Title: Plant responses of an Aeschynomene americana--Hemarthria altissima association to grazing management /
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00099342/00001
 Material Information
Title: Plant responses of an Aeschynomene americana--Hemarthria altissima association to grazing management /
Alternate Title: Hemarthria altissima
Physical Description: xii, 131 leaves : ill. ; 28 cm.
Language: English
Creator: Sollenberger, Lynn Elwood, 1957-
Publication Date: 1985
Copyright Date: 1985
 Subjects
Subject: Aeschynomene americana   ( lcsh )
Grazing -- Florida   ( lcsh )
Forage plants -- Florida   ( lcsh )
Limpograss
Agronomy thesis Ph. D
Dissertations, Academic -- Agronomy -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis (Ph. D.)--University of Florida, 1985.
Bibliography: Bibliography: leaves 122-130.
Statement of Responsibility: by Lynn Elwood Sollenberger.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00099342
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000527238
oclc - 14525156
notis - ACU9327

Downloads

This item has the following downloads:

plantresponsesof00soll ( PDF )


Full Text












PLANT RESPONSES OF AN Aeschynomene americana--
Hemarthria altissima ASSOCIATION TO GRAZING MANAGEMENT















BY

LYNN ELWOOD SOLLENBERGER


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


1985



































To Norman and Lois Sollenberger,

I learned the most
important lessons
from you.
















ACKNOWLEDGMENTS


The author wishes to express his thanks to those individuals who

have made significant contributions to the doctoral program. The help

of Dr. W. R. Ocumpaugh with planning the course of study and designing

the dissertation research was appreciated. It was the author's

pleasure and great privilege to learn from, and interact with, the

late G. O. Mott, who co-chaired the supervisory committee until his

death. Special thanks are extended to Dr. K. H. Quesenberry, chairman

of the committee, for providing stability, insight, and guidance to

the program. For the many helpful discussions, stimulating lectures,

and lunchtime chats (with the exception of those with political

overtones), the author expresses his appreciation to Dr. J. E. Moore.

Both Drs. Quesenberry and Moore provided many hours of OPS and

technical support to the research effort, without which much of what

was done could not have been accomplished. Thanks are due to Dr. W.

G. Blue for serving on the supervisory committee for two years, and to

Drs. O. C. Ruelke, K. A. Albrecht, and G. Kidder for their criticisms,

comments, and suggestions related to the dissertation. The author

also wishes to thank Dr. C. E. Dean and the Agronomy Department for

the opportunity to study at the University of Florida.

The author is grateful for the cooperation and assistance of Sid

Jones and Dean Kelley at the Forage Evaluation Field Laboratory, and

Art Peplow and Ron Kern at the Forage Evaluation Support Laboratory.

Thanks are expressed to David Moon, Andy Schreffler, and Rebecca


iii











Hoffmann for conscientious and exacting assistance in the field,

separation room, and laboratory. The help and cooperation of Patti

Beede with sample handling, clipper maintenance, etc. is acknowledged.

C. M. Payne and Son Seed Company provided the aeschynomene seed used

in the research, and their generosity is appreciated. For her

excellent work in processing this manuscript, thanks are extended to

Mrs. Patricia French. To Eduardo Canudas, much appreciation is due

for help in the laboratory, assistance with the computer, and being a

special friend and fellow forage manager. The author expresses his

thanks to Debbie Cherney for making the Nutrition Lab a friendlier and

more interesting place to work. Some have surely been omitted, but to

Gerardo Morantes, Jaime Velasquez, Chris Deren, Steve Taylor, Kate

Geiger, Judson Valentim, Andrea Givens, Ken Woodard, and Greg Rusland,

the author is grateful for your friendship and wishes you the best in

your careers.

To the author's parents, Norman and Lois Sollenberger, and to

each of his siblings, Allen, Anne, Joyce, and Lane, your interest,

love, and concern have been priceless. To his extended family at

Emmanuel Mennonite Church, your support has been a great

encouragement. Lastly, to his wife Andrea, your smile and cheerful

nature have salvaged many days, your patience with this program has

rarely been exhausted, and your love, interest, and companionship have

made life much richer than it otherwise would have been.









iv

















TABLE OF CONTENTS


PAGE

ACKNOWLEDGMENTS ................................................ iii

LIST OF TABLES..................... .......................... vii

LIST OF FIGURES........................ ............ ............. x

ABSTRACT...................................................... xi

CHAPTER I INTRODUCTION...................................... 1

CHAPTER II LITERATURE REVIEW................................. 4

General Description of Limpograss and Aeschynomene 4
Establishment.................................... 7
Responses to Clipping and Grazing................. 11
Forage Quality.................................... 16
Mineral Status .................................... 23
Fertility Requirements............................ 25
Disease and Pest Problems......................... 27
Effects of Defoliation on Seed Production......... 29

CHAPTER III FACTORS AFFECTING THE ESTABLISHMENT OF AESCHYNOMENE
IN FLORALTA LIMPOGRASS SODS....................... 32

Introduction.............................. ... 32
Materials and Methods............................. 33
Results........................................... 36
Discussion........................................ 46

CHAPTER IV EFFECTS OF GRAZING MANAGEMENT ON PRODUCTIVITY AND
BOTANICAL COMPOSITION OF AN Aeschynomene americana-
Hemarthria altissima ASSOCIATION.................. 51

Introduction...................................... 51
Materials and Methods............................. 52
Results........................................... 56
Discussion....................................... 65












PAGE


CHAPTER V FORAGE QUALITY RESPONSES OF AN Aeschynomene
americana-Hemarthria altissima ASSOCIATION TO
GRAZING MANAGEMENT................................ 74

Introduction..................................... ....... 74
Materials and Methods............................. 76
Results........................................... 81
Discussion....................................... 92

CHAPTER VI SEED PRODUCTION RESPONSES OF Aeschynomene americana
TO GRAZING MANAGEMENT............................. 99

Introduction ............................... ... 99
Materials and Methods............................. 100
Results........................................... 103
Discussion....................................... 107

CHAPTER VII SUMMARY AND CONCLUSIONS........................... 113

Aeschynomene Establishment........................ 113
Productivity and Forage Quality of Aeschynomene,
Limpograss, and Their Association............... 114
Aeschynomene Seed Production...................... 116
Implications of the Research...................... 117

APPENDIX....................................................... 119

LITERATURE CITED................................................ 122

BIOGRAPHICAL SKETCH ............................................ 131

















LIST OF TABLES


TABLE PAGE

3.1 Rainfall data for 1983 and 1984 recorded at the Beef
Research Unit, northeast of Gainesville, Florida....... 37

3.2 Dry matter harvested and botanical composition of
aeschynomene-limpograss associations as affected by
aeschynomene establishment method (1983)............... 39

3.3 The effect of aeschynomene establishment method
on N harvested in total and legume herbage from
aeschynomene-limpograss associations and on
limpograss crude protein (CP) concentration (1983)..... 40

3.4 Soil and climatic conditions for the first 20 days
following seeding of aeschynomene in limpograss sods
(1984).................................................. 42

3.5 Legume seeding date x establishment method
interaction means for total and legume dry matter (DM)
harvested in aeschynomene-limpograss associations
(1984).................................................. 44

3.6 Dry matter and N-harvested responses to legume seeding
rate in aeschynomene-limpograss associations (1984).... 45

4.1 Changes in aeschynomene seedling population as affected
by legume seedling growth stage (LSGS) when early-
season grazing ended................................... 58

4.2 The effect of legume height at initiation of summer
grazing (HI) on cycle 1 (Cl), cycle 2 (C2), and annual
(AN) total dry matter accumulation of aeschynomene-
limpograss associations................................ 60

4.3 The effect of legume seedling growth stage (LSGS) when
early season grazing ended on aeschynomene dry matter
(DM) accumulation and percentage legume in aeschynomene-
limpograss associations............................... 61

4.4 The effect of aeschynomene height at initiation of
summer grazing (HI) on cycle 1 (Cl), cycle 2 (C2), and
annual (AN) aeschynomene (A) and limpograss (LG) dry
matter accumulation in aeschynomene-limpograss
associations........................................... 62















4.5 The percentage of total herbage accumulated (HA),
total herbage mass at initiation of grazing (HM), and
total herbage consumed (HC) that was aeschynomene,
limpograss, and weed dry matter........................ 66

4.6 Regression equations relating the percentages of three
botanical components (aeschynomene, limpograss, and
weeds) in total herbage in the upper layers of the
sward (UL), in total herbage mass present at the
beginning of a grazing cycle (HM), and in total herbage
consumed during the grazing period (HC)................ 67

5.1 The effect of aeschynomene height at initiation of
summer grazing (HI) on crude protein (CP) concentrations
in aeschynomene (A), limpograss (LG), and total (T)
herbage accumulated (HA) and herbage consumed (HC)..... 86

5.2 The effect of aeschynomene seedling growth stage (LSGS)
when establishment period grazing ended on crude
protein (CP) concentration of total herbage accumulated
(HA) and total herbage consumed (HC) in aeschynomene-
limpograss swards...................................... 88

5.3 Regressions of crude protein (CP) concentration in
herbage consumed (HCCP) vs. CP concentration in herbage
accumulated (HACP), of HCCP vs. percentage aeschynomene
in herbage accumulated (PCAHA), and of in vitro
digestible organic matter (IVDOM) concentration in
herbage consumed (HCDOM) vs. IVDOM concentration in
herbage accumulated (HADOM). These data are for
aeschynomene-limpograss swards for 1983 and 1984, and
separate regression equations were fit for each
aeschynomene height at initiation of summer grazing (HI). 89

5.4 The effect of aeschynomene height at initiation of
summer grazing (HI) on in vitro digestible organic
matter (IVDOM) concentrations in aeschynomene (A),
limpograss (LG), and total (T) herbage accumulated
(HA) and herbage consumed (HC)......................... 91

5.5 The effect of aeschynomene height at initiation of
summer grazing (HI) on organic matter accumulation
(OMA), in vitro digestible organic matter
accumulation (DOMA), and in vitro digestible organic
matter consumption (DOMC) of aeschynomene-limpograss
associations........................................... 93


viii












PAGE


6.1 Mean aeschynomene seed production (unadjusted for
closure date) for sites that were ungrazed (U), that
were last grazed before flowering (BF), or that were
last grazed after flowering (AF)....................... 106

6.2 Mean high and low temperatures for 15-day periods from
1 Sept. to 30 Nov. 1983 and 1984....................... 109

A-1 The effect of aeschynomene seeding rate on numbers of
aeschynomene seedlings in limpograss sods at 10, 20,
30, and 60 days after seeding (1983)................... 119

A-2 Legume seeding date x establishment method (D = disk
and B = broadcast) interaction means for aeschynomene
seedling number at 10, 20, and 30 days after seeding
in limpograss sods (1984).............................. 120

A-3 Legume seeding date x seeding rate (kg ha- ) inter-
action means for aeschynomene seedling number at 10,
20, and 30 days after seeding in limpograss sods (1984). 121

















LIST OF FIGURES


FIGURE PAGE

5.1 First grazing cycle responses of aeschynomene leaf (L),
stem (S), pregraze whole plant (WP), postgraze whole
plant (PWP), and herbage consumed (HC) crude protein
(CP) concentrations to aeschynomene height at initiation
of summer grazing in 1983 (n = 20 for each response
curve) ................................................. 82

5.2 First grazing cycle responses of aeschynomene leaf
(L, Y = 769 1.43x), stem (S, Y = 810 10.07x +
0.06x ), pregraze whole plant (WP, Y = 780 3.88x),
postgraze whole plant (PWP2 Y = 914 18.89x + 0.138x ),
and herbage consumed (HC, Y = 788 2.97x) in vitro
digestible organic matter (IVDOM) concentrations to
aeschynomene height at initiation of summer grazing in
1983 (n = 20 for each response curve).................. 83

5.3 The responses of aeschynomene leaf:stem ratio to
aeschynomene height at first grazing in 1983 and 1984.. 84

6.1 The response of aeschynomene seed yield plant in 1983
and 1984 to closure date of fall grazing. Negative
closure dates are days before first flower, and
positive closure dates are days after first flower
when grazing ended..................................... 104

6.2 The response of ungrazed dry matter (DM) remaining at
seed harvest to closure date for 20-, 40-, and 60-cm
initiation height treatments (1984). Negative closure
dates are days before first flower, and positive closure
dates are days after first flower when grazing ended... 108
















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


PLANT RESPONSES OF AN Aeschynomene americana--
Hemarthria altissima ASSOCIATION TO GRAZING MANAGEMENT

By

LYNN ELWOOD SOLLENBERGER

August 1985

Chairman: Kenneth H. Quesenberry
Major Department: Agronomy

Limpograss [Hemarthria altissima (Poir.) Stapf et C. E. Hubb.]

was selected for use in Florida because it has higher digestibility

than do most tropical grasses. Crude protein (CP) concentration of

limpograss may be quite low, however. The summer-annual legume

aeschynomene (Aeschynomene americana L.) can be high in forage

quality, and it has potential for use in association with limpograss.

Experiments were conducted in 1983 and 1984 to study aeschynomene

establishment in 'Floralta' limpograss sods and to evaluate the

effects of grazing management on the aeschynomene-limpograss

association.

Small plot studies of aeschynomene establishment suggested that

soil moisture at seeding, rainfall within approximately 10 days of

seeding, and control of grass competition were most critical in

determining success or failure. In 1983 climatic conditions were

excellent, and legume germination and emergence were not affected by

seedbed preparation or by seeding method. Subsequent legume dry


xi












matter (DM) harvested was greater for seedbed preparation treatments

that minimized limpograss competition. Under adverse environmental

conditions in 1984, use of higher legume seeding rates (14 vs. 7 kg

ha- ) and a seeding method which incorporated seeds into the soil (vs.

broadcast seeding) produced superior legume stands.

Under grazing, grass stubble height (7.5 or 15 cm) did not affect

aeschynomene germination or early seedling survival, but subsequent

legume DM accumulation tended to be higher for the 7.5-cm treatment.

Continued grazing of the grass until aeschynomene seedlings were 5 cm

tall decreased grass competition to the seedlings and improved legume

performance. Initiating summer grazing of the association when

aeschynomene was 80 (1983) or 60 cm (1984) tall maximized seasonal

legume DM accumulation. Earlier grazing (20 cm) resulted in more

uniform accumulation of total and aeschynomene DM throughout the

season, higher efficiency of grazing, more vigorous legume regrowth,

and a trend toward greater total herbage consumption.

Aeschynomene leaves had CP and in vitro digestible organic matter

concentrations of 250 g kg- DM and 700 g kg- organic matter.

Limpograss CP concentration was 25 to 40 g kg-1 DM. Forage quality of

the association was highest when grazing was initiated at 20 cm.

Crude protein concentration of herbage consumed was generally above 70

g kg- DM when percentage legume in total herbage accumulated was 15%

or greater.

Aeschynomene seed production decreased as closure date of fall

grazing was delayed. Legume height at initiation of grazing did not

affect seed yield plant-1.


xii
















CHAPTER I
INTRODUCTION


The beef cattle industry in Florida has been based on perennial

grass pastures for many years. Grass-N systems are highly productive

in the humid subtropics, and they require less intensive management

than do associations of grasses and legumes (Burton, 1976).

Unfortunately, forage quality of tropical grasses is often quite low

due to their high cell wall content, low concentrations of digestible

organic matter and crude protein (CP), and long retention times in the

reticulo-rumen (Wilson and Minson, 1980). Adapted and higher quality

grasses and grass-legume associations are needed to increase animal

production in the region.

Limpograss [Hemarthria altissima (Poir.) Stapf et C. E. Hubb.] is

one tropical grass that has been evaluated for use in Florida

pastures. It has yield potential comparable to that of other tropical

grasses, and cultivar 'Bigalta' has higher concentrations of in vitro

digestible organic matter (IVDOM) than do most tropical species at

similar growth stages (Quesenberry et al., 1978). 'Floralta'

limpograss is the most recently released cultivar. It is more

persistent and generally higher yielding than Bigalta, but IVDOM

concentrations for the two cultivars were not found to be different

(Quesenberry et al., 1981). Despite its higher digestibility, average

daily gains of animals grazing Floralta have not been superior to

those for bahiagrass (Paspalum notatum Flugge) (Quesenberry et al.,












1984). Euclides (1985) determined that CP concentrations of N-

fertilized, continuously stocked Floralta pastures were lower than 50

g kg-l dry matter (DM) during large portions of the grazing season.

These results suggest that low CP concentration may be limiting

voluntary intake and animal performance on Floralta pastures.

Options for improving the CP concentration of limpograss swards

include N fertilization and associating forage legumes with the grass.

The cost of N applications to maintain grass CP concentration above 70

g kg- DM throughout the season is likely to be prohibitive. Forage

legumes, though requiring additional management skill, may be a useful

alternative.

Aeschynomene (Aeschynomene americana L.) is a widely adapted

legume (Hodges et al., 1982) that is high in forage quality

(Gildersleeve, 1982) and has potential for use in association with

limpograss (Gomes, 1978). Aeschynomene is a summer-annual in Florida,

with peak productivity occurring from late July through mid-October.

This is a critical time in the state as grass quality is generally

quite low and animal performance poor. Use of a legume like

aeschynomene may help to avoid the so-called "summer slump."

Little information is available regarding the establishment of

aeschynomene into limpograss sods. Grazing management effects on the

legume and the legume-grass association also need to be studied. A

series of experiments were conducted in 1983 and 1984 to provide

additional information on the potential of the aeschynomene-limpograss

association. The general objectives of this research were 1) to

evaluate aeschynomene establishment in Floralta limpograss sods under










3


a range of management strategies and environmental conditions, 2) to

determine the effects of grazing management on legume and sward

productivity, pasture botanical composition, and forage quality of the

association, and 3) to compare grazing management effects on

aeschynomene seed production.

Results of these experiments are presented in a series of four

papers (Chapters III through VI) as follows: factors affecting

aeschynomene establishment in limpograss (Chapter III), productivity

and botanical composition of an aeschynomene-limpograss association in

response to grazing management (Chapter IV), forage quality responses

of the association to grazing (Chapter V), and aeschynomene seed

production responses to defoliation management (Chapter VI).
















CHAPTER II
LITERATURE REVIEW


The objectives of this review are 1) to provide a thorough

overview of the species Hemarthria altissima and Aeschynomene

americana, and 2) to supply the reader with the framework from which

the dissertation research evolved and on which it was based. No

attempt will be made to present a comprehensive discussion of the

principles of pasture management, utilization, or quality. Instead,

the majority of papers cited will contain information that is directly

related to the species and/or problems to be investigated.


General Description of Limpograss and Aeschynomene


Limpograss


Limpograss [Hemarthria altissima (Poir.) Stapf et C.E. Hubb.]

belongs to the tribe Andropogoneae of the family Poacea. It was

introduced into the United States in 1964 from plant collections made

in South Africa by Dr. A. J. Oakes of the USDA Plant Introduction

Service. The original collection consisted of four plant

introductions, three of which have proven to be agronomically useful.

These were released in Florida as 'Redalta', 'Greenalta', and

'Bigalta' (Quesenberry et al., 1978). Additional plant exploration

(Oakes, 1973) led to the introduction and subsequent release of a

limpograss line called 'Floralta' (Quesenberry et al., 1984).




4












The limpograsses are stoloniferous tropical grasses that

perennate in Florida. They exhibit an erect growth habit, and the

grass canopy tends to be quite stemmy, with leaves that are generally

small and narrow. The inflorescence is a single, spike-like raceme,

and more than one raceme may emerge from a node (Quesenberry et al.,

1978). Seed set is very low, and all of the limpograsses are

established vegetatively.

All cultivars are highly productive in Florida and are well

adapted to wet soils (Quesenberry et al., 1978; Quesenberry et al.,

1984). Redalta and Greenalta are diploid types with good persistence

but low digestibility. Bigalta is a tetraploid with high

digestibility (Quesenberry and Ocumpaugh, 1980; Schank et al., 1973),

but it does not persist well under intensive grazing management

(Pitman et al., 1984). Floralta, also a tetraploid, was specifically

selected for persistence under grazing. Like Bigalta it is highly

productive, and digestibility values are quite good for a tropical

grass (Quesenberry et al., 1981). Unfortunately, N concentrations in

Floralta forage may be so low that animal performance is limited

(Christiansen, 1982; Euclides, 1985), and this possibility has

generated interest in legume-limpograss associations.


Aeschvnomene


A legume that has shown potential for use in association with

limpograss is aeschynomene (Aeschynomene americana L.). Aeschynomene

is of the tribe Aeschynomenea of the Papilionoidae subfamily of

Leguminosae (Allen and Allen, 1981). Aeschynomene is native to the

subtropics of the Western Hemisphere (Hodges and McCaleb, 1972), and












it is indigenous to peninsular Florida (Hodges et al., 1982). No

improved cultivars of A. americana have been released in Florida, but

the legume is the most widely adapted warm-season legume available for

grazing in the state (Hodges et al., 1982).

Aeschynomene is a tall-growing, woody plant that behaves as a

summer annual in this region. Fully grown, it reaches heights of 1 to

2 m. The leaves are pinnately compound, with each blade divided into

25 to 60 leaflets. The leaves measure approximately 2 by 8 cm and are

sensitive to touch, with the two rows of leaflets folding together

when disturbed (Hodges et al., 1982). Aeschynomene exhibits a short-

day floral initiation habit (Kretschmer and Bullock, 1980), and

flowering of the Florida common ecotypes occurs about 10 September in

north-central Florida (Quesenberry and Ocumpaugh, 1981). Flowers are

yellow to violet in color, and they develop on axillary racemes

(Bogdan, 1977). The seed pods have four to eight segments or joints,

and they are generally straight at the upper margin and deeply

indented between joints on the lower margin (Bogdan, 1977).

Aeschynomene is well adapted to poorly drained soils and occasional

standing water (Albrecht et al., 1981; Miller and Williams, 1981).

Moore and Hilmon (1969) have observed the legume growing in water 30

to 45 cm deep. In Florida, almost 60% of the soils are subject to

seasonal water-logging, so tolerance of such conditions is a very

desirable attribute (Allen, 1977).

Until recently aeschynomene was thought to be of minimal

agronomic value (Bogdan, 1977). Studies in Florida have shown that

the legume can be quite high in forage quality and that maximum forage

production occurs during a time when grass quality is too low to












sustain high levels of animal performance (Gildersleeve, 1982; Hodges

et al., 1982; Ocumpaugh and Dusi, 1981). These findings have spurred

interest in overseeding perennial grass pastures with aeschynomene.


Establishment


Perennial pasture grasses, particularly bahiagrass (Paspalum

notatum Flugge), are the mainstay of the Florida beef cattle industry.

Bahiagrass, though of low forage quality (Euclides, 1985; Moore et

al., 1981), is used extensively because it is easy to establish and

persists well. Establishment of potential pasture grasses and legume-

grass associations must likewise be relatively easy and of low risk.

This portion of the review considers some of the important factors

affecting limpograss and aeschynomene establishment.


Limpograss


Schank (1972) reported that seedset occurred in 8 of 11

limpograss introductions evaluated, but in the African introductions

seedset averaged only 2%. Because the cultivars released in Florida

are African lines, these results indicate that propagation of

limpograss will depend upon vegetative plantings.

/Ruelke et al. (1979) evaluated Redalta, Bigalta, and Floralta

establishment using three different planting techniques. ,Best results

were obtained when sprigging was following by light disking and

cultipacking. Bigalta and Floralta were generally superior to Redalta

in establishment-year ground cover and dry matter (DM) harvested.

jQuesenberry et al. (1984) noted that a bermudagrass [Cvnodon

dactylon (L.) Pers.] sprig planter could also be used to establish












limpograss. Using the sprig planter set for a 50-cm row spacing, 750

to 1250 kg of vegetative material were required to plant 1 ha. Moist

or wet soil is essential for successful limpograss establishment

(Quesenberry et al., 1978), and irrigation may be necessary if summer

rains are inadequate. Establishment fertilizer should be low in N and

high in P and K to encourage root development without promoting

excessive weed growth. Limpograss is slower to establish than

digitgrass (Digitaria spp.) or bermudagrass, so broadleaf weed control

may be necessary (Quesenberry et al., 1984). Despite slower

establishment, grazing can be initiated in the spring following

planting.


Aeschynomene


Aeschynomene is a summer annual that is capable of producing

large quantities of seed (Hodges et al., 1982; Moore and Hilmon,

1969). Seeds generally remain in the pod, whether they drop from the

plant or are combined (Ruelke et al., 1974). Germination of unhulled

seed has been reported to be 2 to 6%, with a high proportion of hard

seed (Hanna, 1973). Pod removal and mechanical scarification

increases germination from around 5% to between 78 (Hanna, 1973) and

98% (Ruelke et al., 1974).

Seed with or without the pod can be obtained commercially, and

both types may be useful depending on conditions at seeding. If

adequate moisture is available, then naked seed is desirable due to

its rapid, uniform germination. If moisture is limiting or is likely

to become limiting, seed in the pod or a mixture of the two types

should be used. This practice provides a viable reserve should early












germinating seedlings die due to environmental stress (Hodges et al.,

1982).

Aeschynomene is typically seeded at the onset of the rainy season

(1 June to 1 July) in Florida (Hodges et al., 1982). Germination is

rapid in warm, moist soils, but seedling growth tends to be relatively

slow. Competition from weeds or rapidly growing tropical grasses must

be minimized during the establishment period (Hodges et al., 1982).

Less risk is involved when seeding legumes into conventionally

prepared seedbeds than into sods because plant cover is destroyed,

moisture infiltration and retention may be increased, and seed and

rooting environments are improved (Cook and Lowe, 1977). If seeding

on a conventionally tilled area, aeschynomene seed should be rolled or

disked into the soil to no greater than a 2-cm depth (Hodges and

McCaleb, 1972). Cultivated land seeded to aeschynomene should not be

grazed from the time of germination until plants reach the desired

height for initiation of grazing (Hodges and McCaleb, 1972).

Oversowing of grass sods with legumes, either by sod seeding or

surface broadcasting, has potential as a pasture improvement tool in

Florida. Establishment costs can be low when compared to conventional

tillage and seeding systems (Cook, 1981), but conditions for the

establishing seedling may be much more severe (McWilliams and Dowling,

1970). Successful establishment is dependent on the ability of the

seedling to compete with the old sod for light, plant nutrients, and

moisture (Robinson and Cross, 1960). To favor the developing legume

seedling, the seedling's micro-environment and its competitive

relationships with other plants must be manipulated by management.












Methods of establishing aeschynomene in grass sods have been

evaluated in Florida. Kalmbacher et al. (1978) reported that

herbicidal suppression of bahiagrass facilitated aeschynomene

establishment and increased yield and crude protein of the grass-

legume mixture. Kalmbacher and Martin (1983) achieved excellent

aeschynomene establishment by grazing bahiagrass to 3- to 5-cm

stubble, drilling legume seed into the sod, and continuing to graze

until aeschynomene was 2.5 cm tall. Herbicide applications also were

successful in controlling grass competition, but their cost may be

prohibitive for beef production systems in Florida. Gomes (1978)

evaluated five methods for establishing legumes in Bigalta limpograss.

Methods included 1) no tillage, seed broadcast, 2) light disking, seed

broadcast, followed by cultipacking, 3) seed broadcast followed by

disking and cultipacking, 4) sod seeding with a Zip seeder, and 5)

complete seedbed preparation, seed broadcast, and cultipacking.

Despite very high soil temperatures (380C) in the completely prepared

seedbeds, there were more aeschynomene seedlings 6 weeks after

planting in this treatment than in any other. Gomes and Kretschmer

(1978) suggested that high temperatures scarified depodded

aeschynomene seed and increased germination percentage. Gomes (1978)

found no differences in subsequent aeschynomene DM harvested due to

establishment method. Of five legumes planted with Bigalta,

aeschynomene DM production was greatest in the first year. High seed

yields enabled aeschynomene to successfully reseed, and stands were

excellent in 2 following years.

Methods of stimulating germination and growth of naturally

reseeding aeschynomene stands have been investigated. Moore (1978)












reported that heavy disking, which involved turning the sod and

incorporating surface litter, increased aeschynomene DM harvested four

times over that observed with no sod treatment. Tang and Ruelke

(1976) obtained higher legume seedling densities and legume DM

harvests in burned than in mowed bahiagrass plots. Ruelke et al.

(1974) burned an aeschynomene-bahiagrass sod after the legume had

produced seed. Compared to mowing the grass sod and removing the

residue, burning resulted in more aeschynomene seedlings, greater

mixture DM production, greater percentage legume in the mixture, and

higher yields of digestible organic matter (DOM). The authors

suggested that burning scarified the hard seeds of the legume enabling

them to imbibe water and germinate.


Responses to Clipping and Grazing


Successful growth of legume-grass associations demands a much

higher level of management than that required for well-adapted

perennial grasses receiving N fertilizer (Burton, 1976). If tropical

and subtropical production systems utilizing mixed swards are to be

successful, a thorough understanding of plant responses to cutting and

grazing managements will be needed. This portion of the review

considers the effects of defoliation on forage DM harvested,

subsequent plant regrowth, and stand persistence of limpograss and

aeschynomene.


Limpograss


Limpograss is a highly productive forage grass in Florida. Early

evaluation showed Bigalta to be more productive than 'Coastcross-l'












bermudagrass and a number of digitgrasses (Digitaria decumbens Stent.)

(Quesenberry et al., 1978). From 1971 to 1973, three limpograsses

were compared with other perennial, tropical grasses (Hodges and

Martin, 1975). Redalta and Greenalta out-yielded 'Pensacola'

bahiagrass, but they were comparable in yield to 'Coastal'

bermudagrass. Bigalta productivity was comparable to that of

bahiagrass, but it was lower yielding than bermudagrass. Harvested

every 4 to 5 weeks during the wet season, the limpograsses did not

persist well, and plots were invaded by weeds in the third year. At

Quincy AREC, Redalta, Greenalta, and Bigalta were compared to other

tropical forage grasses including Pensacola bahiagrass and Coastcross-

1 bermudagrass (Quesenberry et al., 1978). Grasses were cut to 2.5-,

7.5-, and 15-cm stubbles at 4- to 5-week intervals. Bigalta out-

yielded Redalta and Greenalta, and limpograss yields were generally

higher than those of bahiagrass, but less than those of bermudagrass.

Stands of all limpograsses were severely reduced in the second year

suggesting that the dry, upland soils of north and west Florida are

not well suited for limpograss production.

Ruelke et al. (1976) compared the four limpograss cultivars in a

small-plot clipping trial and found Floralta to be the top-yielding

genotype. Christiansen (1982) compared Floralta with Bigalta and

Redalta at 3-, 9-, and 18-week defoliation frequencies. Forage DM

harvested increased with increasing length of rest period, and

Floralta was generally either first or second in total DM harvested.

In another study, the limpograsses were grazed or clipped at 5-week

intervals (Quesenberry et al., 1984). Floralta was generally superior

to the other varieties in DM accumulated, and it was more persistent












than Bigalta. After 3 years of mob grazing, Ocumpaugh (1982) observed

that percent ground cover ranged from 75 to 95 for Floralta compared

with 5 to 85 for Bigalta. Christiansen et al. (1981) have suggested

that the difference in persistence between Bigalta and Floralta is

related to reserve carbohydrate status. Floralta partitions more

carbohydrates to stem bases than does Bigalta, presumably giving it an

advantage after defoliation.

No grazing recommendations have been published for Floralta. The

following are specific suggestions for managing Bigalta. Bigalta

should be grazed rotationally to maximize efficiency of utilization

and to increase persistence. During the warm season, grazing periods

should end when grass stubble is 20 to 30 cm (Chambliss, 1978;

Quesenberry et al., 1978). Bigalta maintains a relatively high level

of digestibility with increasing maturity (Quesenberry and Ocumpaugh,

1980), so it may be useful as a stockpiled forage. Caution must be

exercised, however, as late fall applications of N fertilizer followed

by intensive grazing can reduce stands (Chambliss, 1978).


Aeschynomene


Early plantings of aeschynomene were made in 1952 at the Ona

Agricultural Research Center (Hodges et al., 1982). The area was

grazed for 2 years, but no management data were found in the

literature. Moore and Hilmon (1969) published results from work done

in 1963 through 1966. They observed that small stems and leaves of

aeschynomene were very palatable to deer and cattle. Heavy browsing

of the legume in the establishment year resulted in a more prostrate

growth habit and profuse branching.












Tang and Ruelke (1976) evaluated aeschynomene production in

response to plant height at cutting (45 or 90 cm) and to stubble

height (15 or 30 cm). There were no differences in legume DM

harvested due to cutting treatment, but trends favored defoliation at

45 cm. The authors proposed that regrowth of plants cut first at 90

cm was poor due to death of shaded axillary bud sites on the lower

stem. Higher stubble heights were recommended to insure persistence

when mature plants are cut.

Albrecht and Boote (in press) evaluated the effect of stubble

height (9 or 18 cm) on regrowth characteristics of aeschynomene plants

that were 60 cm tall or taller when first cut. Virtually all leaf

area was removed from the plants at both stubble heights. The 18-cm

stubble had more viable axillary buds than did the 9-cm stubble, and

this resulted in a more rapid recovery of leaf area and canopy carbon

exchange rate for the 18-cm treatment.

Mislevy et al. (1981) compared aeschynomene DM production when

plants were cut at 30-, 60-, or 90-cm heights. Highest seasonal DM

harvests were from plots cut first at 30 cm, with regrowth cut at 90

cm. Lowest DM production occurred with the 30-cm initial, 30-cm

regrowth cut system. High plant mortality was observed in both years

when plants were allowed to grow to 90-cm heights before initial

harvest. This was attributed to shading and death of axillary buds on

lower stems. In general an 8-cm stubble height treatment was more

productive than an 18-cm treatment, but it was noted that increasing

initial harvest height while maintaining the 8-cm stubble height

decreased total yield due to loss of bud sites for regrowth.

Branching was most vigorous when plants were cut initially at 30 cm,












and regrowth was more rapid for plants cut to an 18-cm versus an 8-cm

stubble.

Gildersleeve (1982) evaluated aeschynomene response to grazing.

Initiation of grazing occurred when plants were 28, 45, or 54 cm tall.

All treatments were subsequently grazed using a 4-week defoliation

interval. Plants grazed initially at 28 cm were most persistent and

continued to make vigorous regrowth throughout the season. In

addition, these plants tended to accumulate more DM over the season

than did those where grazing was initiated later. Gildersleeve (1982)

noted that early grazing opened up the canopy, decreased light

competition, and stimulated axillary bud development and secondary

branching.

Recommendations for grazing aeschynomene have been made by a

number of authors (Chambliss, 1982; Hodges, 1977; Hodges and McCaleb,

1972; Hodges et al., 1982; Kalmbacher and Mislevy, 1978; Mislevy et

al., 1981). It has generally been suggested that grazing be initiated

when aeschynomene is 45 to 60 cm in height. Grazing should cease when

75% of leaves and small stems are removed, and remaining stubble

should be 15 to 24 cm. Subsequent grazing should be on a rotational

basis when regrowth is 45 to 60 cm tall.

Gildersleeve (1982) cautioned that results obtained from clipping

should not be the basis for grazing recommendations. Data from

clipping trials overestimate DM harvested and underestimate forage

quality. Responses observed with aeschynomene under grazing

(Gildersleeve, 1982) suggest that current recommendations may need to

be revised, but additional evaluation with grazing animals is needed.












Forage Quality


The most useful definitions of forage quality are expressed in

terms of output per animal or voluntary intake of digestible energy

(Moore, 1980; Mott, 1959). It is important to note that animal

potential and forage availability act independently of forage quality

to restrict output per animal (Mott and Moore, 1970). As a result,

forage quality can be expressed in terms of output per animal only

when 1) the forages being compared are the sole sources of energy and

protein, 2) the amount offered exceeds consumption by 5 to 15%, and 3)

the animals have potential for production (Moore, 1980; Moore and

Mott, 1973).

When these conditions are met, the main factors controlling

output per animal (forage quality) are voluntary intake,

digestibility, and efficiency of utilization of forage nutrients

(Milford and Minson, 1965; Minson, 1980; Moore and Mott, 1973).

Voluntary intake varies to a much greater extent than does

digestibility (Minson, 1971), and it is considered to be the most

important factor accounting for differences in forage quality (Moore

and Mott, 1973).

For the purpose of routine laboratory evaluation of forage

quality the present method of choice is in vitro determination of

digestibility (Moore, 1980). When tropical grasses are being

evaluated, N determinations are also important, as N concentrations of

10 to 13 g kg-1 DM or lower can limit animal performance (Minson,

1980).












This portion of the review examines limpograss and aeschynomene

forage quality responses to a range of management practices. Forage

quality in this context includes results from 1) laboratory procedures

used to estimate quality, 2) feeding trials in confinement, and 3)

large-scale animal production experiments.


Limpograss


Intake of tropical grasses is usually less than that of temperate

grasses harvested at similar growth stages (Minson, 1980). This

response to tropical species has been associated with their higher

concentration of cell wall, lower dry matter digestibility, larger

percentage of indigestible cell wall, and longer retention time in the

reticulo-rumen (Minson, 1980).

Early screening of limpograss lines showed that Bigalta, a thick-

stemmed tetraploid, had higher in vitro organic matter digestibility

(IVOMD) than did the fine-stemmed diploids, Redalta and Greenalta

(Schank et al., 1973). These differences were attributed to stem

anatomy as the more digestible cultivar had a lower percentage of

vascular bundle per stem cross-sectional area.

Carvalho (1976) showed that Bigalta limpograss was 70 to 180 g in

vitro digestible organic matter (IVDOM) kg-1 OM more digestible than

were Pensacola or 'Argentine' bahiagrass for regrowth periods ranging

from 1 to 22 weeks. Moore et al. (1981) reported Bigalta IVDOM

concentration to be 60 to 80 g kg-I OM higher than that of bahiagrass

at 4 or 6 weeks of regrowth. Organic matter intake of Bigalta fed to

sheep (expressed as a percent of body weight) was 0.20 to 0.46 units

higher than that observed for Pensacola bahiagrass (Moore et al.,












1981). Bigalta IVDOM concentration declines more slowly with

advancing maturity than does that of many other tropical grasses

(Quesenberry et al., 1981). Schank et al. (1973) reported that IVDOM

concentration of 5-week regrowth of Bigalta was 684 g kg-1 OM, while

that of mature plants (first harvested 17 September) was 660 g kg-1

OM. In vitro digestible organic matter concentration in 14-week

Bigalta regrowth was 620 g kg-1 OM (Quesenberry and Ocumpaugh, 1980).

These responses and Bigalta's relatively good frost tolerance have led

to suggestions that it be used as a stockpiled (Chambliss, 1978;

Quesenberry and Ocumpaugh, 1980), or off-season forage (Kretschmer and

Snyder, 1979; Ruelke and Quesenberry, 1982).

Floralta is also a tetraploid type, and digestibility of Floralta

was reported to be similar to (Quesenberry et al., 1981), or somewhat

less than (Christiansen, 1982) that of Bigalta. When clipped or

grazed at 5-week frequencies, IVDOM concentration for Floralta was 670

and 650 g kg-1 OM, respectively (Quesenberry et al., 1984). Under

continuous grazing, Floralta IVDOM concentration averaged 100 g kg-1

OM higher than that of Pensacola bahiagrass (Euclides, 1985). The

IVDOM concentration of esophageal extrusa from the same pastures

differed by 50 g kg-1 OM.

Animal production trials with Bigalta and Floralta have given

mixed results. Centro Internacional de Agricultura Tropical (CIAT)

experiments in Colombia have shown continuously grazed limpograss

(cultivar name not given) to be superior to Andropogon gayanus Kunth.,

Cynodon nlemfuensis Vanderyst, Brachiaria decumbens Stapf, and

Brachiaria humidicola (Rendle) Schweickt, in steer gain ha-1 (Tergas

et al., 1982). Gain per animal for limpograss was superior to only A.




____











gayanus during a 10-month grazing season. Pitman et al. (1984)

reported animal gains on Bigalta of 0.71 kg d-1 during May at Ona,

Florida, but gains were only 0.12 kg d-1 during August. Limpograss

IVDOM concentration was 46 g kg-1 OM lower (526 vs. 572 g IVDOM kg-1

OM) in August than it was in May. In the first year of a production

trial near Gainesville, Florida, average daily gains from Floralta and

Pensacola bahiagrass pastures were 0.35 and 0.33 kg, respectively

(Quesenberry et al., 1984). Gain ha- was twice as great for Floralta

because of higher carrying capacity and a longer grazing season. In 3

subsequent years, daily gains on Floralta were similar to or slightly

lower than those for bahiagrass, but 100-kg live-weight days ha-I and

gain ha- were greater for Floralta (G. 0. Mott and C. S. Jones,

unpublished data; Quesenberry et al., 1984). That limpograss failed

to produce higher average daily gains than bahiagrass was

disappointing in light of the large differences in laboratory

estimates of quality.

A possible explanation for the lower than expected gains is low N

concentration in limpograss forage. As previously discussed, low N

concentration in forages may limit voluntary intake, and intake is the

primary determinant of animal production (Minson, 1980). Intake of

tropical grasses has been observed to decline rapidly when crude

protein (CP) concentration in the consumed feed fell below 70 g kg-

DM (Milford and Minson, 1965; Minson and Milford, 1967). Minson

(1980) suggests that for most forage diets the critical CP

concentration is in the range of 60 to 80 g kg-1 DM.

Carvalho (1976) reported that CP concentration of Bigalta was

generally lower than that of digitgrass or bahiagrass. Nine-week












regrowth had a CP concentration of 58 g kg- DM. In work done by S.

W. Coleman (Quesenberry et al., 1978), 6-week Bigalta regrowth on

organic soils had 114 g CP kg-1 DM. Data from north Florida indicated

that Greenalta and Bigalta CP concentrations may be low at mid- and

late-season (Quesenberry et al., 1978). Results from Hawaii suggest

that limpograss CP concentration may be lower than that of other

tropical grasses tested (Quesenberry et al., 1978). Bigalta grown

during the cool season in south Florida had less than 70 g CP kg- DM

for a range of N fertilization rates from 112 to 336 kg ha-1 and

cutting intervals from 4 to 12 weeks (Kretschmer and Snyder, 1979).

Quesenberry and Ocumpaugh (1980) recommended N supplementation for

stockpiled Bigalta, as CP concentration was near or below 70 g kg-

DM. In association with legumes and receiving no N fertilizer, Gomes

(1978) reported mean Bigalta CP concentrations of 36, 53, and 37 g

kg- DM for forages harvested on 16 June, 28 July, and 10 Sept. 1977.

Christiansen (1982) reported that the CP concentration of

Floralta limpograss was particularly low during mid-summer. Plants

harvested in July had CP concentrations ranging from 12 to 47 g kg-

DM depending on cutting frequency (6 to 18 weeks) and N fertilizer

rate (0 to 480 kg ha- ). Only when 3-week regrowth was harvested were

CP concentrations above 70 g kg- DM. March applications of 75 kg N

ha-1 resulted in 4- and 6-week Floralta regrowth containing greater

than 120 and 80 g CP kg- DM, respectively (Ruelke and Quesenberry,

1983). By 12 May, 8 weeks after N application, CP concentration was

approximately 60 g kg- DM. Delaying N application to 14 April

resulted in forage with approximately 80 g CP kg-l DM on 12 May. Late

summer (4 August) application of 75 kg N ha-1 increased grass CP












concentration to 100 g kg-I DM by 14 September. By 29 September, CP

concentration had fallen to just above 40 g kg-1 DM. In the second

year, 34 kg N ha applied in March increased CP to just 70 g kg-1 DM,

and by 10 May Floralta CP was less than 40 g kg-' DM. In a

continuously grazed, N-fertilized Floralta pasture, mean grass CP

concentration was 47 g kg-1 OM (top 20 cm of sward) from 21 June

through 28 October (Euclides, 1985). Mean CP concentration of

esophageal extrusa of grazing steers was 58 g kg-1 OM. By

approximately 20 July, CP concentration of the extrusa was less than

60 g kg-1 OM, and it dropped to 50 g kg-1 OM during parts of July and

August.


Aeschynomene


Thornton and Minson (1973) have observed that legumes generally

have a higher voluntary intake than do grasses of the same

digestibility. At IVDOM concentrations of 600 g kg-l OM, they

reported that mean daily intake of six legume species was 28% greater

than that observed for eight grasses (included both C3 and C4

species). Another advantage of legumes is their contribution of N to

the animal diet. Minson and Milford (1967) increased voluntary intake

of poor quality Pangola digitgrass (40 g CP kg- DM) by including 100

to 200 g legume kg-' diet DM.

Paul (1951) described the nutritive value of aeschynomene as

being comparable to that of alfalfa, but he indicated that it was less

palatable. Moore and Hilmon (1969) observed that aeschynomene was

readily consumed by both deer and cattle in Florida. Animals selected












young plants and the terminal branches and leaves of more mature

forage.

Kalmbacher and Mislevy (1978) obtained highest yields of CP and

DOM when aeschynomene was first cut at 30 cm and regrowth was cut at

90 cm. Crude protein concentration was 185 g kg-I DM for plants cut

initially at 30 cm, and 151 g kg-1 DM for those cut first at 90 cm.

Spicer et al. (1982) harvested aeschynomene for haylage at late-bloom

stage. Crude protein and neutral detergent fiber concentrations were

103 g kg-1 DM and 742 g kg-1 OM, respectively. Harvesting mature

aeschynomene for haylage was determined to be a poor way to utilize

the legume forage. When aeschynomene was cut at 30-cm heights

throughout the season, CP and IVDOM concentrations averaged 175 g kg-

DM and 700 g kg- OM, respectively (Mislevy et al., 1981). Delaying

initial harvest reduced IVDOM concentration 80 g kg-I OM for each 30-

cm increment of legume height. Cutting plants to a stubble of 18
-1
rather than 8 cm increased IVDOM concentration by 20 to 55 g kg- OM,

indicating that higher quality forage was concentrated toward the top

of the canopy (Mislevy et al., 1981). Hodges and McCaleb (1972)

reported CP concentrations of 241, 213, and 61 g kg-1 DM for plant

fractions that included the top 15 cm of tall plants, leaves and fine

stems only, or coarse stems (3 to 6 mm in diameter), respectively.

Gildersleeve (1982) evaluated aeschynomene under mob grazing.

Higher leaf to stem ratios were observed when grazing was initiated at

28 rather than 45 or 54 cm. This was important as animals harvested

essentially leaves, seeds, and stems less than 4 mm in diameter. The

leaf plus seed fraction averaged 250 g CP kg-1 DM and 750 g DOM kg-1

OM from August through October (Gildersleeve, 1982). These data












indicate that aeschynomene is capable of furnishing high quality

forage during a period when perennial grass quality does not meet the

needs of a lactating beef cow or a grazing calf (Hodges et al., 1982).

Animal performance has been measured using aeschynomene under a

variety of grazing managements. Hodges et al. (1974) seeded

aeschynomene into a bahiagrass sod to supplement a digitgrass pasture

system. An average of 82% of the cows weaned calves for the system

including aeschynomene, and 67% weaned calves when grazing digitgrass

alone. Calf weaning weights were similar for the two systems, and

average calf production ha-1 was slightly higher when using the

legume. Hodges (1977) reported yearling heifer daily gains of 0.55 kg

when grazing aeschynomene from 26 July to 18 October. Hodges et al.

(1976) recorded summer gains of 0.45 kg d-1 for a bahiagrass-

aeschynomene association compared to 0.26 kg d-1 for a bahiagrass-N

system. Beef production ha-1 was also slightly higher when the legume

was included. Summer-annual forages have been evaluated as creep

graze for young calves (Ocumpaugh and Dusi, 1981). Aeschynomene was

the best forage in both years of the study, and average daily gains

were 0.90 kg, compared to 0.68 kg for the bahiagrass control.


Mineral Status


McDowell et al. (1984) stated that forages can rarely satisfy all

mineral requirements of grazing ruminants in warm-climate countries.

Concentrations of mineral elements in forages are dependent upon the

interaction of a number of factors including soil fertility, plant

species, stage of maturity, yield, pasture management, and climate

(McDowell et al., 1983). In this portion of the review, data












describing the mineral status of limpograss and aeschynomene forages

is discussed.


Limpograss


Rojas-Osechas (1985) evaluated the mineral status of 4-, 6-, and

8-week Bigalta regrowth. Potassium, P, Mg, Ca, S, Mn, Fe, and Mo were

found to be of adequate concentration for beef cows with average

milking ability and body weight of 450 kg. Micronutrients were not

included in the grass fertilization program, and Cu, Se, Zn, Co, and

Na were deficient. Potassium and Cu were the only minerals to

significantly decrease in concentration with increasing maturity.

Quesenberry et al. (1978) determined nutrient composition for

Greenalta and Bigalta forages which had been fertilized with 55, 24,

and 46 kg ha- of N, P, and K, respectively. Mineral concentrations

in limpograss were described as being similar to those of other warm-

season perennial grasses. Quesenberry and Ocumpaugh (1982) reported

that stockpiled Bigalta limpograss could be used by mature pregnant

beef cows without mineral supplementation until late October to early

November. After this date, P, K, and Ca may be required if the forage

constitutes the total diet.


Aeschynomene


Mineral status of 30- to 90-cm regrowth of aeschynomene has been

determined by Kalmbacher et al. (1981). With the exception of Cu, the

needs of growing calves and yearlings should be met without

supplementation. Potassium concentration declined linearly as

initiation of cutting was delayed. Calcium, P, and Mg concentrations












showed little response to increasing maturity. Iron concentration

increased linearly with plant age while that of Mn decreased. Zinc

and Cu concentrations were not affected by delaying cutting. The

authors suggest that these data be interpreted with caution as they

represent 1 year's results at two locations.


Fertility Requirements


An adequate treatment of soil fertility relationships to pasture

production in the sub-tropics is not possible in the context of this

review. The comments that follow are very brief and relate only to

the species under consideration.


Limpograss


Productivity of Floralta limpograss increases with increasing N

rates through at least 480 kg N ha- yr- (Christiansen, 1982). When

cut four times at 9-week intervals, Floralta DM harvested was 5.9,

12.1, and 24.4 Mg ha-1 for N rates of 0, 120, and 480 kg ha-1. Dry

matter harvested was 29.3 Mg ha-1 for plots receiving 480 kg N ha-
-1
yr and cut every 18 weeks. The high cost of N fertilizer and

minimal profit margins for beef producers make high rates of N

impractical in Florida pastures. Quesenberry et al. (1984) suggested

that 120 kg N ha- yr- will provide for high productivity at

reasonable cost.

In general, grass pastures on flatwoods soils are fertilized in

the spring of the year with approximately 60 kg N, 25 kg P, and 50 kg

K ha- (Dantzman, 1978). Another application of 60 kg N ha- should

be made later in the summer. Sulfur may also be deficient in












Spodosols (Mitchell and Blue, 1981), and if needed, S should be

applied at approximately the same rate as P (Dantzman, 1978).

Soil pH of 5.5 to 6.0 is desirable for grass pastures (Dantzman,

1978). Increasing soil pH decreases the availability and uptake of

some micronutrients, so applying excess lime is not recommended

(McDowell et al., 1983). Iron deficiency has been observed when

limpograss was fertilized with high rates of N and defoliated

frequently (Quesenberry et al., 1978).


Aeschynomene


The effects of P and K fertilization of aeschynomene on DM and

seed production were evaluated by Moore (1978). Treatments in one

experiment included N-P-K rates (in kg ha- yr ) of 0-0-0, 0-20-0, 0-

0-74, and 0-20-74. Another study compared rates of 0-0-0, 0-10-37, 0-

20-74, and 0-40-148. On a Pompano fine sand (no value given for soil

P level), there was no advantage to applying P. Dry matter production
-1 -1
doubled with K application of 37 compared to 0 kg ha yr but there

was no yield increase from subsequent K additions. Similar results

were obtained for seed production.

A P and lime-rate experiment was conducted at Ft. Pierce on newly

cultivated soil with pH of 4.7, Ca concentration of 40 g kg- and P

concentration of 1 g kg-l (extractant not given) (Hodges et al.,

1982). In 2 years of observations, 2.0 Mg lime ha-1 before seeding

(first year only) and 60 kg P ha- yr-1 resulted in legume DM

harvested of 5.6 Mg ha- yr- An additional 1.0 Mg ha- of lime with

no increase in rate of P, increased aeschynomene production 0.9 Mg

ha-1 yr-1. Plots receiving 0.2 Mg lime ha-1 and 20 kg P ha-1 yr-1












yielded just 13% as much as those receiving 3.0 Mg ha-1 of lime and 60

kg P ha- yr Reporting on the same study, Kretschmer and Snyder
-1 -1
(1981) noted that aeschynomene receiving 60 versus 20 kg P ha yr

produced 2.9 Mg ha-1 of additional DM over a 2-year period.

Recommendations for sustained aeschynomene production include

liming to pH 5.5 to 6.0 (Hodges et al., 1982). Hodges et al. (1982)

suggest that P and K should be applied at rates of 15 and 56 kg ha-1

despite responses to higher levels as described above. Soils not

routinely receiving micronutrients should have a micronutrient mixture

applied (22 kg ha-1) that is similar in composition to F-503 Oxide

(Hodges et al., 1982).

Land not previously cultivated should be limed to pH 5.5 to 6.0,

and 120 kg ha-I of P205 and K20 should be applied prior to seeding

(Hodges et al., 1982). Micronutrients also may be needed, but it has

been observed that soils producing good bahiagrass supply enough

micronutrients for aeschynomene, with the possible exception of B

(Hodges and McCaleb, 1972).


Disease and Pest Problems


Limpograss


No major diseases have been associated with limpograss in Florida

(Quesenberry et al., 1984). Both Pythium and Fusarium root rot were

diagnosed on Bigalta grown on an organic soil, but these diseases have

not been significant problems (Quesenberry et al., 1978).

Insect and nematode damage have also been minimal. The armyworm

complex (family Noctuidae) that attacks many grasses in Florida can












cause severe, but generally very isolated damage (Quesenberry et al.,

1984). Carvalho (1976) observed that there was less insect feeding on

Bigalta than there was on bahiagrass or digitgrass. Greenalta and

Floralta were found to be less susceptible to the yellow sugar-cane

aphid (Sipha flava Forbes) than was Pangola digitgrass (Oakes, 1978).

Boyd and Perry (1969) reported that yields of Redalta, Greenalta, and

Bigalta were less than those of most tropical grasses when grown in a

sandy soil that was infested with sting nematodes (Belonolaimus

longicaudatus Rau). Floralta had significantly less tolerance to

sting nematode than the most tolerant Hemarthria genotype, but it was

not different from Bigalta and Redalta in tolerance (Quesenberry and

Dunn, 1978). These findings suggest that limpograss should not be

planted on sandy soils infested with sting nematode (Quesenberry et

al., 1978). Damage from this nematode does not seem to be a problem

on flatwoods soils, possibly due to high water levels during the rainy

season (Quesenberry et al., 1978).


Aeschynomene


Sonoda and Lenne (1979) have compiled a bibliography on diseases

of the genus Aeschynomene. Viral diseases of the genus include

tobacco streak virus and legume little leaf disease. Fungal pathogens

include the genera Uredo, Colletotrichum, Cercospera, and Physoderma

(Kretschmer and Bullock, 1980). Colletotrichum spp. anthracnosee) has

been isolated from Aeschynomene spp. at Ft. Pierce, but there was no

clear relationship between given species and levels of resistance.

Armyworm defoliation of aeschynomene has been reported at Ona

(Kalmbacher et al., 1981). Kretschmer and Bullock (1980) have












documented leaf and stem feeding of aeschynomene by a looper [Selenis

monotropa (Grote)]. Velvetbean caterpillar (Anticarsia gemmatalis

Hubner) has also been found feeding on aeschynomene (Moscardi, 1979),

but these larvae are not often a severe problem as they succumb to a

fungal disease [Nomuraea rileyi (Farlow) Samson] in September and

October. Kretschmer and Bullock (1980) also observed moderate damage

due to a microlepidopteran leaf binder. A major pest to legume

seedlings in grass sods is the snail (Kalmbacher et al., 1979).

Burning the sod after dessication with paraquat was found to kill 98%

of the snail population.

Rhoades (1980) and Kretschmer et al. (1980) indicated that

Florida common aeschynomene was resistant to southern rootknot

nematode. Pasley (1981) found both susceptible and resistant ecotypes

in A. americana, and the author cautioned that all sources of Florida

common aeschynomene may not be supplying resistant material.


Effects of Defoliation on Seed Production


Limpograss


As described earlier in the review, seed set on limpograss lines

is quite low (Schank, 1972), and propagation in Florida has depended

upon vegetative plantings.


Aeschynomene


Aeschynomene is capable of producing large quantities of seed

(Hcdges et al., 1982; Moore and Hilmon, 1969). This is a potentially

important trait for an annual, but Humphreys (1979) has indicated that












the effects of grazing and mowing on flowering and seed production of

tropical pasture species are not well understood.

Loch and Humphreys (1970) reported that defoliation of

Stylosanthes humilis H.B.K. at floral initiation, flower appearance,

or advanced flowering sharply reduced seed production. The main

effect of defoliation was to reduce the proportion of florets setting

seed. This happened because flowering time was delayed, and

inflorescences from defoliated plants matured under conditions of

decreased temperature and radiation. In another experiment, seed

production of Stylosanthes guyanensis was reduced if grazing continued

up to the time of floral initiation (Loch et al., 1976). There were

no differences in seed production, however, between an undefoliated

control and plants that were defoliated during vegetative growth.

Wilaipon and Humphreys (1976) reported a 61% increase in seed yield,

over that observed in an undefoliated control, when Stylosanthes

hamata (L.) Taub. was defoliated at early flowering. In this

situation, the legume was grown in association with a tall grass, and

grazing decreased grass competition. Increased branching and more bud

sites for inflorescence development were also attributed to

defoliation.

Gildersleeve (1982) evaluated aeschynomene seed production

response to plant density and to plant height when grazing was

initiated. Higher plant density resulted in increased seed yield, but

seed yield decreased as initiation of grazing was delayed. She

concluded that early grazing removed apical dominance, encouraged

axillary regrowth, and provided more sites for inflorescence

development.








31


Pitman and Kalmbacher (1983) evaluated aeschynomene seed

production under grazing. Seed yield was 27.7, 17.7, and 12.8 g m2

when grazing was terminated on 2 September (vegetative growth), 8

October (first flower), or 4 November (cessation of plant growth),

respectively.
















CHAPTER III
FACTORS AFFECTING THE ESTABLISHMENT OF
AESCHYNOMENE IN FLORALTA LIMPOGRASS SODS


Introduction


Many of the tropical grasses adapted to Florida are low in forage

quality during most of the growing season, and performance of grazing

animals may be limited as a result. Highly productive and better

quality forage species and associations must be developed in order to

increase animal production.

Limpograss [Hemarthria altissima (Poir.) Stapf et C. E. Hubb.]

has shown potential for use in Florida as a forage. A persistent and

productive limpograss variety with higher in vitro organic matter

digestibility (IVOMD) than many tropical grasses at comparable growth

stages was recently released as 'Floralta' (Quesenberry et al., 1984).

Carrying capacity (animal days ha-1) of Floralta pastures has

consistently been greater than that of bahiagrass (Paspalum notatum

Flugge) pastures, but average daily gain was similar for the two

grasses (Quesenberry et al., 1984). There is evidence to suggest that

low N concentration in limpograss forage may be the major factor

limiting animal gains (Euclides, 1985).

To address the problem of low N content, forage legumes have been

evaluated in association with limpograss (Gomes, 1978). A legume that

has shown potential for use with Floralta is aeschynomene

(Aeschynomene americana L.). When well managed, aeschynomene can












provide a high quality forage for the grazing animal (Gildersleeve,

1982; Sollenberger et al., 1985), and like limpograss, aeschynomene is

well adapted to poorly drained soils (Albrecht et al., 1981; Miller

and Williams, 1981).

Aeschynomene has been successfully established in bahiagrass sods

with the aid of herbicides, burning, and grazing for control of grass

competition (Kalmbacher et al., 1978; Kalmbacher and Martin, 1983).

Aeschynomene is typically seeded after the onset of summer rains in

Florida (Hodges et al., 1982), but this limits the productive period

of the legume. Development of management strategies to aid legume

establishment in limpograss sods is of importance, and information

regarding the potential of spring seeding of aeschynomene is also

needed.

Experiments were conducted in 1983 and 1984 to evaluate

aeschynomene-Floralta limpograss associations. The objectives of this

research were to compare the effects of seeding rates, seeding dates,

types of seedbed preparation, and seeding methods upon legume

establishment, legume and total dry matter (DM) and N production, and

sward botanical composition.


Materials and Methods


The research was carried out during 1983 and 1984 at the

University of Florida's Beef Research Unit located northeast of

Gainesville, Florida. The site was a 10- by 50-m block of Floralta

that had been established in 1978. The soils were of the Wachula

series (sandy, siliceous, hyperthermic Ultic Haplaquod), a poorly











drained flatwoods type. Prior to initiating the experiment in 1983,

soil pH at the site was 5.6, and Mehlich I extractable P and K levels

were 9 and 40 mg kg- respectively. Guided by soil test results, P

and K were broadcast applied at rates of 25 and 90 kg ha-1 in 1983 and

45 and 165 kg ha-1 in 1984.

In 1983, four establishment methods and two legume seeding rates

were arranged as a complete factorial set of treatments in six

replications of a randomized block design. The methods of

establishment were 1) no disturbance of the grass sod and broadcast

seeding of the legume (broadcast), 2) light disking, broadcast

seeding, and cultipacking (disk), 3) application of 1,1'-dimethyl-

4,4'-bipyridinium ion (paraquat) at 0.56 kg ha-1 (active ingredient)

followed by broadcast seeding (herbicide), or 4) seeding with a Powr-

till seeder seederr). Seeding rates were 7 or 14 kg ha-1 of dehulled

seed that had been inoculated with rhizobia from the cowpea group.

Plot size was 2 by 4.5 m.

Paraquat was applied to appropriate plots on 2 June 1983.

Tillage and seeding operations were carried out the following day.

Grass height at seeding was approximately 15 cm. Eighty millimeters

of rain fell within 7 days of seeding, and mean daily high and low

temperatures during the period were 30 and 180C, respectively. Stand

counts of legume seedlings were made 10, 20, 30, and 60 days after

seeding. Four, 20- by 30-cm areas were permanently marked in each

plot, allowing the same quadrats to be counted at each date.

A single harvest of all plots was taken on 22 Sept. 1983. A 1-

by 4-m area was cut to an 8-cm stubble height using a Carter flail-












type harvester. The cut forage was weighed wet, and a 1-kg subsample

was dried at 600C to determine percent dry matter. Botanical

composition of harvested DM was determined by hand separation of

freshly harvested herbage. The sample for separation was obtained

using a sickle-bar mower to cut a 5- to 8-cm strip along the harvested

portion of each plot. Separated forage was dried at 600C, weighed,

and ground to pass a 1-mm screen.

Legume seeding rates and only the broadcast and disk

establishment methods used in 1983 were evaluated again in 1984. Four

seeding dates, 1 April, 20 April, 10 May, and 20 June were also

compared. The design used was a split-plot with seeding dates as the

whole plot treatment, and all treatments were replicated three times.

Soil moisture and temperature at the soil surface (to 2 cm) were

determined three times a week for 3 weeks following each seeding date

in 1984. Two-gram soil samples were weighed wet then dried overnight

at 1050C to determine percent moisture. Soil temperature was measured

with a mercury-bulb thermometer. Although early morning

determinations may have been more useful, logistical constraints

required that sampling and temperature readings be taken between 1300

and 1400 h. Values reported are seeding date means over methods,

rates, and replications.

Legume seedling counts were made at 10, 20, and 30 days after

each seeding date according to the procedure described for 1983. All

plots were harvested on 12 July and 3 Oct. 1984. A 1- by 4-m area was

cut to a 12- (12 July) or 8-cm (3 October) stubble using a sickle-bar

mower. Cut forage was collected and weighed immediately. Subsampling











for percent DM determinations and methods for determining botanical

composition were the same as in 1983.

After grinding, legume and grass samples were analyzed for N.

Forages were digested using a modification of the aluminum block

digestion procedure of Gallaher et al. (1975). Sample weight was

0.3 g, catalyst used was 3.2 g of 9:1 K2SO4:CuS04, and digestion was

conducted for 4 h at 4000C using 10 ml H2S04 and 2 ml H202. Ammonia

in the digestate was determined by semiautomated colorimetry

(Hambleton, 1977). Crude protein (CP) was calculated as N X 6.25.


Results


Germination and Seedling Development: 1983


Legume germination and emergence occurred within 72 hours

following seeding in 1983. Legume seedling number was not affected by

establishment method at any of the count dates (P > 0.15), but there

were approximately twice as many seedlings for the 14 compared to the

7 kg ha-1 legume seeding rate on all dates (P < 0.01, Table A-I).

Plant populations declined over time for both rates. For example,

seedling number for the high rate decreased from an average of

120 m-2, 10 days after seeding, to 90 m-2 at the 60-day count.


Dry Matter and N Harvested: 1983


Rainfall was well distributed throughout the 1983 growing season

(Table 3.1) and aeschynomene was approximately 1.5 m tall by harvest

date. There was no establishment method x seeding rate interaction

(P 0.28) for any response variable, so main effects are considered














Table 3.1. Rainfall data for 1983 and 1984 recorded at the Beef
Research Unit, northeast of Gainesville, Florida.


70-yeart
Month 1983 1984 mean

mm

Jan. 82 35 72

Feb. 112 107 94

Mar. 182 92 108

Apr. 163 82 77

May 80 122 90

June 141 95 173

July 94 258 204

Aug. 194 62 210

Sept. 190 67 144

Oct. 61 29 93

Nov. 109 81 49

Dec. 192 16 74

Total 1600 1046 1388


tSeventy-year means are for Gainesville as
of Florida Agronomy Department.


recorded by the University












directly. Total DM harvested was affected by method of establishment

(P < 0.01) but not by legume seeding rate (P = 0.30). Broadcast and

seeder treatments maximized grass and total DM harvested, and disk was

superior to herbicide (Table 3.2). Grass DM harvested was not

affected by legume seeding rate (P = 0.53). Legume DM harvested and

percentage legume in total DM were affected by establishment method

(P < 0.01), with the herbicide treatment resulting in maximum legume

performance (Table 3.2). There was a strong trend favoring the 14

kg ha- seeding rate (P = 0.08), as legume DM harvested was 0.7 Mg

ha-1 higher (5.8 vs. 5.1 Mg DM ha-) than that observed with a 7

kg ha-I rate.

Total N harvested (Table 3.3) was highest using the herbicide

treatment (P < 0.05), and was positively correlated with legume DM

harvested (r = 0.94), but not correlated with total DM harvested

(r = -0.06). Mean legume CP concentration was 94 g kg- DM and was

not affected by method of establishment. Aeschynomene forage from the

7 kg ha-1 seeding rate treatment was higher in CP than that from the

14 kg ha-1 rate (P = 0.02), but the 6 g kg-1 DM (97 vs. 91 g CP kg-1

DM) difference may be of little biological significance. Limpograss

CP concentration was very low (Table 3.3), and it was affected by

establishment method (P < 0.01). There was a strong positive

correlation between grass CP concentration and percentage legume in

total DM harvested (r = 0.83).














Table 3.2. Dry matter harvested and botanical composition of
aeschynomene-limpograss associations as affected by
aeschynomene establishment method (1983).


1983 DM harvested Legume
Method Total Grass Legume in total DM

Mg ha-1

Seeder 12.9 at 9.2 a 3.7 c 28 c

Broadcast 12.7 a 9.0 a 3.7 c 30 c

Disk 11.3 b 4.8 b 6.5 b 58 b

Herbicide 9.8 c 1.8 c 8.0 a 82 a

SEO 0.36 0.44 0.36 2.9


TMeans within columns followed by the same letter are not
significantly different (Duncan's multiple range test, P = 0.01).
tStandard error of a treatment mean.














Table 3.3. The effect of aeschynomene establishment method on N
harvested in total and legume herbage from aeschynomene-
limpograss associations and on limpograss crude protein
(CP) concentration (1983).


Nitrogen harvested
Method Total Legume Grass CP

kg ha g kg-1 DM

Seeder 87 ct 57 c 20 c

Broadcast 84 c 55 c 20 c

Disk 115 b 96 b 25 b

Herbicide 135 a 125 a 37 a

SEt 5.3 5.8 1.2


tMeans within columns followed by the same letter are not
significantly different (Duncan's multiple range test, P = 0.05).
tStandard error of a treatment mean.












Germination and Seedling Development: 1984


In 1984, seeding at four different dates provided a wide range of

soil and climatic conditions for evaluation of aeschynomene

establishment (Table 3.4). Rainfall totaled 60 mm in the first 4 days

after the 1 April seeding, but nighttime temperatures were as low as

40C. Little rain fell following the 20 April seeding, but soil

moisture levels remained adequate for germination and emergence if the

seed had been incorporated into the soil by disking and cultipacking.

By 10 May, soil moisture was severely depleted and few seeds

germinated before rains on 23 May. Both temperature and rainfall were

favorable for the 20 June seeding, and seedlings emerged within 2 days

of planting. Limpograss was 4, 6, 10, and 25 cm tall on 1 April, 20

April, 10 May, and 20 June, respectively, so responses to seeding date

include the effects of differing levels of grass competition.

Legume seedling number was affected by date, rate, and method

main effects at all stand count dates (P < 0.01), and interactions

were numerous. Ten days after seeding there were more aeschynomene

seedlings in disked plots than in the broadcast plots (P < 0.05) for

the April seedings only (Table A-2). At 20 and 30 days after seeding,

the disk treatment was superior over all dates (P < 0.01). There was

a date x rate interaction for all three stand counts (P < 0.01). The

14 kg ha- rate generally had more seedlings than did the 7 kg ha-1

rate, but differences were most pronounced for the 20- and 30-day

counts of the last two seeding dates (Table A-3). The number of

seedlings that emerged tended to increase as date of seeding was









42




Table 3.4. Soil and climatic conditions for the first 20 days following
seeding of aeschynomene in limpograss sods (1984).


Days 1-10 Days 11-20
Seeding Initial Mean Rain- Mean Mean Mean Rain- Mean Mean
date SWCt SWC fall ST AT SWC fall ST AT

g kg-1 g kg- mm C C g kg- mm C C

1 Apr. 210 220 70 23 23-7 200 5 24 23-7

20 Apr. 170 150 7 29 28-8 120 11 31 30-13

10 May 90 90 0 31 29-10 180 111 30 30-15

20 June 180 170 93 30 33-15 140 68 33 33-16


tSWC = soil water content (0 to 2 cm), ST = soil temperature (2-cm
depth), AT = daily high and low air temperatures.












delayed, and emergence was greatest for the 20 June seeding

(P < 0.05).


Dry Matter and N Harvested: 1984


Total and legume DM harvested were less in 1984 than in 1983 due

to the two-cut system and to dry weather during August and September.

There were date x method interactions for both legume and total DM

responses in 1984 (P < 0.01). Aeschynomene DM harvested was greater

with the disk treatment than with the broadcast treatment for seeding

dates 1 April, 20 April, and 20 June (Table 3.5). Total DM harvested

was greater using the disk treatment for the two April seedings, but

for the 10 May and 20 June seedings there were no differences due to

establishment method. Legume DM harvested was greatest for the 20

April seeding when comparing date means within the disk method

(P < 0.05). There was a similar trend for the broadcast treatment, but

seeding date responses were not significantly different. With the

disk treatment, total DM harvested was greatest for the April seedings

(P < 0.05), but with the broadcast treatment there was no date effect.

Legume DM harvested in 1984 was 0.5 Mg ha-1 greater with the 14 as

opposed to the 7 kg ha-1 seeding rate (P < 0.01, Table 3.6). Total DM

harvested followed a similar pattern (P < 0.01), but limpograss DM

harvested was not affected by seeding rate (P = 0.26).

Total N harvested in 1984 was correlated with both aeschynomene

(r = 0.98) and total DM harvested (r = 0.84). There was a method x

date interaction for total N harvested (P < 0.01), and specific

responses closely paralleled those of total DM harvested. Increasing














Table 3.5. Legume seeding date x establishment method interaction means
for total and legume dry matter (DM) harvested in
aeschynomene-limpograss associations (1984).


Total DM harvested Legume DM harvested
Seeding Method Method
date Disk Broadcast Disk Broadcast

Mg ha-1

1 Apr. 6.8 at 4.8 b 1.9 a 0.5 b

20 Apr. 7.5 a 5.1 b 3.9 a 1.0 b

10 May 5.3 a 4.9 a 1.1 a 0.6 a

20 June 5.6 a 5.3 a 1.1 a 0.3 b

SE$ 0.17 0.11


by the same letter


tEstablishment method means within dates followed
are not significantly different (LSD, P = 0.05).
tStandard error of an establishment method mean.









45




Table 3.6. Dry matter and N-harvested responses to legume seeding rate
in aeschynomene-limpograss associations (1984).


Seeding DM harvested Nitrogen harvested Grass
rate Total Grass Legume Total Grass Legume CP

kg ha-1 Mg ha-1 kg ha- g kg-1 DM

7 5.3 4.2 1.1 40 16 24 24

14 6.1 4.5 1.6 54 19 35 26

F test NS *

SE$ 0.17 0.14 0.11 3.3 0.7 3.1 0.5


tNS = not significant, and ** indicate probability levels of P < 0.05
and P < 0.01.
tStandard error of a treatment mean.











aeschynomene seeding rate from 7 to 14 kg ha1 increased total (P <

0.01), legume (P = 0.02), and grass N harvested (P = 0.02, Table 3.6).

Limpograss CP concentration was not affected by seeding rate at the 12

July harvest (P = 0.26), but by 3 October grass CP was 2 g kg-1 DM

higher in plots where the legume was seeded at 14 kg ha-1 (P = 0.05).

As in 1983, grass CP concentration was positively correlated with

percentage legume in total forage DM (r = 0.72). Grass from disked

plots was higher in CP concentration than grass from the broadcast

plots (P = 0.03). Legume CP concentration averaged 140 g kg- DM

using the two-cut system in 1984. This was nearly 50 g kg- DM

greater than in the very mature, woody plants harvested in September

1983, and more closely represents concentrations observed under

grazing (Chapter V).


Discussion


Germination and Seedling Development


Oversowing of legumes into grass pastures, either by sod-seeding

or surface broadcasting, has the potential to improve forage quality

while minimizing pasture renovation costs (Cook, 1980). In the

current study, favorable rainfall and temperature conditions after

seeding on 3 June 1983 resulted in no differences in legume

establishment between four oversowing methods. These results are

similar to those reported by Gomes (1978) for an aeschynomene-Bigalta

association. By contrast, seeding occurred at four dates in 1984, and

under the influence of a wide range of climatic and soil conditions,












incorporating seeds into the soil (disk method) was generally superior

to broadcasting seeds on the soil surface.

Some factors that may have limited germination and seedling

development at the respective 1984 seeding dates can be proposed. The

combination of low nighttime temperatures and wet soils following the

1 April seeding may have been detrimental. No reports of cold

temperature effects on aeschynomene establishment have been found in

the literature. Minimal rainfall following seeding on 20 April

rendered the broadcast treatment ineffective, and emphasized the

importance of incorporating the seed into a moist soil if rainfall is

sporadic or unpredictable. Almost no germination occurred after the

10 May seeding until rain fell nearly 2 weeks later. Some seeds that

were broadcast on the soil surface may have imbibed water from dew on

litter and grass and initiated germination, only to desiccate and die

when no rain fell. Conditions were excellent on 20 June and

aeschynomene seedlings emerged quickly. However, limpograss was 25 cm

tall by this date, and grass competition limited legume establishment.

Similar responses have been reported for Townsville stylo

(Stylosanthes humilis H.B.K.) over-seeded in tall perennial grass

pastures (Gillard, 1977).


Dry Matter and N Harvested


Results from 1983 indicate that aeschynomene production was

strongly related to the degree of limpograss canopy reduction. The

grass was very slow to initiate new growth after treatment with

paraquat, thus allowing the legume 4 to 5 weeks of growth without












significant above-ground grass competition. Disking opened up the

canopy but did not curb grass competition to the extent that the

herbicide treatment did. The sod seeder had little effect on grass

competition, and broadcast seeding had none. Legume DM responses were

comparable to those reported by Kalmbacher et al. (1978). They

observed that herbicidal suppression of bahiagrass enhanced

establishment and productivity of three summer-annual legumes, but

high rates of some compounds reduced bahiagrass stands. In further

work with an aeschynomene-bahiagrass association, Kalmbacher and

Martin (1983) found that a paraquat plus burn treatment prior to

seeding, or a grazing treatment after seeding the legume, controlled

grass competition and allowed the legume to become established. Total

DM and grass DM harvested were inversely related to the amount of sod

disturbance in the current study, and similar results have been

reported (Kalmbacher and Martin, 1983). Increasing legume seeding

rate tended to increase DM harvested in 1983, but it had no effect on

total or limpograss DM harvested.

Nitrogen harvested in 1983 was closely related to legume DM

harvested but was not related to total DM produced. The lack of

relationship between total DM and total N harvested is unusual with

forages in general, but it illustrates the effect of very low

limpograss CP concentrations. Grass CP concentration was highly

correlated with percentage legume in sward DM, but these results must

be interpreted with caution. Highest percentage legume occurred in

paraquat-treated swards where the grass was forced to regenerate new

shoots. The harvested portion of the grass forage was mainly this











younger growth which likely would be higher in CP. Also, disking may

have encouraged mineralization of soil N, making more available to the

N-deficient grass. Some N may have been made available to the grass

by the legume, but other factors probably were involved.

Aeschynomene typically is seeded after the onset of summer

convectional rains in Florida (Hodges et al., 1982), but this practice

severely limits the productive period of the legume. The 1984

experiment provides evidence that earlier seeding may be possible in

some years. Total DM harvested was maximized with April seeding

dates, and legume contribution was greatest when seeded on 20 April.

Earlier seedings were more successful in 1984 because soil moisture

levels were higher in April than for the May seeding, and because

grass competition was minimal in April compared to the level at the

June seeding. Clipping the grass to a 5-cm stubble before each

seeding date may have been a better way to isolate the effects of soil

and climatic conditions.

The disk and broadcast methods were selected for additional

evaluation in 1984 because they require less specialized equipment or

are less costly than the sod seeder or herbicide methods. As a result

they may be more attractive options for pasture improvement in

Florida. As in 1983, the disk treatment was superior to the broadcast

treatment in 1984. Differences due to method were more closely

related to effects on establishment in 1984 than they were in 1983,

when later-season effects of grass competition may have been more

critical. Apparent benefits of the disk-cultipack treatment include

better soil-seed contact, some control of grass competition, and more











light penetration to the base of the canopy. Broadcasting dehulled

aeschynomene seed on the soil surface has been used successfully

(Chapter IV), but adequate rainfall after seeding and control of grass

competition by grazing were necessary.

In contrast to 1983, legume seeding rate was important in 1984,

probably due to less favorable conditions at establishment. The

higher seeding rate increased total and legume DM harvested, and

total, limpograss, and aeschynomene N harvested. These results are in

contrast with work done on prepared seedbeds in Australia. Middleton

(1970) reported that increasing the rate of sowing of one component of

a grass-legume mixture raised its contribution to DM and N harvested,

but decreased that of its associated species. In Pennsylvania, higher

legume seeding rates increased legume contribution, grass DM

harvested, and grass CP concentration (Sollenberger et al., 1984).

From this and other experiments, several major factors affecting

aeschynomene establishment are indicated. These include soil moisture

at seeding, rainfall in the first 10 days after seeding (Kalmbacher

and Martin, 1983), soil temperature, light penetration to the base of

the canopy (Kalmbacher and Martin, 1983), and control of grass

competition until the legume seedling is 3 to 5 cm tall (Hodges et

al., 1982; Kalmbacher and Martin, 1983; Sollenberger et al., 1985).

High legume seeding rates appear to be beneficial, particularly when

establishment conditions are unfavorable. A large number of methods

can be used to establish aeschynomene, but there is less risk involved

when the seed is incorporated into a warm, moist soil, and when the

method utilized offers extended control of grass competition.















CHAPTER IV
EFFECTS OF GRAZING MANAGEMENT ON PRODUCTIVITY
AND BOTANICAL COMPOSITION OF AN
Aeschynomene americana-Hemarthria altissima ASSOCIATION


Introduction


Limpograss [Hemarthria altissima (Poir.) Stapf et C.E. Hubb.] is

a highly productive tropical grass with potential for use in the lower

Southeast (Quesenberry et al., 1978). Digestibility (IVOMD) and

organic matter intake of limpograss were generally superior to that of

bahiagrass (Paspalum notatum Flugge), the predominant pasture species

in Florida (Carvalho, 1976; Moore et al., 1981), but N concentration

in limpograss forage may be very low (Euclides, 1985).

Generally, the most economical alternative for overcoming a

protein deficiency in pastures is to include a legume (Minson, 1980).

Aeschynomene (Aeschynomene americana L.) is the most widely adapted

warm-season legume available for grazing in Florida (Hodges et al.,

1982). Kalmbacher et al. (1978) reported that herbicidal suppression

of bahiagrass sods facilitated aeschynomene establishment and

increased total dry matter (DM) harvested and crude protein

concentration of the association. Excellent aeschynomene stands have

also been obtained by grazing bahiagrass to a 3- to 5-cm stubble,

drilling legume seed into the sod, and occasionally grazing the area

until aeschynomene seedlings were 2.5 cm tall (Kalmbacher and Martin,

1983).












Mislevy et al. (1981) reported highest seasonal DM harvests of

aeschynomene if the legume was cut initially at a height of 30 cm,

with the regrowth cut at 90 cm. Lowest production occurred when the

legume was clipped at 30 cm throughout the season. Under grazing,

aeschynomene was most productive if defoliated first at 28-, instead

of 45- or 54-cm heights (Gildersleeve, 1982). Early grazing was

credited with opening up the canopy for better light penetration and

stimulating axillary bud development and secondary branching.

Establishment of aeschynomene into limpograss sods and subsequent

management of the association have not been evaluated under grazing

conditions. Since both species are well adapted to the large expanses

of poorly drained flatwoods soils in Florida, this type of information

is needed. The objectives of this research were 1) to evaluate the

effects of spring grazing management of limpograss on aeschynomene

establishment and subsequent productivity, and 2) to compare the

effects of summer grazing management of the association on legume and

total DM accumulated and sward botanical composition.


Materials and Methods


An experiment was conducted in 1983 and 1984 at the University of

Florida's Forage Evaluation Field Laboratory, located northeast of

Gainesville, Florida (lat 290 60' N). The research site was a 2-ha

'Floralta' limpograss pasture that had been established in 1981.

Soils were of the Pomona series (sandy, siliceous, hyperthermic Ultic

Haplaquod). Prior to initiating the experiment in 1983, soil pH at

the site was 5.6, and Mehlich I extractable P and K levels were 4 and












18 mg kg-l, respectively. In 1984, soil pH was 5.6, and P and K

levels were 6 and 34 mg kg-1. Guided by soil test results, P and K

were broadcast applied at rates of 44 and 166 kg ha-1 in 1983, and 70

and 133 kg ha-1 in 1984. Twenty kilograms ha-1 of a micronutrient

mixture (F-503 Oxide) were included with the P and K in both years.

Lime was applied at a rate of 2.2 Mg ha-1 in April 1984 because soil

pH was nearly 5.0 in some areas of the pasture.

Experimental variables included 1) limpograss stubble height when

aeschynomene was overseeded (SH), 2) legume seedling growth stage when

early-season grazing was discontinued (LSGS), and 3) legume height

when summer grazing was initiated. Early-season limpograss regrowth

was grazed to a 7.5- or 15-cm stubble. At the onset of summer rains

(7 June 1983 and 20 June 1984) dehulled, scarified, and inoculated

aeschynomene seed was broadcast at a rate of 20 kg ha-1 using a

cyclone seeder. Following overseeding, cattle were rotated among the

pastures to maintain the prescribed grass stubble heights. Grazing

was discontinued 1) when legume cotyledons were exserted, 2) when two-

true leaves were present, or 3) 2 weeks after the two-leaf stage.

Pastures were not grazed subsequently until the legume reached heights

of 20, 40, or 80 cm in 1983 and 20, 40, or 60 cm in 1984. Following

initiation of grazing at the respective heights, all pastures were

grazed rotationally with a rest period of 5 weeks. Defoliation was by

mob grazing to an 8- to 12-cm stubble. Exceptions were pastures where

grazing was not initiated until the legume was 60 to 80 cm tall. The

yearling and 2-year-old steers (Bos spp.) would not graze the mature












herbage to a low stubble, so a more subjective visual appraisal of the

pasture was made to determine when grazing should end.

The complete factorial set of treatments (2 x 3 x 3) was arranged

in a split-split plot design in 1983 and a split-plot design in 1984.

Grass SH was the whole-plot treatment in both years, and LSGS was the

subplot treatment in 1983. The design was changed in 1984 to provide

a stronger test for LSGS differences, despite the original design

being more convenient for animal management. Legume seedling growth

stage means were compared using the single degree of freedom

contrasts, cotyledon vs. two-leaf stage, and copyledon plus two-leaf

stage vs. 2 weeks after the two-leaf stage initiation height effects

were evaluated using orthogonal-polynomials to determine the nature of

response curves. Percentage legume in total herbage accumulated and

in total herbage consumed was transformed (square root) prior to

analysis of variance (Steel and Torrie, 1960). All treatments were

replicated twice in each year, and pasture size was 500 m2

Stand counts were taken at 10, 20- by 20-cm locations in each

pasture when legume seedlings reached the cotyledon stage (3 days

after seeding), the two-leaf stage (10 days after seeding), and the

two leaf plus 2-week stage (24 days after seeding). In 1984 the

cotyledon count was delayed until 5 days after seeding due to

logistical constraints. The same areas were counted at each growth

stage, so that population trends could be monitored.

In 1983, pastures defoliated initially when the legume was 20,

40, and 80 cm were grazed four, three, and two times (cycles),

respectively. In 1984, both 20- and 40-cm treatments were grazed












three times, and the 60-cm treatment was grazed twice. Grazing

seasons for the legume-grass association extended from 11 July through

4 Nov. 1983 and 23 July through 2 Nov. 1984.

All pastures were sampled before and after each grazing period.

Response variables measured include herbage mass, percentage

aeschynomene, percentage limpograss, and percentage weeds. A double

sampling technique was used that combined visual estimates made by a

single observer, and actual values determined by cutting forage to

ground level and hand separating the fresh herbage (Stockdale, 1984).

Five double sampling sites were selected in each pasture before

and after grazing. The 0.5-m2 sites were not chosen at random, but

were chosen to represent the range of the response variables present

in a given pasture. Visual estimates of the four response variables

were recorded at each site. The forage was then clipped, hand

separated into aeschynomene, limpograss, and weed fractions, dried at

600C for 48 hours, and weighed. Actual values were calculated and

regression equations fit for actual values vs. visual estimates

(regression was through the origin). The slope of the regression line

was used to correct the mean of 30 visual estimates that were taken at

randomly selected locations in each pasture. This corrected mean is

the reported value for each response variable.

Generally, data from all pregraze or all postgraze sites of a

given week (30 to 40 double samples) were combined to form a

regression equation. Occasionally, higher r2 values and lower

standard errors of the estimate were obtained if the pastures grazed

in 1 week were divided into two or more groups and separate equations












fit. Similarly, there were times when single equations were used for

2 or more weeks of sampling.

As recommended by Hodgson (1979), terms such as herbage yield and

herbage production are avoided in this report. Suggested terms that

are used include 1) herbage mass, the instantaneous measure of total

weight of herbage per unit area of ground, 2) herbage accumulation,

the change in herbage mass between successive measurements and summed

over time where appropriate, 3) herbage consumed, the mass of herbage

per unit area removed by grazing animals at a single grazing or series

of grazings, and 4) efficiency of grazing, the herbage consumed

expressed as a proportion of herbage accumulation (Hodgson, 1979).

Botanical composition data, unless otherwise specified, are expressed

as a percentage of herbage DM accumulated. When botanical composition

is expressed on a yearly basis, it was calculated as the summation of

component DM accumulated (over grazing cycles) x 100 and divided by

the summation of total DM accumulated. As grazing periods were

generally 48 hours, plant growth during grazing was not accounted for

('t Mannetje, 1978).


Results


Aeschynomene Seedling Counts


Changes in legume seedling populations throughout the

establishment period were used to draw conclusions about the effects

of SH and LSGS on aeschynomene seedling survival. There was no effect

of SH on seedling emergence (P = 0.13 in 1983, and P = 0.80 in 1984),

or seedling survival (P > 0.55) in either year. In contrast, LSGS did












affect seedling survival in both years (Table 4.1). In general,

continuing to graze pastures beyond the cotyledon stage resulted in

greater losses of aeschynomene seedlings than that observed in

pastures where grazing ended at cotyledon exsertion.


Total DM Accumulation


Measurement of DM accumulation started when summer grazing of the

grass-legume associations was initiated. As a result, treatments

which minimized grazing during legume establishment tended to have

higher seasonal accumulation of herbage DM. The 15-cm SH treatment

accumulated 9.59 Mg DM ha-1 in 1983 compared to 8.15 Mg ha-1 for the

7.5-cm treatment (P = 0.02). In 1984, the same treatments accumulated
-1
8.10 and 6.62 Mg of herbage ha-l,respectively (P = 0.16). If grazing

ended early during the legume establishment period, i.e., at the

cotyledon or two-leaf stage, DM accumulation was greater than that

observed if grazing continued for 2 weeks after the two-leaf stage (P

= 0.05 in 1983, and P < 0.01 in 1984). Total DM accumulation was

9.88, 8.95, and 7.79 Mg ha-1 for the cotyledon, two-leaf, and two leaf

plus 2-weeks treatments in 1983, and 7.99, 7.47, and 6.62 Mg ha-1 in

1984. There was no SH x LSGS interaction in either year (P > 0.62).

Legume height at initiation of summer grazing also influenced

annual DM accumulation (P = 0.01 in 1983 and P = 0.03 in 1984, Table

4.2). The effect of height on total DM was quite different in the

first as compared to the second grazing cycle (Table 4.2). In

general, delaying the date of first grazing increased cycle 1 DM













Table 4.1. Changes in aeschynomene seedling population as affected
by legume seedling growth stage (LSGS) when early-season
grazing ended.


1983 1984
Periodt Period
LSGS C-2L 2L-2L2 C-2L2 C-2L 2L-2L2 C-2L2

Change in seedling number m-2

Cotyledon 68t -31 37 5 3 8

Two leaf 50G -49 Ig -38G 6 -32g

Two leaf plus
2 weeks 48G -80G -32G -23G -23G -46G

SE 11 13 10 8 10 10

F test NS NS ** NS **

Contrasts S,* **,NS **,


tPeriod C-2L is the interval between the cotyledon and two-leaf stages.
Period 2L-2L2 is the interval between the two-leaf and two leaf plus
2-weeks stages, and period C-2L2 is the combined interval between the
cotyledon and two leaf plus 2-weeks stages.
tPositive and negative numbers reflect increases or decreases in
legume seedling population during a given period. Mean seedling
number at cotyledon stage was 156 and 217 m in 1983 and 1984.
Following a number, the letter G indicates treatments that were
grazed throughout a period, while g indicates those that were grazed
for part of a period. No letter following the number means no grazing
occurred during the period.
Standard error of a treatment mean.
NS = not significant, S, *, and ** indicate significance at P < 0.10,
P < 0.05, and P < 0.01, respectively.
Probability levels are given for single degree of freedom contrasts
of cotyledon vs. two-leaf stage, and cotyledon plus two-leaf stage vs.
2 weeks after the two-leaf stage.












accumulation, decreased that observed in cycle 2, but increased annual

DM accumulation.


Component DM Accumulation and Sward Canopy Botanical Composition


Limpograss SH did not affect legume DM accumulation in either

year (P = 0.82 in 1983, and P = 0.35 in 1984). Mean legume DM

accumulations in 1983 and 1984 were 1.35 and 1.30 Mg ha-i for the 7.5-

cm SH, compared to 1.18 and 1.04 Mg ha-1 for the 15-cm treatment. The

proportion of total forage accumulated that was legume DM was 15.5 and

20.7% for the 7.5-cm height, while that of the 15-cm treatment was

12.7 and 13.2% in 1983 (P = 0.74) and 1984 (P = 0.17), respectively.

In both years, percentage aeschynomene in total DM accumulated

increased with longer periods of grazing during legume establishment

(Table 4.3). Legume DM accumulation followed a similar pattern in

1984 (P < 0.01). In 1983 there was only a trend in legume DM

accumulation favoring continued grazing beyond the cotyledon stage (P

= 0.30).

Aeschynomene height when plants were first defoliated affected

annual accumulation of legume DM (Table 4.4). The response of legume

DM accumulation to initiation height was very different in cycle 1

than it was in cycle 2 (Table 4.4). This response was consistent in

both years, as plants that were first grazed at 60 or 80 cm produced

almost all of their seasonal total of aeschynomene DM in the first

grazing cycle, while those plants grazed initially at 20 or 40 cm

provided a more uniform distribution of legume DM throughout the

season.














Table 4.2. The effect of legume height at initiation of summer
grazing (HI) on cycle 1 (Cl), cycle 2 (C2), and annual(AN)
total dry matter accumulation of aeschynomene-limpograss
associations.


1983 1984
HI C1 C2 AN C1 C2 AN

--cm-- Mg ha-

20 3.24 3.77 8.57 4.39 1.48 6.85

40 5.03 2.49 8.28 5.63 1.33 7.73

80,601 9.60 0.17 9.77 7.05 0.44 7.49

SEO 0.31 0.21 0.31 0.14 0.15 0.14

F test L** L** L**,Q L* L**,Q L*,Q*


tAeschynomene HI was 80 cm in 1983 and 60 cm in 1984.
$Standard error of a treatment mean.
Linear (L) or quadratic (Q) effects with probability of P $ 0.01 (**),
P < 0.05 (*), or P < 0.10 (letter listed, but not followed by a
symbol).














Table 4.3. The effect of legume seedling growth stage (LSGS) when
early-season grazing ended on aeschynomene dry matter (DM)
accumulation and percentage legume in aeschynomene-
limpograss associations.


Legume DM accumulation Percentage legume in sward DM
LSGS 1983 1984 1983 1984

Mg ha-1 -

Cotyledon 0.83 0.96 8.6 12.1

Two leaf 1.54 1.10 16.4 15.6

Two leaf
plus 2 weeks 1.43 1.46 17.4 23.2

SEt 0.30 0.09

F test NS ** S **

Contrasts -- NS,** S,NS S,**


tStandard error of a treatment mean. Standard errors are not
presented for percentage data, as analysis of variance was performed
on transformed values (square root), and the means presented have
been transformed back to the original scale.
NS = not significant, S and ** indicate probability levels of P < 0.10
and P < 0.01.
tSingle degree of freedom contrasts of cotyledon vs. two-leaf stage,
and cotyledon plus two-leaf stage vs. 2 weeks after the two-leaf stage.























-*^ -^ 0




o c^- o^
Q -4 0
ini N- 01
ouo

ri~o


u U


















>0
-




H U


N








0f
Cn



h 0H

0 u




C n












-'-- 0
Scn





















0 aN
O C c
N ri














6U =










EN-
U N





ci 0
4-i CM i

N N


N^N

cnl




Hu


-T -'t
O 0

O !




m3 c

cn o








in o
M O


m o


0 0 CN 0 -K
1-J


in -01
0m
0 0
o O


00 0 c O
-K
0 0 t4 O O -K
.4


4-

C
0
-3

0c


4-+ m
rJ 0
w cr
N-i-


0 r1














N LCN
C in
0 0


n in


I r u-i n


O

-/
rl

O






Vi







a





*
*KU
p-

















>1
0 H cH




m

0
o ri-i










p C
CO




00 -
4 -i














c e
ON












S 4- 0




W U
w N0














w Ul
- 0 -

UC O-'









o an






0 .

















4 c.c
S++

O UN


0 tl 0












In both years, cycle 1 limpograss DM accumulation increased

linearly as initiation of grazing was delayed, but annual DM

accumulation was not affected by initiation height (P = 0.48 in 1983,

and P = 0.16 in 1984, Table 4.4). Grass DM accumulation for the 1983

season was not affected by LSGS (P = 0.14) or by SH (P = 0.34). In

1984, there was no response of limpograss DM accumulated to SH (P =

0.33), but LSGS effects were important (P < 0.01). Pastures grazed

through the cotyledon, two-leaf, and two leaf plus 2-weeks stages

accumulated 5.12, 4.44, and 3.43 Mg of grass DM ha-1 (for the contrast

cotyledon vs. two-leaf stage, P = 0.14, and for the contrast cotyledon

plus two-leaf stage vs. 2 weeks after two-leaf stage, P < 0.01).

Annual totals for weed DM accumulation were not affected by SH (P

= 0.63 in 1983, P = 0.84 in 1984), LSGS (P = 0.72 in 1983, and P =

0.84 in 1984), or initiation height (P = 0.12 in 1983, and P = 0.47 in

1984). Major weed species in the pastures included vaseygrass

(Paspalum urvillei Steud.) and various sedges (Cyperus spp.).


Herbage Consumed and Grazing Efficiency


Total DM consumed in 1983 was not affected by grazing management,

but trends were evident. Animals consumed 6.62 Mg DM ha-1 on 15-cm SH

pastures compared to 5.36 Mg ha-1 for the 7.5-cm treatment (P = 0.16).

Dry matter consumed tended to be higher when establishment-period

grazing ended early (P = 0.11), as totals were 6.79, 6.07, and 5.12 Mg

ha-1 for the cotyledon, two-leaf, and two leaf plus 2-weeks

treatments. There was no response of total DM consumed to initiation

height (P = 0.26). Trends were, however, quite different from those











of DM accumulation (Table 4.2). Dry matter consumption for 20-, 40-,

and 80-cm treatments was 6.36, 6.05, and 5.56 Mg ha-1, respectively.

Grazing efficiency decreased linearly (P < 0.01) as initiation height

increased. Grazing efficiency means for 20-, 40-, and 80-cm

initiation heights were 74, 66, and 61%, respectively.

In 1984, total DM consumed for 7.5- and 15-cm SH treatments was
-1
4.51 and 3.95 Mg ha-1, respectively (P = 0.42). Similar to the

response observed in 1983, total DM consumed for cotyledon, two-leaf,

and two leaf plus 2-weeks LSGS levels was 4.66, 4.18, and 3.85 Mg

ha- respectively (cotyledon > two-leaf stage, P = 0.07; cotyledon

plus two-leaf stage > 2 weeks after two-leaf stage, P = 0.02). There

were linear (P < 0.01) and quadratic (P = 0.10) effects of

aeschynomene height at first grazing on total DM consumed in 1984.

Treatment means for 20-, 40-, and 60-cm initiation heights were 4.14,

4.85, and 3.71 Mg of DM consumed ha-1, respectively. Grazing

efficiency responded to initiation height in 1984 (linear and

quadratic effects, P < 0.01), and treatment means for 20-, 40-, and

60-cm heights were 61, 63, and 49%, respectively.


Animal Selection


As well as evaluating the effects of grazing management on the

association, it was also of interest to determine if animals were

selecting for or against any of the botanical components. The

botanical composition of total herbage accumulated (HA), total herbage

mass (HM) at initiation of grazing (summed over all cycles), and total

herbage consumed (HC), was calculated for both years, and the means












over all grazing treatments were determined (Table 4.5). These data

suggest selection for aeschynomene and against weeds, but without

knowing the vertical distribution of component DM in the canopy,

conclusions regarding animal selection could not be drawn.

As part of a concurrent experiment evaluating ingestive behavior

of cattle (Moore et al., in press), 24 pastures were characterized in

1983 using a stratified clipping technique. For these pastures, forage

from the bottom of the canopy (the bottom 10 cm for most pastures, the

bottom 30 cm for pastures 60 cm tall or taller) was not included in

calculating the botanical composition of the upper layers (UL). Using

data from both experiments for these 24 pastures, regression analysis

was used to evaluate the relationship of percentage of a component

(e.g., aeschynomene) in UL vs. that in HM, percentage of the component

in HC vs. that in HM, and percentage of the component in HC vs. that in

UL (Table 4.6). It is apparent from these equations that the

relationship between UL botanical composition and HM botanical

composition was similar to the relationship between HC botanical

composition and HM botanical composition. It is apparent that

intercepts approach zero and slopes approach one for the regression of

diet botanical composition (HC) vs. upper layers botanical composition

(UL). This suggests that the botanical composition of the diet was

very similar to that of the upper layers of the sward.


Discussion


Excellent legume stands were obtained in both years, as adequate

rainfall following seeding (> 3 cm within 48 hours) insured rapid









66




Table 4.5. The percentage of total herbage accumulated (HA), total
herbage mass at initiation of grazing (HM), and total
herbage consumed (HC) that was aeschynomene, limpograss,
and weed dry matter.


Botanical 1983 1984
component HA HM HC HA HM HC



Legume 14 10 18 17 12 27

Grass 60 61 61 56 55 58

Weeds 26 29 21 27 33 15


tValues presented are means over all treatments for the grazing
seasons of 1983 and 1984.














Table 4.6. Regression equations relating the percentages of three
botanical components (aeschynomene, limpograss, and weeds)
in total herbage in the upper layers of the sward (UL), in
total herbage mass present at the beginning of a grazing
cycle (HM), and in total herbage consumed during the
grazing period (HC).


Dependent Independent 2
variable Intercept b variable r


Aeschynomene

ULt 3.26 1.52 HM 0.94

HC 2.74 1.48 HM 0.94

HC 0.43 0.93 UL 0.91

Limpograss

UL -23.78 1.28 HM 0.90

HC -24.33 1.28 HM 0.75

HC 3.25 0.94 UL 0.74

Weeds

UL -4.35 0.95 HM 0.90

HC -5.00 1.07 HM 0.58

HC 0.81 1.07 UL 0.57


tData for the equations were from 24 pastures and were means of five
sites per pasture (corrected for actual pasture means that had been
determined by double sampling).
SUpper layers include herbage from 10 cm to the top of canopies less
than 60 cm tall, and herbage from 30 cm to the top for canopies 60 cm
tall or taller.












seedling emergence and prevented loss of seedlings due to desiccation.

pGass SH did not affect aeschynomene germination, emergence, or

seedling survival during the 24 days following seeding. Kalmbacher and

Martin (1983) reported that numbers of aeschynomene seedlings emerging

and surviving for 30 days were not different in ungrazed bahiagrass (20

to 30 cm tall) and in sods that were grazed to a 5-cm stubble until

seedlings were 2.5 cm tall., Data from these experiments suggest that

Saeschynomene germination and early seedling survival is not

significantly affected by moderate levels of grass competition. Long-

term survival and subsequent productivity may suffer, however, as

evidenced by the SH and LSGS trends described below.

Limpograss SH did not significantly influence legume productivity

in either year, but there was a consistent trend over both years

favoring the 7.5-cm SH. Studies have shown that tall grass limited the

productivity of Stylosanthes humilis H.B.K. overseeded in native

pastures (Gillard, 1977). Aeschynomene was more productive if seeded

into a grazed bahiagrass sod than if seeded into bahiagrass that was 20

to 30 cm tall (Kalmbacher and Martin, 1983). Kalmbacher and Martin

(1983) hypothesized that seedling development in a low-light

environment is slower, increasing the length of time that plants are

susceptible to disease and to insect or snail feeding. Decreased

seedling root growth has also been associated with low-light

environments (Groya and Sheaffer, 1981), making plants less able to

tolerate drought or to compete effectively for nutrients. Thus, it

appears that although aeschynomene seedlings survive for some time

under taller grass, they are weakened by grass competition and












eventually lost from the stand in greater numbers than seedlings in

closely grazed grass sods. It is also likely that more extreme levels

of grass competition lead to earlier legume seedling loss, as was the

case for S. humilis in Australia (Gillard, 1977). Although data from

the current research are not conclusive, grazing limpograss to an SH of

approximately 7.5 cm may be advantageous, as even small improvements in

legume performance could be important in N-deficient grass swards.

,Data from both years indicate that LSGS was an important factor
V
affecting legume seedling survival, legume DM accumulation, and

percentage legume in pasture DM. Continued grazing after legume

cotyledon exsertion decreased legume stands, suggesting that

postemergence grazing should be strictly limited to only the amount

required for control of grass competition. Stocking of the 500 m2

pastures was approximately 2 animal days (yearling steers) week-i of

establishment period grazing. It was noted that extending the period

of grazing decreased the number of legume seedlings but increased

subsequent legume DM accumulation. Apparently, the beneficial effect

of long-term control of grass competition was greater than the short-

term effect of reduced aeschynomene stands.

/ Similar responses of aeschynomene DM accumulation to prolonged

establishment-period grazing have been reported (Hodges et al., 1982;

Kalmbacher and Martin, 1983) Results of the current study suggest an

advantage in maintaining closely grazed sods until aeschynomene

seedlings are approximately 5 cm tall. In 1983, grazing through the

two-leaf stage was adequate to maximize legume performance, but in 1984

highest aeschynomene DM accumulation occurred when grazing continued












for an additional 2 weeks. This may have been related to the

experimental area being burned prior to the 1983 season but not prior

to 1984. Grass regrowth after burning was vigorous in 1983, but the

canopy was more open at seeding than it was in 1984. With a less dense

grass sod in 1983, extending the period of early grazing past the two-

leaf stage may not have been as critical as it was in 1984.

Legume DM accumulation, legume DM consumed, and percentage legume

in sward DM were maximized by initiating grazing when aeschynomene was

60 or 80 cm tall. The taller initiation height treatments accumulated

more than 95% of their seasonal total of legume DM in cycle 1. This is

in contrast to the 20-cm treatment that accumulated 56% of its 1983

total in cycles 1 and 2, and 44% in cycles 3 and 4. Grazing

aeschynomene when it was 60 to 80 cm tall resulted in extensive

trampling damage and stand loss. UThose plants which survived had

limited capability for regrowth, probably due to shading and death of

lower axillary bud sites and to selective, and almost complete,

removal of leaf tissue by animals. t should be stressed that these

pastures were heavily grazed, and this undoubtedly compounded the

problem of survival for tall, stemmy legume plants. Fisher (1973)

reported that S. humilis withstood repeated defoliation to a 5-cm

stubble if cutting started early in the season, but delaying cutting

for 1 month resulted in 82% stand loss. 'Poor regrowth of tall

aeschynomene after clipping or grazing has frequently been cited

(Albrecht and Boote, in press; Gildersleeve, 1982; Mislevy et al., i

1981; Tang and Ruelke, 1976), and these authors have suggested that

shading and death of axillary buds on the lower portion of the stem is\












a major factor in this response. Gildersleeve (1982) observed that

plants grazed initially at 28 cm were more productive, more persistent,

and had more vigorous regrowth throughout the season than did plants

that were first grazed at 45 or 54 cm. The author suggested.that early

grazing removed apical dominance, improved light penetration to the

base of the canopy, and stimulated axillary bud development and

secondary branching.

In the current study, measurement of total DM accumulation was

limited to the summer grazing period. This provided an advantage to

pastures that were grazed to the 15-cm SH, or that were grazed just

through the cotyledon or two-leaf stages during establishment. If

grass forage consumed during establishment was accounted for in total

DM, differences due to SH and LSGS would likely be small or possibly

not significant.

Aeschynomene height at initiation of summer grazing had a marked

effect on total DM accumulation. /In both years, cycle 1 DM

accumulation increased linearly as initiation of summer grazing was

delayed, but cycle 2 DM accumulation was lowest for the taller

initiation height treatments.YSeveral factors were likely responsible

for lower productivity of the 80- and 60-cm treatments in the second

cycle.4/First, during the initial grazing period large amounts of very

mature forage were trampled, and stems, particularly of the legume,

were broken and plants killed. Dead plant matter decayed during the 5-

wek rest period, partially offsetting production of new green tissue.

ISecondly, much of limpograss regrowth and all aeschynomene regrowth is

from axillary buds. With grass stems trampled and matted on the soil












surface, regrowth potential of the sward was limited. The effect of

initiation height on legume regrowth has already been discussed.

In pasture situations, herbage accumulation is secondary in

importance to herbage consumption. In 1983, there was a trend toward

higher consumption with early initiation of grazing. In 1984,

initiation of grazing when the legume was 20 or 40 cm tall was

effective in.increasing DM consumed above that observed for the 60-cm

treatme t.' Grazing efficiency was superior at legume initiation

heights of 20 or 40 cm in both years. Better utilization of forage

also made_)asture management decisions, i.e., the end of a grazing

period, easier.

Calculations of mean botanical composition of herbage accumulated,

pregraze herbage mass, and herbage consumed, suggested that animals may

be selecting for aeschynomene. When the botanical composition of the .

upper layers of the canopy was compared to that of herbage consumed, it

seemed more likely that the vertical distribution of forage species in

the canopy, rather than animal selection for specific components, was

influencing the botanical composition of the diet. The upper layers

had a higher percentage legume than did the entire canopy (HM), but it

"was similar to that in herbage consumed. It should be noted that these

data are for entire'grazing periods, and animals may have preferred a

given component and selected for it early in a grazing period.

K/nder conditions similar to those present during this research,

aeschynomene can be successfully established into limpograss sods using

only grazing management to control grass competition. Grazing the

grass sod closely (approximately 8 cm) until legume seedlings are 5 cm









73


tall (after the two-leaf stage) appears to maximize subsequent legume

productivity. Initiating summer grazing early, when aeschynomene is 20

Sto 30 cm tall, is likely the best grazing management practice with this

association. Early grazing results in more uniform accumulation of

total and legume DM throughout the season, higher efficiency of

grazing, more vigorous legume regrowth, and a trend toward greater

total herbage consumption. 4 5-week defoliation interval seems very

appropriate for aeschynomene-limpograss pastures, but this research did

not compare it to other intervals. Animal selection between species

does not appear to be of major importance with this association, as all

components were consumed in similar proportion to that observed in the

upper layers of the canopy.
















CHAPTER V
FORAGE QUALITY RESPONSES OF AN Aeschynomene americana-
Hemarthria altissima ASSOCIATION TO GRAZING MANAGEMENT


Introduction


Forage quality evaluation of 'Bigalta' and 'Floralta' limpograss

[Hemarthria altissima (Poir.) Stapf et C. E. Hubb.] has shown that

these cultivars have higher concentrations of in vitro digestible

organic matter (IVDOM) than do most other tropical grasses at similar

growth stages (Quesenberry et al., 1981). Average daily gains of

animals grazing N-fertilized (200 kg N ha-1 yr-1) Floralta, however,

have not been superior to gains of animals grazing the less digestible

bahiagrass (Paspalum notatum Flugge) (Quesenberry et al., 1984).

Euclides (1985) sampled N-fertilized limpograss pastures and

found that mean crude protein (CP) concentration of the top 20 cm of

the canopy was 47 g kg-1 organic matter (OM). Crude protein

concentration of esophageal extrusa of grazing steers was 58 g kg-I OM

during the same 4-month period. Milford and Minson (1965) have

reported that intake of tropical grasses declines rapidly when diet CP

concentration falls below 70 g kg-1 dry matter (DM), so one possible

explanation for lower than expected animal performance on limpograss

may be a protein deficiency.

Minson (1980) has suggested that the most economical way to

overcome a protein deficiency in pastures is to include a legume.

Aeschynomene (Aeschynomene americana L.) is a summer-annual legume












that performs well on the large expanses of poorly drained soils in

Florida, and it has shown potential for use in overseeding perennial

grass pastures (Kalmbacher et al., 1978). Aeschynomene can provide

high quality forage for grazing animals, with leaf tissue CP and IVDOM

concentrations reported to be 250 g kg-1 DM and 750 g kg-1 OM,

respectively (Gildersleeve, 1982).

Grazing and clipping management play a critical role in

determining aeschynomene forage quality. Crude protein concentrations

were 185 and 151 g kg-1 DM for plants clipped initially at 30- and 90-

cm heights (Kalmbacher et al., 1978). Delaying initial harvest
-1
reduced IVDOM concentration 80 g kg OM for each 30-cm increment of

legume height (Mislevy et al., 1981). Clipping aeschynomene when it

was 30 cm tall throughout the season resulted in CP and IVDOM

concentrations of 175 g kg- DM and 700 g kg-1 OM, respectively

(Mislevy et al., 1981). In the same study, delaying the initial cut

until the legume was 90 cm, followed by one regrowth cut at 30 cm,

reduced quality to 119 g CP kg-1 DM and 525 g DOM kg-1 OM.

Most evaluations of aeschynomene forage quality have been under

clipping management. Information on forage quality responses to

grazing are needed. In addition, no data are available which indicate

whether associating aeschynomene with limpograss can overcome what

apparently is a protein deficiency in grass-N swards. The objectives

of this research were 1) to quantify the effects of grazing management

on aeschynomene leaf:stem ratio, and leaf, stem, and whole plant N and

IVDOM concentrations, 2) to evaluate the effects of grazing management

on limpograss and total sward forage quality, and 3) to determine if












overseeding of aeschynomene can eliminate a suspected protein

deficiency in limpograss pastures, and if so, to suggest an effective

and practical grazing management system for the association.


Materials and Methods


An experiment was conducted in 1983 and 1984 at the University of

Florida's Forage Evaluation Field Laboratory, located northeast of

Gainesville, Florida (lat 290 60' N). The research site was a 2-ha

Floralta limpograss pasture that had been established in 1981. Soils

were of the Pomona series (sandy, siliceous, hyperthermic Ultic

Haplaquod). Prior to initiating the experiment in 1983, soil pH at

the site was 5.6, and Mehlich I extractable P and K levels were 4 and

18 mg kg -, respectively. In 1984, mean soil pH was 5.6, and P and K

levels were 6 and 34 mg kg-1. Guided by soil test results, P and K

were broadcast applied at rates of 44 and 166 kg ha-1 in 1983, and 70

and 133 kg ha- in 1984. Twenty kilograms ha1 of a micronutrient

mixture (F-503 Oxide) were included with the P and K in both years.

Lime was applied at a rate of 2.2 Mg ha-1 in April 1984 because soil

pH was nearly 5.0 in some areas of the pasture.

Experimental variables included 1) limpograss stubble height when

aeschynomene was overseeded (SH), 2) legume seedling growth stage when

early-season grazing was discontinued (LSGS), and 3) legume height

when summer grazing was initiated. Early-season limpograss regrowth

was grazed to a 7.5- or 15-cm stubble. At the onset of summer rains

(7 June 1983 and 20 June 1984), dehulled, scarified, and inoculated

aeschynomene seed was broadcast at a rate of 20 kg ha-1 using a












cyclone seeder. Following overseeding, cattle were rotated among the

pastures to maintain the prescribed grass stubble heights. Grazing

was discontinued 1) when legume cotyledons were exserted, 2) when two

true leaves were present, or 3) 2 weeks after the two-leaf stage.

Pastures were not grazed subsequently until the legume reached heights

of 20, 40, or 80 cm in 1983 and 20, 40, or 60 cm in 1984. Following

initiation of grazing at the respective heights, all pastures were

grazed rotationally with a rest period of 5 weeks. Defoliation was by

mob grazing to an 8- to 12-cm stubble. Exceptions were pastures where

grazing was not initiated until the legume was 60 to 80 cm tall. The

yearling and 2-year-old steers (Bos spp.) would not graze the mature

herbage to a low stubble, so a more subjective visual appraisal of the

pasture was made to determine when grazing should end.

The complete factorial set of treatments (2 x 3 x 3) was arranged

in a split-split plot design in 1983 and a split-plot design in 1984.

All treatments were replicated twice in each year, and pasture size

was 500 m2. Grass SH was the whole-plot treatment in both years, and

LSGS was the subplot treatment in 1983. The design was changed in

1984 to provide a stronger test for LSGS differences, despite the

original design being more convenient for animal management. Legume

seedling growth stage means were compared using the single degree of

freedom contrasts cotyledon vs. two-leaf stage, and cotyledon plus

two-leaf stage vs. 2 weeks after the two-leaf stage. Initiation

height effects were evaluated using orthogonal polynomials to

determine the nature of response curves. Regression models for













aeschynomene leaf-stem data did not include quadratic terms unless

they were significant at P < 0.05.

In 1983, pastures defoliated initially when the legume was 20,

40, and 80 cm tall were grazed four, three, and two times (cycles),

respectively. In 1984, both 20- and 40-cm treatments were grazed

three times, and the 60-cm treatment was grazed twice. Grazing

seasons for the legume-grass association extended from 11 July through

4 Nov. 1983 and 23 July through 2 Nov. 1984.

All pastures were sampled before and after each grazing period.

A double sampling technique was used to determine herbage mass and

sward botanical composition. It included taking visual estimates of

herbage mass, percentage legume, percentage grass, and percentage

weeds at five sites in each pasture. The same 0.5 m2 areas were then

clipped to ground level, and the fresh herbage was hand-separated into

aeschynomene, limpograss, and weed components. The separated forage

was dried at 60C for 48 hours before weighing. Actual values for

each response variable were calculated and regressed on visual

estimates (forced through the origin). These equations were used to

correct the mean of 30 visual estimates of each response variable that

were taken at randomly selected locations over entire pastures. The

procedure was discussed in more detail in Chapter IV.

The hand-separated botanical components were used for forage

quality analyses. For each pasture at each grazing cycle there were

six types of samples. These included pregraze aeschynomene,

limpograss, and weeds, and postgraze aeschynomene, limpograss, and

weeds. Because sampling was done at five sites pregraze and












postgraze, there were five different samples of each of the six types

generated each time a pasture was grazed. To simplify the analytical

process, the five samples of each type (e.g., postgraze legume) were

composite during grinding (weighted composites). Exceptions were the

pregraze grass component in 1983, and the pregraze legume and pregraze

grass components in 1984. In 1983, pregraze grass samples from each

site were analyzed, and values reported are means over the five sites

in a given pasture. In 1984, pasture composites were formed within

species after grinding forage from pregraze legume and pregraze grass

sites separately. A constant percentage of the ground forage from

each site was removed and mixed with that taken from the other four

sites to form the pasture composite.

For each initiation height treatment represented in the set of

pastures to be grazed in a given week, two pregraze legume samples

were selected for leaf-stem separation. The only restrictions to

random selection were that the two samples from one height treatment

could not be from the same pasture, and that there must be sufficient

legume herbage to complete the analytical processes on both leaf and

stem fractions. Whole plant legume samples were dried prior to leaf-

stem separation, and after separation the fractions were redried

overnight before weighing. The leaf fraction consisted of petiole,

rachis, and leaflets. Where whole plant CP concentration is compared

to that of legume leaf or stem, the whole plant value was obtained by

laboratory analysis of the composite pregraze legume samples from the

other four sites in that pasture.












Samples for analyses were ground to pass a 1-mm screen using a

Wiley mill. In 1983, a subset of 50 samples from each sample type was

selected and analyzed for N using a modification of the aluminum block

. digestion procedure of Gallaher et al. (1975) and for in vitro organic

matter digestibility (IVOMD) using a modification of the two-stage

technique (Moore and Mott, 1974). Nitrogen and IVOMD values for the

remainder of the samples were predicted by near-infrared reflectance

spectroscopy (Norris et al., 1976). In 1984, all analyses were by wet

chemistry procedures as described for 1983. Legume leaf and stem

samples were analyzed using wet chemistry methods in both years.

Crude protein was calculated as N x 6.25.

Unless otherwise indicated, CP and IVDOM concentrations were

expressed in terms of the annual accumulation of DM for the specific

fraction being discussed (e.g., aeschynomene, limpograss, weeds, or

total herbage). Values reported for CP and IVDOM concentration of

herbage accumulated and herbage consumed were calculated using known

quantities for pregraze and postgraze herbage mass, pregraze and

postgraze sward botanical composition, and pregraze and postgraze

component (aeschynomene, limpograss, and weeds) CP, OM, and IVDOM

concentrations for each grazing cycle. For example, when calculating

the CP concentration in aeschynomene herbage consumed, the first step

was to calculate legume herbage consumed and legume CP consumed in

each grazing cycle. Values were summed over cycles generating annual

totals for legume DM and legume CP consumed. Dividing kg of legume CP

consumed ha-1 by kg of legume DM consumed ha-1 provided a value for CP

concentration in aeschynomene herbage consumed.












Results


Aeschynomene Leaf and Stem Quality


Forage quality responses of aeschynomene leaf and stem tissues to

legume height at initiation of grazing are illustrated in Figs. 5.1

and 5.2. These data from the first grazing cycle in 1983 show that

legume leaf quality was minimally affected by increasing maturity. In

contrast, stem and whole plant forage quality declined markedly as

initiation of grazing was delayed. Quality of aeschynomene forage

consumed during grazing was consistently higher than that for pregraze

whole plants, while legume herbage remaining after grazing was

generally lower in quality than pregraze stem tissue.

In subsequent grazing cycles, leaf quality was not significantly

affected by legume height when plants were first grazed [leaf CP = 248

+ 0.48 (height), r2 = 0.19; leaf IVDOM = 700 + 0.20 (height), r2 =

0.03]. Leaf CP concentration remained above 230 g kg-1 DM through the

end of the season in both years, but IVDOM concentration declined to

approximately 650 g kg-1 OM during late October and early November.

Stem quality was lower throughout the grazing season in pastures where

initiation of grazing had been delayed [stem CP = 104 + 0.51 (height)

- 0.014 (height)2, r = 0.82; stem IVDOM = 528 -2.46 (height), r2 =

0.57]. Responses observed in 1984 were similar to those reported for

1983 (data not presented).

In both years, legume leaf:stem ratio decreased as initiation of

grazing was delayed (Fig. 5.3). Plants that were grazed at 20-cm

heights in the first cycle continued to have higher leaf:stem ratios












270

2 240- O L,^=254-0.048X,r2=0.01
m HC Y=222-0.806 X,r =0
S210 OWP,Y=215-1.036X,r =0.76

180 -
z
I 150
2
<: *PWP,Y= 246-5.58X+0.040X2
120 r2=092
z
S90
z
o 60
0- A S,Y= 180-2.83X-+0.019 X2,r2 =078
0 30 I I I 1
20 30 40 50 60 70 80

INITIATION HEIGHT (cm)


Fig. 5.1. First grazing cycle responses of aeschynomene leaf (L),
stem (S), pregraze whole plant (WP), postgraze whole
plant (PWP), and herbage consumed (HC) crude protein
(CP) concentrations to aeschynomene height at initiation
of summer grazing in 1983 (n = 20 for each response
curve).







83





0 8 r OL r=0.44
I a HC r = 0.77
r 720

z 640

<560


z
OWP,r2=QD89
S 480- s ,r2.Q91

8 400
400PWP, r2=0.82
o 320 -

240 ,
20 30 40 50 60 70 80

INITIATION HEIGHT(cm)


Fig. 5.2. First grazing cycle responses of aeschynomene leaf
(L, Y = 769 1.43x), stem (S, Y = 810 10.07x +
0.06x2), pregraze whole plant (WP, Y = 780 3.88x),
postgraze whole plant (PWP1 Y = 914 18.89x + 0.138x ),
and herbage consumed (HC, Y = 788 2.97x) in vitro
digestible organic matter (IVDOM) concentrations to
aeschynomene height at initiation of summer grazing in
1983 (n = 20 for each response curve).












1.8
1.8 1983CYCLE I,r2=0.87

1. Y= 2.65-0.057 X + 0.0004 X2
0 1.6 1983 CYCLES 2ncd3,r2=Q71
o = 1.134-0.0099X
F- 1.4 0 1984 CYCLE I,r2 =0.73
Y= 1.56-Q0129X
2 1.2

1 1.0-
8-

06-


1984 CYCLES 2and3
04 -
Y=1.46-0.0148X, r2=0.63
02 -
20 30 40 50 60 70 80

INITIATION HEIGHT (cm)


Fig. 5.3. The responses of aeschynomene leaf:stem ratio to
aeschynomene height at first grazing in 1983 and 1984.












throughout the grazing season than did plants first defoliated at 40

to 80 cm (Fig. 5.3).


Whole Plant and Total Sward Canopy CP Concentrations


Limpograss CP concentration was very low in both years of the

study, although in 1984 it was approximately 10 g kg- DM higher than

in 1983 (36 vs. 27 g CP kg-1 DM, Table 5.1). Grass SH affected CP

concentration of total grass herbage accumulated in 1983, as values

for 7.5- and 15-cm treatments were 28 and 25 g kg-1 DM (P = 0.07).

There was no LSGS effect in 1983 (P = 0.58), but it was important in

1984 (P = 0.03). There was no SH main effect in 1984 (P = 0.39), but

there was an SH x LSGS interaction (P = 0.02). The LSGS when spring

grazing ended did not affect CP concentration of grass grazed to a 15-

cm SH (P = 0.18), but limpograss CP concentrations when SH was 7.5 cm

were 44, 39, and 31 g kg-I DM for the two leaf plus 2-weeks, two-leaf,

and cotyledon treatments, respectively (P = 0.05 for cotyledon vs.

two-leaf stage, and P = 0.03 for cotyledon plus two-leaf stage vs. 2

weeks after two-leaf stage). Limpograss CP concentration was

significantly (P < 0.01), but not highly correlated with percentage

legume in herbage accumulated (r = 0.57 in 1983 and r = 0.55 in 1984).

Correlations were higher for the 80- (r = 0.80) and 60-cm (r = 0.84)

treatments than for the 20- (r = 0.59 in 1983, and r = 0.26 in 1984)

and 40-cm (r = 0.32 in 1983, and r = 0.63 in 1984) heights.

Aeschynomene CP concentration was highest in pastures defoliated

initially when the legume was 20 cm tall (Table 5.1). The






















i- O Ca
O, r o
-4~


0 M
CO

O
JU


.0
i


m c




p "
tu





e M






00
0 C

' a
to

"O

































00 (


e a
01 0






















0 0


Q al
0













M
u 0
OU
.4-00




















o) e
Oc m
u a


n c


oo













-4


a, 00 1-




















o U) U






ON -4O c







-4 I
co cor^


DI- I -


U0 0'

ON U) -T en 40
C-l Cl .-I i4




iW
+-- ^-'
0++ N
SOO 0 01
Cl U3 4-1

UO Ui


00 r- m w


r-T




en en en z


El











rl

o
0 u
-4










a a
0













10
C





























U)
m N
4-4
- 4 -














*4- ++











establishment variables, SH and LSGS, had no effect on legume CP

concentration in either year (P > 0.37).

Major weed species in the pastures included vaseygrass (Paspalum

urvillei Steud.) and various sedges (Cyperus spp.). Weed forage

quality was very low in 1983 and 1984, averaging 54 and 64 g CP kg-

DM and 440 and 380 g DOM kg-I OM, respectively.

Crude protein concentration in total herbage accumulated (HA) was

not affected by initiation height (P = 0.12 in 1983, and P = 0.15 in

1984, Table 5.1). The SH effect was not important in 1983 (P = 0.54),

but in 1984, CP concentration for the 7.5-cm treatment was 76 g kg-1

DM compared to 60 g kg- DM for the 15-cm treatment (P = 0.06). In

both years, the major factor affecting CP concentration of HA was LSGS

(Table 5.2). None of the LSGS treatments resulted in CP

concentrations in HA greater than 70 g kg- DM in 1983, and only the

two leaf plus 2-weeks treatment surpassed that level in 1984.

Animals consistently selected a diet higher in CP concentration

than that of HA (Tables 5.1 and 5.2). In 1983, only the two leaf plus

2 weeks LSGS treatment resulted in diet CP concentrations above 70 g

kg- DM. All LSGS means were greater than that level in 1984, and the

two leaf plus 2-weeks treatment again resulted in the highest CP

concentration in herbage consumed (HC).

Regression analysis was used to evaluate the relationship of CP

concentration in HC vs. that observed in HA (Table 5.3). Equations

were fit for each initiation height treatment, as height affected the

response (i.e., intercepts were different between heights within a

year) observed in both years (P < 0.01). Slopes of lines were similar











Table 5.2. The effect of aeschynomene seedling growth stage (LSGS)
when establishment period grazing ended on crude protein
(CP) concentration of total herbage accumulated (HA) and
total herbage consumed (HC) in aeschynomene-limpograss
swards.


1983 1984
LSGS HA HC HA HC

g CP kg-1 DM

Cotyledon 43 47 59 75

Two leaf 54 63 64 86

Two leaf plus
2 weeks 62 72 81 112

SEt 4.0 5.9 3.0 6.3

F test S S ** **

Contrast NS,S NS,S NS,** NS,**


tStandard error of a treatment mean.
NS = not significant, S and ** indicate significance at P 0.10 and
P < 0.01.
$Single degree of freedom contrasts of cotyledon vs. two-leaf stage,
and cotyledon plus two-leaf stage vs. two leaf and 2-weeks stage.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs