GROWTH ANALYSIS OF FLORIGRAZE
RHIZOMA PEANUT (Arachis glabra ta Benth.)

BY

ABELARDO JOSE SALDIVAR FITZMAURICE

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

UNIVERSITY OF FLORIDA
1983

TO MY PARENTS:
OLGA AND ABELARDO

ACKNOWLEDGEMENTS

The author sincerely thanks Dr. William R. Ocumpaugh, chairman of
the supervisory committee, for his continuous help throughout the conduc-
tion of the field work and during the preparation of this manuscript.
Thanks are also extended to Drs. William G. Blue, John E. Moore, Gerald
0. Mott, and Gordon M. Prine, members of the supervisory committee, for
their counseling, for reviewing the manuscript and for their constructive
criticism towards its improvement.

Special recognition is given to Consejo Nacional de Ciencia y Tec-
nologta (CONACYT), Mexico, for providing the funds for the studies of
the author. Recognition is extended to Colegio de Postgraduados, Chapingo,
for the financial support.

The author feels greatly indebted to his fellow graduate students
and all those who helped in the field: Juan Aguirre, Epigmenio Castillo,
Scott Christiansen, Ciro DSvila, Jonas Bastos de Veiga, Bel son Dzowela,
Rhonda Gildersleeve, Liana Jank, Dean Kelly, Jorge Rodrfguez, Luis
Rodrfguez, Marco Soto, Isabel Valencia, and Jorge V^squez. Very special
thanks go to Susan Sladden for the help in the laboratory. Dr. Kenneth
Boote for lending the laboratory for TNC analyses. Dr. John E. Moore for
the use of the in vitro facilities. Dr. Raymond N. Gallaher for the use
of technicon autoanalyzer and computing facilities, and Dr. Kenneth H.
Quesenberry for allowing the intrusion in his laboratory and for his
valuable help.

iii

The author is very grateful to Dr. Manuel Cuca Garcia for the
support and encouragement of five years and to Maria Ignacia Cruz for
being like a mother and also for patiently typing this manuscript.

Finally, the author wishes to express his deepest gratitude to
his parents, Olga and Abelardo, and to his sisters, Olga, Irma and
Susana for their financial help and the constant encouragement through
all these years.

iv

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS iii

LIST OF TABLES vi

LIST OF FIGURES ix

ABSTRACT x

INTRODUCTION 1

CHAPTER I - LITERATURE REVIEW 3

Arachis glabrata Benth 3

Agronomic Characteristics of Flori graze Rhizoma Peanut 4

Potential Use 5

Forage Management and Utilization 6

Forage Regrowth 7

Forage Quality 15

CHAPTER II - GROWTH ANALYSIS OF AERIAL AND UNDERGROUND PLANT

PARTS OF FLORIGRAZE RHIZOMA PEANUT 25

Introduction 25

Materials and Methods 27

Results and Discussion 32

Shoot Growth 32

Rhizome Growth 39

Total Growth 43

Summary 46

CHAPTER III - DIFFERENTIAL ACCUMULATION OF TOTAL NONSTRUCTURAL
CARBOHYDRATES AND NITROGEN IN PLANT PARTS OF

FLORIGRAZE RHIZOMA PEANUT 55

Introduction 55

Materials and Methods 57

Results and Discussion 59

Ice vs Dry Ice 59

Total Nonstructural Carbohydrates 61

Nitrogen 67

Summary 73

V

1

Page

CHAPTER IV - EFFECT OF DEFOLIATION ON GROWTH, TOTAL NON-
STRUCTURAL CARBOHYDRATES AND NITROGEN IN
FLORIGRAZE RHIZOMA PEANUT RHIZOMES 76

Introduction

Materials and Methods.
Results and Discussion
Rhizome Growth

Percent Nitrogen

CHAPTER V - EFFECT OF GRINDING METHOD ON in vitro ORGANIC
DIGESTIBILITY OF FORAGES

Introduction

Materials and Methods

Grasses ,
Legume
Summary . .

LITERATURE CITED.

76

77
80
80

Total Nonstructural Carbohydrates 86

91)

Summary

102

102
103

Results and Discussion ]9?

11)1
118
125

CHAPTER VI - GENERAL DISCUSSION "1 23

133

BIOGRAPHICAL SKETCH

vi

i

Table

LIST OF TABLES

Page

1 Sampling schedule for three data sets to record
the growth and regrowth of Fl or i graze rhizoma

peanut

2 Mathematical equations to describe the growth
of Flori graze rhizoma peanut components in

three data sets

3 Regression analyses of complete data and individual
comparisons between data sets for growth components

in Flori graze rhizoma peanut

4 TNC percentage in cooled (ice) and frozen (dry
ice) samples in 1981 regrowth (II) and 1981 (III)

of Flori graze rhizoma peanut °"

5 Regression analyses of complete data and individual
comparisons between data sets for TNC percent in
plant parts of Florigraze rhizoma peanut

6 Regression analyses of complete data and individual
comparisons between data sets for N percent in

plant parts of Florigraze rhizoma peanut d9

7 Soil mineral analysis in samples from field A and
field B in February 1981

8 Sampling and defoliation schedule for 1981 treatments
imposed on Florigraze rhizoma peanut

9 Regression analyses of complete data and individual
comparisons between defoliation treatments for growth
components of Florigraze rhizoma peanut 82

10 Mathematical equations to describe the effects of
frequency of defoliation on the growth of rhizomes
in Florigraze rhizoma peanut °^

n Regression analyses of complete data and individual
comparisons between defoliation treatments for TNC
percent in plant parts of Florigraze rhizoma peanut . . 88

vii

Table Page

12 Regression analyses of complete data and individual
comparisons between defoliation treatments for N

percent in plant parts of Florigraze rhizoma peanut 93

13 Weight of shoots per plant for defoliation treatments

at sampling dates in Florigraze rhizoma peanut. . . 97

14 List of grasses and sampling dates for Florigraze
rhizoma peanut

15 Differential IVNDFD in relation to grinding
treatment

16 Diagram of the laboratory experiment for grasses
and legume

17 Analyses of variance for all forages, grasses,

and legume samples

18 Analyses of variance for grasses separated by genus,

and for legume samples separated into leaf and stem. 114

19 Effect of grinding treatment and incubation time
on IVOMD by grass, and IVOMD means by grass by
treatment

20 Regression analyses per year and plant fraction
for treatments and incubation effects upon IVOMD
of Florigraze rhizoma peanut, and treatment means

by year and plant fraction 121

21 Regression and correlation coefficients for
prediction of Organic Matter Digestibility (OMD)
and Organic Matter Intake (OMI) from IVOMD of nine
grasses with four grinding treatments and with

three incubation times '23

viii

LIST OF FIGURES

Total growth and components of growth in

Flori graze rhizoma peanut for three data sets. . .

Growth rates of plant components in Flori graze
rhizoma peanut for three data sets

Relative growth rates of plant components

in Flori graze rhizoma peanut for three data sets .

Production of shoots in Flori graze rhizoma
peanut for three data sets

Temperature and rainfall recorded in Gainesville
in 1980 and 1981

Primary and secondary rhizomes

Percent TNC (organic matter basis) in plant
parts of Flori graze rhizoma peanut for three
data sets

Percent Nitrogen (organic matter basis) in plant
parts of Flori graze rhizoma peanut for three
data sets

Effect of defoliation treatments on growth of
rhizomes in Flori graze rhizoma peanut

Effect of defoliation treatments on TNC percent
in plant parts of Flori graze rhizoma peanut. . . .

Effect of defoliation treatments on N percent

in plant parts of Flori graze rhizoma peanut. . . .

Effect of defoliation treatments on the number
of shoots in Flori graze rhizoma peanut

Effect of grinding treatment and incubation on
IVOMD of nine grasses . . .

Effect of grinding treatment and incubation on
IVOMD of seven samples of Flori graze rhizoma
peanut

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

GROWTH ANALYSIS OF FLORIGRAZE
RHIZOMA PEANUT (Arachis glabrata Benth.)

By

Abelardo Jose Saldfvar Fitzmaurice
April 1983

Chairman: William R. Ocumpaugh
Major Department: Agronomy

A series of studies was conducted to determine growth characteris-
tics and responses to defoliation under field conditions; and quality
characteristics of Flori graze rhizoma peanut. Growth of shoots and
rhizomes was monitored in undefoliated plants for two consecutive years
(1980 growth and 1981 regrowth). The first year growth was replicated
(1981 growth) in another field. During the second year (1981 regrowth),
growth of rhizomes was recorded in plants defoliated every 2, 6, and 8
weeks. Total nonstructural carbohydrates (TNC) and nitrogen (N) were
determined in plant parts. Rhizoma peanut forage samples (leaf and stem)
were compared to nine tropical grass samples, as to the effect of four
grinding treatments (Udy vs Wiley mills, 1.0- and n.5-mm screen) and
three incubation times on in vitro organic matter digestibility (IVOMD).

In undefoliated plants relative growth rates were higher in 1980
and 1981 growth than in 1981 regrowth, indicating a reduction in relative
growth potential. The cause of this appears to be N deficiency in early

X

regrowth. In both years rhizome growth paralleled shoot growth until
September. From then on, shoot growth declined and rhizomes continued
to grow, at least until the first frost. Defoliation reduced plant
biomass, especially rhizome growth. Regardless of defoliation fre-
quency, little growth was observed in rhizomes after September. In the
rhizomes of undefoliated plants, TNC accumulation was higher in 1981
regrowth than in 1980 and 1981 growth. In November 1981 the rhizomes
of undefoliated plants (1981 regrowth) contained about 70% TNC compared
to 45 to 50% TNC for defoliated plants; rhizomes of 8-week frequency
had higher TNC than 2-week and 6-week frequencies (P = 0.0141 and
P = 0.0001, respectively). The TNC accumulation in rhizomes started
with the reduction of growth rates of shoots and rhizomes in September
and continued to increase until the end of the growing season regardless
of defoliation treatments.

All grasses and legume samples responded to grinding treatment.
Grinding treatment had a greater effect on low quality forages than on
high quality forages. In grasses, 0.5-mm grinding increased IVOMD.
In the legume samples 0.5-mm grinding increased IVOMD also, but 1.0 mm
Udy gave higher IVOMD than 0.5-mm Wiley. The IVOMD of legume samples
was affected very little by incubation time. Fine-grinding in grasses,
increased accuracy of prediction of organic matter digestibility (in_
vivo) from IVOMD.

x1

INTRODUCTION

Forage legumes may be the most economical way of introducing ni-
trogen (N) in pastures to enhance animal production. This might be par-
ticularly important in those production systems where economic returns
are not sufficient to justify the use of N fertilizer. As opposed to
temperate areas, legumes in the tropics constitute a relatively new area
of research. In the last 20 years, researchers in tropical areas have
devoted considerable effort to understand the dynamics of legume-grass
associations. However, the results are controversial. The failure in
a lot of instances is associated with lack of complete understanding
of the physiological capabilities of the legumes. The purpose of the
association is to provide N to the system while increasing the nutritive
level of the animal diet. Therefore, to achieve this goal it is neces-
sary to study the management or alternate managements in the associates
to determine their compatibility.

Intensity and frequency of defoliation are the two most important
factors in forage management. Forage plants respond differently, physio-
logically and morphologically, to these two factors. These responses de-
termine the survival of the plants. Morphology and regrowth mechanisms
in grasses allow them to survive a relatively wide range of defoliation
managements, whereas legumes are more restricted. In North Florida, most
perennial tropical legumes do not survive the winter and the annuals
do not start growing until late June. Florigraze rhizoma peanut

1

2

has a great potential for North Florida because it survives the winter
and starts growing very early in the spring. Some observations on
this legume suggest that it might be associated with bahiagrass
(Paspalum notatum Fl'ugge.) under grazing conditions. The persistence
of this legume to grazing is related to its rhizomatous nature. The
rhizomes are also used for vegetative propagation since it does not
produce seed. This fact represents an inconvenience because up to
2 years are required for establishment of pastures.

The objectives of this research were to study the growth charac-
teristics of Florigraze rhizoma peanut, the role of carbohydrate re-
serves and N, the responses to defoliation, and the comparison of its
forage quality to tropical grasses of known in vivo quality data.

CHAPTER I
LITERATURE REVIEW

Arachis glabra ta Benth .

Peanuts (Arachis spp), also known as groundnuts, are the most wide-
spread and potentially the most important food legume species of the
world; they are grown in some 82 countries (Varnell and McCloud,
1975).

The genus Arachis originated in South America where it is widely
distributed from the Amazon River through Brazil, Bolivia, Paraguay,
Uruguay and northern Argentina to about 35° south latitude (Higgins,
1951). The genus Arachis comprises several species. Herman (1954)
described eight species of wild peanuts from which seven are perennial.
Gregory et al . (1973) stated that on the basis of collected material
the genus could be grouped into 30 to 50 different species. The three
perennial Arachis species most widely distributed in South America are
qlabrata. marginata, and villosa. Their limit of distribution is
about 35° south latitude which in the northern hemisphere would almost
correspond to the northern boundary of the state of Georgia in the
U.S.A. This feature suggested that some ecotypes among those perennial
species could be adapted to North Florida (Prine, 1973).

Arachis glabrata Benth. was introduced into the United States in
1936, by W. Archer,, as PI 118457 and released to farmers with the variety
name 'Arb' (Prine, 1964). Since then other introductions of rhizoma

4

peanut have been named and distributed to farmers for testing (Prine
et al,, 1981). In 1962, a rapidly spreading plant was observed in an
Arb rhizoma peanut plot, by G. M. Prine, who isolated it and tested
it as Gainesville Selection No. 1 (GS-1) (Prine, 1973). It is sus-
pected that this selection is a seedling from Arb. This selection was
released with the cultivar named 'Florigraze' in 1979 (Prine et al . ,
1981).

Agronomic Characteristics of Florigraze Rhizoma Peanut

Florigraze rhizoma peanut is a rhizoma tous legume. It produces
very little seed; therefore, its establishment is accomplished by vege-
tative propagation (sprigging) or by planting rhizomes. The latter is
the most recommended method of propagation. A potato digger was one
of the first mechanical devices used for harvesting the rhizomes from a
well-established rhizoma peanut field (Adjei, 1975). One square foot
(930 cm^) mats were planted in hills 7 to 10 cm deep separated 1.8 m,
but better coverage is obtained if the mats are broken up into four
pieces and planted 0.9 m apart in every direction (Adjei, 1975; Adjei
and Prine, 1976). Recently, the rhizome material is being dug with a
bermudagrass sprig digger. Using this machine the rhizomes come out cut
in small pieces, and according to Prine et al . (1981), this may reduce
the efficiency of propagation. Hence, they suggested that a modification
to the machine may help to obtain larger pieces of rhizomes. However,
at the present time there is no experimental evidence to conclude such
a loss in efficiency (Prine, 1982. Personal communication).

Florigraze is adapted to well -drained soils. Prine et al . (1981)
emphatically stated that it is not adapted to flatwoods soils or to

5

any soil subjected to high water tables. Like any other legume, phos-
phorus (P) and potassium (K) should be provided for good growth, but
special attention should be focused on calcium (Ca) and magnesium (Mg)
(Prine et al., 1981). To ensure good nodulatfon, it is recommended to
inoculate rhizomes at planting with Rhizobium .iaponicum (Adjei , 1975).

Florigraze is a slow establishing crop with fastest growth during
the summer months. It takes from 2 to 3 years to get a solid sward
(Prine, 1979). It is drought resistant, and if drought is severe, top
growth may die and then regenerate from rhizomes following rainfall
(Prine et al . , 1981). Some reports indicate that well established plots
resisted frequent defoliation (Breman, 1980; Beltranena, 1980).

The rhizomatous habit of A. glabra ta makes it suitable for grazing.
Some studies have shown that it produces approximately the same amount
of underground dry matter as it does above ground (Breman, 1980). Even
though other reports indicate that it tolerates grazing and trampling
by cattle, alone and in association with grasses (Blickensderfer et
al., 1964; Otero, 1946), there is no information regarding its persis-
tence in relation to its physiological or morphological responses to
defoliation.

Potential Use

Prine (1964) planted Arb rhizoma peanut in association with
'Pensacola' bahiagrass, 'Coastal' bermudagrass (Cynodon dactyl on L.)
and 'Pangola' digitgrass (Digitaria decumbens Benth.); he found that
the highest dry matter yield was produced when associated with bermuda-
grass (5,040 kg/ha) or in pure stand (4,360 kg/ha). In another study, in
pure stands, Florigraze produced more dry matter yield (10,300 kg/ha)

6

than Arb (8,740 kg/ha) (Prine, 1973). Quality determinations of
Florigraze rhizoma peanut in vivo (Prine et al., 1981) indicate no
palatability problems; voluntary intake was well above the maintenance
requirement in two hays with different maturity. In a preliminary
study, Saldivar et al . (1981) reported that suckling calves creep
grazing Florigraze rhizoma peanut gained 160 g/day more than calves
without creep grazing.

The persistence, high quality, and yield of Florigraze rhizoma
peanut make it a forage with a great potential. Its slow establishment
may be overcome by its perennial ity and persistence under diverse
production systems. Currently, research is being conducted at the
University of Florida to determine the feasibility, from both practical
and economical standpoints, of intercropping Florigraze rhizoma peanut
with cash crops to reduce establishment costs. Preliminary results on
these investigations indicate that some cultural practices may enhance
the establishment rate during the second year (Franca-Dantas, 1982).

Forage Management and Utilization

Forage utilization is determined by type, intensity, and economics
of livestock production operations. In any production system, forage
utilization involves the removal of foliage; this is technically
defined as defoliation, which means removal of leaves, but it more
commonly refers to removal of shoots (leaves and stems) in forage
species. The frequency and degree of defoliation determine the in-
tensity of forage utilization (Humphreys, 1981). Yet, time of defolia-
tion, in relation to plant development, may be as important (Humphreys,
1966). Degree of defoliation for mechanized harvesting can be defined

7

as cutting height in cm because the homogeneity of defoliation, with
respect to plant structure, promotes a high correlation with leaf area.
In grazing situations, defoliation is not homogeneous; therefore, two
parameters have been used. The first one defines degree of defoliation
as residual leaf area index (LAI) after grazing or cutting (Humphreys
and Robinson, 1966; Brown et al . , 1966a; Brown and Blaser, 1968) whereas
the second one takes into account age effect in photosynthetic activity
of leaves (Brown et al . , 1966b) as well as photosynthesis of stems
(Ludlow and Wilson, 1971) and defines degree of defoliation as residual
dry matter/ha (FDM/ha) after grazing (Mott, 1973). On the other hand,
frequency of defoliation is defined as the period of time between defolia-
tions. Time of defoliation deals with the physiological changes occurring
in the plants that may affect growth.

The response of forage plants to the interaction of frequency, degree
and time of defoliation may be of physiological and/or morphological im-
portance, and it has been a primary concern to pasture scientists because
the concurrent manipulation of all three factors determines yield and
persistence of forage. Forage utilization involves the animal or the
transformer of forages into products for human consumption (Blaser et al . ,
1966); therefore, a compromise between plant and animal yield may be per-
tinent (Humphreys, 1966).

Forage Regrowth

Plant growth depends on photosynthesis as a source of energy, and
nutrients from the soil. Any factor affecting photosynthesis will also
affect growth. Complete defoliation temporarily eliminates the photo-
synthetic apparatus (shoots), stopping growth until new shoots are

8

developed. This replacement of shoots after defoliation is termed
regrowth. The regrowth is associated with meristematic tissue, which
is responsible for cell division, and is found in growing centers known
as buds (Jewiss, 1972a; Humphreys, 1981). Buds remain dormant until
stimulation is produced. Development of buds into shoots is hormone
regulated (Phillips, 1969). When a shoot develops, the apical meristem
produces new leaves and new buds are formed which will remain dormant
until stimulation occurs. Every bud has potential to produce a new
shoot, or branch, depending on the growth habit of the plant. However,
apical dominance restrains the development of these buds by an inhibi-
tory mechanism controlled by hormones (Thimann, 1937) and metabolic
sinks (Phillips, 1969). When the apical meristem is removed, regrowth
is stimulated and axillary buds develop into shoots. The number of
buds that grow and the rate of growth of these newly developed shoots
will depend on nutritional factors (Sachs and Thimann, 1964). In the
decapitated shoot, leaf and bud formation is stopped and further leaf
and bud formation will depend on the new shoots. In grasses, this
process is less understood, especially in sod-forming species where
apical dominance is difficult to track (Humphreys, 1981); however,
it is better understood in dicotyledonous species (Phillips, 1969).
Defoliation of grasses above the apical meristem does not stop growth
since mutilated leaves continue to elongate (Jewiss, 1972a) unless
a leaf has completed expansion (Langer, 1954). In grasses with deter-
minate growth, the apical meristem undergoes a series of changes and in-
florescence formation initiates at the onset of reproductive phase.
This process stops leaf formation in the shoot and inhibits the

9

development of axillary buds. In legumes the shoot apex remains
vegetative whereas axillary buds develop inflorescences. The inhibi-
tion of axillary buds development receeds shortly after stem elonga-
tion is almost complete and inflorescence emerges (Jewiss, 1972a). When
shoot development is initiated in the buds, energy is required for
cell division and elongation. The source of energy could be nutrient
reserves in the plant, which are translocated from the storage organs
(Smith, 1981) to the growing sink, or from photosynthesis in the tissue
remaining after defoliation (Blaser et al . , 1966). The degree of con-
tribution of the two sources may be quite variable, and they will be
discussed later in more detail.

Forage yield and persistence are often correlated with a vigorous
regrowth, and in turn three factors are recognized to affect forage
regrowth: 1) Total non-structural carbohydrates (Graber, 1931;
Weinmann,1961 ; Smith, 1981); 2) Number of growing points (Booysen et
al., 1963; Jewiss, 1972b; Kretschmer, 1976; Knight, 1970); 3) Leaf
area left after defoliation (Brown and Blaser, 1968; Valle, 1977;
Andrade, 1979). The first two are forage species dependent whereas the
third one is an external factor that modifies the other two.

Non-sitructural carbohydrates. Non-structural carbohydrates, com-
monly referred as total non-structural carbohydrates (TNC) (Smith,
1981) or total available carbohydrates (TAC) ( Wei nmann, 1961 ) are
the primary source of energy to initiate regrowth in perennial and
biennial forage plants (Smith, 1981); particularly after a period of
dormancy where no photosynthesis occurs. The accumulation of TNC in
undisturbed plants follows a particular seasonal trend in every

10

species (Weinmann, 1948, 1961; Smith, 1975, 1981). It usually increases
in summer and decreases during periods of active growth (Humphreys, 1981).
The levels of concentration of TNC in storage organs (roots and/or crowns)
can be reduced by liberal applications of N fertilizer (Brown et al.,
1966a) because more assimilate is utilized in shoot growth and less is
mobilized to roots and stem bases (Hojjati et al., 1968), Moderate
shading also decreases the concentration of TNC (Burton et al . , 1959;
Blaser et al . , 1966) without decreasing growth, unless liberal N is
supplied (Burton et al . , 1959; Colby and Drake, 1966). Conversely,
water stress and suboptimal temperatures favor TNC accumulation (Eaton
and Ergle, 1948; Blaser et al . , 1966). This indicates that when photo-
synthesis exceeds carbohydrate utilization, TNC accumulates. The op-
posite occurs when growth exceeds photosynthesis (Wardlaw, 1968). Blaser
et al. (1966) stated that low temperature affects growth more than it
affects photosynthesis. Some tropical grasses are especially sensitive
to low temperatures (West, 1969; Carter et al., 1972; Chatterton et al.,
1972) and carbohydrate produced in leaves is not translocated to stems
and roots, resulting in reduction of photosynthesis. Another example
of this relationship can be seen in some legumes where N fixation does
not meet the N demand for growth, thus carbohydrate may accumulate in
roots but low applications of N maintain low levels of root TNC (Barta,
1979). In some circumstances, concentration of TNC may not be a good
parameter because a decrease in concentration might be produced by in-
creased plant mass (Weinmann and Goldsmith, 1948; Humphreys and Robin-
son, 1966). A drop in TNC has been observed after defoliation (Graber,
1931; Weinmann, 1947, 1952) and it has been positively correlated with

n

plant regrowth (Wolf, 1978). For these reasons the accumulation of high
levels of TNC in roots and basal stems of forages has been the objective
in grazing and cutting practices (Humphreys, 1966; Smith, 1981). How-
ever, other studies have failed to correlate regrowth with levels of
TNC (Davidson and Milthrope, 1965; Weinmann, 1947; Humphreys and Robinson,
1966). There is a general agreement in that a minimum level of TNC is
required to initiate regrowth in the spring. The controversy arises in
regard to regrowth after defoliation because some grasses do not show a
marked reduction in concentration (Weinmann, 1947) unless frequently
defoliated (Weinmann and Goldsmith, 1948). It seems that the interac-
tion between growth habit and carbohydrate levels might be the answer to
this matter because complete defoliation is seldom achieved in decumbent
plants, whereas plants with erect growth are more easily defoliated
(Rossi ter and Collins, 1980),

Growing points. As it was discussed previously, regrowth mechanisms
are essentially the same in all plants but their morphology is highly
variable. The morphology of grasses varies from bunch-type to
rhizomatous types, and intermediate types that produce stolons and
types that modify their morphology from upright growth to horizontal
growth (Humphreys, 1981). Legumes, on the other hand, present very
simple forms with one main stem that branches as it grows. These
simple types might be of erect growth or climbing type of growth.
They also present forms that have crowns from where several shoots are
produced (Oakes, 1968; Brewbaker, 1976). Little is known about actual
numbers of buds, but more research has been done on the position

12

of growing points and the hierarchical distribution. When shoots
elongate during the reproductive phase, lateral buds arise also; there-
fore, if shoots are cut or grazed, most lateral buds may be removed;
hence regrowth will initiate from basal buds (Jewiss, 1972b) which
means that fewer shoots may be formed. This type of relation varies
between species (Branson, 1953) due to the proportion of vegetative
to reproductive shoots. This proportion changes also with the season
(Knight, 1970). In some grasses, early removal of vegetative shoots
increases the proportion of reproductive shoots, resulting in a con-
siderable increase in seed production. In some grasses, internodal
elongation occurs during the vegetative stage resulting in the elevation
of vegetative buds above the soil surface and susceptibility to removal
(Booysen et al . , 1963) .

In legumes, the hierarchical distribution of buds is better under-
stood (Phillips, 1969). However, growing points move away from the
soil along with the shoot apex, and the growth suppression on lateral
buds depends on nutritional status and light. Defoliation of legumes
close to the ground eliminates most growing points; therefore, regrowth
must start at the crown of the plants. Variation has been found in the
numbers of shoots that legumes produce after defoliation (Kretschmer,
1976; Imrie, 1971). Poor regrowth has been related to deficiency of
basal bud sites in some erect types of Stylosanthes guianensis (Grof
et al., 1970). These results indicate that some legumes may show
little resistance to defoliation. This may be particularly important
in legumes with climbing habit unless they are able to form rooting
stolons where new growing points are generated (Hutton and Beall, 1977).

13

Some woody legumes have an excellent regrowth capacity from buds near
the crown (Brewbaker, 1976; Gildersleeve. 1982). Grasses and legumes
with rhizomatous habits maintain growing points below cutting height;
therefore, more vigorous growth after defoliation is often observed. The
pattern of shoot formation is typical of every plant; therefore, defolia-
tion at a particular time may reduce shoot production.

Degree of defoliation. As discussed previously, defoliation means
the removal of shoots, but from the standpoint of forage regrowth,
what is left of the plant is important. Single plants may not reflect
the response of a plant community (Alexander and McCloud, 1962) because
light interception (Brougham, 1955), plant morphology (Ludlow and
Wilson, 1970), and plant age (Brown and Blaser, 1968) interact in the
rate of regrowth. The plant material that remains after defoliation
contributes to the process of regrowth through photosynthesis and for
nutrient reserves. The extent of photosynthesis will depend primarily
on leaf material. It has been shown that increasing leaf area
increases the growth rate until 95% of the light was intercepted
(Brougham, 1955, 1956). The point where growth rate reaches its maximum
was termed optimum LAI (Davidson and Donald, 1958), and at this point
the leaves under the canopy are at their compensation point. However,
different growth rates can be observed with similar LAI's when the
canopies differ in age (Begg and Wright, 1964; Brown et al., 1966b).
This means that a lenient defoliation does not necessarily result in
a higher regrowth (Bryant and Blaser, 1961). It has been observed that
higher productions are obtained with less frequent but complete
defoliations than with several partial defoliations (Ridgman, 1960;

14

Reid, 1966; Vickery, 1981). Blaser et al . (1966) suggested that high
degree of defoliation is recommended, especially after a long period
of growth. Vickery (1981) concluded that lenient defoliations are
not favorable to plant regrowth since animal and mechanical defolia-
tions remove most of the newer leaves which are in the top of the
canopy. These leaves are more photosynthetically active than the
ones in the lower canopy. Plant morphology also affects the degree of
defoliation and consequently the rate of regrowth. Plants with
prostrate habits are more difficult to defoliate (Rossi ter and
Collins, 1980; Humphreys, 1981) but are less affected by changes in
photosynthetic efficiency since most of their leaves are equally ex-
posed to light. Therefore, lenient defoliations can maintain a higher
growth rate (Ludlow and Charles -Edwards, 1980).

Interaction of factors. From the previous examination of factors
affecting regrowth, it is clear that no one by itself can be considered
as the key to forage management. In this discussion, very little was
said about environmental factors which undoubtedly play an interactive
role. With the understanding of the physiological mechanisms of re-
growth in plants, it can be easier to understand their potential adapt-
ability to a given environment which will encourage changes in the
plant to ensure its survival. From an agronomist's point of view,
survival of plants is not enough; they must produce under defoliation,
but management of defoliation may be the key to plant survival. There-
fore, knowing when and how much can be defoliated constitute the
parameters in forage production. But with the animal as utilizer,
forage production per se may not be very meaningful; thus animal produc-
tion becomes the determinant in forage utilization.

15

Forage Quality

Barnes (1981) defined the term forage as "the vegetative portions
of plants that are consumed by animals" (p. 2). From a nutritional
point of view, Van Soest (1973) defines it as "coarse livestock food
that is composed of leaves, stems, and sometimes grain" (p. 53).
Then, forage quality refers to the efficiency of edible plant materials
to provide nutrients for growth, reproduction, etc. Mott (1959) sug-
gested that forage quality must be expressed ultimately in terms of
output per animal. The use of output per animal as a measure of forage
quality takes into consideration the interactions, known and unknown,
between animal, plant, and environment that determine animal production.
According to Mott (1973), forage quality comprises the nutritive value
and the consumption or "intake" of forages. Nutritive value is a func-
tion of the chemical composition of a forage, its digestibility and the
nature of digested products. On the other hand, intake is a function of
the acceptability, rate of passage, and availability. With these con-
cepts in mind, Moore and Mott (1973) pointed out that forage quality
can be expressed in terms of output per animal provided that forage
availability is not limiting and animal potential does not vary between
treatments. In a strict sense, forage quality cannot be defined in
terms of a single component; however, the quantification and/or qualifi-
cation of one or more of those components may lead to an approximate
estimation of forage quality.

The qualitative evaluation of forages calls for comparison methods
that allow the selection of the most productive forages. Those methods
must be simple, inexpensive„and easy to reproduce. Over the years it

16

has been widely accepted that the best measure of forage quality is
animal performance; however, in very many instances it is not possible
to compare several forages in animal trials; this is particularly
critical in plant breeding programs where a large number of varieties
and small samples are evaluated (Marten, 1981). Minson (1981a) stated
that two other reasons for using alternate methods of forage evaluation
are the formulation of rations to feed animals, and identification of
problems when animals fail to achieve the expected performance. For
these reasons, several approaches have been taken to elucidate the
relative influence of the known components of forage quality. Digest-
ibility and intake of forages seem to be the most widely accepted param-
eters of forage quality. Raymond (1969) combined digestibility and
intake to develop the parameter termed nutrient intake, which is defined
as the intake of feed x the digestibility of feed x efficiency of uti-
lization of digested feed. Although this particular index requires
animal trials to determine the consumption of the forages under evalua-
tion, it can be very useful if its components were estimated from labo-
ratory analyses. Since both intake and nutritive value are in turn
affected by various factors, they will be discussed separately.

Forage intake. Different approaches must be taken to define the
components of forage Intake, i.e., in the pasture situation intake is a
function of the acceptability, rate of passage, and availability (Mott,
1973, whereas, in confinement, availability may not be limiting. It has
been postulated that intake of forage has a greater effect on forage
quality than does nutritive value (Ventura et al., 1975) and it is
attributed to the "distension" mechanism by which certain level of

17

fin in animals causes the cessation of consumption (Campling, 1970).
According to Van Soest (1973), the attempts to establish measurements
of intake as a feed evaluation characteristic have been difficult be-
cause the urge to eat varies among animals depending upon their physio-
logical requirements. For comparison purposes, this can be overcome
by utilizing several animals per treatment (Heaney, 1970) and animals
of the same age and condition. Abrams (1980) summarized six factors
responsible for the control of forage intake, when the control mechanism
is distension. The six factors are l)rumen fill, 2) potentially digest-
ible and indigestible pool sizes, 3) initial particle size distribution,
4) rate of particle size reduction, 5) rate of passage, and 6) rate of
digestion. In the discussion of these factors Abrams pointed out that
they do not operate independently and their interrelationships will
vary from forage to forage. This type of fractionation has been utilized
in modeling for prediction of intake and digestibility. The models
tested have been promising although the results so far are not satis-
factory (See Waldo et al., 1972; Martens and Ely, 1979; Golding, 1976;
Abrams, 1980).

The prediction of intake from laboratory analyses has not been very
successful, and this, in the opinion of Donefer (1970), has been
due to lack of complete understanding as to what physiochemical mecha-
nisms, plant and animal, influence the voluntary intake by ruminants.

It is a general agreement that if intake is controlled by disten-
sion, the intake must be regulated by the unload rate of forage from
the reticulo-rumen. Grovum and Williams (1973) provided experimental

18

evidence to confirm that the reticulo-rumen is the limiting compart-
ment in the passage of dry matter. Furthermore, the physical form of
the material found in the subsequent compartments indicates that most ]
physical breakdown of forage occurs in the same compartment (Bailey
et al., 1976). This suggests that the rate of forage unload depends
upon mechanical and chemical degradation (Laredo and Minson, 1973).
So far, it has not been possible to delineate the relative influence of
these two mechanisms. In an attempt to characterize the physical degra-
dation in the rumen, Chenost (1966) proposed a "f ibrousness" index,
obtained by measuring the energy required to grind forage samples through
a l-irm screen. He found a 0.90 correlation coefficient with six temper-
ate grasses between the f ibrousness index and dry matter intake. Laredo
and Minson (1973) utilized the same method on five tropical grasses,
separated in leaf and stem. They reported that more energy was required
to grind stems than leaves. Their overall correlation coefficient be-
tween grinding energy and dry matter intake was 0.70 for leaves and
stems together. Worrell (1982) found negative correlations between
energy required for regrinding and organic matter digestibility and
neutral detergent fiber digestibility.

Nutritive value. The second major component of forage quality is
the nutritive value, and it refers to the relative supply of available
nutrients contained in a forage. Mott (1973) summarized the constit-
uents of nutritive value into three: chemical composition of the for-
age, its digestibility, and the nature of digested products. Due to
the nature of forages, digestibility is the most widespread parameter
of forage quality. It can be applied not only to dry matter, but also

19

to the various chemical fractions of the forage; e.g., organic matter,
cell wall (Neutral Detergent Fiber), lignin and energy (Hacker and Min-
son, 1981).

Digestibility. Digestibility refers to the disappearance of forage
material due to the truly chemical breakdown of the chemical constituents
of forages into compounds absorbed by the animal. Estimation of digest-
ibility with animals, along with forage intake, is the best parameter
to estimate forage quality. In most cases, in vivo digestibility refers
to apparent digestibility, since true digestibility is seldom estimated;
the difference between true and apparent digestibility of dry matter is
approximately 12.9% units (Van Soest and Moore, 1965). Therefore, in
most cases, the digestiblity values reported for forages are underestima-
tions of true digestibility. It is known that digestibility varies among
species of plants (Moore and Mott, 1973) and among genotypes of both
temperate and tropical grasses (McLeod and Minson, 1969; Burton et al.,
1967) as well as tropical and temperate legumes (Moore and Mott, 1973).

It is unknown what causes digestibility to change, but the changes
associated with maturity of a forage have supported the conclusion that
digestibility is more a function of plant factors than animal factors,
though it is recognized that efficiency of digestion is related to
many animal factors such as species, level of intake, etc. (Minson,
1976). The relationship between digestibility and chemical composition
of forages has been approached through two general methods — the first
one involves the fractionation of feed and feces into chemical com-
pounds (Gaillard, 1962; Jarrige, 1965), the second one fractionates the
forage into cell wall and cell contents (Van Soest, 1967). Both

20

methods demonstrated that the availability of the cell contents is not
limited by the cell wall. According to Van Soest (1976), these find-
ings support the conclusion that cell wall digestibility is the limit-
ing factor in forage quality. Chemical changes occur when a plant
matures, the cell walls become relatively more important, and the
fibrous constituents increase and become more lignified. These changes
result in progressively lower dry matter digestibilities (Sullivan,
1969). Regressions have been used to relate digesbitil ity of forages
to lignin, acid detergent fiber (Van Soest and Moore, 1965; Van Soest,
1965; McLeod and Minson, 1971), cellulose and hemicellulose (Russo,
1981), cell wall (Abrams, 1980) but none of them explains the vari-
ability found among and within species. Van Soest (1976) stated that
the error with the use of lignin is particularly increased through
the interaction of legumes and grasses, since grasses tend to have a
higher cell wall content and lower lignin content than legumes at
the same digestibility.

Unlike forage intake, the estimation of in vivo digestibility
through laboratory analysis has been very successful. This has been
accomplished by several methods; however, two of them are the most
widely accepted by the scientific community. The one developed by
Tilley and Terry (1963) is known as two-stage in vitro technique.
It is based on the simulation of the digestive processes. The first
stage involves the incubation of the forage sample with rumen fluid-
buffer solution; and the second stage, a treatment with acid-pepsin.
This method has been considered the most accurate laboratory method
to predict dry matter digestibility in vivo of temperate and tropical

21

forages (Barnes, 1973; Minson, 1981a); and it has been used in plant
breeding and selection programs in many parts of the world (McLeod and
Minson, 1978). Another method to estimate in vivo digestibility was
proposed by Donefer et al . (1963) and it is based on the utilization
of purified enzyme extracts containing cellulase. Jones and Hayward
(1975) described a two-stage technique in which a sample is incubated
first with acid-pepsin to remove cell contents, followed by the incuba-
tion of the residue with the cellulase. According to Minson (1981a),
the application of this method became a successful routine when an
inexpensive "broad spectrum" cellulase became commercially available;
this enzyme had cellulase, hemicellulase and pectase activity.

The accuracy of estimation is measured in terms of correlation coef-
ficients between the in vivo digestibility and the in vitro values of
digestibility. Minson (1981a) stated that the most useful criterion for
comparing the laboratory methods is the residual standard deviation
(RSD) of the regressions. He further stated that RSD is the minimum
error that must always be applied to any in vivo digestibility value
predicted from the regression.

Relationship between forage intake and digestibility. The diges-
tion and passage of forage through the animal depends upon its chemical
and physical breakdown. From the factors listed by Abrams ( 1 980 ) , the
rate of particle reduction is the only factor that may exert the most
influence over the other factors. It is also the one factor where most
overlapping occurs between plant and animal effects. Much attention
has been focused on the chemical aspects of forage degradation and very
little to the mechanical counterpart. The latter involves chewing.

22

rumen motility, and rumination which may directly affect the rate of
passage and digestion. This is illustrated with the results of Blaxter
et al. (1956) who reported that rate of passage increased as particle
size decreased, regardless of feeding level. In the same study, they
found that digestibility of dry matter also decreased as particle size
decreased, attributing it to lower retention time. This is an example
of the continuous process occurring in the reticulo-rumen, where par-
ticles from a piece of forage may have different digestibilities due to
reduced or increased fermentation times. Rodrigue and Allen (1960) ob-
served reduced digestibility of cellulose and/or crude fiber, in relation
to fineness of grinding. They concluded that the finer the hay was ground,
the greater was the depression in digestibility of the total ration and
the faster the rate of excretion. Numerous examples show that for both
temperate and tropical forages, digestibility and intake show a variable
degree of correlation (Minson, 1972; Moore and Mott, 1973; Ventura et al.,
1975; Chaves, 1979; Abrams, 1980). This means that forages with low di-
gestibility may have higher intake than forages with high digestibility.

A number of studies have demonstrated that for a given forage
species, the leaf fraction is more digestible than the stem fraction
(Clements et al . , 1970; Hacker, 1971; Laredo and Minson, 1975; Hacker and
Minson, 1981). Therefore, if digestibility affects intake, it would be
expected to find higher intake of leaf than stems. Some in vivo trials
with sheep have proved this hypothesis right; however, the digestibility
of the stem fraction has been slightly higher or equal to the leaf
(Laredo and Minson, 1973; 1975). The same results were obtained with
sheep and cattle (Poppi et al . , 1981a, b). These studies explained
the differences in intake in terms of retention time and rate of

I

23

excretion. If digestibility is the same but retention time is not,
then the rate of particle size reduction must be increased. In vitro
observations indicate that longer fermentation times result in higher
digestibilities; also, the rate of digestion can be increased by reduc-
ing the particle size (Troelsen and Han el , 1966; McLeod and Minson,
1969; Laredo and Minson, 1975). The study by Poppi et al . (1981a)
showed that the digestibility of the fiber (NDF) was lower in the leaf
than in the stem fraction. All these observations indicate that the
rate of particle-size reduction is a more important factor than di-
gestibility in the control of forage intake. Furthermore, they sug-
gest that the rate of particle-size reduction will determine the ex-
tent of digestion and hence digestibility. Histological observations
(Akin et al., 1977; Hanna et al., 1976; Akin and Burdick, 1975) of
different forages indicate that there are marked differences in the
extent of digestion in the different tissues. It is possible that the
disappearance of the most digestible cells, in combination with mechan-
ical action of the animal, increase the fragility of the tissue and
stimulate the particle reduction.

It appears that more research is needed to gain understanding in
the animal mechanisms that regulate forage intake. In the vegetal
counterpart, it is necessary to understand the changes that occur in
the plants which alter digestibility. It should also be pointed out
that the in vitro digestibility must be used cautiously and should not
be used alone as the only parameter of forage quality. Therefore,
with the level of understanding of the factors of forage quality, it
would be safer to base forage evaluation on in vitro digestibility.

24

mineral balance, and leafiness plus all the desirable agronomic charac-
teristics (Minson, 1981b).

■j

CHAPTER II

GROWTH ANALYSIS OF AERIAL AND UNDERGROUND PLANT PARTS
OF FLORI GRAZE RHIZOMA PEANUT

Introduction

Growth analysis is a tool to assist in the interpretation of morpho
physiological changes occurring in plants throughout their life cycles.
It is based on the mathematical description of the growth of a plant,
either as a whole or as its components. Several concepts of growth
analysis have been proposed and were summarized by Radford (1967) and
Evans (1972). In general, all concepts are analyzed as a function of
time with the purpose of studying the seasonal growth which in turn is
a function of the environment. The selection of the form of analysis
depends on the objectives of the study. For plant breeding purposes,
Snyder and Carlson (1978) used the ratio of taproot to leaf weight in
sugarbeet (Beta vulgaris L.) to select for higher taproot yield. Dun-
can and Hesketh (1968) found genetic differences in net photosynthetic
rates of different races of maize (Zea mays L.) with respect to tem-
perature regimes. Comparing several peanut (Arachis hypogaea L.) cul-
tivars, Duncan et al . (1978) suggested that the yield of cultivar 'Early
Bunch' could be increased by improving crop growth rate whereas that of
'Florunner' could be increased by improving partitioning. McCollum
(1978) studied the effect of P -fertilizer regimes on the growth and
yield of potato (Solanum tuberosum L.)

25

26

Milthorpe and Moorby (1979) suggested that fitting growth data
into curves, for comparison purposes, gives a greater overall understand-
ing of the effect of the experimental variables. They also pointed
out that the experimenter must keep in mind that drastic changes in
growth caused by changes in the environment may not show up in the
fitted curves. Radford (1967) indicated that fitting curves with regres-
sion has the advantage of pooling errors inherent in sampling plants
grown in the field.

Rhizoma peanut, one of several perennial species in the genus
Arachis, has shown considerable potential for improvement of livestock
production in North and Central Florida. The most outstanding selec-
tion was released by the University of Florida in 1979 with the cultivar
named 'Flori graze' (Prine et al., 1981). The rhizoma tous habit of the
plant makes it able to withstand grazing. A serious disadvantage of
Florigraze rhizoma peanut is the lack of seed production; therefore,
its establishment must be achieved by vegetative propagation. The most
rapid method is by planting rhizome pieces late in the winter (Prine et
al., 1981). There is little information regarding the seasonal growth
and development of the aerial parts in relation to the underground rhi-
zome system.

In North and Central Florida, the aerial part dies in the winter
whereas the rhizome system remains dormant and initiates growth early
in the spring. Since both perennation and persistence depend on the
rhizome part of the plant, it is necessary to understand the dynamics
of growth in the plant. With this in mind, the objectives of the pres-
ent study were to describe the growth pattern of Florigraze rhizoma

27

peanut and the proportion of plant parts relative to the recovered
plant biomass under field conditions.

Materials and Methods

The experiments were carried out in 1980 and 1981 in two areas of
land (A and B) of approximately 0.2 ha each, at the University of Florida
Agronomy Department's "Green Acres" farm located about 18 km west of
Gainesville, FL. The experiment in field A was conducted for 2 years,
starting in 1980, whereas the one in field B was conducted only during
1981. The soil type in both areas was Arredondo loamy sand (loamy,
siliceous, hyperthermic, Grossarenic Paleudult). In 1980, preparation
of the land area A included harrowing and disking; treatment with 2.3
kg active ingredient (ai) ha'^ of vernolate (S-propyl dipropylthiocar-
bamate) and 3.0 kg ai ha"^ of benfluralin (N-butyl-N-ethyl-a,a,a-
trifluoro-2,6-dinitro-p-toluidine) for preemergence control of nutsedges
(Cyperus spp) and annual grasses (Diqitaria sanquinalis L. and Cenchrus
pauciflorus, Benth.) respectively. The soil was fertilized in January
1980 with approximately 52 kg of P, 200 kg of K, and 2 kg of minor
elements (FTE 503 mix) per ha. Rhizome material was planted on 27 Feb.
1980 in hills every 2 m in rows 3 m apart. Each hill was composed of
three to five rhizome pieces, to make up a weight of 30±2 g (fresh
weight). The rhizome material used for the planting was dug with a
potato digger (Prine et al . , 1981). Each rhizome piece was of approxi-
mately the same length (25 to 30 cm) and diameter (3 to 4 mn) in an at-
tempt to maintain morphological uniformity. Rhizomes were planted at
about 7-cm depth and were inoculated with a commercial granular peanut

28

inoculum by sprinkling a few mg of granules over the rhizomes prior to
covering them with soil. Weed control during 1980 included spraying with
0.96 kg ai ha'^ of bentazon (3-(l-methylethyl )-lH-2,l ,3-benzothiadiazin-
4(3H)-one 2,2-dioxide) for control of broadleaf weeds 1 month after
emergence (April 1980). After the April spraying, hand weeding was done
in and around the rhizoma peanut plants to allow the use of a rotary
mower between rows without physical damage to the plants. The spaces
between plants were mowed in June and August 1980. Field A was uniformly
mowed in January 1981 to eliminate dead material. In 1981, field A was
sprayed with 1.15 kg ai ha"^ of alachlor {2-chloro-2' .S^diethyl-N-
Cmethoxymethyl )-acetanilide) and 0.34 kg ai ha"^ of 2,4-DB (4-(2,4-
dichlorophenoxy) butyric acid) shortly after emergence of rhizoma peanut
(late March) to control broadleaf weeds. Mowing between the plants was
the only weed control practiced in field A during the growing season in
1981 .

Field B was planted in 1981 as a duplication of the 1980 planting
in field A. In field B the land area was rototilled on 15 Dec. 1980.
On 20 Feb. 1981 the area was sprayed with 2.0 kg ai ha'^ of vernolate
and 0.75 kg ai ha"^ of trifluralin (a,a,a-Trifluoro-2,6-dinitro-N, N-
dipropyl-p-toluidine). The herbicides were incorporated with a harrow
imnediately after application. On 17 Feb. 1981 the soil was fertilized
with 18 kg of P and 67 kg of K per ha. Rhizomes were planted in hills
on 4 March 1981 with distribution of hills, selection, planting and inoc-
ulation of rhizomes similar to the 1980 planting on field A, except that
digging of rhizomes was performed with a bermudagrass sprig digger
(Prine et al . , 1981). In field B, in 1981, weed control consisted of

29

postemergence application of 0.34 kg ai ha"' of 2,4-DB to control broad-
leaf weeds and spot applications of ghyphosate (isopropyl -amine salt of
N-(phosphonomethyl) glycine) to eliminate perennial and annual grasses.

In both 1980 and 1981, plant spacing was designed to ensure that
all plants had enough land to grow without competition from the adjacent
plants and to be able to differentiate one plant from another. The center
of the plants was marked with a wooden stake at planting, A sampling
schedule (Table 1) was arranged for both years. In field A the 1980 sam-
pling was termed 1980 growth (or data set I) and the 1981 sampling was
termed 1981 regrowth (or data set II). The 1981 sampling in field B was
termed 1981 growth (or data set III). During 1980 in field A, 10 randomly
selected hills were dug at each sampling period. The randomization was
restricted to one half of the plants, digging in alternate rows and col-
umns such that at the end of 1980 one half of the initial number of
plants were left in the field with a spacing of 4 x 3 m between plants.
In 1981, in both fields, the number of hills per sampling was reduced to
seven. The plants were dug by loosening the soil with a garden fork,
around and within the hill; then from the center of the hill outward the
rhizomes were lifted carefully by hand following its direction. One
radial section was dug at a time to avoid breaking the hill into pieces.
After digging the hills (which remained intact), the sand was shaken off
and all foreign plant material was removed. When digging was completed,
the plants were carried to a nearby water faucet and carefully, but thor-
oughly, washed on perforated aluminum trays to remove any soil adhering
to the plants. After washing, each plant sample was separated into
shoots and rhizomes, and the number of shoots recorded. The two

30

Table 1. Sampling schedule for three data sets to record the growth and
regrowth of Florigraze rhizoma peanut.

Data sets

Field A

Field B

1980 Growth (I) 1981 Regrowth (II)

1981 Growth (III)

Sampling Dates

28 Feb (

59)"^

3

Mar (62)

4 Apr (

94)

27

Mar ( 86)

2 May (

122)

23 May (

143)

20

May (140)

20

May (140)

6 Jun

[157)

8

Jun (159)

9

Jun (160)

20 Jun

:i71)

26

Jun (177)

4 Jul

[185)

19 Jul

(200)

15

Jul (196)

9

Jul (190)

1 Aug

(213)

3

Aug (215)

7

Aug (219)

15 Aug

(227)

4 Sept

(247)

1

Sept (244)

5

Sept (248)

26 Sept

(269)

3

Oct (276)

2

Oct (275)

17 Oct

(290)

7 Nov

(311)

1

Nov (305)

31

Oct (304)

11 Dec

(345)

27

Nov (331)

26

Nov (330)

Julian days.

31

sections were dried at 70 C until weight equilibration (about 36 hours).
When dried, the sections were weighed and each one further separated into
two parts — shoots into leaf and stem and rhizomes into primary and second-
ary rhizomes — and their weight recorded. Secondary rhizomes were con-
sidered those cylindrical in form, rather thin (less than 3 inn in diam-
eter), whereas primary rhizomes are conical and show a thickening from
where branches are produced. After the third sampling period of the 1981
regrowth in field A, a plant subsample was taken from the hill according
to the following procedure: the initial rhizome pieces, planted the year
before, had developed into taproot-like structures from 10 to 15 mm in
diameter. Therefore, by untangling one of these rhizomes from the re-
mainder of the hill, it was possible to recover all its branches and ac-
ceptably maintain the proportion of shoots to rhizomes. This subsample
was then processed as previously described; the remainder of the plant
sample was freed from other plant materials, washed, dried, and weighed
to record total growth and to estimate the total growth of the different
parts. The 1981 growth in field B was processed like 1980 growth in
field A.

Total growth and growth of the components were fitted to an exponen-
tial function of the form

W(t) = W e^^ ^ ""^^ (1)

where:

W(t) = Predicted dry weight (g)
W = observed dry weight (g)
t = days
e = 2.78

32

a,b,c = linear, quadratic, and cubic coefficients.

Comparisons between growth and regrowth curves were performed with
General Linear Model procedure of SAS (SAS Institute Inc., 1982) using
the logarithmic form of equation (1)

Ln W (t) = Ln W + at + bt^ + ct^ (2)
and data set as class variable to test differences in linear, quadratic,
and cubic coefficients among curves. The first partial derivative of equa-
tion (1) was used to estimate growth rates (g hill"^ day" ) in total
growth and growth of components (Radford, 1967). The first partial deriv-
ative of equation (2) was used to estimate relative growth rates (g g of
hill"^ day'b also for total growth and growth of components (Radford,
1967).

Results and Discussion

The 1980 growth (data set I), 1981 regrowth (data set II) and 1981
growth (data set III) were plotted against time (Fig. 1). Prediction
equations for the data-set curves are shown in Table 2. The regression
analyses to compare data sets of all plant fractions is presented in
Table 3. To facilitate understanding, the results will be discussed by
individual plant fractions and total growth.

Shoot Growth

The shoot growth presented a growth curve characterized by a rapid
increase in growth from June to September (Figs. lA, IB, IC) followed by
a plateau which maintained until late October. The comparisons among
shoot growth curves (Table 3) indicated that all curves were different
from each other especially with respect to their linear coefficients.

33

Table 2. Mathematical equations to describe the growth; of Florigraze
rhizoma peanut components in three data sets.

Coefficients^

Timet

Time

2

Time'

Total growth

E-4§

E-4

E-4

1980 growth

1981 regrowth
1981 growth

(I)

(II)

(III)

3.8166
6.6244
3.4934

-0.0413*
-0.0168*
-0.0252*

3.29
1.31
1.76

-5.576
-2.202
-2.476

E-7*

E-7

E-7

Shoot growth

1980 growth

1981 regrowth
1981 growth

(I)

(II)

(III)

-6.2341
8.7398
-7.5550

0.0822*
-0.0707*
0.0890*

-1.74
4.16
-2.21

E-4*
E-4*
E-4*

8.360
-6.780
1.910

E-8

E-7*

E-7

Rhizome growth

1980 growth

1981 regrowth
1981 growth

(I)

(II)

(III)

5.0050
7.2453
8.0344

-0.0644*
-0.0235*
-0.0799*

4.07
1.20
3.48

E-4*

E-4

E-4

-6.231
-1 .564
-4.039

E-7*

E-7

E-7

Number of shoots

1980 growth

1981 regrowth
1981 growth

(I)

(II)

(III)

8.4953
8.0961
-4.9340

-0.0794*
-0.0172

o.noi*

4.13
6.04
-4.74

E-4
E-5
E-4

-6.191
-7.126
6.804

E-7*

E-8

E-7

Leaf growth

1980 growth

1981 regrowth
1981 growth

(I)

(II)

(III)

-.2781
7.4579
-7.9557

2.47 E-3*
-0.0552*
0.0904*

1.47
3.39
-2.29

E-4*
E-4*
E-4*

-3.378
-5.678
1.982

E-7*
E-7*
E-7

Stem growth

1980 growth

1981 regrowth
1981 growth

(I)

(II)

(III)

-6.6587
9.3843
-3.4561

0.0731 *
-0.1024*
0.0821 *

-1.37
5.75
-1.84

E-4*
E-4*
E-4*

3.951
-9.110
1.407

E-7
E-7
E-7

Primary rhizomes

1980 growth

1981 regrowth
1981 growth

(I)

(II)

(III)

4.0790
5.9507
5.1432

-0.0610*
-0.0412*
-0.0502*

3.71
2.41
2.09

E-4*

E-4

E-4

-5.649
-3.500
-2.011

E-7*

E-7

E-7

Secondary rhizomes

1980 growth

1981 regrowth
1981 growth

(I)

(II)

(III)

4.2245
6.7326
8.2322

-0.0616
-0.0159
-0.0912

4.01
0.75
4.01

E-4*

E-4

E-4

-6.176
-0.909
-4.821

E-7*

E-7

E-7

Linear, quadratic and cubic
♦Significant (P > 0.05).

ays

§1Q-x

34

Table 3. Regression analyses of complete data and individual comparisons
between data sets for growth components of Florigraze rhizoma
peanut.

Data set comparisons

Plant part Data sets Data set I Data set I Data set II

I, II, III vs vs vs

Data set II Data set III Data set III

PR = F§

Total Growth

Data set+
Data set X time*
Data set X time^
Data set x time^

0.0001
0.0001
0.0461
0.2261

0.0001
0.0001
0.6063
0.0980

0.0001
0.0439
0.0036
0.3959

0.0001
0.0001
0.4362
0.9583

Shoot Growth

Data set
Data set x time
Data set X time^
Data set xtime^

0.0001
0.0001
.0.1954
0.3656

0.0001
0.0001
0.0768
0.1385

0.0001
0.0528
0.2836
0.8092

0.0001
0.0001
0.5973
0.2909

Rhizome Growth

Data set

Data set X time

Data set xtime^

Ua La ocL A u 1 iHc

0.0001
0.0001
0.0008

U. U060

0.0001
0.0001
0.9691

0.0001
0.0001
0.0001

0.0001
0.0002
0.1042

Number of Shoots

Data set
Data set X time
Data set x time^
Data set x time

0.0001
0.0001
0.4984
0.0032

0.0001
0.0001
0.1712
0.2211

0.0001
0.0022
0.3703
0.0002

0.0001
0.0043
0.9384
0.2979

Leaf

Data set
Data set X time
Data set x time^
Data set x time-^

0.0001
0.0001
0.6652
0.4995

0.0001
0.0001
0.3934
0.6955

0.0001
0.0496
0.9195
0.3099

0.0001
0.0001
0.5287
0.3447

Stem

Data set
Data set x time
Data set x time^
Data set x time-^

0.0001
0.0001
0.5813
0.2712

0.0001
0.0027
0.3053
0.1043

0.0001
0.0090
0.4101
0.8469

0.0001
0.0002
0.7254
0.2183

35

Table 3 - continued.

Data set comparisons

Plant part Data sets Data set I Data set I Data set II

I II III vs vs vs

Data set II Data set III Data set III

PR = F

Primary Rhizome

Data set 0.0001 0.0001 0.0005 0.0001

Data setx time 0.0003 0.0001 0.0065 0.5459

Data setx time2 0.0002 0.9122 0.0001 0.0268

Data setx time3 0.3501 0.2528 0.2769 0.7503

Secondary Rhizome

Data set 0.0001 0.0001 0.0001 0.0001

Data setx time 0.0001 0.0001 0.0001 0.0001

Data setx time2 0.0030 0.8546 0.0001 0.1448

Data setxtime3 0.0831 0.0196 0.7404 0.5126

^Intercept
^Time in days
§Significance level

36

Since a logarithmic model was employed to compare the curves, a
significant difference among quadratic coefficients would suggest differ-
ences in relative growth rates (Figs. 3A, 3B, 3C). This is so. because
to obtain the first partial derivative of the growth curve, the linear
coefficient becomes the intercept of relative growth rate curve; also,
two times the quadratic coefficient becomes the linear coefficient in
the relative growth rate curve, and so on. In 1980 growth (I) and 1981
growth (III) the quadratic coefficients were negative (Table 2) while in
1981 regrowth (II) it was positive. Although among quadratic coefficient
there was a difference only between 1980 growth (I) and 1981 regrowth
(II), the relative growth rates can be ranked in the following order:
1980 growth, 1981 growth, and 1981 regrowth. At the time of the last
sampling, the shoot growth of 1981 growth (III) (Fig. IC) was still on
the plateau; in the same year but in separate fields, this was clearly
not the case of the 1981 regrowth (II). The plant canopies in 1980 growth
(I) and 1981 regrowth (II) were very dense and began to show signs of
senescence in October, particularly at the base of old shoots where light
was probably limiting. In other temperate and tropical species, it has
been shown that photosynthesis is lower at the base of a canopy (Brown
and Blaser, 1968) because less direct sunlight penetrates to the base of
closed canopies (Ludlow and Wilson, 1970). The 1981 growth (III) plants
did not close the canopy; in fact, they remained rather prostrate. No leaf
senescence or leaf loss was observed in spite of low minimum temperatures
(Fig. 5B) recorded in November 1981. In contrast, frost damage was ob-
served in the plants of 1981 regrowth (II); there was no explanation to
this, since the fields were about 600 m apart.

37

The shoot growth was closely associated with the number of shoots
(r^ = 0.98, 0.63, and 0.88 for 1980 growth (I), 1981 regrowth (II). and
1981 growth (III), respectively). These are plotted against time in
Fig. 4A. The weight of individual shoots also increased with time. The
shoot growth rates increased until September in all three data sets
(Figs. 2A, 2B, 2C). The shoot production rates (Fig. 4B) also increased
for the 1980 growth (I) and 1981 regrowth (II) until September whereas
the 1981 growth (III) reached its minimum shoot production rate in Sep-
tember and started to increase from then on. These results suggest
that after September, plants started to reduce number of shoots and the
individual shoot weight. In the case of 1981 growth (III), the increase
in total shoot weight was due to the increasing number of shoots and not
to an increase in the weight of individual shoots. Nineteen eighty-one
was a dry year, and a severe drought occurred until the end of June
(Fig. 5C). In the 1981 growth (III), newly formed shoots died during
the drought, leaving only the old ones. These old shoots branched abun-
dantly after the drought and formed a thick stem. This branching occurred
from lateral buds on the stem, as opposed to the new shoots which origi-
nate from buds on the rhizomes. This helps to explain the decreasing
rates of shoot production in the 1981 growth (Fig. 48) where the canopy
was formed as a result of shoot branching. During the drought it was ob-
served that rhizomes produced numerous roots and no shoots. A response
like this was not observed in 1981 regrowth (II), probably because old
primary rhizomes were very well rooted; although shoot production
was affected (Fig. 4A). This was reinitiated at an increasing rate
after the dry period.

38

When looking at the growth of leaves and stems (Figs. ID, IE, IF),
it can be noticed that the leaf fraction is the shoot component that
declines earlier in the year, while the stem fractions remain high for
a longer period. The statistical comparisons between coefficients (Table
3) gave the same results as shoots. Curves depicted in Figs. 2D, 2E, and
2F showed sharper decline in growth rates (September) in leaf fractions
than in stem fractions. The declining rates of growth in the leaf
fraction were related to the size and structure of the plant canopies.
The highest rate of decline was observed in 1981 regrowth (II) followed
by 1980 growth (I) and 1981 growth (III). Even though leaf area was not
measured, it was observed that plant canopies were fuller in 1980 growth
(I) or 1981 regrowth (II) than in 1981 growth (III). Probably, a full
canopy was reached, causing a steep decline in leaf growth rates, as
opposed to the 1981 growth (III) where the rate of decline was very small.
The rate of increase of shoot numbers in the 1981 growth (III) partially
accounts for the slow decline in leaf and stem growth rates, although
the size of shoots did not increase after September. Early in the year,
when shoots started to emerge, stems were heavier than leaves, but in 2
to 3 weeks, the weight of the leaves accounted for up to 75% of the weight
of the shoots. As the plants grew older, the leaf fraction decreased
steadily due to leaf loss and to stem development. Furthermore, the
shoots produced later in the year had very small leaves, and even though
they were counted as shoots, the leaf mass was not greatly increased.
The fact that stem growth rates did not decline at the same time that the
leaf growth rates, may have been due to a slower rate of decomposition in
the stems, which means that after a shoot looses its leaves, the stem

39

remains longer. In this experiment, dead shoots were not counted, but
shoots which remained green with at least one leaf were counted. In
corn, Williams et al . (1965) reported that crop growth rates increased
with increasing plant populations with concurrent reduction in the size
of individual plants. Similar results were reported by Knight (1970) for
orchardgrass (Dactyl is glomerata L.). where he found that removal of
tillers enhanced the development of others, making up for the space left.
In this context, with respect to the ranking of relative rates of shoot
production (Fig. 4C) and relative growth rates of shoots (Figs. 3A, 3B,
and 3C), it appears that low shoot growth in 1981 regrawth (II) was due
to a low rate of shoot production. The 1981 regrowth (II) showed an in-
creasing relative growth rate of shoots reaching a maximum in August, and
declining thereafter. The reason for this was that the emergence of
shoots did not occur at the same time; therefore, growth was not a result
of development of the initial shoots, but the result of shoot growth and
shoot emergence.

Rhizome Growth

As opposed to shoot growth which reached a plateau in September,
rhizome growth continued at least until the last samples were taken at
the end of the year (Figs. lA, IB, and IC for 1980 growth (I), 1981 re-
growth (II), and 1981 growth (III), respectively). The regression anal-
yses (Table 3) showed differences between rhizome growth curves, partic-
ularly between linear coefficients (Table 2), suggesting differences in
growth rates. There were no differences among quadratic coefficients,
except between 1980 growth (II) and 1981 growth (III), but the linear
coefficients determined the differences in relative growth rates

40

(Figs. 3A, 3B, and 3C). The ranking in relative growth rates can be
determined with the quadratic coefficients and it follows the same trend
as shoot relative growth rate. Even though the linear coefficient of
1981 regrowth (II) was the smallest, it ranked last because the quadratic
coefficient (Table 2) was also the smallest and did not counteract the
effect of the linear coefficient. The relative growth rates of rhizomes
in 1980 growth (I) (Fig. 3A) almost doubled those of 1980 regrowth (II)
(Fig. 3B) with the ones of 1981 growth (III) (Fig. 3C) in the middle.
Rhizomes grew almost parallel to shoots from the beginning of the year
until September. It was previously believed that rhizomes grew only in
the fall (G. M. Prine, 1980. Personal communication). The results of
these studies indicate that shoots stop growing early in the fall and
rhizomes continue to grow, indicating an increase in partitioning of as-
similates towards rhizome growth. This phenomenon is clearly illustrated
with the growth rate curves (Figs. 2A, 2B, and 2C) where the growth rates
of shoots are declining in September while rhizome growth rates either
continue to increase (1980 growth and 1981 growth) or are reaching a
plateau (1981 regrowth). Throughout the year rhizomes were continuously
produced, but in September of both years it was more evident; probably
because fewer shoots were produced and it changed the proportion of aerial
and underground parts. Figures 2A, 2B and 2C also show that growth rates
of rhizomes in 1980 growth (I) and 1981 growth (III) were comparatively
higher than shoots. This did not occur in 1981 regrowth (II). Various
degrees of assimilate partitioning have been reported for plants with
determinate growth. The variation can be genetic (Duncan et al . , 1978;
Snyder and Carlson, 1978) or can be caused by nutrient imbalances that.

41

in turn, cause alteration of 'sinks" (McCollum, 1978). Rhizoma peanut
is perennial and day neutral plant. In this experiment it flowered all
year and very few pods were found. Most pods were empty, although some .
were found in different stages of pod filling, indicating that pod fill
was completed early in October. With the observations of these experi-
ments it is not possible to determine whether there was a physiological
change that made rhizomes a stronger sink for assimilates, or if it
was only induced by reduction in growth in the aerial part.

From the rhizomes, two fractions were separated, primary and second-
ary rhizomes. In Figs. ID, IE, and IF it can be seen that secondary rhi-
zomes constituted the bulk of the rhizome network and the largest compo-
nent of the entire plant. The secondary rhizomes (cylindrical) seemed
to play a role of spreading the plant and maintaining it as one unit
(Fig. 6A). Most nodules were found on secondary rhizomes and small roots
associated with them (Fig. 6B). On the other hand, primary rhizomes
(conical) developed from secondary rhizomes at points where shoots sur-
face out. These conical structures were the point of origin of new
rhizomes and shoots (Fig. 6C). Rhizomes grew radially, from the center,
occasionally some were observed growing perpendicularly to the center,
but almost never towards the center. In fact, in the 1981 regrowth (II)
some plants started to overlap and it was easy to determine the origin of
a rhizome by determining the direction of its growth. Any rhizome pro-
duced will grow several centimeters and its tip will emerge as a shoot
(Fig. 6A). As the shoot grows, a primary rhizome is produced, but only
if the shoot develops. Otherwise, the rhizome will branch under the
ground in one or more directions. These branchings do not produce primary

42

rhizomes. Primary rhizomes thicken according to the number of branches
produced from this point. Roots also develop from these points. Most
roots were found associated with primary rhizomes but not all primary
rhizomes produced deep roots. In the 1980 growth (I) the ratio of second-
ary to primary rhizomes increased as the growing season progressed; the
same trend was observed, to a lesser extent, in the 1981 growth (III).
In the 1981 regrowth (II) the ratio was maintained rather constant
throughout the year. The growth rates (Figs. 2D, 2E, and 2F) for both
rhizome fractions presented the same trends as the growth curves; which
means that in 1981 growth (I) and 1981 growth (II) the growth rates of
secondary rhizomes increased faster and were much higher than in primary
rhizomes. In the 1981 regrowth (II), growth rates of both fractions were
very similar, indicating that more primary rhizomes were produced during
the regrowth phase than during growth. This was further confirmed by
the relative growth rates, which were higher for the primary rhizomes.
Data in Table 3 show, for primary rhizomes, that the quadratic coeffi-
cient of 1981 growth (III) was different from both 1980 growth (I) and
1981 regrowth (II) (P = 0.0001 and P = 0.0268, respectively). These
results indicated that relative growth rates for primary rhizomes in 1980
growth (I) and 1981 regrowth (II) were very similar and at the same time
higher than in 1981 growth (III). Conversely, for secondary rhizomes,
Table 3 shows differences in linear and quadratic coefficients between
1980 growth (I) and 1981 growth (II), (Table 2) and differences in linear
and cubic coefficients between 1980 growth (I) and 1981 regrowth (II).
These results indicated differences in relative growth rates, with the
highest for 1980 growth (I) followed by 1981 growth (III) and 1981

43

regrowth (II). These results can be visualized in Figs. 3D, 3E and 3F;
in 1980 growth (I) and 1981 growth (III) relative growth rates were
higher for secondary rhizomes than for primary. The reverse effect was
observed in the 1981 regrowth (II).

The reduction in relative growth rates of secondary rhizomes, along
with the reduction in relative rates of shoot production, in the 1981
regrowth (II), strongly suggested that efficiency of both plant propaga-
tion and production was lost in the second year. On the other hand, the
increase in primary rhizomes may indicate that an equilibrium is reached
although rhizome growth was reduced.

Total Growth

Total growth comprised shoot and rhizome fractions (Figs. lA, IB,
and IC) and it was characterized by a rapid increase from June until
October, when top growth reduced its growth rates. A regression analysis
including all three data sets (Table 3) showed differences among inter-
cepts and among time (linear) and time (quadratic) coefficients of the
curves. However, the individual comparisons showed only differences in
quadratic coefficients between 1980 growth (I) and 1981 growth (III)
(P = 0.0036); indicating that sets reached their plateau at different
times. The cubic coefficients were not different (P = 0.2261) among
total growth curves, indicating that a decline in total growth occurred
approximately at the same time, although in the 1981 growth (II) the
inflexion point did not occur within the observed range of time. Nine-
teen eighty-one was very dry, and a drought was experienced until the
end of June. Consequently, 1981 growth (III) (Fig. IC) was delayed until
July, and thereafter rhizoma peanut grew at a nearly linear rate until

1

44

at least 27 November when the last samples were dug. The drought effect
is seen in both 1981 regrowth (II) (Fig. IB) and 1981 growth (III)
(Fig. IC), in the solid line at the beginning of the growing season.
It occurred earlier in the 1981 regrowth (II) because the shoots emerged
about 3 weeks before the 1981 growth (III). In addition, the plants were
much larger and occupied more space; therefore, water in the upper profile
was probably depleted and the plants showed the effect. Conversely, the
plants of the 1981 growth (III) had almost no competition and they were
very small. Drought stress was observed in some plants after it started
raining at the end of June. The amount of rainfall was not enough for
the incipient root system of young plants in 1981 growth (III), but it
was enough for the well -rooted plants in the 1981 regrowth (II). The num-
ber of shoots (Fig. 4A) also followed the same trend as that of total
weight, indicating that the early season decrease in both 1981 sets were
associated with losses in shoot numbers. This confirmed the observations
in the field of shoots dying during and after the drought.

Maximum growth rates for total growth occurred between mid-September
and mid-October (Figs. 2A, 2B, and 2C), suggesting a decline in photo-
synthetic activity. The decline in growth rates coincided with the sea-
sonal decline in temperature (Figs. 5A and 5B); however, other effects
must have been associated with it because high growth rates in 1981
growth (III) (Fig. 2C) were maintained over a longer period of time.
It was indicated that the structure of the canopy could have affected
photosynthesis. An illustration of this can be found in growth rates of
shoots (Figs. 2A, 28, and 2C). In 1980 growth (I) and 1981 growth (III),
the decline in total growth rates was closely associated with the decline

45

in rhizome growth rates. In the 1981 regrowth (11) the decline was as-
sociated with the declining rates of shoot growth, indicating that net
assimilation was reduced. It has been reported that Pangola digitgrass
reduces photosynthesis with night temperatures around 10 C (West, 1969)
particularly if low temperature occurs for 3 consecutive days (Karbassi
etal., 1972. In Siratro (Macroptilum atropurpureum [D.C.] Urb.) very
little growth occurs below 21 C and 14 C, maximum and minimum, respec-
tively (Jones, 1967; Whiteman and Lutham, 1970). It seems that rhizome
peanut is sensitive to the decline in temperature that occurs in September
and October, and reduces shoot growth. Rhizome growth is totally depen-
dent on the supply of sugars from the shoots; and soil temperature does
not decline as fast as air temperature. Therefore, rhizome growth will
continue for as long as the shoots remain photosynthetically active.
In these experiments, the foliage was not disturbed, so it was thought
that the age of the shoots could decrease photosynthesis regardless of
the structure of the canopies, resulting in a reduction of shoot growth
rates due to constant respiration. Prine et al . (1981) provided data
that refutes this hypothesis. In one of their experiments, Florigraze
rhizoma peanut was harvested three times in 2 consecutive years. In
both years, the third harvest which represented growth from mid-August
to mid-October produced 50% less than the previous two harvests, indi-
cating that indeed the plant reduces its shoot growth in September. To
understand further, the differences among growth curves, it was necessary
to examine the relative growth rates because the obvious differences in
plant size. Data in Table 3 show that for total growth, the quadratic
coefficients of 1980 growth (I) and 1981 growth (III) were different

46

(P = 0.0036) indicating differences in relative growth rates. The
quadratic coefficient for 1980 growth (I) was almost double the coeffi-
cients of either 1981 regrowth (II) or 1981 growth (III), and despite a
low intercept, it was translated into higher relative growth rates.
Data in Fig. 3A, 3B and 3C show that highest relative growth rates were
achieved in the 1980 growth, followed by 1981 growth and 1981 regrowth.
These resul ts suggest that Florigraze rhizoma peanut grows more effi-
ciently during the first year. The observations in these studies in-
dicate that for a given mass of rhizome material, more rhizomes are
produced in thefirstyear of growth than in the second. Also, more
shoots, relatively speaking, are produced during the first year. What
causes these changes in the proportion of plant components cannot be
explained on the basis of experimental evidence. Visual observations of
plants in September 1982, in their third growing season, indicated that
the rate of spreading decreased every year. Nevertheless, it takes less
time to fill in the canopy when the growing season starts. This makes
the plant competitive with weeds since it starts growing early in April
unless it has just been planted; in newly planted fields, growth will be
started a month or so later.

Summary

Growth analysis is a valuable tool for interpretation of morpholog-
ical changes in plants as a result of environment or treatment effects.
Two field experiments were conducted in 1980 and 1981 to describe the
growth pattern of Florigraze rhizoma peanut and the proportion of plant
parts relative to recovered plant biomass. The experiments were

47

conducted at the University of Florida Agronomy Department's "Green
Acres" farm, 18 km west of Gainesville. Experiment A was conducted for
2 years (1980 growth and 1981 regrowth). Experiment B was conducted
only in 1981 (1981 growth) as a replication of 1980 growth of experiment
A. Ten and seven plants (1980 and 1981, respectively) were dug at all
samplings. Sampling schedule covered the entire growing season in both
experiments. Plants were dug, washed, separated into shoots and rhi-
zomes, and dried. The dry weights were recorded and plant fractions fur-
ther separated — shoots into leaves and stems, and rhizomes into primary
and secondary rhizomes. Data were fitted into logarithmic form of ex-
ponential function and analyzed with regression analyses.

Statistical differences were associated with environmental effects.
The 1981 regrowth had lower relative growth rates than the 1980 and 1981
growth. Shoots and rhizomes grew parallel to each other until Septem-
ber. From then on, shoot growth declined and rhizome growth continued
to increase. The decreasing rates of shoot growth as opposed to the in-
creasing rates of rhizome growth suggested that photosynthates were
being translocated to the rhizomes. The cessation of growth in shoots
with the concurrent growth in rhizomes suggested that differences in
gradient temperature were affecting shoot formation but not rhizome
formation, indicating that photosynthesis had not been stopped.

Secondary rhizomes constitute the bulk of rhizoma peanut bio-
mass. Like total growth, the relative growth rate of secondary rhizomes
was lower in 1981 regrowth than in 1980 or 1981 growth. The same
occurred with the relative rates of shoot production (number of shoots),
indicating a strong relationship between shoots and secondary rhizomes.

48

which confirmed field observations that secondary rhizomes emerged as
shoots. Roots were found associated with primary rhizomes and nodules
with secondary rhizomes. Higher relative growth rates in rhizomes of
1981 growth than in 1981 regrowth indicated that efficiency of propaga-
tion was lost in the second year of growth.

!

49

480

1980 Growth ( 1 )

.•.

m

• " .

400

•• Total
■ ■ Rhizome

320

GO Shoot *

240

1 ou

800

lA

M

A M J J A S

0 N

TIME

1980 Growth ( I)

240 '

•

200 **Leof •

ooSt«m
160 ■■Primary Rhizom* •
2Q ••Secondary " ^

I

80
40

M A M J J A S 0 N

TIME

0».

TIME

1000

1981 Regrowthdl )

800

•

600

•

* A

400 •

0

200 *■

0

■

AOS

1 E

M

A M J J A S

0 N

TIME

■ 160

1981 Growth (HI)

■ 1,20

•

•

' 80

•

• A

■ 40

1 F

M

A M J J A S

0 N

• ■ ■ ■ TIME

Fig. 1. Total growth and components of growth in Florigraze rhizotna
peanut for three data sets.

lA, IB, IC = Total growth, rhizome, and shoot growth.
ID, IE, IF = Leaf, stem, primary rhizome, and secondary
rhizome growth.

50

o

5

1980 Growth( 1 )

4

••Total • -

3

■ ■ Rhizome , ■

2

oo Shoot • '

■
•

1

• 0 a ' 0

• 0 ■ 0

0

■
•

o

A M J J A S 0

0

N

TIME

I

>«

o,

T

1980 Growth ( I )
AA Leaf

oo stem <

2 Primary •
Rhizome •
Secondary •
Rhizome •

0

A M J

A S 0 N

TIME

o

T3

1981 Regrowth(ll)

20
15

10

5

0

• o

• -8 ■

• o

o

o •

2 B o.

AMJ JASON

TIME

a
■o

3
2
I

• 0

1981 Growth (III)

" 2C

AMJJASON

TIME

I

o.

1981 Regrowth(ll)

■0 ■ !b

A M J

* ° •

o ■ »
* • • O

4 °-

2E •

J A S 0 N

TIME

I

o

I

1981 Growth (III)

1.0
-10.5

■ •

•0

;.9iP 2F *^

AMJ J ASON

TIME

Fig. 2. Growth rates of plant components in Flori graze rhizoma
peanut for three data sets.

2A, 2B, and 2C = Total, shoot and rhizome growth rates.
2D, 2E, and 2F = Leaf, stem, primary rhizome, and
secondary rhizome growth rates.

51

a

o •

0.05
0.04
0.03
0.02
0.01
0.0 •
M

1980 Growth 1 1 )

o ••Total

o

°o •■Rhizome

o ^ oo Shoot

■ 3A

O • I

o

A M J J A S 0 N

o
•o

o
at

0.05 *

4

0.04
0.03
0.02 o °
0.01

0.0 ,^

M A

1980 Growth ( I )

* oo stem

*^ ■■Primary Rhiz.
A ••Seconaary Rh,

• ■

• ■
•■

M

3 D

J J A

Oa

0 N

TIME

TIME

o

o
at

ai

0.02
0.01
0.0 •

M

1981 Regrowth(ll)

o o o 0

• •a'

■ o

A M

3B

J A

o •
o •

0 N

r

TIME

o
•a

I

o

ao3

0.02
.01
.0

1981 Regrowth(ll)
. A * * .

3E

M A M J J A

c5*

o*

TIME

S 0

I

o

13

.C.
<^

o
en

1981 Growth (III)

0.04

0.03

0.02

• • • o • • ■

0.01

0.0

■

M

AMJJASON

TIME

Fig. 3. Relative growth rates of plant components in Florigraze
rhizoma peanut for three data sets.

3A, 3B, and 3C = Total, rhizome and shoot relative growth
rates .

3D, 3E, and 3F = Leaf, stem, primary rhizome, and secondary
rhizome relative growth rates.

52

(LIIMOjfiQj) .Xop s;ooL|S

q q o o o

^ (O cvj — d

So'

° •

CD

q in q irt o
cvj — — d d

o

CO

<

-3
<

■

•0

■ *o

■ • 0

• 0

o

• ao

• <■

- o«

<

.LlI

• c

0

0

'l-

■ • 0

■

• 0

z

■

• 0

• °

<

•

fO <M

o o

O O

d d

d d

Aop jooMS spoons

(Lj|MOj59J) Siooqs 'ON

(l|;moj6) s400L|s -qn

4J

<d

■a

a;

a>

i.

+->

(.

3

e

«

o.

£

o

N

c

o

•r~

*•>

U

0)

3

N

-o

10

O

S.

• i.

cn

c a.

T-

o

t- 4->

O

■•-> O

o o

Ll-

(/)

3 ^

C

O

o

O

s- «»-

Q. O

</)

I/)

+J

•t-> (/)

o »*-

O (U

o

o

O ■•->

«o

c

to s.

o

«t-

•t—

o

o >

u

c

3

</>

O -D

0) lO

•r-

o

■M 1—

■t->

s.

U Q. C3£ a£

3

■o

II

II II

o

<C OQ O

Q. ^ ^ ^

53

54

Fig. 6. Primary and secondary rhizomes.

6A - branching in a secondary rhizome;
6B - nodules in secondary rhizome

6C - primary rhizome with shoots and secondary rhizomes.

CHAPTER III

DIFFERENTIAL ACCUMULATION OF TOTAL NONSTRUCTURAL
CARBOHYDRATES AND NITROGEN IN PLANT PARTS OF FLORIGRAZE

RHIZOMA PEANUT

Introduction

It has been shown that forage species have a typical pattern of
total nonstructural carbohydrate (TNC) accumulation and mobilization
(Weinmann, 1948; Smith, 1975, 1981). This pattern can be modified
by the environment and by management practices (Eaton and Ergle, 1948;
Brown et al., 1966). A typical pattern shown a decrease in TNC accumu-
lation in periods of rapid growth (Brown and Blaser, 1965) and an in-
crease in periods when growth is reduced, indicating that photosyn-
thesis is less affected by temperature than is growth (Brown and Blaser,
1968). The nonstructural carbohydrate fraction stored in a particular
plant varies with the forage species (Smith, 1968); therefore,
it is imperative to know what type of carbohydrate is accumulated in
a particular species to choose the proper laboratory analysis. Also,
it is necessary to determine the plant part which accumulates and
stores carbohydrates in order to properly sample a plant.

The persistence of perennial forage species that remain dormant
during the winter depends, in part, on the availability of TNC for
regrowth in the spring (Humphreys, 1966). Nitrogenous compounds are
also stored, and utilized for regrowth (Davidson and Milthorpe, 1966).
The regrowth mechanis;m is a complex one and energy is required to

55

56

activate it until photosynthetically active tissues are developed.
The absence of a ninimum level of TNC may result in poor yields and
persistence because the initial regrowth can be retarded, thus making
the plant less competitive for light, water, and nutrients.

Flori graze rhizoma peanut is a rhizomatous plant introduced into
the United States by W. Archer in 1936 (Prine, 1964) from the state of
Mato Grosso, Brazil (Prine et al., 1981). It is drought tolerant and
is very well adapted to Florida's well -drained soils (Prine et al.,
1981). It is sensitive to low temperatures and above ground
plant tissue is frost-killed in the winter, but the rhizomes remain
dormant until the spring. The main disadvantage of Flori graze rhizoma
peanut is that it does not produce seed, hence, it has to be vegetativ-
ely propagated. Due to the slow rate of establishment, maximum yield
and utilization may take from 2 to 3 years (Adjei and Prine, 1976).
Franca-Dantas (1982) suggested that propagation methods should be
improved or rhizoma peanut should be intercropped during the first
and second years of establishment in order to maintain a profitable
production system.

In order to examine the possibility of improving the propagation
methods in Florigraze rhizoma peanut and to design management systems
to increase the efficiency of utilization, a series of studies was
conducted to gain some understanding in the growth characteristics of
this cultivar. In the previous chapter, results were reported pertain-
ing to morphological changes in Florigraze rhizoma peanut during the
first 2 years of establishment.

57

The objectives of the study reported herein were to determine
the TNC and N accumulation patterns of undisturbed plants. A second
objective was to determine the plant part where carbohydrates accu-
mulate. Since plants keep respiring after being cut or dug, and sampling
in the field usually takes several hours, a third objective in this
study was to determine whether there was an effect of preservation
method, during sample processing, on the level of TNC recovered.

Materials and Methods

Two field experiments in two locations were carried out in 1980
and 1981 at the University of Florida Agronomy Department's "Green
Acres" farm, 18 km west of Gainesville, FL, In one of the locations,
termed field A, an experiment was conducted for 2 years. In this ex-
periment 300 rhizoma peanut hills were planted with rhizome pieces in
February 1980. Approximately 150 hills were dug in 1980 and about
70 in 1981. These samples will be referred to as 1980 growth or data
set (I) and 1981 regrowth or data set (11). In the other location,
or field B, 300 rhizoma peanut hills were planted in March 1981. Approx-
imately 70 hills were dug in 1981. These samples will be referred to
as 1981 growth or data set (III). The details of field design, soil
type, and general practices were described in Chapter II. In 1980
and 1981, 10 and 7 samples, respectively, were taken at all sampling
periods (Table 1, Chapter II) for laboratory plant tissue analyses.
The samples were taken immediately after digging the rhizoma peanut
hills. The procedure for sampling was similar to that previously des-
cribed (Chapter II) to determine the relative percentage, in large

58

hills, of the four plant fractions characterizing the growth. The
procedure consisted of selecting one well rooted primary rhizome, then
two branches of the selected rhizome were untangled from the rhizome
mat. All other branches of that primary rhizome were pruned off at
the base. The sample was quickly separated into shoots and rhizomes.
In 1980, the plant fractions were put in paper bags and cooled by plac-
ing them on crushed ice in a polyure thane ice chest. The ice chest was
constantly drained to prevent samples from soaking up water. In both
the 1981 experiments the samples were separated into two similar sub-
samples. One subsample was cooled on ice as above, and the other was
frozen in situ by placing it in a plastic bag with crushed dry ice.
These plastic bags were stored in another ice chest. In both years,
when digging was completed in the field, the samples were taken to a
laboratory and transferred to dry paper bags. The paper bags with the
samples were immediately placed (standing open) in an oven at 110 C for
20 to 40 minutes, depending on the load. The bags were then transferred
to a forced-air oven set at 60 C where the samples remained for 36 hours.
Once dried, the samples were separated into four plant fractions, as
indicated in Chapter IV and weighed. Leaf and stem fractions were put
back together, leaving three plant fractions (shoots, primary rhizomes,
and secondary rhizomes). Rhizoma peanut hills were categorized into
small, medium, and large according to their weight. For laboratory
analyses, samples were somposited according to the size of the hill
they came from. Samples from small hills were combined into one, as
were samples from medium and large hills. Hence, three composite
samples (small, medium, large) represented every plant fraction at

59

every sampling period. The composite samples were ground in a Udy
cyclone mill to pass a 0.5 mm screen and stored in plastic containers
until analyzed. The samples were analyzed for total nonstructural car-
bohydrate (TNC) following the procedure outlined by Smith (1981) with
the modifications reported by Valle-Melendez (1981) and Christiansen
(1982). Nitrogen analyses were performed with a micro-Kjeldahl , alumi-
num block diqestor and a Technicon Autoanalyzer" (Gallaher et al., 1975)
at the Agronomy Research Support Laboratory of the University of Florida.
Organic matter was determined by placing a 0.5 g sample in a muffle fur-
nace for 5 hours. Nitrogen and TNC concentrations were expressed on
organic matter basis. The data were analyzed by fitting second degree
regression equations, with time as a continuous variable. Seasonal
variation among TNC curves and among N curves was tested by comparing
linear and quadratic regression coefficients. The statistical tests
were performed using General Linear Model (GLM) procedure of SAS (SAS
Institute Inc., 1982) with data set as class variable.

Results and Discussion

Ice vs Dry Ice

Data in Table 4 show the results of the comparison between cooled
and frozen samples. Total nonstructural carbohydrates, as percentage of
organic matter, were slightly higher in frozen samples than in cooled

samples; however, there were no statistical differences between TNC values

2

(P = 0.3940) and no significant time x ice and time x Ice interactions
(P = 0.6662 and P = 0.6089, respectively). These small differences might
have been caused by broken cell membranes in the frozen samples, which

60

Table 4. TNC percentage in cooled (ice) and frozen (dry
of 1981 regrowth (II) and 1981 growth (III) of
rhizoma peanut.

ice) samples
Fl ori graze

Sampling date

Shoots

Ice

Dry
ice

Plant parts

Primary
rhizomes

Ice

Dry
ice

Secondary
rhizomes
Dry
ice

Ice

1981 Regrowth (II)

TNC %

27 March

— t

50.2

56.9

47.6

54.1

20 May

15.1

16.7

28.3

34.6

28.7

33.0

9 June

11.9

14.0

43.4

44.1

44.0

42.7

9 Julv

18.6

19.5

22.5

26.4

21 .4

22.4

1 Aim

14.3

16.8

37.4

36.2

33.9

32.6

11 4

10 7

36.3

38.6

48.0

45.5

16 0

44 2

■t"t •

47.5

52.1

50.8

31 Oct.

15.8

15.1

59.0

61.2

66.1

70.3

26 Nov.

16.5

15.7

66.6

65.8

72.1

72.8

1981 Growth (III)

20 May

-X

9.8

16.0

18.2

8 June

11.0

16.6

16.8

26 June

13.3

11.0

10.6

15 July

6.4

8.6

8.6

10.4

9.9

8.1

3 Aug.

15.6

16.9

16.5

17.1

12.2

17.6

1 Sept.

9.8

8.4

16.7

16.1

17.4

16.9

3 Oct.

9.6

15.1

24.2

29.0

26.6

33.7

1 Nov.

12.1

13.3

37.5

38.1

39.2

40.9

27 Nov.

12.6

15.0

39.8

38.4

42.2

40.1

fNo shoot growth.
INot enough sample.

61

increased the solubility of nonstructural carbohydrates; or it may have
been an actual reduction in respiration. Nevertheless, these results
suggest that very little carbohydrate was lost when the samples were
cooled in crushed ice. Haslemore et al. (1980) used a range of tempera-
tures from -15 C to 40 C in postharvest storage and found very little
effect on N-soluble sugars, and starch analyses. Nevertheless, they
suggested that plant samples should be cooled until they can be dried.
Since the 1980 samples for TNC analysis were cooled, the results from
here on will only include the TNC values obtained with samples cooled
in ice.

Total Nonstructural Carbohydrates

The TNC values as percentage of organic matter were plotted against

time and are shown in Fig. 7 for shoots, primary and secondary rhizomes.

In shoots from all three data sets, TNC varied within a relatively

narrow range from 8 to 19% (Fig. 7A). A regression analysis showed

2

nonsignificant Data set x time and Data set x time interaction effects
(Table 5) or TNC percent in shoots. The curves, then, were very similar
in shape, and although there were peaks and valleys in all of them (Fig.
7A), TNC remained relatively constant throughout the year in every
experiment. However, the intercepts of the curves were different
(P = 0.0077). The intercept of 1981 regrowth (II) was significantly
higher than the intercepts of 1980 growth (I) and 1981 growth (III).
This difference was translated into a higher overall TNC percentage in
the shoots of 1981 regrowth (II) than in the other two data sets. The
trends of TNC in the shoots were not linear, but the increases and
decreases cannot be explained on the basis of diurnal variations. Holt

62

SHOOTS

Ui

u

IT
UJ

a.

(J
z

30
20
10

7A

* 1980 GKMlii

• 1981 R«qRM«i
■ 1981 Graolh

M A M J J A S 0 N 0

TIME

7D

* *• +13870 -ail KTW r*..32

• 211 - .1230 +(2.711050? r»..l4
■ 1 2.0- .012 D + (2.9iKriO' r'< .01

M A

0 N

TIME

PRIMARY RHIZOMES

TIME TIME

SECONDARY RHIZOMES

UJ

u

UJ

a.

u
z

M AM

•99.1 - .730 + 00190* r*-.84
■SS.O - .470 + .00130' r'>.94

MAM

30
20

lOf

TO
60
50
40

30
20
10

TIME

TIME

Fig. 7. Percent TNC (organic matter basis) in plant parts of
Flori graze rhizoma peanut for three data sets.
7A, 7B, 7C = observed data
70, 7E, 7F = estimated data

63

Table 5. Regression analyses of compplete data and individual comparisons
between data sets for TNC percent in plant parts of Flori graze
rhizoma peanut.

Data set comparisons

Plant part Data sets Data set I Data set I Data Set II

I II III vs vs vs

Data set II Data set III Data set III

PR = Ft

Shoots

Time^

0.4939

0.5286

0.2729

0.7308

Time^

0.2699

0.1469

0.1606

0.1125

Data sets

0.0077

0.0109

0.5497

0.0001

Data set x time

0.3884

0.3551

0.2115

0.9510

Data set x time^

0.2157

0.1424

0.3797

0.3474

Primary rhizome

Time

0.0001

0.0001

0.0001

0.0001

Time^

0.0001

0.0001

0.0001

0.0-01

Data set

0.0001

0.0001

0.3062

0.0001

Data set x time

0.0001

0.0001

0.0113

0.0037

Data set x time2

0.2914

0.1847

0.4356

0.7919

Secondary rhizome

Time 2

0.0001

0.0001

0.0002

0.0001

Time

0.0001

0.0001

0.0001

0.0001

Data set

0.0001

0.0001

0.5021

0.0001

Data set x time

0.0650

0.0091

0.0466

0.1783

Data set x time^

0.0007

0.0833

0.8571

0.0734

tSignificance level

$Time in days (Juli

an) .

§Intercept.

64

and Hilst (1969) showed diurnal fluctuations in carbohydrate percentage,
but minimum levels were found early in the morning. In this study the
samples were taken at relatively wide intervals, and all samples were
taken on sunny days between 9 and 12 A.M. Therefore, those cyclic
changes in shoot TNC must be related to morpho-physiological changes
occurring in the plant.

In the trends corresponding to TNC in the two rhizome fractions
(Figs 7B and 7C), there was an increase in mid- June, while shoots showed
a decrease (Fig. 7A) ; once more in early July, TNC in rhizomes decreased
and shoot TNC increased. These cycles corresponded to cycles of shoot
emergence observed in the field, when rhizome development preceded shoot
formation and emergence. Wardlaw (1967) reported high levels of carbo-
hydrate assimilates in leaves of wheat (Triticum aestivum L.) with
drought stress. A similar trend was found with wheat and oats (Avena
sativa L.) by McCaig and Clarke (1982). These reports may partially
explain the high TNC percentage in the shoots of 1981 regrowth, since
1981 was a rather dry year and drought stress was observed at the onset
of the growing season and at the end. In the middle of July, when rapid
growth started, the trends of TNC in shoots and rhizomes seemed to follow
a similar pattern; however, the magnitude of the changes in rhizomes were
by far bigger than in shoots. This was expected since shoots are not
storage organs and even though some physiological changes may increase
the demand for carbohydrates, these results indicated that rhizoma pea-
nut shoots maintained a low level of TNC; this suggested that, if photo-
synthesis remained constant, the excess of assimilates was translocated
to the rhizomes.

65

Some reports in the literature state that carbohydrates accumu-
late during the cool temperatures of the fall, when growth rates have
decreased (Brown and Blaser, 1965; Wardlaw, 1969; Christiansen, 1982);
the TNC levels in rhizomes of rhizoma peanut also began to increase
at the end of September when rates of growth were decreasing. The TNC
accumulation patterns in both rhizome fractions were very similar; how-
ever, primary rhizomes had higher TNC than secondary rhizomes (P = 0.050).
For practical purposes, either rhizome can be sampled for TNC since the
overall mean differed by less than 2.0 percentage units. Furthermore,
considering the size of the fractions (see Chapter II), more TNC was
found in secondary rhizomes than in primary ones. The fractions seemed
to play different roles; secondary rhizomes spread the plant while
primary rhizomes were centers of growth and development of the root
system. Nevertheless, secondary rhizomes are capable of branching and
producing more shoots and rhizomes, and can eventually become primary
rhizomes. Therefore, either one of them would give an indication of
the status of the plant.

The regression analysis of TNC in primary rhizomes (Table 5)
showed significant Data set x time interaction in all three comparisons
between data sets. The fitted curves (Fig. 7E) show that less TNC
was mobilized in the 1981 regrowth (II) than either the 1980 growth (I)
or 1981 growth (III). Also, more TNC was accumulated in the 1981 re-
growth (II) than in the other two data sets, which were very similar.
Therefore, the rates of TNC accumulation were highest for the 1981
regrowth (II) followed by 1981 growth (III) and 1980 growth (I). The

66

same trends with the same ranking or data sets were observed for
TNC percentage in secondary rhizomes (Fig. 7F), although statistical
results of the comparisons between curves of data sets were not the
same (Table 5). In the 1980 growth (I) and 1981 growth (III) curves,
the linear coefficients were not different (P = 0.0466) indicating
higher TNC accumulation in 1981 growth (III) than in 1980 growth (I)
(Fig. 7F). Between 1981 regrowth (II) and 1981 growth (III) there was
no difference in the rates of accumulation (linear P = 0.1783 and
quadratic P = 0.0734). This result suggested that differences among
TNC patterns in the secondary rhizomes of those two data sets were due
to initial TNC percentage (intercepts); however, the observed curves
for primary and secondary rhizomes (Figs. 7B and 7C) show that initial
TNC percentages, in all three data sets, were very similar. Further-
more, the percent TNC curves of secondary rhizomes (Table 5) for the
1980 growth (I) had significantly different intercept and linear
coefficient from 1981 regrowth (II). Therefore, TNC accumulation rates
in secondary rhizomes also ranked highest for 1981 regrowth (II) fol-
lowed by 1981 growth (III) and 1980 growth (I). The nature of these
slight discrepancies between TNC trends in primary and secondary rhi-
zomes seemed to originate between July and September (Figs. 78 and 7C)
when cyclic changes in percent TNC were more drastic in secondary rhi-
zomes. During early and later parts of the year both fractions followed
similar patterns. These results suggested that an equilibrium existed
between the two rhizomes. Since most growth occurred in secondary rhi-
zomes, the percent TNC was higher in secondary than in primary rhizomes
in periods of rapid growth. However, an equilibrium was reached later

67

between the two fractions. To illustrate this point, the period between
July and September (Fig. 7C) shows a sharp TNC increase and decline in
the 1980 growth (I); in the same period, 1981 regrowth (II) and 1981
growth (III) show an increase followed by a constant period (Fig. 7C).
The same trend is observed, to a lesser degree, in primary rhizomes
(Fig. 7B). This period corresponds to the period of fast growth in
rhizomes reported in Chapter II, indicating that photosynthetic rates
were exceeded by rhizome growth rates in 1980; whereas in 1981 rhizome
growth rates may have been smaller than photosynthetic rates, because
TNC increased in quantity, although it is not reflected on a percent
basis. These observations clearly demonstrated that, at this point in
time, the rhizomes became a strong sink for assimilates in undisturbed
plants.

Nitrogen

The N percentage in shoots was plotted against time; regression
equations were also generated and plotted. The observed fitted curves
for N percentage (organic matter basis), for shoots, primary and secon-
dary rhizomes are shown in Fig. 8. The regression analyses for shoots
(Table 6) showed that the intercept, linear coefficient (time), and
quadratic coefficient (time ) of 1981 regrowth (II) were different
from those of 1980 growth (I) and 1981 growth (III). On the other hand,
between 1980 growth (I) and 1981 growth (III) there were only significant
differences between linear coefficients and intercepts (P = 0.0008 and
P = 0.0086, respectively). Therefore, 1981 growth (III) had the highest
N percent followed by 1980 growth (I) and 1981 regrowth (II). Nitrogen
concentration in shoots (Fig. 8A) declined until September and then

68

SHOOTS

5.5

" \

4.6 \

4.3 \

Z

4.0 '

Ui

o

3.7

cc

3.4

UJ

3.1

2B

z

2.5

Z2

1.9

1.6

M A

8A

•I9ei R<«re«Hi(M)
■ 1961 Grawtti (IH)

•49.45-.2J0 +.00090* r*'.e2
DQ • 14.92 -.OOTO-olzilCf^ f*-.9S
■ 54.7B-.^0 ')-.0(^40'r'*.9S

55i

5i2

4.9

4.6

A3

40

3.7

3.4

3.1

2M

25

22

1.9

1.6

TIME

TIME

PRIMARY RHIZOMES
8E

*3.6-X)IJ0 + (S.5jlO*)D'r''.90
• 2.9- 0120')' (2.^Kr*JD^ r*..60
■ 5.4 -.0250 * (5.1 iKTW t^.l9

3.4
3.1

2j8
25
2
.9

TIME

TIME

SECONDARY RHIZOMES

*35-jOt20 ♦(27«tO*)0' r». 4€
8 F • 34 - dj7D t (4.6*ilCf*)0* r«. .60
=^ ■5.l-.02IO + t3.6"lO-^0*r'«.ei

3.4

3.1
2B
25
22
1.9
1.6

TIME

TIME

Fig. 8, Percent Nitrogen (organic matter basis) in plant parts of
Florigraze rhizoma peanut for three data sets.
8A, SB, SC = observed data.
8D, 8E, 8F = estimated data.

]

69

Table 6. Regression analyses of complete data and individual compari-
sons between data sets for N percent in plant parts of
Flori graze rhizoma peanut.

Data set comparisons

Plant part Data sets Data set I Data set I Data set II

I II III vs vs vs

Data set II Data set III Data set III

PR = Ft

Shoots

Time*

0.0001

0.0001

0.0001

0.0001

Time^

0.0001

0.0001

0.0001

0.0345

Data set§

0.0001

0.0097

0.0008

0.0001

Data set x time

0.0002

0.0315

0.0086

0.0001

Data set x time^

0.0081

0.0048

0.6889

0.0001

Primary rhizome

Time

0.1769

0.0001

0.0130

0.6288

Time 2

0.0001

0.0001

0.0001

0.4192

Data set

0.0001

0.1257

0.0001

0.0003

Data set x time

0.0016

0.0005

0.3212

0.0598

Data set x time^

0.3769

0.6676

0.4824

0.5368

Secondary rhizome

Time

0.0519

0.0748

0.0001

0.5814

Time2

0.0001

0.0001

0.0001

0.0266

Data set

0.0001

0.0019

0.0134

0.0001

Data set x time

0.0001

0.0001

0.0019

0.0001

Data set x time^

0.8502

0.0908

0.5023

0.6179

fSignificance level.
JTime in days (Julian).
§Intercept.

70

remained constant until the end of the experimental periods. The 1981
regrowth (II) showed a rather constant pattern from start to end, prob-
ably because the first sampling was taken when shoots were over a month
old, although the subsequent samples were also lower than those of the
1981 growth (III). The N percent in plants declines with age, as cell
walls thicken and lignin deposits. The N levels found for shoots are
within the range reported by Prine et al. (1981) for Flori graze rhizoma
peanut during several growing seasons and several locations in Florida.
But the 1981 regrowth (II) N percent was maintained at the lowest extreme
of the reported ranges. Only the 1980 growth (I) and 1981 growth (III)
followed the decreasing pattern reported by Breman (1980) and Beltranena
(1981).

In both rhizome fractions, N percentage (Figs. 8B and 8C) also
declined until about August and then it slowly increased until the end
of the experiments. The fluctuations that occurred throughout the growing
season did not correlate with any of the morphological changes observed
in the plant. The fitted regressions (Figs. 8E and 8F) showed that N
percentage followed a declining trend at the beginning of the growing
seasons, followed by a relatively constant stage and ended with a slight
increase. The regression analyses for primary rhizomes (Table 6) showed
that 1981 growth (III) had similar linear and quadratic coefficients to
1980 growth (I) and 1981 regrowth (II), although the intercepts were
different. Between 1980 growth (I) and 1981 regrowth (II) only the linear
coefficients were different. In secondary rhizomes the intercepts and
linear coefficients were different among all three curves (Table 6).
These results indicated, on one side, that more N was found in secondary
rhizomes than in primary rhizomes. On the other side, the results also

71

indicated that the concentration of N in rhizomes decreased during periods
of rapid growth and increased at the end of the year. The decline of N
at the beginning of the growing season, without concurrent increase in
rhizome mass, indicated that N was mobilized to shoots. Davidson and
Milthorpe (1966) reported mobilization of nitrogenous compounds from
the roots of Dactyl is glomerata L., whereas Smith and Silva (1969) found
the same in alfalfa roots.

The plants in this study were grown on Arredondo loamy sand soil
which is characterized by low organic matter and poor N retention (USDA,
1976). Soil pH was 5.9 and 6.0 in field A and field B, respectively
(Table 7). No N fertilizer was applied during the experimental periods;
therefore, the N sources for plant growth were residual N and N fixation
from the symbiosis of rhizoma peanut with Rhizobium japonicum (Adjei and
Prine, 1976). Maximum N fixation occurs when nodules have enough avail-
able carbohydrates (Bergersen, 1975). Data in Figs. 7D, 7E, and 7F show
that consistently more TNC was accumulated in the plant parts of the 1981
regrowth (II) than either the 1980 growth (I) or 1981 growth (III). On
the other hand, the N results (Figs. 8D, 8E, and 8F) show that N percen-
tage was also consistently lowest in the 1981 regrowth (II). Mclntyre
(1967) found that low N levels favored rhizome formation in couchgrass
(Agropyron repens L. Beauv.). Also, low temperatures had a similar effect.
In potato (Solanum tuberosum L.), increasing N supply increased leaf area,
but there was a strong interaction with P (Watson, 1963). In these ex-
periments the low levels of N observed in the 1981 regrowth (II) suggest
that N deficiency, particularly at the onset of the growing season, did
not allow maximum growth. Soil analysis (Table 7) indicated that P was

72

Table 7. Soil mineral analysis in samples from Field A and Field B, in
February 1981 .

4"

Field A^

4,

Field B+

Analysis

ppm

s.d.

c.v.

ppm

s.d.

c.v.

P

48.8

10.2

20.9

50.7

11.8

23.2

K

40.3

10.5

26.2

36.4

11.2

30.6

Ca

138.0

42.0

30.4

140.6

58.8

41.8

Mg

20.3

9.9

48.5

10.7

6.1

57.7

Fe

14.9

4.8

32.2

15.0

-3.5

23.1

Mn

3.9

0.6

16.1

3.1

0.6

20.7

Cu

0.5

0.1

31.3

0.2

0.03

11.5

Zn

0.6

0.2

29.6

0.8

0.3

33.0

pH (H2O)

5.87

0.142

2.42

5.98

0.176

2.95

pH (KCl)

4.53

0.137

3.02

4.61

0.197

4.27

Means of 12 composite samples.
"^Vleans of 20 composite samples.

73

within acceptable range for this type of soil, although K, Ca, and Mg
were low with respect to recommended standards to grow peanuts (Arachis
hypoqaea L.) (Rhue and Sartain, 1979). The same levels were present in
both fields. These observations suggested that N limited growth. It has
been reported that N fixation does not fulfill the demand for growth in
some legumes (Alios and Bartholemew, 1959). This might be the explana-
tion for low relative growth rates found in 1981 regrowth (Chapter II).
This was probably not the case in the 1980 growth (I) and 1981 growth
(III) because some residual N from previous crops was available to the
plants. The N deficiency seemed to have occurred early in the growing
season of 1981. Immediately after emergence of the 1981 regrowth plants,
a drought was experienced and very few nodules were found. In the 1981
growth (III), no nodules were found either, but the plants developed a
very branched root system and very few shoots (Chapter II). Field ob-
servations suggest that rhizome growth of rhizoma peanut is reduced in
heavy soils. Some coiling of rhizomes has occurred in this type of soils
(G. M. Prine, personal communication). In these experiments, at the
end of 2 years, plants remained together in spite of multiple ramifica-
tions and rooting; therefore, it might be possible that rhizome growth
was reaching a plateau. In fact, the primary rhizomes developed more
than secondary rhizomes in the 1981 regrowth as compared to 1980 growth
or 1981 growth (Chapter II). Also, observations during 1982 seemed to
indicate that the rate of spreading was reduced even more than in 1981.

Summary

All plants show typical patterns of TNC accumulation. Usually
there is a decrease of TNC in periods of slow growth. The objectives

74

of the present study were to determine TNC and N accumulation patterns
of undisturbed Florigraze rhizoma peanut plants; to determine where
TNC accumulates; and to determine the effect of cooling and freezing
forage samples in the field on TNC recovery. Two field experiments
(A and B) were carried out in 1980 and 1981, 18 km west of Gainesville,
FL. Experiment A was conducted during 2 years (1980 growth and 1981
regrowth), experiment B only 1 year (1981 growth). Samples were taken
for TNC and N analyses according to a sampling schedule that covered
the entire growing season. Ten and seven samples (1980 and 1981, res-
pectively) were taken at all samplings. The sample was split in two,
one sample was cooled (ice) and the other frozen (dry ice). Samples
were separated into shoots, primary rhizomes and secondary rhizomes.
Samples were composited later into three samples using plant size as
compositing criterion. Samples were analyzed for TNC and N. The
results were fitted into a quadratic model and statistically analyzed
with regression analyses.

There was no statistical difference between cooling and freezing
samples, although there was a tendency of frozen samples to show higher
TNC recoveries. The TNC percentage was consistently higher in 1981
regrowth in all plant fractions than 1980 and 1981 growth, whereas the
opposite occurred with N, suggesting that low relative growth rates of
1981 regrowth were due to N deficiency. More TNC was found in second-
ary rhizomes in periods of rapid growth than in primary rhizomes. With
the same trend for N, these results suggested that secondary rhizomes
are a more active metabolic center; therefore, in periods of rapid
growth they will not give an indication of the status of the plant.

75

The TNC accumulation increased at the end of the growing season indi-
cating that photosynthesis exceeded growth, and it continued to
increase at least until the last samples were taken. Initial decline
of TNC was not related to increase in rhizome weight, indicating that
TNC was respired and/or used in shoot growth. This kind of response
was observed in cycles but the extent of reduction was very small
indicating that enough photosynthates were produced at all times.
This maintained increasing levels of TNC at the end of the growing
season, in spite of the growth of rhizomes.

CHAPTER IV

EFFECT OF DEFOLIATION ON GROWTH, TOTAL NONSTRUCTURAL
CARBOHYDRATES, AND NITROGEN IN FLORI GRAZE
RHIZOMA PEANUT

Introduction

Regrowth in perennial plants occurs after defoliation or after
the top growth is killed by low temperatures in the winter. Plants
need active meristematic tissue and nutrient reserves to initiate
regrowth. Some researchers have stressed the importance of total non-
structural carbohydrates (TNC) to initiate regrowth (Weinmann, 1961;
Smith, 1981) while others have questioned it (May and Davidson, 1958;
Blaser et al., 1966). However, this controversy only arises with
respect to the initiation of regrowth after defoliation because it is
generally accepted that TNC is needed in regrowth after winter
(Humphreys, 1981). It has been demonstrated that stored N also plays
an important role in regrowth (Davidson and Milthorpe, 1966) although
its relative contribution in relation to TNC may be low (Smith and
Silva, 1969). The number of shoots is species dependent; however,
the removal of apical meristems triggers shoot formation from lateral
meristems (Thimann, 1937), increasing the number of shoots. In legumes
during the reproductive stage, lateral meristems develop inflorescences
(Phillips, 1969); therefore, if apical meristems are removed, new
shoots must develop from basal meristems which is a long and ineffi-
cient process.

76

77

Florigraze rhizoma peanut, a legume native to South America, is
an important forage crop in Florida. However, a serious disadvantage
is the lack of seed production; therefore, it has to be vegetatively
propagated (Prine et al., 1981), This process is costly and slow,
because rhizomes must be dug for its propagation. The establishment
usually takes from 2 to 3 years (Adjei and Prine, 1976). Due to its
rhizomatous habit, numerous lateral buds remain underground; therefore,
shoot formation after defoliation should only be limited by nutrient
reserves. Reports by Beltranena et al. (1981) and Breman (1980) indi-
cate that maximum yields of Florigraze rhizoma peanut are obtained
when it is defoliated every 8 weeks. In Chapters II and III, results
were reported on the total growth, morphological changes throughout
the year, and storage and translocation of TNC and N in undisturbed
plants. With those patterns as reference, an objective of this study
was to quantify the effects of frequency of defoliation upon the growth
and development of rhizoma peanut rhizomes and the translocation and
accumulation of TNC and N. Another objective was to determine the
effect of defoliation upon shoot formation.

Materials and Methods

A field experiment was conducted in 1980 and 1981 at the University
of Florida Agronomy Department's "Green Acres" farm, 18 km west of
Gainesville, FL. The details on field design, soil type and general
practices were described in Chapter II. The plants used in this study
were planted in February 1980 (Field A, see Chapter II) and they were
allowed to grow throughout 1980 without disturbance or plant competition.

78

The foliage was frost-killed in December 1980, and all dead material
was removed by mowing in January 1981. Growth data for these plants
during 1980 and 1981 were reported in Chapter II. The 1981 regrowth
data, which comprised the growth of undefoliated hills during the
second season of establishment were used in this study as an undefoli-
ated control treatment. In addition to this, in the same field, three
defoliation frequencies were randomly assigned to 91 hills in 1981.
The frequencies were every 2, 6, and 8 weeks. The 2-week frequency
was assigned to 42 hills, the 6-week to 28 hills, and the 8-week to
21 hills. Defoliation was done by mowing at 3.5 cm height with a
rotary lawn mower. All 91 hills were defoliated on 10 June 1981 to
initiate the experiment. Within every treatment a sampling schedule
was randomly assigned. Seven hills per treatment were dug at all
sampling periods. The sampling and defoliation schedules are presented
in Table 8. The sampling schedule was designed to match, as much as
possible, that of the undefoliated control. At all sampling periods
all hills were dug; a sample was taken from every hill, as described
in Chapter II, to determine the relative weight of shoots, primary
rhizomes, and secondary rhizomes. The sample and the remainder of
the hill were freed of extraneous plant materials, washed to eliminate
sand, and dried in cloth bags at 60 C for 48 hrs. Immediately after
digging, samples were also taken of all three plant fractions and
placed on ice. These samples were taken to a laboratory, dried at
110 C for 30 minutes and transferred to a forced-air oven set at 60 C,
where they remained for 36 hours. All samples were ground in a Udy
cyclone mill with a 0.5 mm screen, and stored in plastic containers.

79

Table 8. Sampling and defoliation schedule for 1981 treatments imposed
on Flori graze rhizoma peanut.

Defoliation frequency

Undefoliated 8 weeks 6 weeks 2 weeks

St Dt S D S D

Date

9 June 10 June 10 June

9 July 12 July

23 July 23 July

7 Aug. 4 Aug. 4 Aug. 5 Aug.

5 Sept. 5 Sept. 6 Sept. 6 Sept

2 Oct. 2 Oct. 3 Oct. 3 Oct.

13 Oct. 14 Oct.

31 Oct. 31 Oct.

26 Nov. 26 Nov. 25 Nov. 25 Nov.

10

June

26

June

11

July

23

July

4

Aug.

20

Aug.

6

Sept.

18

Sept.

3

Oct.

14

Oct.

1

Nov.

12

Nov.

tSampling date (S).
jOefoliation date (D).

80

Analyses for TNC were performed on shoot, primary and secondary rhizomes
with the procedure outlined by Smith (1981) with the modifications
reported by Valle-Melendez (1980) and Christiansen (1982). Nitrogen
analyses were performed on the same plant fractions with a micro-Kjeldahl
aluminum block digestor and a Technicon Autoanalyzer" (Gallaher, 1975).

Primary and secondary rhizome weight data as well as percent TNC
and N (organic matter basis) of all three plant fractions were fitted
into regression equations. All defoliation frequencies were compared
to the undefoliated control and among the defoliation frequencies. All
statistical comparisons were carried out with General Linear Model
procedure with time as continuous variable and defoliation frequencies
as class variables (SAS Institute Inc., 1982).

Results and Discussion

Rhizome Growth

The growth of rhizomes for all defoliation treatments was plotted
against time and is shown in Fig. 9. A regression analysis (Table 9)
which included all defoliation treatments showed treatment differences
and treatment x time effect, indicating differences among intercepts
of the curves and differences between linear coefficients. Individual
comparisons bewteen each defoliation treatment and the undefoliated
control showed that only the linear coefficients of the 2-week and
6-week defoliation treatments were different from the linear coeffi-
cient of the control (P = 0.0100 and P = 0.0482, respectively). This
indicated that the rhizome growth rates of the 2- and 6-week defolia-
tion frequencies were lower than in the undefoliated plants. The

81

82

Table 9. Regression analyses of complete data and individual comparisons
between defoliation treatments for growth components of Flori-
graze rhizoma peanut.

Comparisons of defoliation treatmentst

0-2-6-8 0-2

0-6

0-8

2-6

2-8

6-8

Rhizomes

Time§
TRU

TR X time

PR = Fi-

0.0001 0.0002 0.0002 0.0042
0.0001 0.0001 0.0001 0.0095
0.0208 0.0100 0.0482 0.1724

0.0103 0.0185 0.0571
0.6427 0.0363 0.1977
0.7233 0.4911 0.8071

Primary rhizomes

Time
TR

TR X time

0.0001 0.0001 0.0001 0.0001 0.0001
0.0001 0.0001 0.0001 0.0013 0.2695
0.0007 0.0015 0.0260 0.0887 0.2523

0.0001 0.0001
0.0010 0.1256
0.0413 0.5239

Secondary rhizomes

Time
TR

TR X time

0.0012 0.0024 0.0056 0.0279 0.1175
0.0001 0.0001 0.0015 0.0288 0.8091
0.0983 0.0511 0.1656 0.2507 0.9252

0.1466 0.2643
0.0894 0.2471
0.7483 0.9032

Number of shoots

Time, 0.5848 0.7902 0.3893 0.6330

Time'^ 0.2333 0.2287 0.2830 0.3630

TR 0.8949 0.8959 0.5295 0.5948

TR X time 0.5288 0.4441 0.4299 0.6277

TR X time2 0.9722 0.7033 0.9785 0.8666

0.7889
0.4506
0.4514
0.1111
0.7811

0.8848
0.6322
0.4481
0.3356
0.6775

0.3434
0.4329
0.9411
0.9172
0.8757

fPrequency of defoliation in weeks,
jsignificance level.
§Time in days (Julian).
^Intercept.

83

regression analyses also showed that the intercepts of the curves of
all three defoliation frequencies were different from the undefoliated.
Even though the 8-week frequency seemed to have similar rhizome growth
rate to the undefoliated, the total growth was higher in the undefoliated
(P = 0,0095). When comparisons were made among defoliated treatments,
there was only significant difference between the intercepts of the
2-week and 8-week frequencies; indicating that all defoliation frequen-
cies gave similar rhizome growth rates, although relative growth rates
were higher in the 8-week frequency than in the 2-week frequency. The
6-week frequency was not different to either 8- or 2-week frequencies.
The fitted curves (Table 10) give a better understanding of these results
and clearly show the ranking of treatments. The ranking indicates that
rhizome growth decreases as defoliation frequency increases. These
results were expected since it had been reported that other tropical
legumes reduced yields with increasing cutting frequencies (Jones, 1967;
Bryan et al., 1971). Smith and Graber (1948) reported reduction on the
size of coronal rhizomes in biennial sweet clovers (Melilotus officina-
lis Lam. and Melilotus alba Desr.) when defoliated early in the fall.

The separation of rhizomes into primary and secondary rhizomes
did not show any particular change (Fig. 9) in the relationship re-
ported in Chapter II for the undefoliated plants in the 1981 regrowth.
This means that the secondary rhizomes continued to be the bulk of the
rhizome mat, but the ratio of secondary to primary rhizomes decreased
with time. The regression analyses to compare production curves of
primary or secondary rhizomes, among defoliation treatments, gave
essentially the same results and ranking that was reported for rhizome

84

Table 10. Mathematical equations to describe the effect or frequency
of defoliation on the growth of rhizomes in Flori graze
rhizoma peanut.

Component Intercept Coefficient

timet

g hiir"" day-l

Rhizome

Undefoliated

-1034.59

7.51

8 weekt

- 48.16

2.03

6 week

- 40.40

1.53

2 week

- 30.19

1.17

Primary rhizomes

Undefoliated

- 564.65

3.33

8 week

- 175.17

1.15

6 week

- 129.57

0.85

2 week

69.40

0.57

Secondary rhizomes

Undefoliated

- 469.93

4.17

8 week

127.00

0.87

6 week

89.17

0.67

2 week

99.60

0.59

tTime in days (Julian).
tFrequency of defoliation.

85

growth (Tables 9 and 10). Undefol iated plants produced more primary
and secondary rhizomes than any defoliation treatment; and the 8-week
frequency produced more of both rhizomes than the 2-week frequency.
The dry-matter accumulation of the primary or secondary rhizomes of
the 2-week and 6-week defoliation treatments were small and not dif-
ferent (P = 0.2659 and P = 0.8091 for primary and secondary rhizomes,
respectively). However, more development was observed in the 6-week
plants than in the 2-week plants. The regression coefficients (Table
10) show that defoliations every 2 or 6 weeks had a tendency to slightly
increase rhizome grwoth (6 = 1.17 g day'^ and 6 = 1.53 g day ^ for 2-
and 6-week, respectively). The difference between them was a higher
number of rhizomes in the 6-week frequency than in the 2-week. In
the latter, rhizomes decreased in number but the remaining became thicker
and rougher in texture; even secondary rhizomes increased in diameter.
Under 6-week frequency, the plants produced secondary rhizomes and
maintained the same morphology as the 8-week or undefol iated plants.
It had been reported that rhizoma peanuts frequently defoliated developed
rosette-like structures (Prine et al., 1981). The same thing was ob-
served in this experiment in the 2-week defoliation treatment. Primary
rhizomes developed into what seemed to be a taproot, but it only happened
in deep-rooted primary rhizomes. Other primary rhizomes, without roots,
remained connected to the "taproot- like" primary rhizomes but rarely
branched. Most new shoots arose from the taproot, mostly vertically,
although a few secondary rhizomes produced shoots. When plants from all
defoliation treatments were visually compared to undefoliated plants,
there was a clear reduction in area covered, except for the 8-week

86

defoliation frequency where plants kept spreading, although the amount
of rhizomes per area seemed less than in the undefoliated plants.

In this experiment, three defoliations (8-week frequency) de-
creased rhizome production to about half of the undefoliated control,
whereas 4 or 12 defoliations reduced it to one third of the control.
These results suggest that if rhizome growth is desired, a more lenient
(one defoliation) or no defoliation should be practiced. It should
be pointed out, however, that visual observations of the plants early
in 1982 did not indicate failures in regrowth in plants under the
defoliation treatments of 1981. The canopies in all plants were closed
very quickly, but the area covered reflected the defoliation treatments
of the year before. These observations further indicated that defolia-
tion will reduce rhizome production and spreading.

Total Nonstructural Carbohydrates (TNC)

Total nonstructural carbohydrates percentage of the organic matter
was plotted against time for shoots, primary and secondary rhizomes
(Figs. lOA, lOB, and IOC). In all three plant fractions, TNC was
ranked in the same order: undefoliated > 8-week > 2-week > 6-week. A
regression analysis (Table 11) showed that undefoliated plants had higher
TNC in all three plant fractions than in any of the defoliation treat-
ments. The comparisons between defoliation treatments showed that the
8-week frequency had higher TNC in shoots than the 6-week frequency.
In primary and seocndary rhizomes, the 8-week frequency had higher TNC
than both the 2- and 6-week frequencies. The 2- and 6-week frequencies
had similar TNC in rhizomes but differed in shoots, although the 2-week
frequency on the average had higher TNC than the 6-week frequency. The

87

■ ' ■ ■ ■ ■ ■ — I I I I I

JASON JASON

TIME TIME

g, 10. Effect of defoliation treatments on TNC percent
in plant parts of Florigraze rhizoma peanut.
lOA, lOB, IOC = observed data.
lOD, lOE, lOF = estimated data.

88

Table 11. Regression analyses of complete data and individual comparisons
between defoliation treatments for TNC percent in plant parts
of Flori graze rhizoma peanut.

Comparisons of defoliation treatments t

Plant parts 0-2-6-8 0-2 0-6 0-8 2-6 2-8 6-8

PR = Ft

Shoots

Times

U.Uoob

U.

n n^Ri

n 0047

Time'-

n nnm

u . UUU 1

n nnm

n nnm

n mQ?

n 0001

0 0078

0 5836

TRf

0.0001

0.0001

0.0001

0.0001

0.0084

0.6796

0.0065

TR X time

0.2240

0.5805

0.1020

0.9688

0.2212

0.4096

0.0940

TR X time2

0.0016

0.3638

0.9847

0.0004

0.3468

0.0002

0.0002

Primary rhizomes

Time

0.0001

0.0001

0.0001

0.0001

0.0001

0.0001

0.0001

Time 2

0.0002

0.1917

0.0074

0.9958

0.0001

0.0031

0.0001

TR

0.0001

0.0001

0.0001

0.0001

0.1304

0.0141

0.0001

TR X time

0.0143

0.0037

0.1301

0.2480

0.1888

0.8728

0.1170

TR X time^

0.0006

0.0160

0.0001

0.0143

0.0944

0.5026

0.5441

Secondary rhizomes

Time

0.0001

0.0001

0.0001

0.0001

0.0001

0.0001

0.0001

Time^

0.0001

0.0004

0.0001

0.2811

0.0001

0.0005

0.0001

TR

0.0001

0.0001

0.0001

0.0001

0.1762

0.0759

0.0070

TR X time

0.0007

0.0136

0.3174

0.0064

0.1476

0.0319

0.0005

TR X time2

0.0556

0.1358

0.0183

0.7732

0.1416

0.4001

0.0549

tFrequency of defoliation in weeks.
^Significance level.
§Time in days.
^Intercept.

89

fitted curves for all treatments and the three plant fractions are
shown in Figs. lOD, TOE, and 10F. All four treatments showed an in-
creasing TNC accumulation trend, although the defoliated treatments
presented a reduction of TNC at the beginning and an increasing rate
afterwards. The highest reduction in rhizome TNC occurred at 2- and
6-week frequencies (Figs. lOB and IOC). In the 2-week frequency this
reduction in TNC occurred after three defoliations whereas in the 6-
week frequency it happened after one defoliation. It should be pointed
out that, as compared to the undefoliated control, the 8-week frequency
also reduced the rhizome TNC percentage after one defoliation. Since
the 8-week had a higher TNC percentage than the 6-week frequency after
one defoliation, the results indicated that plants needed about 8 weeks
to recover their TNC level after one defoliation. Some reports in the
literature indicate that TNC levels in grasses (Weinmann, 1947) and
legumes (Smith, 1981; Barta, 1979) are reduced after defoliation. The
extent of this TNC reduction depends on the degree of defoliation and
the photosynthetic activity of the remaining tissue. It also depends
on the time needed for the regrowth to photosynthestze enough assimi-
lates to satisfy the demand for growth. The TNC contribution to re-
growth in grasses occurs only between 2 to 18 days after defoliation
(Davidson and Milthorpe, 1966; Sullivan and Sprague, 1943; Blaser et al.,
1966).

In this experiment the short-term effects on rhizome TNC content
after defoliation were not measured; however, the TNC percentage in
initial rhizome samples and the weight of the plants (Fig. 9) indicated
that TNC was used in regrowth. Furthermore, in the 6- and 8-week fre-
quencies almost no leaf area was left after defoliation, because the

90

shoots were elongated enough that the first leaves were above cutting
height (3,5 cm). The difference in TNC accumulation between 2- and
6-week frequencies may be due to the leaf area left after defoliation.
After five defoliations, the shoots in the 2-week frequency grew almost
parallel to the ground and leaves remained very small, so considerable
leaf area remained close to the ground. Nevertheless, some material
was removed at every mowing, thus, less TNC was mobilized for shoot
growth. Conversely, in the 6- and 8-week frequencies, almost all leaf
area was removed; therefore, some TNC must have been mobilized for
initial shoot regrowth. Beltranena (1980) reported LAI of about 2.0,
at the peak of the growing season, when Florigraze rhizoma peanut was
defoliated every 2 weeks; LAI decreased from the end of August until
October. Those leaf area measurements only included leaves above the
3.8 cm cutting height. Therefore, it could be speculated that a
higher leaf area is attained at the end of 2 weeks. For the 6-week
frequency, Beltranena's results show an LAI of about 3.5, indicating
a slower recovery of LAI in the 6-week frequency; possibly because the
leaf area under 3.8 cm cutting height was almost zero.

In rhizomes, all three defoliation frequencies had approximately
the same TNC levels at the end of the year (50% in 8- and 6-week fre-
quencies and 45% in 2-week frequency). The TNC percentage was higher
in undefoliated plants (70%) than in defoliated. This difference in
TNC accumulation between defoliated and undefoliated plants does not
seem to explain the reduction in rhizome mass in defoliated plants,
but it explains the excellent regrowth observed in 1982 on the plants
defoliated during 1981.

91

Percent Nitrogen

Nitrogen percentage was also plotted against time and is presented
in Figs. 11 A, IIB, and IIC for shoots, primary rhizomes, and secondary
rhizomes, respectively. In shoots, N percentage had a declining trend
in all treatments. The ranking of N percentage was 2-week = 6-week >
undefoliated > 8-week and it can be seen in Fig. lOD. The regression
analysis (Table 12) showed that the intercepts of the lines were differ-
ent, except between 2- and 6-week frequencies, indicating that more fre-
quently defoliated plants maintained a higher N percentage in the
shoots than the less frequently defoliated ones. This is in agreement
with results comnonly found in the literature for defoliated plants.
In general, the N levels observed in shoots are lower than those re-
ported by Breman (1980) and Beltranena (1980); however, they only in-
cluded the plant material above cutting height. In this experiment,
the samples included shoots cut at ground level; therefore, the stubble
remaining below cutting height had a diluting effect. This explains
the declining shoot N percent observed in all defoliation treatments,
even though no dead stubble was included in any sample.

In primary rhizomes there were also significant differences be-
tween treatments (Table 12) but the ranking was undefoliated > 2-week >
6-week = 8-week (Fig. HE). In secondary rhizomes the differences be-
tween undefoliated plants and all defoliation treatments were more
marked, and there was no difference among N percent in all three
defoliation frequencies (Table 12). These results, on one side,
point out that primary rhizomes are a less active metabolic center
since levels of N are in general lower than in secondary rhizomes.

92

SHOOTS

2.9

UJ

O
K

=2.1

ZI.7
UJ
O
(C
Ui 1.3

a.

II A

■Ct H UndefoMalad
• • 8-We«k
O O 6-W««k
■ ■ 2-We«k

ON J A

PRIMARY RHIZOMES
II E

z

o
cr

z'-7
UJ

o
a.

2.9

2.5

1.7

1.3

2.1

1.7

1.3

-I I I

2.5 ■

Ui
(9
O

—2.1

1.7-

UJ
O
(C
Ui

a.

1.3

SECONDARY RHIZOMES
II F

A S 0

TIME

_i u

-I i_

2.1

1.7

1.3

S 0

TIME

Fig. n

Effect of defoliation treatments on N percent in plant
parts of Flori graze rhizoma peanut
IIA, TIB, nc = observed data.
IID, TIE, IIP = estimated data.

93

Table 12. Regression analyses of complete data and individual comparisons
between defoliation treatments for N percent in plant parts of
Flori graze rhizoma peanut.

Comparisons of defoliation treatments!

Plant parts 0-2-6-8 0-2 0-6 0-8 2-6 2-8 6-8

PR = Ft

Shoots

Times

U.UUU 1

U . UUU 1

n

nnm

, UUU 1

n nnm

U . UUU 1

n nnm

n 0001

0 0001

Ti me^

U . 13 13

n

u <

n 0665

0.1312

0.0141

0.3028

TRU

0.0001

0.0004

0,

.0012

0.0224

o!4054

0.0045

0.0012

TR X time

0.0001

0.0001

0.

.1332

0.0742

0.0153

0.0526

0.6521

TR X time2

0.0001

0.7806

0,

.0896

0.0001

0.4455

0.0005

0.0001

Primary rhizomes

Time

0.0001

0.0129

0,

.0001

0.0006

0.2742

0.4475

0.0001

Time^

0.0608

0.3594

0,

.1966

0.1497

0.4004

0.0916

0.0089

TR

0.0001

0.0001

0,

.0001

0.0001

0.0014

0.0061

0.5649

TR X time

0.0859

0.0384

0

.3613

0.5617

0.1815

0.1834

0.6375

TR X time^

0.7279

0.8609

0

.8173

0.3542

0.9602

0.2913

0.0297

Secondary rhizomes

Time

0.0001

0.0001

0

.0001

0.0001

0.0001

0.0001

0.0001

Time'^

0.2819

0.0139

0

.0307

0.0005

0.4754

0.3103

0.0137

TR

0.0001

0.0001

0

.0001

0.0001

0.1743

0.2572

0.7997

TR X time

0.0211

0.0121

0

.0198

0.0240

0.9162

0.5046

0.3187

TR X time^

0.0002

0.0162

0

.0009

0.0001

0.3870

0.0477

0.1785

fFrequency of defoliation in weeks.
^Significance level.
§Time in days (Julian),
illntercept.

94

On the other hand, differences in N percentage due to different treat-
ments can more easily be detected in primary rhizomes than in second-
ary ones. Most nodules were found in secondary rhizomes or in short
primary roots arising from secondary rhizomes; consequently, more N
in secondary rhizomes would be expected. Data in Figs. HE and 11 F
clearly show that more N was mobilized from primary rhizomes than from
secondary ones. It can also be seen that the 2-week frequency maintained
a rather constant N percentage in primary rhizomes, indicating that N
demand in other tissues was being supplied by other sources. The N
percent in rhizomes reported by Adjei and Prine (1976) was 1.91%.
It was very comparable to the values ranging from 1.87 to 1.96% re-
ported by Breman (1980) for several defoliation treatments, fireman's
results came from rhizome samples taken at the end of the year, whereas
the results of Adjei and Prine were obtained with 3-month old rhizomes.
In the experiment reported herein, very similar values (1.84-2.0%)
were found in secondary rhizomes of defoliated plants at the end of
the year; whereas in primary rhizomes they were lower (1.64%). In
undefoliated plants N varied from 1.60 to 2.36% for July and November,
respectively. In the experiments reported in Chapter III it was found
that N in rhizomes was 2.24 and 2.36% in 1980 growth and 2.56 and 2.33%
in 1981 growth for July and November, respectively.

In 1981, rhizome peanut depended almost entirely on N fixation
since no N fertilizer had been added for at least two growing seasons.
In Chapter III it was postulated that the growth of undefoliated plants
during the second growing season had been hampered by either a lack of

95

N or by reduction in growth potential in the plants. The comparison
of undefoliated and defoliated plants in regard to TNC and N percentage
does not clarify the problem because it gives evidence in both direc-
tions. However, it gives a perspective of the relationships between
TNC accumulation, N availability and plant growth. In rhizomes of
undefoliated plants, TNC and N percent (Figs. lOE, lOF, and IIF) in-
creased throughout the experimental period, indicating that enough
carbohydrates were available for N fixation. It also indicated that
photosynthesis and N fixation were occurring at a faster rate than
growth; therefore, TNC and N accumulated. In defoliated plants two
situations seem to prevail. First, plants defoliated every 2 weeks
increased percent TNC in rhizomes with a steady but slow increase in
N, although the N level remained rather low throughout most of the
growing season. The stress on these plants stopped growth and develop-
ment of rhizomes. Hence, the small increment of weight in both primary
or secondary rhizomes was probably due to TNC and somewhat to N accu-
mulation. Likewise, the initial reduction in weight might have been
due to TNC and N removal, along with death of some secondary rhizomes.
The second situation is given by 6- and 8-week frequencies. In both
frequencies, N remained at a lower percentage than both 2-week and
undefoliated plants, while percent TNC increased like in 2-week
frequency plants. This suggested that more growth was occurring in
these two frequencies; in fact, the rhizomes looked more developed
than those of the 2-week frequency. These observations suggested
that TNC and N accumulation could not account for the slight weight
increment in the 6-week frequency and the weight increment in the

96

8-week frequency. The explanation of these results may have been
found in the dry matter percent of the rhizomes but unfortunately it
was not measured.

Defoliation of legumes often results in reduction of N fixation,
due to reduction in the supply of sugars to the nodules. In this
respect, Whiteman (1970) reported reduction in size and number of
nodules, with a subsequent reduction in yield of 'Siratro' [Macrop-
tilium atropurpureum (D.C.) Urb.l with increasing cutting frequencies.
Jones (1974) with the same species found that plant density and yield
increased linearly with cutting interval. In this experiment, all
defoliation treatments reduced biomass production vnth increased fre-
quency of defoliation. Table 13 shows the weight of shoots per hill
for every sampling date and defoliation frequency. These results in-
dicated taht the 6- and 8-week frequencies produced almost as much
as the undefoliated plants. Although no real measurement of yield
was taken, an estimation of yield in the 2-week frequency suggested
that it was not appreciably lower than with longer frequencies and in
undefoliated plants. The corollary of these results is that plants
modified their growth pattern when put under stress.

The number of shoots (Fig. 12) did not show a significant dif-
ference (Table 9), although the general trend indicated that all de-
foliation frequencies promoted early increase in shoot numbers which
may have been responsible for the drop in N percent in primary rhizomes.
It also indicated that a 2-week frequency caused an early decline in
shoot numbers. Since the importance of N in bud development and shoot
formation in grasses is recognized (Jewiss, 1972b), it seems that the

97

Table 13. Weight of shoots per plant for defoliation treatments at
sampling dates in Flori graze rhizoma peanut.

Undef ol i ated Defoliation frequency

8 week 6 week 2 week

Date Weight Date Weight Date Weight

9 Juy

304

23 July

12 July

180

7 Aug.

—t

4 Aug.

519

5 Aug.

161

5 Sept.

908

5 Sept.

301

6 Sept.

187

2 Oct.

1177

2 Oct.

341

13 Oct.

141

3 Oct.

149

31 Oct.

940

31 Oct.

115

26 Nov.

826

26 Nov.

103

25 Nov.

118

25 Nov.

53

tg hill-',
ilost sample.

98

99

drought experienced early in 1981 lowered the levels of N in the
plants, and impaired later shoot formation. This in turn reduced
potential growth but not photosynthesis; therefore, TNC accumulated.
Defoliation with longer intervals reduced rhizome growth but, rela-
tively it increased the production of forage. In other words, the
proportion of rhizome to top growth was changed in favor of the top
growth. At least during one growing season, no defoliation treatment
reduced the rhizome mass. This finding explains the observed excel-
lent persistence of Flori graze rhizoma peanut, and it has particular
relevance since it can almost guarantee its survival under the most
stressful conditions. It had been suggested by Breman (1980), that
rhizome production paralleled forage yield. The results in this
experiment indicate that any defoliation frequency would change the
ratio of rhizome growth to forage growth. Therefore, for establish-
ment purposes, rhizomes should be given preference. These results
confirm the observations of Prine et al. (1981) who stated that no
defoliation should be practiced during the first year to improve estab-
lishment. In a better year, probably N levels would be higher than in
this experiment and growth would be higher, with lower TNC levels.
Rhizoma peanut is a tropical plant; therefore, with the onset of
cooler temperatures in the autumn, growth is reduced more than photo-
synthesis. The results in Chapter II indicated that it changed its
partitioning and assimilates were translocated to the rhizome system
(Chapter III). Defoliated plants in this study do not confirm that
hypothesis, probably because the stress and lack of N also inhibited
bud formation and development in rhizomes.

100

Summary

Plants require active meristematic tissue and/or nutrient reserves
to initiate regrowth after winter or after defoliation. Defoliation
may alter plant morphology. The objectives of this study were to quan-
tify the effect of frequency of defoliation upon the growth and develop-
ment of rhizomes, shoot formation and accumulation of TNC and N. A field
experiment was conducted in 1980 and 1981, 18 km west of Gainesville, FL.
Florigraze rhizoma peanut was planted in 1980. In 1981 three frequencies
of defoliation (2, 6, and 8 weeks) were randomly assigned to 91 hills.
In addition to this, undefoliated plants also planted in 1980 were used
as control. Samples were taken from every defoliation treatment according
to a sampling schedule to record weight of rhizomes and shoots. Samples
were also taken for TNC and N analyses. The growth results were fitted
to a linear model, TNC and N results were fitted to a quadratic model.
All data were statistically analyzed with regression analyses.

Rhizome growth decreased as frequency of defoliation increased.
The number of shoots was not significantly altered by frequency of defol-
iation. Defoliation stimulated shoot growth, and changed the ratio of
shoots to rhizomes. This was also reflected in lower TNC accumulation
in defoliated than in undefoliated plants. The 6- and 8-week defoliation
frequencies decreased N percent in rhizomes indicating that N was mobi-
lized for shoot production. The 2-week defoliation stopped rhizome
growth, although TNC accumulation took place. The latter probably
accounted for the slight increase in weight of the rhizome of plants
under 2-week defoliation frequency. The 6- and 8-week defoliation

101

frequencies did not stop rhizome spreading. Secondary rhizomes main-
tained the same proportions observed with primary rhizomes in unde folia-
ted plants. These results suggested that more growth can be expected
in defoliated plants if enough N is available. Nitrogen percent in-
creased in all defoliation treatments towards the end of the growing
season; however, a deficiency at the beginning might have reduced the
growth potential of the plants.

CHAPTER V

EFFECT OF GRINDING METHOD ON in vi tro ORGANIC
MATTER DIGESTIBILITY OF FORAGES

Introduction

Precise estimation of forage quality with laboratory procedures
greatly facilitates the evaluation of forages. Digestibility of forages
is an important parameter of forage quality (Raymond, 1969). Tilley and
Terry (1963) proposed a two-stage technique that attempts to simulate
the digestive process in ruminants. This in vitro procedure, with a
number of modifications, has become the most widely accepted method of
forage evaluation because it predicts in vivo digestibility with more
accuracy than other procedures (Barnes, 1981; Minson, 1981a). The use-
fulness of the technique is based on the recognition of its limitations.
If the sources of variation are not properly controlled, interpretation
of results may be impossible or useless. Rumen inoculum is the main
source of variation in the technique itself. Therefore, it has been
suggested that comparison among a given set of forage samples should be
made with the same batch of rumen fluid (Clark and Mott, 1960). Baum-
gardt et al., 1962, recommended the inclusion of a standard forage in
every laboratory run to account for variability among batches. Another
source of variation is the length of incubation time (Troelsen and Hanel,
1966; McLeod and Minson, 1969). Special reference is made by McLeod and
Minson (1969) as to the particle size of forage samples since they found
that coarsely ground samples were digested in vitro less rapidly than

102

103

finely ground ones. Worrell (1982) found variability among forage species
for the effect of particle size in vitro digestibility of cell wall, using
only one fermentation time. However, McLeod and Minson (1969) indicated
that the effects of particle size tended to disappear when incubation
time was increased. Grinding of samples for digestibility analyses at
the University of Florida's Animal Nutrition Laboratory is usually per-
formed with a Model 3 Wiley Mill and 1.0-mm screen (J. E. Moore. Per-
sonal communication). However, grinding of similar forage samples with
the same screen size in a Udy Cyclone mill, the particle size appeared
to be smaller and more uniform than with the Wiley mill. Therefore, the
objectives of this study were to determine the effects of mill type and
screen size upon the in vitro organic matter digestibility (IVOMD) of
rhizoma peanut and nine selected grasses, and upon variability in the
prediction of in vivo organic matter digestibility in tropical grasses.

Materials and Methods

A laboratory experiment conducted in 1982 at the University of
Florida's Animal Science Nutrition Laboratory, in Gainesville, FL.
Forage material of Florigraze rhizoma peanut (Arachis glabrata Benth.)
was obtained from a previous field experiment conducted during 1980 and
1981. The plant material represented four stages of maturity of un-
defoliated plants, for 1980 and three stages of maturity for 1981
(Table 14). For every sampling, three composite samples of rhizoma
peanut forage were separated into leaf and stem; therefore, three
samples of leaves and three of their corresponding stems represented
each stage of maturity. The samples had been dried at 70 C for 36 hrs.

104

Table 14. List of grasses and sampling dates for Fl on graze rhizoma
peanut.

No. Grass Samples"'' Regrowth

- weeks -

I Paraguay 22 bahiagrass (Paspalum notatum FlOgge) 4

II Paraguay 22 bahiagrass (Paspalum notatum Flu'gge) 6

III Paraguay 22 bahiagrass (Paspalum notatum Flu'gge) 8

IV Argentina bahiagrass (Paspalum notatum Flugge) 2

V X124-4 digitgrass (Digitaria spp.) 4

VI Survenol a digitgrass (Digitaria X umfolozi Hall) 8

VII Survenola digitgrass (Digitaria X umfolozi Hall) 6

VIII XI 24-4 digitgrass (Digitaria spp.) 2

IX Coastcross 1 bermudagrass (Cynodon dactylon Pers.) 2

Legume Samples'^
Sampling Dates

1980

I 19 July

II 4 Sept.

III 17 Oct.

IV 11 Dec.

One composite sample per grass (from two field res.).

Three composite samples per date (from ten field reps, in 1980 and

7 field reps, in 1981)

1981

V 9 July

VI 5 Sept.

VII 31 Oct.

105

and ground in a Model 3 Wiley mill (4-mm screen) and stored in Whirl pak
bags. Details on field sampling procedures and compositing of samples
were reported in Chapters II and III. From each leaf or stem sample,
four 5-g subsamples were weighed and stored in capped plastic bottles.

In addition to the rhizoma peanut samples, nine grasses (Table 14)
of known in vivo digestibility were selected for this study (Abrams,
1980). The selection was based on results reported by Worrell (1982)
(Table 15). All grasses had higher in vitro NDF digestibility, when
reground in an Intermediate Wiley mill with a 475-nm screen, than
subsamples of forage previously ground through a 4-mm screen in a
Model 3 Wi 1 ey mi 1 1 .

To select the grasses in Table 15, three ranges were established
based on the differences in NDF digestibility between 4-mm and 475 my.
The average digestibility differences were 0.5, 2.1 and 5.0 percentage
units for small, medium, and large differences, respectively. Forage
production details for the grasses were reported by Abrams (1980).
The grass samples had been ground (Model 3 Wiley mill, 4-mm screen)
and stored in plastic containers under controlled temperature and
humidity (J. E. Moore. Personal communication). Twelve 5-g subsamples
were taken from each grass sample and stored in capped plastic bottles.

Grass and legume subsamples were randomly assigned to four grind-
ing treatments with three replications (Table 16). The grinding treat-
ments were 1) Udy cyclone mill 0.5-mm screen, 2) Intermediate Wiley
mill, 0.5-mm screen, 3) Udy cyclone mill 1.0-mm screen, 4) Standard
(Model 3) Wiley mill 1.0-mm screen. The 0.5-mm screen of the Inter-
mediate Wiley mill was modified by replacing the original screen

106

Table 15. Differential IVNDFD in relation to grinding treatment
(Worrell, 1981).

Grinding Treatments
Group"^ GrasI 4 mm 475 lim Difference

IVNDFD {%)

I

Par. -4-

49.0

54.0

5.0

1

II

Par. -6

50.1

55.2

5.1

III

Par. -8

48.1

53.3

5.2

IV

Arg: -2

53.0

55.9

2.9

2

V

'X124-4'-4

62.3

64.1

1.8

VI

Sur. -8

52.7

54.4

1.7

VII

Sur. -6

57.3

58.2

0.9

3

VIII

'XI 24-4 '-2

71.1

71.5

0.4

IX

CCl-2

55.0

55.3

0.3

In vitro Neutral Detergent Fiber Digestibility

"Group numbers denote large (1 ) , medium (2), and small (3) IVNDFD
differences.

See Table 14 for grass identification.

107

•i->

^ 0)

3 _

O •!-

i-

3

o

JZ

^ 00 CVJ
CM ^

"a- 00 cvj
CM ^ r>-.

* 00 CM

CM "a- 1>-

1/1

01

ro ro 00 f*^

CO CO CO CO

00 CO 00 CO

I

3
CO

Ol

c

•a
c

C3

E
E

c

LO IT) O O

LO Lft O O

Lf) in o o

0)

Ol

O O 1— •—

O O f- •—

O O 1— 1—

s.

o

00

>> >^

>>

>. >>

>>•— S^f-

>,,- >,.-

T3 •!- T3

■O •!- TJ •!-

"O "O

=3 3 =) 3

=) 3 =J 3

^ 3 =3 3

(— CM CO ^

1— CSJ CO ^

r— CM CO ^

(/)

(U

"q.

E

(/)

3

(/>

Oi

_l

to

o
o

(U

0)

c
o

(/)
a;
■fj
<a
i.

to

0)
r—

a.
E

oo

<u
o»
la
s-
o

(/)
1/1

CS f—

a.
>, E
c <a

_i i—
a.

>i E

<: 00

a.
E

(O
X
O)

1Q8

with a standard 0.5-nini screen of the Udy cyclone mill (144 holes cm ).
The subsamples were randomly ground and returned to the containers. A
timer was attached to the Wiley mills to grind the samples 2 and 4 minutes
in the Intermediate and Model 3 mills, respectively.

The IVOMD analysis was performed on all samples in three laboratory
runs, confounding the treatment replications with runs. In each run,
three sub-subsamples were taken from every subsample and randomly as-
signed to three lengths of fermentation in stage 1 of the standard ijx
vitro procedure used at the Nutrition Laboratory (Moore and Mott, 1974).
The fermentation times were 24, 48, and 72 hours. The entire run was
started the same day, and 24 hours later, the first sub-subsamples (24
hours incubation) were acidified, and at 72 hours the third sub-
subsamples were acidified. The second stage of the procedure (acid-
pepsin) lasted 48 hours in all cases. Therefore, recovery of organic
matter was carried out in three consecutive days. A diagram showing
how each sample was processed and evaluated in the laboratory is pre-
sented in Table 16. Two standard forages (low and high quality) were
included in each run.

In grasses and legume statistical analysis were done with analysis
of variance and General Linear Model procedures (SAS, Institute Inc.,
1982).

Results and Discussion

The low and high quality standards included in every run had 42.2%
(S.D. = 2.22) and 63.8% (S.D. = 2.19) IVOMD, respectively, at 48 hours
incubation. The variation came from the second run (39.7 and 61.31%

109

IVOMD for low and high quality, respectively). These results confirmed
the observations of Clark and Mott (1960) and Bawmgardt et al . (1962)
who stated that variation should be expected from batch to batch. It
should be pointed out, however, that at 72 hours incubation (no 72
hours samples were included in the first run) the differences due to
batch seem to disappear (52.5 and 71 .1 ; 52.2 and 71.6% IVOMD, for
low and high quality in second and third laboratory runs).

The results of the analyses of variance, for all forage samples,
for grasses, and for legume samples are presented in Table 17. In
the analysis of variance, that included all forage samples (grass and
legume), there were significant grinding- treatment and forage x
grinding-treatment effects on IVOMD. The separation of grasses from
legume samples indicated that, in grasses, the interaction forage x
grinding-treatment remained significant, whereas in legume samples
the same interaction was non-significant (P = 0.0819). The IVOMD
means per grinding treatment (Table 17) indicated that grasses and
legume samples responded differently to grinding treatments. In
grasses, higher IVOMD was obtained with 0.5-mm screen than with l.O-mm
screen. In legume samples, higher IVOMD was obtained with Udy mill
than with Wiley mills regardless of screen-size; although 0.5-mm
screen gave also higher IVOMD than l.O-mm screen within mill-type.
These results indicated that the responses of grasses were related
more to the size of the screen than to the type of mill; while the
responses of the legume samples were related to both mi 11 -type and
size of screen. This might be the reason why part of the variation

no

ape o

0) '

CO

B
(O

I/)

<u
B
3

o>

0)

CO

a>

</)
I/)

re

s.

a;
o

o

0)

o
c

•1"

>

1/1
(U
(/I

>>

c

(U

C9

J- O

II z:

a: o

Q. >

I/)

I.
o

J. o

11 s
a: o

*

*

*

1

Lf)

o

o

[

o

00

o

o

s:

O

o

o

o

1

a

CM

o

o

o

o

1

*

*

*

*

1

CO

o

cr>

o

Lf)

o

o

o

CM

Ol

a\

1

CO

r>.

o

o

o

CM

vo

1

o

o

o

o

o

CO

o

o

1

>

1

r—

o

o

o

o

o

o

o

o

1
1
1

VO

CM

CO

CO

CM

<£)

CO

1
1

CO

CM

1^

1
1
1

*

*

*

1
1

I

00

00

o

o

•a-

1

cn

o

o

CO

1

s:

o

o

o

CO

1

o

•

•

•

•

1

CM

o

o

o

o

cn

(/)

c

*

*

*

*

*

*

*

<o

Oi

00

s:

o

CO

o

o

o

CO

o

o

o

CO

s:

o

o

o

o

o

o

o

cn

o

o

o

o

o

o

o

o

CO

c

>

CM

0)

o

o

o

o

o

o

o

o

E
■P

<0

0)

l~

CM

CO

CO

CM

VO

CO

\-

CM

*

*

*

CO

O

<NJ

o

o

LD

o

CM

o

o

CO

s:

CO

O

o

o

O

o

CM

O

c>

o

O

•— o
o o
o o

o o
o o
o o

* * *

■— I— ^ en

o o CM r»»

O O O C7<

o o o cri

o o

CM CM
CM

o o

CO CD
CO

o o o o

^ CO CM

<a- CO

0)

CD
lO

C i~
3 O

a: LL.

(U

■»-»

o

•

+J

le

01

X

•l->

(0

c

&.

<u

w

o

o

s-

o

u.

1—

U-

(O

X

X

ja

3

O

o

o

u

c

c

c

c

t — (

1 — I

►—4

a
o

ift «a-

CM CM CM

O r- O

CO CM CO
CM CTl CM

ir> Lf) IT)

ir> CO I—

^ O CO

CO o

CM r—

CM 00 CO

CO in ^

C3 I— * C3

1— CM

ir> in

00

c

B

r>» CD LO CO
m in in ID

>>

>»

•o

=)

3

=)

3

E

E

1

E
1

1

E
1

in

m

o

O

1

CD

o

-lO

u

c

•r-

(/I.

to

CJ

■(->
oo

Ill

among forage samples was removed by separating the grasses from the
legume samples.

The results in Table 17 also show that there were differences
in dry-matter content among forage samples which was expected because
of the differences in maturity among forage samples. The analyses of
variance (Table 17) show that separating the forage samples, into
grasses and legume samples, eliminated the forage x grinding-treatment
interactions. The means of dry-matter content in grasses and legume
samples are presented in Table 17. These results indicated that the
dry matter contents in grasses and legume subsamples, ground in the
Udy mill, were consistently higher (1 to 2 percentage units) than in
subsamples ground in the Wiley mills. The loss of moisture caused by
Udy mill is probably due to the air that drives the ground forage off
the grinding chamber into the collection bottle. More moisture was
lost with 0.5-mm screen than with 1.0-mm screen, probably because
of the increased surface area in the fine-ground samples. In Wiley
mills no difference was observed in dry-matter content.

Since the response to grinding -treatments of legumes and grasses
were different, they will be discussed separately.

Grasses

The grinding treatment effects (mill and screen-size) on IVOMD
were plotted against incubation time for each individual grass (Fig.
13). The analyses of variance. Table 18, showed that the separation
of grasses by genus (Paspalum and Digitaria) eliminated the forage x
grinding-treatment interaction. The same thing was observed when the
grasses were separated into the three groups reported in Table 15.

Fig. 13. Effect of grinding treatment and incubation on IVOMD of
nine grasses. (See Table 14 for grass identification).
Dotted line denotes in vivo organic matter digestibility.

113

INCUBATION TIME (hours)

114

Table 18. Analyses of variance for grasses separated by genus, and for
legume samples separated into leaf and stem.

All

Grasses

Paspalum

Diqitaria

D.F.

IVOMD

D.F.

IVOMD

D.F.

IVOMD

Kun

o
c

n nnm*

u . UUU 1

2

0.1147

2

0.0023

Grass

8

0.0001*

3

0.0001*

3

0.0001*

Treat.

3

0.0001*

3

0.0001*

3

0.0001*

Grass x Treat.

24

0.0031*

9

0.3912

9

0.4014

Incubation

2

0.0001*

2

0.0001*

2

0.0001*

Inc. X Grass

16

0.0001*

6

0.0213*

6

0.0023*

Inc. X Treat.

6

0.0008*

6

0.0736

6

0.0457

Inc. X Grass x Treat.

48

0.8936

18

0.4973

18

0.8739

All Legume Samples

Leaf Samples

Stem Samples

D.F.

IVOMD

D.F.

IVOMD

D.F.

IVOMD

Dl in

Kun

2

0.0759

2

0.5987

2

0.0956

Leg

13

0.0001*

6

0.0009*

6

0.0001*

Treat.

3

0.0001*

3

0.0001*

3

0.0001*

Leg. X Treat.

39

0.3173

18

0.0071*

18

0.6590

Incubation

2

0.0001*

2

0.0001*

2

0.0001*

Inc. X Leg.

26

0.0220*

12

0.0588

12

0.3969

Inc. X Treat.

6

0.1699

6

0.2969

6

0.6192

Inc. X Leg. x Treat.

78

0.9999

36

0.9986

36

0.9922

* statistically significant.

115

At first it was thought that these results were due to a loss in sen-
sitivity, in the analyses of variance, caused by the reduction of
degrees of freedom to test the interaction. However, the random for-
mation of groups with the nine grasses showed that the interaction
could be significant with three grasses. These observations suggested
that grasses with similar characteristics (genetic or IVNDFD) responded
to grinding treatments in the same form. The results in Table 18 also
show that the interaction incubation-time x grinding-treatment dis-
appeared when the grasses were separated by forage genus; the inter-
action incubation-time x forage remained significant. These results
indicated that grasses responded differently to incubation time regard-
less of grinding treatment. This kind of response was expected since
the grasses were selected within a wide range of digestibilities;
and it has been reported that there is variation in rates of in vitro
digestion within and between forage species (Gill et al., 1969; Smith
et al., 1972).

The results of the regression analyses to compare the effects of
grinding, as a function of incubation time, on IVOMD of the nine
grasses are presented in Table 19. There were no statistical dif-
ferences among the rates of digestion of grinding treatments, within
grasses; and only Coastcross-1 bermudagrass resulted in no grinding-
treatment effects (P = 0.1804). These results indicate that grinding
treatments did not increase the rate of digestion after 24 hours.
Therefore, the largest effect of grinding must have occurred within
24 hours. In this experiment, the grass responses to incubation can
be classified into three categories. Category 1 includes Paspalum

116

> CM

m ^ f— •— o

i#> o o o f— <T>

m 00 o O CNJ 00

n f— o o tr> CO

o o o o o o

IT) O <— <— 00

rr »— O O LO wo

CM O O r-.

O O O O ^ "—

o o d d d d

to CM

m o

»— <— LO O

O O 00

O o ^

O O rn o>

o o o o o

CO r— r- CM

LD o o O lO f»«»

n o O O »—

O O O O

d d d d d d

Lf) .— ^ ^

CO O O O r-

O O O O

»— o o o ^

d d d d d

>— cn
O O O

O O O lO

o o o m

d d d d

^ r—

O O r*v

O O ro

O O o

o o o o

O O O

o o o

uo O O O

o o o o o

C3> <—.— «— O CM

CM o O O 0^ CM

O O O O 1^ o

o o o o CM

d d d d d d

CM ro »— ^ m m cm

I— oj ^ ro

I— CM CO ^

^ ^ lO

ro CM ^ CT»
in

»— cm CO

I— CM ^ ^

^ CM n

CM ^ ro ^

00 oo vD O

CM t— ro

o <— r--

CM m ^

un \a ^ O
^ O

CM .— rn ^

T3 •-- -O -r-

117

species (Figs. 13-1, 13-11, 13-III.and 13-IV) which showed a nearly
linear response to incubation time. The IVOMD at 24 hours was about
30%; and increasing the incubation from 48 to 72 hours had a small
decrease in the rate of digestion. Category 2 includes digitgrasses
(Figs. U-V, 13-VI, 13-VII and 13-VIII) which showed a response until
48 hours and then digestion rate declined between 48 and 72 hours.
The IVOMD at 24 hours varied between 35 and 45% depending on maturity.
In Category 3 Coastcross-1 bermudagrass stands alone (Fig. 13-IX).
It showed a rapid rate of digestion within 24 hours, with little ad-
ditional digestion between 24 and 48 hours. The IVOMD at 24 hours
was similar to the 2-week regrowth of X124-4 digitgrass; and increasing
incubation from 48 to 72 hours slightly increased IVOMD.

The IVOMD means for grinding treatments (Table 19) show that,
except for Coastcross-1 bermudagrass, grinding with 0.5-mm screen
increased IVOMD regardless of mill. Although, IVOMD with 1 .0-mm
screen was higher with the Udy mill than with Wiley mill. The dif-
ferences between 0.5 and 1.0-mm screens can be as large as 8 percentage
units (Fig. 13-III), or only 2 percentage units (Fig. 13-VII). In
most grasses there was less than one percentage unit between Udy (0.5)
and Wiley (0.5) mills. Conversely, with 1.0-mm screens there were in-
crements of up to 4 percentage units when grinding with Udy compared
to the Wiley mill. . The responses to grinding treatment followed the
pattern adopted as selection criteria for the grasses (Table 15).
This confirmed the results of Worrell (1981) who found variability
among grasses in their response to in vitro NDF digestibility when
finely ground (475 my) vs coarsely ground (4-mm) .

118

The results in this experiment indicated that low quality forages
have low initial rates of digestion, but they can be altered by decreas-
ing the particle size. Medium and high quality forages responded very
little to fine-grinding indicating high initial rates of digestion.
McLeod and Minson (1969) showed that mean in vitro digestibility of
four tropical grasses was increased by reducing the particle size of
the sample. In the same study they found that increasing length of
incubation time increased digestibility and reduced the magnitude of
differences caused by particle size. In this study there was a
tendency to reduce grinding- treatment differences at 72 hours incuba-
tion, but in only the 2-week regrowth of X124-4 digitgrass (Fig. 13-
VIII) were the differences completely eliminated

Legume

The results of grinding- treatment effects (mill and screen size)
were plotted against plant maturity and incubation time (Fig. 14).
Since there were marked differences between leaf and stem IVOMD, both
plant fractions were plotted separately on the same figure. The
results in Table 18 show that the separation of legume samples into
leaf and stem brought up a significant forage x grinding- treatment
interaction in the leaf fraction. This separation also eliminated the
interaction incubation-time x forage. These results indicated that
IVOMD in leaves had the same responses to grinding treatments and to
incubation than stems.

The regression analyses (Table 20) showed grinding- treatment ef-
fects in both years for the leaf and stem fractions. In the legume,
like the grasses, there were no grinding-treatment differences in rates

119

g. 14. Effect of grinding treatment and incubation on IVOMD
of seven samples of Flori graze rhizoma peanut. (See
table 14 for legume sample identification).

120

of digestion. Also, there were no grinding-treatment differences for
linear and quadratic coefficients of plant maturity. These results
indicated that grinding treatments affected IVOMD as in the grasses;
that is, the effect of grinding occurred within 24 hours, and further
incubation did not show an increase in the rates of digestion due to
grinding treatments. The magnitude of the responses, however, was
much smaller in the legume than in the grasses. The ranges of varia-
tion seldom exceeded five percentage units. This is comparable to
the results found for high quality grasses (Figs. 13-VIII, 13-IX).
The response of legume samples to incubation time was also small,
increasing incubation time from 24 to 48 hours yielded about 5
percentage units increase in IVOMD whereas increasing from 48 to 72
hours increased IVOMD less than 2 percentage units. These trends
were similar in both leaves and stems. These results indicated that
about 90% of the digestion occurs within 24 hours. Similar results
were reported by Troelsen and Hanel (1966) who found that alfalfa
(Medicago sativa L.) IVOMD reached a plateau at 24 hours, increasing
only 3.2 percentage units when the incubation time was extended to
48 hours and 0.5 percentage units from 48 to 96 hours. In the same
study, Troelsen and Hanel (1966) found that wheat straw (Triticum
aestivum L.) IVOMD did not reach a plateau at 96 hours.

Since the objectives of this study were to determine the effects
of grinding treatment, the effects of legume age will not be discussed
here. The ranking of treatment means (Table 20) shows that with the
exception of the 1980 leaf fraction, the Udy mill gave higher IVOMD
than the Wiley mills. The reason for this is not completely under-
stood. This response did not occur in the grasses, suggesting that

121

Table 20. Regression analyses per year and plant fraction for treat-
ments and incubation effects upon IVOMD of Flori graze rhizoma
peanut, and treatment means by year and plant fraction.

1980 1981

Leaf

Stem

Leaf

Stem

D.F.

P = F

P = F

D.F.

P = F

P = F

Run

2

0.8223

0.0099

o
c

Treat

3

0.0001

0

n 0324

Inc

1

0.0001

0.0001

1

U.UUU 1

U.UUU 1

l2

1

0.0001

0.0001

1

0.0005

0.0133

I XT

3

0.2328

0.8600

3

0.2191

0.5975

I^.T

3

0.3484

0.6596

3

0.6067

0.8717

AGE

1

0.0001

0.7442

1

0.0001

0.0001

a2

1

0.0199

0.0001

1

0.0001

0.0001

IX A

1

0.0031

0.9546

1

0.4298

0.9908

A XT

3

0.1215

0.2700

3

0.4939

0.8578

A^xT

3

0.6688

0.8913

_3

0.4973

0.0550

MSE

121

85

+

Treatments Treatment Means

Udy (.5mm) 70.3(1) 50.6(1) 69.7(1) 54.0(1)

Wiley (.5 mm) 69.3 (2) 48.8 (3) 67.0 (3) 52.0 (3)

Udy (1.0mm) 68.9(3) 49.7(2) 68.5(2) 53.0(2)

Wiley (1.0 mm) 68.4(4) 47.7(4) 65.6(4) 51.4(4)

Means of three incubation times.

122

the mode of grinding of the mills produced different textures (particle
size distribution) in the samples of legumes. Chenost (1966) reported
grass differences in ease of pulverization as measured by electrical
power. Therefore, it is possible that Flori graze rhizoma peanut is
more fragile than the grasses included in this study, so when it is
processed through a Udy mill, it is pulverized. The answer to this
problem may be found by categorizing the textures of the legume and
grass samples run through the mills. The dry matter content in legume
and grass subsamples ground in the Udy mill were consistently higher
(1 to 2 percentage units) than in subsamples ground in the Wiley
mills. These observations indicated that some moisture was lost with
the Udy mill, especially with the 0.5 mm screen. The grasses in this
study also lost moisture, when ground with a Udy mill, in the same
magnitude as the legume. However, the IVOMD results did not show the
same responses as the legume. Therefore, these results suggest that
ease of physical degradation was a major determinant in the response
of the legume to grinding. These results also point out that the ob-
served in vivo performance (Prine et al . , 1981) of Florigraze rhizoma
peanut may be due to ease of degradation, both physical and chemical.

The IVOMD percentages of the nine grasses were correlated for
each grinding treatment and incubation time to the in vivo data
(Abrams, 1980) of the same grasses. Prediction equations were generated
for organic matter digestibility (OMD) and organic matter intake (OMI)
(Table 21). The results in Table 21 show that incubation for 24 hours

123

Table 21. Regression and correlation coefficients for prediction of

Organic Matter Digestibility (OMD) and Organic Matter Intake
(OMI) from IVOMD of nine grasses with four grinding treat-
ments and with three incubation times.

Mills

Inc

Screen

Udy

Wiley

3

Intercept

r

Intercept

r

hrs

- mm -

IVOMD

vs OMD

24

0.5
1.0

0.72
0.64

32.0
37.6

0.70
0.71

4.3
4.3

0.89
0.53

26.2
42.5

0.78
0.73

3.8
4.1

48

0.5

0.82

12.7

0.95

1.8

1.03

1.3

0.95

1.9

1.0

0.76

19.2

0.95

1.9

0.70

24.5

0.94

2.0

72

1.0

0.99

-3.8

0.88

2.8

0.84

5.7

0.89

2.7

0.5

0.88

5.5

0.90

2.6

0.82

10.1

0.97

1.6

IVOMD vs OMI

0.5

0.55

24.2

1.0

0.51

28.4

0.5

0.59

17.7

1.0

0.54

22.3

0.5

1.03

5.3

1.0

0.88

15.8

0.5

0.68

19.5

1.0

0.59

26.2

0.75 3.4 0.67 17.1 0.74 3.5

0.75 3.4 0.47 31.6 0.75 3.4

0.61 4.1 0.48 24.8 0.60 4.1

0.65 3.9 0.50 25.1 0.69 3.7

— (X IVOMD vs OMD

0.94 2.1 1.12 0.6 0.97 1.5

0.92 2.3 0.75 24.0 0.92 2.3

(X IVOMD) vs OMI

0.73 3.5 0.71
0.72 3.6 0.51

18.2
31.0

0,72 3.6
0.74 3.5

Mean of three incubation times.

124

gave realtively low correlation coefficients regardless of grinding

treatment. Incubation for 48 hours gave high correlation coefficients

in all grinding treatments; there was a slight tendency in the Udy

mill to reduce the standard error of the estimate (s„ ). The

y-x

72 hours incubation gave medium to high correlation coefficients; and
particularly important was the correlation between IVOMD with 1.0-mm
Wiley and OMD, because it was the highest correlation coefficient
obtained in this set of comparisons. The explanation to this can be
found in Fig. 13; 48 hours incubation for all grinding treatments
and almost all grasses, consistently underestimated iji vivo digest-
ibility. At 72 hours most grinding treatments overestimated in vivo
digestibility but l.O-mn Wiley seemed to be closer to the target at
72 hours than any other grinding treatment. In spite of the apparent
better estimation of in vivo digestibility at 72 hours, the prediction
equations with 48 hours incubation gave lower standard errors of
estimate. Similar results were reported by McLeod and Minson (1969).
Data reported by Hartadi (1980) for the same grass samples used in
this study generated similar correlation coefficients to the reported
herein. Hartadi (1980) reported a standard error of estimate of
3.3 using 78 grass samples.

The correlations of grinding treatments and incubation times
with OMI followed the same pattern as the correlations with OMD.
There were not marked differences among grinding treatments in the
prediction of OMI from IVOMD. The incubation at 24 and 72 hours re-
duced the correlation coefficients as compared to 48 hours incubation.

125

The mean IVOMD of three incubation times (X IVOMD) was also
regressed against OMD and OMI. The correlation coefficients obtained
were very similar to the ones obtained with 48 hours incubation.

The results in this study indicated that for comparison purposes
all forage samples must be ground in the same mill and also that the
grinding technique must be standardized if prediction of in vivo data
is required. Weller (1973) found that crude protein was a good predic-
tor of OMD, with a small number of samples; however, the prediction
failed when a large number of samples was involved. In this experiment,
only nine grasses were studied, and even though they represented dif-
ferent quality forages (Moore et al., 1981) more forage species are
needed to corroborate these findings. The results in Figs. 13-VII
and 14-IX for Survenola digitgrass, and Coastcross-1 bermudagrass, re-
spectively, gave conclusive evidence that 48 hours incubation does not
give an indication of the "average" digestibility of forage particles
that occurs in the rumen of animals. Histologic work by Akin et al .
(1977) has shown that the extent of digestion in tissues varies a great
deal .

Summary

Digestibility of forages is an important parameter of forage qual-
ity. Its precise estimation through laboratory procedures facilitates
the evaluation of forages. A laboratory experiment was conducted at
the University of Florida's Animal Science Nutrition Laboratory in
1982. The objectives were to compare Flori graze rhizoma peanut (leaf
and stem) to nine grasses as to the effect of grinding treatments and

126

incubation time (stage I) upon the IVOMD, and to explore the possibility
of reducing variability in the prediction of OMD from IVOMD.

Grinding treatments were 1) Udy Cyclone mill (0.5 mm screen), 2)
Wiley mill (0.5-mm screen), 3) Udy mill (1.0-mm), 4) Wiley mill (1.0-
mm). Experimental design was a split-plot with three replications
(laboratory runs). Grass and legume subsamples were weighed (5 g) and
ground (4 treatments). Three sub-subsamples were randomly allocated
to three incubation times (24, 48, and 72 hours). In every laboratory
run, all incubations were started the same day and recovered on three
consecutive days. Results were statistically analyzed with analyses
of variance and linear regression.

All grasses and legume samples responded to grinding treatments.
In grasses 0.5-mm grinding increased IVOMD regardless of mill. In the
legumes samples 0.5-mm grinding increased IVOMD also, but 1.0-mm Udy
gave higher IVOMD than 0.5-mm Wiley. These results indicated differ-
ences in ease of degradation between grass and legume samples. In both
grasses and legume samples, Udy mill decreased the moisture content in
samples. Grasses responded differently to incubation time, three
categories of responses were summarized. The first one included grass
samples in the genus Paspalum; the second included grass samples in the
genus Digitaria; and the third included Coastcross-1 bermudagrass .
Grinding effects were large in the first group (>10 percentage units);
medium in the second (<10 percentage units); and little in the third
(<5 percentage units). The response of Florigraze rhizoma peanut to
incubation time was very small, with a pattern similar to the third
group of grasses, but at a much higher digestibility. The response

127

to grinding was very small (<5 percentage units) for leaf and stem.

The response of grasses and legume samples indicated that grinding

increased the rate of digestion within 24 hours.

The s for prediction of OMD from IVOMD were lowest with fine-
y-x

grinding and 48 hours incubation and also with medium-grinding and

72 hours incubation, low s„ ^ were also obtained with X IVOMD (mean

y-x

of three incubation times) and fine-grinding.

CHAPTER VI
GENERAL DISCUSSION

The growth of Florigraze rhizoma peanut was studied for 2 years
(growth and regrowth) with a replication of the growth phase in another
year and different field. Laboratory analyses were performed in plant
parts for TNC and N. In 1981 three frequencies of defoliation were
imposed on 1 -year-old plants and their growth monitored. Forage sam-
ples were compared to nine grasses of known in vivo organic matter
digestibility as to the effect of four grinding treatments and three
incubation times on IVOMD.

Dry matter accumulation in plant parts indicated that rhizome
growth paralleled shoot growth until September when shoot growth de-
clined and rhizome growth continued to increase. The rates of growth
in rhizomes were still increasing which indicated that net photosyn-
thetic rates were still above the compensation point. In these studies
it was not completely understood what caused rhizomes to grow more than
shoots. Rhizome formation preceded shoot formation. Therefore, it
would have been expected to observe an increment of shoots as well;
however, this was not observed. The observations of 2 years indicate
that in the growth phase (first year) rhizome growth from September on
was not only related to an increase in the size of rhizomes. The in-
crease was also related to rhizome formation and development. At this
time, shoots emerged and remained small or died. These shoots were

128

129

easily observed in the periphery of the plants. Shoots emerging within
the canopy were also observed; most of them did not develop and usually
died within 2 weeks.

These observations suggested that above ground temperature was
limiting shoot growth. The underground growth was not affected as
much probably because temperature gradients in the soil do not dras-
tically change early in the autumn. In the 1981 regrowth, these trends
were very similar although the extent of changes was altered. It was
suggested that rhizome development was possibly hampered by a defi-
ciency of N, which reduced the potential of rhizome and shoot growth.
Although this hypothesis remains to be tested, it was formulated indi
rectly from the following observations: 1) In summer and autumn,
rhizomes in the 1981 regrowth appeared thicker, although they were not
counted. 2) The N concentration in shoot and rhizome tissues was
rather low as compared to the 1980 and 1981 growth experiments. 3)
The analyses of TNC indicated that the accumulation of TNC accounted
for about 30% of the total weight increment. This comparison was made
with the weight reached by rhizomes in the first year of growth,
which had only about 40% TNC. 4) The reduction in shoot growth was
rather drastic and relatively early in the growing season, indicating
low rates of shoot replacement. Therefore, leaf senescence was probably
responsible for low rhizome growth rates in 1981 regrowth.

In addition to the above mentioned observations, it can be added
that defoliation of plants reduced the allocation of dry matter to
rhizomes, indicating that rhizome growth is enhanced by the absence of
shoot growth. Defoliation stimulated the plants to produce shoots.

130

therefore, rhizome production was reduced. On the other hand, in
defoliated plants, TNC accumulation did not occur in the same pro-
portion as in plants in the 1981 regrowth. This stimulation was
observed in the removal of N from the rhizomes defoliated every 6 or
8 weeks. Therefore, one could expect to find lower levels of TNC in
defoliated plants with no N starvation. These plants would produce
more shoots, and probably more rhizomes; the latter depending on the
frequencies of defoliation. In extreme cases of stress, like in con-
tinuously defoliated plants, rhizome production and development can
be stopped, although TNC and N will accumulate in the remaining
tissues. This experiment did not involve observation of defoliation
on several consecutive years; therefore, questions on rhizome longev-
ity effects, on subsequent regrowths cannot be answered. However, it
is clear that Flori graze rhizoma peanut can withstand the expected
abuse in a dry year.

The observations in these studies suggest that in some years
Flori graze rhizoma peanut may need N fertilizer at the beginning of
the growing season, although this consideration must be studied on
economical basis. From the standpoint of establishment, it would
probably increase the rate of spreading, which in turn may reduce
pasture maintenance costs. Adjei and Prine (1976) reported that
rhizoma peanut did not respond to N; however, the results herein do
not disagree with those observations because those observations were
not made on extremely dry years.

The comparison of Flori graze rhizoma peanut with nine tropical
grasses of various quality degrees indicated that Florigraze rhizoma

131

peanut stands among the best quality forages available for North
and Central Florida. In vivo data had already indicated high organic
matter intake. The results herein indicated that part of the reason
for this may be the high rates of digestion as observed with in vitro
procedure. The responses of Flori graze to grinding treatments and
incubation times are similar to those reported for high quality
temperate legumes.

The responses of Florigraze to defoliation indicated that per-
sistence will not be directly affected, and the quality comparisons
highlighted some of the reasons of the observed performance of
Florigraze rhizoma peanut. Therefore, the results in this study
confirmed and helped to explain some of the morphological and
physiological reasons which make rhizoma peanut a species with great
potential for the tropical areas.

LITERATURE CITED

Abrams, S. M. 1980. Evaluation of kinetic models of ruminant intake
and digestibility utilizing tropical grasses. Ph.D. Dissertation.
University of Florida, Gainesville, FL.

Adjei, M. B. 1975. Establishment techniques for perennial peanuts
(Arachis glabrata Benth.). M.S. Thesis. University of Florida,
Gainesville, FL.

Adjei, M. B., and G. M. Prine! 1976. Establishment of perennial

peanuts (Arachis glabrata Benth.). Proc. Soil Crop Sci . Soc. Fla.
35:50^53.

Akin, D. E., and D. Burdick. 1975. Percentage of tissue types in tropical
and temperature grass leaf blades and degradation of tissues by
rumen microorganisms. Crop Sci. 15:661-668.

Akin, D. E., E. L. Robinson, F. E. Barton, and D. S. Himmelsbach. 1977.
Changes with maturity in anatomy, histochemistry, chemistry, and
tissue digestibility of bermudagrass plant parts. J. Agr. Food
Chem. 25:179-186.

Alexander, C. W., and D. E. McCloud. 1962. CO^ uptake (net photosyn-
thesis) as influenced by light intensity of isolated bermudagrass
leaves contrasted to that of swards under various clipping regimes.
Crop Sci. 2:132-135.

Alios, H. F., and W. V. Bartholomew. 1959. Replacement of symbiotic
fixation by available nitrogen. Soil Sci. 87:61-66.

Andrade, J. M. S. 1979. Smutgrass (Sporobolus poiretii [Roem and

Schult.] Hitchc.) control with grazing management systems. Ph.D.
Dissertation. University of Florida, Gainesville, FL.

Bailey, R. W., J. A. Monro, S. E. Pickmere, and A. Chesson. 1976. Her-
bage hemicellulose and its digestion by the ruminant. In Carbo-
hydrate Research in Plants and Animals. Miscellaneous~Fapers 12;
Landbouwhogeschool Wageningen, The Netherlands.

Barnes, R. F. 1973. Forage testing and its applications. In For-
ages. M. E. Heath, D. S. Metcalfe and R. F. Barnes (eds.Tlowa
State University Press. Ames, Iowa.

Barnes, R. F. 1981. The role of forage in the United States. In:

Forage Evaluation Concepts and Techniques. J. L. Wheeler and R. D.
Mochrie (eds.). Griffin Press Ltd. Netley. South Australia.

132

133

Barta, A. L. 1979. Effect of nitrogen supply on photosynthate par-
titioning, root carbohydrate accumulation and acetylene reduction
in birdsfoot trefoil . Crop Sci . 19:715-718.

Bawmgardt, B. R., M. W. Taylor, and 0. L. Cason.1962. Evaluation of

forages in the laboratory. II. Simplified artificial rumen pro-
cedure for obtaining repea table estimates of forage nutritive
value. J. Dairy Sci. 45:62-68.

Begg, J. E., and M. J. Wright. 1964. Relative effectiveness of top
and basal leaves for the growth and vegetative shoots of reed
canarygrass (Phalaris arundinacea L. Crop Sci. 4:607-609.

Beltranena, R. 1980. Yield, growth and quality of Flori graze rhizoma
peanut (Arachis glabrata Benth.) as affected by cutting height
and frequency. M.S. Thesis. University of Florida, Gainesville,
FL.

Beltranena, R., J. Breman, and G. M. Prine. 1981. Yield and quality
of Flori graze rhizoma peanut (Arachis glabrata Benth.) as affected
by cutting height and frequency. Proc. Soil Crop Sci. Soc. Fla.
40:153-156.

Bergersen, F. J. 1975. Factors controlling nitrogen fixation by

Rhizobia. Biological Nitrogen Fixation in Farming Systems of

the Tropics. A. Ayanaba, P. J. Dart (eds.).John Wiley and Sons
Ltd. New York.

Blaser, R. E., R. H. Brown, and H. T. Bryant. 1966. The relationship
between carbohydrate accumulation and growth of grasses under
different microclimates. Proc. 10th Int. Grassld. Congr. Helsinki,
Finland, pp. 147-150.

Blaxter, K. L., N. M. Graham, and F. W. Wainman. 1956. Some observa-
tions on the digestibility of food by sheep, and on related
problems. Brit. J. Nutr. 10:69-91.

Blickensderfer, C. B., H. J. Haynsworth, and R. D. Roush. 1964. Wild
peanut is promising forage legume for Florida. Crops and Soils.
17:19-20.

Booysen, P. de V., N. M. Tainton, and J. D. Scott. 1963. Shoot-appex
development in grasses and its importance in grassland management.
Herbage Abst. 33:209-213.

Branson, F. A. 1953. Two new factors affecting resistance of grasses
to grazing. J. Range. Mangt. 6:165-171.

Breman, J. W. 1980. Forage growth and quality of Florigraze perennial
peanut (Arachis glabrata Benth.) under six clipping regimes. M.S.
Thesis, University of Florida, Gainesville, FL.

134

Brewbaker, J. L. 1976. The woody legume, leucaena; promising source
of feed, fertilizer and fuel in the tropics. Sem. Int. Ganadena
Trop. Acapulco Gro. Mexico. (Forr). P. 13-27.

Brougham, R. W. 1955. A study in rate of pasture growth. Aust. J.
Agric. Res. 6:804-812.

Brougham, R. W. 1956. Effect of intensity of defoliation on regrowth
of pasture. Aust. J. Agric. Res. 7:377-387.

Brown, R. H., and R. E. Blaser. 1965. Relationships between i^serye
carbohydrate accumulation and growth rate in orchardgrass and tall
fescue. Crop Sci . 5:577-582.

Brown, R. H., and R. E. Blaser. 1968. Leaf area index in pasture
growth. Herb. Abstr. 38:1-9.

Brown, R. H., R. E. Blaser, and H. L. Dunton. 1966a. Leaf-area index
and apparent photosynthesis under various microclimates for dif-
ferent pasture species. Proc. 10th Int. Grassld. Congr. Helsinki,
Finland. P. 108-115.

Brown, R. H., R. B. Cooper, and R. E. Blaser. 1966b. Effects of leaf
age on efficiency. Crop Sci. 6:206-209.

Bryan, W. W., J. P. Sharpe, and K. P. Haydock. 1971. Some factors affect-
ing the growth of lotononis (Lotononis bainesii). Aust. J. Exp.
Agric. Anim. Husb. 11:29-34.

Bryant, H. T., and R. E. Blaser. 1961. Yields and stands of orchard-
grass compared under clipping and grazing intensities. Agron. J.
53:9-11.

Burton, 6. W., R. H. Hart, and R. S. Lowrey. 1967. Improving forage
quality in bermudagrass by breeding. Crop Sci. 7:329-332.

Burton, G. W., F. E. Jackson, and F. E. Knox. 1959. The influence of
light reduction upon the production, persistence, and chemical
composition of coastal bermudagrass, Cynodon dactyl on. Agron. J.
51:537-541.

Campling, R. C. 1970. Physical regulation of voluntary intake. In
Physiology of digestion and metabolism in the ruminant. A. T.
Phillipson (ed.). Oriel Press, Newcastle-upon-Tyne, England.

Carter, J. L., L. A. Garrard, and S. H. West. 1972. Starch degrading

enzymes of temperate and tropical species. Phytochem. 11:2423-2428.

Chatterton, N. J., G. E. Carlson, VI. E. Hungerford, and D. R. Lee. 1972.
Effect of tillering and cool nights on photosynthesis and chloroplast
starch in Pangola. Crop Sci. 12:206-208.

135

Chaves, C. M. 1979. Lignin saponification products and their relation
to forage quality. Ph.D. Dissertation, University of Florida,
Gainesville, FL.

Chenost, M. 1966. Fibrousness of forages: Its determination and its

relation to feeding value. Proc. 10th Int. Grassld. Congr. , Helsinki,
Finland. P. 406-411.

Christiansen, S. 1982. Energy reserves and agronomic characteristics
of four limpograsses [Hemarthria altissima (Poir) Stapf et C. E.
Hubb] for Florida's flatwoods. Ph.D. Dissertation, University of
Florida, Gainesville, FL.

Clark, K. W., and G. 0. Mott. 1960. The dry matter digestion in. vitro
of forage crops. Can. J. Plant Sci . 40:123-129.

Clements, R. J., R. N. Oram, and W. R. Snowcroft. 1970. Variation among
strains of Phalaris tuberosa L. in nutritive value during suirmer.
Aust. J. Agric. Res. 21:661-675.

Colby, W. G., and M. Drake. 1966. Carbohydrate reserves in Orchard-
grass. Proc. lOth Int. Grassld. Congr., Helsinki, Finland. P. 151-
155.

Davidson, J. L., and C. M. Donald. 1958. The growth of swards of

subterranean clover with particular reference to leaf area. Aust.
J. Agric. Res. 9:53-72.

Davidson, J. L., and F. L. Milthorpe. 1965. Carbohydrate reserves in
the regrowth of cocksfoot (Dactyl is glomerata L.). J. Brit.
Grassld. Soc. 20:15-18.

Davidson, J. L., and F. L. Milthorpe. 1966. The effect of defoliation
on the carbon balance in Dactyl is glomerata. Ann. Bot. , N.S.,
30:185-198.

Donefer, E. 1970. Forage solubility measurements in relation to nutri-
tive value. Proc. Natl. Conf. Forage Qual . Util., Nebraska Center
Continuing Ed., Lincoln, NB. P. Q1-Q6.

Donefer, E. , P. J. Niemann, E. W. Crampton, and L. E. Lloyd. 1963.

Dry matter disappearance by enzyme and aqueous solutions to predict
the nutritive value of forages. J. Dairy Sci. 46:965-970.

Duncan, W. G., and J. D. Hesketh. 1968. Net photosynthetic rates,
relative leaf growth rates, and leaf numbers of 22 races of maize
grown at eight temperatures. Crop Sci. 8:670-674.

Duncan, W. G., D. E. McCloud, R. L. McGraw, and K. J. Boote. 1978.

Physiological aspects of peanut yield improvement. Crop Sci. 18:
1015-1020.

136

Eaton, F. M., and D. R. Ergle. 1948. Carbohydrate accumulation in the
cotton plant at low moisture levels. Plant physiol. 23:169-187.

Evans, 6. C. 1972. The quantitative analysis of plant growth. Univ.
of California Press. Davis. CA.

Franca-Dantas, M. S. 1982. Establishment of 'Florigraze' rhizoma
peanut (Arachis glabrata Benth.) intercropped with corn, wheat
and sunflower. Ph.D. Dissertation. University of Florida,
Gainesville, FL.

Gail lard, B. D. E. 1962. The relationship between the cell -wall
constituents of roughages and the digestibility of the organic
matter. J. Agric. Sci . 59:369-373.

Gallaher, R. N., C. 0. Weldon, and J. 6. Futral . 1975. An aluminum
block digester for plant and soil analysis. Proc. Soil Sci. Soc.
Am. 39:803-806.

Gildersleeve, R. R. 1982. Agronomic evaluation of Aeschynomene
species under mob grazing conditions. M.S. Thesis. University
of Florida, Gainesville, FL.

Gill, S. S., H. R. Conrad, and J. W. Gibbs. 1969. Relative rate of in
vitro cellulose disappearance as a possible estimator of digestible
dry matter intake. J. Dairy Sci. 52:1687-1690.

Golding, E. J. 1976. Rational methods for predicting quality and

digestible energy concentration of warm-season forages for ruminants.
Ph.D. Dissertation. University of Florida, Gainesville, FL.

Graber, L. F. 1931. Food reserves in relation to other factors limiting
the growth of grasses. Plant Physiol. 6:43-71.

Gregory, W. C, M. P. Gregory, A. Krapovickas, B. W. Smith, and J. A.
Yarbrough. 1973. Structures and genetic resources of peanuts.
Jjl Peanuts-Culture and Uses. Amer. Peanut Res. Assoc., Inc.
Stillwater, OK. p. 47.

Grof, B., W. A. T. Harding, and R. F. Woolcock. 1970, Effects of cut-
ting on three ecotypes of Stylosanthes guyanensis. Proc. 11th
Int. Grassld. Cong. Surfer's Paradise, Australia. P. 226-230.

Grovum, W. L., and U. J. Williams. 1973. Rate of passage of digesta
in sheep. 3. Differential rates of passage of water and dry
matter from the reticulo-rumen, abomasum and caecum and proximal
colon. Brit. J. Nutr. 30:23U240.

Hacker, J. B. 1971. Digestibility of leaves, leaf sheaths and stems
in setaria. J. Aust. Ins. Agric. Sci. 37:155-156.

137

Hacker, J. B., and D. J. Minson. 1981. The digestibility of plant
parts. Herbage Abstracts. 51:459-482.

Hanna, W. W.. W. G. Monson, and G. W. Burton. 1976. Histological
and in vitro digestion study of 1 and 4 week stems and leaves
from~high and low quality bermudagrass genotypes. Agron. J. 58:
219r222.

Hartadi, Hari . 1980. Prediction of the quality of tropical grasses
for ruminants by laboratory analyses and sumative equations. M.S.
Thesis, University of Florida, Gainesville, FL.

Haslemore, R. M., I. J. Warrington, and P. G. Roughan. 1980. Influence
of drying method and post-harvest conditions on total nitrogen,
soluble sugar, and starch levels in plant tissue. N.Z. J. of
Agric. Res. 23:355-359.

Heaney, D. P. 1970. Voluntary intake as a component of an index to
forage quality. Proc. Nat. Conf. Forage Quality Eval . Util.
Lincoln, NB. P. Cl-ClO.

Hermann, F. J. 1954. A synopsis of the genus Arachis. USDA Monograph
No. 19.

Higgins, B. B. 1951. Origin and early history of the peanut. In.

The Peanut-The Unpredictable Legume. The National Fertilizer As-
sociation, Washington, D.C. p. 18.

Hojjati, S. M., R. A. McCreery, and W. E. Adams. 1968. Effects of
Nitrogen and Potassium fertilization, irrigation and clipping
interval on the chemical composition of Coastal Bermudagrass
(Cynodon dactyl on (L.) Pers. I. Total Available Carbohydrates.
Agron. J. 60:617-619.

Holt, D. A., and A. R. Hilst. 1969. Daily variation in carbohydrate
content of selected forage crops. Agron. J. 61:239-242.

Humphreys, L. R. 1966. Pasture defoliation: A review. J. Aust. Inst.
Agric. Sci . 32:93-105.

Humphreys, L. R. 1981. Environmental adaptation of tropical pasture
plants. Macmillan Pub. Ltd., London, England.

Humphreys, L. R., and A. R. Robinson. 1966. Interrelations of leaf

area and nonstructural carbohydrate status as determinants of the

growth of sub- tropical grasses. Proc. 10th Int. Grassld. Congr.
Helsinki, Finland. P. 113-116.

Hutton, E. M., and L. B. Beall. 1977. Breeding of Macroptilium
atropurpureum. Trop. Grassld. 11:15-:31.

138

Imrie, B. C. 1971. The effects of seventy of defoliation and soil
moisture stress on Desmodium intortum. Aust. J. Exp. Agric. Anim.
Husb. 11:521-524.

Jarrige, R. 1965. The composition of sheep faeces and its relation

to forage digestibility. Proc. 9th Int. Grassld. Congr. , Sao Paulo,
Brazil. P. 809-814.

Jewiss, 0. R. 1972a. Vegetative growth. In Grasses and legumes in
British agriculture. C. R. W. Spedding and E. C. Diekmahns (eds.)
Bull. 49 CAB. England.

Jewiss, 0. R. 1972b. Tillering in grasses— its significance and control.
J. Br. Grassld. Soc. 27:65-82.

Jones, D. I. H., and M. V. Hayward. 1975. The effect of pepsin pre-
treatment of herbage on the predictions of dry matter digestibility
from solubility in fungal cellulase solution. J. Sci. Food Agric.
26:711-718.

Jones, R. J. 1967. Effects of close cutting and nitrogen fertilizer
on growth of Siratro (Phaseolus atropurpureus) pasture at Samford,
south-eastern Queensland. Aust. J. Exp. Agric. Anim. Busb. 7:157-
161.

Jones, R. J. 1974. Effect of an associate grass, cutting interval, and
cutting height on yield and botanical composition of Siratro pastures
in a subtropical environment. Aust. J. Exp. Agric. Anim. Husb. 14:
334-342.

Karbassi, P., S. H. West, and L. A. Garrard. 1972. Amilolytic activity
in leaves of a tropical and temperate grass. Crop Sci. 12:58-60.

Knight, R. 1970. The effects of plant density and frequency of cutting
on the growth of cocksfoot (Dactyl is glomerata L.) 1. The
production of vegetative and reproductive tillers. Aust. J. Aric.
Res. 21:9-17.

Kretschmer, A. E. 1976. Growth and adaptability of Centrosema species
in south Florida. Proc. Soil Crop Sci. Soc. Fla"! 36:164-168.

Langer, R. H. M. 1954. A study of leaf growth in timothy (Phleum
pra tense). J. Br. Grassld. Soc. 9:275-284.

Laredo, M. A., and D. J. Minson. 1973. The voluntary intake, digesti-
bility and retention time by sheep of leaf and stem fractions of
five grasses. Aust. J. Agric. Res. 24:875-888.

Laredo, M. A., and D. J. Minson. 1975. The voluntary intake and

digestibility by sheep of leaf and stem fractions of Lolium perenne.
J. Br. Grassld. Soc. 30:73-77.

I

139

Ludlow, M. M., and D. A. Charles-Edwards. 1980. Analysis of regrowth
of a tropical grass legume sward subjected to different frequencies
and intensities of defoliation. Aust. J. Agric. Res. 31:673-692.

Ludlow, M. M., and G. L. Wilson. 1970. Studies on the productivity
of tropical pasture plants. II. Growth analysis, photosynthesis,
and respiration of 20 species of grasses and legumes in a con-
trolled environment. Aust. J. Agric. Res. 21:183-194,

Ludlow, M. M., and G. L. Wilson. 1971. Photosynthesis of tropical
pasture plants. 1. Illuminance, carbon dioxide concentration
leaf temperature, and leaf-air vapour pressure difference. Aust.
J. Biol. Sci. 24:449-470.

Marten, G. C. 1981. Chemical, in vitro and nylon bag procedures for
evaluating forage in the U.S.A. Iji Forage Evaluation Concepts
and Techniques. J. L. Wheeler, R. D. Mochrie (eds.) Griffin Press
Ltd., Netley, South Australia.

Martens, D. R., and L. 0. Ely. 1979. A dynamic model of fiber diges-
tion and passage in the ruminant for evaluating forage quality.
J. Anim. Sci. 49:1085-1095.

May, L. H., and J. L. Davidson. 1958. The role of carbohydrate reserves
in regeneration of plants. Aust. J. Agric. Res. 9:767-777.

McCaig, T. N., and J. M. Clarke. 1982. Seasonal changes in nonstruc-
tural carbohydrate levels of wheat in oats grown in a semi arid en-
vironment. Crop Sci. 22:963-970.

McCollum, R. E. 1978. Analysis of potato growth under differing P
regimes. I. Tuber yields and allocation of dry matter and P.
Agron. J. 70:51-57.

Mclntyre, G. I. 1967. Environmental control of bud and rhizome develop-
ment in the seedling of Agropyron repens L. Beauv. Can J. Bot.
45:1315-1326.

McLeod, M. N., and D. J. Minson. 1969. Sources of variation in the in.
vitro digestibility of tropical grasses. J. Brit, Grassl . Soc.
24:244-249,

McLeod, M,N.,andD. J. Minson. 1971. The error of predicting pasture
dry matter digestibility from four different methods of analysis
for lignin. J. Br. Grassl. Soc. 26:251-256.

McLeod, M. N., and D. J. Minson. 1978, The accuracy of the pepsin-
cellulase technique for estimating the dry matter digestibility
in vivo of grasses and legumes. Animal Feed Sci. and Tech. 3:277-

-m-. —

140

Milthorpe, F. L., and J. Moorby. 1979. An Introduction to Plant
Physiology. Cambridge Univ. Press. Cambridge.

Minson, D. J. 1972. The digestibility and voluntary intake by
sheep of six tropical grasses. Aust. J. Exp. Agr. Anim. Husb.
12:21r27.

Minson, D. J. 1976. Relation between digestibility and composition

of feed— a review. In Carbohydrate research in plants and animals.
Miscellaneous Papers 12; Landbouwhogeschool , Wageningen, The
Netherlands.

Minson, D. J. 1981a. An Australian view of laboratory techniques

for forage evaluation. In Forage Evaluation Concepts and Tech-
niques. J. L. Wheeler, R. D. Mochrie (eds .)Griffin Press Ltd.,
Netley, South Australia.

Minson, D. J. 1981b. Characteristics of the nutritionally ideal graz-
ing plant. In Forage Evaluation Concepts and Techniques. J. L.
Wheeler, R. D7 Mochrie (eds.). Griffin Press Ltd., Netley, South
Australia.

Moore, J. E., and G. 0. Mott. 1973. Structural inhibitor of quality

in tropical grasses. In Anti -quality Components of Forages. Spec
Pub. No. 4, Crop Sci . Soc. Amer., Madison, WI.

Moore, J. E., and 6. 0. Mott. 1974. Recovery of residual oi^lf J ^ mat-
ter from in vitro digestion of forages. J. Dairy Sci . 57:lZ58-lZby

Moore, J. E., M. A. Worrell, S. M. Abrams, W. R. Ocumpaugh, and G. 0.
Mott. 1981. Quality of tropical perennial grass hays. Florida
Beef Cattle Research Report. IFAS. University of Florida, Gaines-
ville, FL.

Mott, G. 0. 1959. Animal variation and measurement of forage quality.
Symposium on forage evaluation IV. Agron. J. 51:223-226.

Mott, 6. 0. 1973. Evaluating forage production. In: Forages. M. E.
Heath, D. E. Metcalfe, and R. F. Barnes(eds.) Iowa State University
Press, Ames, 10.

Cakes, A. J. 1968. Leucaena leucocephala. Description, culture, uti-
lization. Advancing Frontiers of Plant Sci. ( India). 20:1 -14.

Otero, V. R. 1946. 0 capim Kikuyu (Pennisetum clandestinum Horscht.)
2nd ed. Ministerio de Agricultura. Brazil. Departamento Nacio-
nal de Produ?ao Animal .

Phillips, I. D. J. 1969. Apical dominance. In The Physiology of
Plant Growth and development. M. B. Wilkins (ed.). McGraw Hill.
New York.

141

Poppi, D. P., D. J. Minson, and J. H. Ternouth. 1981a. Studies of
cattle and sheep eating leaf and stem fractions of grasses. I.
The voluntary intake, digestibility and retention time in the
reticulo- rumen. Aust. J. Agric. Res. 32:99-108.

Poppi, D. P., D. J. Minson, and J. H, Ternouth. 1981b. Studies of
cattle and sheep eating leaf and stem fractions of grasses. II.
Factors controlling the retention of feed in the reticulo- rumen.
Aust. J. Agric. Res. 32:109-121.

Prine, G. M. 1964. Forage possibilities in the genus Arachis. Proc.
Soil Crop Sci. Soc. Fla. 24:187-196.

Prine, G. M. 1973. Perennial peanuts for forage. Proc. Soil Crop
Sci. Soc. Fla. 32:23-35.

Prine, G. M. 1979. Utilization of legumes to minimize nitrogen fer-
tilization. Proc. 28th Beef Cattle Short Course. I FAS. Univer-
sity of Florida. P. 147-154.

Prine, G. M., L. S. Dunavin, J. E. Moore, and R. D. Roush. 1981.

Flori graze rhizoma peanut a perennial forage legume. Univ. Fla.
Agric. Exp. Stn. Circ. S-275.

Radford, P. J. 1967. Growth analysis formulae— Their use and abuse.
Crop Sci. 7:171-175.

Raymond, W. F. 1969. The nutritive value of forage crops. Adv. in
Agron. 21:1-108.

Reid, D. 1966. Studies on cutting management of grass-clover swards.
4. The effects of close and lax cutting on the yield of herbage
from swards cut at different frequencies. J. Agric. Sci., Camb.
66:101-106.

Rhue, R. D., and J. B. Sartain. 1979. A survey of the fertility status
of Florida soils as indicated by selected soil tests results. Proc.
Soil Crop Sci. Soc. Fla. 38:112-116.

Ridgman, W. J. 1960. The effect of height of cutting on the subsequent
growth and yield of lucerne. J. Br. Grassld. Soc. 15:291-295.

Rodrigue, C. B., and N. N. Allen. 1960. The effect of fine grinding of
hay on ration digestibility, rate of passage, and fat content of
milk. Can. J. Anim. Sci. 40:23-29.

Rossiter, R. C, and W. J. Collins. 1980. The responses to defoliation
of two strains of subterranean clover, differing in growth habit,
when grown in swards. Aust. J. Agric. Res. 31:77-87.

142

Russo, Sandra L. 1981. Comparison of methodologies for determination
of content and digestibility of hemicellulose and cellulose in
tropical grass hays. Ph.D. Dissertation. University of Florida,
Gainesville, FL.

Sachs, T., and K. V. Thimann. 1964. Release of lateral buds from
apical dominance. Nature, Lond. 201:939-940.

Saldivar, A. J., W. R. Ocumpaugh, G. M. Prine, J. E. Hentges. 1981.
Creep grazing 'Flori graze' rhizoma peanut with beef calves.
Southern Branch ASA. p. 13.

SAS Institute Inc. SAS User's Guide: Statistics, 1982 Edition. Cary,
NC: SAS Institute Inc.

Smith, D. 1968. Classification of several North American grasses as
starch or fructosan accumulators in relation to taxonomy. J. Brit.
Grassld. Soc. 23:306-309.

Smith, D. 1975. Trends in nonstructural carbohydrates in the stem
basis of switchgrass. J. Range Mgmt. 28:389-391.

Smith, D. 1981. Removing and analyzing total nonstructural carbohy-
drate from plant tissue. Wisconsin Agric. Bull. R2107.

Smith, D., and L. F. Graber. 1948. The influence of top growth

removal on the root and vegetative development of biennial sweet-
clover. J. Amer. Soc. Agron. 40:818-831.

Smith, D., and J. P. Silva. 1969. Use of carbohydrate and Nitrogen
root reserves ih the regrowth of alfalfa from greenhouse experi-
ments under light and dark conditions. Crop Sci . 9:464-467.

Smith, L. W., H. K. Goering, and C. H. Gordon. 1972. Relationships of
forage compositions with rates of cell wall digestion and in-
digestibility of cell walls. J. Dairy Sci. 55:1140-1147.

Snyder, F. W., and G. E. Carlson. 1978. Photosynthate partitioning
in sugarbeet. Crop Sci. 18:657-661.

Sullivan, J. T. 1969. Chemical composition of forages with reference
to the needs of the grazing animal. ARS USDA 34-107.

Sullivan, J. T., and V. G. Sprague. 1943. Composition of the roots
and stubble of perennial ryegrass following partial defoliation.
Plant Physiol. 18:656-670.

Thimann, K. V. 1937. On the nature of inhibitions caused by auxin. Am.
J. Bot. 24:407-412.

143

Tilley, J. M. A., and R. A. Terry. 1963. A two-stage technique for the
in vitro digestion of forage crops. J. Br. Grassld. Soc. 18:104-111.

Troelsen, J. E., and Donna J. Hanel . 1966. Ruminant digestion in_ vitro
as affected by inoculum donor, collection day, and fermentation
time. Can. J. Anim. Sci. 46:149-156.

USDA. 1976. Special soil survey report maps and interpretations,
Alachua County, Florida. SCS/USDA and FAES. University of
Florida, Gainesville, FL.

Valle, L. S. 1977. Changes in smutgrass CSporobolus poiretii (Roem.
and Schult.) Hitchc.l ground cover induced by spraying with
molasses and grazing management. Ph.D. Dissertation. University
of Florida, Gainesville, FL.

Valle-Melendez, R. R. 1981. Physiological aspects of maize (Zea mays
L.) yield. Ph.D. Dissertation. University of Florida, Gainesville,
FL.

Van Soest, P. J. 1965. Symposium of factors influencing the voluntary
intake of herbage by ruminants: Voluntary intake in relation to
chemical composition and digestibility. J. Anim. Sci. 24:834-843.

Van Soest, P. J. 1967. Development of a comprehensive system of feed
analyses and its application to forages. J. Anim. Sci. 26:119-128.

Van Soest, P. J. 1973. Composition and nutritive value of forages. In_
Forages. M. E. Heath, D. S. Metcalfe, and R. F. Barnes (eds.).
Iowa State University Press, Ames, 10.

Van Soest, P. J. 1976. The estimation of digestibility from chemical
composition. In_ Carbohydrate research in plants and animals.
Miscellaneous Papers 12. Landbouwhogeschool , Wageningen, The
Netherlands.

Van Soest, P. 0., and L. A. Moore. 1965. New chemical methods for

analysis of forages for the purpose of predicting nutritive value.
Proc. 9th Int. Grassld. Congr. , Slo Paulo, Brazil. P. 783-789.

Varnell, R. J., and D. E. McCloud. 1975. Germplasm preservation and
genotype evaluation in Arachis (Peanuts). Report of a Workshop.
University of Florida, Gainesville, FL.

Ventura, M. , J. E. Moore, 0. C. Ruelke, and D. E. Franke. 1975. Effect
of maturity and protein supplementation on voluntary intake and
nutrient digestibility of Pangola digitgrass hays. J. Anim. Sci.
40:769-774.

144

Vickery, P. J. 1981. Pasture growth under grazing. In Grazing
Animals. F. H. W. Morley (ed.). Elsevier Sci . Pub. Co.,
Amsterdam.

Waldo, D. R.. L. W. Smith, and E. L. Cox. 1972. Model of cellulose
disappearance from the rumen. J. Dairy Sci. 55:125-129,

Wardlaw, I. F. 1967. The effect of water stress on translocation in
relation to photosynthesis and growth. I. Effect during grain
development in wheat. Aust. J. Biol. Sci. 20:25-39.

Wardlaw, I. F. 1968. The control and pattern of movement of carbo-
hydrates in plants. Bot. Rev. 34:79-105.

Watson, D. J. 1963. Some features of crop nutrition. In The Growth
of the Potato. J. D. Ivins and F. L. Milthorpe (edj. Butterworths
London, England.

Weinmann, H. 1947. Determination of total available carbohydrates in
plants. Plant Physiol . 22:279-290.

Weinmann, H. 1948. Underground development and reserves of grasses.
A review. J. Br. Grassld. Soc. 3:115-140.

Weinmann, H. 1952. Carbohydrate reserves in grasses. Proc. 6th
Int. Grassld. Congr. College Park, PA, USA. pp. 655-660.

Weinmann, H. 1961. Total available carbohydrates in grasses and legumes
Herbage Abstr. 31 :255^261.

Weinmann, H., and E. P. Goldsmith. 1948. Underground reserves of
Cynodon dactyl on. Better turf by research. African Explosives
and Chemical Industries Ltd., and S. African Turf Research Fund.
Johannesburg, South Africa.

Weller, F. E 1973. Prediction of voluntary intake and nutrient digesti-
bility of warm-season grasses by laboratory methods. M.S. Thesis.
University of Florida, Gainesville, FL.

West, S. H. 1969. Carbohydrate metabolism in tropical plants subjected
to low temperatures. Soil Crop Sci. Soc. Fla. Proc. 29:264-267.

Whiteman, P. C. 1970. Seasonal changes in growth and nodulation of
perennial tropical pasture legumes in the field. II. Effects of
controlled defoliation levels on nodulation of Desmodium intortum
and Phaseolus atropurpureus. Aust. J. Agric. Res. 21 :207^214.

Whiteman, P. C, and A. Lulham. 1970. Seasonal changes in growth and
nodulation of perennial tropical pasture legumes in the field. I.
The influence of planting date and grazing and cutting on Desmodium
uncinatum and Phaseolus atropurpureus. Aust. J. Agric. Res. 21:
195r206.

145

Williams, W. A., R. S. Loomis, and C. R. Lepley. 1965. Vegetative

growth of corn as affected by population density. II. Components
of growth net assimilation rate and leaf-area index. Crop Sci.
5:215-219.

Wolf, D. D. 1978. Nonstructural carbohydrate and dry matter relationships
'in alfalfa tap roots. Crop Sci. 18:690-692.

Worrell M. A. 1982. Structural degradation of tropical grasses for
the prediction of digestibility and voluntary intake by sheep.
M.S. Thesis. University of Florida, Gainesville, FL.

BIOGRAPHICAL SKETCH

Abelardo Jos€ Saldfvar Fitzmaurice was born in Nuevo Laredo,
Tamaulipas, Mexico, on June 13, 1953. His parents are Abelardo Saldfvar
Silva and 01 ga Fitzmaurice de Saldfvar. He graduated from Escuela Nacio-
nal de Agricultura, Chapingo, in 1976 with an "Ingeniero Agrdnomo Zoo-
tecnista" degree. He has attended the University of Florida since
September, 1977, where he was awarded a Master of Science degree in
1979. He is currently a Ph.D. candidate. Upon completion of his grad-
uate work, he will return to Mexico where he will teach and conduct
research in forages at the Colegio de Postgraduados, Chapingo.

146

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

William R. Ocumpaugh, Chairmait'
Associate Professor of Agronoioy

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

William G. Blue
Professor of Soil Science

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

'^j/Dhn E. Moore
.professor of Animal Science

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

/

, Gerald 0. Mott
Professor of Agronomy

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.

Gordon M, Prine
Professor of Agronomy

This dissertation was submitted to the Graduate Faculty of the College
of Agriculture and to the Graduate Council, and was accepted as partial
fulfillment of the requirements for the degree of Doctor of Philosophy.

April 1983

Dean, C^bllege of Agricu^Tture

0

Dean for Graduate Studies and
Research