangeland communities. Weed Sci. 18:697-703.

Evans, R.

A., and J.

A. Young. 1972a. Microsites requirements for establishment of annual rangeland weeds. Weed Sci. 20:350-356.

Evans, R.

A. and

J.

A. Young.

1972b.

Germination and establishment of Salsola

in relation to seedbed environment. II. Seed distribution, germination, and seedling growth of

Salsola

and microenvironmental monitoring of the seedbed. Agron. J. 64:219-224.

Fay, P. K., and W. A. Olson. 1978.

Technique for separating weed seed from soil. Weed Sci. 26:530-

533.

184

Ferguson, H. and G. O. Boatwright. 1968.

Effects of environmental factors on the development of the crown node and adventitious roots of winter wheat

(Triticum aestivum).

Agron.

J.

60:258-260.

Fisher, C. E., C. H. Meadors, R. Behrens, E. D. Robinson,

P. T. Marion and H. L. Morton 1959. Control of mesquite on grazing lands. Tex.

Agr.

Exp. Sta. Bull.

935.

Fowler, N. L.

1986. Microsite requirements for germination and establishment of three grass species. Amer. Midi. Nat. 115:131-145.

Frasier, G. W., D. A. Woolhiser, and J. R. Cox.

1984.

Emergence of two warm-season grasses as influenced by the timing of precipitation: a greenhouse study.

J. Range Manage. 37:7-11.

Fulbright,

T. E., A. M. Wilson and E. F. Redente. 1984.

Effects of temporary dehydration on growth of green needlegrass

(Stipa viridula Trin.) seedlings. J.

Range Manage.

37:462-464.

Fulbright,

T. E., A. M. Wilson and E. F. Redente. 1985.

Green needlegrass seedling morphology in relation to planting depth. J. Range Manage. 38:266-270.

185

Glendening,

G. E. 1941.

Development of seedling of

Heteropocion contortus as related to soil moisture and competition. Bot. Gazette 102:684-698.

Glendening,

G. E.

1942.

Germination and emergence of some native grasses in relation to litter cover and soil moisture.

Agron.

J.

34:797-804.

Granstrom,

A. 1982.

Seed banks in five boreal forest stands originating between 1810 and

1963.

Can. J.

Bot.

60:1815-1821.

Griffiths, D. 1901.

Range Improvements In Arizona.

U.S.D.A.

Bureau of Plant Industry, Bull.

4. 31 p.

Griffiths, D. 1904.

Range Investigations In Arizona.

U.S.D.A.

Bureau of Plant Industry, Bull. No. 67.

62 p.

Hadjichristodoulou,

A., A. Della, and J.

Photiades. 1977.

Effects of sowing depth on plant establishment, tillering capacity and other agronomic characters of cereals. J. Agric.

Sci. 89:161-167.

Haferkamp,

M.

R.,°D.

C.

Ganskopp,

R. F. Miller, and F. A.

Sneva. 1987.

Drilling versus imprinting for establishing crested wheatgrass in the sagebrushbunchgrass steppe. J. Range Manage. 40:524-530.

186

Harper, J. L., J. N. Clatworthy,

I. H.

McNaughton, and G.

R.

Sagar. 1961.

The evolution and ecology of closely related species living in the same area. Evolution

15:209-227.

Harper, J. L., J. T. Williams and G. R.

Sagar. 1965.

The behavior or seeds in soil. I. The heterogeneity of soil surfaces and its role in determining the establishment of plants from seed. J. Ecol. 53:273-

286.

Harper, J. L. 1977.

Population Biology of Plants.

Academic Press, London. 892 p.

Hassanyar,

A. S. and A. M. Wilson. 1978.

Drought tolerance of seminal lateral root apices in crested wheatgrass and Russian wildrye.

J. Range Manage. 31:

254-258.

Hauser, V.L. 1986.

Water injection in grass seed furrows.

Trans. Amer. Soc. Agric. Eng. 29:1247-1253.

Herbel,

C. H., G. H. Abernathy, C. C. Yarbrough, and D.

K. Gardner.

1973. Rootplowing and seeding arid rangelands in the Southwest. J. Range Manage.

26:193-197.

187

Herbel,

C.

H. and

R. E. Sosebee. 1969. Moisture and temperature effects on emergence and initial growth of two range grasses.

Agron. J. 61:628-631.

Hillel, D. 1982.

Introduction to Soil Physics. Academic

Press, Orlando, Florida. 364 p.

Hormay,

A. L. 1970. Principles of Rest-rotation Grazing and Multiple-use Land Management. USDI,

Bureau of

Land Management.

26 p.

Howard, W. E. 1950. Wildlife depredations on broadcast seedings of burned brushlands. J.

Range Manage.

3:291-298.

Hull,

H.

M., L.

N. Wright and C. A. Bleckmann. 1978.

Epicuticular wax ultrastructure among lines of

Eragrostis lehmanniana Nees developed for seedling drouth tolerance. Crop Sci. 18:699-704.

Humphrey, R. R. 1958.

The desert grassland.

Bot. Rev.

24:193-252.

Hyder, D.

N., A. C. Everson, and R. E. Bement. 1971.

Seedling morphology and seeding failures with blue grama. J.

Range Manage. 24:287-292.

188

Jerling,

L.

1983.

Composition and viability of the seed bank along a successional gradient on a Baltic sea shore meadow.

Holarctic Ecol. 6:150-156.

Johnson, D. A.

1980.

Improvement of perennial herbaceous plants for drought-stressed western rangelands. In:

N. C. Turner and P. J. Kramer (eds.). Adaptations of

Plants to Water and High Temperature Stress. John

Wiley, New york.

Jordan, G. L.

1981.

Range Seeding and Brush Management On

Arizona Rangelands. Bull.

T81121,

Agri. Exp. Sta.,

Univ. of Ariz. Tucson.

88 p.

Jordan, G. L. and M. L. Maynard.

1970.

The San Simon

Watershed: revegetation. Prog.

Agric. in Ariz.

22:4-7.

Judd, I. B. and L. W. Judd.

1976.

Plant survival in the arid southwest

30 years after seeding. J. Range

Manage.

29:248-251.

Kaviani.

N., R. N. Gibbs, and R. F.

Horrell. 1985.

Seed placement measuring technique for direct drilling machines

- a note. New Zealand J. Exp. Agric.

13:191-194.

189

Kay, B. L.

1987.

Modifications of seedbeds with natural and artificial mulches. p. 221-224.

In: G. W.

Frasier and R. W. Evans (eds.) Proceeding of

Symposium, Seed and Seedbed Ecology of Rangeland

Plants.

21-23

April

1987,

Tucson, Arizona.

U.S.D.A.,

Agricultural Research Service.

Kinsinger,

F. E.

1962.

The relationship between depth of planting and maximum foliage height of seedlings of

Indian ricegrass. J. Range Manage. 19:279-283.

Knipe, O. D.

1968.

Effects of moisture stress on germination of Alkali sacaton, Galleta, and blue grama.

J. Range Manage.

21:3-4.

Malone, C. R.

1967.

A rapid method for enumeration of viable seeds in soil. Weed

Sci. 15:381-382.

Mathis, G. W., M. M.

Kothmann and W. J.

Waldrip. 1971.

Influence of root plowing and seeding on composition and forage production of native grasses. J. Range

Manage. 24:43-47.

McGinnies,

W. J.

1960.

Effects of moisture stress and temperature on germination of six range grasses.

Agron.

J.

52:159-162.

190

McGinnies, W. J. 1974. Effect of planting depth on seedling growth of Russian wildrye (Elymus lunceus

Fisch.) J. Range Mange. 27:305-307.

McKenzie, R. E., D. H. Heinrichs and L. J. Anderson 1946.

Maximum depth of seeding eight cultivated grasses.

Scientific Agric. 26:426-431.

McWilliam, J. R., R. J. Clements, and P. M. Dowling.

1970.

Some factors influencing the germination and early seedling development of pasture plants. Aust.

J. Agric. Res. 21:19-32.

Moore, R. P. 1943. Seedling emergence of small-seeded legumes and grasses. Agron. J. 35:370-381.

Moore, J. M., and R. W. Wein. 1977. Viable seed populations by soil depth and potential site recolonization after disturbance. Can. J. Bot.

55:2408-2412.

Morrison, D. F. 1976. Multivariate Statistical Methods.

2nd ed. McGraw-Hill Book Company, New York.

Mutz,

J. L. and C. J. Scifres. 1975. Soil texture and planting depth influence bufflegrass emergence. J.

Range Manage. 28:222-226.

191

Nelson, J.

K. and S. Gabel. 1987.

Initial establishment of 14 forage species on root plowed creosotebush

(Larrea tridentata) rangeland in Presidio County,

Texas. Tex. J. Agr. Nat. Resources. 1:49-52.

Nelson, J.

R., A. M. Wilson, and C.

J. Goebel. 1970.

Factors influencing broadcast seeding in bunchgrass range. J.

Range Manage. 23:163-169.

Newman, P.

R. and L. E. Moser. 1988. Grass seedling emergence, morphology, and establishment as affected by planting depth. Agron. J. 80:383-387.

Olmsted,

C. E. 1941. Growth and development in range grasses.

I. Early development of

Bouteloua curtipendula in relation to water supply.

Bot.

Gazette 102:499-519.

Olmsted, C. E. 1942. Growth and development in range grasses.

II. Early development of Bouteloua curtipendula as affected by drought periods.

Bot.

Gazette

103:531-542.

Oomes,

M. J.

M. and W. T. Elberse. 1976.

Germination of six grassland herbs in microsites with different water contents. J. Ecol. 64:745-755.

192

Oppenheimer,

H. R.

1960. Adaptation to drought:

Xerophytism.

In: Plant-Water Relationships in Arid and

Semi-Arid

Conditions. UNESCO,

Switzerland.

Pareja, M. R.,

D.

W. Staniforth, and

G.

P. Pareja. 1985.

Disturbance of weed seed among soil structural units. Weed

Sci. 33:182-189.

Pearson, C.

J. and

R. L.

Ison. 1987.

Agronomy of

Grassland Systems. Cambridge

Univ.

Press. Cambridge.

169 p.

Plummer, P. A., A. C. Hull,

Jr., G. Stewart, and J. H.

Robertson. 1955.

Seeding Rangelands in Utah, Nevada,

Southern Idaho, and Western Wyoming.

USDA,

Forest

Service, Agric. Handbook

71. 73 p.

Roundy, B. A. and

G. L. Jordan. 1988.

Vegetation changes in relation to livestock exclusion and rootplowing in southeastern Arizona. The

Southwestern Nat.

33:425-436.

Salim, M. H.,

G.

W.

Todd and A. M. Schlehuber. 1965.

Root development of wheat, oats, and barley under conditions of soil moisture stress. Agron J. 57:603-

607.

193

Savory,

A. 1978. A holistic approach to ranch management using short duration grazing.

Proc.

First Int.

Rangeland Congress. pp. 555-557.

Schmutz,

E. M. and M. F. Al-Rabbat. 1969.

Crested wheatgrass and winterfat emergence under simulated drouth. Prog. Agric.

Ariz.

21:3-5.

Sheldon, J. C. 1974. The behavior of seeds in soil. III.

The influence of seed morphology and the behavior of seedlings on the establishment of plants from surface-lying seeds. J. Ecol. 62:47-66.

Simanton,

J. R. and G. L. Jordan. 1986.

Early root and shoot elongation of selected warm-season perennial grasses.

J.

Range Manage. 39:63-67.

Sokal,

R. R. and F.

J.

Rohlf. 1981.

Biometry

(2nd ed.).

W. H. Freeman and

Co.,

New York. 859 p.

Staaf,

H., M. Jonsson, and

L. Olsen. 1987.

Buried

Germinative seeds in mature beech forests with different herbaceous vegetation and soil types.

Holarctic Ecol. 10:268-277.

Stoddart,

L.

A., A.

D. Smith, and T. W. Box. 1975. Range

Management.

McGraw-Hill, New York. 532 p.

Tadmor, N. H. amd Y. Cohen. 1968.

Root elongation in the preemergence stage of Mediterranean grasses and legumes. Crop Sci. 8:416-419.

194

Taylor, H. M.

1971.

Soil conditions as they affect plant establishment, root development and yield. p. 292-

312.

In: K. K. Barnes, W. M. Carleton, H. M.

Taylor, R. I.

Throckmorton and G. E. Vanden Berg

(eds.). Soil compaction of agricultural soils. Amer.

Soc.

Agr.

Eng., St. Joseph, MI.

Thornber,

J. J.

1905.

Range improvement. Arizona Agric.

Exp. Sta. Annual Report.

16:17-22.

Thornber,

J. J. 1907.

Range conditions. Arizona Agric.

Exp. Sta. Annual Report.

18:226-229.

Thornber,

J. J.

1908.

Range improvement. Arizona Agric.

Exp. Sta. Annual Report. 19:353-355.

Thornber, J. J. 1909.

Range condition. Arizona Agric.

Exp. Sta. Annual Report.

20:575-582.

Thornber,

J. J.

1910.

Condition on the ranges. Arizona

Agric. Exp. Sta. Annual Report.

21:371-376.

Thornber,

J. J.

1911.

Range condition. Arizona Agric.

Exp. Sta. Annual Report.

22:533-540.

195

Tiedemann, A.

R. and

E.

M. Schmutz. 1966.

Shrub control and reseeding effects on the oak chaparral of

Arizona.

J. Range Manage. 19:191-195.

Tischler, C.

R. and P. W. Voigt. 1983.

Effects of planting depth on vegetative characteristics of three forage grasses at days post emergence. Crop

Sci. 23:481-484.

Trevis,

L. 1958. Interrelations between the harvester ant

Veromessor pergandei (Mayr) and some desert emphemerals.

Ecology

39:695-704.

U.S. Dept. of Commerce -

Weather Bureau. 1957-87.

Climatological data - Arizona,

Vols.

61-91

,

Vallentine,

J.

F. 1963. Range

Seeding in Utah. Extension circular

307.

Utah

State University, Logan. 20 p.

Vallentine,

J.

F. 1989. Range

Development and

Improvements.

(3rd ed.) Academic Press. San Diego.

524 p.

Van Der Sluijs

D. H. and D. N. Hyder. 1974.

Growth and longevity of blue grama seedlings restricted to seminal roots. J. Range Manage. 27:117-119.

196

Wesson, G., and P. F. Wareing.

1969.

The role of light in the germination of naturally occuring populations of buried weed seeds. J. Exp. Bot.

20:402-413.

Whalley, R. D. B., C. M.

McKell, and L. R. Green. 1966.

Seedling vigor and the early nonphotosynthetic stage of seedling growth in grasses. Crop

Sci. 6:147-150.

Wilson, A. M. and D. D.

Briske. 1979.

Seminal and adventitious root growth of blue grama seedlings on the Central Plains. J. Range Manage.

32:209-213.

Wright, L. N.

1964. Drouth tolerance-program-controlled environmental evaluation among range grass genera and species. Crop

Sci. 4:472-474.

Wright, L. N.

1971.

Drought influence on germination and seedling emergence. In: K. L. Larson and J. D.

Eastin

(eds.). Drought Injury and Resistance in

Crops, Spec.

Publ.

No.

2

Crop

Sci.

Soc. Am.,

Madison, Wisconsin. 88 p.

Wright, L. N. and A. K.

Dobrenz. 1973.

Efficiency of water use and associated characteristics of Lehmann lovegrass.

J. Range Manage.

26:210-212.

197

Wright, N. and

L. J. Streetman. 1960.

Grass Improvement for the

Southwest. Tech.

Bull.

143. Agr. Exp. Sta.,

Univ. Arizona, Tucson.

Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement