Group size in wedge-capped capuchin monkeys (Cebus olivanceus) :vulnerability to predators, intragroup and intergroup fe...

MISSING IMAGE

Material Information

Title:
Group size in wedge-capped capuchin monkeys (Cebus olivanceus) :vulnerability to predators, intragroup and intergroup feeding competition
Physical Description:
vii, 126 leaves : ill. ; 28 cm.
Language:
English
Creator:
Srikosamatara, Sompoad, 1955-
Publication Date:

Subjects

Subjects / Keywords:
Cebus olivaceus -- Behavior   ( lcsh )
Cebus olivaceus -- Feeding and feeds   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1987.
Bibliography:
Includes bibliographical references (leaves 116-125).
Statement of Responsibility:
by Sompoad Srikosamtara.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000950282
notis - AER2466
oclc - 16927937
System ID:
AA00003381:00001

Full Text













GROUP SIZE IN WEDGE-CAPPED CAPUCHIN MONKEYS (Cebus olivaceus):
VULNERABILITY TO PREDATORS,
INTRAGROUP AND INTERGROUP FEEDING COMPETITION


By

SOMPOAD SRIKOSAMATARA


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


UNIVERSITY OF FLORIDA


1987















ACKNOWLEDGEMENTS


To study in one foreign land and carry out my research in

another, I owe my favor and gratitude to a number of people.

Dr. John G. Robinson deserves to be mentioned first, not

only as my major academic advisor but also for his personal

kindness to me. He introduced me to both the University of

Florida and the llanos of Venezuela. Without him this research

would have been impossible.

I would like also to thank my supervisory committee, in

alphabetical order, Drs. John F. Eisenberg, John H. Kaufmann,

Richard A. Kiltie and Francis E. Putz, for their supervision and

encouragement.

In the field site in Venezuela, many people helped to create

a good atmosphere for research. Tomas Blohm was a tireless host.

He helped me in numerous ways in the llanos, in Caracas and in

between. Stu Strahl, an American friend in a foreign land, also

helped me in numerous ways, especially with the cars. Renee

Rondeau carried me out of the ditch and managed to get me to the

hospital after a car accident. Researchers and their associates

at the ranch made living there a joy, including Stu and Lisa

Strahl, Theresa Pope, Dennis Daneke, Mark Ludlow and Renee

Rondeau.








A number of sources supported my stay in Gainesville. My

aunt, Ms. Malee Poungpat paid for my air ticket to go to the

University of Florida. The Primate Conservation Program at the

University of Florida provided a scholarship for the first two

years. My stay during the last two years was supported by

scholarships provided by the Jessie Smith Noyes Foundation and

DSR-B research assistantship from the University of Florida to

Dr. Robinson.

This research was supported by a Sigma Xi Grant-in-Aid for

Research, a Friend of National Zoo award to Dr. Eisenberg and Dr.

Robinson, and a National Science Foundation grant (BSR-8300035) to

Dr. Robinson. The Smithsonian Institution allowed me to use its

vehicles.

My data analysis was supported by both personal and public

computer facilities. Gustavo Fonseca, Jody Stallings, and Dr.

Robinson allowed me to use their APPLE IIe microcomputers. I

also used an IBM PC belonging to the Department of Zoology, the

IFAS VAX Computer Network, and the Northeast Regional Data Center

(NERDC) IBM mainframe of the State University System of Florida.

For the ability to stay in foreign lands for five years in a

row I give all credits to the foregoing. I am grateful to my

grandparents for their gentle care and for giving me patience and

strength.


iii


















TABLE OF CONTENTS


ACKNOWLEDGEMENTS.ooooo..... o.................................oo o o ii

CHAPTERS

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

Group Size in Primates.............................. 2
Factors Affecting Group Size in Social Animals...... 4
Group Living in Primates: Food or Predators?....... 6
Cebus olivaceus: a Candidate for Testing
Wrangham (1980) and van Schaik (1983)'s
Hypotheses........................................ 8

II STUDY SITE AND METHODS............................... 9

Study Site.......................................... 9
Study Animals....................................... 10
Methods.......................... ................... 17

III GROUP SIZE AND INTRAGROUP FEEDING COMPETITION....... 20

Introduction........................................ 20
Methods .................. ..................... ...... 21
Results ....... .............. .. .. .... ............. 27
Discussion.......................................... 54

IV GROUP SIZE AND VULNERABILITY TO PREDATORS........... 56

Introduction........................................ 56
Methods............................................. 57
Results............................................. 59
Discussion.......................................... 67









page

V GROUP SIZE AND INTERGROUP FEEDING COMPETITION ....... 69

Introduction........................................ 69
Methods........ ...... .......................... ..... 70
Results.......................... ....... .......... 76
Discussion.......................................... 110

VI CONCLUSION.................................. ........ 114

BIBLIOGRAPHY.............................................. 116

BIOGRAPHICAL SKETCH.......... ............................. 126
















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



GROUP SIZE IN WEDGE-CAPPED CAPUCHIN MONKEYS (Cebus olivaceus):
VULNERABILITY TO PREDATORS,
INTRAGROUP AND INTERGROUP FEEDING COMPETITION

By

Sompoad Srikosamatara

May, 1987


Chairperson: John G. Robinson
Major Department: Zoology


Most primates live in social groups. Do primates in social

groups benefit from increased feeding advantages or reduced

vulnerability to predators? This study addressed these

questions.

Wedge-capped capuchin monkeys, Cebus olivaceus, which occur

in the llanos of Venezuela, are ideal subjects for studies of the

advantages of group living in primates. The animals live in

groups that range in size from five to more than 40 animals.

Group ranges overlap completely. The abundance of their

resources varies spatially and seasonally. This monkey

population has been the subject of a long-term study since 1977.

Two groups of C. olivaceus were studied for a 17 month

period from May to July 1982 and from May 1983 to June 1984. One








group (26-36 animals) was about three times as large as the other

(9-5 animals).

Intragroup feeding competition was more evident in large

groups than in small groups. Animals were more spread out

vertically. Agonistic interactions were more frequent. Animals

spent less time feeding on fruit resources. Large groups moved

farther, had an ongoing movement pattern, backtracked less, and

used a larger area. However, the interpretation of these

differences was difficult because animals in the two groups used

different resources.

Seasonal comparisons indicated that feeding competition

occurred in both groups. During periods of food scarcity,

animals in both groups shifted from clumped food sources (fruit)

to more evenly distributed food sources (insects), thus

alleviating the effect of intragroup feeding competition.

Animals in large groups spent less time in vigilance and more

time on the ground. These results are consistent with

expectations based on decreased vulnerability to predators in

large groups.

There is also among-group variation in foraging success.

Large groups displace smaller groups, altering the use of time,

movements, and use of space of these groups. Large groups have

access to more resources, and are able to monopolize rare,

spatially restricted food sources.


vii
















CHAPTER I
INTRODUCTION



Most primates live in social groups of various sizes. There

are solitary animals such as lorises, monogamous groups such as

gibbons and siamangs, and multifemale groups such as baboons,

macaques and capuchin monkeys. There are two general ecological

hypotheses that address the question of why most primates live in

social groups. One suggests that animals in groups have foraging

advantages. The other is that group living animals are less

vulnerable to predators. The first has been championed recently

by Wrangham (1980), who has argued that animals living in groups

will have feeding advantages relative to solitary animals. The

second has been championed by van Schaik (1983), who proposed

that predator vulnerability is the primary factor encouraging

living in groups. Van Schaik also pointed out that social

animals will have to cope with higher intragroup feeding

competition and as a result, animals in groups will have a lower

feeding efficiency than solitary animals.

Wedge-capped capuchin monkeys, Cebus olivaceus

(=nigrivittatus, following Honacki et al., 1982) were used to

test the two hypotheses outlined above. First I review the

literature on group size in primates and the factors affecting

group size in social animals in general. The two hypotheses








proposed by Wrangham (1980) and van Schaik (1983) are discussed

and a number of predictions based on these hypotheses are tested.

Lastly, the characteristics of C. olivaceus that make this

species ideal for a test of these hypotheses are presented.



Group Size in Primates


Most primates live in social groups of different sizes

(Jolly, 1985; Chalmers, 1980; Terborgh, 1983). Species that live

in large groups tend to be found in more open areas, while forest

living primates seem to live in smaller groups (Crook and

Gartlan, 1966; Crook, 1970; Eisenberg et al., 1972). Within a

species, group size is often smaller at lower population

densities (examples, Table 1-1). When population size decreases,

group size also decreases. This has been found in yellow

baboons, Papio cynocephalus, vervet monkeys, Cercopithecus

aethiops, and Toque macaques, Macaca sinica (Table 1-1), but not

in Hanuman langurs, Presbytis entellus (Table 1-1).

Early workers suggested that particular group sizes were

typical for a given species (Carpenter, 1942, 1954; Crook and

Gartlan, 1966). More recent work has indicated that there is

considerable intraspecific variation in group size, both in

different areas (e.g., baboon, DeVore and Hall, 1965; howler

monkey, Eisenberg, 1979b) and at specific sites (e.g., baboon,

Altmann and Altmann, 1970). This variation has been examined

statistically by van Schaik (1983). Van Schaik found that in

interspecific comparisons, variation (measured using coefficient

of variation) in group size increased with mean group size. The


















N
r *4
OW
*i-4

0D ^

3*S
O..


w co
1 00
to cn


-4 00
V-4


e o





N N


C'a C
m N
O0-


-M'-I


t- 00
to
*e


0t
00 00
*m
0 0










tO
OOT-


& I


>0
z0


n o )

i I 6
0 C c Q
0m a

8 I


0 *
4 0*0
v-4











GIi
OBr @3 t.4










4*- a
.4 1



4J 44

I) a
844



Err.
4J M 0 B 4 4
M CO 4- 9


*1 2
45 4.
4. *
9-4 W


m t c t~
CM q-4 V-4 r-4


4-)
be to
11













IM C
0





4l)








interpretation of this result remains unclear, however, because

high variation in group sizes tended to be found in certain

genera, e.g., Maca and Papio (van Schaik, 1983; Jolly, 1985;

Caldecott, 1986). Variation in group size might be more

pronounced in these genera, or it might be purely a consequence

of the number of studies on these genera.

Great variation in group size may also be found in species

that occur in a range of different habitats (Eisenberg et al.,

1972) e.g., baboons and macaques. Some species (e.g. Propithecus

verreauxi or white sifaka) however, seem to be able to adapt to

many kinds of habitats, without any associated variation in group

size (Richard, 1974, 1978).

The variation in group size may be larger when average group

size decreases (Table 1-1) but seems to be independent of the

spacing system reported for the species (Table 1-1; van Schaik,

1983). However, this trend is not universal (see the vervet

monkey, C. aethiops; Table 1-1).



Factors Affecting Group Size in Social Animals


Most primates are social animals. Group size in social

animals may be a result of (1) selection directly on group size

(Rodman, 1981; Brown, 1982; Terborgh, 1983; Pulliam and Caraco,

1984). Alternatively, group size may be a consequence of

selection on other traits (Lewontin, 1970, 1977, 1979; Gould,

1982; Gould and Lewontin, 1979), such as (2) simple demographic

processes: birth, immigration, death and emigration (Cohen,








1969, 1975) or (3) individual ontogenetic trajectories (Wiley,

1981).

Arguments that group size is an adaptive trait (Rodman,

1981; Brown, 1982; Terborgh, 1983; Pulliam and Caraco, 1984;

Sibly, 1983; Giraldeau and Gillis, 1985) generally assume that an

individual's inclusive fitness varies as a function of the size

of the group in which that individual is found. An animal may

benefit directly from group living through an increased ability

to find or monopolize food (Eisenberg et al., 1972; Wilson, 1975;

Clutton-Brock, 1974; Leighton and Leighton, 1982; Schoener, 1971;

Wrangham, 1980) or by a reduced vulnerability to predators

(Alexander, 1974; van Schaik, 1983; Terborgh, 1983). However, by

living in groups, animals must cope with higher intragroup

feeding competition (van Schaik, 1983; Janson, 1985; Watts,

1985). Group size may be set by an optimal trade-off between

increased predator protection and decreased foraging opportunity

(Terborg, 1983; Wilson, 1975).

Other authors argue that group size is a consequence of

selection on other traits (e.g. Lewontin, 1970). Cohen (1969),

for instance, proposed a model to account for the observed

distribution of group sizes in wild primates. He proposed that

group size is the result of selection on one or all of four main

demographic processes: birth, immigration, death and emigration.

The distributions predicted by his model matched the observed

distribution of group sizes in a number of primates including

gibbons (genus Hylobates), yellow baboons (Papio cynocephalus)

and howler monkeys (Alouatta palliata). These results suggest

that natural selection may act directly on population parameters









and not on group size per se (Altmann and Altmann, 1979).

Alternatively, selection may act on individual ontogenetic

trajectories or individual schedules of social development

(Wiley, 1981), and group size is an indirect consequence of this

selection. Each individual animal changes its social status

according to its age or seniority and vacancies in the social

hierarchy. This model has successfully accounted for the

distribution of group sizes in the communally breeding stripe-

backed wren (Campylorhynchus nuchalis) (Wiley and Rabenold,

1984; Rabenold, 1985).



Group Living in Primates: Food or Predators?


Wrangham (1980) proposed that group size affects access to

patchily distributed resources. He argued that males and females

are subject to different evolutionary forces. Mate acquisition

is more important for the male, while food acquisition is more

important for the female. Females should track the distribution

of food while males should track the distribution of females. If

resources are patchily distributed in space, and if large groups

of females are able to displace small groups from resources, then

female kin should tend to stay together and form long-term

relationships. The result is the "female-bonded group".

Wrangham (1980) did not reject reduced vulnerability to

predators as a benefit of social living, but he viewed it as less

important than intergroup feeding competition. Wrangham (1980)

also accepted that there is a cost to living in a group because

of intragroup feeding competition. However, he regarded this








cost as compensated by benefits obtained from success in

intergroup feeding competition. Wrangham noted that there are

two possible responses to food scarcity within such a group. As

females in the female-bonded group are mostly kin and intragroup

feeding competition should be most intense during periods of food

scarcity, females should either emigrate to reduce the cost to

their kin, or they should switch from high to low quality food.

On the other hand, van Schaik and van Hooff (1983) and van

Schaik (1983) proposed the alternative view that predator

avoidance is the primary cause of group-living in primates. Van

Schaik proposed that living in a group always carries the cost of

high intragroup feeding competition. Large groups will therefore

have lower feeding efficiencies, and this cost must be balanced

by the benefit of reduced predator vulnerability. This

hypothesis ignores any benefit of intergroup feeding advantage.

Van Schaik's (1983) hypothesis also does not address the effect

of season on the use of time, space and resources. No matter

what the season is, he predicted that intragroup feeding

competition will be higher in the large group than in the small

one.

Van Schaik (1983) predicted that (i) larger groups should be

better at detecting predators and (ii) in forest primates, small

groups should spend more time higher off the ground to avoid

predators.








Cebus olivaceus: a Candidate for Testing
Wrangham (1980) and van Schaik (1983)'s Hypotheses


Wedge-capped or weeper capuchin monkeys (Cebus olivaceus) in

the llanos of Venezuela demonstrate characteristics that are

ideal to test the predictions derived from Wrangham (1980) and

van Schaik (1983)'s hypotheses.

(1) They live in groups that range in size from five to more

than 40 animals (Robinson, 1986; this study).

(2) They live in female-bonded groups, in which females tend

to stay in their natal group while males emigrate (Robinson, in

preparation).

(3) They inhabit a seasonal forest in which fruit trees are

clumped (Robinson, 1986). This allows groups to defend discrete

patches of food (Robinson, 1985). The availability of fruit and

certain invertebrate prey also varies seasonally with a peak in

the wet season (Robinson, 1986).

(4) Group ranges overlap completely (Robinson, 1986) and

groups must compete directly for the same resources.

(5) Last, but most important, is that this monkey population

has been well studied (Robinson, 1981, 1984a, 1986; Fragaszy,

1986; de Ruiter, in press) and basic information on ecology,

behavior and demography are already available.

















CHAPTER II
STUDY SITE AND METHODS



Study Site


The study site was Fundo Pecuario Masaguaral in the central

llanos of Venezuela, a cattle ranch owned by Sr. Tomas Blohm.
o / o /
Its location is 8 34 N and 67 35 W, about 45 Km south of

Calabozo. The nearest village is Corozo Pando. It has been an

active site of wildlife research (e.g. Montgomery and Lubin,

1977; Eisenberg, ed., 1979a; Wiley and Wiley, 1980; Rabenold,

1985; Austad and Rabenold, 1985).

My study site was in the riparian gallery forest that

covered the eastern part of the ranch. This forest is bordered

by the Cano Caracal and the Rio Guarico. The forest can be

classified broadly as dry tropical forest (Ewel et al., 1976).

The annual rainfall is about 1450 mm and is highly seasonal with

most of the rain falling between May and October (Robinson,

1986).

Three broad types of vegetation can be classified within the

forest according to soil types (Tayler, 1978; Robinson, 1986):

forests on loamy sand, loam, and clay soils, respectively.

Certain species of trees occur broadly in all forest types, e.g.,

the palm Copernicia tectorum and the guacimo Guazuma tomentosa.









Others have a highly clumped distribution, e.g., the guarataro

Vitex orinocensis.

Most tree species in this forest are patchily distributed in

space, and bear fruit in a very seasonal manner (Robinson, 1986).

More species fruit during the rainy than the dry season. The

fruit availability therefore is highly patchy in both space and

time. A detailed description of the study site appears in

Robinson (1986). The descriptions of the forests in general in

this llanos region can be seen in Monasterio and Sarmiento

(1976), Troth (1979), and Sarmiento (1984).



Study Animals


The study animals are wedge-capped capuchin monkeys, Cebus

olivaceus (=nigrivittatus following Honacki et al., 1982). This

is a New World primate belonging to the family Cebidae (Jolly,

1985). This species is distributed from the Orinoco River basin

to the northern bank of the lower Amazon and Negro Rivers (Freese

and Oppenheimer, 1981).

The study population on this ranch was studied very briefly

by Oppenheimer and Oppenheimer (1973) in 1969. Robinson started

long-term research in 1977. Researchers working with this

population include Robinson (1981, 1984a, 1986), Fragaszy (1986)

and de Ruiter (in press). This study is a part of a long-term

study on demography and social organization of this species.

About 14 groups were contacted during the study period

(Table 2-1). Group sizes and composition within the study area

changed little between Robinson's and my study period.











li II

P11,1 l~ Coo -.
*- *. O 0 *
4 0000o-4 44$ e
o T7-i
0 0 0 Ow 0 0 1 0C
3 > h~r u g (Q Q uoj M(w^''9-
3^ S CCCC'-1 c', y i


+ +


CV) C C
4. 4


++ +


W O 0 O- 0


C qr-lOC 0 0 0 ON (D '-1INO-4N
N 0mQMWtr a 0oi w m to N 0



+r +

+ +





+ U
| _ - M,- - -


4)


'-4)

4.4


(U

I4-
4)




"i









Most observations were concentrated on two groups, Main and

White. Main group was studied intensively by Robinson (1977-

1979), Fragazy (1980-1981) and de Ruiter (1981-1982) and I will

call it the LARGE group. White group has been oensused by

Robinson since 1977 and was a subject of preliminary comparative

behavioral study by de Ruiter (in press). I will call this group

the SMALL group.

Both groups were habituated to me at the beginning of my

study. Animals in both groups were individually recognized (see

Figure 2-1) and my identifications were checked for accuracy in

the field by Dr. Robinson and Mr. de Ruiter. The LARGE group

increased in size from 26 to 36 while the SMALL group decreased

in size from 9 to 5 during my study period. The changes in

composition of the groups can be seen in Table 2-2 and 2-3. The

LARGE group increased in size mostly as the result of births

while the SMALL group decreased in size as the result of deaths

of old animals.

Studies of primate ecology and behavior cope with special

kinds of problems. Few studies of primate ecology and behavior

include more than one group in comparable detail (Clutton-Brock,

1977). In this study, all inferences were based on one small and

one large group. It is hardly possible for the study of this

sort to follow the principles of experimental design suggested

for ecological studies (see Hurlbert, 1984; Hawkins, 1986).

























Figure 2-1. Example of individual monkeys in the SMALL
group. The following list is from the left to the right and from
the top to the bottom: (1) an adult male (Scar Lip, SL), (2) an
old adult female (Beret, BT), (3) an adult female (Furry Face,
FF), (4) an older juvenile male (Benno, BN), (5) an older
juvenile female (Geje, GE), (6) a young juvenile female (Irma,
IR), (7) younger juvenile female (Petra, PT) and (8) a very young
juvenile female (Frances, FR) sitting next to the large adult
male (SL).





14










wr.
I w Y-- "3




















4 0 .
ft- j~iV : A'K 8^^14













TABLE 2-2

DEMOGRAPHIC CHANGES IN THE LARGE GROUP, 5 MAY 1982 29


Date Change


9 July 82 Birth


Infanticide

Birth

Immigration

Birth

Birth

Birth

Birth

Birth

Birth

Birth


Animal


F Penny (born to Pinnocchio)

M IH (born to Hi)

IH (by Big Male)

F Puffy (born to Pointed Face)

Subadult M Rudiger

F Hanna (born to Hi)

M Bobo (born to Butte)

M Angel (born to Amelia)

M Mana (born to Mary Francis)

F Malee (born to Mo)

? BE's infant (born to Becky)

? WH's infant (born to Whity)

? CR's infant (born to Crawley)


JUNE 1984


New group
size

27

28

27

28

29

30

31

32

33

34

35

36

37


F, M and ? stand for female, male and unknown sex, respectively


July

July

May

May





June

June

May

June









TABLE 2-3

DEMOGRAPHIC CHANGES IN THE SMALL GROUP, 21 MAY 1982 16 JUNE 1984


Change


Birth

Death

Immigration

Death

Emigration

Death

Death?

Emigration


Animal


F France (born to Furry Face)

Adult F Margaret

Subadult M Charlie

Adult F Bare Shoulder

Subadult M Charlie

Adult F Beret

Younger Juvenile F Petra

Older Juvenile M Benno


New group
size

10

9

10

9

8

7

6

5


F and M stand for female and male respectively


Date


July

May

July

Aug

Sept

Nov

Feb

Feb









Methods


The observations were made from May to July 1982 and from

May 1983 to July 1984. Observation periods were paired, one with

the LARGE and the other with the SMALL group, so that results

could be compared. I minimized the interval between observation

periods as much as possible. In July 1982 I tried 5-day

continuous observations on each group, following a tradition of

primate field studies (e.g. Struhsaker, 1975). To reduce the

time between the paired observation periods with the LARGE and

the SMALL group, I reduced the observation period on each group

to 2 days in June 1983. The dates of systematic observation

periods during 1983-1984 can be seen in Figure 2-2.

Observations on other groups were made opportunistically.

Animals in each group were described by noting the unique

characteristics of individuals in each particular group. The

identity of groups was double-checked with Dr. Robinson during

his annual censusing period. The observations on other groups

also allowed me to examine relations among all groups in the

study area.






















*41

&
4 .










ff4 c 00
e*I

CQC
*6I co


4-4 )
W8
OWS.4

iet




i'"^



Q 03
*


*M 00
I 0) 0
Csl'-I

i 0 *j










D0D wO Ca
4 00 Z a c
"III
z .a ( f- I- > o z a, cr c,
, t


[


I



L


10


121


I


[

I


I

I:
El


I

L


F


I
L


L

I


0
ro
CD

_0
N
CM

O -

0

CM
-0.
N 0


I
L


















CHAPTER III
GROUP SIZE AND INTRAGROUP FEEDING COMPETITION



Introduction


Intragroup feeding competition was recognized as a

potentially important factor determining group living in primates

by both Wrangham (1980) and van Schaik (1983). Van Schaik (1983)

viewed it as a direct cost while Wrangham (1980) viewed it as a

side-effect of group living. Demonstration of higher intragroup

feeding competition in large groups will not support or reject

either hypothesis.

Two approaches will be used to test if there is intragroup

feeding competition among animals in groups. One approach is to

compare groups of different sizes. There should be higher

intragroup feeding competition in larger groups. This intragroup

feeding competition will affect the use of time, movement, use of

space and access to resources. The other approach is to compare

the same group at different seasons. There should be higher

intragroup feeding competition during the period of food

scarcity. To reduce this intragroup feeding competition, animals

should use lower-quality food with more homogeneous distribution

during the period of resource scarcity. These changes, in use of









time and space, group movements, and use of resources, should be

greater in large groups than small groups.

This chapter will demonstrate that there is higher

intragroup feeding competition in large groups. The use of time

and space in groups of different sizes and in the same groups at

different seasons will be compared.



Methods


Sample Periods


During this study, I completed 16 paired observation

periods. Three pairs were made in the summer of 1982 and the

remainder in 1983-1984. Each pair lasted for two days in each

group, except for two periods in summer 1982 in which five days

of observations were made on each group.

The wet season use-of-time data were derived from six paired

periods from June to November, 1983. The dry season data were

derived from five paired observation periods from December 1983

to April 1984.

Observations on group movement and use of space were from

the 14 paired sample periods. The sample periods for movement in

the wet and dry seasons were the same as the observations on the

use of time. For comparative reasons, the use of space was

calculated from 5 observation periods in each season. They were

from June to October 1983 in the wet season and December 1983 to

April 1984 in the dry season.









Behavioral Sampling and Use of Time


Each systematic observation day (0600-1800) was divided into

24 half-hour sample periods for the LARGE group and 48 fifteen-

minute periods for the SMALL group. Instantaneous samples of

behavior (Altmann, 1974) were made on each individual, five

seconds after the animal was first located in each sample period.

No individual was sampled more than once in a sample period.

Except for defining a "vigilance" category, I adopted the same

categories of behavior used by Robinson (1986). These included

nine broad categories: foraging, feeding, moving, resting, self

cleaning, social grooming, playing, vigilance, and other social

and non-social behavior. Foraging is defined as looking for and

feeding on invertebrate prey. Feeding is defined as actually

ingesting food and is virtually restricted to items of plant

origin. The foraging category has more subcategories than any

others (see details in Robinson, 1986). During foraging I

recorded whether an item was ingested during the five seconds

immediately following a record, and this allowed me to quantify

capture success. Vigilance was defined as the activity when

animals were alert and looking around but were not foraging or

feeding.

All social interactions were also noted opportunistically

using the ad libitum method (Altmann, 1974). In other words, I

recorded all social interactions whenever they were noted. I

paid special attention to aggressive interactions. Most

aggressive interactions were associated with vocalizations

(trills, whistles, screams) (Robinson, 1984b), making them easy to









detect. Some aggressive interactions that were not associated

with loud vocalizations (e.g., open mouth threat, see also

Oppenheimer, 1973; Klein, 1974; Weigel, 1978) might have been

missed.



Movement and Use of Space


Before I began my study, an extensive trail system was

established by Robinson, Fragaszy and de Ruiter. The trails were

marked with colored plastic flags and marked aluminum tags every

25 meters. The trail system was approximately 60 km long.

During my study period, I maintained this system and also

extended trails to areas that were used by the SMALL group,

mainly in the northern part of the study site (Figure 3-1).

Locations of the study groups were noted every half-hour

starting at 0600 h until 1800 h on all systematic observation

days. The center of mass of the group was carefully placed on

the map. The center of mass of the group is defined as the

approximated location of a center of a group.

To study group movements, two variables were measured from

the observed group paths. These were half-hour movements and

half-hour turning angles. Half-hour movements measure the

distances from two centers of mass in consecutive half-hour

periods. Half-hour turning angles measure angular deviations

from straight ahead between consecutive half-hour movements.

Both half-hour movement and turning angle have been used before

in primate studies (Waser, 1976; Robinson, 1986).





































4i










02
491






































































Sz *


0

0 <
.j a
Sg
, 0









To study the use of space of the two groups, one-hectare

quadrats were used. The quadrat locations and sizes are identical

to those of Robinson (1986). Each half-hour location of the

center of mass was considered as a single score when the use of

area of the two groups was calculated. Home range was calculated

by the total number of quadrats occupied during the half-hour

locations in a specified period of time.

To study the vertical use of space of the group, the height

above ground was noted every time a behavioral observation was

made. The heights were lumped into 5 categories: ground (=0), 1-

5, 6-10, 11-15, and 16-20 m from the ground. The lumped

categories may reflect different foraging microhabitats.

Every time a behavioral observation was made, the distance

and identity of nearest neighbors was noted whenever possible.

The distance was estimated in 1 m intervals up to 10 m.



Use of Resources


During behavioral observations, I noted the object types in

which animals were foraging. When the monkeys fed, the food was

classified as to animal or plant origin.

Plant material used by the animals was identified by various

methods. Most of the important plant species used by the animals

were shown to me by Dr. Robinson at the beginning of my study.

Some plants were identified and tagged. I also consulted the

local people or resident botanists at the ranch. Three

references which I used for plant identification were Ramia

(1974), Hoyos (1979) and Steyermark and Huber (1978). Most










plants used by the monkeys were identified to species. In

contrast, food of animal origin was rarely identified.



Data Analysis


Statistical analysis relied on SPSSx version 2.1 (SPSS,

Inc., 1986). Most analyses used non-parametric statistics

(Siegel, 1956) except where indicated. Use of parametric tests

followed Sokal and Rohlf (1981). When non-parametric tests were

used with a large set of data, a sample was randomly selected

from the appropriate data using option 4 in SPSSx (SPSS, Inc.,

1986).

To compare variation such as in vertical use of space,

coefficients of variation were compared. The logarithm to the

base 10 was also taken for each datum to estimate relative

variability (Lewontin, 1966). This allowed me to use Kruskal-

Wallis one-way ANOVA to test for significant differences (Sokal

and Braumann, 1980).



Results



Intergroup Comparison

Effect on Spacing


Prediction 1: The higher intragroup feeding competition

among animals in larger groups will affect individual feeding

efficiency so in order to reduce intragroup feeding competition,

animals in larger groups should spread themselves out more.








Nearest-neighbor-distances among animals should be greater in

larger groups.

Nearest-neighbor-distances among animals in both large and

small groups were comparable. The mean nearest-neighbor-distance

among animals in the LARGE group was 3.8 m while in the SMALL

group it was 3.9 m (Table 3-1), a difference that is not

statistically significant (Mann-Whitney U-test, n =527
L
instantaneous samples, n =378 instantaneous samples, p>0.1).
S

Prediction 2: To reduce intragroup feeding competition,

animals in large groups should spread themselves vertically more

extensively. This may be because by spreading themselves

vertically, animals in a group will use different microhabitat or

different resources. Consequently, animals avoid intragroup

feeding competition.

The vertical spread of the animals in the LARGE group was

more extensive than that in the SMALL group. This is shown both

by coefficients of variations and relative variabilities. The

coefficient of variation in height for animals in the LARGE group

was higher (69.8% in the LARGE group versus 44.1% in the SMALL

group). The relative variabilities in height for animals in both

groups was significantly different (Kruskal-Wallis one-way ANOVA,

n =1929 instantaneous samples, n =1578 instantaneous samples,
L S
p<0.01).












TABLE 3-1


THE NEAREST NEIGHBOR DISTANCE (m) AMONG ANIMALS IN BOTH GROUPS


Season Mean


Overall

Wet

Dry


Overall

Wet

Dry


3.8

3.6

3.0


3.9

3.9

3.5


Median



4.0

3.0

3.0


4.0

4.0

3.0


SD Total
records


2.31

2.44

1.79


2.35

2.26

1.89


2689

478

474


2315

692

316


Group





LARGE






SMALL









Effect on Social Interactions


Prediction 3: The higher intragroup feeding competition may

be exhibited as a higher frequency of agonistic interactions. I

expected that frequency of agonistic interaction would be greater

in large groups.

The animals in the LARGE group interacted aggressively more

frequently with one another than those in the SMALL group. The

mean rate of agonistic interactions among animals in the LARGE

group was 23.75 per day (n =8 days) while in the SMALL group it
L
was 7.75 per day (n =8 days). This frequency of agonistic
S
interactions was significantly different (t-test, 14df, p<0.01).

Note that because nearest-neighbor-distances did not differ

significantly between the groups, this is not a consequence of

more potential interactants being closer together in the LARGE

group.



Effect on Time Budgets


Prediction 4: The higher intragroup feeding competition in

larger groups should result in lower invertebrate foraging

success. This is because the number of animals in large groups

should reduce the invertebrate resource base and thereby reduce

capture success. It is predicted that the animals in large

groups should have lower invertebrate foraging success.

Overall, animals in both groups had equal capture success

(13.6%, n =3449 instantaneous samples versus 12.5%, n =2547
L S
instantaneous samples; chi-square test, ldf, p>0.1). This effect









was consistent across all microhabitats (Table 3-2; chi-square

test, ldf, p<0.01 for all categories).


Prediction 5: The higher intragroup feeding competition

should result in lower access of animals in a group to fruit

resources. This may be the result of higher agonistic

interactions in large groups or animals may avoid each other. I

expected that the animals in large groups would spend less time

feeding.

Table 3-3 shows how both groups used their time. The LARGE

group spent less time feeding as predicted (chi-square test, ldf,

p<0.01).


Prediction 6: If higher intragroup feeding competition

results in less access to resources, then animals in larger

groups should allocate more time to foraging (sensu latu) than

animals in smaller groups. I expected that the animals in large

groups would spend more time feeding, foraging and moving

combined together.

Contrary to prediction, the animals in the LARGE group spent

less time feeding, foraging and moving combined (73.3% versus

75.9%, Table 3-3) than animals in the SMALL group (chi-square

test, ldf, p<0.01).

This result may be a consequence of animals in the two

groups using different resources, or of animals in the LARGE group

having higher priority of access to clumped and restricted

resources (Chapter V).













DIFFERENTIAL


TABLE 3-2

CAPTURE SUCCESS (%) IN DIFFERENT MIC"HJABITATS


Category


Dry twig

Dry branches

Leaves

Insects

Bark

Palm leaves

Palm nuts

Twigs

Dry palm leaves

Dry leaves

Palm boots

Leaf litter

Palm tops

Dry boughs

Palm branches

Branches


Capture
LARGE
group

33.3

34.6

50.8

67.3

20.0

37.1

21.5

9.7

23.7

52.4

16.9

2.4

20.0

13.3

51.2

87.5


success (%)
SMALL
group

36.2

31.5

36.8

62.9

21.0

21.3

20.0

19.5

16.3

22.2

34.6

0

13.6

25.0

72.2

80.0


Total sample
LARGE SMALL
group group

189 221

81 111

59 117

52 35

40 62

89 47

330 50

31 41

59 43

21 36

71 26

206 32

25 22

23 24

43 18

24 20











TABLE 3-3

TIME BUDGET (%) OF ANIMALS IN THE


LARGE group

Forage 28.6

Feed 14.1

Rest 10.1

Move 30.6

Drink 0.9

Socialize 13.7

Non-socialize 1.0

Vigilance 0.9


Total
instantaneous
samples


13,630


LARGE AND SMALL GROUP

Time
SMALL group

29.7

19.6

12.8

26.6

0.4

8.2

1.2

1.5


9,419









Prediction 7: The higher intragroup feeding competition

should allow less time for animals in larger groups to rest.

This may be the result of animals in larger groups spending more

time interacting agonistically or avoiding one another. I

expected that animals in larger groups would spend less time

resting.

The animals in the LARGE group spent less time resting

(10.1% versus 12.8%, Table 3-3) than animals in the SMALL group

(chi-square test, 1df, p<0.01).


Prediction 8: If animals in larger groups spend more time

avoiding one another, then they should spend more time moving. I

expected that animals in larger groups would allocate more time

to moving.

The animals in the LARGE group spent more time moving (30.6%

versus 26.6%, Table 3-3) than animals in the SMALL group (chi-

square test, ldf, p<0.01).



Effect on Group Movement and Use of Space


Prediction 9: Avoiding intragroup feeding competition by

moving should also result in more extensive group movements. I

expected that larger groups would move farther than smaller

groups.

The length of the day-range in the LARGE group was longer.

The mean length of day-range in the LARGE group was 2,400 m

(SD=506 m, n =29 days) while that of the SMALL was 2,000 m
L









(SD=468 m, n =24 days), a difference that was statistically
S
significant (t-test, t=3.07, p<0.01).

The mean half-hour step length of the LARGE group was also

longer. The mean half-hour step length in the LARGE group was

99.8 m while that of the SMALL group was 81.7 m (Table 3-4), a

difference that was statistically significant (Mann-Whitney U-

test, n =734 records, n =725 records, p<0.01, Figure 3-2).
L S

Prediction 10: If animals in groups of different sizes used

the same resource base, then animals in larger groups should

exploit their resource faster. The disadvantages of turning

around and returning to previously occupied areas or previously

exploited fruit trees should be greater in larger groups. As a

result, larger groups should tend to move forwards more and

backtrack less than smaller groups. I predicted that half-hour

turning angles in larger groups would be smaller on average than

turning angles in smaller groups.

The mean half-hour turning angle of the LARGE group was

smaller than that of the SMALL group. In the LARGE group, it was
o o
50.9 while in the SMALL group it was 64.9 (Table 3-5), a

difference that was statistically significant (Mann-Whitney U-

test, z=-5.06, n =680 records, n =622 records, p<0.01). In
L S
addition, Figure 3-3 indicates that the overall distribution of

turning angles of the LARGE group is more skewed towards smaller

turning angles.

Another interpretation of this result is that the larger

half-hour turning angles of smaller groups is a consequence of











TABLE 3-4

THE HALF-HOUR STEP LENGTH (m) IN BOTH GROUPS


Group Season Mean Median SD Total
records


Overall 99.8 84.0 67.91 734

LARGE Wet 97.4 85.0 60.84 284

Dry 90.2 73.0 68.33 234


Overall 81.7 65.0 69.21 726

SMALL Wet 85.8 65.0 72.29 284

Dry 71.7 58.0 71.74 227






























C:- 0o
PER
4 -i


4J C -- 4)





0 0
*sF











1.'8




r-I
G)4
.-4
5 t





4-;
0 &0
+3 *^ -p

*M 0
Co '5
8- O


34 0) )*

S.C JS3





38







E
00




0-
O-


Z
to 0
o M

z



0
r
0


-0-
o I
O








0-









I --w^ U)^ --------
00 00











Si 0 0 0









SNOISVOO 00 0UsiNnN













TABLE 3-5


THE HALF-HOUR TURNING

Group Season


LARGE






SMALL


Overall

Wet

Dry


Overall

Wet

Dry


ANGLE

Mean



50.9

48.5

53.4


64.9

68.3

60.2


(DEGREES) IN BOTH GROUPS


Median



39.0

37.0

41.0


53.0

58.0

46.0


SD Total
records


44.32

42.82

45.89


50.17

51.20

49.85


680

269

214


622

242

199



























4.4














0







fi


4 r4
4-)
C4




















































I | I I I I i
0 0 0 0 0 0 0 0
Ia- (D tn o o N


CO
0 a.




o- o
0 5

0o 0..

~0 0




o0





II-

o-
0J0










o --
0 0 0 0 0





N C M q* .


S3ldVIVS AVO-TIV ENIUnG
SNOISVOOO _40 ui:AfN









intergroup encounters or animals in smaller groups avoid

intergroup encounters (Chapter V).


Prediction 11: The greater number of animals in larger

groups should result in higher total food requirements. If

larger areas contain larger amounts of food, then larger groups

should use larger areas than smaller groups. I expected that

larger groups would have larger group ranges than smaller groups.

At the end of the study the group range of the LARGE group

had entered 209 one-hectare quadrats while the SMALL group had

entered 174. At comparable lengths of sampling time, the range

of the LARGE group was always larger (Figure 3-4).



Seasonal Comparison

This set of predictions seeks to demonstrate the importance

of intragroup feeding competition by comparing the use of time

and space by animals during periods of resource abundance and

resource scarcity.


Prediction 12: To reduce intragroup feeding competition,

animals in both groups should emigrate more frequently during the

periods of food scarcity (Wrangham, 1980). Robinson (1986)

demonstrated that the dry season is the period of resource

scarcity. Emigrants should be adult females.

However, no instances of adult female emigration from either

group (Table 2-2, 2-3) were recorded.















*h



-is

U.1
U0
A








I
in4 0
*g m

S*^e
II .-
|.)




44

0
.0


Q









II

000
LU






z
\0%
_


0. CL
0 0

w .
09 0


0 0 0 0 0 0 0 0 0 0
0 0 COD 0( N
N(: (S3IV103H)

3ZIS 39)NV: 3AI.liV. nlqno









Effect on Use of Resource


Prediction 13: Alternatively, animals in social groups

should shift onto "low quality" food during periods of resource

scarcity (Wrangham, 1980). Wrangham argued that low quality food

is generally homogeneously distributed in the environment and

therefore not competed over. I expected that animals in both

groups would spend more time foraging for animal material during

the dry season. This is because animal material such as insects

is more homogeneously distributed than plant material such as

fruit (Robinson, 1986).

Both groups spent significantly more time feeding on animal

material during the dry than during the wet seasons (Table 3-6;

chi-square test, ldf, p<0.01 for both groups).



Effect on Spacing


Prediction 14: Animals may respond spatially to intragroup

feeding competition by spreading out through the forest. I

expected that nearest-neighbor-distances among animals in social

groups would be higher during the period of food scarcity.

Contrary to prediction, the mean nearest-neighbor-distance

among animals in both groups was less during the dry season. The

mean nearest-neighbor-distances among animals in the LARGE group

during the wet and dry seasons were 3.6 m and 3.0 m respectively

(Table 3-1). The seasonal difference was statistically

significant (Mann-Whitney U-test, n =87 instantaneous samples,
W
n =92 instantaneous samples, p<0.1). Mean nearest-neighbor-
D





























co0 CV) C4

jb t 0 C

CM m




Q t o r






'Y, sa
aa* 0m









C,;
lw




m
v\~ N co a.






S00 C*4 0






*1-
m CQ CD









*P4 $4C


r o O
r* ~a & i S *








distances among animals in the SMALL group during the wet and dry

seasons were 3.9 m and 3.5 m respectively. The seasonal

difference was not statistically different (Mann-Whitney U-test,

n =134 instantaneous samples, n =56 instantaneous samples,
W D
p>0.1).

This failure to verify the prediction may be the result of

animals switching their food during the period of food scarcity

(Prediction 13).


Prediction 15: Animals may spread themselves out more

vertically during the period of food scarcity. I expected that

variation in vertical use of space would be higher among animals

during the dry season.

The coefficients of variation in height were higher in the

dry season than the wet season for both groups. The coefficient

of variation in height for the LARGE group was 85.8% in the dry

season and 59.9% in the wet season. The coefficient of variation

in height for the SMALL group was 49.7% in the dry season and

40.7% in the wet season. The relative variabilities in height

both groups spent in both seasons were significantly different

(Kruskal-Wallis one-way ANOVA, n =629 instantaneous samples,
W
n =459 instantaneous samples, p<0.01 in the LARGE group and
D
n =579 instantaneous samples, n =304 instantaneous samples,
W D
p<0.01 in the SMALL group).

This same result may be a consequence of animals switching

their food during the dry season (Prediction 13). Low quality

food, by definition, is more dispersed in the environment.









Effect on Time Budget


Prediction 16: If intragroup feeding competition affects

the types of food eaten by animals (Prediction 13), animals in

both groups should spend less time actually feeding but more time

foraging (sensu latu) during the dry season.

Table 3-7 shows how animals in both groups spent their time

in both seasons. Animals in both groups spent more time foraging

(sensu latu) during the dry season. The animals in the LARGE

group spent 22.9% of their time foraging (sensu latu) in the wet

season and 44.3% in the dry season (chi-square test, ldf,

p<0.01). The animals in the SMALL group spent 23.1% of their

time foraging (sensu latu) in the wet season and 51.7% in the dry

season (chi-square test, ldf, p<0.01).

The LARGE group spent less time feeding in the dry season

but the SMALL group spent more time. The animals in the LARGE

group spent 15.0% of their time feeding in the wet season and

12.4% in the dry season (chi-square test, ldf, p<0.1), while

animals in the SMALL group spent 20.6% of their time feeding in

the wet season and 24.3% in the dry season (chi-square test, ldf,

p<0.1).


Prediction 17: The higher intragroup feeding competition

should allow less time for animals in both groups to rest during

the dry season.

The animals in both groups spent less time resting in the

dry season (Table 3-7). The animals in the LARGE group spent

10.6% and 8.7% of their time resting in the wet and dry season














TABLE 3-7

THE TIME BUDGET OF BOTH GROUPS DURING

% Time
Activity Large group
Wet Dry
season season

Forage 22.9 44.3

Feed 15.0 12.4

Rest 10.6 8.7

Move 31.9 25.2

Drink 0.7 0.6

Socialize 18.1 7.3

Non-socialize 0.2 1.5

Vigilance 0.6 0.0


THE WET AND DRY SEASONS

% Time
Small group
Wet Dry
season season

23.1 51.7

20.6 24.3

14.5 6.6

31.7 12.1

0.3 0.4

6.7 2.6

0.6 2.1

2.5 0.2


Total
instantaneous
samples


3,909 4,070


3,352 2,039









respectively (chi-square test, ldf, p<0.01). The animals in the

SMALL group spent 14.5% and 6.6% of their time resting in the wet

and dry seasons respectively (chi-square test, ldf, p<0.01).


Prediction 18: To reduce intragroup feeding competition

during the dry season, animals should spend more time moving.

Contrary to prediction, animals in both groups spent less

time moving in the dry season. The animals in the LARGE group

spent 31.9% and 25.2% of their time moving in the wet and dry

season respectively (chi-square test, ldf, p<0.01), while the

animals in the SMALL group spent 31.7% and 12.1% of their time

moving in the wet and dry season respectively (chi-square test,

ldf, p<0.01).



Effect on Group Movement and Use of Space


Prediction 19: If a shift to food with more homogeneous

distribution is an important mechanism to reduce intragroup

feeding competition, groups should move shorter daily distances

during the dry season. This follows from Robinson's (1986)

demonstration that daily travel distances are lower when animals

are foraging on homogeneously distributed resources.

The average half-hour step length was lower in the dry

season than in the wet, for the LARGE group but not for the SMALL

group. The mean half-hour step length for the LARGE group was

97.4 m and 90.2 m in the wet and dry seasons respectively (Table

3-4), a difference that was statistically significant (Mann-

Whitney U-test, n =284 records, n =234 records, p<0.01). The
L S









mean half-hour step length in the SMALL group was 85.8 m and 71.7

m in the wet and dry seasons respectively. These means are not

significantly different (Mann-Whitney U-test, n =284 records,
W
n =227 records, p>0.05).
D
This suggests that the shift of food is an important

mechanism of reducing intragroup feeding competition especially

during the dry season.


Prediction 20: The shift to homogeneously distributed food

should also affect half-hour turning angles between seasons.

Robinson (1986) reported a weak relation across months between

average turning angle and the proportion of homogeneously-

distributed food in the diet.

The half-hour turning angles in different seasons were not

significantly different in either group. The mean turning angles
o
in the LARGE group during the wet and dry season were 48.5 and
o
53.4 respectively (Table 3-5; Figure 3-5). Those in the SMALL
o o
group were 68.3 and 60.2 in the wet and dry seasons

respectively. Seasonal comparisons of turning angles were not

significantly different in either group (Mann-Whitney U-test,

n =269 records, n =214 records, P>0.1 in the LARGE group; n =242
W D W
records, n =199 records, p>0.1 in the SMALL group).
D

Prediction 21: Animals should occupy a smaller area in the

dry season, because they spend more time feeding on food

homogeneously distributed in space.

The SMALL group used a smaller area during the dry season

but the LARGE group used a slightly larger area. The areas















1
I.


I!






o4l














0 00
r- U- w c"



0 z
"
OO OO <


I z
LI-
I L-


.. J -m
O








LJ m :DNu
000









o<(, W o 3d- V- OD N (0

SOO0O38 %









covered by the SMALL group were 87 and 47 ha during a comparable

period in the wet and dry seasons respectively. Home range areas

in the LARGE group were 98 and 107 ha in the wet and dry seasons

respectively.



Discussion


There have been a number of studies showing the importance

of intragroup feeding competition on social living. With the

increase in group size, animals spend less time feeding (van

Schaik et al., 1983a) and more time moving (Green, 1978; van

Schaik et al., 1983a). Animals in large groups also stayed away

from each other to avoid intragroup feeding competition (Green,

1978). Intragroup feeding competition may also result in longer

day-range lengths (Waser, 1977b; Green, 1978; van Schaik et al.,

1983a, Dunbar, 1984; Sharman and Dunbar, 1982; de Ruiter, in

press), and larger group ranges in larger groups (Iwamoto and

Dunbar, 1983; Takasaki, 1981; Stacey, 1986; Suzuki, 1979;

Davidge, 1978; Makwana, 1978; van Schaik and van Hooff, 1983).

Animals in a group may also switch their food in order to avoid

intragroup feeding competition during periods of food scarcity

(Oppenheimer, 1968; Hladik, 1975; Dittus, 1977; Robinson, 1986;

de Ruiter, in press).

This study suggests that intragroup feeding competition is

important in structuring vertical use of space by a group of

monkeys. However, variation in vertical use of space may also

result from variation in resource abundance at different heights

in different seasons. Nevertheless there is still more









agonistic interaction in larger groups and this might structure

vertical use of space.

Intragroup feeding competition did not affect invertebrate

capture success but did affect fruit feeding.

This study suggests that moving is a behavioral mechanism to

reduce intragroup feeding competition. Larger groups move

farther and have a greater tendency to move forwards than smaller

groups. Large groups also use larger areas.

The effect of group size on the types of food eaten require

special attention. The types of food eaten by large groups may

be the result of intragroup feeding competition or the priority

of access to resources (see Chapter V).

The results in this chapter cannot be used to support or

reject Wrangham (1980) or van Schaik (1983)'s hypotheses. Both

hypotheses accept the effect of intragroup feeding competition on

social living. The differential use of resources by animals in

different groups suggests the importance of intergroup feeding

competition on social living (Chapter V).














CHAPTER IV
GROUP SIZE AND VULNERABILITY TO PREDATORS



Introduction


Both Wrangham (1980) and van Schaik (1983) recognized a

reduction of vulnerability to predators as an important factor

for group living in primates. Van Schaik (1983) considered

vulnerability to predators as a primary factor while Wrangham

(1980) considered it to be of secondary importance. To prove that

living in groups can reduce vulnerability to predators of animals

in a group will not prove or disprove either Wrangham's or van

Schaik's hypothesis.

The presence of other individuals in a group 1) can reduce

the probability that a given animal will be detected by a

predator (Hamilton, 1971; Treisman, 1975), 2) increase the

probability that predators will be detected by some animals in

the group (Pulliam, 1973; Treisman, 1975; Kenward, 1978), 3)

increase the probability that the animals will deter the attack

of predators (DeVore, 1965; Eisenberg et al., 1972) and that even

if the group is attacked, an individual can avoid being a victim

(Hamilton, 1971; Jarman, 1974).

In forest primates, animals in larger groups may detect

predators from a greater distance or faster than animals in

smaller groups (van Schaik et al., 1983a, 1983b). In an area










with only terrestrial predators, animals in small groups may

spend more time off the ground (van Schaik et al., 1983b; de

Ruiter, in press).

This chapter will show indirect evidence that by living in

groups, animals reduce their vulnerability to predators. Animals

in large groups spend less time vigilant and more time on the

ground. This evidence, however, is not a conclusive

demonstration that by living in groups animals reduce their

vulnerability to predators.




Methods



Sample Periods


During this study, I completed 16 paired observation

periods. Three pairs were made in the summer of 1982 and the

remainder in 1983-1984. Each pair lasted for two days in each

group, except for two periods in summer 1982 in which five days

of observations were made on each group.

The wet season data were derived from six paired periods

from June to November, 1983. The dry season data were derived

from five paired observation periods from December 1983 to April

1984.



Behavioral Sampling and Use of Time


Each systematic observation day (0600-1800) was divided into

24 half-hour sample periods for the LARGE group and 48 fifteen-









minute periods for the SMALL group. Instantaneous samples of

behavior (Altmann, 1974) were made on each individual, five

seconds after the animal was first located in each sample period.

No individual was sampled more than once in a sample period.

Except for defining a "vigilance" category, I adopted the same

categories of behavior used by Robinson (1986). This included

nine broad categories: foraging, feeding, moving, resting, self

cleaning, social grooming, playing, vigilance, other social and

non-social behavior. Foraging is defined as looking for and

feeding on invertebrate prey. Feeding is defined as actually

ingesting food and is virtually restricted to items of plant

origin. The foraging category has more subcategories than any

others (see details in Robinson, 1986). During foraging I

recorded whether an item was ingested during the five seconds

immediately following a record, and this allowed me to quantify

capture success. Vigilance was defined as the activity when

animals were alert and looking around but were not foraging or

feeding.



Use of Vertical Space


The height above ground was noted every time a behavioral

observation was made. The heights were lumped into 5 categories:

ground (=0), 1-5, 6-10, 11-15, and 16-20 m from the ground.









Data Analysis


Statistical analysis relied on SPSSx version 2.1 (SPSS,

Inc., 1986). Most analyses used non-parametric statistics

(Siegel, 1956) except where indicated. Use of parametric tests

followed Sokal and Rohlf (1981). When non-parametric tests were

used with a large set of data, a sample was randomly selected

from the appropriate data using option 4 in SPSSx (SPSS, Inc.,

1986).



Results


Cebus olivaceus monkeys were seen to mob and to give alarm

calls in the presence of aerial and terrestrial predators.

Potential terrestrial predators included jaguars (Panthera onca),

ocelots (Felis pardalis), tayras (Eira barbara), and boas (Boa

constrictor). Monkeys also gave alarm calls to caimans (Caiman

crocodilus) and collared peccaries (Tayassu tjacu). Monkeys

gave alarm calls to the following birds: hook-billed kites

(Chondrohierax uncinatus), black vultures (Coragyps atratus),

green ibises (Mesembrinibis cayennensis) and rufous-vented

chachalacas (Ortalis ruficauda).


Prediction 1: If the animals in large groups have a higher

probability of detecting terrestrial predators, then the animals

in large groups should spend less time in vigilance. By living

in groups, the probability of some animal detecting a predator is

higher (Bertram, 1980), and therefore individuals can be less

vigilant.









Animals in large groups spend less time in vigilance. Table

3-3 shows the use of time by animals in both groups. Animals in

the LARGE group spent less time in vigilance (0.9% versus 1.5%)

(chi-square test, ldf, p<0.01).


Prediction 2: The animals in large groups should spend more

time on the ground. The high probability of detecting

terrestrial predators by animals in large groups should allow

them to exploit a foraging microhabitat where they might be more

vulnerable to predators.

The animals in the LARGE group spent more time on the ground

than the SMALL group (19.2%, n =13630 instantaneous samples
L
versus 3.2%, n =9419 instantaneous samples; chi-square test, ldf,
S
p<0.01). The average heights off ground of animals in the LARGE

and SMALL groups were 7.3 and 9.9 m respectively (Table 4-1).

The animals in the LARGE group also spent more time on the ground

than the SMALL group throughout the day (Figure 4-1).

An alternative hypothesis that makes the same prediction is

that the vertical distribution of space is related, not to the

predator vulnerability, but to the use of resources. Animals in

both groups spent more time on the ground when they were foraging

for invertebrates (Figure 4-2). Animals in large groups foraged

more for invertebrates, perhaps because of increased intragroup

feeding competition (see chapter III).


Prediction 3: The animals in both groups should spend less

time in vigilance during the dry season than the wet season.

Because trees are deciduous during the dry season, the visibility













TABLE 4-1

THE HEIGHT ABOVE GROUND (m) SPENT BY BOTH GROUPS


Group Season Mean Median SD Total
records


Overall 7.28 5.0 5.078 13630

Large Wet 9.24 10.0 5.530 3909

Dry 4.23 5.0 3.631 4070


Overall 9.89 10.0 4.362 9419

Small Wet 10.40 10.0 4.254 3352

Dry 7.32 5.0 3.635 2039






















Figure 4-1. Change through the day in the median height off
ground and % of records on ground in both groups.















0 o o


LARGE GROUP


000


00000


-00


0000


SMALL GROUP

0 00 000000


000000


30
0

20D
z
0

10 r

o

a


- 10


0


8 10 12 14 16 18

HOUR OF DAY


I


0


.................. ........




























Figure 4-2. Histogram showing % of records in relations to
height off ground for overall, during foraging for insects and
feeding on fruit.













OVERALL


FORAGING


FEEDING


FOR INSECTS


ON FRUIT

M-I


LII"IIII


LARGE GROUP


SMALL


GROUP


50 40 30 20 10


0 10 20 30 40 50


% RECORDS


10
5
0
ground


I


0
ground


20-
15-
10 -


0
ground


l


i r i l l i f i f f


20
15^


1 |


I i I i I A .


I










during this season is higher, and animals need not be as

vigilant.

Animals in both groups spend more time in vigilance during

the wet season than during the dry season. The animals in the LARGE

group spent 0.6% and 0% of their time in vigilance during the wet

and dry season respectively, a significant difference (chi-square

test, ldf, p<0.1). The animals in the SMALL group also spent

more time in vigilance during the wet season (2.5%) than during

the dry season (0.2%) (chi-square test, ldf, p<0.01).


Prediction 4: The animals in both groups should spend more

time off the ground in the wet season. I reasoned that it is

harder to detect predators during the wet season (Prediction 3)

and animals in both groups can reduce vulnerability to predators

by spending more time off the ground.

Animals in both groups spent less time on the ground during

the wet season. The animals in the LARGE group spent 12.5% and

33.2% on the ground in the wet and dry season respectively, a

significant difference (chi-square test, Idf, p<0.01). The

animals in the SMALL group spent 1.8% and 8.5% on the ground in

the wet and dry seasons respectively, a seasonal difference which

was significant (chi-square test, Idf, p<0.01). The animals in

both groups spent more time higher in the trees in the wet season

than the dry season (Table 4-1).

One alternative explanation is that resources were more common

on the ground during the dry season. Another explanation is that

the ground was flooded during the wet season, and therefore was

less available as a foraging microhabitat.










Discussion


A number of studies have shown indirect benefits from social

living by reducing vulnerability to predators. By living in

social groups, animals may detect predators at a greater distance

(van Schaik et al., 1983b), spend less time in vigilance (van

Schaik et al., 1983b; de Ruiter, in press), and spend more time

on the ground (van Schaik et al., 1983a; de Ruiter, in press).

There is no study of primates showing the direct effect of group

size on vulnerability to predators. A study on birds, however,

showed no effect of group living on vulnerability to predators

(Page and Whitacre, 1975, p. 82, reanalysing by using group size

of one and more than one: chi-square test, ldf, p>.0.1).

In this study area, the only potential aerial predator that

has been reported to take Cebus olivaceus is the Harpy Eagle,

i harpyja (Rettig, 1978), but this species has not been

observed at the study site. The smaller Ornate Hawk-Eagle,

Spizaetus ornatus, however, is present. Terrestrial predators

include jaguars (Panthera onca) and ocelots (Felis pardalis) but

most are primarily active during the night (Schaller and

Crawshaw, 1980; Ludlow, 1986). Boas (Boa constrictor) may be

able to take the monkeys (Chapman, 1986).

This study showed that animals in large groups spent less

time in vigilance. But this does not mean that animals in large

groups are less vulnerable to predators. It is debatable that

the degree of arboreality is solely a product of vulnerability to

predators. It may be the result of intragroup feeding

competition.





68



The results reported in this chapter cannot be used for

supporting or rejecting either hypothesis. This study does not

show any conclusive correlation between group size and

vulnerability to predators.















CHAPTER V
GROUP SIZE AND INTERGROUP FEEDING COMPETITION



Introduction


Intergroup feeding competition has been recognized by

Wrangham (1980) as a main selective pressure for living in a

social group. Larger groups should be able to displace smaller

groups. Van Schaik (1983), on the other hand, downplayed the

importance of intergroup feeding competition in selecting for

social living. A demonstration that animals in large groups have

feeding advantages because of the size of their group will

support Wrangham's hypothesis.

Good examples for intergroup feeding competition in animals

can be seen in animals that defend space. In primates, these

species usually emit loud vocalizations that maintain spacing.

Examples include gibbons (Hylobates lar, Carpenter, 1940), titi

monkeys (Callicebus moloch, Mason, 1968), howler monkeys

(Alouatta palliata, Carpenter, 1934), Hanuman langur (Presbytis

entellus, Jay, 1965), white-cheeked mangabey (Cercocebus

albigena, Waser, 1975, 1977a), and black and white colobus

monkeys (Colobus guereza, Marler, 1969). Of the other species

that do not defend exclusive areas, some demonstrate an

intergroup hierachy and a group rank. Group rank is determined









by group size in rhesus monkeys (Macaca mulatta, Southwick

et al., 1965), but not in C. albigena (Waser, 1977a).

This chapter will show that there is intergroup feeding

competition in capuchin monkeys, Cebus olivaceus, and that

animals in large groups have a greater access to resources.

Large groups dominate smaller groups and intergroup encounters

affect animals' use of time, group movement, use of space and

access to resources.



Methods



Group Sizes and Censusing


The identity and sizes of groups were obtained from regular

censuses. Whenever I encountered other groups, I followed them

for as long as possible. Some of the groups were totally

unhabituated while others had been censused regularly since 1977.

I noted all distinctive individual characteristics of group

members and gave names to any newly recognized individuals. A

final group count was arrived at only after repeated censuses. I

noted the locations of followed groups every half-hour.

Establishing group identities was essential to the study of

intergroup relations.



Intergroup Relations


Intergroup encounters were noted both during the systematic

observation periods and opportunistically when I was following










other groups. An intergroup encounter is defined behaviorally

as one in which animals from at least two groups were in visual

contact. The time and location of each intergroup encounter was

noted. Such opportunistic sampling is unbiased with respect to

the outcome and identity of the group interaction. However, data

on the relative frequency of interactions involving the LARGE and

SMALL groups depended only on counts taken during systematic

observation periods.



Sample Periods


During this study, I completed 16 paired observation

periods. Three pairs were made in the summer of 1982 and the

remainder in 1983-1984. Each pair lasted for two days in each

group, except for two periods in summer 1982 in which five days

of observations were made on each group. Observations on use of

space were from the 14 paired sample periods.

To quantify the effect of intergroup encounters on use of

time, I compared time budgets on days with intergroup encounters

and days without encounters in both wet and dry seasons. Wet

season months were from June to November, 1983, while dry season

months were from December 1983 to April 1984. Three and two

observation periods were obtained for the LARGE group during the

wet and dry seasons respectively, while two and three periods were

obtained for the SMALL group during the wet and dry seasons

respectively.









Behavioral Sampling and Use of Time


Each systematic observation day (0600-1800) was divided into

24 half-hour sample periods for the LARGE group and 48 fifteen-

minute periods for the SMALL group. Instantaneous samples of

behavior (Altmann, 1974) were made on each individual, five

seconds after the animal was first located in each sample period.

No individual was sampled more than once in a sample period.

Except for defining a "vigilance" category, I adopted the same

categories of behavior used by Robinson (1986). This included

nine broad categories: foraging, feeding, moving, resting, self

cleaning, social grooming, playing, vigilance, and other social

and non-social behavior. Foraging is defined as looking for and

feeding on invertebrate prey. Feeding is defined as actually

ingesting food and is virtually restricted to items of plant

origin. The foraging category has more subcategories than any

others (see details in Robinson, 1986). During foraging I

recorded whether an item was ingested during the five seconds

immediately following a record, and this allowed me to quantify

capture success. Vigilance was defined as the activity when

animals were alert and looking around but were not foraging or

feeding.

All social interactions were also noted opportunistically

using the ad libitum method (Altmann, 1974). In other words, I

recorded all social interactions whenever they were noted.









Movement and Use of Space


Before I began my study, an extensive trail system was

established by Robinson, Fragaszy and de Ruiter. The trails were

marked with colored plastic flags and marked aluminum tags every

25 meters. The trail system was approximately 60 km long.

During my study period, I maintained this system and also

extended trails into areas that were used by the SMALL group,

mainly in the northern part of the study site (Figure 3-1).

Locations of the study groups were noted every half-hour

starting at 0600 h until 1800 h on all systematic observation

days. The center of mass of the group was carefully placed on

the map. The center of mass is defined as the approximated

location of a center of a group. When intergroup encounters

occurred I noted down time and location of encounters.

To study group movements, two variables were measured from

the observed group paths. These were half-hour movements and

half-hour turning angles. Half-hour movements measure the

distances from two centers of mass in consecutive half-hour

periods. Half-hour turning angles measure angular deviations

from straight ahead between consecutive half-hour movements.

Both half-hour movement and turning angle have been used before

in primate studies (Waser, 1976; Robinson, 1986).

To study the use of space of the two groups, one-hectare

quadrats were used. The quadrat locations and sizes are

identical to those of Robinson (1986). Each half-hour location

of the center of mass was considered as a single score when the

use of area of the two groups was calculated. Home range was









calculated by the total number of quadrats occupied during the

half-hour locations in a specified period of time.

As the time and location of intergroup encounters were also

noted, I was able to analyse the effects of intergroup encounters

on group movement. The locations of intergroup encounters were

also assigned in the same quadrat system as above.

To correlate the use of space with the distribution of food,

data on the distribution of fruit trees from the study by Dr.

Robinson and new data obtained at the end of my study were used.

The numbers of trees of different species were counted along the

existing North-South trails (Robinson, 1986). The transect lines

were then divided into sample quadrats. Each sample quadrat was

100 m in length and 2.5 m in width on either side of the trail

(Robinson, 1986). This 500 sqm quadrat running North-South

through the center of each 1 ha quadrat was taken as

representative of that quadrat. We successfully covered 125 of

209 quadrats or 59.8 % in the LARGE group's range and 88 of 174

quadrats or 50.6 % in the SMALL group's range.

To examine more closely the group overlap in range, I spent

as much time as possible following all the possible groups in the

study area during May 23 to June 29, 1984. This overlapped with

the annual census period of Dr. Robinson, which allowed me to

match the identity of the groups with Robinson (1986). This

period was at the end of my study period so that most groups were

habituated. Information on the location of other groups was also

obtained from Dr. Robinson. When I followed other groups, I

noted the location every half-hour.









To study the vertical use of space of the group, the height

above ground was noted every time a behavioral observation was

made. The heights were lumped into 5 categories: ground (=0), 1-

5, 6-10, 11-15, and 16-20 m from the ground.

Every time a behavioral observation was made, the distance

and identity of nearest neighbors was noted whenever possible.

The distance was estimated in 1 m intervals up to 10 m.



Use of Resources


During behavioral observations, I noted the object types in

which animals were foraging. When the monkeys fed, the food was

classified as to animal or plant origin.

Plant material used by the animals was identified by various

methods. Most of the important plant species used by the animals

were shown to me by Dr. Robinson at the beginning of my study.

Some plants were identified and tagged. I also consulted the

local people or resident botanists at the ranch. Three

references which I used for plant identification were Ramia

(1974), Hoyos (1979) and Steyermark and Huber (1978). Most

plants used by the monkeys were identified to species. In

contrast, food of animal origin was rarely identified.

Morisita's index was used as a measure of the consistency or

patchiness in space of a particular tree species in the forest

(Robinson, 1986; Poole, 1974). Morisita's index for each species

of trees was obtained from Robinson (1986).









Sucrose density of fruits was measured with a Bausch and

Lomb sucrose refractometer (see also Janson, 1985; White and

Stiles, 1985).



Data Analysis


Statistical analysis relied on SPSSx version 2.1 (SPSS,

Inc., 1986). Most analyses used non-parametric statistics

(Siegel, 1956) except where indicated. Use of parametric tests

followed Sokal and Rohlf (1981). When non-parametric tests were

used with a large set of data, a sample was randomly selected

from the appropriate data using option 4 in SPSSx (SPSS, Inc.,

1986).

To compare variation such as in half-hour movements,

coefficients of variation were compared. The logarithm to the

base 10 was also taken for each datum to estimate relative

variability (Lewontin, 1966). This allowed me to use Kruskal-

Wallis one-way ANOVA to test the significant differences (Sokal

and Braumann, 1980).




Results



Intergroup Relations


Intergroup encounters occurred on 51.6 % and 67.7 % of the

systematic observation days in the LARGE and SMALL groups

respectively (n =n =31 days). The behavior of group members
L S
during intergroup encounters has been described by Robinson










(1986). Intergroup encounters occurred during all times of the

day but mostly frequently in the morning (Figure 5-1). The

probability of an intergroup encounter was higher during the wet

season in both groups ((LARGE group: 75% of days in the wet season

(n =12 days) versus 20% of days during the dry season (n =10
W D
days); SMALL group: 83.3% of days in the wet season (n =12 days)
W
versus 50% of days during the dry season (n =10 days)).
D
Intergroup encounters occurred 20 times in the LARGE group

but 44 times in the SMALL group, during a comparable sample of 31

observation days.


PREDICTION 1: If intergroup feeding competition has

contributed significantly to the evolution of primate sociality,

then larger groups must spatially displace smaller groups. If

the outcome of intergroup encounters between pairs of groups is

predictable, then this can be used to define the relative rank of

groups.

Between 1977 and 1982 the SMALL group displaced other groups

only once during 12 intergroup encounters (8.3%; Robinson, 1986,

Table 16). During my study period it won three of 65 encounters

(4.7%; Table 5-1). The only group that the SMALL group

displaced was the BROWN group, another small group of 10 animals.

Of the five encounters, BROWN displaced SMALL twice (40%)

indicating that the ability of these groups to displace one

another was unclear.

Between 1977 and 1982 the LARGE group displaced other groups

on 49 of 60 intergroup encounters (81.7%) (Robinson 1986, Table















4-4
0












it














0)



O0





"I



0---- 0





L-- -L-





I I -i I iI

8V0 W (0 : :D8- Nr N
















ON N








tO- NW T-x m o
-4-
03V4 C4-1 >4 0


-44







C-3



> 3








X n
-4









CQ4
M N




00


+
cvQ
CO



4 +


0


g to M0 0 0 *) W
snd 'i :


8









0
















i3


0













-4
+

T-4
































-3
-q



T-4













4-4












C0
+







+










-


CO





t-4



T-4



C'J



-4
























-4
t,







in



It-



Cn









16). During my study period the LARGE group won 30 of 35

encounters (85.7%) (Table 5-1). The LARGE group consistently

lost to the GREAT NORTHERN and CINNAMON groups, two very large

groups of more than 40 members. On occasion, it also lost to the

CHESTNUT and ORANGE groups. The ORANGE group was equal in size

to LARGE group, the CHESTNUT group was smaller.

The number of different groups that SMALL group encountered

was greater than the number encountered by LARGE group: 10

different groups versus 4. This suggests that SMALL group was

pushed into interstitial areas occupied at least occasionally by

a large number of other groups.

During the eight year period from 1977 to 1984, the LARGE

group always displaced the SMALL group.



Use of Time


Prediction 2: If intergroup encounters are a form of

intergroup feeding competition, then intergroup encounters will

affect the use of time of animals in small groups. I expected

that on days when intergroup encounters occurred, smaller groups

would spend more time moving and foraging for invertebrates than

larger groups. Smaller groups should also rest less, and feed

less on fruit. Intergroup encounters should have little or no

effect on time budgets in larger groups.

Intergroup encounters negatively affect the use of time in

some activities in the SMALL but not in the LARGE group. Table 5-2

compares allocation of time to different activities by animals in

both the LARGE and SMALL groups, during both the wet and dry













TABLE 5-2

THE EFFECT OF INTERGROUP ENCOUNTER ON TIME BUDGETS (%)
OF ANIMALS IN BOTH GROUPS DURING THE WET AND DRY SEASONS


Activity




Forage

Feed

Rest

Move

Drink

Socialize

Non-socialize

Vigilance

Total
instantaneous
samples

Number of days


% Time
LARGE group


Wet season
Intergroup
Yes No

33.1 30.7

16.9 16.0

8.5 8.0

28.9 31.8

0.1 0.1

11.5 13.2

0.3 0.2

0.7 0.0


952


924


Dry season
encounter
Yes No


38.8

15.8

8.4

26.7

0.6

8.8

0.8

0.1


52.2

11.5

5.4

22.7

0.7

6.2

1.3

0.0


855 753


3 3 2 2


% Time
SMALL group
Wet season Dry season
Intergroup encounter
Yes No Yes No


27.5

25.0

10.7

31.2

0.5

4.0

0.2

0.9


22.5

25.6

14.0

26.5

0.2

9.1

1.1

1.0


45.9

24.3

8.4

15.3

0.2

3.3

2.4

0.2


53.1

24.4

8.9

10.0

0.3

1.7

1.3

0.3


552 614 633 631


2 2 2 2










seasons. Animals in the SMALL group spent more time moving and

foraging (sensu latu) on encounter days during the dry and wet

seasons respectively (chi-square test, ldf, p<0.01). But

intergroup encounters did not affect % time animals in the SMALL

group spent feeding and resting in either season, moving in the

wet season, and foraging (sensu latu) in the dry season (chi-

square test, ldf, p>0.01). Intergroup encounters did not

negatively affect % time animals in the LARGE group spent on

moving, foraging (sensu latu), feeding and resting in either

season (chi-square test, ldf, p>0.01).



Group Movement


Prediction 3: Intergroup encounters should affect the

pattern of group movement. This effect should be greater in

smaller groups. I expected that day-ranges and half-hour

movements would be longer on days with intergroup encounters,

especially in small groups. The direction of movement of small

groups should also be altered by intergroup encounters.

The length of the day-range correlated with the number of

intergroup encounters in the SMALL group (r =0.63, n=26 days,
s
p<0.01; Figure 5-2). Ranges were longer on days on which

intergroup encounters occurred. There was no such correlation in

the LARGE group (r =0.14, n=29 days, p>0.1). Average half-hour
s
movements were longer on days that included intergroup encounters

in the SMALL group (Mann-Whitney U-test, n=234 records, n=492

























Figure 5-2. The regression lines of the length of day
ranges and the number of intergroup encounters per day in both
groups.












*


w S
I S
5.


3r


S SMALL GROUP
* 3


3h


LARGE GROUP


21


0 I 2 3 4
NUMBER OF ENCOUNTERS/DAY








records, p<0.05) but not in the LARGE group (Mann-Whitney U-test,

n=355 records, n=379 records, p>0.1).

Intergroup encounters also influenced the direction of group

movements in the SMALL group more than the LARGE group. The

half-hour turning angles immediately following an intergroup

encounter were larger in the SMALL group. The mean half-hour
o
turning angle following encounters in the SMALL group was 87.8
o o
(with median=86.5 SD=57.74 n=35 records) while in the LARGE
o o o
group, it was 54.3 (with median=39.0 SD=49.40 n=14 records).

These angles were significantly larger than mean angles of

turning outside the context of intergroup interactions only in

the SMALL group (Mann-Whitney U-test, n =14 records, n =35
L S
records, p<0.01; Figure 5-3).



Use of Space


Groups in the study area have highly overlapping home ranges

(Robinson, 1986, Figure 23). Figure 5-4 shows that areas used by

the two main study groups were also used by at least eight other

groups during a month's sample in 1986. During the study period

an index of overlap (Holmes and Pitelka, 1968) between the LARGE

and SMALL groups is 0.4099.


Prediction 4: Animals in smaller groups should restrict

themselves to areas away from clumped food resources. The

foraging pattern of the non-tufted capuchin monkeys (Terborgh,

1985; Robinson, 1986) has been described as extended movements

between clumped food resources. If they are denied access to
























Figure 5-3. Distribution of turning angles between half-
hour steps after intergroup encounters in both groups.









-


LARGE GROUP


6b


120


n7


[1


TURNING ANGLES BETWEEN
HALF-HOUR STEPS AFTER
INTERGROUP ENCOUNTERS


180


SMALL GROUP


32r


28
24


20-
16 -
12 -


4
0 -


IL


Or


12 L



























Figure 5-4. Locations of ten groups of monkeys in relation
to combined ranges of the LARGE and SMALL group (265 ha) during
May 23 and June 29, 1984.



















ORANGE


GREY


BROWN


PALE


PINK


COCOA


MAIN


WHITE









clumped resources, animals in smaller groups will not range over

a large area but instead concentrate their activities in a few

specific areas. I expected that quadrats of intergroup

encounters would be more localized (at specific clumped

resources) in the range of large groups, and small groups would

concentrate their activities in certain areas.

Locations of intergroup encounters were more localized in

the range of the LARGE group (Figure 5-5). Figures 5-6 and 5-7

show how both groups used their ranges. The LARGE group spread

its time more evenly over its range than did the SMALL group.

The quadrats of heavy use were also clumped in certain areas of

the SMALL group's range. The use of space of the LARGE group was

more homogeneous. There are more quadrats with higher frequency

of use in the SMALL group (Figure 5-8). The LARGE group occupied

8.1, 21.1 and 40.7 per cent of its home range in 25, 50 and 75

per cent of its time respectively (Figure 5-9). The SMALL group

occupied 5.2, 17.8 and 39.1 per cent of its range in 25, 50 and

75 per cent of its time respectively (Figure 5-9).


Prediction 5: If animals in different groups avoid one

another or if groups displace one another from certain areas, the

use of space by different groups should be negatively correlated.

I expected that the use of area of overlapping by large and small

groups would be negatively correlated.

In the area of overlap, there was no negative correlation

between the quadrat occupancies of the LARGE and SMALL groups

(r =0.0407, n=118 quadrats, p>0.1). Both groups used these
s
areas of mutual overlap as heavily (Wilcoxon-pairs signed-ranks


























Figure 5-5. Locations of intergroup encounters in both
groups.

























LARGE GROUP


o 0


0 0


00
0 0
0 0
10 0


SMALL GROUP