Persistence of seven forage legumes under three grazing regimes

MISSING IMAGE

Material Information

Title:
Persistence of seven forage legumes under three grazing regimes
Physical Description:
xiv, 150 leaves : ill. ; 28 cm.
Language:
English
Creator:
Muir, James Pierre, 1958-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1989.
Bibliography:
Includes bibliographical references (leaves 140-148).
Statement of Responsibility:
by James Pierre Muir.
General Note:
Typescript.
General Note:
Vita.
General Note:
Includes abstract.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001549206
notis - AHG2778
oclc - 22484859
System ID:
AA00003774:00001

Full Text














PERSISTENCE OF SEVEN FORAGE LEGUMES UNDER
THREE GRAZING REGIMES













By

JAMES PIERRE MUIR


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1989















ACKNOWLEDGMENTS


There are, I am certain, many more constructive ways

for a university researcher to further his or her career

than to sacrifice the thousands of hours needed to guide a

disciple through the tortuous paths of graduate research.

Nevertheless, Dr. W. D. (Buddy) Pitman has been willing to

invest that time in this work. I am grateful for that

sacrifice and hope that his satisfaction will come as I, his

student, fulfill the scientific and professional

expectations which he has consciously and subconsciously set

for me.

I owe many thanks to Dr. Ken Quesenberry as well. His

efforts not just in reviewing the work presented herein but

in preparing me academically for this exercise have been

invaluable. His wry humor in times of stress and

admonishing prods when energies flagged made the goal

attainable.

I believe that no three people so fully incorporate

what I wish to accomplish in agriculture as do the remaining

three reviewers of this text. Drs. George Tanner, Joseph

Conrad and Loy Crowder provided me with not simply the

direction of committee members or the instruction of









classroom teachers but with symbols of what agricultural

researchers should strive for. My thanks go to them for

their contributions to this work as well as their

inspiration to my career.

I cannot forget the periodic revitalization I have

received over the past five years from my two co-

conspirators, Graham Knox and Steve Calhoun. By hook and by

net they managed to keep my spirit and my mind together as

we all became agronomists.

What is a person's career and professional

accomplishments without the reassuring stability of a

family? Although my wife Kaycie did not write a chapter and

my son Petie was not allowed to scribble a crayon marking on

this manuscript, their imprint is there. Put simply,

without them, the efforts invested in this dissertation

would have been without meaning. Although sacrifices were

made in terms of time spent together as a family, I can say

with full confidence that they have been, and always will

be, far more important to me than any professional endeavors

I should undertake on behalf of my career. I thank them for

being there from the onset to the final crystallization of

this dissertation.


iii

















TABLE OF CONTENTS


Page


ACKNOWLEDGEMENTS . . .

LIST OF TABLES . . .

LIST OF FIGURES . . .

ABSTRACT . .

CHAPTER ONE INTRODUCTION . .

CHAPTER TWO LITERATURE REVIEW . .

Plant Adaptations to Foraging .
Regrowth Capacity . .
Internal Composition . .
Availability . .
Associated Species . .
Growth Habit . .

Effects of Direct Grazing on Legume Production


ii

vii

ix


xiii


Methods for Measuring Plant Acceptability

Previous Studies in Tropical Forage Legume
Persistence . .
Common Grazing . .
Frequency of Defoliation .
Defoliation Intensity .
Put-and-take Management .
Deferred Grazing . .

CHAPTER THREE METHODS AND MATERIALS .


. 18


18
S. 19
21
25
S. 29
S. 32

S. 36


Field Study . .
Experimental Layout .
Grazing Management .
Persistence Evaluation
Statistical Analysis .

Pot Study . .


S 36
S 37
S 39
S 40
S 41











CHAPTER FOUR RESULTS . .

Field Study . .
Height of Perennials, 1987 .
Persistence Within Perennials, 1987-1989
Persistence Among Perennials, 1987-1989
Persistence Within Annuals, 1987-1989 .
Persistence Among Annuals, 1987-1989 .


Pot Study . .
Winter Harvest .
Root mass .
Root total non-structural
carbohydrate percent .
Root total non-structural
carbohydrate mass .
Herbage mass .
Herbage nitrogen percent
Herbage nitrogen mass .
Leaf mass .
Leaf-stem ratio .
Flower and pod mass .
Spring Harvest .


Root mass: species X winter harvest
Root mass: autumn clipping X
winter harvest . .
Herbage mass: species X
winter harvest . .
Herbage mass: autumn clipping X
winter harvest . .
Herbage nitrogen percent: species X
autumn clipping . .
Herbage nitrogen mass: species X
winter harvest . .
Correlations Between Winter and
Spring Factors . .

CHAPTER FIVE DISCUSSION . .


Species Persistence under Grazing Management
Aeschynomene americana .
Alysicarpus vaqinalis . .
Desmanthus virgatus . .
Desmodium heterocarpon .
Galactia elliottii . .
Macroptilium lathyroides .
Vigna adenantha . .

Factors Affecting Persistence .
Climatic Adaptation . .
Moisture Stress . .


Page
46


S 81
S 81
S 81

S 85

S 87
S 89
S 92
S 94
S 96
S 99
S 99
S 102


103

103

106

107

107

109

111

114


114
114
116
116
117
118
120
121

122
122
122










Page
Temperature Stress . .. 125
Microenvironment . .. 127
Management Factors . .. 129
Direct influences .. 129
Indirect influences . 132

Relationships between Defoliation and
Plant Composition . .. 132

CHAPTER SIX CONCLUSIONS .. 135

LITERATURE CITED . .. 140

BIOGRAPHICAL SKETCH . .. 149















LIST OF TABLES


Table Page

1 Number of subsamples per experimental
unit in the pot study ... . 45

2 Plant height means of Alysicarpus vaqinalis
(AV), Desmodium heterocarpon (DH), Galactia
elliottii (GE) and Desmanthus virgatus (DV)
under zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing
treatments during the 1987 growing season .47

3 Height of Alysicarpus vaqinalis (AV)
Desmodium heterocarpon (DH), Galactia
elliottii (GE) and Desmanthus virgatus (DV)
under zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing treatments
in December 1987 as a percent of ungrazed
plots . .... 56

4 Mean percent survival comparison within
and among perennials Alysicarpus
vaqinalis (AV), Desmodium heterocarpon (DH),
Galactia elliottii (GE), Desmanthus virgatus
(DV), and Vigna adenantha (VA) under zero (Z),
spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments using May 1987 as
a base date . ... .58

5 Mean percent survival comparison within and
among Macroptilium lathyroides (ML) and
Aeschvnomene americana (AA) under zero (Z),
spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments using May 1987 as
a base date ...... .. . 75

6 Mean root mass, root total non-
structural carbohydrate (TNC) percent and
root TNC mass of Galactia elliottii,
Desmodium heterocarpon and Desmanthus
virgatus following autumn clipping
treatments . ... 82


vii









Table Page

7 Mean herbage mass, herbage nitrogen percent
and herbage nitrogen mass of
Galactia elliottii, Desmodium heterocarpon
and Desmanthus virgatus following autumn
clipping treatments . ... 90

8 Mean leaf mass, leaf-stem ratio and flower
and legume mass of Galactia elliottii,
Desmodium heterocarpon and Desmanthus
virgatus following autumn clipping
treatments . .... .. 98

9 Mean root mass and herbage mass
of Galactia elliottii, Desmodium
heterocarpon and Desmanthus virgatus
allowed 16 wk recovery following winter
harvest at 3-cm height and an unharvested
control . ... 104

10 Mean root mass and herbage mass of
Galactia elliottii, Desmodium
heterocarpon and Desmanthus virqatus allowed
16 wk recovery after superimposing a winter
harvest at 3-cm height and an unharvested
control on autumn clipping treatments 105

11 Mean herbage nitrogen percent of
Galactia elliottii, Desmodium heterocarpon
and Desmanthus virgatus allowed 16 wk
recovery following autumn clipping
treatments . ... 108

12 Mean herbage nitrogen mass of Galactia
elliottii, Desmodium heterocarpon
and Desmanthus virgatus allowed 16 wk
recovery after being submitted to a winter
harvest at 3-cm height and an unharvested
control . . 110

13 Correlation between pre-winter root and
root total non-structural carbohydrate (TNC)
mass with post-winter herbage mass and
herbage nitrogen mass in three forage
legumes allowed 16 wk recovery after being
subjected to three autumn clipping
treatments and winter harvested at 3-cm
heights . . 112


viii















LIST OF FIGURES


Figure Page

1 Effect of zero (Z), spring/summer (S/S)
and spring/summer/fall (S/S/F) grazing on
height of Alysicarpus vaginalis (A.V.)
beginning 20 May 1987 with cattle added day
0 and taken off S/S day 115 ... .49

2 Effect of zero (Z), spring/summer (S/S)
and spring/summer/fall (S/S/F) grazing on
height of Desmodium heterocarpon (D.H.)
beginning 20 May 1987 with cattle added day
0 and taken off S/S day 115 ... .50

3 Effect of zero (Z), spring/summer (S/S)
and spring/summer/fall (S/S/F) grazing on
height of Desmanthus virgatus (D.V.)
beginning 20 May 1987 with cattle added day
0 and taken off S/S day 115 ... .52

4 Effect of zero (Z), spring/summer (S/S)
and spring/summer/fall (S/S/F) grazing on
height of Galactia elliottii (G.E.)
beginning 20 May 1987 with cattle added day
0 and taken off S/S day 115 .. 53

5 Mean percent height remaining in spring/
summer (S/S) and spring/summer/fall (S/S/F)
grazed plots of Alysicarpus vaginalis (A.V.),
Desmodium heterocarpon (D.H.), Desmanthus
virgatus (D.V.) and Galactia elliottii
(G.E.) compared to ungrazed plots on
December 1987 .. . 55

6 Mean percent persistence of Alysicarpus
vaginalis starting 22 May 1987 under zero
(Z), spring/summer (S/S) and spring/summer/
fall (S/S/F) grazing treatments ... 59

7 Mean percent persistence of Desmodium
heterocarpon starting 22 May 1987 under zero
(Z), spring/summer (S/S) and spring/summer/
fall (S/S/F) grazing treatments .. 61











8 Mean percent persistence of Desmanthus
virgatus starting 22 May 1987 under zero
(Z), spring/summer (S/S) and spring/summer/
fall (S/S/F) grazing treatments ... 62

9 Mean percent persistence of Galactia
elliottii starting 22 May 1987 under zero
(Z), spring/summer (S/S) and spring/summer/
fall (S/S/F) grazing treatments ... 64

10 Mean percent persistence of Vigna
adenantha starting 22 May 1987 under zero
(Z), spring/summer (S/S) and spring/summer/
fall (S/S/F) grazing treatments ... 66

11 Mean persistence of perennials Alysicarpus
vaginalis (AV), Desmodium heterocarpon (DH),
Desmanthus virgatus (DV), Galactia elliottii
(GE) and Vigna adenantha (VA) under zero (Z),
spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments on December 1987
using May 1987 as a base date 68

12 Mean persistence of perennials Alysicarpus
vaginalis (AV), Desmodium heterocarpon (DH),
Desmanthus virgatus (DV), Galactia elliottii
(GE) and Vigna adenantha (VA) under zero (Z),
spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments on December 1988
using May 1987 as a base date . 69

13 Mean percent persistence of Macroptilium
lathyroides starting 22 May 1987, under
zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing
treatments .. . 76

14 Mean percent persistence of Macroptilium
lathyroides (ML) and Aeschynomene americana
(AA) on December 1987 and December 1988
under zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing
treatments using May 1987 as a base 78

15 Mean percent persistence of Aeschynomene
americana starting 22 May 1987, under
zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing
treatments . . 79


Figure


Page











16 Mean root mass of Galactia elliottii
(GE), Desmodium heterocarpon (DH) and
Desmanthus virgatus (DV) under extended
clipping (EX), early clipping (ER) and
unclipped (U) treatments ... .84

17 Mean root total non-structural carbohydrate
(TNC) percent of Galactia elliottii
(GE), Desmodium heterocarpon (DH) and
Desmanthus virgatus (DV) under extended
clipping (EX), early clipping (ER) and
unclipped (U) treatments ... .86

18 Mean root total non-structural carbohydrate
(TNC) mass of Galactia elliottii
'(GE), Desmodium heterocarpon (DH) and
Desmanthus virgatus (DV) under extended
clipping (EX), early clipping (ER) and
unclipped (U) treatments ... .88

19 Mean herbage mass of Galactia elliottii
(GE), Desmodium heterocarpon (DH) and
Desmanthus virgatus (DV) under extended
clipping (EX), early clipping (ER) and
unclipped (U) treatments ... .91

20 Mean herbage nitrogen percent of Galactia
elliottii (GE), Desmodium heterocarpon (DH)
and Desmanthus virqatus (DV) under extended
clipping (EX), early clipping (ER) and
unclipped (U) treatments ... .93

21 Mean herbage nitrogen mass of Galactia
elliottii (GE), Desmodium heterocarpon (DH)
and Desmanthus virgatus (DV) under extended
clipping (EX), early clipping (ER) and
unclipped (U) treatments .. 95

22 Mean leaf mass of Galactia elliottii
(GE), Desmodium heterocarpon (DH) and
Desmanthus virgatus (DV) under extended
clipping (EX), early clipping (ER) and
unclipped (U) treatments ... .97

23 Mean leaf-stem of Galactia elliottii
(GE), Desmodium heterocarpon (DH) and
Desmanthus virqatus (DV) under extended
clipping (EX), early clipping (ER) and
unclipped (U) treatments .. 100


Figure


Page











24 Mean seed and pod mass of Galactia
elliottii (GE), Desmodium heterocarpon (DH)
and Desmanthus virgatus (DV) under extended
clipping (EX), early clipping (ER) and
unclipped (U) treatments .. 101


xii


Figure


Page















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

PERSISTENCE OF SEVEN FORAGE LEGUMES
UNDER THREE GRAZING REGIMES

By

JAMES PIERRE MUIR

August 1989

Chairman: William D. Pitman
Cochairman: Kenneth H. Quesenberry
Major Department: Agronomy



A degraded flatwoods pasture at Ona, Florida, was

planted in 1986 with strips of Aeschvnomene americana L.,

Alysicarpus vaginalis D.C., Vigna adenantha (G. F. Meyer)

Marechal, Macherpa and Stainier, Desmodium heterocarpon (L.)

D.C. cv. Florida, Galactia elliottii Nuttal., Macroptilium

lathyroides (L.) Urb. and Desmanthus virgatus (L.), Willd.

Late spring through summer (S/S) grazing, late spring

through fall (S/S/F) grazing and ungrazed treatments were

imposed during 1987 and 1988. Grazing pressure was 2.2

yearling heifers per ha in the summer, 1.2 yearling heifers

per ha during the fall. Plant heights of the four upright-

growing perennials were measured during 1987 and plant


xiii









population survival using May 1987 numbers as a base were

taken during 1987, 1988 and 1989.

Height measurements indicated that D. virgatus,

Galactia elliottii, Desmodium heterocarpon and Alysicarpus

vaqinalis in the two grazed treatments had reduced plant

heights by June. Differentiation between all treatments was

apparent in the last two species by mid-December.

Despite high persistence of A. vaqinalis and Desmodium

heterocarpon in May 1988, populations declined by May 1989.

Desmanthus virgatus and Viqna adenantha had persistence

values near 25% for the S/S treatment May 1989. Galactia

elliottii, in May 1989, had the highest persistence of all

entries with no differences between the grazed treatments.

A fall clipping trial indicated that Desmodium

heterocarpon invested in seed production when not clipped.

Desmanthus virgatus partitioned photosynthate to both

herbage and roots while unclipped Galactia elliottii showed

a marked increase in root mass and root non-structural

carbohydrate but not effect on herbage production.

Aeschynomene americana, Macroptilium lathyroides and

Desmanthus virgatus persistence mechanisms included

unpalatable stem and seed production. Vigna adenantha

survived through rapid regrowth, Galactia elliottii by

storing nutrient reserves for early spring growth, while

Alysicarpus vaqinalis and Desmodium heterocarpon survived

via prostrate growth and seed production.


xiv















CHAPTER ONE

INTRODUCTION



Grazing and browsing animals have had a large influence

on the evolution of many plants from which they derive

sustenance. Depending on grazing intensity and/or cycle as

well as other environmental factors which affect herbage

recovery, range and pasture plants have developed

morphological or chemical factors which assist them in

either deterring or surviving predation (Hodgkinson and

Williams, 1983). Some species outgrow these terrestrial

herbivores while others become inconspicuous via decumbent

growth. A few have developed chemicals in the herbage which

discourage ingestion via taste or anti-digestive factors. A

third survival mode makes use of simple regenerative

capacities in which plants acquire the ability to resprout

and regrow faster than neighboring species after being

damaged. In forage science it is this last group which

primarily interests the pasture manager because these plants

are acceptable to herbivores, nutritious and capable of

permanence in the system.

It is natural, then, that these same adaptations in

turn influence the acceptability of the plants as forage to











herbivores. Depending on the feed amount and species

available, harvesting animals will place differing pressures

upon specific species within the area open to ranging. The

rancher who plans to cultivate various legume and grass

species to enhance domesticated animal productivity should

therefore understand which species will be preferentially

selected in forage mixtures and which will not benefit

his/her program by being completely unacceptable to the

animals.

In peninsular Florida very little of this information

is available on tropical forage legumes. Pitman and

Kretschmer (1984) have conducted initial work with a number

of introduced legumes at the Ona Agricultural Research and

Education Center (Ona AREC) encompassing 19 accessions from

17 species as well as 50 entries in a further work (Pitman

et al., 1988). Grazing evaluation of single rows of these

legumes was conducted in Paspalum notatum Flugge swards.

Several species survived the grazing while others were

completely eliminated. Of the survivors, only those for

which seed was available could be studied further in this

experiment. All those which did not survive, however,

should not be excluded from further evaluation. The high

degree of variability and complete lack of persistence of

some legumes (for example Desmanthus virgatus (L.) Willd.,

which was utilized heavily by wildlife) cannot be attributed

solely to climatic or edaphic factors in these studies since











all plants were subjected to dense competition from grasses

and little range in grazing management levels. Under less

competition from grass or with more intense management,

results might well have been different. For example, Pitman

and Kretschmer (1984) reported very erratic establishment of

Viqna adenantha (G.F. Meyer) Marechal, Mascherpa and

Staineir and Vigna parkeri Bak. while under different

management Pitman and Singer (1985) obtained good pasture

coverage from the same species. The question yet to be

answered, therefore, is not only will cattle graze or avoid

species but what grazing management is needed to ensure

proper establishment and maximum persistence while still

maintaining high productivity.

Another aspect of tropical forage legumes which

concerns ranchers in Florida is pasture fertilization needs.

The purpose of the legume component in a grazing system is

to provide improved mineral content, especially nitrogen in

crude protein form. But must plant nutrients be added to

native flatwoods soils to ensure proper establishment and

persistence of these species? Producers, attempting to

economize as much as possible in a financially-strapped

cattle market, would prefer not to invest in fertilizer for

native pastures. Information on which forage legume species

might do well in unfertilized flatwoods pastures with native

grasses is therefore of interest.











Frequently, researchers look at legume persistence in

pastures by varying the stocking rate to which the species

are subjected. Florida, however, with its distinct seasonal

conditions where spring growth is often limited due to low

rainfall, demands additional approaches. Summer growth in

grasslands provides relatively abundant forage while fall

deferment may be needed to allow annual legumes to set seed

and perennial species to store energy for regrowth following

winter frost. The period in which tropical forage legumes

could be of greatest benefit to Florida ranchers is

therefore the so-called "summer slump" of August and

September. During these months herbage is abundant but low

in quality, causing reduced animal weight gains or even loss

despite excess forage (Pitman et al., 1984). One aspect of

this study was therefore centered on evaluating selected

legume species for potential to persist under grazing during

the late spring forage quantity deficit and the summer

forage quality slump.

This study sought to answer several questions pertinent

to forage legumes in South-central Florida:

Objective 1. Of the species which have done well in

initial introduction studies, determine which will persist

under the various grazing regimes selected.

Objective 2. Among those that survive, discern what

mechanisms are involved which allow species to tolerate the









5

various grazing treatments under conditions found in Central

Florida flatwoods.

Objective 3. Determine some of the relationships of

defoliation to plant composition as these relate to regrowth

in selected entries.















CHAPTER TWO

LITERATURE REVIEW



Plant Adaptations to Foraging



Plants may achieve persistence under grazing in various

ways. Although climate, soils, pathogens and insect

herbivory also affect plant survival mechanisms, animal

herbivory has had an extensive effect on plant growth

mechanisms such as reproductive cycles, regrowth

characteristics or internal biochemical composition. In

some more obvious strategies, plants may expand their

populations under herbivory via basal regrowth, prostrate

leaves and stems, well developed root nutrient storage,

rhizome and stolon growth, rapid regeneration from various

non-apical meristimatic tissue following trampling or

defoliation, rapid seeding, seed dormancy, morphological and

biochemical characteristics unfavorable to herbivore

ingestion and digestion, and many other means (Hodgkinson

and Williams, 1983).

Forage expansion via recruitment was reported by many

researchers either through vegetative propagation or

seeding. Vegetative propagation examples include Medicago











sativa L. (Campbell, 1974) and Macroptilium atropurpureum

(D.C.) Urb. (Hodgkinson and Williams, 1983). Jones and

Evans (1977) reported on soil seed reserves which were high

even in such perennials as Lotononis bainesii Baker,

Desmodium intortum (Mill.) Urb. Thell. and Trifolium repens

L. Taylor (1972) reported on several years' seed production

by annual legumes and noted that different cultivars of the

same species produced differing seed amounts under varying

climatic conditions, greatly affecting seedling numbers the

following years. Depending on previous seed crops and

hardseededness, many species survive non-consecutive, poor

seeding seasons according to Hagon (1974).



Regrowth Capacity



Individual plant regeneration after grazing also

contributes to persistence in many species. If associations

are grazed at a time when one component can regrow more

rapidly than another yet both are grazed equally, then that

species which recovers more quickly will likely exhibit

greater dominance. In Florida, Pitman (personal

communication) observed that many tropical legumes continued

to grow further into the fall than did Paspalum notatum.

The advantage which this C-4 grass possesses in superior

growth rates is reduced perhaps due to decreased daylength

or temperatures. In the spring, however, the roles are











reversed when the P. notatum tends to start responding to

warming trends faster than these legumes. Unlike the fall

situation, this difference may be a result of more extensive

fibrous root systems possessed by the grasses which give

them an advantage in this moisture-scarce period. The

normal legume advantage derived from deep taproots does not

exist for many legumes in these spodosol pastures due to

shallow hardpans in the soil.

Regrowth capacity is naturally more related to

hereditary factors, especially when considered across a

broad spectrum of environments. This is usually more

important in perennials, for example Leucaena leucocephala

(Lam.) de Wit (Hodgkinson and Williams, 1983) but can also

be important in annuals or short lived perennials such as

Stylosanthes humilis H.B.K. and Stylosanthes hamata (L.)

Taub. (Gardner, 1981). Differing environments and grazing

situations, of course, would greatly affect how well these

adaptations would serve population survival.



Internal Composition



Plant biochemical factors which deter grazing before it

even occurs are, at times, as important as recovery from

grazing. Forage digestibility has been shown in many cases

to influence passage rate and, therefore, acceptability to

herbivores. This can be easily observed in such









9

characteristics as coarse-stemmed morphologies implying high

fiber contents as Donnelly and Hawkins (1959) saw in

Lespedeza cuneata (Dumont) G. Don or as Hodges and McCaleb

(1972) reported with Aeschynomene americana L. Often,

however, these characteristics may be only a contributing

factor or may even be masked by other anti-nutritional

factors.

Other less superficially-apparent factors may involve

tannin, which is thought to lower digestibility (Donnelly

and Anthony, 1969 with Lespedeza cuneata), or such compounds

as alkaloids which appeared to be toxic to cattle (Thomas et

al., 1985). The alkaloid content can at times be such a

factor that animals will not eat the legumes at all. Thomas

et al. (1985) observed that in a mixed pasture of Andropogon

gayanus Kunth and Zornia brasiliensis Vog. in Brazil, cattle

grazed so selectively that the pasture was 100% legume after

3 yr. Not even during the dry season when the legume was

the only green material available did cattle eat the high

alkaloid-containing Zornia. There is some suspicion that

the problem in this species may also be related to sulfur

content (Lascano et al., 1981). Thomas et al. (1985)

observed that Calopogonium mucunoides Desv. and Desmodium

ovalifolium Walp. were also thought to have similar

characteristics since, when planted in associations with

grasses, they soon became the dominant component. Middleton

and Mellor (1982) discovered that Calopogonium caeruleum











Hemsl., had similar characteristics. In a pasture with

Panicum maximum Jacq this legume soon became the dominant

species and average daily liveweight gain declined from 0.5

to 0.2 kg head-1 over 2 yr.

Other plant biochemicals may actually attract

herbivory. Crude protein may raise digestibility, for

example with Lespedeza cuneata (Donnelly and Anthony, 1969).

Sugars and soluble carbohydrates may also raise palatability

(Cowlishaw and Alder, 1960) and thereby enhance selection.

There are, however, many studies which indicate that these

correlations do not hold or are at most confusing, a view

Warmke et al. (1952) held after they found no significant

correlation between soluble sugars and palatability in some

tropical legumes.



Availability



Herbage availability of particular forages may be an

important factor in forage selection. Reid (1951) stated

that forage accessibility has considerable influence on

utilization. Both Reid et al. (1967b) and Raymond and

Spedding (1966) pointed out than increased intake of

N-fertilized grasses may occur simply because these are more

abundant that the unfertilized plants. The same may be true

for P- and K-fertilized legumes. Bite size, the amount of

herbage harvested with each bite, may be more important in











these cases then palatability as Clark and Harris (1985)

indicated in their study of white clover spatial

distribution and its relationship to content in sheep

grazing. Other studies, however, indicate no relationship

between yield and grazing preference (for example Warmke et

al., 1952).



Associated Species



If legume preference is the targeted information, the

other legumes and associated grasses in which all are grown

may influence cattle selectivity. Lascano et al. (1981),

when comparing grazing selectivity of Centrosema pubescens

Benth. in Andropogon gayanus, Panimcum maximum and

Brachiaria decumbens Stapf. swards, determined that the

legume was grazed more heavily as grass defoliation

decreased in vitro digestibility, leaf/stem ratios and crude

protein of the different grasses. Likewise, Carvalho et al.

(1984) determined that Neonotonia wightii (R. Grah. ex

Wightii and Arn.) was selectively grazed in Panicum maximum,

Paspalum notatum and Brachiaria mutica Forsk. pastures only

when grass components decreased considerably from grazing.

Not only cattle selectivity but associated grass growth

habit may affect legume persistence. When compared to pure

grass or pure legume swards, legume components of grass-

legume mixtures are more likely to be inferior (Crowder and











Chheda, 1982). Tall bunch grasses may shade out some

legumes while allowing others to thrive in the open ground

between tufts. Dense mat-forming grasses, in contrast, may

choke out some low-growing legume species yet allow other

viney types greater productivity due to structural support

which allows more access to sunlight. In either of these

cases, cattle access may also be affected by associated

grass morphology (Strange, 1960). Although Strange (1960)

in Kenya found that there was simply no substitute for

general adaptability to the local environmental conditions,

twining legumes did better with taller, erect grasses while

more prostrate legumes were inferred to do best with lower

growing grasses.



Growth Habit



Perhaps the most important factor in forage persistence

under grazing is growth habit. Canopy structure, whether

composed of one species or an association, heavily

influences what and how much a grazing animal can collect as

Moore et al. (1985) showed with Aeschynomene americana and

Hemarthria altissima (Poir.) Stapf and C.E. Hubb in Florida.

These authors concluded that as herbage concentration

increased in the upper canopy, intake per bite of those

species present also increased.


1











Twining legumes, in particular, have difficulties

tolerating direct grazing and trampling. Thomas et al.

(1985) pointed out that some Galactia and Centrosema species

have stems and buds which are very vulnerable to this

pressure and therefore do not persist well in directly

grazed pastures. Those climbing species which can root at

the nodes, however, may overcome this limitation as do

Centrosema macrocarpum Benth., Centrosema pubescens (Thomas

et al., 1985) and Viqna parkeri (Cook and Jones, 1987).

Another morphological type which has difficulties under

direct and heavy grazing is the group containing upright

growing species with a limited capacity for regrowth from

non-apical meristimatic tissue (Thomas, 1986). Annuals

especially figure heavily in this group. Once grazed close

to the ground, these often are unable to recover

sufficiently to set seed, especially if harvested late in

the growing season. Thomas et al. (1985) and Thomas (1986)

also included in this group species of the genera

Stylosanthes and Centrosema.

In a study of 50 legume accessions, Pitman et al.

(1988) concluded that prostrate species such as a perennial

Alysicarpus vaqinalis (L.) D.C., Vigna parkeri and Desmodium

barbatum Benth. had better persistence under heavy grazing

than under lighter grazing pressure. They also pointed out

that light grazing produced the opposite effect in some of

these low-growing types.











There is often a trade-off of yield for persistence

when decumbent varieties are compared to erect ones. Leach

et al. (1982) found that in terms of numbers, spreading

Medicago sativa types were more persistent. In actual dry

matter production, however, the erect lines were higher

despite lower numbers of individuals.

An aspect which few research trials address is the

length of time different legumes take to establish and the

indication this might give of subsequent persistence. It

appears that those legumes, especially perennials, which

take longer to establish, are often more persistent due to

deeper, more developed root systems. Wong and Eng (1983)

grazed recently established pastures at 3.8 cattle ha'1 over

3 yr and found that quickly establishing perennials such as

Stylosanthes quianensis (Aubl) Sw. cv. Cook did not maintain

vigor and ground cover as did more slow to establish

Desmodium ovalifolium (L.) Benth. A prime example of the

slow to establish legumes, one adapted to well drained

soils, is Arachis glabrata Benth. According to Prine et al.

(1986) and Prine et al. (1981), germplasm of this legume is

well suited to drought sands but should not be grazed at

all the season of establishment after being planted from

rhizomes. Even faster establishing annuals such as

Aeschynomene americana L. (Kalmbacher et al., 1988) or

short-lived perennials like Macroptilium lathvroides (L.)

Urb. (Pitman et al., 1986) should not be grazed after


1









15

seedlings enter the upper canopy. Heavy grazing just prior

to this stage to avoid closed canopies and grass competition

does benefit the stands, however. Once well established,

grazing may commence.

Kretschmer (1988) mentioned two other factors which

assisted some legumes in overcoming temporary, if not

necessarily continuous, over-grazing. The first is simply

woodiness. Unpalatable stem will naturally discourage total

destruction of some species such as Leucaena spp. The other

is the capacity to store energy and resprout via crowns or

rhizomes. The main factor involved may be that these

anatomical structures are unavailable to the animal for

consumption as in the case of some Arachis spp.



Effects of Direct Grazing on Legume Production



Many studies have gathered information on the

above-ground effects of grazing on legumes. The most

obvious and most often documented, of course, is the

abundance of leaf and stem. Davidson and Brown (1985)

working with Neonotonia wightii and Desmodium intortum

described but one example in which excessive grazing

pressure decreased green matter while moderate stocking

rates maintained or increased the proportion of legume in

the pasture.











Fewer studies, however, have documented the actual

numbers of plants that survive differing grazing regimes.

Gardner (1981) working with Stylosanthes hamata indicated

that this species survived normal defoliation pressures via

weak perennation but resorted to soil seed reserves when

mature plants were destroyed by overgrazing. That author

found that only 0.03% of all seedlings survived to a third

growing season. Jones et al. (1980) discovered in a study

of grazed Macroptilium atropurpureum lines that plant

numbers declined differently for various lines as the

pasture matured. Increasing the stocking rate from 2 to 3

steers ha'1 resulted in plant densities of 5.3 and 1.9

plants m-2, respectively, in 5 yr. Jones (1979), working

with the same species and a range of grazing treatments,

however, found that grazing frequency at differing stocking

rates had no effect on plant density and seedling

regeneration.

Roberts (1980) reviewed several publications reporting

the effect of botanical composition on animal gain due to

differing stocking rates. His general conclusions were that

botanical composition in some pastures is not affected by

increased stocking rates which do, however, result in

decreased liveweight gain per head after a certain level.

Higher stocking rates on other pastures did result in

distinctive botanical composition changes but, surprisingly,

did not affect animal gain. This later finding would











indicate that the pastures resulting from poor management

were as good or better than the original mixture.

Very few studies, however, have concentrated on

determining what happens beneath the soil surface to grazed

and over-grazed plants. Even fewer studies have focused on

the effect of differing grazing regimes on root

carbohydrates. For information on this area the literature

is limited to studies on clipped plants. Trejos and Borel

(1985), for example, studied the effect of different cutting

heights and intervals on total non-structural carbohydrates

of Stylosanthes capitata Vog. No differences were found,

even in root and base content although percent and not total

carbohydrate was reported. They did note, however, that the

longer the plant rested after cutting, the greater the

percent carbohydrate recovery in the roots and bases. They

felt that they should shorten the recovery periods (27 d was

the shortest) to discern differences between treatments and

species.

Whiteman and Lulham (1970) studied the effect of

defoliation, both mechanical and animal, on nodule number

and weight. They found that in Macroptilium atropurpureum

mean weight per nodule was reduced whereas in Desmodium

uncinatum (Jacq.) D.C. grazing and cutting reduced nodule

number rather than size.











Methods for Measuring Plant Acceptability



The amount of forage which is removed during grazing is

termed herbage utilization (Heady, 1964). Preference, then,

can be measured in the relative utilization of the various

forages compared if availability is equal among species. As

Cook and Stoddart (1953) pointed out in the case of

rangelands, plant utilization (and by deduction preference)

is most commonly measured by using length or weight of the

grazed versus the ungrazed pasture portions. Reid et al.

(1967a) in their work with temperate grasses used a

palatability index calculated as the proportional dry matter

consumption from each treatment compared to that consumed in

all treatments combined.

Marten (1970) warned against not taking growth during

grazing into account when calculating utilization. Heady

(1964) used grazing exclosures to ensure this difference did

not affect the accuracy of the results obtained in his work.



Previous Studies in Tropical Forage Legume Persistence



Most studies in forage legume persistence have involved

plant survival under various grazing pressures and/or

frequencies rather than grazing periods as was studied in

this dissertation. These studies, however, hold some

pertinent information which can be utilized here.













Common Grazing



Although simply planting various legumes side by side

in a pasture and allowing cattle continuous, unlimited

access to them is a rather simple study method, it

encompasses several possible draw-backs. Foremost among

these is the danger of assuming that any plots with superior

persistence have the best species, cultivars or accessions.

The survival may simply be due to lower acceptance by the

animals. More palatable legumes may provide higher animal

weight gains in more appropriately managed situations when

compared to persistent but unpalatable lines.

It may be useful, therefore, to plant an ungrazed

control next to the grazed plots to determine whether

disappearance, if it occurs, is due to animal preference or

genetic limitations on the part of the plant. Even then,

however, too much competition from ungrazed grass or no

pressure from direct grazing and trampling can give

misleading or at least incomplete information.

Pitman and Kretschmer (1984) studied seventeen tropical

legumes under common grazing at one grazing pressure and

frequency in Florida with measurable results. After the

third year growth and second year grazing from May to

November, only four species showed any significant survival.

In terms of original planted area, Macroptilium lathyroides











covered 9% and Viqna luteola (Jacq) Benth. 6% of the

pasture, a non-significant difference from the other 13

species. Aeschynomene americana covered 28% and Vigna

parkeri topped the list at 42% cover, both significantly

higher than the others but also different from each other.

Most species were lost due to low vigor and failure to

regenerate. According to the authors, at least one,

(Macroptilium atropurpureum), succumbed to its incapacity to

tolerate direct grazing, while others disappeared at least

in part due to selective grazing (e.g. Desmathus virqatus).

The main reason for failures in persistence according to the

authors, however, was the heavy competition from the

associated Papalum notatum as well as the pressures of

direct grazing.

Pitman et al. (1986) also used common grazing pastures

to compare nine Stylosanthes quianensis var. quianensis

accessions, three Stylosanthes hamata and Stylosanthes

humilis accessions along with Aeschynomene americana in

peninsular Florida. Grazing pressure varied from 2 animals

ha1 during June and July the first year on two replications

to 3 animals ha- on all four replications the second year

to no animals the third year. Proportion of original cover

surviving at the end of the trial was highest for

Aeschynomene americana at 80% with the next closest being

two different accessions of Stylosanthes quianensis at 10%.









21

The last two were not significantly different from the rest

of the legumes which had no measurable survival.

Difficulties in this type of evaluation, especially its

repeatability in dissimilar edaphic and climatic zones, is

exemplified in the 1985 Pitman et al. publication which

covered four different sites in Florida and Costa Rica.

Thirty-six legumes were studied (although not all at every

place) under grazing. The authors did not claim to have

identical grazing pressures at all four sites since this

would be virtually impossible. Perhaps because of this and

the variation in edaphic and climatic conditions, the

results varied considerably from site to site.

The above-mentioned report exemplifies the great

variability which exists in the tropical and sub-tropical

areas when dealing with forage adaptability. A cautious

approach to this research area would entail initial

observations such as Pitman et al. (1985) did for each new

climatic or ecotypic zone followed by more specific work to

be done on the three or four species which show the most

promise in distinct environments.



Frequency of Defoliation



General forage management wisdom indicates that legumes

are far less able to survive repeated defoliation than are

grasses (Kretschmer, 1988). Smith (1970), working with









22

sheep and Medicago sativa in Australia, for example, showed

that persistence was much higher at both high and low

stocking rates when rotational grazing was utilized, thereby

limiting defoliation frequency. He also found that the more

paddocks were subdivided in the rotations, the higher the

productivity. Unfortunately, the economics of such highly-

divided pastures may be rather prohibitive in some

conditions. Leach et al. (1982) used a flexible grazing

frequency approach to study different M. sativa lines in

Australia. Their study compared lines of different

morphologies under a system in which plots were grazed to

the ground and then allowed to recover for 6 wk. Over a

period of 3 yr, persistence was better for spreading lines

than for erect ones although actual winter production was

higher for the locally developed erect cultivar 'Hunter

River.'

Lazier (1981) used an unusual grazing regime in which a

6-wk interval between grazing was the only set factor. At

grazing time cow-calf pairs were allowed to graze plots of

native Belizean Calopogonium caeruleum, Desmodium canum

(Gmel) Schinz and Thell and Desmodium gyroides (D.C.) Hask.

to an unspecified but even degree. Desmodium gyroides,

although it had a 34% mortality over the 3-yr period, proved

to have the highest grazing index (derived by multiplying

amount grazed by degree grazed) and the greatest dry matter

availability after the trial.











Whiteman (1969) in Australia also used an unusual

variation on the frequency theme in Chloris gavana Kunth

pastures planted to Macroptilium atropurpureum, Lotononis

bainesii, Glycine javanica L. and Desmodium uncinatum. For

2 yr, plots were grazed to a 6- to 10-cm stubble height by

sheep. What was unusual was that the intervals between

grazing were determined by seasonal, genetic and animal

directed capacities of the plots to recover. This turned

out to be approximately 6 wk during the warm season and 9 wk

in the cooler periods. The author did not state what

criterion was used to determine full recovery and regrowth.

Under this regime, Glycine appeared to have persisted best

while Lotononis exhibited the lowest survival and

productivity.

Jones and Clements (1987) studied various introductions

and lines of Centrosema virginianum (L.) Benth. as well as

Macroptilium atropurpureum cv. Siratro, Desmodium intortum

cv. Greenleaf, Centrosema pubescens cv. Belalto and Vigna

parkeri cv. Shaw under 3-wk rest, 4-d graze regimes. For 4

yr only 1.5 animals ha'1 were used but the last 4 yr the 4-d

grazing was extended on half the experiment to produce a 2.3

animals ha-1 stocking rate. Results varied for different

species under different conditions but after 8 yr at the low

pressure, only the Centrosema virginianum lines still

comprised significant portions of the pasture with the

highest line totaling 18% cover. All the other plots, with









24

the exception of Macroptilium atropurpureum, were persistent

during the first 5 yr at this level although differences did

exist. At the high stocking rate nothing persisted after 4

yr.

Jones (1979), examined not only different grazing

pressures but different grazing frequencies as well. In his

study, M. atropurpureum pastures were rested for 3, 6 and 9

wk between 4-d grazing regimes at stocking rates ranging

from 0.8 to 2.8 head ha' He found that the 3-wk rest was

inadequate and legume yield declined dramatically. He

further noted that although decline at the higher stocking

rates was greater than at the lower rates, the longer rest

period allowed much more effective recovery at all stocking

rates.

Less-frequent grazing, however, did not always result

in higher legume percentage in pasture studies. This was

especially true when low frequencies were combined with low

grazing pressure, as Santillan (1983) found in Ecuadorian

pasture mixes including Neonotonia wightii, Centrosema

pubescens, Panicum maximum and Pennisetum purpureum

Schumach. In these low-use situations the erect growing

grasses outcompeted the viney legumes and shaded them out.

Maraschin (1975) found the same general rule to be

applicable in a Florida study utilizing both viney and erect

legumes (Macroptilium atropurpureum and Desmodium intortum)

in a more decumbent type grass (Cynodon dactylon (L.) Pers.











cv. Coastcross-l). A balance avoiding over-use and

under-use, therefore, seems to work best when grazing

frequency can be varied in a management situation.



Defoliation Intensity



Most legumes have shown a decline in persistence with

an increase in stocking rate (Cowan et al., 1975).

Humphreys (1980) and Jones (1979) both indicated that viney

legumes were especially susceptible to an increase in

grazing pressure. Bryan and Evan (1973) agreed with this

general observation but added that trailing legumes

encountered difficulties not only under heavy but under

moderate stocking rates as well.

Of course, there is a limit to which even the hardier

species can withstand excessive grazing. Smith (1970),

working with Medicaqo sativa in a subtropical setting with

sheep, found that 2.0 wethers ha"' was the ideal stocking

rate in a continuous system but plants still survived even

at 4 animals ha1'. When he used 4.9 wethers ha"1 in a six

paddock rotational system, however, there were plant losses

from "digging."

There are some species which appear to thrive under

heavier grazing. Normally these are varieties with a

prostrate morphology which benefit from the removal of

upright growing competition (Bryan and Evan, 1973). Native











or naturalized legumes that have adapted to local heavy

grazing especially seem to fit into this category.

Partridge (1980) studied a locally prevalent Desmodium

heterophyllum (Willd.) D.C. in Fiji and discovered that it

persisted better and contributed more to cattle feed at

stocking rates over 3 head ha 1 where introduced species

like Macroptilium atropurpureum disappeared.

Sometimes, however, simply fostering better seedling

establishment, especially in the case of annuals, early in

the establishment of the pasture greatly increases

establishment and persistence rates. Stobbs (1969), for

example, found that Stvlosanthes qracilis H.B.K. did better

as stocking rates increased from 1.65 to 5.0 head ha".

Shaw (1978), working with the same genus but another

species, Stylosanthes humilis, found the same general rule

to be true and related the phenomenon directly to reduced

competition from native grasses early in establishment.

In another angle on the grass competition problem,

Hutchinson (1970), working with sheep and Trifolium repens

in a subtropical setting, found that the legumes did poorly

not only under heavy stocking rates but under light pressure

as well due to heavy competition from grasses. A medium

rate seemed most effective in maintaining persistence.

Davidson and Brown (1985) conducted a grazing study in

which a pasture of Panicum maximum, Neonotonia wightii and

Desmodium intortum was deliberately overgrazed until the











legume component was only 3%. Pasture rest, reduced

stocking rate (1 head ha'1) and reduced stocking rate plus

phosphate fertilization all resulted in legume recovery to

over 50% of the dry matter component after 2 yr. A third

treatment in which the original heavy stocking rate was

maintained (2 head ha 1) showed no recovery over a 2-yr

period. Milk yield and weight change of the grazing animals

were positively correlated with the status of the legumes in

the respective plots.

Other researchers have found that stocking rates do not

seem to influence persistence in some species. Rika et al.

(1981) varied stocking rates between 2.7 and 6.3 animals

ha'1 with various legume-grass mixtures and concluded that

pasture botanical composition was not related to grazing

pressure.

Santillan (1983) found that grazing durations varying

from 1 to 28 d on a pasture of Centrosema pubescens,

Neonotonia wightii, Panicum maximum and Pennisetum purpureum

likewise had little effect on legume persistence. Unlike

the above study, however, this researcher found that grazing

pressures of 1.6, 3.3, 5.0, 6.6, and 8.3 kg dry matter on

offer/100 kg body weight and rest periods between grazing of

0, 14, 28, 42, and 56 d did have a significant effect.

Especially at combinations of high grazing pressures and

short rest periods the legume percentages tended to decrease

in this relatively high rainfall Central American region.











Alcantara and Abramides (1984) tested five legumes in

grass mixtures and also found that the legumes that did well

at low intensity grazing thrived at high levels as well. In

their case Macroptilium atropurpureum and Neonotonia wiqhtii

seemed most adapted to the particular Brazilian situation

studied. Not surprisingly, Cunha et al. (1984), in the same

region found that the same species with the addition of

Centrosema pubescens did equally well at low, medium and

high grazing intensities utilizing a seasonally adjusted

grazing system. Wilson et al. (1982) did a wide survey of

Aeschynomene falcata (Poir) D.C. in the Australian

subtropics and discovered that it also fit into this

omni-surviving group. Whether under light periodic grazing

or continuous heavy pressure (kept to 5 cm year round) this

species survived and actually spread in all cases except one

waterlogged site. This would indicate that there are some

species so well adapted to the local conditions and direct

grazing that overgrazing to the point of destroying stands

may be difficult. The studies did not state, however,

whether animal gain on these persistent legumes was higher

than on other less tolerant species.

Where grazing sensitive species are used, the resultant

decline in animal output per area should not be surprising.

Watson and Whiteman (1981) subjected Centrosema pubescens,

Macroptilium atropurpureum and Sylosanthes guianensis cv.

Endeavor mixtures in various grasses to 1.8, 2.7, 3.6 and









29

4.5 animals ha-1 over 4 yr. In the pasture with the most

productive grass species, live-weight gain per ha per yr

showed a definite quadratic relation ranging from just under

400 kg at the low pressure to 600 kg at 3.6 animals haI1 and

then back down to 500 kg at the highest stocking rate. Less

productive grasses did not show this relationship as

distinctly. The relationship between animal gain and

percent legume component was also quadratic for all

mixtures.



Put-and-take Management



Some researchers have bypassed the stocking rate

dilemma by using a variation of continuous grazing in which

numbers of cattle theoretically are maintained at the

optimum stocking rate such that the legume component had a

good chance to persist. Under these conditions species'

survival or lack thereof should result from genetic traits

rather than management.

Buller et al. (1970) implemented this system in Brazil

to study Sylosanthes gracilis, and Glycine javanica in

association with Digitaria decumbens (Stent). Year round

grazing resulted in Sylosanthes gracilis disappearance and

Glycine javanica persistence despite good animal acceptance

of both legumes. This might indicate that even at carefully

set stocking rates, those species which continue to grow











year round (even in the dry season as does Sstylosanthes

gracilis) will likely suffer more losses than those which

are dormant part of the year as was Glycine javanica in this

study. Production of rhizomes, stolons or rooting nodes in

trailing legumes, although not documented in this case,

might also give viney species advantages over those which

are completely dependent on seed production for

reproduction. This should be especially true in

continuously grazed systems.

Hodges et al. (1976) studied two annual legumes in

Florida, Aeschynomene americana and Indegofera hirsuta L.,

under a put-and-take system with several different grass

associations. These authors found that productivity of the

two legumes varied widely year to year but that Aeschynomene

americana had a higher potential pasture yield. Both

species were found to be equally productive in animal weight

gain per area when compared to nitrogen fertilized

grass-only pastures when at least 25% legume cover was

obtained (Hodges et al., 1977).

In a variation of the put-and-take management system,

Thomas (1976) subjected paddocks of Desmodium uncinatum,

Macroptilium atropurpureum, Desmodium intortum, Macroptyloma

axillare (E. Mey.) Verdc., Neonotonia wightii, and

Stvlosanthes quianensis cvs. Schofield and Endeavor to

grazing by Malawian fat-tailed sheep to a constant 10-cm

height. The results showed a markedly higher persistence











and productivity by Desmodium uncinatum and Macroptyloma

axillare. The two Stylosanthes cultivars were the most

productive the first year but were out-produced by the

others in subsequent years.

Thomas and Andrade (1984) repeated this general

evaluation scheme using cattle on Brazilian savannah. In

this study only the genus Stylosanthes was studied, using

eight accessions of Stylosanthes quianensis, Stylosanthes

macrocephala Ferr. and Costa and Stylosanthes capitata.

What is noteworthy in this study is that different species

of the same genus and different varieties of the same

species responded to grazing in markedly different ways. In

their particular situation Stylosanthes macrocephala CIAT

1582 and both Stvlosanthes capitata CIAT 1019 and 1097

outproduced the other species and entries after 4 yr grazing

to a constant 10-cm height. In a later trial, Thomas and

Andrade (1986) again found differences within species

(Stylosanthes spp. and Zornia spp.) under both equal and

different grazing pressures.

Pitman et al. (1988) also utilized a put-and-take

system to study 50 legume accessions planted in common

pastures. The authors studied persistence under a heavy

stocking rate defined as 4 to 6 head ha"1 and a light rate

ranging from 1 to 3 head ha1. Perhaps due to a combination

of various other factors including heavy grass competition

at establishment, intermittent winter frosts and summer











flooding, the results showed there were no outstanding

legumes among those studied. Of those that did survive

after 3 yr, Macroptilium atropurpurem had higher persistence

(3.5 %) under the low stocking rate when compared to the

high rate. Desmodium barbatum was the opposite, exhibiting

minimal but higher persistence (1.5 %) under the high as

compared to the low stocking rate.



Deferred Grazing



Davidson and Brown (1985), in a previously mentioned

experiment with dairy cattle on Panicum maximum, Neonotonia

wightii and Desmodium intortum pastures, showed that

deferred grazing at critical times sometimes could result in

overall legume yield increases reflected in higher milk

production over the year and decreased weed problems. By

allowing no grazing during the spring season, critical

winter yields were higher than in those treatments under

continuous use. Jones (1979) likewise found that in

pastures of Macroptilium atropurpureum where productivity,

but not plant density, had declined from overgrazing

(excessive frequency and stocking rate), prolonged rests

were very effective in pasture regeneration. In this study

an entire growing season was allowed for recovery prior to

use again in the autumn. It was noted, however, that









33

pastures in already reasonable condition recovered far more

effectively than those in overgrazed treatments.

Gutteridge (1985b) allowed Stylosanthes spp. and

Macroptilium atropurpureum pastures under a 2.5 to 6.5

animal units ha-1 stocking range to rest during the dry

season not so much for management purposes but to imitate

indigenous grazing systems in Thailand. Four-day grazing

periods and 16-d rests were also used. Although this

experiment unfortunately did not have a year-round grazed

control, the author found that the effects of different

stocking rates were far less distinct after than before each

rest period. Gutteridge (1985a) found that Macroptilium

atropurpureum was the only entry which showed strong

perennation although it, like all the Stylosanthes entries,

tended to spread or survive (mostly at lower stocking rates)

more via seeds than vegetatively. The author surmised that

the seed dependent entries had greater difficulty surviving

under the deferred grazing of the dry season because they

did not have the water extracting root capacity of the

Macroptilium atropurpureum pastures. It would appear, then,

that for shallow-rooted annuals or perennials which act as

annuals in some conditions to benefit from deferment that

rest should occur during late growing seasons when moisture

is still available to those roots.

Annuals particularly seem to benefit from intensive

grazing during some periods and no grazing in others.











Stockwell (1984a) found this to be true in Australia with

Centrosema pascuorum Martimus ex. Benth. cv. Bundey. His

recommendations included heavy grazing during the early

rainy season to limit grass growth and limited grazing from

late wet to early dry season to allow seed set. The same

author (Stockwell, 1984b) from work with another annual

legume, C. pascuorum cv. Cavalcade, recommended slightly

different management for a species to be used primarily

during the dry season. Heavy early grazing to keep grasses

under control during establishment was still recommended but

thereafter use of the pasture was to be deferred until the

dry season when it was most critically needed.

Sollenberger et al. (1987a) studied the effect of early

season deferment on the annual legume Aeschynomene americana

in Hemarthria altissima cv. Floralta pastures. These

researchers found that grazing the grass early in the spring

until the legume seedlings had reached at least the two-leaf

stage and then withholding grazing until they were at least

60 cm tall gave the highest dry matter production. The

authors pointed out, however, that when grazing was

initiated in the 20- to 40-cm height cattle seemed to be

able to utilize the more uniform and less lignified plants

more efficiently. This was illustrated by decreased stem

quality indicators digestibilityy and nitrogen content) and

leaf/stem ratio as grazing initiation was delayed

(Sollenberger et al., 1987b).









35

In contrast to annuals, most perennials establish more

effectively with deferred grazing during the early stages.

Andrews and Comudom (1979) found that subjecting legumes

such as Desmodium intortum and Trifolium repens to light

pressure gave far better establishment. They found that the

perennial, Stylosanthes guianensis, in particular suffered

if grazed heavily during early establishment. The

recommendation to graze annuals heavily is more likely to

assist establishment where faster-growing sod grasses are

stronger and less so where pure legume stands or bunch

grasses are present.















CHAPTER THREE

METHODS AND MATERIALS



Field Study



The site for the experiment was a deteriorated 2-ha

pasture at the Ona Agricultural Research and Education

Center (Ona AREC). Vegetative cover within the pasture was

highly variable. Portions contained primarily native range

vegetation including such grass species as low panicums

(Panicum spp.), creeping bluestem (Schizachyrium

stoloniferum Nash.), and broomsedge bluestem (Andropogon

virginicus L.), while others were dominated by vasey grass

(Paspalum urvillei Steud.) and common bermudagrass (Cynodon

dactylon (L.) Pers.).

The soil was Immokalee fine sand (sandy, siliceous,

hyperthermic Arenic Haplaquod) with a composite pH of 5.6,

and nutrient elements at the following levels (mg kg'1):

phosphorus 4.1, calcium 655, potassium 8, copper 0.54, iron

10.3, magnesium 203, manganese 1.0 and zinc 1.1.

The pasture was chopped in the fall of 1986 with a

Marden rolling chopper and sprayed in April 1987 with 2-4-D

(2, 4-dichlorophenoxyacetic acid, butoxyethyl ester)


I











selective herbicide to control broadleaf weeds in the

pasture. Individual blocks were rotovated to a 30-cm depth

in May 1986 and planting took place throughout the June,

July and August period.



Experimental Layout



The overall experimental design was a randomized

complete block design with six blocks. Treatments were

arranged in a strip-plot as described by Gomez and Gomez

(1984). Legume entries were assigned as north-south strips.

East-west strips were made up of the three grazing

treatments of ungrazed, grazed from May through December and

grazed only during late spring and summer (fall deferment).

Common grazing was used for the entire experiment with

cattle excluded from the grazing treatments at the

appropriate seasons.

Legume main plots (north-south strips) measured 7.0 by

15.0 m consisting of five plant rows each with 84 individual

plants spaced 30 cm apart. Each row was separated by 1.0 m

and an additional 2.0 m was inserted between plots. Grazing

treatments (east-west) measured 5.0 by 49.0 m. Each grazing

treatment was separated from the others when appropriate by

a five-strand barbed-wire fence.

The legumes evaluated were: Aeschvnomene americana L.,

Alysicarpus vaqinalis D.C., Desmanthus virgatus (L.),


L











Willd., Desmodium heterocarpon (L.) D.C. cv. Florida,

Galactia elliottii Nuttal., Macroptilium lathyroides (L.)

Urb. and Vigna adenantha (G. F. Meyer) Marechal, Mascherpa,

and Stainier.

Annuals and short-lived perennials, for which unlimited

seed was available, Aeschynomene americana and Macroptilium

lathyroides, were broadcast throughout the individual plots

on 13 April 1987. Seeding rate was 10 kg ha 1, 5 kg haI1 of

which was unhulled seed for the Aeschynomene americana. For

all other species except Desmodium heterocarpon, seed was

limited. These were therefore initially planted during the

summer months of 1986 in peat cups or directly transplanted

from native stands to 30-cm spacings. All peat cups were

inoculated with cowpeaa" type Rhizobium at seeding to avoid

a disadvantage in nodule formation compared to transplanted

native species.

Heavy rains in May followed by a dry period in June

forced replanting of many individual plants in 1986. Those

seeded directly were especially affected by waterlogging in

early summer. Aeschynonome americana and Macroptilium

lathyroides suffered complete establishment failure. These

were therefore reseeded in April, 1987 at the original rates

after light discing of the specific plots.

Any plants which died from among the other species were

replaced during 1986 up through August. Due to

unavailability of seed, Galactia elliottii plots were not











completely filled in with plants in peat cups. The G.

elliottii plots were subsequently completed in the fall by

transplanting plants from a nearby range site.

Deer (Odocoileus virginianus seminolus Goldman and

Kellogg) and rabbit (Sylvilaqus spp.) consumption of the

legumes, especially Aeschynomene americana, Macroptilium

lathyroides and Desmanthus virqatus, was a problem. Rabbit

fencing was placed around the latter plots but no effort was

made to exclude deer. Instead, a large area of Macroptilium

lathyroides and Aeschvnomene americana was planted in a

neighboring pasture to divert the deer in the summer of

1987.



Grazing Management



On 22 May 1987 eight crossbred yearling heifers were

placed in the pasture. On 28 May, 4 head (2.2 head ha"1

remaining) of these were removed after the initial excessive

herbage growth had been reduced. On 15 September animal

numbers were reduced to 2 head (1.2 head ha 1) on the s/s/f

treatment strips due to forage reduction. These last 2 head

were taken off the pasture on 1 December when cold and frost

effectively stopped forage regrowth.

Cattle were excluded from the zero graze treatments by

a permanent five-strand barbed-wire fence. Cattle were left

on the pasture in a continuous grazing system, but were











excluded from the spring/summer-only grazing strips on 15

September 1987 by barbed-wire fences.

In 1988, the wire around the fall-deferment strip was

removed and two yearling steers were added to the pasture in

May. The wire was put up once again on 15 September 1988

and cattle removed on 1 December.



Persistence Evaluation



Established plant populations were determined on 21 May

1987 by counting surviving plants before cattle were added

to the study on 22 May 1987. The numbers gathered in each

subplot at this date were then used in computing persistence

percent at subsequent dates.

Plant numbers were taken on December 1987, May 1988,

December 1988, and May 1989. Persistence was calculated as

plant counts on specific date / plant count of May 1987 *

100. This allowed for determination of a population change

after May 1987. An increase in population would register as

over 100% persistence.

In subsequent months of 1987, heights of individual

plants of the center rows in each subplot were determined.

This was discontinued during the winter months when little

or no growth occurred.











During the 1988 growing season, species numbers were

determined only in May and December. In May 1989 the final

count was taken on the inside three rows.



Statistical Analysis



The statistical analysis included an analysis of

variance (AOV) of percent persistence as well as average

plant heights recorded within species at different grazing

pressures.

Since annuals and perennials had distinctive growth,

seeding and regrowth habits, these were analyzed in separate

groups. Only the perennials, with the exception of the

viney Viqna adenantha, were measured for effect of grazing

on individual plant heights during 1987. The annuals were

not sufficiently established for data collection at this

time.



Pot Study



A pot study to observe species physiological responses

to varying defoliation stresses was conducted. In order to

parallel the field trial, this experiment was conducted

during late fall, winter and early spring of 1987-1988. The

entries were subjected to clipping stress just prior to the

normal dormant period. By observing the regrowth potential











and biochemical composition of the plants during and after

clipping, it was hoped that factors might be found to

explain field trial results. Regeneration after the short

cold days of January, February and March was also observed

to determine the effect pre-winter clipping stress had on

post-winter regrowth.

Species were selected based on their initial field

establishment success and seed availability. Of those that

showed promise, three were selected for their upright or

climbing growth habits, conducive to height-related clipping

regimes. The species employed were Desmanthus virgatus,

Galactia elliottii, and Desmodium heterocarpon.

Germinated seeds of these species were placed in peat

cups and seedlings were allowed to establish. Plants of

uniform size were then inoculated with cowpea inoculum and

transplanted into a pot containing 1 kg of unfertilized

Immokalee fine sand (sandy, silicious, hyperthermic Arenic

Haplaquod). Soil used was taken from the field trial

pasture and consisted of the top 20 cm sifted through a 1-cm

screen. No amendments were added and the plants were

watered from above whenever necessary to keep the soil moist

throughout. All pots were allowed an adaptation period of 6

wk during which any dead or weak seedlings were replaced.

The experiment was conducted on tables with opaque

fiberglass roofing for protection from direct precipitation.









43

The experimental design was a randomized complete block

with four replications. Three clipping treatments were

imposed during the autumn period. These will be referred to

as 'autumn clipping treatments.' The autumn clipping

treatments were imposed every 2 wk for periods of 0, 6, and

12 wk beginning on 15 October 1987. Thus, treatments

consisted of an unclipped control, three clippings during

the initial 6 wk (early clipping), and six clippings over a

12 wk period (extended clipping). Initial clipping heights

in the early and extended clipping treatments were set at

50% of the blocks' average height for each species using the

tallest or longest point as reference. Subsequent clippings

were made at that same height. At each clipping, material

was separated by leaf, stem and reproductive organs, dried

at 720C for 48 h, weighed and composite with other

clippings from that same pot. Each experimental unit

consisted of 16 pots.

During 8 to 13 January 1988, plants in half of the pots

in each experimental unit were sacrificed. Herbage in these

pots was clipped at a 3-cm height above the soil surface,

separated into leaf, stem and flowers/pods, dried, weighed

and composite with previous clippings where appropriate.

The remaining plant portions, consisting of roots and stem

bases, were washed free of soil, dried at 720C for 48 h and

prepared for total non-structural carbohydrate (TNC)

analysis.











Of the remaining eight pots per experimental unit, a

strip plot arrangement of treatments was imposed with four

pots in each experimental unit harvested to a 3-cm height

during 8 to 13 January 1988. This was done to represent the

normal above-ground herbage destruction which occurred in

the field due to frosts and freezes. The four remaining

pots in each experimental unit were not subjected to a

winter harvest. After 16 wk (May 1988) all plants were

harvested to a 3-cm stubble height. Both fractions, above-

ground herbage and roots, were recovered, dried, and

weighed.

Herbage from the autumn clipping treatments and the

winter harvest from four pots within each experimental unit

was composite and ground through a Wiley mill equipped with

a 1-mm screen to provide sufficient material to analyze

crude protein. Roots were likewise weighed, ground and

composite except that only roots of two pots were used per

sample for TNC analysis. Above-ground herbage from the

spring harvest was treated in the same manner except that

material from only two pots was composite to form a

laboratory wet chemistry sample. See Table 1 for a

breakdown on experimental unit subsamples.

Nitrogen content was determined by an auto-analyzer

method employing a modified aluminum block digestion

procedure described by Gallaher et al. (1975). Sample














Table 1. Number of subsamples per experimental unit in
the pot study.

Treatment Variable Subsample number


Autumn clipping

Winter harvest
Winter harvest
Winter harvest
Winter harvest


Spring harvest
Spring harvest
Spring harvest


Herbage mass

Herbage mass
Herbage nitrogen
Root mass
Root total non-
structural carbohydrate

Herbage mass
Root mass
Herbage nitrogen


weight was 0.25 g, catalyst used was 3.2 g of 9:1

K2SO4:CuSO4 and 2 ml H202. Ammonia in the digestate was

determined by semiautomated colorimetry (Hambleton, 1977).

Roots from the winter harvest were analyzed for TNC

following a modified enzymatic extraction procedure adapted

from Smith (1981). Reducing sugars were analyzed with

Nelson's (1944) colorimetric approach to the copper

reduction method first described by Somogyi (1945).















CHAPTER FOUR

RESULTS


Field Study



Height of Perennials, 1987



Inter-species height differences were not compared

since species morphologies differed and responses of

individual species to grazing were the primary interest. Of

the perennials in this study, Desmanthus virgatus was the

only upright species, Alysicarpus vaqinalis and Desmodium

heterocarpon were normally prostrate while Galactia

elliottii displayed upright growth in early stages and a

viney habit latter in maturity. Vigna adenantha displayed

essentially only viney growth.

Analysis of intra-species height differences are shown

by date for the 1987 grazing season in Table 2. No

differences (P=0.36) existed between grazing treatments for

any species before cattle were added on 22 May.

After 38 grazing days (30 June) there was a distinct

difference (P=0.01) between the zero grazing and the two












Table 2. Plant height means of Alysicarpus vaqinalis (AV),
Desmodium heterocarpon (DH), Galactia elliottii
(GE) and Desmanthus virgatus (DV) under zero (Z),
spring/summer (S/S) and spring/summer/fall (S/S/F)
grazing treatments during the 1987 growing season.


Date Grazing AV DH DV GE
treatment


----------------cm----------------
21 May Z 10.2at 15.6a 23.3a 16.3a
S/S 10.9a 16.7a 22.6a 19.1a
S/S/F 11.7a 16.2a 21.5a 20.4a

30 June Z 13.2a 18.7a 24.3a 20.9a
S/S 10.3b 12.3b 18.8b 12.3b
S/S/F 9.6b 11.0b 16.9b 12.6b

29 July Z 20.la 24.5a 29.1a 25.3a
S/S 9.6b 11.0b 18.4b 13.4b
S/S/F 10.7b 12.2b 17.2b 10.8b

1 Sept Z 24.3a 27.9a 31.5a 25.8a
S/S 1l.lb 10.5c 16.4b 12.9b
S/S/F 11.0b 13.4b 14.6b 10.9b

1 Oct Z 26.3a 31.2a 31.5a 23.8a
S/S 13.6b 10.6c 18.6b 11.9b
S/S/F 10.1c 13.5b 15.1b 1l.lb

3 Nov Z 29.0a 28.4a 30.7a 19.4a
S/S 17.2b 11.4b 19.3b 11.5b
S/S/F 10.1c 10.5b 14.6c 14.8ab

15 Dec Z 27.2a 27.5a 28.3a 19.9a
S/S 16.6b 12.1b 18.5b 15.3ab
S/S/F 9.0c 9.1c 14.5b 9.3b


tMeans at each date within columns differ (P<0.05) if
not followed by a common letter according to Duncan's
Multiple Range Test.











grazed treatments for all four species measured. This

difference persisted throughout the grazing season except in

the case of Galactia elliottii on 15 December (P=0.14).

During September (P=0.0002) and October (P=0.0006) an

unexplained difference between the spring/summer (s/s) and

spring/summer/fall (s/s/f) grazing treatments in Desmodium

heterocarpon appeared. By November, due perhaps to

differing grazing treatment, this unexplained difference

disappeared.

The removal of grazing animals on 15 September from the

fall-deferred treatment produced differences (P=0.005)

within 15 d between the s/s and s/s/f treatments in

Alysicarpus vaginalis. This difference became more and more

pronounced as fall progressed (Fig. 1).

Fall deferment from grazing of Desmodium heterocarpon

took a little longer to affect plant heights but became

apparent by December (P=0.0001). This delayed effect of

deferment may have been in part due to the unexplained

differences between the s/s and s/s/f treatments which

existed prior to animal removal (Fig. 2). Within two months

after grazing deferment the relative order in height between

the s/s grazing treatment and the s/s/f grazing treatment

had been reversed.

Desmanthus virgatus height reacted in much the same

manner as Desmodium heterocarpon except that at the December










49





40

K Z o S/S A S/S/F


E
30
I30
CD
LU
I

z 20 -
S -----------o



10. L -- A ... .. .. ..- -





0I
0 50 100 150 200 250
DAYS INTO GRAZING SEASON



Fig. 1. Effect of zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing on height
of Alysicarpus vaqinalis (A.V.) beginning 20 May
1987 with cattle added day 0 and taken off S/S
day 115.
















40

Z o S/S AS/S/F


E
S -30










0-1- --- --------
I
LU

z 20
-J



10 -A




0 --I I
0 50 100 150 200 250
OAYS INTO GRAZING SEASON



Fig. 2. Effect of zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing on height of
Desmodium heterocarpon (D.H.) beginning 20 May
1987 with cattle added on day 0 and excluded
from S/S on day 115.









51

reading the differences (P=0.0007) apparent between s/s and

s/s/f which appeared in November (Fig. 3) became

statistically non-significant. This was perhaps due to

heavy deer predation since the cattle-excluding fences were

not a deterrent for these browsing animals. It was not due

to senescence since plants in these plots continued to

generate new growth until killed back by frost.

Galactia elliottii, despite an unexplained increase in

s/s/f plant height in November, showed perhaps the most

interesting trend by December. Although there was no

difference (P=0.14) between the s/s and the s/s/f

treatments, there also was no difference between the s/s and

zero grazed plots. This indicated that G. elliottii either

benefited sufficiently from the fall rest to catch up with

the zero treatment or the zero treatment senesced sooner due

to fewer recently produced leaves. It was noted to shed

well over 50% of its leaves in the range during the cold

months. This became apparent when the 33% height increase

for s/s was compared with only 3% increase for the zero

graze treatment during the 40-d period in which neither was

grazed (see also Fig. 4 and the near-steady height decrease

throughout the fall period illustrated by the decline in the

zero graze line from day 100 on).

Height measurements indicated, at least in the second

growing year and first grazing season, that these four
















40

SZ o S/S A S/S/F



,- 30 *
H-
CDLU
H-i

z 20- 0-
S---- ..o------------

10
A-a


10




0 --- I
0 50 100 150 200 250
DRYS INTO GPAZING EEASCirJ



Fig. 3. Effect of zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing on height
of Desmanthus virgatus (D.V.) beginning 20 May
1987 with cattle added day 0 and excluded from
S/S day 115.

















40

Z o S/S A S/S/F


E
30


LI


20




10 -
H-






2 :II

LU .

0-
0 -- A-0






0 50 100 150 200 250
ORYS INTO GRAZING SEASON



Fig. 4. Effect of zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing on height
of Galactia elliottii (G.E.) beginning 20 May
1987 with cattle added day 0 and excluded from
S/S day 115.









54

forage legumes were grazed by the yearling heifers. Figures

1 through 4 illustrate graphically, however, that plant

herbage growth did respond to fall deferment from grazing

except in the case of Desmanthus virgatus. In this entry,

heights followed this general trend but the s/s and s/s/f

means were not different (P=0.14) from each other. All

others, particularly the more decumbent Alysicarpus

vaqinalis, showed a positive response to the pre-winter rest

by addition of new foliage.

In Fig. 5, plant heights of both s/s and s/s/f

treatments as a percent of the ungrazed treatment are shown

for all perennials measured in December (Table 3).

Alysicarpus vaginalis suffered the least in the s/s

treatment (P=0.11), recovering to 70% of the zero treatment

height. Desmanthus virgatus at 66% was not different from

either Alysicarpus vaginalis or Desmodium heterocarpon.

This last species, which registered 53%, was not

statistically different from the most shortened entry,

Galactia elliottii at 49%.

Figure 5 also illustrates the uniform grazing

defoliation which occurred among most perennial species in

the s/s/f grazing (P=0.47). Only Desmanthus virgatus, at

50%, differed from the others which ranged between 34 and

36% of the ungrazed treatment. Woody stem development early

in establishment of D. virgatus may have limited the degree

of defoliation of this upright growing legume.


~
















100



75
N
Z

50



25

0



AV

GE


S/S S/S/F
GRAZING TREATMENT




Fig. 5. Mean percent height remaining in spring/summer
(S/S) and spring/summer/fall (S/S/F) grazed
plots of Alysicarpus vaqinalis (A.V.),
Desmodium heterocarpon (D.H.), Desmanthus
virgatus (D.V.) and Galactia elliottii (G.E.)
compared to ungrazed plots on December 1987.











Table 3.


Height of Alysicarpus vaginalis (AV),
Desmodium heterocarpon (DH), Galactia
elliottii (GE) and Desmanthus virgatus (DV)
under spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing treatments
in December 1987 as a percent of ungrazed
plots.


Grazing AV DH DV GE
regime


-----------------%-----------------

Z 100at A 100a A 100a A 100a A
S/S 70a B 53bc B 66ab B 49c B
S/S/F 35b C 36b C 50a C 34b B

tMeans within lines differ (P<0.05) if not followed by
a common lower case letter according to Duncan's
Multiple Range Test.

Means within columns differ (P<0.05) if not followed
by a common upper case letter according to Duncan's
Multiple Range Test.











Persistence Within Perennials. 1987-1989



Since the interactions between species and grazing

treatment existed for all dates (P<0.001), the discussion of

persistence of both within and among perennials, as well as

within and among the annually reseeding group will be

limited to the simple effects.

After one grazing season, Alysicarpus vaginalis

appeared to have suffered few losses under grazing and

increased substantially where protected completely or in the

fall (Table 4). There was an apparent difference (P=0.09)

between s/s/f and zero grazing, however. The increase to

214% in the zero grazing treatment, (Fig. 6), was thought to

be at least in part due to artificially low plant counts at

the base date in 1987. This species was noted to regrow

slowly from frosted plants early in the growing season.

Winter stress affected population dynamics considerably

in this species. By May 1988, the zero graze treatment was

lower (P=0.09) than the s/s treatment as seen in Fig. 6.

This may have been due to A. vaqinalis's failure to either

store sufficient nutrients in roots when forced to compete

with ungrazed grasses or due to an enhanced susceptibility

to frost and freeze when forced to grow upright in heavy

competition.

By the end of the second grazing season, December 1988,

the same basic trends held except that the s/s plots had











Table 4.


Mean percent survival comparison within and among
perennials Alysicarpus vaqinalis (AV), Desmodium
heterocarpon (DH), Galactia elliottii (GE),
Desmanthus virqatus (DV), and Vigna adenantha (VA)
under zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing treatments
using May 1987 as a base date.


Date Grazing AV DH DV GE VA
treatment


-------------------%------------------------
Dec 1987 Z 214at A 86a B 68a BC 38a C 101a B
S/S 153ab A 85a B 71a B 47a B 53b B
S/S/F 93b A 95a A 54a B 12b C 13c C

May 1988 Z 47b B 88a A 55a B 86a A 110a A
S/S 109a A 91a AB 36b D 51b CD 71b BC
S/S/F 89ab A 98a A 25b C 51b B 20c C

Dec 1988 Z 57b B 66a B 62a B 32a C 109a A
S/S 155a A 61a B 25b B 20b B 42b B
S/S/F 75b A 55a B 10b C 4c C 5c C

May 1989 Z 2a C 10a C 47a B 136a A 63a B
S/S 12a B 12a B 26ab B 52b A 24b B
S/S/F 8a B 18a B 9b B 55b A Ic B

tMeans within columns at each date differ (P<0.05) if not
followed by a common lower case letter according to Duncan's
Multiple Range Test.

Means within lines at each date differ (P<0.05) if not
followed by a common upper case letter according to Duncan's
Multiple Range Test.


























SLUU
z \12/88
LU

0n -. -





\
00 ----------- -----------------








0 10 20
MONTHS GRRZEO



Fig. 6. Mean percent persistence of Alysicarpus
vaginalis starting 22 May 1987 under zero
(Z), spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments.











higher persistence than the s/s/f treatment. During the

winter of 1988-1989 there was a late frost on 26 February

followed by an unseasonably dry spring with only 38% of the

normal rainfall at Ona AREC. This latter climatic stress

more than any may have caused the dramatic decline in A.

vaqinalis plant population throughout the experiment. Both

regenerating plants and new seedlings were destroyed. As a

result, this species was essentially eliminated so that the

differences between treatments seen earlier no longer held

(P=0.46).

Desmodium heterocarpon, on the other hand, showed less

effect (P>0.65 for all dates) of grazing regime on

individual plant survival (Table 4). This held true

throughout the experiment as can be seen in the lack of

differentiation between treatments in Fig. 7. It, along

with Alysicarpus vaqinalis, seems to have suffered the most

from the late winter frost and early spring drought of 1989.

Desmanthus virgatus showed a similar lack of response

(P=0.33) to grazing at the end of 1987 (Table 4). Following

both the winter die-back and another grazing season,

however, the differences between the two grazed treatments

and the ungrazed treatment became more apparent (P=0.004) as

illustrated in Fig. 8. During the two readings in 1988 the

s/s and s/s/f treatments were not different (P>0.05)

although there was a trend for greater survival in the s/s

treatment.

























LU 10-
C-)-
S100---.. 12/88
z ----- k--.

U)


0 .
A\
50 -,,




N '



0 10 20
MONTHS GRRZEO



Fig. 7. Mean percent persistence of Desmodium
heterocarpon starting 22 May 1987 under zero
(Z), spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments.























LU 100
S12/88





50 -
-n -- I











MONTHS Rt-ED



Fig. 8. Mean percent persistence of Desmanthus virgatus
starting 22 May 1987 under zero (Z),
spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments.











After the stressful winter and early spring of 1989,

Desmanthus virgatus continued to show a treatment effect

(P=0.01). The notable exception was a lack of difference

(P>0.05) between the s/s and the ungrazed treatments.

Perhaps due to condition improvement from fall rest, plants

in the s/s grazing regime were able to maintain their vigor

as well as the plants in the ungrazed treatments.

There was a noticeable response (P=0.01) to fall

grazing by Galactia elliottii at the end of 1987 (Table 4

and Fig. 9). Although plant survival was not high for the

zero and s/s treatments at this time, the 12% survival for

the s/s/f plots was especially low. No differences (P>0.05)

existed between the zero and s/s treatments at this time

indicating a response on the part of this native legume to

fall deferment from grazing in the first year.

Winter stress showed some interesting results for this

species. In May 1988, many individual plants which were not

visible in December resprouted. This resulted in a 126%

increase at the zero graze level, a 9% increase with the

fall deferred treatment and a 325% increase for the s/s/f

treatment. No seedlings were observed. These results

indicate that the early spring growth of this species

(earlier than most grasses) might substitute for fall

deferment. Differences (P=0.03) between grazing treatments

still existed.



























LU

n)
LU


10 20
MONTHS GRAZED


Fig. 9. Mean percent persistence of Galactia elliottii
starting 22 May 1987 under zero (Z),
spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments.











By the end of a second grazing season, 1988, both

grazed plots were inferior to the ungrazed treatment. The

s/s, however, was more persistent (P=0.004) than the s/s/f,

indicating, again, a benefit from fall deferment. This

benefit did not allow the numbers to maintain the 50%

recovery measured at the end of 1987.

Despite the harsh winter and early spring of 1989, G.

elliottii showed the same recovery after the cold months as

it did in 1988. As happened a year before, the difference

that existed between the two grazing treatment populations

before winter disappeared in the spring although differences

between these and the control were still apparent (P=0.001).

In fact, all three groups showed an increase in numbers over

a year previous with the ungrazed population propagating

itself to 36% over the original plant number in May 1987.

Vigna adenantha showed the most consistent response to

fall deferment, showing higher persistence than the s/s/f

treatment and lower than the ungrazed control for all dates

(P=0.001 for all dates). At every date measured except May

1989 (Table 4), the zero treatment also showed over 100%

persistence, completely covering each plot and invading

adjacent borders. This species also showed that during the

cold months it could regenerate from completely denuded

tops. During the 1987-88 winter, the s/s treatment

especially showed improvement with an increase of 34%

(illustrated in Fig. 10). As in the case of Galactia









66





150

O Z o S/S A S/S/F

12/87


1 00
12/88
z


0 LU 0
cn -


\





_--A-._



0 10 20
MONTHS GRAZED



Fig. 10. Mean percent persistence of Vigna adenantha
starting 22 May 1987 under zero (Z),
spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments.











elliottii, this recovery may have been due to stored non-

structural carbohydrate reserves which were used to put out

new growth before heavy grass competition and grazing

occurred in early spring.

By May 1989, the late frost and early spring drought

which severely affected several of the other species also

combined to reduce Vigna adenantha populations in all

grazing treatments. The s/s/f population, already small in

December 1988, essentially disappeared. The s/s treatment

had only 24% of its original population remaining, lower

than the ungrazed treatment. With the onset of summer

rains, however, both the s/s and the ungrazed treatments

were expected to regain a considerable amount of their

original vigor.



Persistence Among Perennials, 1987-1989



Although of limited interest to grazed pastures, it is

interesting to note that by the end of the 1987 grazing

season Alysicarpus vaginalis and Viqna adenantha populations

increased while the others decreased in the ungrazed

treatments (refer to Table 4 for actual figures as well as

to Fig. 11 for December 1987 and Fig. 12 for December 1988

during the discussion in this section). Desmanthus virgatus

and Galactia elliottii especially showed considerable

















300







100








/ / -




Z S/S S/S/f
GRAZING TRERTTENT





Fig. 11. Mean persistence of perennials Alysicarpus
vaginalis (AV), Desmodium heterocarpon (DH),
Desmanthus virgatus (DV), Galactia elliottii
(GE) and Viqna adenantha (VA) under zero (Z),
spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments on December 1987
using May 1987 as a base date.













































Fig. 12.


100
LU


ix 150 N



0

iIr






S/S S/S/f
GRAZING TREATMENT




Mean persistence of perennials Alysicarpus
vaqinalis (AV), Desmodium heterocarpon (DH),
Desmanthus virqatus (DV), Galactia elliottii
(GE) and Vina adenantha (VA) under zero (Z),
spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments on December 1988
using May 1987 as a base date.









70

decline the first grazing season under nothing more than the

local climatic and edaphic conditions, native herbivores as

well as grass competition. Desmanthus virgatus, however,

was not different (P=0.01) from the other perennials with

the exception of Alysicarpus vaqinalis.

In the s/s treatment at this first date, the decumbent

A. vaginalis showed a greater ability to persist and

increase than any of the others (P=0.0007). Again, this

might be an artificial increase due to its late sprouting

and therefore low numbers at the base 1987 spring date.

Desmodium heterocarpon, equally capable of exhibiting

prostrate growth habits, had the next highest persistence

although it and the remaining three perennials were not

different (P>0.05) from each other. Not surprisingly, the

two decumbent species were the most persistent in 1987 under

the s/s/f grazing, with both showing values in the nineties.

Desmanthus virgatus was lower (P>0.05) at 54% but also

higher (P<0.05) than the two remaining perennials, Galactia

elliottii and Vigna adenantha. It is important to note that

these last two, the species which suffered most under

continuous growing-season grazing, were both viney climbers.

The five-month rest from grazing during the winter

months, a period when little grass grows in Florida

pastures, changed the picture considerably in the zero

grazing treatment (P=0.04 for that date). By May 1988,

Desmodium heterocarpon as well as the early spring growing











Galactia elliottii and Vigna adenantha showed high

persistence values of over 85% while the remaining

Alysicarpus vaginalis and Desmanthus virgatus both exhibited

an inferior presence.

After the winter period and before cattle were added to

the trial in the second grazing year, the s/s plots remained

essentially the same in terms of relative persistence. The

two major exceptions were a considerable decrease in D.

virgatus, making it inferior to all other species, and an

equal degree of increase in Vigna adenantha.

Alysicarpus vaginalis, Desmodium heterocarpon and Vigna

adenantha populations remained constant relative to each

other in the post-winter, May 1988 s/s/f treatment

(P=0.001). Desmanthus virgatus, 25%, and Galactia

elliottii, 51%, switched positions on the relative

persistence scale with the first species decreasing

considerably to make it numerically indistinguishable

(P>0.05) from Vigna adenantha's 20% and the latter's 51%

making it less (P<0.05) persistent than the decumbent

Alysicarpus vaqinalis and Desmodium heterocarpon.

At the December 1988 reading, following three year's

growth and two seasons of grazing, Alysicarpus vaqinalis

lost the most number of plants relative to the readings one

year earlier in the ungrazed treatment. Only 27% of the

plants survived. This put Vigna adenantha's 110%

persistence higher than all the rest at the ungrazed









72

treatment level and Galactia elliottii's 32% lower (P=0.005)

than any of the other perennials.

Although there was a general decrease in numbers for

all species except Alysicarpus vaqinalis at this date, the

general picture remained the same after the second year of

s/s grazing (P=0.02). Overall, the average persistence

showed a steady decline from an average 82% persistence at

the end of 1987 down to an average 61% survival after 1988.

In the s/s/f treatment, Galactia elliottii and Vigna

adenantha continued to decline and were joined by Desmanthus

virgatus at the bottom of the scale on December 1988.

Desmodium heterocarpon also declined but at 55% was higher

than the above three species. The other species with a

tendency for decumbent growth, Alysicarpus vaqinalis, had

the highest (P=0.001) survival at 75% although that too was

inferior to its December 1987 showing.

The late frost and early spring drought prior to the

May 1989 reading changed the picture considerably in the

ungrazed control. Alysicarpus vaqinalis and Desmodium

heterocarpon suffered further population reductions and the

first species essentially disappeared. Unless these two

were able in the subsequent months to recover dramatically

from hidden crowns or seed reserves, this would indicate

that these species were unable to survive without

defoliation of competing grasses.











At the May 1989 date Desmanthus virqatus and Vigna

adenantha under no grazing showed a higher (P=0.0001)

persistence than the two species discussed in the preceding

paragraph. In the case of the V. adenantha especially,

effects of the frosts and drought were apparent. Only

Galactia elliottii managed to exceed its original May 1987

numbers to show that, as a native, it is adapted to local

conditions and periodic stresses once well established.

While other broadleaf species and grasses displayed visible

signs of drought stress, this viney species actually

produced new shoots and covered its wilted neighbors.

In the s/s treatment G. elliottii again topped the list

at 52% survival, over twice the persistence of any other

species (P=0.01). Among the remaining perennials, no

differences (P>0.05) in population persistence existed under

the fall deferment.

Under the s/s/f treatment, the native G. elliottii

again showed a strong regeneration from roots that were

essentially denuded of all top-growth by grazing the

previous fall. This became readily apparent when the 4%

survival of December 1988 was compared to the greatly

improved 55% following five months of cold, frosts and low

precipitation. Individual plants in this treatment were

considerably less vigorous than those of the two other

treatments, however. None of the other perennial species'

population matched this recovery rate in either numbers or









74

vigor. Desmodium heterocarpon, at 18%, was the closest but

was not different (P>0.05) from the other three perennials.



Persistence Within Annuals, 1987-1989



Although Macroptilium lathyroides is a weak perennial,

stands in Florida survive from year to year primarily on the

basis of abundant seed production and subsequent

germination. For the purposes of this discussion, then, it

will be included in the same group as Aeschynomene

americana. Since interaction between grazing and species

existed at all dates (P<0.01), simple effects are discussed

below.

Persistence of Macroptilium lathyroides after one

growing season and one grazing season showed distinct

treatment effects (P=0.0002; Table 5). Zero grazing showed

the best cover at 60%. The s/s grazing treatment, at 36%,

was lower than the ungrazed but also higher than the s/s/f

regime which showed only a 14% persistence. Plant counts

made after the winter months were comprised mainly of

perennating plants which persisted despite frost kill of the

upper growth and very young seedlings. Other than a near

50% population loss in the ungrazed treatment, little change

occurred (Fig. 13) over the winter. Again, etiolated and

exposed growth in the grass-choked, ungrazed treatment or











Table 5. Mean percent survival comparison within
and among Macroptilium lathyroides (ML) and
Aeschynomene americana (AA) under zero (Z),
spring/summer (S/S) and spring/summer/fall
(S/S/F) grazing treatments using May 1987 as a
base date.

Date Grazing ML AA
regime


Dec. 1987 Z 60at A 24ab B
S/S 36b A 26a A
S/S/F 14c A 14b A

May 1988 Z 31a A 4b B
S/S 26a A 16a B
S/S/F 13b A 16a A

Dec 1988 Z 32a A Oa B
S/S 27a A 4a B
S/S/F 3b A Oa B

May 1989 Z lla A Oa B
S/S 3b A Oa B
S/S/F Ob A Oa A

tMeans within columns at each date differ (P<0.05) if
not followed by a common lower case letter according to
Duncan's Multiple Range Test.

Means within lines at each date differ (P<0.05) if
not followed by a common upper case letter according
to Duncan's Multiple Range Test.


















0 S/S/F


12/87


12/88


--------------- -
---0-------- 0,^
-- -- ----------------.


o C~^ a -


0-
'0


MONTHS GRAZED


Fig. 13.


Mean percent persistence of Macroptilium
lathyroides starting 22 May 1987 under zero
(Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing treatments.


50


o S/S


I









77

actual colder conditions in the dense herbage may have made

this species more vulnerable to cold temperatures.

Other than a steady decrease in s/s/f plant

persistence, the December 1988 survival values of M.

lathyroides were little changed from the May 1988 readings.

The early and late frosts, as well as the unusually drought

early spring, may have contributed to a continued population

decline by the May 1989 plant count. At that date no plants

survived in the s/s/f treatment and only 3% of the original

population either survived or had been replaced through

reseeding. The ungrazed control, however, maintained an 11%

persistence which was higher (P=0.004) than the two grazed

treatments.

Aeschynomene americana had the highest mean persistence

under the s/s grazing regime at the December 1987 reading

(Table 5 and Fig. 14), although this was not different

(P>0.05) from the zero grazed plots. The s/s/f plots showed

the lowest persistence although they were not different

(P>0.05) from the zero grazed plants. In May 1988, after

the winter stress period, seedlings in the s/s and s/s/f

treatments were not different (P>0.05) from each other and

were both superior (P<0.05) to the zero treatment.

After two grazing seasons, by December 1988, there was

essentially no persistence of A. americana (Fig. 15 shows

this steady decline). Only the s/s treatment had any

survivors at 4%. But this was not different (P=0.29) from


1



















50
40 L


30 0



10

10





ML-12/87 AA-12/87 ML-12/88 AR-12/88
SPECIES-ORTES



Fig. 14. Mean percent persistence of Macroptilium
lathyroides (ML) and Aeschynomene americana
(AA) on December 1987 and December 1988
under zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing treatments
using May 1987 as a base.























1002






50


o1 -----------
0 10 20
MONTHS GRAZED


Fig. 15.


Mean percent persistence of Aeschynomene
americana starting 22 May 1987 under
zero (Z), spring/summer (S/S) and
spring/summer/fall (S/S/F) grazing treatments.









80

the other two treatments which both showed no plants at all.

This situation was not changed by the winter of 1988-1989.

All treatments showed no persistence at this date. It

should be noted, however, that seeds of this species do not

normally germinate and persist in numbers until well into

the wet, hot summer months.



Persistence Among Annuals, 1987-1989



In the ungrazed treatment, Macroptilium lathyroides

exhibited superior persistence (P=0.01) to Aeschvnomene

americana at the December, 1987 reading (Table 5). There

were no differences (P>0.33) in either of the two grazed

treatments for both these species.

Following winter cold stress, the relative survival of

the two species remained the same at the ungrazed level

despite a decrease in the numbers of both entries. In the

s/s strip plots, however, Macroptilium lathyroides showed a

higher (P<0.05) survival and perennation although this trend

did not appear in the s/s/f treatment (E=0.64).

By the end of the second year, December 1988, both

annuals showed a steady decline in population compared to a

year earlier. This decline was not as severe for M.

lathyroides as with Aeschynomene americana (Fig. 14) since

the first exhibited superior (P<0.05) persistence through

reseeding in both grazed treatments.


~











After the hard frosts and low precipitation of the

intervening months, May 1989 plant populations in all

treatments for both species declined except where they were

already zero. Macroptilium lathyroides in the ungrazed

treatment was the only appreciable plant population

remaining with 11% persistence. Caution should be used in

drawing conclusions from the May data since germination and

establishment of these species is typically limited prior to

summer months.



Pot Study



Winter Harvest



Since there was an interaction between species and

treatment in all dependent variables (P<0.02), except

herbage nitrogen mass (P=0.54), the discussions and tables

in this section will be concerned solely with the simple

effects of this experiment.



Root mass



Galactia elliottii and Desmanthus virgatus showed

similar treatment effects on root mass with no difference

(P>0.05) between the early and extended clipping treatments

in either of these two species (Table 6). There was,











Table 6. Mean root mass, root total non-structural
carbohydrate (TNC) percent and root TNC mass
of Galactia elliottii, Desmodium heterocarpon
and Desmanthus virqatus following autumn
clipping treatments.


Species Clipping Root TNC% TNC
treatment mass mass


g % g
G. elliottii Extendedt 2.76b 25.5b 0.73b
Early 2.86b 25.7b 0.72b
Control 5.67a 33.7a 1.92a

D. virgatus Extended 2.36b 23.0ab 0.55b
Early 2.55b 22.1b 0.57b
Control 4.77a 25.3a 1.20a

D. heterocarpon Extended 3.12a 12.la 0.37ab
Early 3.80a 12.0a 0.46a
Control 3.40a 9.8b 0.32b

G. elliottii Extended 2.76a 25.5a 0.73a
D. virgatus 2.36a 23.0b 0.55ab
D. heterocarpon 3.12a 12.1c 0.37b

G. elliottii Early 2.86b 25.7a 0.72a
D. virgatus 2.55b 22.1b 0.57ab
D. heterocarpon 3.80a 12.0c 0.46b

G. elliottii Control 5.67a 33.7a 1.92a
D. virgatus 4.77b 25.3b 1.20b
D. heterocarpon 3.39c 9.8c 0.32c

tExtended clipping was six clippings at 2-wk intervals,


early clipping included only the first three
and the control was never clipped.


clippings


Means within columns for each division differ (E<0.05)
if not followed by a common letter according to
Duncan's Multiple Range Test.











however, a large difference (P<0.05) between these two

treatments and the unclipped control. Desmodium

heterocarpon showed no differences (P=0.14) between

treatments (Fig. 16). This may in part be because this

species was observed to flower, seed and senesce much more

than either of the other two. Desmanthus virgatus was not

observed to senesce at all throughout the experiment but

continued to put out new green growth continuously.

All three species reacted in much the same way to the

extended clipping treatment (P=0.15). In the early clipping

treatment, however, Desmodium heterocarpon had a higher

(P=0.002) root mass than either of the other two species,

perhaps paralleling the earlier senescing which favored

reserves in the roots (Table 6). Galactia elliottii had the

largest root mass in the unclipped treatment, developing a

xylopod (enlarged taproot, Schultze-Kraft and Giacometti,

1979) even in weakly developed plants. Desmanthus virqatus

had the next largest mass, smaller than Galactia elliottii

but also heavier than Desmodium heterocarpon which had the

smallest root system. This last species, in marked contrast

to the other two, exhibited a much more extensive secondary,

fibrous root system.










84










4





2O






U

ER

EX
DV DH GE
SPECIES



Fig. 16. Mean root mass of Galactia elliottii
(GE), Desmodium heterocarpon (DH), and
Desmanthus virgatus (DV) under extended
clipping (EX), early clipping (ER) and
unclipped (U) treatments.











Root total non-structural carbohydrate percent



Galactia elliottii had a higher (P=0.0001) TNC percent

in the unclipped treatment with no differences (P>0.05)

between the two clipped treatments (Table 6). This last

phenomenon may in part be due to the tendency of this

species tended to drop its leaves and stop growth in all

treatments about the time that the clippings in the early

clipping treatment were terminated. If the treatments had

begun a little earlier in the season, differences between

the early and the extended clipping treatments may have

developed.

This was also the case with Desmodium heterocarpon

except that in this species the unclipped treatment was

lower (P=0.002) than the two clipped treatments (Fig. 17).

This difference between the two species may be explained by

the fact that the clipped plants were less dormant at the

winter harvest since they were unable to complete seed

production as well as did the unclipped control. There was,

perhaps, a much larger percentage of dead root component in

the unclipped plants which may have brought the TNC percent

down.

Desmanthus virgatus, being more metabolically active

during these late fall months than the other species, showed

less distinct differences (P=0.09) between the clipped and

the unclipped treatments. The unclipped control had higher














































Fig. 17.


35

30

25

20
z
15
0
0
10

5


U

ER

EX
DH GE
SPECIES



Mean root total non-structural
carbohydrate (TNC) percent of Galactia
elliottii (GE), Desmodium heterocarpon
(DH), and Desmanthus virgatus (DV) under
extended clipping (EX), early clipping
(ER) and unclipped (U) treatments.