ABSTRACT

BEHAVIORAL ECOLOGY OF THE
PANAMANIAN TAMARIN, Saguinus oedipus
(CALLITRICHIDAE, PRIMATES)

BY

Gary Alan Dawson

The objective of this study was to gather information regarding
the autecology and behavior of the Panamanian tamarin (Saguinus oedipus),
a biomedically-important primate which inhabits areas of second growth
forest from the Department of Choco, in Colombia, to at least central
Panama. Toward this end, a holistic approach to data gathering was
employed. Techniques utilized in order to collect this information
include: observations in the field, the trapping, marking, and
observations of marked tamarins, the tracking of radiotagged groups,
and the examination of the diet, parasites, sex-age-class composition,
and reproductive parameters of a collected sample consisting of 131
tamarins collected at biweekly intervals over the course of one year.

Estimates of population density, and the Problems which arise in
making such estimates, are discussed. Implications regarding the
usage of these census techniques for primates inhabiting large home
ranges under conditions of dense foliage and consequent low visibility
are given.

Food habits, as evidenced by the stomach and intestinal contents
of 129 tamarins taken over the course of a year, are discussed in

relation to information regarding the relative abundance of fruits and

Gary Alan Dawson

insects over seasons, and between upland and lowland sites. Both
diversity and equitability values for components of the diet is
presented according to season. The relationship between the
distribution and abundance of food sources, and tamarin ecology and
behavior, is discussed.

Quantitative information regarding the use of time and Space by
this tamarin are presented. Contrary to earlier reports, this tamarin
exhibits a definite rhythm of activity. Differential usage of
habitats is also explored. The use of both space and time is
correlated with salient ecological conditions and behavioral
peculiarities. A detailed account of nesting or roosting behavior,
hitherto undescribed, is also included.

Data were also collected on the size, composition, and stability
of tamarin social groups. While group size and composition remained
relatively stable over the course of the study, some group constituents
were found to be unstable. The concept of the tamarin social group
as an extended family unit is discarded; it is suggested that a
"typical" social group consists of a breeding pair, the young of the
year, and a revolving complement of sexually mature, but socially
immature, animals. The probable evolutionary advantages of this
latter system are discussed. The correlation between group stability
and the degree of deciduousness of trees on a group's home range site
is also explored.

The absence of pronounced sexual dimorphism, the duality of
male-female roles, evidence for a one male-one female dominance
system, high incidence of intragroup aggression, intraspecific

communication system, foraging strategy, and distribution of resources

Gary Alan Dawson

are considered in defining the species' social structure and the
adaptive significance of that structure in a seasonal environment.

Data regarding tamarin reproduction result from the examination
of the reproductive tracts of sacrificed animals and observations in
the field. While pregnant females or newborn young may be encountered
every month of the year, most births occur from March to June, with
the height of the peak being in April and May. The numerous, newly-
pregnant females collected in May and June indicate another potential
birth peak in August or September. The potential is rarely realized,
however, and young, in fact, are seldom seen outside the March-June
birth peak. No seasonal differences were noted in the reproductive
tracts of either males or females. The possible significance of this
pattern of reproduction is discussed.

The parasites of S. oedipus were also investigated. In general,
the loads of intestinal parasites appear to be relatively low when
compared with data for animals which have been in captivity for some
time. Data regarding the incidence of blood parasites and ectoparasites

are also presented and discussed.

BEHAVIORAL ECOLOGY OF THE
PANAMANIAN TAMARIN, Saguinus oedipus
(CALLITRICHIDAE, PRIMATES)

By

Gary Alan Dawson

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Fisheries and Wildlife

1976

ACKNOWLEDGMENTS

This study could not have succeeded without the unselfish support
of a number of people and institutions. First and foremost, I would
like to extend my sincere appreciation to Dr. R. H. Baker for his
ceaseless guidance and encouragement throughout the course of this
arduous study. I also wish to thank Drs. E. Roelofs, G. Mouser,

P. Murphy, and J. Hunter for their advice and assistance in this
undertaking. 1H; Gorgas Memorial Laboratory I received the able
support of the entire scientific and administrative staff. I would
especially like to thank Dr. M. D. Young for his avid support of my
primate project, and the current director, Dr. P. Galindo, for his
generous support and the wealth of ecological information which he

gave me. The late Mr. A. Clark ably administered the logistic portion
of my study. Dr. R. Beumer, my advisor in Panama, was instrumental in
the planning phases of the study. To Mr. E. Mendez, also of Gorgas
Laboratory, I owe an everlasting debt of gratitude for his generous,
abiding friendship, his assistance in technical matters, particularly in
the identification of insects and ectoparasites, and for the use of his
laboratory facilities. I also wish to thank Mr. Mendez's assistant,

M. Morales, for his many kindnesses. Dr. 0. Sousa and his staff were
responsible for the data on trypanosomes and microfilariae which
accompany this report. I would also like to thank Dr. C. Johnson for

his successful efforts in keeping me alive for the duration of the study.

ii

I also received support from a number of other agencies.
Dr. K. Johnson at the former Middle American Research Unit (MARU)
offered advice and assistance for portions of this study.
Dr. L. Hendricks, who was attached to the U. S. Army contingent at
MARU, kindly identified the tamarin endoparasites. The Panama Canal
Company's Hydrographic Branch provided me with meteorological data and
maps of the area. I am also indebted to Drs. R. Thorington, M. Moynihan,
G. Montgomery, N. Smith, and N. Smythe of the Smithsonian Institution
for their advice on ecological and other technical matters. Ms. B. Cochran
of the AVM Instrument Co. is also to be thanked for her interest in
this project and her promptness in handling my requests regarding
telemetry equipment. The late Dr. R. Ewer, a visiting scientist at
the Smithsonian Tropical Research Institute (STRI), offered both her
friendship and many fascinating discussions on animal behavior. I am
also very grateful to Dr. R. Foster of the University of Chicago for
his brilliant work in identifying the plant components of the tamarin's
diet. Early work in the identification of plants on the study site
was undertaken by Ms. M. Correa of the University of Panama. The
identification of mites found on the tamarins was kindly done by
Dr. F. Lukoschus of the Katholieke Universiteit of the Netherlands
and Dr. A. Fain at the Institute de Medicine Tropicale, Antwerp,
Belguim; ticks were identified by Dr. R. Yunker, Rocky Mountain
Laboratory. Work with saguinus reproductive tracts was facilitated by
Dr. W. Dukelow and his staff at the Endocrine Research Unit, Michigan
State University.

A special debt of gratitude is owed to Mr. G. Barrett Jr. for his
able field work, dependability, and faithful friendship during the many
long days spent in the field.

iii

I would also like to thank L. Jenkins and J. Boger for secretarial
help, and P. Stinson for her artwork. Ms. J. Baglien was a source of
constant help and inspiration throughout the study and in the writing
of this tome. I also wish to thank the D. Baglien family, the
K. Mahler family, and Sra. A. de Linares for their many kindnesses.

This study was supported by grants to the author by the Midwestern
Universities Consortium for International Activities (MUCIA) and the
Lister Hill Fellowship administered by Gorgas Memorial Laboratory. I
wish to thank those responsible for securing this funding, and I extend
my sincere appreciation to Dr. J. Hunter and the staff of the Latin
American Studies Center at Michigan State University, who ably
administered the former grant.

Computer time necessary for the analysis of data on tamarin
reproduction was provided by Dr. W. Magee and Dr. W. Dukelow. Other
calculations herein were facilitated by the Wang 600—lh calculator in

the Michigan State University Museum.

iv

TABLE OF CONTENTS

LIST OF TABLES . . .

LIST OF FIGURES .

INTRODUCTION
An Overview of Neotropical Primate Studies
Medical Importance of the Callitrichidae
Range of Saguinus oedipus geoffroyi
Morphology . . . . .
Coloration . . . . . . . . .

THE STUDY AREA . . . . . . . . . . . . . .

Climate and Photoperiod . . . . .
Topography . . . . . . . . . . . . . . .
Vegetation .

Description of the Rodman Area .
Vertebrate Fauna . . .
Mammalian competitors . . . . . . .
Avian competitors . . . . . .

Mammalian predators

Avian predators . . . .
METHODS AND MATERIALS . . . . . . .
Introduction . . . . . . . . . . .
Trapping of Live Tamarins . .

10
10
15
16
16
17
17
18
20
20

21

TABLE OF CONTENTS (Cont'd) Page

Anesthesia . . . . . . . . . . . . . . . . . . . . . . . . 22
Radio-location Study . . . . . . . . . . . . . . . . . . . 22
Group Composition and Stability . . . . . . . . . . . . . 28
Estimates of Population Density . . . . . . . . . . . . . 29
Data from Collected Tamarins . . . . . . . . . . . . . . . 29
Weight and standard measurements . . . . . . . . . . 30
Sex and reproduction data . . . . . . . . . . . . . . 3O
Sex-age classes . . . . . . . . . . . . . . . . . . . 31
Food-habits analysis . . . . . . . . . . . . . . . . 32
Parasites and their preservation . . . . . . . . . . 33
Preservation of skin and skeletal materials . . . . . 3h

RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . . . . . 35
Results: Estimates of Population Density . . . . . . . . 35
Discussion: Population Density . . . . . . . . . . . . . 39

Results: Analysis of the Diet of the Panamanian
Tamarin . . . . . . . . . . . . . . . . . . . . A3

Discussion: Analysis of the Diet of the Panamanian

Tamarin . . . . . . . . . . . . . . . . . . . AA
Temporal distribution of food resources . . . . . . . AB
Spatial distribution of food resources . . . . . . . 52
Components of the diet . . . . . . . . . . . . . . . Sh

Orthoptera . . . . . . . . . . . . . . . . . . . 57
Lepidoptera . . . . . . . . . . . . . . . . . . 58
Coleoptera . . . . . . . . . . . . . . . . . . . 58
Homoptera . . . . . . . . . . . . . . . . . . . 59
Hemiptera . . . . . . . . . . . . . . . . . . . 59

vi

TABLE OF CONTENTS (Cont'd)

Hymenoptera . . . . . . . . . . . . .
Diptera . . . . . . . .
Thysanoptera . . . . . . . . . . . .

Araneae (Spiders) . . . . . . . . . . . .
Vertebrata .
Plant complement of the diet .

Miscellaneous materials

Results: Utilization of Time and Space

Initiation of daily activity

Cessation of daily activity . . . . . . .
Total activity time . . . . . . .
Quantification of daily activity .
Characteristics of trees used as roosts
Patterns of movement within the home range
Daily travels within the home range

Differential habitat usage as measured by time
spent within habitats . . . . . . . . .

Differential habitat usage by distance travelled
within habitats . . . . . . . . . .

The lowland group vs. the upland group: A
comparison of home range size and utiliZation .

Discussion: The Use of Time and Space .

Daily activity pattern . . . . . . . . . . .

Quantification of the daily activity pattern
and behavioral attributes . . . . . . . . .

Roost trees: Locations and associated tamarin
behaviors . . . . . . . . .

Effects of rain on tamarin activity

vii

59
59
59
59
6O
6O
62
62
63
6h
6h
65
65
68

71

73

76

76
8O
81

8h

89

91

TABLE OF CONTENTS (Cont'd) Page

Patterns of home range usage . . . . . . . . . . . . 91
Daily travels within the home range . . . . . . . . . 92
Differential utilization of habitat types . . . . . . 93
The lowland group vs. the upland group -- a

comparison of home range size, home range

utilization, and territoriality . . . . . . . . . . . 95

Results: Group Composition and Stability . . . . . . . . 98

Trapping success . . . . . . . . . . . . . . . . . . 98
Group size . . . . . . . . . . . . . . . . . . . . . 99
Population sex ratio . . . . . . . . . . . . . . . . lOl
Composition of monitored groups . . . . . . . . . . . 101

Group stability . . . . . . . . . . . . . . . . . . . 10h
Comparison of group stability . . . . . . . . . . . . 10h
Stability among sex-age classes . . . . . . . . . . . 107
Home range data . . . . . . . . . . . . . . . . . . . 109
Importance of group components . . . . . . . . . . . 110

Discussion: Group Composition and Stability . . . . . . . 11h

Results: Reproduction . . . . . . . . . . . . . . . . . . 125
Discussion: Reproduction . . . . . . . . . . . . . . . . 130
Results and Discussion: Parasites . . . . . . . . . . . . 1A2

Helminths . . . . . . . . . . . . . . . . . . . . . . lh2
Ectoparasites . . . . . . . . . . . . . . . . . . . . lhh
Blood parasites . . . . . . . . . . . . . . . . . . . lhh
SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1&7

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . 155

viii

Number

10

ll

12

LIST OF TABLES

Common vegetation on the forested Pacific Slope
of the Panama Canal Zone.

Estimates of population densities for the S. oedipus
population in Rodman, Panama Canal Zone. January,
1973, to May, 197h.

Values for diversity (H') and equitability (E) of
tamarin gastrointestinal contents, April, 1973 -
April, l97h.

Identified plant components of the diet of S. oedipus
based on observations in the field and the stomach
andiIWEStinalcontents of 129 individuals collected
from April, 1973, to April, 197A.

Seasonal distribution of animal components in the
stomach contents of S. oedipus.

Activity pattern of the Panamanian tamarin, S. oedipus
based on the linear distance travelled during bihourly
periods in 26 wet-season days.

Night roosts of the Panamanian tamarin: Tree species
and frequency of usage.

Daily path length and home range Size for selected
Neotropical primates.

Habitat utilization by S. oedipusz Habitat preference
determined by time Spent among 8 habitats during 26

wet-season days.

Variability in daily habitat usage by S. oedipus
during 26 wet—season days.

Habitat utilization by S. oedipus: Habitat preference
determined by distance travelled among 8 habitats

during 26 wet—season days.

Sex ratios within tamarin age classes.

ix

56

66

67

72

7h

75

78

102

LIST OF TABLES (Cont'd)

Number

13

1h

15

16

17

l8

Sex-age class composition of five social groups of
the Panamanian tamarin, S. oedipus.

A comparison of mean group Size and the number of
associated individuals for five tamarin social groups
whose home ranges lie along a moisture-vegetation
gradient.

Emigrations and disappearances among sex-age classes:
Expected values based on sex—age class proportions

in the five monitored groups and the collected sample.

Comparison of body weight, body length, and selected
cranial and dental measurements of mature male and
female S. 0. geoffroyi.

Somatic and reproductive tract measurements of
mature male S. 0. geoffroyi, with a comparison of
measurements over trimesters.

Somatic and reproductive tract measurements of
mature female S. 0. geoffroyi, with a comparison
of measurements over trimesters.

108

111

127

129

Number

2a

2b

LIST OF FIGURES

Rainfall and temperature data for Balboa, Panama
Canal Zone: Readings taken during the study vs.
a 72 year mean.

Number of tamarins seen per census of a h3h0 x 50 m
strip of forest. Rodman, Panama Canal Zone. Dry
season of 1973.

Number of tamarins seen per census of a h3h0 x 50 m
strip of forest. Rodman, Panama Canal Zone. Dry
season of l97h.

Number of tamarins seen per census of a h3h0 x 30 m
strip of forest. Rodman, Panama Canal Zone. Wet
season, 1973.

A comparison of home range size and degree of overlap
of a group of tamarins inhabiting a lowland area and
and group inhabiting an upland area.

Distribution of habitat types and roost trees within
the home range of the lowland group.

Cumulative daily movements of the lowland group over
approximately 500 hours of radio-tracking.

Group Size for 71 distinct groups of the Panamanian
tamarin (S. oedfipus) encountered on the Pacific Slope
of the Panama Canal Zone from January, 1973, to May,

l97h.

Shifts in the time of sunrise and sunset at Balboa,
Panama Canal Zone.

xi

Page

ll

37

37

38

7O

77

100

135

INTRODUCTION

An Overview of Neotropical Primate Studies

Despite the difficulties in studying arboreal primates (Aldrich—
Blake, 1970), many field investigations of New World monkeys (all of
which are arboreal) have been conducted in recent years. The howler
monkey, genus Alouatta, which served as the subject for the first
comprehensive field investigation of New World primates (Carpenter,
193A) has been the object of the majority of studies (e.g. Chivers,
1969; Mittermeier, 1973). Field investigations have also been under-
taken on species of the genera AteZes, Cebus, and Saimiri (8.9.
Thorington, 1967; Thorington, 1968a; Oppenheimer, 1969; Baldwin and
Baldwin, 1972). Both the quantity and the quality of the information
regarding the ecology and behavior of these Species are Substantial.
However, due to the limitations on observation imposed by the densely-
foliated environment, the information regarding platyrrhines (New World
primates) cannot compare with the data base that has been established
for the terrestrial catarrhines of the Old World.

Nevertheless, certain generalizations can be made regarding
Species of the Neotropical genera named above: 1) Under pristine
conditions, these Species tend to form large, relatively stable social
units; 2) these social units are structurally complex, with several
sex and age classes represented within the unit; 3) the principal

dominance hierarchy is predicated on male dominance; and h) the dominance

hierarchy may be characterized as being either uni—male or age-graded
multi-male in structure (Eisenberg et al., 1972).

In contrast, little is known of the ecology or social structure
of those platyrrhines which form small social units, 9.9., Aotus spp.,
CaZZicebuS mOZOCh, or any of the Callitrichidae. The information at
hand regarding their habits and their social organizations is at best
fragmentary, and often anecdotal.

Mason's (1966, 1968) study of CaZZicebuS maloch may be considered
the pioneer field study of a Neotropical primate Species which occurs
in small social units. Moynihan (196A, 1970) discussed the behavior
patterns of Aotus trivirgatus and Saguinus geoffroyi, but his primary
emphasis lay in laboratory study rather than in field observation. The
contributions of Coimbra—Filho and Mittermeier (1973) and Thorington
(1968b) add to our knowledge of callitrichid ecology, but result from
casual observations rather than from comprehensive field study. While
several field studies on members of the Callitrichidae are currently
in progress, none of their findings has yet been published. The
present study is a substantial addition to our understanding of the

ecology and behavior of the medically-important Callitrichidae.

Medical Importance of the Callitrichidae

Marmosets and tamarins (Callitrichidae) have become increasingly
popular in recent years as animals for biomedical research. Their
small Size, their relative tractability when compared to larger Species,
their tendency to produce twins, and, in some Species, their abundance
and relatively low cost are responsible for their popularity in today's

budget-conscious Scientific community. Moreover, while other monkey

populations continue to decline due to over-exploitation and the
destruction of their habitats, some callitrichid Species, 8.9.,

S. oedipus, thrive in the dense, second-growth forests which result
from man's activities in the Neotropics. Thus, potentially, some
callitrichid species offer at least a partial solution to the current
shortage of primates for biomedical purposes.

Marmosets and tamarins have proven to be useful laboratory
animals for a variety of purposes. The presence of chimerism in
their twin offspring make them particularly suitable for some genetics
studies (Benirschke 8t aZ., 1962). They are also widely used in the
studies of virology, protozoology, psychology, and dentistry (8.9.
Sandler 8t aZ., 1966; Deinhardt 8t al., 1967; Hunt 8t al., 1970;
Young, 1970; Dreizen 8t 81., 1972; Patterson 8t aZ., 1973).

The Panamanian tamarin, S. oedipus, is an important reservoir
host of several human diseases. For example, they are known to carry
relapsing fever (Clark 8t aZ., 1931), leishmaniasis (Herrer 8t 82.,
1966), trypanosomiasis (Sousa 8t aZ., 197A), and enteropathogenic
bacteria (Kourany and Porter, 1969). Thus this tamarin, which is
often found in second growth forests close to human habitations, is

a potential source of sylvatic infections.

Range of Saguinus oedipus 980ffr09i

The Panamanian tamarin, Saguinus oedipus 980ffroyil is the only
member of the family Callitrichidae endemic to Central America. The

ancestors of the present form apparently immigrated from South America

 

lFollows the classification by Hershkovitz, 1966.

during the late Pliocene (Hershkovitz, 1969). The modern range extends
from the Colombian province of Choco north to at least central Panama
(Hershkovitz, l9h9; Handley, 1966; Moynihan, 1970). Moynihan (1970)
indicates that S. oedipus (=S. geoffroyi) is characteristic of areas

of moderate humidity, and is almost absent from the humid Atlantic
coast. However, the trapping records of Telford 8t a1. (1972) indicate
that it is extremely abundant in a strip of disturbed vegetation about
5 km wide along the Atlantic coast in the Province of San Blas.
Furthermore, S. oedipus is reported to be abundant on edges of gallery
forest in the humid Province of Darien (P. Galindo, pers. comm.). It
appears then, that a lack of habitat in the appropriate seral stages,
rather than a climatologically-determined habitat type, restricts the

range of this tamarin in the humid lowlands of Panama.

Morphology

Saguinus oedipus geoffroyi is a small monkey; very few individuals
exceed 650 g in weight. Only slight sexual dimorphism exists -- adult
females (pregnant females excluded) in this study averaged h.3 percent
heavier than adult males. Average body length for adult animals of
both sexes is about 235 mm; average body weight is about 500 g. The
Panamanian tamarin, then, while delicate in structure, is a relatively
robust animal.

As in all callitrichids, the tail is not prehensile, and is used
primarily as a balancing organ as the animal forages on the tips of
fragile branches and progresses in long, arching leaps from tree to tree.

The position of the tail, and the presence or absence of piloerection at

a discrete point in time are expressive of the monkey's behavioral
state (Moynihan, 1970).

The Panamanian tamarin has well-developed claws on the front feet;
the digits on the hind feet also possess claws, with the exception of
the hallux, which has a nail and is apposable. The combination of
claws and relatively Short fore and hind limbs equip this animal for
a mode of locomotion which may be characterized as squirrel-like. The
apposable great toe is often used in conjunction with the rear foot
in graSping a branch while the animal is hanging upside down while
eating such pendant fruits as those of Cecropia spp. and Genipa americana.

The dentition reflects the tamarin's insectivorous and frugivorous

2 1 3 2

diet. The dental formula is I—- C—- P— M—

2, 1’ 3, 2, x2=32. The canines are

especially well-developed, and are used as offensive and defensive
weapons as well as for procuring food. Unlike many Old World primates,
notably the baboons and drills, there are no significant differences
in canine size and shape between males and females. The molars are
trituberculate, a dental condition which Hill (1957) regards as primitive.
The short gastrointestinal tract reflects the demands of an
insectivorouS-frugivorous diet. The following measurements (in mm) were
taken from the gastrointestinal tracts of 92 adults: 1) Small intestine,
2': A76 i 8(SE); 2) large intestine 2': 186 i b(SE); 3) caecum ii: 33 :
1(SE). Comparative information regarding the gastrointestinal propor-
tions of other primates may be found in Fooden, 196A; Hladik, 1967;
and Jones, 1972. A comparison of S. 0. geoffroyi intestinal tract

Inicrostructure with that of other New World primates which exhibit

ciifferent feeding habits may be found in Hladik (1967).

The suprapubic gland is larger and more highly developed in
females than in males (Wislocki, 1930a; Moynihan, 1970; Epple, 1972).
In this study the average suprapubic gland width was 32 percent greater
in adult females than in adult males (26.6 mm vs. 20.11mn, respectively).
Seasonal changes were apparent in the gland width of females, but not
males (See section on reproduction). The mean width of female
suprapubic glands was greater during the peak season of parturition
(March—June) than in the remainder of the year. Apparently, endocrino-
logical changes during that period, together with increased frequency of
maternal.'behavior, are responsible for this difference (H. Box, pers.

comm.).

Coloration

The color pattern of the Panamanian tamarin, Saguinus oedipus
geoffroyi, has been adequately described by Moynihan (1970), and is
quoted here. "Individuals of both sexes have the same color pattern,
which is quite complex. The face is largely bare (more so in adults
than in young) and a dark gray in color, but there are stripes of
white hair on the cheeks and the Sides of the nose or snout. There
also is a triangular patch or 'blaze' of white on the forehead and
front of the crown, and the Sides of the bare face are framed by
another pair of whitish stripes (passing behind the naked black ears).
The rear part of the head and the nape are chestnut or rufous red.

The upper part of the body is largely brindled black and grayish-yellow
(the precise arrangement of the black and yellow differs considerably
in different individuals). The underparts and most of the arms and

legs are white, sometimes tinged with yellowish-orange on the breast

and belly. The proximal part of the tail is deep rufous, and the
distal part is black (with a white tip in all or most infants and

young juveniles)."

The following can be added to this description:

The yellowish tinge on the normally white underparts can be quite
pronounced. The variable character is so striking that Elliot (1912)
made this a character for his Species salaquiensis, which he later
withdrew. The ears are dark grey or black, as are the planar surfaces
of the feet and the claw and nails. The iris of the eyes is chocolate;
the pupil black.

The above description iS accurate and agrees closely with my
observations. However, I did not observe a white tail tip in infant
or juvenile animals, nor does Mendez (1970) mention this character
in his description of the animal. It is possible that white tail tips
may be characteristic of some individuals in restricted, local
populations, and that Moynihan's animals came from such a source.

There is no record in the literature of color mutants in S. 0.
geoffroyi. During the course of this study two types of color mutants,
four animals in all, were encountered. The first mutant was a
melanistic animal seen near Gamboa, Panama Canal Zone; this animal
was not collected. It appeared to be entirely black in color. The
other three animals exhibited incomplete pigmentation of the skin on
the face, the planar surfaces, the palmar surfaces, the nails, and
claws. These tamarins were collected between 6 and 7 km west-southwest
of Balboa, Panama Canal Zone, and are housed in the Michigan State
University Museum under catalogue numbers 22916, 22963, and 23038.
0f the three, the male offers the most striking variation from normal

coloration. Pigment is present in only about 10 percent of the nearly

naked facial Skin, and its distribution is spotty. The right and

left ears are approximately 50 and 30 percent pigmented, respectively.
The plantar and palmar surfaces are entirely devoid of pigment; the
dark claws are noticeably lighter than normal; one claw and both nails
are also devoid of pigment. Pigment is absent on the suprapubic gland

of the male; the suprapubic glands of the females are blotchy.

THE STUDY AREA

The Pacific Slope of the Panama Canal Zone was selected as the
site for this study because it: 1) Supports high densities of tamarins,
2) contains second growth forests which have not been disturbed recently,
and 3) contains a system of hard surfaced roads which provides easy
accessability to the study area during both seasons. The entire study
area lies within the Dry Tropical Forest life zone (Holdridge and
Budowski, 1956; Holdridge, 1967). The focal point of the study, the
former Rodman Naval Ammunition Depot (referred to hereafter as Rodman)

is located at 8057'N, 79°57'w.

Climate and Photoperiod

Mean monthly temperaturesimithe Canal Zone average about 800E
(26.7OC) throughout the year. Rainfall, however, is markedly seasonal
with a four month dry season from mid-December to mid—April and an
eight month wet season from mid-April to mid-December. 0n the Pacific
side, rainfall over the dry season averages 5.h8 inches (139.19 mm);
rainfall during the wet season averages 62.83 inches (1595.88 mm)

(based on 72 years of continuous records, Panama Canal Company). Rain-
fall during the 1973 dry season was 3.99 inches (101.35 mm) below the

mean; rainfall for the 1973 wet season exceeded the mean by l6.hh inches
(817.58 mm). Rainfall during the 197A dry season was also below normal,

3.16 inches (80.26 mm) below the mean. The dry season is accompanied

10

by desiccating winds which average 10 mph (16.1 km/hour) and often
gust to over 25 mph (b0.2 km/hour). Rainfall and temperature during
the study are compared with historical averages for these parameters
in Figure 1.

Annual changes in photoperiod at this latitude are Slight --
approximately 63 minutes difference between summer and winter solstices.
Other accounts of Canal Zone climate and photoperiod can be found in

Kaufmann (1962) and Fleming (1971).

Topography

The land rises from sea level to 1180 feet (360 m) and consists
of Small hills and valleys. Milennia of intense weathering on a
lateritic substrate have produced hills which rise abruptly from the
surrounding environment. The flats between the hills are dissected
by small streams, most of which are seasonally intermittent. The
lower reaches of the two main rivers in the study area, the Rio Cocoli
and the Rio Velasquez, are meandering, sluggish streams characteristic
of peneplain surfaces. Both streams run northeastward to the Panama

Canal.

Vegetation

The forest on the Pacific Side of the Canal Zone is primarily a
Second growth forest. Physiognomically, the forest ranges from brushy
vegetation of 1-2 m in height on the edge of the forest to trees about
25 m in height in the gallery forests lining the streams. The bulk of
'the vegetation consists of trees 10-20 m in height. There are, however,

ljirge areas on hillsides and ridges where edaphic factors prohibit the

Figure l. Rainfall and temperature data for Balboa, Panama Canal
Zone: Readings taken during the study vs. a 72 year mean.

11

Figure l.

nV.
ad
em
m .s
c
_
_
.

 

mean
-—-- study

 

 

 

 

r
a
‘II“III‘||.I| e
.m y
I! f
.0
\ \ h
‘\\ t
‘1'”, n
I’I O
m
. 4 . . - . - 1 . . J . 4 A - J . _.J
u n mmmmmmuumuuuuuw
A": maawEQQEmu Ammzoc: :359

12

growth of large trees. Lower growth forms, particularly low brush
and grasses, predominate here and render foot travel difficult.

Structurally, the forest consists of a Single stratum which
follows the contours of tree height. The beginning of a two to three
stratum system is apparent in gallery forests, but even here the
dominants are not uniformly distributed, and the canopy is incompletely
closed.

A table from Fleming (1971) describing the vegetation on his
study grid in Rodman has been expanded to include the common vegetation
(predominantly tree Species) from the study area as a whole (Table 1).

Gallery forests of Anacardium excelsum line both permanent and
intermittent streams. Two palms, Bactris balanoides and Corozo 0Z8if8ra,
are commonly found in lower, wetter areas of the gallery forest. A
fern, Adiantum Zucidfim, iS common along stream banks.

Where a canopy exists in upland Sites, its common components are
CalycophyZZum candidissimum, Cordia alliodora, Enterolobium cyclocarpum,
Ficus sp., Guazuma uZmibeia, and Luehea seemanii. Other emergent
trees which are important on some sites include Spondias mombin,
Cedrela sp., Chrysophyllum cainito, and Scheelea rostrata.

As Fleming (1971) indicated, the shrub layer in this forest is
well developed. Rank growth of vines, grasses and Shrubs are common
on certain sites where tree growth is inhibited by edaphic factors;
additionally, fallen trees are commonplace -- often whole stands are
toppled in high winds. These areas of windfalls and other areas
without a uniform canopy are covered by thickets containing bamboo
grass (Chusquea simplicifZOra), a shrub (HirteZZa racemosa), and, if
the opening is completely unshaded, growth of grasses such as Olyra

Zatifblia and Panicum maximum.

13

Table 1. Common vegetation on the Pacific Slope of the Panama Canal
Zone.

 

TREES

Anacardium excelsum (Bert. and Balb.) Skeels
Andira inenmis H.B.K.

Annona reticulata

Annona spraguei Safford

Apeiba tibourbou Abul.

Bunchosia cornifblia H.B.K.

Bursera simaruba (L.) Sarg.

Byrsonima crassifblia (L.) H.B.K.
Chlycophyllum candidissimum (Vah1.) DC.
Cassia moschata

GastiZZa panam8nsis

Cecropia 8ximia Cuatr.

Cecropia obtusifblia Bertol.

Cecropia peltata L.

Cedrela sp.

Chrysophyllum cainito L.

Cbchlospermum vitifblium (Willd.) Spreng.
Cordia aZZiodbra (Ruiz and Pav.) Roem and Schult.
Ficus Sp.

Genipa americana L.

Guazuma uZmibeia Lam.

Inga Sp.

Lonchocarpus Sp.

Luehea seemanii Tr. and P1.

Mungifera indica L.

Miconia ar98nt8a (Swartz.) DC.
Pittoniotis trichantha Griseb.

Shpindus saponaria L.

Spondias mombin L.

Terminalia Sp.

Vbchysia férruginea Mart.

xylopia frutescens Aubl.

PALMS

Bactris balanoidea (Oerst) Wendl.
Corozo 018if8ra (H.B.K.) L.H. Bailey
Scheelea rostrata (Oerst) Barret.

SHRUBS

Hirtella americana

Hirtella racemosa Lam.

Lacistema aggregatum (Berg) Rusby
Piper reticulatum L.

PSychotria cuSpidbta Bredem.
Paychotria undata Jacq.

Randia Sp.

1h

Table l (Cont'd):

GROUND LAYER

.Adiantum Zucidum Sw.
Chusquea simplicifiora Munro
Gyperus ZuzuZae (L.) Retz.
Olyra Zatifblia L.

Panicum maximum Jacq.

 

15

Many trees, Such as Bursera simaruba, Calycophyllum candidissimum,
and Spondias mombin, Shed their leaves during the dry season. As a
result, the forest floor is generally covered with a leaf layer
several inches thick. In the dry season the leaves are particularly
dry, and quiet foot travel becomes impossible. During the wet season
conditions are more favorable for decay, and the leaf layer diminishes
somewhat, but it is never completely absent. Where the canopy is
dense enough to inhibit the passage of light and thus prevent the
growth of Shrubs, the forest floor is relatively open. In general,
however, unimpeded travel is possible only in stream beds and the

gallery forests.

Description of the Rodman Area

Rodman, which served as the site for behavioral and radio-tracking
portions of this study, is a 1015 ha fenced reserve administered by
Gorgas Memorial Laboratory for field biomedical studies. The topOgraphy
of Rodman does not differ from that of the surrounding area. The
forest cover is representative of the study area as a whole, and has
nOt been disturbed to any great extent since World War II. The tamarin
population in Rodman also appears to be representative of the tamarin
populations on the Pacific Slope. A dendritic system of hard surfaced
roads allows Gorgas personnel access to most of the area during both
wet and dry seasons. As a study Site, Rodman offered the additional
benefit of being close (within 15 minutes drive) to Gorgas Memorial
Laboratory.

Although hunting is prohibited in the area, Rodman is heavily

utilized by local poachers. Fortunately, rather than tamarins, their

16

prime targets are whitetail deer (0dbcoi18us virginianus), collared
peccary (Tayassu tejacu), agouti (Dasyprocta punctata), and paca
(Agouti paca). While tamarins are eaten in some areas of Panama, their
small Size and the high cost of ammunition make them an economically—
marginal target at best. Hunting for meat or Sport is probably a
negligible factor in tamarin mortality; however, capture of young for
the local pet and medical trades may be responsible for substantial

losses.

Vertebrate Fauna

Fleming (1971) indicated that nearly 70 species of mammals are
present in Rodman. The avifauna is also rich and varied, as is the
herpetofauna. Rather than discuss these faunas in detail, I wish to
touch on those Species which are of probable importance to tamarins
as either competitors or predators.

Mammalian competitors: To my knowledge, only three Species of
primates occur in the study area: Saguinus 08dipus, Aotus trivirgatus,
and C8bu8 capucinus. Aotus is not uncommon, and may be a nocturnal
ecological equivalent of Saguinus (Moynihan, 196A). Cebus is locally
rare -- only three were seen during the entire study, and the Cebus
diet, which consists primarily of the larger fruits (Oppenheimer, 1969;
Eisenberg 8t al., 1972), does not broadly overlap that of the tamarin.
Thus, the night monkey, Aotus, may be the only important primate competitor.

Other mammals which have diets overlapping that of Saguinus
include: Marmosa robinsoni, COZuromys dérbianus, Didelphis marsupialis,
Nasua nasua, Bassaricyon gabbii, Potus fiavus, Sciurus granatensis,

and Sciurus variegatoides. The last two mammals are squirrels which

l7

compete directly with Saguinus for fruits, and, because they are also
diurnal, are frequently involved in interspecific conflict. Fruit
bats, such as Artibeus spp., are also competitors, particularly for the
fruits of Ficus spp., C88r0pia spp., and Anacardium excelsum.

Avian competitors: Moynihan (1970) considered several tyrannid
flycatchers to be probable competitors of tamarins for both fruits
and insects. My observations indicate that some flycatchers do eat
the same fruits, but that the primary insect types taken by tamarins
(8.9. Tettigoniid grasshoppers, Lepidopteran larvae, and large
Coleopterans) are not prime flycatcher foods. Flycatchers usually
Specialize on smaller, flying insects. It is more likely that gleaning
Species such as motmots (8.9., Baryphthengus martii), the squirrel
cuckoo (Piaya cayana), and the puffbird (MbZacoptiZa panamensis) are
stronger competitors for insects than are the flycatchers. Oropendolas
(Zarhynchus wagleri) and chachalacas (OrtaZis cinereiceps) are important
competitors for the Small fruits used by tamarins, and are often
chased from fruit trees by this primate.

Mammalian predators: The tayra, Eira barbara, is an excellent
climber and is capable of capturing tamarins in the treetops as well
as on the ground. A tayra has been observed carrying a dead tamarin
(Moynihan, 1970). A second predator, the yaguarundi (F8Zis yagouroundi)
may catch tamarins when the latter are foraging on the ground or
travelling in the very low brush. On two occasions, yaguarundis
approached the investigator while he was carrying traps which contained
vocalizing tamarins. Finally, the ocelot (Felis pardalis), an arboreal

cat, is also capable of taking tamarins.

 

 

 

l8

Avian predators: The large hawks and hawk-eagles are probably
the most important predators of S. oedipus. Important raptors in
this area include: the black hawk—eagle (ButeogaZZus urubitinga),
the white hawk (Leucopternis albicoilis), the tyrant hawk-eagle
(Spizaetus tyrannus) and the ornate hawk-eagle (Spizaetus ornatus).
The tyrant hawk-eagle has been observed to prey upon tamarins (N. G.
Smith, pers. comm.). I observed the black hawk and the ornate hawk-eagle
praying upon the variegated squirrel (Sciurus variegatoides), which is
about tamarin Size and, as noted above, occupies the same habitat.
All of the hawks listed above are certainly potential predators of
S. 08dipus.

Moynihan (1970) indicates that S. oedipus are "... very much

afraid of most birds of prey," and that, .. the only bird of prey
that does not provoke such reactions is the Double-toothed Kite
(Harpagus bidentatus)." Moynihan (ibid.) noted that even perched

birds of prey evoked an antipredator response on the part of the
tamarins. I observed tamarins vocalizing at a Grey—headed Kite
(Leptodon cayanensis), which was perched in a tree at least 50 m from
the group. Tamarins vocalize and initiate escape behavior upon the
appearance of most large, flying objects, including low—flying aircraft.
Moynihan (1970) observed that S. 08dipus, "... burst into high intensity
escape behavior when a Turkey Vulture (cathartes aura) flies by, even
though the vulture, ... probably could do them little damage even if

it wanted to." In general, this statement is probably true. However,

N. Smythe (pers. comm.) once saw Turkey Vultures diving at a group of

tamarins, apparently in an attempt to dislodge twin infants from the

19

back of one of the adults. I once observed a toucan (Ramphastos
sulfuratus) attempting to do the same. (All names of birds cited

herein follow Eisenmann, 1955).

METHODS AND MATERIALS

Introduction

The field study of forest—dwelling primates requires Some rather
Specialized methods, particularly when the monkey involved is small
and easily alarmed, as are tamarins. Aldrich—Blake (1970) discussed
some of the problems involved in gathering ecological and ethological
data on arboreal Species. These problems include: 1) Extreme
restrictions in contact time occur due to interference of thick
vegetation and the altitudinal hiatus which occurs between the arboreal
primate and the earth-bound observer. Habituation of the animal to
the observer is thus more difficult than it might be in an open,
terrestrial environment; 2) Heavy foliage makes it difficult to see
all animals involved in a given behavioral interaction. In addition,
the stimulus which releases various behaviors, or the complete
response of the part of all or even one group member, may not be
visible; 3) Animals are often difficult to observe initially, Since
they remain motionless in heavy foliage; A) Once the monkeys have
been located, it is difficult to follow them in the forest habitat;
5) Some sex and age classes, because of behavioral biases, are more
easily observed than others. This bias is reflected in the data
regarding species behavior; 6) Accurate accounts of group Size and/or

composition are difficult, and often impossible to make; and 7) The

20

21

use of various resources by the animals cannot be readily quantified.
All of these problems were present in the current study, and are

responsible for the development of much of the methodology that follows.

Trapping of Live Tamarins

Tomahawk Live Traps (Tomahawk Live Trap Co., Box 323, Tomahawk,
Wisc. 5AA87) measuring Al cm x 1A cm x 1A cm were used in trapping
live tamarins. Traps were baited either with banana (Musa paradaisica)
or mango (Mangiféra indica), and wired to trees and vines 3-10 m above
the ground in areas frequented by tamarins. Traps were generally set
in clusters of from 3-5, Since the vocalizations of trapped individuals
often drew the attention of other tamarins, which were then captured
in the remaining traps. Traps were washed in stream water, as
necessary, to remove accumulated bait and the odor of trapped animals.

Tamarins are adept at stealing bait from traps. They may enter
the trap only a few inches and then stretch forward and remove the
bait without tripping the pan, or they may insert their paws from
the outside of the trap and extract the bait a fingerful at a time.

I found that wrapping the bait in l/A" (7.6 mm) mesh hardware cloth
and suspending it from a wire above and posterior to the pan, or
placing it underneath the pan so that it was more difficult to
extricate, increased my success.

One attempt was made to capture an entire group through the
use of a banana and brown sugar "stew" liberally laced with ethyl
alcohol. The bait was accepted readily, but the alcohol did not
have the desired effect, i.8. the monkeys did not lose their

locomotory capabilities, albeit they were obviously impaired.

22

Anesthesia

Ketamine anesthesia (Drake, 1972) was used routinely before
handling live-trapped animals in the laboratory. Ketalar (Parke Davis
Co.) was administered intramuscularly in dosages of: 0.2 cc for
infants, 0.3 cc for young juveniles, 0.5 cc for older juveniles and
adults. This dosage proved sufficient for the half hour period
necessary to perform the needed operations. Complete recovery
occurred 6-8 hours following injection. No mortality occurred in

anesthetizing 10A tamarins.

Radio-location Study

Two tamarin social groups were chosen as subjects for detailed
studies regarding the utilization of time and Space within their
respective home ranges. The first group, the lowland group by
designation, occupied a lowland area on one of the tributaries of
the Rio Velasquez. The second group, the upland group, occupied an
upland area which became xeric during the height of the dry season.

I intended to compare the use of time and Space between two groups
inhabiting these disparate sites in order to test a hypothesis
formulated by Janzen and Schoener (1968) which holds that insectivorous
vertebrates (tamarins require large quantities of animal protein)
inhabiting seasonally-xeric upland areas must radically alter their
patterns of Spatial and temporal usage in order to exploit the

seasonal insect resources, while insectivores occupying lowland areas
in which insect populations are more stable maintain their pattern

of usage throughout the year.

23

The area occupied by the lowland group is typical of the lowland
sites where S. 08dipu8 is commonly found. A seasonally-intermittent
stream, a branch of the Rio Velasquez, bisects the home range. A
gallery forest of Anacardium excelsum (to 23+ m) occurs along the
watercourse; trees typical of increasingly drier sites appear as one
moves out of the river valley and onto the uplands bordering the
river. The area is relatively flat in aspect, rising from 12 m to
55 m above sea level at the highest point. The flatness of the
area lent itself to the construction of a grid system of trails,
which divided the area into squares 100 yards (91.5 m) on a side.

The 12,500 ft (3813 m) trail system supplemented the existing road
system in creating a well-mapped study area in which precise estimates
of group location were possible.

The elevation of the area inhabited by the upland group ranged
from A3 m to 127 m above sea level. The vegetation here was markedly
different from that in the lowland group's area. Many of the tree
species on this Site were deciduous in the dry season; the average
height of the forest here was also lower due to edaphic factors and
the seasonal lack of moisture. The terrain was rugged in aspect,
which necessitated the construction of a modified trail system, rather
than a strict grid system, for the purpose of determining group
location. Where the terrain permitted, the grid system outlined
above was constructed. However, when the grade became too steep or
the undergrowth and windfalls prohibitively thick, paths were cut
following passible routes which lay as close to the intended grid line
as possible. The approximate location of these paths was determined

through the use of a measuring tape and a lensatic compass. Over

2A

16,000 ft (A880 m) of trail were cut in the upland group area to
supplement the existing system of roads. This trail system permitted
the monitoring of the upland group over much of its home range. The
large home range of the group, however, and a lack of time and funds
precluded the construction of a trail network which encompassed the
complete home range.

Radio location telemetry equipment used in this study was
manufactured by the A. V. M. Instrument Company (808 W. Springfield Ave.
Champaign, Ill. 61820). Model St-l transmitters mounted on bioperm
plastic collars equipped with whip antennas constituted the transmitter
package. Originally, both continuous Signal and pulsating transmitters
were used, but the former were abandoned in favor of the lighter
(20-25 g), longer-lasting, pulsating models. The weight of the
pulsating transmitter package was approximately A-S percent of animal
body weight, a permissable load for the study of the movements of
free ranging animals (Brander and Cochran, 1969). The output frequency
of all transmitters was controlled by a crystal which oscillated in
the 50 MHz range. The system was designed for maximum output at the
third harmonic (150 MHz), and permitted the reception of weak Signals
at the ninth harmonic (A50 MHz) for close range locations (see
Montgomery 8t al., 1973, for details).

A custom—made LA-ll receiver mounted with a two element yagi
antenna was used to receive directionally the 150 MHz signal. The
effective range of the 150 MHz Signal was much less than the 1200 m
indicated by Montgomery 8t al. (1973) due to the presence of signal-
impeding vegetation and topographic discontinuities. In most cases,
reception of the 150 MHz signal was not possible at distances greater

than 300 m; reception of the A50 HMz Signal was limited to about 100 m.

25

After the trail system had been constructed in the study areas,
the A50 MHz frequency was used almost exclusively Since it allowed
more precise determination of location than the 150 MHz frequency.

In most cases the location of the group (or, at times, only the
telemetered animal) could be determined to within 10 m of the actual
location. Vocalizations and the sound of moving foliage often aided
in obtaining precise locations. When the group was hidden in dense
foliage, triangulation from the grid lines determined location to
within 20 m.

AS noted previously, only the lowland group was radio-monitored
for whole-day periods. In preparation for a day of radio-tracking,
the lowland group was located about 2 hours prior to sunset on the
preceding evening and then followed until it reached its roost tree
for that night. The tamarins were watched for at least 30 minutes
as they sat on the roost. I then returned to the roost before sunrise
on the following day to begin the day's tracking routine.

The behavior in the roost tree was observed, and the exact time
the first tamarin left the tree was noted. Locality and behavioral
data were then recorded throughout the day at 15 minute intervals.

Since I wished to observe the behavior of undisturbed animals,
and since I wished to monitor the natural movements of a free-ranging
group rather than the movements of a harassed group of tamarins fleeing
from an observer, I found it necessary to employ certain tactics which
rendered my presence less obtrusive: 1) Olive drab and camouflage
clothing was worn in order to merge more perfectly with the forest
environment. I also applied mud to my face to avoid facial "shine";

2) I attempted to remain at least 30 m from the group when following

26

it to avoid alarming its constituents. This was usually sufficient
under dense foliage conditions. Careful stalking of the group from
this position was frequently rewarded by glimpses of undisturbed
behavior; 3) once the direction of travel had been ascertained, it
was often possible to use the path system to "circle" ahead of the
tamarins and observe them from a hidden position to the Side of their
path of travel; A) binoculars were used in observing their behavior
at a distance; and 5) when, deSpite precautions, my presence was
detected, I immediately retreated and hid myself from view. Almost
invariably, the tamarins resumed their activity and did not flee from
the area.

In addition to elucidating the differences in temporal and spatial
utilization between two groups inhabiting upland and lowland Sites,
the radio-tracking segment of the study had the following purposes:
1) To determine the activity pattern of S. 0. 980ffroyi; 2) to
quantify variability in usage of habitat types within the home range
and to integrate these findings within the scheme of home range
usage; and 3) to observe and interpret intra- and intergroup behavior.

The linear distance travelled by the lowland group within two-
hour time periods was used as a measure of activity. Location of the
group was determined every fifteen minutes, and the information was
transferred to a map of the area. Linear distances travelled in each
two-hour period were compiled from these maps after the study had
been completed. In analyzing the activity data, I adopted the null
hypothesis that the distance travelled in any given time period would
be proportional to the total time amassed for that period. Expected

values for each time period were obtained by multiplying the proportion

27

of total time in each time period by the total linear distance travelled
over all time periods. A Chi-square test (Sokal and Rohlf, 1969) was
used a priori to test whether or not significant differences occurred
among hourly periods. An a posteriori test employing the Bonferroni Z
statistic (Neu 8t aZ., 197A) was used to identify significant

components of the Chi-square analysis.

The home range of the lowland group was divided into eight habitat
types using height of vegetation, associated tree species and physiognomy
of the upper strata as criteria. My null hypothesis stated that habitat
usage would be proportional to the area encompassed by each habitat.

The amount of time Spent within each habitat and the distance
travelled therein were used as measures of habitat usage. A Chi—square
test (Sokal and Rohlf, 1969) was used in testing the null hypothesis
of proportional usages. Following the rejection of that hypothesis,
an a posteriori test Suggested by Neu 8t aZ. (197A) for usage-availability
data was used to identify those habitats which were preferred, those
which were used in proportion to habitat size, and those which were
used significantly less than expected.

The acceptance of an hypothesis which predicts proportional
usage of habitats which are not uniformly distributed in size or
over space requires the assumption that usage is independent of
location. I feel that this assumption is reasonable Since: 1) the
tamarin is capable of moving rapidly enough to traverse the home range
in any 15-minute observation period, and thus may be expected to use
those habitats within the range by preference rather than by accident
of location; and 2) a Chi-square test testing the distance travelled

and time Spent in the inner vs. outer one-half of the home range

28

indicated that usage did not differ Significantly between the two
elements -- i.8., there was no trend apparent which indicates higher
usage either of a central or core area, or of the peripheral portion

of the home range.

Group Composition and Stability

Live trapping of study group constituents began in Rodman in
January, 1973. Groups were selected whose estimated home ranges lay
at fairly discrete intervals along a moisture-vegetation continuum
which extended from the moist bottomlands with evergreen gallery
forests to seasonally-xeric uplands which support Scrubby, highly
deciduous forests. This sampling scheme was based on an hypothesis
suggested by Janzen and Schoener (1968) which holds that insectivorous
vertebrates inhabiting seasonally xeric tropical uplands must employ
different patterns of spatial utilization than their lowland counter-
parts in order to exploit the more seasonally variable insect resources
characteristic of upland areas.

Sixty-eight tamarins were captured during the study. All were
marked with leather collars wrapped with from one to three strips of
brilliantly-colored plastic tape according to a predetermined code
relating the animal to its group. Longevity of the collars and
readability of the tape proved excellent; only 3 of 68 collars required
replacement during the study period, and all collar codes were readable
throughout the study. Small mammal ear tags (Salt Lake Stamp Co.,

Salt Lake City, Utah) were attached to the left ear of each of the

color-marked animals as an additional means of identification.

29

Five complete social groups were trapped, marked, and observed
during the study. Immigrants were trapped and marked after they had
been observed with the group. Groups were retrapped and observed
at regular intervals from February 1973 to December 1973. From
January to May, 197A, each group was observed at monthly intervals.
Observations were made from blinds while the group fed at pre-baited
platforms 3-5 m above the ground. Identification was aided by the
use of 7-1AX binoculars. Each group was observed for two consecutive
days during each of the observation periods in 197A. No changes in

group composition were noted on any of the second day counts.

Estimates of Population Density

Tamarin densities in Rodman were estimated using a modified version
of the strip-census method described by Hayne (19A9). Tamarins were
censused by two observers from a car travelling slowly (less than
10 mph or 16 kmh) on the Rodman road system. The effective width of
the strip was 30 m during the wet season and 50 m during the dry season.
Estimates were further divided by time of day, 06A5—0830 hours and
0830-10A5 hours in order to compensate for differences in activity
levels and choices of early morning Sites noted while observing
animals. In all, 1,712 km were traversed during the censusing.
Estimates of population density obtained by this strip census were
compared to animal densities based on home range Size as indicated by

telemetry data.

Data From Collected Tamarins

One-hundred thirty-one tamarins were collected with a 20 gauge

30

Shotgun in areas peripheral to Rodman at the rate of five animals
every two weeks over the course of one year. Collection Sites were
changed frequently to avoid the decimation of local populations.

All collecting sites were located at least 1 km from the nearest
observation blind; most were located at least 3 km from Rodman.
Individuals, which had been under observation for 5—15 minutes in
order to ascertain group Size, were collected without known bias.
Upon collection, tamarins were placed in a portable cooler to prevent
tissue degeneration. Data obtained from this collection, and the
methods employed, are outlined below.

Weight and standard measurements: The body weight of intact
individuals was determined to the nearest 0.1 g onennOhauS triple
beam balance (OhauS Scale Corp., Union, N. J. 07083). Standard
mammalogical measurements were taken (in mm) and include: 1) Total
length, 2) tail length, 3) length of ear from notch, and A) length
of hind foot.

Sex and reproduction data: Males were divided into two classes,
mature and immature (see section on sex-age classes). Testes were
preserved in AFA prepared according to a formula from Mosby 8t al.
(1969). Measurements taken on the preserved material include, for
each testis: 1) Length, 2) width, 3) breadth, and A) weight.
Histological cross sections stained with eosin and hematoxylin were
made of the testes of one male from each month's collection. The
testes of the sampled individuals approximated the mean testis size
of all adult males examined for that month. This was done to determine
whether or not changes in gross morphology occurred even though testis

size remained constant over time.

31

Females were divided into five reproductive classes by means of
obvious and traditional criteria: 1) Imperforate, 2) perforate,

3) pregnant, A) lactating, and 5) post-lactating. Incidence and
crown—rump length of embryos were recorded. Female reproductive

tracts were preserved in AFA as indicated above. Measurements of
preserved female reproductive tracts include: 1) Right and left ovary
length, 2) right and left ovary width, 3) right and left ovary breadth,
A) right and left ovary weight, 5) length of right and left oviducts,
6) cervico—vaginal length, 7) cervix length, 8) uterus depth,

9) uterus width, and 10) uterus weight.

Reproductive measurements, together with body length, body weight,
and the length, breadth, and degree of pigmentation of the suprapubic
gland (see below) were compared over three trimesters: 1) November
to February, the peak breeding season; 2) March to June, the peak
season of parturition; and, 3) July to October, a period of comparative
reproductive quiescence. Differences among catagories were determined
using Single classification analysis of variance with a two-sided
hypothesis. Actual analysis was conducted by computer using the
Michigan State University Bastat system. Immature individuals were
excluded from the entire analysis; weights and uterine measures of
pregnant females were also excluded.

Sex-age classes: Tamarins were divided into four sex-age classes;
mature males, mature females, immature males, and immature females.

The criterion used for the separation of immature from mature animals
was the degree of pigmentation of the suprapubic and the circumgenital
glands (gland length and width was also recorded, in mm). These apocrine

and holocrine structures are unpigmented or very Slightly pigmented in

32

prepuberal tamarins (Epple, 1967a). Tamarins reach puberty at
approximately 18 months of age (Hampton 8t aZ., 1966); my observations
on known-age animals indicate that the suprapubic gland becomes
pigmented at 16-18 months of age. This change in pigmentation is
striking and appears to be a reliable method of determining age. Body
weight can also be used as a criterion of division, but it is less
proximally related to the physiological changes which occur at puberty
than is the degree of pigmentation of the suprapubic gland. Only 3

of 38 juveniles in the collected sample exceeded A25 g, no mature
tamarins weighed less than this.

Food—habits analysis: Gastrointestinal tracts were preserved
in 10 percent formalin as soon as possible upon returning to the
laboratory. The esophagus, jejunum, and rectum were secured with
string to prevent loss of internal materials. Fixed food material
was examined as time permitted.

In analysing the material, the small intestine, large intestine,
and caecum were first measured (in mm). Stomach contents and identifi—
able intestinal contents were then taken from the gastrointestinal
tract and placed in petri dishes. A dissecting scope aided in
separating this material for measurement. Insects were identified to
order; plant material was given a number and preserved in 95 percent
alcohol for subsequent identification (plant materials were identified
by Dr. Robin Foster, University of Chicago). All foods were quantified
volumetrically according to the water-displacement method outlined
by Korschgen (1969). Tabulation follows the aggregate volume method

of Martin et al. (19A6).

33

No attempt was made to assess the availability of foods; hence no
strict assessment of food preference is possible. Foods are listed
according to volume and frequency of occurrence. As Korschgen (1969)
indicates, however, "high frequency and high volume indicate a food
of high quality or preference."

Smythe (1970a) and Fleming (1971) outlined the major phenological
differences which occur annually in the Canal Zone. On the basis of
their findings, I wished to compare the diversity and equitability of
tamarin diets over three periods: 1) The dry season, mid—December to
mid—April; 2) the first half of the wet season, mid-April to mid-August;
and 3) the latter half of the wet season, mid—August to mid-December.
Comparisons of diversity and equitability of both animal and plant
material as measured by both volume and occurrence were made using
the standard Shannon-Weiner function as found in Krebs (1972).

Parasites and their preservation: HelminthS encountered while
examining gastrointestinal contents were further preserved in AFA.
Identification of helminths was undertaken by Capt. L. Hendricks of
the Middle American Research Unit, Balboa, Canal Zone.

Ticks, which were attached to the nasal septum, were removed and
preserved in 95 percent ethyl alcohol. Ticks were identified by
Dr. K. Yunker, Rocky Mountain Laboratory, Hamilton, Montana.

Nasal mites were obtained by irrigating the turbinal bones with
water and examining the effluent. The mites were preserved in 95
percent ethyl alcohol and mailed to Dr. F. Lukoschus, Katholicke
Universitiet, Nijnegan, The Netherlands, for identification.

Blood smears and capillary-tube blood samples were obtained from

the blood of 1A8 tamarins. The presence or absence of trypanosomes

3A

and microfilaria was determined through direct examination of thick
and thin, giemsa—stained Smears by Dr. 0. Sousa and his staff at
Gorgas Memorial Laboratory. The trypanosomes of 15 animals were
identified using the techniques of direct blood smear examination,
animal inoculation, hemoculture, and xenodiagnosis described in
detail by Sousa 8t aZ. (197A).

Preservation of Skin and Skeletal materials: Flat Skins or
study Skins were made of all collected specimens. Borax was used as a
preservative, and Skins were dried in an air-conditioned room to
prevent putrification.

Excess tissue was removed from the skeletons; the skeletons were
then placed in a dry, heated container for desiccation. Dried Skeletons
were Shipped to the Michigan State University Museum, where their
further cleaning was facilitated by the use of dermestid beetles. Both
skins and skeletal material are housed in the mammalogical collection
of the Museum, Michigan State University.

Skeletal measurements herein were made with needle point calipers

to the nearest 0.01 mm. Canine measurements follow Kinsey (1972).

RESULTS AND DISCUSSION

Results: Estimates of Population Density

The strip census technique outlined in the methods section was
used on 181 days during the study as part of the standard operating
procedure. Data presented herein are from a 168—day subset of this
period, and represent the information obtained from days on which the
same A3AO m strip was sampled twice, from 06A5-0830, and from 0830-10A5
hours. These records were collected on A5 days during the 1973 dry
season, 70 days during the 1973 wet season, and 53 days during the
197A dry season. In all, 1A58 km were traversed; 729 km during the
prime activity period from O6A5—O83O hours, and 729 km between the
hours of 0830 and 10A5.

Estimates of population densities (Simply animals seen per square
kilometer censused) are given in Table 2. Such density estimates are
commonly found in the current primate literature. No estimates of
precision are given, since the distribution of individual density
estimates over days is highly skewed (Figures 2a, 2b, and 3). In and
of themselves, these data are of little value; they are included
here for heuristic purposes which will be explained in the discussion
section.

Estimates of population density were also obtained from information
on home range and territory usage by the two telemetered groups.

Population density within the home range of the lowland group was

35

36

Table 2. Estimates of population density for the S. oedipus population
in Rodman, Panama Canal Zone. January, 1973, to May, 197A.

 

 

Season Time Animals/km2 Censuses(n)
Dry season, 1973 06A5-0830 9.7 A5
Dry season, 1973 0830-10A5 A.2 A5
Wet season, 1973 06A5-0830 23.1 70
Wet season, 1973 0830-10A5 6.5 70
Dry season, 197A 06A5-0830 28.0 53

Dry season, 197A 0830-10A5 11.0 53

 

Figure 2a. Number of tamarins seen per census of a A3AO x 50 m strip
of forest. Rodman, Panama Canal Zone. Dry season of 1973.

Figure 2b. Number of tamarins seen per census of a A3A0 x 50 m strip
of forest. Rodman, Panama Canal Zone. Dry season of 197A.

37
Figure 2a and 2b.

 

 

 

 

w .

so

20

In
(I)
G)
U)
3
(I)
C
d)
O

. [I [1|[jir4I[:ll_ll:ll_lI ,ilea , , ,JItjl, - - ,fi
u—
° 40‘
h
S
r.

E
3
5 30<

20< - census from 0645-0330 hours

[:1 census from «30- ms hours

 

 

 

 

 

number of tamarins observed

Figure 3. Number of tamarins seen per census of a A3AO x 30 m strip
of forest. Rodman, Panama Canal Zone. Wet season, 1973.

38

82350 mctmES *0

S S on a. 2 2 2 2 2 2 ~— .—

 

9:9. ...z: - as so: «.350 D

950... 83 - 23 E0: 338 I

a

2

U

n

s w

a

.1

w.
8

3

a

U

S

n

m

3 S
as
f .s

 

 

 

:.

.m mazmflm

39

estimated at approximately 27 tamarins per square kilometer. The
density estimate for the wet season home range of the upland group
was approximately 36 tamarins per square kilometer. The dry season
estimate of population density for the upland group's home range
was about 20 tamarins per square kilometer. These observations,
combined with a knowledge of approximate home range size and usage
patterns for three non-telemetered groups, point to a year-around

population density figure of about 20—30 tamarins per square kilometer.

Discussion: Population Density

As the populations<xfwild primates dwindle in the face of habitat
destruction and overexploitation, the need to estimate primate numbers
becomes increasingly important. If we are to preserve the rarer
primate species from extinction, and if we intend to assure the
existence of the viable populations that currently serve as reservoirs
for those primates used in biomedical programs, then we must begin
to gather data on population densities as the first step in the
conservation program. In order to make appropriate management decisions,
we must first have a knowledge of what Species are still extant, where
they are found, and in approximately what numbers. This basic
information is necessary before the most rudimentary of management
strategies can be employed. Census figures should not be misconstrued
to be estimates of primate availability; allowable rate of "harvest"
or "offtake" cannot be determined without a detailed knowledge of a
species' population dynamics.

The current interests in primate conservation by such organizations

as the Institute for Laboratory Animal Resources of the National Academy

A0

of Sciences-National Research Council, the World Health Organization,
and the International Union for the Conservation of Nature have
provided the impetus for much of the current study of Neotropical
primates. Most studies involve estimates of population density; for
several studies with which I am familiar, this is the major objective
(see Southwick, C. H., 1975).

According to Thorington (1968b), direct visual censusing, i.8.
going out into an area and counting all animals that can be seen, is
the most common census method. This technique is most appropriate
for larger, more visible Species, where the recognition of individual
group members precludes the possibility of repeated counts. Thorington
(1972) also indicates that strip censuses are commonly used visual
censusing techniques, particularly where estimates of relative
abundance are required.

Thorington (1971), on the basis of his earlier work (Thorington,
1968b), concluded that, "... tamarin populations can be assessed in a
meaningful way by strip censuses." In addition to visual estimates,
he indicated that tamarin vocalizations might prove useful both in
locating nonvisible individuals and in attracting tamarins from a
distance through the use of the actual or taped calls of a captive
animal (Thorington, 1972; Eisenberg and Thorington, 1973). He also
suggested (Thorington, 1972), that radio telemetry might prove to be
the most useful technique for studying marmosets (tamarins).

Data from strip censuses conducted during this study are presented
in Table 2 and Figures 2a, 2b, and 3. 0n the surface, the differences
in density estimates appear to be sufficiently gross to merit explana-

tion. Typically, one might explain the presence of uniformly lower

A1

density estimates for the 0830-10A5 census periods by observing
that tamarins are most active and vocal early in the morning, and
therefore density estimates taken after this time are likely to be
lower. One might explain the differences between dry season estimates
by noting that the dry season in 1973 was unusually severe, which
forced groups to leave their upland ranges earlier than in that for
197A, when the moderate dry season allowed them to remain in the
upland portions of their range over a greater number of censuses.
Additionally the differences between wet season densities and the
densities from the 197A dry season could be attributed to increased
foraging activity in response to lower levels of both insect and
plant foods. These figures may not, however, be compared in this manner.
As evident from Figures 2a, 2b and 3, the distribution of the
number of tamarins seen per census is decidedly skewed. Because
this distribution in no way approximates a normal distribution, no
estimates of precision can be meaningfully applied to the estimates
of density (means). Thus the apparent differences between these
estimates cannot be compared by parametric methods. Nonparametric
tests are also inapplicable, Since samples of this size require a
modified comparison with an approximate "t" distribution, which is
in fact an approximation of a normal distribution (Sokal and Rohlf,
1969). Even if nonparametric tests were applicable, a comparison
of means for distributions where the modal values and the bulk of
the observations equal zero would not be of great applied interest.
Because of the variability among censuses taken on different
days, the high incidence of the mode, zero, and the considerable

variability of the density estimates themselves, this method is

A2

impractical for censusing tamarins which inhabit densely-foliated
habitats affording low levels of visibility.

The failure of this census method is due to the low probability
of tamarin-observer contact in a large (20—AO ha) home range under
conditions of low (30—50 m) visibility. If one were studying a Species
or population that occupied smaller home ranges and/or lived in a more
open habitat that allowed greater visibility, the probability of
contact would be increased, and the census technique would be more
precise.

It might be argued that precision might also be gained by either
increasing the number of transect censuses taken or by increasing
the length of the transects. While this is undoubtedly true from a
theoretical point of view, practical constraints prohibit such
sampling in the habitats normally occupied by tamarins. The number
of repetitive census taken during this study, and the length of the
transect itself, exceed the time, energy, and monetary budgets for
most field studies. The length of potential transects is further
limited by chronological considerations, i.8. the transect must be
short enough to be traversed over the period of peak activity if it
is to yield the desired results. In the current study, transect
length would be limited to that distance which could be traversed
during the peak morning activity period, which is less than three
hours in length.

The most reliable density estimates in this study result from
data on the home range usage of two telemetered groups and from
data obtained from the approximate home range Sizes and usage patterns

for three additional groups of marked animals. 0n the basis of these

A3

data, an all-season density figure of about 20-30 tamarins per square
kilometer was obtained. While this figure is not precise, it allows
one to compare the relative abundance of tamarins from this area
with that of other areas. Such an index of relative abundance,
rather than an exact estimate of animal abundance, should be the

goal of most studies.

It should be noted that while some of the density estimates
obtained during the strip census fall within the above range, the
relationship of these figures with the true density are strictly
fortuitous. A glance at Figures 2a, 2b and 3 reveals that on most
days one would conclude that no tamarins existed in the study area.
On other days, density estimates might reach 200 tamarins per square

kilometer.

Results: Analysis of the Diet of the Panamanian Tamarin

A cautionary note must precede the presentation of data. The
majority of tamarins collected were taken during the morning hours
since they were more visible and vocal, and hence more vulnerable,
at that time. Observations of foraging animals in Rodman indicated
that tamarins Spent much of this time foraging for insects rather
than fruit. This particular feeding strategy is most efficient
Since winged insects (8.9. large Orthoptera, Lepidoptera), as
poikilotherms, have a lower metabolic rate in the cool temperatures
of early morning and are less likely to fly or escape, than during
the warmer daylight hours. Janzen (1973) observed that sweep samples
for tropical insects did not yield high numbers of insects until

about 0800 hours (presumably after the insects' metabolic rates had

AA

reached levels conducive to activity). The proportion of insects
in the diet, then, may be exaggerated. In support of this view,
large quantities of fruit, presumably ingested during the afternoon
and evening of the previous day, were found in the intestines.
However, C. M. Hladik, in a personal communication to 1%. Moynihan
(1970), indicated that this tamarin may eat relatively more fruit
in the morning and relatively more insects in the afternoon.

The relative amounts of plant and animal material in the diet,
and the diversity and equitability of these food items in the diet,
are recorded by season in Table 3. The three "seasons", dry season,
the first half of the wet season, and the second half of the wet
season, are distinctly different in regard to the abundance and
diversity of insects and plant phenology. The various components
of the animal complement of the diet, and the seasonal incidence in
the diet, are reported in Table 3. Identified fruits and their

importance in the diet are recorded in Table A.

Discussion: Analysis of the Diet of the Panamanian Tamarin

As Thorington (1970) indicates, the food habits of free-ranging
primates are poorly known, deSpite the fact that foraging behavior
is a major component of the daily behavioral repertoire of most
primate Species. A knowledge of the availability and distribution
of a species' food resources is absolutely essential in both defining
the ecological niche occupied by that species and in interpreting
its social behavior. Deinhardt (1970), in reviewing the nutritional
requirements of callitrichids, commented on both the paucity of data

on the natural diet of marmosets and the contradictory nature of the

. u

a

A5

 

mmm.o mew.o mab.o A.ooov pQOGOQEoo Hmes< .sdoomu use ocwpmmpsH
oow.a mmm.a mmm.m A.ooov pQOQOQEoo Hosfls< .sdommo pew mcwpmoch
mmw.o m:®.o sow.o A.ooov m .pcocomsOo pecan .sdoomo can mewmepeH
mma.: mea.: mom.a A.ooov .m .pcocomsoo pecan .adommo one mcflmech
mm~.o mms.o smm.o A.Ho>v m .peocomsoo pewam .ESOOwo use ocflpmmpeH
m®4.m Hmn.m mam.m A.Ho>v .m .pcoeomsoo pnwam .adommo one mswpmmch
mm.mmm ms.mma oa.mmm noov deflaopme pecan madao> .Edommo one mcflpmoch
mmw.o 000.0 www.o A.ooov m .pnocomsOo pqwam .nowBOpm
:wm.m Haw.m Hmm.m A.ooov .m .peoeoasOo panda .nowsOpm
o:>.o mmm.o mms.o A.Ho>v m .pcoeomSoo pcmam .nowSOpm
mwo.m osw.a :ow.m A.HO>V .m .peocomfioo pcwam .nowsOpm
ma>.o Hmw.o emm.o A.ooov m .pcocomsoo Hmsfls¢ .nowaopm
moo.m gwm.m omm.a A.ooov .m .pQOGOQEOO Hwafiq< .momSOpm
smm.o mmm.o owm.o A.Ho>v m .puoeomSoo Hweflq< .nomsOpm
mom.a mom.H wma.a A.Ho>v .m .pcmsomsOo stfls< .nomsOpm
wo.mw om.mm mm.mm mpcmcomsOo pcoam R .sOmsopm
mm.mm o~.mm :m.:s mpSOSOQSOQ Hesflq< R .nomsOpm
ww.amm.m Hm.H>>.m mm.aom.m mm“ .Hmswcm pom Aoov madao> new: .zomsopm
mm.mmm mw.m>m m.mHH Aoov madao> proe .nowBOpm

 

Aoausvaobeoooo oaz Amsusvpmama< as: Assuavaaaaa oaz

Amanda ea: Hanna ea: seashore on:
powwow hcwwm condom Sufism condom hag
mama ccoomm mam: pagan

 

.aeaa Hanna
I mema Hfiam< .mpeopcoo Heeflpmmpewoapmmw caaoemp Mo m hpflaflpepasvo one .m apamao>fie pom mous> .m manna
. A V . A v .

A6

m< I .q cheeses ESNmeaommagb
owwompommm
.Qm owrmmxm N
.am wammmm
becomes»:
Amv .aam eaowxwmmem
owoowaoahwmmem
Hm .Hm I camcom cocAELOpooe: mo mmfioomm cease
m¢ I .om Ampamsmv embemmad eweoowz
memowsopmdamz
mnemm emefleaopmecs mo mmwoomm oco
.Qmaaflz w .ppwam oesopwzm oweommxm
wmmowfipasoowam
.am msaacoowNoq
owwowficmaaflm
.Hm w .ae weresomm annexe
.Hbs< Beekeepwb seesaw
mwoowoawe
caecum emcHSLOpoocs mo monommm 039
.mm mwuws
ommowpfl>
.am eweeazb
.mm emrwNNxsm
owoowwcfimwm
gum I A.q smasN .mv senses mewproaw
nm¢ I mfiomxm A.Qamm w .pammv Eszmoam Exwpadedxw
m< I .A seeps» samkwmebz
Owoowwoamowcd
.ma< .Haosz eaowaweeN esmeoamwm
memowwpaonmsm

mo I .mm eboNoooou
mmoomcommaom
xooo mememsdeoa ommwameb
.mm msowm
.oaaflz oprwmew enema
mm I .mm ofisoasoa
steam owmoaomb
mm I ewNokwmsubo osmosomm
owmowaoz
mscmm cocflanvoocs mo mmfiommm mco
owOOmSersoq
em I zpmsm Amammv Esbemoamme sedumwoeq
macadampmfiooq
.q ESnSNsowbma ammwm
Amv .aam amawm
mmoomaomflm
caecum oocwsampowsd mo mmfiommm 039
memomambfimcflN
.oa soaesm
memomowHfiSm
msemm UOQwSLOpoocz go mmwoomm oeo
mwoomHHwSonm
mdsom vmcwaampouss mo moflowmm mco
.am rompemponxm
momowa<
hoafiwm .m .q A.2.m.mv samkmoNo omoaom
.poaawm A.pmamov euoaumoa emNmmaom
.Hocmz A.pmsoov emeOSUNUQ mwaboem
.mm macrosmmm
meadow

 

mama Hands eoaa soaooaaoo aaaaeaeaoes omfi go
one Q“ mcofipw>nmmpo no women «emo&%%omm .o

.sema Hanna op
mucopcoo chfipmopefl one nomSOpm can one camwm
.m mo poflc map mo mpGOGOQEoo pceam wmfimfipcmcH .n manna

A7

moasom boom pquaomEH no on Op ©m>a6mpo pen .mpcmpcoo nowEoum 2w pemmwnm 902 I
hmpmmEflAp and em cocoaasooo no madao> an &om swap smpmoao I

pmpmmsfiup and 2H mocmaadooo no wadao> an poflv mo somloa I

nopmmaflap and Ca OOGOMASOOO so madao> an pwflo mo moaIm I

GOmcmm pm: was: vacuum I
condom pea yawn pmufim I
QOmwmm 5am I

<:¢1c)r4cutn.:

 

H< I .am sworopemz w
owoownpcwo<

mmfioomm manflmmom 039
meoowpwnasOso

.am ewEmms
promappso

m< I .Qm dwpameNw

oedemamse agenda

No I .QOmflho 8&u26&0®&u menowiouwwm
.am emseaek
.mm septum
mo I Amy .mam emasmmzou
omwodfipsm

am I mscow umcfienopmccd Mo mmflommm one
am I .mm ExreNom
owoochaom

.m.m.m A.qv ewakwmmeae eEweomamm
owoowfinwflmawz

.am eNmapmb
mwooefiamz

.mm sammaxm a
memowammasm

mssmm woswshopqud mo mmfioomm mco

.x.m.m anemones seamen

mo .mm .Ha I Amy .aan ease
mmmoqwssmoq

.q Groomsmse UNNmnawm
mwcoooedeQOmhnzo

.Haflwm eewuesoae ewmoNAN

mo I .mm seesaw

mo I waomeem wmxmeamm Gretta
memoeeocq<

 

”Av.pcoov : manwe

A8

data that are available. One might assume a priori, on the basis of
their primitive, tri-tubercular molars (Hershkovitz, 1972), their
relatively Short gastrointestinal tracts -- indicative of insectivorous
primates (Jones, 1972), and a highly-developed system of villi in
their small intestines (Hladik, 1967), that callitrichids are, in
general, highly insectivorous, and possibly frugivorous as well.
However, Enders (1930) working with S. 0. geoffroyi, reported only
fruits and seeds from gastrointestinal contents, although he observed
that captive tamarins Showed a fondness for insects. Fooden (196A)
found only fruit in the stomach contents of Saguinus midas collected
in Surinam during the December-February rainy season. In contrast,
Hladik (1967) estimated that the diet of S. midbs contained about
1/3 insect material and 2/3 fruits, flowers, and seeds, and Jones
(1972) considered the callitrichid genera Callithrix, C8bu8ZZa,
Leontideus, and Saguinus to be euryphagus. Furthermore, Hladik 8t al.
(1971), in an admittedly fragmentary description of the diet of
S. 0. geoffroyi, concluded that this tamarin utilized both fruit
and a large number of insects.

Hershkovitz (1969) states that, "Marmosets also eat animals
other than insects, such as worms, or small birds, and also seeds
and much fruit when available, but insects appear to be the chief
food, particularly at times when food is scarce." In summary, the
majority of accounts indicate that most Callitrichidae are rather
catholic in their food preferences.

Temporal distribution of food resources: AS indicated earlier,
the climate is markedly seasonal in the Dry TrOpical Forest Life Zone

(Holdridge and Budowski, 1956). The seasonality of rainfall, as one

A9

might expect, profoundly affects the availability of both fruits
and insects, the two most important items in the tamarin diet.
Rather than attempting to measure the seasonal availability of
either fruit or insects, I relied on data compiled by others
regarding gross phenological characteristics of the forests in
Panama, and insect abundance in the Dry Tropical Forest Life Zone.
Janzen and Schoener (1968) discussed differences in insect
abundance and diversity between wetter and drier sites in a Costa
Rican Dry Tropical Forest during the dry season, and also discussed,
both implicitly and explicitly, the differences between wet and dry
season abundance and diversity. Their more pertinent findings and
observations may be summarized as follows: 1) Insect abundance and
diversity decrease over all habitats during the dry season. Smaller
forms, because they are more subject to desiccation than larger
forms, are most affected; 2) Immature stages, particularly caterpillars,
are uncommon in all habitats during the dry season; 3) Biomass of
all insects, including the large orthopterans and lepidopterans,
increases as one moves from dry, upland Sites to wetter, lowland
sites; A) Total dry weight of all arthropods (including Spiders)
decreases as one moves from wetter to drier areas, although the
decrease is less severe than the decrease in numbers, since many of
the larger forms persist, albeit in reduced numberS,at drier Sites;
5) The greater proportion of some insect groups in respect to biomass
(8.9., Orthoptera, Hemiptera, Coleoptera) means that predictions made
concerning the biomass and adaption of vertebrate predators in the
understory should give disproportionate weight to the properties and

changes in abundance of these more abundant prey groups; and

50

6) During the dry season, an insectivore which specializes on a
particular taxon is more likely to encounter its prey in the wetter,
lowland area than in the drier, upland areas. The reverse situation
is postulated for the wet season, when insect biomass should be
greater in upland areas rather than lowland areas. From this last
suggestion, one might infer that lowland habitats harbor fairly
stable insect populations, while the insect populations inhabiting
upland areas experience periods of population irruption followed by
rapid decline.

An account of the seasonal variation in fruit availability in
tropical monsoon forest on Barro Colorado Island in the Panama Canal
can be found in Smythe (1970a, b). While the forest on Barro Colorado
is more mature and the climate there is somewhat wetter than that
found in the study area, both support many of the same tree Species,
and both areas experience marked seasonal changes in phenology due
to the yearly monsoon pattern of precipitation. The timing of
phenological changes by tree species common to both areas is roughly
the same. This is expected since rainfall is dependent upon the
pervasive trade winds, and these two localities, which are less than
25 km apart, are not separated by major orographical barriers.

Smythe (1970a, b) divided fruits into two categories: Type 1,
fruits with diameters greater than 1.5 cm, and type 2, fruits with
diameters less than 1.5 cm. While tamarins do feed on type 1 fruits,
type 2 fruits are the mainstay of their diet. In general, the seeds
of type 2 fruits are dispersed by birds and small mammals (Hladik, 1967;
Smythe, 1970a; personal observations), and usually pass through tamarin

gastrointestinal tracts intact.

51

The primary fruit peak begins with the coming of the rains in
mid-April or early May and continues until about the middle of
August (Smythe, 1970a; Fleming, 1971; personal observation). The
greater biomass of fruit observed during this period is due to the
increased number of all Species in fruit, but is primarily attributable
to the abundance of the large type 1 fruits. The secondary peak of
fruit abundance noted by Smythe on Barro Colorado Island did not occur
in the study area. Type 1 fruits are almost nonexistent from mid-
August to mid-April. The fruiting of smaller, type 2 fruits is,
however, more evenly distributed throughout the year (Smythe, 1970a, b;
personal observations). Nevertheless, the biomass of type 2 fruits
is quite low in the late wet season (October, November, and the first
half of December), and remains low throughout the dry season (mid-
December to mid-April).

Thus the tamarins in the study area must survive a period of
both insect and fruit scarcity from the latter part of the wet season
through the dry season. Foraging activity comprises a larger proportion
of the tamarin's daily routine, particularly in upland areas, and
patterns of spatial utilization change as upland groups move to lower,
moister habitats in order to secure both insect and plant foods. The
volumes of food found in tamarin stomachs at this time of year tend
to be lower when compared with values for the first and second halves
of the wet season (Table 3). However, variability of individual stomach
volumes renders this comparison rather imprecise (F = 2.792, .1 > p < .2,
using log-transformed data). Adult tamarins collected during the dry
season also weighed considerably less than adults weighed during the

last half of the wet season (;.= A8A g vs. ;-= 530 g; F = 23.732,

52

p < .002). While the occurrence of recently-matured tamarins in the
sample may have been responsible for a portion of the difference, the
loss of weight by adults during this period of stress is a real
phenomenon. The omental fat reserves accumulated during the type 2
fruit peak in the wet season are almost totally absorbed during the
dry season, and individuals lack their earlier robustness. The
effects of the food shortage in the late wet season on various
Panamanian mammals are discussed in Smythe (1970a, b) and Mittermeier
(1973).

Spatial distribution of food resources: The preferred habitats
of the Panamanian tamarin are the early seral stages of second growth.
The distribution of these cover types was patchy, but the patches
of second growth in the study area were very large. In fact, areas
of early second growth were the predominant cover type. The Spatial
distribution of both insect and fruit resources within these patches
may be characterized as being dispersed and of low density. The
insect types favored by tamarins are most abundant in areas of low
brush and in the vine covered trees bordering the forest edge. "Edge"
habitats contain more surface area per unit ground area than level
stands, and the productivity in regard to both plant and animal biomass
is greater here than in the mature forest. This well-known "edge
effect" is discussed in many ecology texts (8.9., Odum, 1971). The
large folivorous insects sought by tamarins are themselves dispersed
in space due to the ubiquitous dispersion of their food resources.
Thus the tamarin must travel widely to capture insect prey. Tamarins
do not forage cooperatively, but rather as individuals in a loose group.

Most foraging for animal foods occurs at 2-15 m above the ground,

53

although at times one may observe them foraging in the tops of such
trees as Anacardium excelsum, 20 m or more above the ground. Similar
observations were made by Moynihan (1970). Once, however, I came
upon a group-of tamarins foraging for grasshoppers in the tall grass
bordering a road, and on another occasion, during the dry season, I
observed a single animal foraging for insects among the leaves on the
forest floor.

As indicated above, the plant foods most commonly utilized by
tamarins are disPersed by birds and small mammals. It follows that
within the large patches suitable for their growth, the plants which
bear these fruits are scattered rather than clumped, and that, Spatially,
these fruit types comprise low density resources.

Altmann (197A) observed that low density, dispersed resources
are most efficiently exploited by small social groups utilizing
partitioned (territorially discrete) home ranges, where insect and
plant resources necessary for sustenance are usually available
throughout the year. In upland areas, both fruit and insect biomasses
alternate between prodigious wet season highs and abyssmal dry season
lows. During the wet season, individual food items remain dispersed,
but their overall density in the patch is so high that the patch
itself may be considered an abundant food source. Such
a temporally delimited food source is most effectively used by groups
with overlapping home ranges. This manner of spatial utilization was
used by S. 0. geoffroyi groups inhabiting a seasonally dry upland
area. Nomadism, another efficient mechanism for using temporally
delimited resources, was not observed, but long-distance travel by

upland groups to lowland areas was observed during the dry season.

5A

Altmann (197A) would predict the existence of large group size given
abundant resources. However, the small group size noted in upland
groups may be a parsimonius concession to conditions during the Six
months of the year when food is scarce. Phylogenetic history may also
be an influencing factor here.

The analysis of the diet of almost any mammal, let alone one
living in a tropical habitat exhibiting a high diversity of plant
and animal life, is an ambitious undertaking, and may require years
of painstaking collection and effort. The results presented herein
must be considered a preliminary estimate of the diet of S. 0. 980ffroyi.
It Should be stressed, however, that this estimate, incomplete as it
is, is the most detailed and quantitative estimate to date of the
diet of any of the Callitrichidae, and, with the possible exception
of the work by Hladik 8t al. (1971), of any Neotropical primate Species.

The stomach contents of this tamarin lend themselves well to
quantitative analysis since, in general, individual food items are
poorly masticated. Many fruits are swallowed whole, and the insects,
while chewed more thoroughly, can be separated easily due to their
colorful and distinctive chitinous exoskeletons. It is also possible
to identify and measure the relative volumes of seeds and fruits in
the intestines; the frequency of occurrence of insects may also be
determined because of the longevity of their durable chitinous remains.
Analysis of the intestinal material greatly increased the number of
fruits in the dietary estimate; two additional orders of insects were
recorded as well.

Components of the diet: Arthropods and small vertebrates comprised

almost one-half (A9.5 percent) of the aggregate volume of the stomach

55

contents examined. While this estimate may be too high due to the
behavioral biases mentioned above, the figure does demonstrate that
animal material is an important component of the diet. AS noted
in Table 3, the importance of the animal complement of the diet varied
seasonally, as did diversity and equitability of the components of
the diet.

Diversity of animal components by volume was lowest during the
dry season, when large orthopterans and coleopterans formed the bulk
of the diet, and increased through the wet season. Equitability by
volume, however, was greatest during the first half of the wet season,
when insects of small and intermediate Size were presumed to be abundant.
The equitability values for dry season and late wet season were lower
due to a dependence on fewer, but larger, prey items in fewer
categories. Diversity and equitability values based on occurrence of
food items in the sample favor the first half of the wet season, but
differences among seasons using this measure are probably non-Significant.
On the other hand, the diversity and equitability values based on the
occurrence of insect types in intestinal contents may be a more
useful measure since this analysis permits the sampling of animal
types taken over a longer time period. The high diversity and
equitability values for the dry season indicate that the tamarins
took a variety of animal foods, and probably captured what they could
find rather than expending time and energy in the search for the
more preferred foods.

The incidence and proportion of arthropod and vertebrate prey
items in the diet of 129 collected tamarins are summarized in Table 5.

The importance of each category in the diet will be discussed separately

56

.mmH u 0 .oo 00.0Hm

oEdHo> mpmmmamm<

 

 

.00 n c .00 ma.aw u madao> opwmopmm<w

.w: u z .oo mm.wsa u madao> mpwwmamm<m

.H: u z .oo m~.mm u madao> mpwmmamw<m

.zaso Hwfinmpes mapwwmwpcmofi so UmmmmH
00.0 H0.0 III III III III :.m 00.0 esopafim
0m.a 00.0 0.0 ma.0 H.m No.0 4.0 00.0 wsmpmonSAm
m®.m Hm.0 m.m 00.0 m.: 00.0 0.: mm.a whopmHsOm
mm.m mm.m m.» ma.o 0.:H H0.m III III esopaosom
m0.mH Hm.m m.es 00.0 s.0a me.m m.» 00.0 Aerosoaav seesaw
mm.a mm.m III III H.m e0.m s.m ms.0a Amnes0000H0 oaoaaq
00.0H sa.e m.» . 00.0 m.mm 0m.0 m.ma 0m.0 esopaooaoo
00.0w m0.ma 0.0m 00.0H 0.ms sm.ma 0.: 0H.0 aaopaoeadoq
as.as 0a.0s 0.00 H0.me 0.0» 00.n0 e.m0 mm.ss neopaospso

mocOAASOOOR oszao>& oocwaasOoog madao>& mocmnssOoow madao>w cocoaa5600& medao>R

m

anew am>o H0909

A.ooanasI.ms<IoHav
eeomoom so: mass 000

m

“.maanaaIHaadeIoaav

condom so: ease and

m

Aasaa<IoHeI.oo0Ioaev
QOmwmm ham

 

H

.wmo&%%omm .o .m mo mucopcoo nomEOpm ea mpcmcomsoo Hesfisw mo soapSQHapmHU

Hmsomwmm .m manna

57

in the following paragraphs. Since no data were gathered regarding
the availability of these food items, no definite statements can be
made regarding food preference. However, as Korschgen (1969) indicates
those foods comprising a high percentage of the aggregate volumes, and
exhibiting a high frequency of occurrence, may be thought of as

preferred foods.

Orthoptera: Orthopterans, which constitutedcnmn~70 percent of the
aggregate volume, were easily the most important prey items. Hladik
(1970) and Hladik 8t al. (1971) also reported that large orthopterans
were major items in the diet of this tamarin. Large long-horned grass—
hoppers (Tettigoniidae) were the most important orthopterans encountered.
During the dry season and late wet season the size of Tettigoniids
taken by tamarins was larger than that observed during the rest of the
year. These larger species (and also mature stages of Species with
smaller, immature forms extant in the early wet season) are probably
more capable of withstanding the rigors imposed by the torrential rains
during the late wet season and the effects of desiccation in the dry
season than are their smaller relatives and immature stages. Walking
sticks (Phasmidae) were the only other orthopterans present in the
stomach samples. One cockroach (Blattidae) was discovered in the
intestinal contents. The paucity of cockroaches in the diet was
surprising, since cockroaches are readily eaten by captive tamarins and
were abundant in old squirrel nests and leaf litter which accumulated
in the crotches of trees. The rarity of cockroaches (which serve as
intermediate hosts for acanthocephalans, Dunn, 1963) in the diet may
help to explain the low incidence of this parasite in the collected

samples (see section on helminths).

58

Lepidoptera: Both larvae and adult moths (Suborder Heterocera) and
butterflies (Suborder Rhopalocera) were represented in the diet.
Larval biomass, however, far exceeded adult biomass. Adult lepidopterans
in flight are elusive prey for the tamarins; probably the majority
of adults are captured in the early morning dampness when they are
unable or less able to fly than during the latter part of the day.
The tamarins do, however, attempt to catch adult lepidopterans
throughout the day. The presence of lepidopteran wing scales in the
turbinal bones of virtually every weaned tamarin attests to this
prey's popularity. Lepidopterans, second only to Orthopterans as
wet season food items, were all but absent from the diet in the dry
season (Table 5). This is in keeping with the findings of Janzen
and Schoener (1968), who noted that lepidopterans, and especially
lepidopteran larvae, were practically nonexistent during the dry

seas on .

Coleoptera: Large Coleoptera, primarily cerambycid beetles, were
important food items in the dry and early wet seasons. Janzen and
Schoener (1968) stated that large coleopterans were among the three
common orders of large insects conspicuous during the dry season.
Their presence in stomach samples in the dry season supports this.
The increased frequency of occurrence of coleopterans in the diet in
the early wet season might be predicted on the basis of the existence
of more equitable environmental conditions for most insect Species

at that time. No explanation is offered, however, for the absence

of these insects in samples taken during the late wet season.

:1 r

uwI‘
70a

59

Homoptera: A number of Small homopterans of the family Cicadidae
were present in stomach samples taken in the wet season. Homopterans

are of very minor importance in this tamarin's diet.

Hemiptera: Hemipterans of the family Reduviidae were found in five
stomachs. Low volumes and low frequency of occurrence of this insect
family should indicate that it is probably not a preferred food. It
Should additionally be observed that the dry season values for this
order are low, despite the fact that Janzen and Schoener (1968) found
Hemiptera to be one of the three fairly abundant orders during the
dry season. No individuals of the Triatominae, which serve as vectors
for trypanosomiasis, were encountered. This implies that the high
incidence of trypanosomiasis present in Panamanian tamarins (see
section on blood parasites) is due more to an active host-parasite
relationship rather than to the ingestion of the vector and associated

pathogens by the monkey.

Hymenoptera: Only ants (Formicidae) were encountered in the sample.

These may have been taken incidently as the tamarins foraged on fruits.

Diptera: The one dipteran found in the sample was probably ingested

incidently with fruit.

Thysanoptera: Thrips, which were probably ingested with fruits, were

occasionally found in the intestinal contents.

Araneae (Spiders): Spiders were relatively common components of the
diet, particularly during the wet season. The majority of spiders

utilized appeared to be large and medium-sized orb weavers.

6O

Vertebrata: Two stomachs contained the remains of Iguanid lizards,
one of which was identifiable to the genus AnoZis. While the animal
biomass gained per lizard captured is high, lizards are apparently
uncommon items in the S. 0. geoffroyi diet.

Tamarins were also observed eating bird eggs, although no evidence
of eggs was seen in the stomach samples. This primate also eats

nestling birds (H. Van Horn, pers. comm.).

Plant complement of the diet: A total of 101 fruit types thought to be
distinct species were obtained from the gastrointestinal tracts of
collected tamarins. The flowers and fruits of seven additional
species were eaten by tamarins in the study area. Of the 108 Species,
69 were identified to family, 5A (at least tentatively) to the generic
level, and 26 to Species. Since many trees fruit only every other
year, and Since Panamanian tamarins occupy a wider range of habitats
than that sampled, the probable number of plants used by this monkey
for food is much higher.

In Table A the identified plants are listed by family; their
relative importance in the diet :hs indicated by footnotes. Those
Species without footnotes are considered to be of minor importance,
at least under the environmental conditions which prevailed during the
study. Most plants listed have fruits, in the broadest sense, or
seeds, which tamarins utilize. However, for three trees, Cassia
moschata, Byrsonima crassifbiia, and Luehea seemanii, only the flowers
were observed to be eaten. For a fourth, Cedrela sp., the new leaf
buds are the only portions of the plant having gustatory appeal for

tamarins.

61

In Table 3, the diversity by volume of plant material in the
stomach contents is Shown to be lowest during the first half of the
wet season and highest in the second half. The first value appears
to be out of place, Since this time period Should mark the time of
greatest diversity for available food Species. This apparent anomaly
is due to the fact that the fruit of one plant, Cecropia obtusifblia,
accounted for 69 percent of the aggregate volume of plant material.
The selection of this fruit over all others at the time of greatest
fruit diversity indicates a very strong preference for this fruit.

A glance at the diversity of occurrence values for stomach contents
reveals that the values correspond with the expected trend based on
the phenological information discussed previously. It is apparent
from the first wet season values that those tamarins which fed heavily
on C. obtusifblia also "snacked", as it were, on a wide variety of
other plant material.

The volume of plant contents in the intestine was lowest during
the dry season and highest during the two wet season trimesters. The
low dry season value is undoubtedly due to the Scarcity of fruits
during that season. The intermediate value recorded for the first
half of the wet season, when fruits are most abundant, indicates, in
light of the plant/animal ratio listed above, that, given an abundance
of animal material and an abundance of plant material, the tamarins
will exploit the former over the latter. The high value recorded for
the third trimester is indicative of the Shift from animal to
vegetable material as insects become scarcer late in the wet season,
and also of the availability of fruits during at least the first

half of the trimester.

62

The diversity of plant material in the intestine appears to be
slightly higher in the dry season than in the other trimesters. This
reflects the usage of a wider variety of plants, each of which is
relatively uncommon. The low equitability value for dry season volume
is attributable to the predominance of Alibertia Sp. (33 percent) and
Miconea argentea (11.2 percent) in the sample. The equitability by
occurrence of dry season plant materials provides a more realistic

assessment of the situation.

Miscellaneous materials: One cubic centimeter of red clay was
recovered from the stomach of one tamarin collected during the dry
season. The clay may have been ingested for its mineral content.
Ticks, primarily nymphs of the genus Amblyomma, were recovered
from the stomachs of eight tamarins. These, and the tamarin body
hair found in most stomachs, are probably ingested during grooming

behavior.

Results: Utilization of Time and Space

Quantitative measures of daily activity and habitat usage reported
herein result from detailed observations of one tamarin social group,
the lowland group, which occupied a Site along a tributary of the Rio
Velasquez. The majority of the quantitative measurements are based
on 26 complete days of radio tracking during the wet season and 3
during the dry season. Other data were obtained from the remainder
of over 500 hours of tracking the lowland group. The home range Size
of the lowland group and degree to which that range was overlapped by

other social groups, are compared with that of the upland group, which

63

occupied a seasonally dry hillside site. Home range Size and usage

for the upland group are based on more than 100 hours of radio-tracking
for 1A wet season days and 17 dry season days. Other data regarding
the upland group home range Size and the degree of overlap by other
groups result from frequent observations in this group's home range
over the course of the study.

Initiation of daily activity: I define initiation of daily
activity as the time when the first group member left the roost tree.
Signs of activity were seen before this, including the movement of
group members in the roost tree and/or urination and defecation.

These behaviors were quite variable; thus the time the first animal
left the tree was used as the criterion for the beginning of daily
activity.

I observed the egress of apparently undisturbed (by me) tamarins
from roost trees on 35 days. The average time of egress was official
sunrise time (local) plus 11.7 minutes, with a standard deviation of
i 22 minutes. Available light at that time of morning varied with the
cloud cover and the density of vegetation surrounding the roost tree.
On a few days a flashlight would have been useful in approaching the
roost tree at the time the tamarins left it, although I avoided using
one for fear of disturbing them. On these days it would not have
been possible to observe animals leaving the tree were it not for the
fact that the roost trees were "backlighted". On other days, it was
light enough to read when the tamarins began their diurnal activity.

Extremely early and extremely late times of activity initiation
appeared to be roughly correlated with conditions which influenced the

cessation of activity on the previous day. For example, when heavy rains

6A

caused the tamarins to cease their movements in mid afternoon, activity
began earlier the next morning. Conversely, when the tamarins foraged
after the time of official sunset, they tended to become active
somewhat later the following day.

Cessation of daily activity: Tamarins of the lowland group
were observed entering the roost trees on A7 occasions. The average
time of entrance was official sunset time (local) minus 3A.2 minutes,
within a standard deviation of i 33.7 minutes. Heavy afternoon rains
which continued until after dark had the effect of depressing this
average. In general, however, group activity ceased at least one
half hour prior to sunset. Light levels on most days would have
allowed foraging for an hour or more following the cessation of daily
activity.

Total activity time: Average activity time (minutes) of the
lowland group for 26 wet season days was 676 i 62 S.D. Daily
activity time ranged from A26 minutes, for a day when heavy rain fell
from lAlO hours until dark, to 731 minutes. Total activity time
for 1A days in the early wet season (May-July) did not differ from
activity time observed for 12 days during the late wet season (August-
November; F = .113, p > .75). Rainfall, which was fairly evenly
distributed within and among days, did not suppress activity in any
one time period. A comparison of the total activity time for each
of 3 dry-season days with mean wet-season total activity—time using
a technique for comparing Single observations with means of samples
(Sokal and Rohlf, 1969) indicated no Significant differences (p > .1

for all tests).

65

Quantification of daily activity: A summary of daily activity
using linear distance travelled (path distance) as the criterion is
presented in Table 6. The highly Significant Chi square value
(p < .001 6 d.f.) indicates that distance travelled over time periods
is not proportional to the available time within periods. A definite
activity pattern is apparent —- tamarins travelled Significantly
farther than one would expect during the early morning hours, began
to slow down in late morning, experienced a lull in activity during
midday, and resumed rapid travel at significantly higher rates than
one would expect on the basis of available time during the late
afternoon and evening hours. The behavioral Significance of this
pattern will be discussed.

Characteristics of trees used as roosts: Tamarins were observed
on their night roosts on 61 occasions. At least seven tree Species
were used as roosts; many were used repeatedly. On six of the 26 days
of radio-tracking, the lowland group returned to the same roost which
it left that morning. A summary of tree types and frequency of usage
is presented in Table 7.

Trees used were primarily of two types: tall, usually evergreen
trees such as Anacardium excelsum which possessed broad, leafy crowns;
or Shorter, often deciduous trees which supported dense tangles of
evergreen vines. The mean height of the roost above the ground was
16.0 i 3.5 m (SD) in the former type and 6.6 t 1.7 m (SD) in the
latter tree type. The mean height above the ground of all roosts
was 1A.O i 5.0 (SD). Tamarins roosted in the crown or within tangles

of vines; tree cavities were not used, although they existed.

66

.Aomxwmv coflpaomosm vopomaxo camp Mommas mapcmoweflcmwm cofipnomoam wm>smmbo **
.AoAVflmv cowpaomoaa wopoomxm can» nozoa hapsmoflmwcmflm coapnomoam Um>smmbo *

 

 

 

00H.sa00n0e
remao M 0ws00. 000. s0m.amms mm0s s0e0 000. 000HI00AH
**00HAM00W0FH. ems. 000.0: 0m00 mmsm ass. 00eHI00mH
sees M 0W00H. 0AA. s0m.00 0000 esmm 00H. oomHIoomH
*asH M mumms. 0aa. 0m0.00s was» seem 00H. oomanoaa
saws Mmmwaoa. mos. 00m.m0 m000 prom 00H. 00HHI0000
00HAM00W0sH. 0AA. 000. 0ma0 memo mes. 0000I00s0
*eoeo M 0w0e0. meo. 0B0.00 000m m0mm 000. 00s0I00m0
Hm so mam>umch Awmvmocdpmwm medad> 8 Aavcmaam>wa9 Aavmocdpmfim Ammvoewa onamm
moemwflmcoo oo>aompo mo 0 coedpmfla amocflq newswq Hepoe MO 0509 maasom
Rm0 mo smaflswm: scapuomonm cm>somno UmpooQNm coapaomoam

 

.0500 commmmIpmz mm ao>o muowamm hahsoz
mcflhso 0maam>wap mocepmflo awocwa no women «msmwpmo .m .cwnwswp cwwwsmcem one mo camppwm mpfi>fipo< .0 60909

Table 7. Night roosts of the Panamanian tamarin:

67

frequency of usage.

Tree Species and

 

 

 

Number of Trees Used "n" Nights Total Total
Nights Trees Nights
Tree Species l 2 3 A 5 9

Anacardium excelsum 5 2 1 3 l O 12 29
Spondias mombin 2 O O O O l 3 11
Cassia Sp. 1 1 O l O 0 3 7
Luehea seemanii O O l O 1 0 2 8
Chpania Sp. 0 1 O O O O 1 2
Ficus sp. 1 O 0 O O 0 1 1
Dead trees, 1 1 O O O O 2 3

Vine covered -—- ——— -—- -—- -——-
Total 10 5 2 A 2 1 2h 61

 

68

The locations of the various roostS are given in Figure 5. While
apparently suitable trees are located throughout the home range,
the distribution of those used for roosts is distinctly clumped.
Most roost trees are located in the northeast corner of the home
range, far from the areas of overlap depicted in Figure A. The
others are clumped near the areas of overlap. This distribution,
the relationship between roost tree site and the surrounding vegetation,
and behavior at or near the roost trees, will be discussed.

Patterns of movement within the home range: The travels of
the lowland group within its range were not random; they appeared to
be influenced by a number of factors. The foremost were territorial
behavior and distribution of food resources. AS noted in Figure A
a total of approximately 15 percent of the lowland group's home range
(territory) was overlapped at four locations by four other social
groups. Of the four groups, three were observed to engage in social
altercations with the lowland group quite frequently; only one
agonostic encounter was recorded between the lowland group and the
fourth group. On 19 of the 26 days monitored, the lowland group
entered, or passed within, vocalizing distance of at least three of
the four areas of overlap which it contested with these three groups.
The observation is not incidental; the lowland group often travelled
rapidly to their areas of overlap, and usually two of the four areas
were visited from early to mid morning, the time of which most
territorial disputes took place. On Six of the remaining seven days,
the lowland group visited the areas of overlap of two of the three

groups. On one day in 26, the group visited only one area of overlap.

Figure A. A comparison of home range Size and degree of overlap of a
group of tamarins inhabiting a lowland area and a group
inhabiting an upland area.

69

macho asmm .Qmfiam>o
msosc wapQESpm ammaao>o
0:000 emoao Boz .mmaao>o
msoao zoaamw .mwaam>o
mzono com 302 .moanm>o

mm: o>wmzaoxm mo mwa<

00000 02<000 .mozem 0200

.: masmflm

 

macaw 00m .mwaao>o o

msoao Hawm .meaam>o U
msono scone .mwaam>o o

@5090 nm>Hm .mwaam>o a

was m>wmsaoxe mo wmh< a

00000 0200200 .moz<m 0:00

Figure 5. Distribution of habitat types and roost trees within the
home range of the lowland group.

70

0.0de E w
s +0m .pmmaom hpmaamu D 0.
389002 63004000. «media E

 

pcmwawsm .S NHI: pmopom w
s mHIm .pmmnom came: 7W.////. 0
5 main. .pmmnom omcm I a
mcflnemo pmmnom mum m
a male 50.800 020x 3 m
e BIA .880 a a
000.3 cased/colon? Home Kr
_ .mm geek U
.00 3.50030 Av
”$5050.05 60:09.5 3
.0m 6.0.0.060 0
£550.: «6002000. 0
:33 Scam 03000866000. 4
m Tasman

71

Food availability also influences movements. Marked changes in
movement patterns were noted when Anacardium excelsum, Lacistema
aggregatum, zuxi a wild grape, Vitis Sp., came into fruit. The response
to fruits and stems of A. excelsum during the late dry season, a time
of fruit scarcity, was particularly impressive. For several weeks the
movements of the group appear to be highly correlated with the
distribution of fruiting A. excelsum in the home range.

The distribution of suitable roost trees may also influence
movement.

Daily travels within the home range: The mean linear distance
or path length travelled per day was 2061 m, with a standard deviation
of A02 m. Mean path length travelled per hour was 183 m (600 ft.). The
mean travel distances (the distance between the two extreme locations
on a given day, defined by Mason, 1968) was A68 m, with a standard
deviation of 66 m. The mean area of the home range encompassed by
the daily path travelled was 9.A ha with a standard deviation of 0.6 ha.
Thus approximately 36 percent of the home range was encompassed by the
group's movement on any given day. Path distance travelled in the
outer 1/2 of the home range did not differ from the path distance
travelled in the inner 1/2 of the home range, $2 = .006, 1 d.f.,

p > .9. Mean daily path length and home range Size are compared with
those of other Neotropical Species in Table 8. Tamarins can, however,
travel much faster than this. A single radio-tagged animal from the
upland group averaged 952 m/hour over a one hour period, and the
lowland group once travelled 221 m in Six minutes enroute to a roost

tree, a rate of 2210 m/hour.

72

 

Aeeaaa eases

0.0 00.0

xooNoE mSQmONNNeb

 

A000Hv mam>wno 00I:.sa 00.0 emoNNwa ennesomw

A0m0av nmpeOmneo omanw 0H.0 omoNNes cubeson

A0m0av seasonemaao 0.00 +0.0 032063080 03000

Aemm0av coumcwaony 0.0H III mamaxwem wawswem

A000HV cascaem can caseanm 0.00 0.:I0.0 «pmnmamo wawsmem

A505pm psmmmpmv comsma mMImm A.0.mvon.wmo.m wh©fi$komm 0300000 exewxmem
moemammmm A030 000m mwcwm 080m Ava 500\npmcma spam mmwoomm

 

.mmpwswam Hmowmoupomz vmpomamm MOM 0000 mmswa 0&0: cam npmcma Spam hafiwm .w wands

73

Differential habitat usage as measured by time spent within
habitats: The home range (=territory) of the lowland group was
divided into eight habitat types (Figure 5). Time spent within each
habitat type is summarized in Table 9. The significant Chi-square
value (p < .001, 7 d.f.) indicates that the time spent in the various
habitats is not proportional to habitat area. Low brush (0 — b.O m),
forest edge (h.0 - 15 m) and the gallery forest of Anacardium excelsum
may be considered preferred habitats, since they were used significantly
more than expected on the basis of habitat area, while xeric forest,
mesic forest (open canopy, 6 — 18 m), sparsely forested openings, and
areas of grass were used significantly less. The habitat which
included vine-covered trees with emergent palms (Scheelia rostrata)
was used in proportion to its area. A Chi-square contingency test
(7 X 26) indicated that time spent within the seven utilized habitats
was highly dependent on the day monitored (x2 = S26h, 6 d.f., p < .001).
The variation in time spent within each of the habitats is quantified
in Table 10. Note that the variability in usage was lower in the
preferred habitats. Cumulative time spent in the outer one half of
the home range was not significantly different from cumulative time
spent in the inner one half of the range ($2 = .80h, l d.f., p > .1).
Thus, there was no apparent tendency to use central areas of the home
range more than peripheral areas. A single sample test (Sokal and
Rohlf, 1969) was used to compare time spent within each habitat during
each of three dry season days with the mean time Spent in each habitat
over the wet season. No significant differences were evident, with

p > .1 for all comparisons.

7h

.AmmAflmv :oHpaomoam cmpommxm swap nonmfiz mapqwoflwficmflm cofippomoam ©m>hompo **
.Ammvwmv Cpraomoum vmpommxm gasp nozoa hapcmofiMHcmfim cowpaomopm oo>pompo *

 

 

 

m:m.smamu a
mapflmmom mpflswa mocmcflasoo oz coo. mmm.:mm o mm.:mm omo. 3m.o m
*xomHaMNMWmoa. maa. mmm.pz Hema mm.mmma woo. mm.m a
Hmom. mummo. wmo. moo. mm: 2.33 mmo. is m
.20.“. mummo. mmo. 89mg: mam 3.8% 0:. 8.: m
**:~:am mwsms. awe. omm.amm moon mm.>mam 3mm. mm.m :
*mooanMWHoo. moo. mow.~m a: >m.aaa woo. Hm.o m
3.8.“. meo. 38. Snow mmm $.08 96. 84 m
immmflamwmam. mmm. 38.5: Hmmm 3.83 :5. mm.» a
«m so mpcmflofimmmoo Afimv 8 Aawsv Anfisv Ammv oops Ammv ampssz
mocmowmcou mo mcopr>pmmno m om>hmmno cmpommxm prop go woa¢ popwpwm
:hafiawm: Rmm mo coproaoam mafia mafia cowpaomopm
.mhwc comwomlpms mm mcfindv
mpwpflpwn m moosm pummm mafia hp confispmpoo mozmpmmmam pwpflnwm uw3mwwoo .m 39 cowpwuwawp: papfipmm .m magma

75

 

 

Table 10. Variability in daily habitat usage by S. oedipus during
26 wet-season days.
_Time (min) _pistance (m)
Habitat x SD x SD
1 216.58 97.8h 559.39 158.70
2 20.69 21.08 67.69 55.h8
3 1.69 5.12 h.28 13.13
9 311.35 83.h9 9h3.39 265.20
5 26.23 27.35 1h5.05 11h.93
6 18.92 23.2h 70.15 88.07
7 75.61 68.65 271.39 121.07
8 0 0 0 o

 

76

Differential habitat usage as measured by distance travelled

within habitats: The home range (territory) of the lowland group was
divided into eight habitat types (Fig. 5). In Figure 6, the paths of
group travel for the entire study period are recorded as lines. The
patterns of usage in this figure are further elucidated by the
information in Table ll, where the distance travelled within each
habitat type is compared with the expected distance travelled per
habitat type, based on the hypothesis that distance travelled within
the habitat is proportional to habitat area. The significant Chi-square
value (p < .00], 7 d.f.) indicates that habitat usage as estimated by
distance travelled among habitats was not proportional to habitat area.
Less travel distance was logged in areas 2, 3, S, and 8 than expected,
while more distance was logged in areas A, 6, and 7 than expected.
On the basis of this analysis, habitats h, 6, and 7 are considered to
be preferred habitats. The distance travelled in area I was proportional
to the area of that habitat type in the home range. A Chi-square
contingency test (7X26) indicated that the distance travelled within
given habitat types was highly dependent of days samples (:2 = 3h,7l3,
6 d.f., p < .00l). The variability of habitat usage within each
habitat type is evident from Table ll. Two of the habitats (l and h)
which were preferred by the tamarins have proportionally smaller
standard deviations than those which were not preferred. The relatively
high standard deviation for habitat 6 is due to its isolation as a
single habitat block near the border of the home range. When habitat
6 was used, it was used intensively, as evidenced by both the time
and distance measurements from Tables 9 and ll.

The lowland group vs. the upland group: A comparison of home

range size and utilization: Home range size and the degree of home

Figure 6. Cumulative daily movements of the lowland group over
approximately 500 hours of radio-tracking.

Figure 6.

77

 

   

. ’44- ‘
'4 ,'-/.‘",..’p’ ‘
.I ’ fly

I

:7

 

.6

 

“ .18347%41-44
' la”. ‘ \‘D’I&?fi"\“—
d-sqlfia..>

    

       
      

100 hi

78

.Aomxwmv vmpooaxm swnp pmnmfls mapcmoflmflcmwm ooaam>map moqmpmwo oo>ammpo **
.AmmVHmv ompommxo swap amzoa mapcmofimflcwflm ooaao>wap mocmpmwo om>pmmpo *

 

 

 

Hmo.amm~um8
mapflmmom mpfiefla moeoowmcoo oz 000. HHH.omoa o Ha.omoa omo. 6m.o w
**6ma M swoma. mma. aeo.oo6 6moa 6m.oaam poo. mm.m a
**6mo M mwmmo. smo. 6oH.66 smoa 63.6oma 6mo. se.o 6
*mao M mw~6o. oeo. Hmo.o6mm Heem ms.oomo 6sa. m6.s m
*xs6snm mwmms. 66:. oms.smaa momsm 6:.maooa Ham. 6m.o s
*moo M mumoo. moo. Hao.osm HHH ma.sms ooo. Hm.o m
*mmoam mwamo. mmo. mos.6oa o6ea 6m.6msm 63o. Hm.a m
6amam mw66m. Hem. 666.3 ssmsa am.oossa 2am. mm.» H
am no mpcwfloflmmmoo Afimv R Aav Aav Ammv mead Ammv nopadz
mocmowmcoo mo mcowpw>ummno N omaao>wpe omaam>whe Hopoe mo wma< pmpflpwm
:mawawm: §mm mo Gprhoaoum monopmflm mocmpmwm cowpuomoum
©m>aomno cmpommxm
.mzmo comwmmIpos mm moansw mpwpwpwz
w wpoEm omaam>wAp mocwpmww kn cocHEpmpmc monopmmmnm pwpfipmm "m3mvhmo .m an coopmuwawps pmpwnmm .HH mange

79

range overlap with other tamarin groups are presented in Figure h.
The home range of the lowland group was approximately 26 ha in area.
The lowland group defended this area as a distinct territory; during
the wet season only about 13 percent of its home range was overlapped
by the ranges of other groups. The home range size of the upland
group during the first six months of the wet season encompassed an
area of about 32 ha. During the late wet season and the dry season the
home range included this area and at least 11 additional hectares.
This home range was not defended as a fixed geographical entity, but
was rather shared with five other groups through spatial and temporal
segregation.

During the first six months of the wet season, upland areas
appeared to support higher densities of tamarins than the lowland
areas. A comparison of population densities in the effective home
ranges of the lowland and the upland groups demonstrates this. If
one apportions all areas of home range overlap according to the size
of the groups using these areas and subtracts that proportion of the
home range used by other groups from the total home range figure, one
obtains a crude estimate of the foraging area, or the effective home
range, available for that group. The density estimates arrived at
in this manner are approximately 27 tamarins/km2 for the lowland range
and approximately 36 tamarins/km2 in the upland range. It must be
kept in mind that these figures result from data for only two groups,
and must be regarded cautiously. They do, however, offer an objective
estimate in support of my subjective impressions regarding population

densities in the two habitat types.

80

Discussion: The Use of Time and Space

No animal exists in a vacuum; all living animals occupy three
dimensional space during even the briefest moment of time. The
location of an animal in space is the result of a multitude of
interactions between that animal and the environmental milieu in
which it finds itself, and, to a greater or lesser extent, chance.
The coevolutionary history of the animal and the associated biota
have a bearing on the use of space, as do more proximate ecological
occurrences. Phylogenetic history is responsible for providing the
animal with a potential behavioral repertoire with which to respond
to environmental stimuli; impinging stimuli, the strength of the
primary stimulus, experience, and the behavioral state of the animal
will determine precisely how the animal reacts. Use of time and
space, then, are dependent on the interplay of many ecological and
ethological parameters. The purpose of this discussion is to attempt
to interpret the patterns of temporal and spatial utilization used by
Panamanian tamarins in light of available ethological and ecological
information.

Before proceeding further, it is necessary to define two terms
which will appear from time to time in the discussion. The first,
home range, is defined as "The area in which an animal normally
lives, exclusive of migrations, emigrations, or unusual, erratic
wanderings" (Brown and Orians, 1970). For the second term, territory,
this paper will follow the usage suggested by Noble (1939) and Burt
(l9h3), which is any area defended against encroachment by extra—
social-unit conspecifics. Definitions, such as that suggested by

Schoener (1968), which require that the territory be an area of

81

exclusive use for the occupant, are not flexible enough to account
for the frequent acts of treSpass which occur on the periphery of
large territories.

Daily activity pattern: Moynihan (1970), in describing the
habits of S. 0. geoffroyi, observed that "unlike other diurnal
monkeys with which I am familiar, and most other diurnal mammals and
birds in the same habitats, these tamarins do not become active at,
or very shortly after, dawn. In the wild they are never seen moving
around until at least a quarter of an hour after it has become light.
In most cases, they apparently do not get up until approximately
three quarters of an hour, or more, after full light." He also
observed that, "It is only approximately one hour or half hour before
sunset that they gradually stop feeding and drift off in the general
direction of their sleeping trees." He further states that with

". . one wonders if the animals do

this unusual activity pattern,
not become torpid or semitorpid at night."

My data support Moynihan's (1970) observations. Unlike the
squirrel monkey (Shimiri sciureus), another small insectivore-
frugivore (Eisenberg et al., 1972) whose activity pattern has been
documented (e.g. Thorington, 1967; Baldwin and Baldwin, 1972), this
tamarin neither begins foraging at dawn nor does it customarily
forage in the twilight following sunset. Furthermore, my observations
of roosting tamarins suggest that they may become torpid, or semi-
torpid, at night, as Moynihan hypothesized.

If, for example, one waits for 10-15 minutes before approaching

the roosting animals, they maintain their formation, and, if they

respond to the observer's presence at all, they turn their heads

82

groggily, stare at the observer, and then resume their former position.
The same response can be elicited in the early morning hours. Apparent
torpor has also been observed in S. 0. geoffroyi housed under
laboratory conditions (8. H. Hampton, pers. comm.).

Whether or not actual torpor does occur, the unusual activity
pattern which precludes activity in the crepuscular hours is a real
phenomenon which begs an explanation. At present, theories relating
to the adaptive significance of this behavior must, because of the limited
knowledge available, be regarded as conjectural. A few hypotheses,
however, will be suggested;

Moynihan (1970) advanced the idea that this activity pattern
was adaptive in that it temporally removed the tamarins from
competition for insects with flycatchers, which forage extensively
and intensively during crepuscular periods. While this view may
have some merit, flycatchers, as indicated earlier, tend to specialize
on smaller insects and flying insects, which are more difficult for
tamarins to obtain and probably prohibitively "expensive" in terms of
the tamarins' energy budget. Given the temporal segregation of the
flycatchers and tamarins and the difference ill the type of insects
utilized, I find his suggestion that the similarities between many
flycatcher calls and the vocal repertoire of the tamarins represent
social mimicry to be unlikely. On the contrary, similarities between
flycatcher and tamarin vocalizations are probably the result of the
convergent evolution of vocalization frequencies which are most
efficient in the general habitat occupied by both forms. If the

vocalizations of the two forms were broadly overlapping in time, one

83

would expect to find distinctly different types of vocalizations
occurring in order to prevent the reception of extraneous auditory
signals by either form.

A more probable explanation might be that the tamarin's activity
may revolve around the time of optimal vulnerability of its insect
prey. Large orthopterans, lepidopterans, and coleopterans, because
of their low surface to volume ratio, warm at a slower rate than do
the smaller insects. Hence, while the small insects favored by
flycatchers are warmed quickly and become active shortly after dawn,
these larger-bodied insects remain farther down in the foliage,
gradually absorbing heat and becoming more metabolically active.

The time of their greatest vulnerability to tamarins must occur when
they are emerging from their nocturnal resting places and have not
yet achieved their optimal capacity to escape.

A similar hypothesis might be advanced regarding their early
cessation of activity. When the sun has dropped from its zenith
and cooler temperatures and higher humidities prevail, flying insects
and insects in the outer layer of the foliage become more numerous,
and because of their locations, more vulnerable to flycatchers and
gleaning birds than they were in the heat of the day. The large
insects which the tamarins prefer lose their heat more slowly than
the smaller forms because of their lower surface to volume ratio.
Thus, they are relatively invulnerable to predation by non-aerial
predators until quite late in the evening. The early cessation of
activity by S. oedipus, may be a feasible response when insect food
cannot be obtained, except at high energetic cost, until late in the

evening. Cessation of activity and the assumption of an apparently

8h

torpid or semitorpid condition may have an additional value in predator
avoidance. Small tamarin groups may be less capable of repelling
predators than, for example, larger Saimiri groups, which may forage
well after sunset (Baldwin and Baldwin, 1972).

The tamarin may also have adopted this behavioral pattern in order
to escape competition from crepuscular (and nocturnal) insectivorous
and frugivorous mammals. The night monkey, Aotus trivirgatus, which
appears to be the nocturnal counterpart of Saguinus (Moynihan, 196A,
1970), is fairly common in the study area, and was occasionally
observed in the crepuscular hours. An evolutionary history of
widespread sympatry may be responsible for the temporal segregation
of these two species. The wooly opossum, Caluromys derbianus, an
abundant animal in Rodman, is another possible competitor that is
temporally segregated from Saguinus.

Quantification of the daily activity pattern and behavioral
attributes: Upon examining Table 9, it is apparent that tamarins
travelled significantly farther than expected during the first two
hours of the morning. The relatively longer distances logged are
partially attributable to the tamarin's predilection for feeding on
insects at this time of day. On the majority of days monitored,
foraging began at or very near the roost; usually in areas of low,
brushy vegetation, but occasionally in the top of Anacardium excelsum,
20 m or more above the ground. On other days, the tamarins travelled
100-200 m as a group and then began foraging after having passed over
other areas which appeared to be equally as productive as the area
in which they began foraging. Upon leaving the roosts, the monkeys

began to forage individually and to move as a loose group through the

85

forest. Group scatter at this time was usually h0-7O m. Foraging
was generally extensive rather than intensive, hence the longer path
length. Insects appeared to be the primary objective during at least
the first half hour of the activity period, although fruit was taken
occasionally. Approximately an hour after leaving the roost the
emphasis in foraging activity shifted to fruits. Contact calls
(called long whistles by Moynihan, 1970) which were commonly heard
early in the time block became less frequent as the tamarins reduced
their group scatter. During the latter portion of this time period
and the beginning of the next the monkeys were often seen eating
fruit in single trees and concentrated food patches. On the majority
of days monitored, the group began to move as a unit toward areas

of home range overlap between 06h5 and 0800 hours. These movements
added, in part, to the significantly higher movements per unit time
before 0700 hours.

Most of the behavior involving intergroup encounters occurred
between 0700 and 1000 hours; the major part of the distance travelled
to areas of intergroup conflict was accomplished between 0700 and 0900
hours on 20 of 26 days. While foraging for both fruits and insects
occurred en route to areas of overlap, the path of travel was much
more linear than during the early morning hours, when the animals
were scattered in their foraging. Upon reaching areas of overlap,
the group often devoted much of its time to foraging. In two areas,
Spondias mombin and Inga sp. were common and, during their fruiting
seasons, were attractive sources of fruit.

The opposing group often did not arrive at the overlap area at
the same time as the lowland group. Often the lowland group remained

in the area for an hour or more before the other group appeared; on

86

some days the other group did not visit the area of overlap at the
same time as the lowland group did, if at all. In following the
lowland group, I obviously could not remain at the area of overlap to
determine whether or not the other group appeared at some other time
during the day.

When the groups did encounter one another in an area of overlap,
the behaviors observed were variable. Often a group would leave the
area before the opposing group entered. One could frequently hear
the "long whistle" vocalizations (Moynihan, 1970) of the incoming
group before they were visible. As Moynihan (1970) indicates, "long
whistles" are a bivalent signal which attract group members and repulse
members of other groups. It seems likely that these calls enable
the tamarin groups to mutually avoid one another. When the groups
did come within visual contact of each other, the confrontations were
accompanied by choruses of long whistles by both groups. Moynihan
(1970) also observed that these vocalizations were the principal ones
heard in intergroup agonistic encounters. In seven instances, chases
occurred between members of opposing groups. In three instances
actual physical contact was made. flfiufixephysical contacts were
punctuated by "rasping" vocalizations (a group of vocalizations
defined by Moynihan, 1970).

Marking behaviors, i.e. "sit rubbing" and "pull rubbing" (Moynihan,
1970), and especially the former, appeared to be more frequent in
areas of overlap than in other parts of the home range. However,
since thick brush in all areas made observation of the animals difficult,
comparable quantitative data were not obtained. Scent marking both

preceded and followed agonostic confrontations, and sit rubbing also

87

occurred in the absence of the opposing group. Pull rubbing, as
Moynihan (1970) indicates, is rare except in the more intense behavioral
situations, such as those involving overt conflict, or possibly the
sudden appearance of terrestrial predator. Pull rubbing was observed
only four times, twice following encounters involving physical
contact with members of the other group and twice when the observer
appeared to be the stimulus. Marking was usually initiated by the
dominant (i.8. reproductive) female. The tamarins tended to perform
their rubbing behaviors on protruding knots, stubs, and rough bark
rather than on smooth branches. No evidence was found for the
existence of fixed, customarily-used territorial markers or sign
posts. While differences of opinion exist regarding the function
and information content of scent marking behaviors (Epple, 1967b;
Epple and Lorenz, 1967; Moynihan, 1970; Epple, 1972), they are an
integral component of territorial behavior.

By approximately 1000 hours the tamarins had began moving out
of areas of low brush and into the shade. (Thorington, 1968b, observed
the same tendency in Shguinus midbs). On 12 days the tamarins
travelled less than 200 m from the area of overlap before slowing
their pace and engaging in intermittent resting and foraging activities;
on 10 days the group travelled relatively rapidly across the range
to another area of overlap before their cadence of activity diminished.
On 19 0f the 26 days mid-day activity revolved around stands of
Anacardium excelsum and the large, shady trees in Habitat h. While
activity did not stop between 1000 and 1500, path distance became
significantly shorter than expected due primarily to frequent periods

of quiescence and to less directionality in group movement. Individuals

88

and subgroups foraged sporadically between rest periods, during
which the tamarins sat relatively motionless in the trees or very
occasionally sprawled full length on a branch. Autogrooming and
allogrooming behaviors were seen more frequently during those hours
than during the remainder of the day, with the possible exception
of the first few minutes following periods of rain, when autogrooming
was a common, and possibly obligatory, comfort behavior.

On most days the tempo of activity began to increase at about
1500 hours. Travel between this hour and dark was directional and
relatively rapid, which resulted in significantly longer distances
travelled over the last two to four hours of the day. While insects
were taken during the late afternoon and evening, most travel seemed
oriented toward fruit trees. The animals might travel 100 m or more
in five or fewer minutes and then spend 15-30 minutes in a fruit
tree. The path followed during late afternoon did not appear to
be random; it appeared to be directed by the ultimate goal, the
roost tree, and the location of fruit trees. That a given roost tree
was the ultimate goal was indicated by: l) The direct nature of the
route of travel; 2) the fact that former roost trees and other
presumably suitable trees were passed en route to the nights' roost;
3) the fact that individual roost trees were often used on several
successive nights; and h) the rapidity of group movement over the
last 100—300 m before arriving at the roost. The rapid movement to
roost trees, the direct movements between fruit trees, and the brief
interludes of multidirectional foraging for insects, are responsible
for the significantly higher rates of travel evident in the ultimate

and penultimate time periods.

89

Roost trees: Location and associated tamarin behaviors:
When the locations of roost trees (Figure 5) are compared with the
map showing areas of home range overlap (Figure A), it is apparent
that most roost trees are located in the northeastern portion of
the home range, while the remainder occur near areas of group overlap.
Since apparently suitable trees are located throughout the area,
the reason for this distribution of roosting sites is not apparent.
One might surmise that the location of roosts is dependent upon a
number of factors, the most important of which are patterns of
territorial travel, which might be affected by intergroup relationships,
and temporal variations in the availability of food.

Roost trees, in 22 of 2h cases, were either much higher than
the surrounding vegetation and/or lacking extensive vine connections
with other trees. This lack of physical connection with the surrounding
strata might lessen the probability of detection by nocturnal,
arboreal predators and might also provide an effective "lookout" from
which to observe the surrounding vegetation before beginning the
next day's activity.

I was able to observe clearly the tamarins on the roost in
only 13 of the trees. In each case, the roost was located toward
the periphery of the tree rather than close to the trunk. In all
13 instances a pad of small sticks, leaves and fibrous material was
located in the multiple branches where the animals roosted. The
presence of this material poses an interesting question, "Do tamarins
build nests?" I saw no evidence of tamarins carrying large pieces
of vegetation in their mouths or otherwise participating in nest

building activity. The five "pads" which I examined, however, appeared

90

to have been constructed masses of material instead of an accumulation
of debris. They may have been constructed by squirrels, or partially
constructed by tamarins. Regardless of their origin, these "pads"

of vegetation did exist, and tamarins sat on them.

The existence of nest pads may be important in the epidemiology
of trypanosomiasis with which these tamarins are heavily infected
(Sousa et al., l97h; see also section on blood parasites). The
nest pad, which may remain moist during the dry season, harbors a
variety of insect life. While no triatomine bugs were encountered
in the five pads examined, the existence of the pads and periodic
visitation by the tamarins suggest that such a vector-host relation-
ship is feasible. This situation merits further study.

As indicated in the prior section, the tamarins often move both
directly and rapidly to the roost tree. They also travel silently,
without vocalizing. Rapid, silent movement to sleeping trees is
probably advantageous in predator avoidance.

Upon reaching the roost the monkeys huddle together either on
the pads discussed earlier or within a mass of thick leaves or vines.
The tails are held forward and close to the body, as described by
Moynihan (1970). Their heads are oriented downward and toward the
center of the huddle, thus obscuring the white blaze on the head.
Small infants are carried on the ventral sides of group members.

This series of behaviors would seem to have the effect of making the
animals less conspicuous while at the same time reducing individual
heat loss through huddling.

If the tamarins were disturbed within approximately 10 minutes

of assuming the huddling posture, they usually left the tree. On three

91

nights I unintentionally disturbed the lowland group while trying
to get close enough to see them. On these three occasions the huddle
veritably exploded, with tamarins exiting in all directions. In
each case the troop regrouped and settled into another roost tree
within 15 minutes. On two other occasions the reproductive female
left the group and wandered slowly away from the group, at the same
time uttering low volume, high pitched, single notes of about a
second's duration. After I followed her away from the roost tree,
she returned to the group. This behavior pattern appears to be
analogous to the wing-dragging of female birds protecting a brood.
Further observations must be made to determine whether or not these
behaviors are real components of the species' behavioral repertoire.

In the morning tamarins began moving in the roost at least 10
minutes before leaving it. Few vocalizations were given until after
they had left the roost. Tamarins sometimes left the roost singly;
at times two or three or the entire group left at once.

Effects of rain on tamarin activity: Measurable amounts of
rain occurred on 21 of 26 tracking days. The loss of foraging time
due to rain falling between initiation and cessation of activity
averaged 15 minutes/day. This is a low figure, since one cannot
determine how early the tamarins would have begun or ceased activity
had it not been for the rain. Heavy rain, and prolonged rain,
definitely affect tamarin activity.

Patterns of home range usage: That patterns of home range usage
exist is evident from the above information having to do with activity
over the day and the distribution of roost trees. The salient factors

which influence these Spatial and temporal patterns include territorial

92

behavior, thermoregulatory behavior, anti—predator behavior, foraging
behavior, temporal and Spatial distribution of resources, and cover
types. Upon comparing the travels of the lowland group (Figure 6)

with the figure depicting those areas of the lowland group's home

range that are overlapped by other groups (Figure h), the relationship
between movements and territoriality are obvious. When the accumulated
travels of the lowland group (Figure 6) are compared with a map

showing the cover types in the lowland group's home range (Figure 5),
the relationship between cover type and usage is also apparent.

Daily travels within the home range: In perusing Table 8, one
observes that the daily path length of S. 0. geoffroyi in this study
was second only to that of the Baimiri oerstedi studied by Baldwin
and Baldwin (1972). Both forms are insectivorous and frugivorous;
the long daily path lengths reflect the longer travel distance
necessary to obtain animal protein. The path length of the white
face monkey, Cébus capucinus, which is primarily frugivorous but
somewhat insectivorous (Eisenberg et aZ., 1972) occupies an inter-
mediate position. Howlers (AZouatta villosa) and the ti-ti
(Gallicebus maloch) are primarily frugivorous and folivorous (Carpenter,
1939; Mason, 1966); their shorter daily path lengths reflect the
nature of their less demanding diet.

Home range size and rate of travel observed are reasonably
close to the figures predicted by Moynihan (1970), but the tamarins
did not, as he suggested, travel throughout their home range several
times a day or even once a day. They did, however, visit most of
the points of overlap, even on extreme ends of their range, on most

days.

93

Differential utilization of habitat types: Chapman (1929)
and Enders (1930, 1935) wrote that this tamarin preferred areas of
second growth, and the latter author indicated that they were common
in areas of low, brushy vegetation not inhabited by other local
primate species. Moynihan (1970) stated that this callitrichid had
a pronounced preference for the vegetation along the edges of taller
forests. He further refined this general statement by saying that
S. 0. geoffroyi prefers to inhabit "the vicinity of the edge" rather
than "the edge itself." Moynihan also indicates that extensive
areas of tall forest are not favorable for tamarin populations, and
cited as evidence the inverse relationship between forest growth and
tamarin abundance on Barro Colorado Island.

The above authors based their opinions of habitat usage on
visual observations made while gathering natural history and behavioral
data. I chose to quantify habitat usage in order to provide additional,
more definitive information regarding habitat preference. The reader
is referred to Figure 5 for a breakdown of the lowland group's home
range by habitat; the cumulative group travels in Figure 6 give a
rough idea of travel patterns within and among habitat types. Tables
9 and 11 present data on time Spent and distance travelled in each
of eight habitat types as indices of habitat preference.

Areas of low brush (Habitat l) and forest edge (Habitat h), as
expected, were clearly preferred habitats. While path distance
travelled in low brush was not greater than expected, the time spent
Within the area was significantly higher than expected. These values
reflect the usage pattern for this habitat: intensive, multidirectional
foraging. Areas of forest edge were used extensively for foraging

activities, resting and grooming areas, and as primary arteries of

9h

travel. Intensive use of edge vegetation in foraging undoubtedly
reflects the greater abundance and diversity of both fruits and
insects which commonly occur in such ecotones (Odum, 1971). The
abundant vines and the high foliage density also provide readily
accessible escape cover from predators, a characteristic described
and appreciated by Moynihan (1970).

The gallery forest of Anacardium excelsum (Habitat 7) was also
a preferred habitat. This habitat type was used extensively as a
loafing area during the heat of the day. Some foraging for insects
occurred in the upper levels of this habitat type during the early
morning hours and on several evenings; when the trees were in fruit
during the latter part of the dry season and the early part of the
wet season, these fruits and the fleshy area immediately above the
fruit were a preferred food. The importance of A. excelsum as a
roost tree influenced early morning and late evening movements within
this habitat, which resulted in higher scores in both distance
travelled and time Spent within the habitat.

Habitat 6, an area which, save for the abundance of emergent
palms, resembled the forest edge habitat, was utilized heavily in
foraging for insects and also as a staging area for disputes with
the river group, which occupied a home range to the south of this
location. The tamarins foraged widely in the palm crowns, and
frequently used the fronds as trail runways. Thus my observations
disagree with those of Moynihan (1970), who stated that these tamarins
avoid palms and other monocots. The trapping records of Telford et al.
(1972), who commonly captured this tamarin on the ground in clumps of

HQZiconia spp., also cast doubt on Moynihan's observation.

95

Four habitat types were definitely avoided by the lowland group.
These included areas of open, seasonally deciduous forest (Habitat 2),
areas of open forest on mesic Sites (Habitat 5), a sparsely forested
opening (Habitat 3), and the grass which bordered roads and former
ammunition bunkers (Habitat 8). The forest types held one characteristic
in common - they were all relatively open, with a scarcity of vines.
Productivity of tamarin foods in these areas was undoubtedly lower
than in the surrounding areas. In addition, these open areas may
not have offered sufficient escape cover. That the areas of grass
were not used is surprising, since these areas abut the edge vegetation
most commonly used by tamarins and, at the same time, contain high
densities of orthopterans. A tamarin could easily dash out into
the grass, capture a few insects, and return to the adjoining cover,
as was observed in a small area of grass bordering an old road in
another part of the area. However, the increased risk of predation,
and particularly the vulnerability to aerial predators, may make the
use of these large areas of grass undesirable.

0n the basis of the above findings, the ideal tamarin habitat
appears to be areas of low brush and vine-covered "edge" vegetation
interspersed with taller, non-deciduous trees. The former habitat
types provide both food and cover; the latter provides night roosts
and shelter from the sun in the heat of the day. Thorington (1968b)
also recognized the importance of this mixed habitat type in his
observations of Shguinus midfis in Guiana.

The lowland group vs. the upland group -- a comparison of home
range size, home range utilization, and territoriality: Bates (1970)

in a review of territorial behavior in primates, stated that, "territorial

96

relations between primate groups are far from universal, even among
those species which do establish territories." Struhsaker (1967a and
b), in his study of the vervet monkey (Cercopithecus aethiops), found
that home range and territory are functions of vegetational composition
and distribution rather than group Size. The validity and widespread
applicability of the concept are recognized by Altmann (197A), who
states, "Home range overlap depends primarily on those essential
resources with the most restricted Spatial distribution: it will be
low in relatively uniform habitats and will be extensive if several
essential resources have very restricted distributions." The temporal
restriction of resources, however, is also important, and the word
"temporal" may be substituted for "spatial" in the above quote. The
observations of Jolly (1972) confirm this. In her study, Lemur catta
groups which had formerly lived in discrete, well—defended territories,
adopted a system of time-space segregation with extensive home range
overlap in response to a scarcity of food resources during a severe
dry season.

The wet season home ranges of the lowland and upland groups,
and the extent to which they were overlapped by other social groups,
are presented in Figure A. The differences in degree of overlap
between the two home ranges are striking. The entire home range of
the lowland group was defended as a territory; hence the small degree
of overlap. The home range of the upland group, on the other hand,
was widely overlapped by five groups. Territorial defense, when
defined as the defense of an exclusive area (Schoener, 1968) did not
occur. Rather, the upland group appeared to defend the integrity of

the area around the group rather than a fixed geographical territory.

97

One might further differentiate between the use of Space by the
two groups by observing that while the travels of the lowland group
conformed to regular patterns of home range usage, those of the
upland group appeared to be more random. The wet season home range
size of upland group exceeded that of lowland group, which implies
that resources in the upland area were less available to individual
animals and thus necessitated more widespread foraging. Furthermore,
while the lowland group maintained a relatively stable home range
and pattern of travel throughout the late wet season and the dry
season, the upland group expanded its home range by at least 11 ha and
wandered widely. Seminomadic behavior was common during the late
wet season and the dry season among those groups inhabiting upland
areas. DeMoor and Steffans (1972) observed Similar contrasts between
vervet monkeys (Cercopithecas aethiops) living in gallery forests and
those inhabiting a sandy hillside.

The key to this disparity in usage patterns between the two
groups appearstx>lie in the temporal distribution of resources.
Lowland mesic areas along streams support relatively stable quantities
of food throughout the year, while upland areas undergo great seasonal
changes in insect abundance and diversity (Janzen and Schoener, 1968).

The territorial pattern of usage exemplified by the lowland group
is the expected pattern for a territorial primate inhabiting a
relatively uniform habitat (Altmann, 1979). He further states that
"The natural tendency of animals to occupy all available parts of
the habitat while minimizing competition with conspecifics, combined
with the advantages of an established familiar area, tends to produce

a mosaic distribution of home ranges, with contiguous or minimally

98

overlapping boundaries." The efficient exploitation of predictably
stable and more or less uniformly distributed, low density food
resources, in conjunction with the advantages accrued by living in
an exclusive area, make territoriality the most feasible utilization
strategy for S. oedipus groups inhabiting lowland areas.

In contrast, the usage pattern of upland groups is a feasible
alternative for exploiting resources which are predictably limited
in time. The formation and defense of territories would probably
be energetically unfeasible in areas which provide adequate resources
for only six months of the year. If food resources are richer in
the uplands than the lowlands during the early wet season (as the
higher density of tamarins per unit area for upland group's home
range suggest), the energetic costs of defending territories in the
time of food abundance would undoubtedly be higher in the uplands
than in the lowlands due to the "attractiveness" of the rich resources

to contiguous groups.

Results: Group Composition and Stability

Trapping success: Trapping success was low, averaging 39.0h
trap days/tamarin (68 captures over 5,388 trap days). Of those tagged
animals, only 50 figured prominently in the study. Six infants in
the five groups died or disappeared before I could capture them.
Juveniles were more susceptible to capture and recapture than adults
(x2 = 10.356, p < .005) with expected values for their capture based

on the proportion of juveniles in the combined collection and study

group samples.

99

Group size: Mean group Size was determined for 71 groups.
Rounded mean size was used for 2% groups seen frequently in Rodman;
the remainder of the data results from single observations. Mean
size was found to be 6.39 i .31 when carried infants were excluded
from the count, and 6.93 i .36 when the infants were included
(Figure 7).

Individual tamarins and aggregates smaller than the mean were
often observed during periods of casual observation. However,
observations of such small groups over a longer time period (5—30
minutes) or in a "mobbing" Situation directed toward the investigator,
revealed that these individuals and aggregates were almost invariably
constituents of larger groups. No "floating" population of extra—
group individuals was evident.

The above data must be further qualified by the observation that
the two largest groups appeared to be casual groupings of tamarins
at concentrated food sources rather than permanent Social groups.

Both "groups" were observed only once, although I often passed
through that area.

Group size was compared over three trimesters: 1) November to
February, the peak breeding season, 2) March to June, the peak season
of parturition, and 3) July to October. Carried infants were excluded
from the analysis since they are incapable of participating actively
in intergroup movement, and, moreover, their numbers could be expected
to increase group size in the third trimester. Analysis of variance
indicated that no differences in group sizes existed over trimesters

(F = .2h9, p > .75, d.f. = 2,68).

Figure 7. Group Size for 71 distinct groups of the Panamanian tamarin
(S. oedipus) encountered on the Pacific slope of the Panama
Canal Zone from January, 1973, to May, l97h.

100

mum
a

966

~

 

 

 

 

 

cm 3.22%

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

U
v D m
a.
9
J
O
I,
6—
6
J
O
n
d
s
6—
.8

mhzwwm

101

Population sex ratio: The sex ratio for S. 0. geoffroyi in
this study did not differ significantly from a 1:1 ratio. This held
true for both immature and mature age classes collected by trapping
or by shooting (Table 12). No sex specific mortality factor appears
to be affecting either age class.

Composition of monitored groups: A summary of the sex-age
class compositions for the five monitored groups is given in Table 13.
No definitive statement on either group or population sex-age class
structure can be made on the basis of the composition of five groups.
Neither can a definitive statement be made regarding the sex-age
class structure of a population on the basis of a collected sample
of 131 tamarins. The small sample size of the first estimate militates
against the probability of it being a representative sample of their
population; hidden biases, particularly behavioral ones, might
influence the latter estimate. A comparison of the sex—age class
distributions arrived at by the two methods, while inconclusive in
a strict sense, would, if favorable, lend credence to the validity
of these estimates.

To this end a Chi-square goodness of fit test was used to
compare the sex-age class distribution of the five groups of known
composition with the sex-age class distribution from the collected
sample. The two distributions did not differ significantly; in fact,
they proved practically identical (x2 = 0.359, p > .9, 3 d.f.). This
supports the contention that the sex-age class structures found in
both samples are representative of the population's sex-age class
structure. On the basis of this evidence both distributions are assumed
to be representative, and, for the purpose of this study, valid esti-

mates of population composition.

Table 12.

102

Sex ratios within tamarin age classes.

 

 

A. Collected Tamarins

1

Mature males

Mature females

Immature males

Immature females

B. Trapped Tamarins

1

Mature males

Mature females

Immature males

Immature females

53
146
18

21
19
18

10

.1495

.500

.100

2.286

 

103

.coHHHmomEoo QSOHm mo mHmEHHmm mm mamme mo owmam>4

 

 

 

 

m

.momdmcmo mo H0985: u cm

.Umpsaoxm mason UmHHHmUH

m.o H m.6 m.o H v.0 H.o H H.H m.o H H.m s.o H s.m mHoHoe

m.0 H 2.0 m.0 H 0.0 m.0 H m.a m.0 H :.m 3.0 H 0.m A00 zoaamw

m.0 H m.0 m.0 H 0.0 3.0 H ~.0 H.0 H 0.0 3.0 H 0.: A00 unmama

m.0 H H.m 0.0 H 0.0 0.0 H H.H 0.0 H m.m H.0 H 0.a Rev 00m

:.0 H 0.» m.0 H :.H m.0 H H.H 0.0 H m.w m.0 H H.m va gemVH0

m.0 H H.m H.0 H m.0 m.0 H m.H H.0 H 0.m H.0 H 0.0 AJHV csmazoq

mm H mm .m mm mm mm H mm m m2: 68.8
Hmpoe monEom mHSHmEeH mmamz oHSHmEEH HmmHmEom madpmz mmamz oases:

mommmao ommeom
.m3mwwmo .m .cflamemp qmflsmemgmm map 00 mmsoam Hwfloom w>Hm mo :oHHHmomEoo mmmao ommeom .ma manme

10h

Group stability: While the five monitored groups maintained
relatively stable group sizes and sex-age class proportions (Table
13) their particular complements of individuals proved unstable.
Forty-one immigrations and emigrations (considering emigration and
subsequent immigration by the same individual as separate events),
fourteen disappearances, and one death were recorded while observing
the five groups over a period of from 10-15 months per group. Since
systematic, i.8. monthly counts, were made only during the last five
months, additional changes in group composition may have gone unnoticed.
Twenty immigrations and emigrations, and Six disappearances were
recorded during 19 observation periods during the wet season.
Twenty-one immigrations and emigrations, eight disappearances, and
one death were recorded over 29 dry season counts, 25 of which were
made during the l97h dry season. Rates of mobility between wet and
dry seasons are not comparable due to the irregular intervals among
wet season censuses. The data demonstrate, however, that changes
in group composition occurred frequently during both seasons.

Comparison of group stability: Two measures were used to compare
stability among groups: 1) The percentage of individuals associated
with a group above the mean for that group; and 2) comparisons of
median stability, where group stability is based on the stability
value of its components, and stability itself is defined as the
frequency of occurrence of individuals over pertinent sampling periods.
The first measure is crude in that it is sensitive only to gross
changes in group association; the second is more finite since it

also measures the tenure of individual constituents.

105

In Table 1h, groups are organized along a gradient according to
the relative availability of moisture and degree of deciduousness
of the vegetation in their home ranges. It is apparent that the
number of individuals associated with groups above and beyond the
mean varied directly with the degree of deciduousness in their home
range and inversely with the availability of moisture.

Median group stability was tested in the l97h dry season, when
each group was censused monthly for five months, and in the 1973
wet season. Median stability values for the 197M dry season are
listed in Table 1h. No values of median stability are presented for
the dry season since uneven, inadequate sampling precluded comparisons
among all groups. For the eight of ten wet season comparisons which
were possible, a separate index of median stability, based on common
census periods, was developed for each paired comparison in order
to correct for unbalanced sampling. Formulae for comparing the
median stabilities of groups are found in Hays (1963).

An a priori test to determine whether differences in median
stability occurred among the five groups during the 197k dry season
indicated that Significant differences did occur (:1:2 = 8.hhh, .1 < p >
.05, h d.f.). A posteriori paired comparisons indicated that two
groups, lowland and upland, exhibited significantly higher dry season
stabilities than yellow group (:2:2 = h.5h9, p < .05, 1 d.f., and
x2 = 5.519, p < .025, 1 d.f., respectively). All other comparisons
Were non-significant, with p > .1 or more for all comparisons.

Paired comparisons of group median stability during the wet season
indicated that the lowland group was significantly more stable than
‘Yellow group (x2 = 3.976, p < .05, l d.f.). Median stabilities among

(Ither groups did not differ, with p > .1 or more for all comparisons.

106

.m0. v a pm zoaaow mo Hosp Soak maomgww hpflafipmpm commom H03 cmwumfl

 

 

 

a
.mmo. v a pm Soaaow mo HmQH Bose mammme szHHpmHm commmm 3H6 cmeozm
.mo. v a we Soaamw Mo Hosp anw msommwo kHHHHQmHm Cowmmm haw cmflcmmm
.QHHHQ an HcmsHHSH00H 6005HQQHH
ovaHHHm
0: ~0.00H PH mam. H mem.0 Boaam» 0:

w

mooa 3m.m0H ma 00m. H :mm.0 wsmamb 0m m.

S

fi+

00 00.me 0 me. H msa.m 00m 0m m

o

oo 2.0. ms 6mg. H oooé 68.5 om fl

8

p

mooa 0>.mm 0H 6mm. H m:a.0 socmazoq 0H m.

w

asma commmm hag smoz 0>op< mamst>H0cH mm H x msou0 :woa :Ommmm

mHHHHpmHm cmemz 00H6H00m6< R Hwopmfloomm< I. hHQImmmsmsoscwomc mmaam>
Hsoohmm omHmEHHmm H0>Hm

 

.HcmemHm coflpmpowm>ImH5HmHoe m wooam mHH omsma 080: moon: mmsosm
Hmwoom swsmsmp 0>H0 How mamsofl>H02H UmHmHoommm mo Honesc one use mNHm msoam some 0o comHHmQEoo .JH manna

107

Stability among sex-age classes: The median test was also used
to test for differences in stability among sex—age classes, both for
the entire study period and the l97h dry season. No significant
differences in stability among sex-age classes were noted for either
period (x2 = h.h02, p > .1, 3 d.f., and x2 = .h83, p > .975, 3 d.f.,
respectively).

In a separate test, emigrations and disappearances for all Sex-age
classes over all counts were compared with the expected number of
emigrations and disappearances per class based on the proportion of
each sex-age class in the overall population (Table 15). The results,
x2 = 9.2h5, p < .05, 3 d.f., indicate that the expected numbers of
emigrations and disappearances per class differed signficantly from
the numbers observed. It can be seen from Table 15 that mature
tamarins of both sexes disappeared and emigrated somewhat less than
expected, while the immature tamarins emigrated and disappeared
more frequently than their numbers in the population would predict.

A comparison of the observed proportion of emigrations and disappearances
per age class with the expected proportions was made using the

Bonferroni Z statistic in order to determine whether or not the
significant Chi-square value could be attributed to extreme differences
between observed and expected values in one or more of the sex-age
classes (see Table 15). The results indicate that while the

differences between observed and expected proportions approached
significance (.2 < p > .1) for each comparison, the significant Chi-
square value was not attributable to significant differences in

any sex-age class.

108

 

mom.onma

 

6mmnmmw6mo. 66H. sso.a Hms.m 6 eos. mmHmSoH asapaaea
om: Mammoa. How. 6oo.s som.s o ems. moses msspaaaH
om: Mammoa. How. mos. mmm.aa o Hmm. mmHanH assess
Hm: MAWoeo. omm. oo6.H eso.ma 6 mos. mmama oHSSas
coHHHomoam oo>Hmmpo mosam> mocwhmommmmwa oosmpwommmmfio AcoHHHomoam mmmao ommemm
00>Hompo so mcoHHHomoam mamswmIH£0 0cm coHHmHmHem 0cm cofipmumfiem vopoomxmv

HsmHoHMHooo oopoomxm 60>Homno coHHmasmom

mocmvfimsoo cw noHHHomoam

zHHENm 000

 

mmmao ommemm no 00669 mosam> Umpommxm

.onEmm wopooaaoo map can mmsoaw oesopwcoe o>Hm map CH mooHHHomoam

“mommmao mwmemm moose moosmamommmch 0cm mcoflpmawfism

.mH mHQmB

109

Immature tamarins, then, tend to emigrate and disappear more
frequently than adults. This observation seems to be at odds with
the stability comparison, which revealed no differences in stability
among sex-age classes. This can be explained, however, by noting
that the prolonged dependence of the young on the group early in
life, which results in high stability scores, compensates for the
higher rates of emigration and disappearance which occur among older
immature tamarins.

Group losses by emigration or disappearance were recouped by
immigration and natality. The sex-age class proportions of 18
immigrants entering the monitored groups for the first time did not
differ Significantly from expected proportions bases on sex-age
class proportions in the collection (x2 = .992, p > .5, 3 d.f.). A
second source of recruitment, immigrating animals which had formerly
emigrated, was found to be substantial (13 animals). Emigrants were
found to reunite with their former groups in proportion to the
numbers of each sex-age class which were known to have emigrated
(x2 = 1.092, p > .5, 3 d.f.).

While emigration to adjacent groups was a common phenomenon (11
observations), long-distance emigrations were also observed. One
male about 18 months of age traveled 6.5 km from the home range of
his original group. Several tamarins, including one female of about
12 months of age, were known to have travelled at least 2 km from
their group of origin.

Home range data: An attempt was made to quantify both home
range size and utilization patterns for two groups, the lowland, on

the humid extreme of the moisture gradient, and the upland, on the

110

dry side. The lowland group occupied a home range of about 26 ha.
This home range was defended as a territory in the strictest sense -
a defended area with a distinct geographical boundary (Bates, 1970).
The approximately 13 percent of this range overlapped by three other
groups was the site of frequent intergroup contact and territorial
marking. The lowland group maintained the boundaries of its home
range throughout the year; in contrast, the boundaries of the upland
group's home range were flexible and were expanded in the late wet
season and the dry season. During the wet season, the upland group
was known to occupy an area of 32 ha; its home range during the dry
season encompassed at least h3 ha. The 32 ha portion of the home
range was not defended as a geographical entity - at least 83 percent
of it was overlapped by at least five other groups. The groups using
this area apparently did so through separation from other groups in
time and/or space. Although frequent aggressive encounters and
agonistic displays between groups occurred, both offense and defense
seemed directed toward defending the immediate area around the group
rather than toward defending a geographic territory per se.
Importance of group components: Perhaps the most important
observation to be made here is that practically no sexual dimorphism
exists in this Species. In Table 16 parameters of nine characters
which are commonly sexually dimorphic in primates are listed, and
the intersexual differences tested (pregnant females were excluded
from the samples). The only parameter which differed significantly
'between sexes was body weight, where females averaged h.3 percent
Iieavier than males. Mature males and females are roughly equal in

faize, weight, and dental armament, and are thus equally capable, in a

111

.mwma .600666 Boaaom mp6080asm60s 066660

 

 

.88 H0. 0H 06096360 00HQHOQI066006 06665 mp60E0Hdm60s mcwsoaaom 666 06.9m

.0003Hox0 6066600 H66600Hmn

.0 30. 6606606 op 0630603 H6606>666Hm

63 mo. H 60.6 0066806 6006069 063660 H63696066s p000

06. A a 636.0 03 mo. H 60.6 60668 6H060Hp 066660 H6HSQH666E H004

33 00. H 03.0 0066800 :Hmc0a 06Hc6o H6HSQH666E H003

o6. A a 666.3 63 oo. H 66.6 00306 666603 063666 H03opHooaa H603

33 mo. H mm.m 0036800 066660 666336368 .6666069 H6smcH3oHp6q

03. A m mam.m 03 m0. H 0m.m 60368 066660 666HHHx68 H£H060HQ H6smcwaoflp6q

63 >0. H m0.0 6036800 npwo0a 066660 666666368 H0mzoq

06. A a 0mm. 03 00. H 30.0 00365 3£H060H 06H66o 6H6HHH36B H0mcoq

03 Ha. H mm.0 0066800 npws0a 066660 6H6HHHx6E ARV 0>Hpo0mmm

ON. A Q 600.6 00 NH. H 63.0 00H6E npmc06 066660 666aafix6e AMV 0>Hpo0mem

a: 00.H H 3m.mm 0066800 3880 np060an oHH6Eo06N

06. A a 3M0. 03 0m.3 H 0m.mm 6036B 3660 6H060Hp oHH6somAN

33 00.6 H 60.00 0066800 3880 zpmc0a 33:30

on. A a 636. 63 6s.3 H 63.36 60366 6Aeav 6H6603 33:36

33 33.6 H 63.6om 0030606 6H36303 soom

o3. v a 666.3 mm 66.6 H 63.663 A6363 3H36303 6306

63 e3.6 H 66.336 0036600 3660 6H6603 3606

o6. A a 636.6 66 63.3 H 63.666 00366 3660 666603 soom
m m 6 mm .m

 

.wmo£%%omm .o .0 036500 666 0368

0H5H6a mo mpc0E0HSm608 H6Hc00 066 H6HQ6H0 00Ho0a0m 066 .npms0a hcon .930603 6609 mo somHH6mEo0 .06 06969

112

physical sense, of overt participation in aggressive encounters,
both inter and intragroup, and in predator defense.

On the basis of morphological similarities, one might expect a
more equal apportionment of roles between sexes than is found in
most primate societies. My field observations bear this out. Both
males and females participated actively in the care of infants, a
fact which is already well substantiated in the literature (e.g.
Hampton, l96h; Shadle et aZ., 1965; Epple, 1967a, 1970a; Weber, 1972).
Both sexes were observed to actively participate in territorial
defense. In the majority of instances, this consisted of uttering
"long whistles" (Moynihan, 1970) at the other group, with perhaps a
few rushes toward the opposing group being made by individuals of
either group. However, six intergroup encounters were observed where
opponents made physical contact; of these, three took place near
artificial feeding sites and three took place in natural situations.
Males tended to be more aggressive in the six altercations involving
contact. Both sexes were seen to mark vegetation with their supra—
pubic and circumgenital glands following both the aggressive encounters
and those involving only agonistic display. Females appeared to mark
more frequently than males, and to initiate marking behavior following
an encounter. Both sexes exhibit behavior indicative of "control"
animals in the sense of Bernstein (196b, 1966).

The absence of pronounced sexual dimorphism and the duality of
roles noted above might lead one to suspect the existence of a social
structure with a one male, one female dominance system such as the
one described by Epple (1967a, 1967b, 1972) for laboratory groups

of Callithrix jacchus. Observations at feeding platforms during the

113

dry season indicated that certain males and females occasionally
dominated members of their own sex in competition for food. I also
observed that a single female per group had priority to reproduction -
i.e. only one infant or set of twins was found in any one group at a
given time, although the group might contain two or more sexually
mature females. Epple (1970a), who apparently observed the same
phenomenon in laboratory groups of S. 0. geoffroyi, indicated

that only alpha females bore young.

A further test of Epple's hypothesis that one and only one alpha
(reproductive) female occurs per group was made by comparing the
expected number of reproductive females (18.9/131) based on 1 female
per average group of 6.93 animals, with the observed number found in
the collected sample (25/131). A Chi—square test yielded a value of
x2 = 1.820, p > .1, 1 d.f., which indicates that the hypothesized
number of reproductive females did not differ significantly from the
number observed.

Reproductive females appear to be older than nonreproductive
females as evidenced by the criteria of a higher frequency of scarring,
eSpecially on the ears and the face, and the high incidence of damaged
canine teeth (7/22 vs. 1/26, x2 = 5.60M, p < .025, 1 d.f., and 8/22 vs.
3/26, x2 = h.h33, p < .05, 1 d.f., respectively). Hershkovitz (1970)
observed that dental caries and damaged teeth occurred only in the
older marmoset age classes. The incidence of a highly developed
suprapubic gland was higher in reproductive females than in nonrepro—
ductive females (21/22 vs. 6/26, 62 = 11.112, p < .001, 1 d.f.).

Males did not differ from females in regard to the number of torn

ears and facial Scars (12/52 vs. 8/h8, x2 = .512, p > .1, 1 d.f.) or

11h

broken canines (16/52 vs. 11/h8, x2 = .570, p > .1, l d.f.). When
combined with observational data, these findings offer further support
for the coexistence of a single dominant male per group and also the

lack of sex-specific mortality.

Discussion: Group Composition and Stability

Carpenter (1952, l95h) initiated the synthesis of primate studies
toward the end of presenting an integrated classification system for
primate social structures. More recent reviews (e.g. Crook and
Gartlan, 1966; Bernstein, 1970) attempted to order the various primate
taxa according to somewhat arbitrary evolutionary grades based on
social "complexity." Eisenberg et al. (1972) reassessed this system
and developed a more workable system which relates social structure
with the ecology of a given Species. The following discussion has as
its goal the redefinition of the position of S. 0. geoffroyi on this
continuum. Group size and structure of the Panamanian tamarins will
be discussed in relation to habitat, diet, and methods of communication.
The transitory nature of some group constituents, and its implications,
will also be discussed. A new category of social organization for
primates, the age-dependent, one-male, one-female dominance system
will be presented. It is Suggested that other Species of Saguinus and
other callitrichids occupy similar positions, although further studies
are needed to confirm this hypothesis.

Epple (1972), in summarizing the accounts from older literature,
reported that callitrichid group size varied from 2—12 animals.
Thorington (1968b) observed groups of Saguinus midas to vary from

2—6 individuals, and reported that other researchers in Surinam had

115

seen groups of S. midas numbering up to 20 animals. Hladik and

Hladik (1969) and Moynihan (1970), working in Panama with S. 0. geoffroyi,
found groups numbering 2—9 tamarins, with single individuals also in
evidence. The last author considered an account of group sizes of

up to 12 animals (Chapman, 1929) to be grossly inaccurate.

My own observations of 71 groups yielded a mean of 6.93 i .36
animals per group, with a range from 1 to 19. The few single animals
observed were almost always seen to join a group if watched for a
half hour or more. The two largest groups, one of lb and one of 19,
were observed only once each at localized food sources, and thus
may represent casual groupings of tamarins rather than permanent
social groups. Most groups observed during the study ranged from
3—9 animals.

The mean Size of specific social groups is determined by a
multitude of factors: e.g. phylogenetic history, population density,
diet, foraging system, spatial and temporal distribution of resources
in the environment, relative distribution and the magnitude of
population depressants such as predation and parasitism, and the
recent history of epizootics (Crook and Gartlan, 1966; Bernstein,
1970; Eisenberg et al., 1972). It would be pointless to pursue a
discussion of Saguinus group size as determined by these factors,
since so many of them remain unqualified. However, some general
comments can be made.

For the most part, callitrichids inhabit "edge" vegetation in
the broadest sense, which can generally be considered dense vegetation
offering low levels of visibility (Thorington, 1968b; Moynihan, 1970;

Coimbra-Filho and Mittermeier, 1973). Under these conditions, it is

116

not surprising that the visual system of communication is rather
primitive and rudimentary (Moynihan, 1970), and that the auditory
and olfactory systems are highly developed (Epple, 1967a, 1967b, 1968,
1970a, 1972; Moynihan, 1970). High frequency vocalizations, which
form the bulk of the acoustical repertoire (Epple, 1968; Moynihan,
1970), have the probable advantage of being absorbed rapidly by the
foliage, thus enabling the passage of Specific information to group
conspecifics at short range, without attracting predators, of which
there are many, from a distance (Marler, 1965; Moynihan, 1967, 1970).
As noted previously, tamarins, particularly Saguinus, are highly
insectivorous and frugivorous. The fruits most often used are small,
and their seeds are generally dispersed by birds and small mammals.
These small fruits are more uniformly distributed over time and Space
than are the large, fleshy fruits (Smythe, 1970b), which are the
mainstay of larger primate frugivores. The callitrichid food sources
are more or less uniformly distributed rather than clumped.
Observations from this study indicated that little, if any,
cooperation occurs in the capture of individual food items, although
at times an insect flushed by one tamarin might be captured and eaten
by another. Of more significance is the observation that individual
group members act in a sentinel capacity, warning those engaged in
feeding of the approach of an aerial or terrestrial predator.
Increased probabilities of individual survival through group cooperation
in predator detection, and through the "mobbing" of predators by the
group are probably the primary advantages accrued to individuals in
the Saguinus social system. The number of "alarm" vocalizations in

the acoustical repertoire of these small, vulnerable primates (Epple, 1968;

117

Moynihan, 1970), and the high frequency with which they are uttered

at a variety of stimuli on any given day seem to support this view.
Mobbing of natural predators was not observed. However, tamarins often
rushed toward the investigator while he was removing live—trapped
individuals from the trees in which they were captured. These rushes,
which were both vigorous and accompanied by intense vocalizations,
sometimes ended only a few feet from the observer. It thus seems
probable that tamarins would mob a smaller predator. Other possible
advantages of group behavior, such as thermoregulatory benefits which
may be gained in nocturnal huddling behavior, should not be ignored.

When taken together, the limited range of intragroup communication,
the relatively uniform distribution of food resources within a given
season, and the space required by foragers in a non-cooperative,
individually-oriented foraging system may also militate against the
formation of large social groups. The high levels of aggression
exhibited by group dominants (Epple, 1967a, 1970b, 1972) and terri-
torial behavior noted by Thorington (1968b) and Moynihan (1970), are
probably instrumental in maintaining both individual and group
foraging space required in the efficient exploitation of dispersed,
low density food resources.

Carpenter (1952, l95h) observed that troop size tended to be
standard for a given Species, and that sex-age ratios also appeared
to be stable. Eisenberg et al. (1972) indicated that habitat
differences cause widespread variation in intraspecific social
structure within some taxa and not others. One may infer from their
discussions that, within a given habitat at a given point in time,

the factors governing modal group size are fairly constant, although

118

the habitat type itself may be quite "patchy" (as defined by MacArthur
and Levins, 196M). This appeared to be true in this study where,
although home range sites varied greatly in regard to the availability
of food and moisture during the dry season, group size and group
sex-age ratios remained relatively constant.

The mechanism necessary for the maintenance of stability in both
group numbers and sex-age class proportions is obvious - immigration
and natality must compensate for losses (emigration and mortality)
in proportion to the loss in each sex-age class. Natal recruitment
alone is insufficient to compensate for these losses. The data
demonstrate that these vacancies are indeed filled, and that
immigration is the primary mechanism.

The source of immigrating animals appears to be from other
social groups rather than a population of extra-group individuals.
Evidence for this is partly negative - single animals and small
groups were almost always associated with larger groups and rarely
remained isolated for extended periods. Moreover, baited feeding
platforms were never visited by single animals. In contrast, ten
immigrants, roughly one half of the total, were marked animals from
other groups. It might be surmized that immigrants come from two
sources: 1) Groups which maintain their integrity while exhibiting
a turnover of constituents, and 2) groups which for some reason prove
to be nonviable and thus dissipate. The existence of the latter
source is necessary if one is to explain the maintenance of group
size with no existing extra—group population as a source of replace-

ments.

119

Four of the five groups each contained a reproductive female
which remained with its respective group for the course of the study.
Epple (1970a), working with S. 0. geoffroyi, observed that only alpha
females bore young, although "inferior" but sexually mature females
copulated with the fertile male (alpha male). It appears likely, then,
that the reproductive females in these free-ranging groups are alpha
animals. In addition, at least one adult male remained a permanent
member of each group throughout the study. In three of five groups
the resident males were observed to supplant other associated males
at feeding platforms and in some natural feeding situations. They
were also observed to "lead" in chasing an intruding group from the
feeding platform or territory. Evidence exists,tflunl,for a stable
group core of one reproductive or alpha female and one resident,
probably dominant, male. The other stable element of the group is
the dependent young, usually the survivor of twins, which does not
leave the group for extended periods until it is roughly a year or
more of age. In light of their temporary existence, the remainder
of the group constituents may be considered transients.

The above observations, when combined with those in the prior
section regarding the relative roles of male and female tamarins,
suggest that the typical S. 0. geoffroyi group consists of an alpha
male, an alpha female, sedentary young of the year, and a transient
complement of presumably subordinate animals of both sexes, which
appear to be younger than the alpha female, and probably the alpha
male as well. These subordinate animals may or may not be sexually
mature (see section on reproduction). Using terminology parallel to

that used by Eisenberg et al. (1972), this configuration may be

120

considered an age-related, male-female troop. The existence of the
dual dominance system is plausible given the lack of sexual dimorphism
and the duality of roles discussed previously. The statistically
equivalent rates of intergroup movement by males and females add
credibility to the concept - viz. unisexual dominance systems should
generate higher rates of movement, and perhaps extra—group individuals,
of the dominated sex.

As noted previously, the existence of the dual dominance system
in free-ranging callitrichids was predicted by Epple (1967a, 1967b,
1970a, 1970b, 1972) on the basis of laboratory observations in which
Callithrix jacchus was the primary model. She garnered supportive
evidence from less intensive observations of other callitrichids,
including 3. 0. geoffroyi (which she referred to as Oedipomidas Spixi,
after Hill, 1957). She thought of this system as existing within a
family group consisting of one or two adult pairs and their offspring
of successive years. This was in keeping with the traditional concept
regarding the social structure of the Callitrichidae, and was
strengthened by her observation that, while dual dominance systems
did form in groups of unrelated C. jacchus and Saguinus fascicollis
as well, they were rarely stable over long periods of time, whereas
family groups remained peaceful over the three-year study. The social
configuration which I observed appeared to be identical to those in
Epple's "artificial" i.8. unrelated groups.

Epple (1967a, 1967b) indicated that the alpha animals "tolerated"
juveniles. A "tolerance" for juveniles was also observed in an
artificial feeding situation by Nelson (1975). In the current study

a juvenile female approximately 12 months of age was tagged in the

121

upland group in June, 1973. During the two subsequent weeks She was
alternately a member of the upland group and adjacent Yellow group.
She then disappeared until August, when she was seen with another
group hOO m south of the upland group's home range. In September She
became a member of Green group, whose home range was located 2 km from
that of the upland group. It would appear then, that juveniles may
move from group to group with a minimum of interference.

The integration of new group members, and the maintenance of
dominance in the group may at times, however, require overt physical
encounters. Epple (1972) observed that the introduction of a strange
conSpecific into the laboratory group precipitated aggressive behavior,
particularly on the part of the alpha animal of the same sex. She
also observed that serious fights between alpha animals and their
subordinates occurred frequently, and often after the group had
been stable for a year or more. In the current study, the "long rasps"
(Moynihan, 1970) which accompany aggressive contact were heard
frequently, although it was often difficult to determine whether the
conflicts were intra- or intergroup in nature. The field observations
of Moynihan (1970) indicate that intragroup altercations are more
frequent, and, as evidenced by the vocalizations, the more intense.

It is possible that high levels of intragroup aggression are the
driving force behind the intergroup flow of transient animals.

A group social structure composed of a single dominant pair and
a Shifting complement of subordinates would have several probable
advantages over the traditionally-accepted "extended family" group
with an identical dominance system:

1. Gene flow in the population would be greatly increased since

individuals from other groups, rather than siblings, would ascend

122

to dominant, i.8. breeding positions. This would limit genetic
drift and assist in maintaining the integrity of the species;
Immigrants ascending to breeding positions in an open social
system would contribute a more varied genotype (i.e. heterozygous
at more loci), tested over a wider variety of ecological condi-
tions, to the offspring. It follows that offspring exhibiting more
variable, more broadly-tested genotypes would be genetically

more "fit" for survival under a wider variety of environmental
conditions than offspring from isolated social groups subject to
a limited range of selective pressures. This consideration is
especially important Since callitrichids in general are
r-selected (as defined by Pianka, 1970) relative to other primate
species, and may be thought of as colonizing forms which inhabit
widely-scattered, temporally-discrete areas of second growth

over a wide range of environmental conditions;

Replacement of individuals lost to mortality would be more rapid,
thus facilitating the maintenance of group Size at optimal levels;
The higher incidence of intragroup aggression in non-family
groups would favor selection of more aggressive alpha animals.
(It might well be argued to the contrary that high levels of
aggression might not be adaptive Since they might prove to be
socially disruptive. However, Bernstein and Gordon (l97h)
suggested that "Rather than disrupting social relationships,
aggression serves to enforce regulated Social interactions which
maintain primate societies"). It ispmflsible that tamarins which
are highly aggressive in intragroup conflicts might also prove

highly aggressive in intergroup conflicts. The capacity for

123

high levels of intergroup aggression would be adaptive in

maintaining the large foraging territory necessary to exploit

dispersed food resources; and

5. Long distance movements by transient individuals, and high levels
of individual aggression fostered in non-family groups might aid
in the discovery and colonization of small, widely-separated
areas of second growth such as those generated by slash and

burn agriculture.

Gartlan (1968) and Eisenberg et al. (1972) discussed recent
literature in which intraspecific group structure and function were
found to vary over environmental gradients. Jolly (1972) found that
lemur groups which normally defended fixed territories abandoned these
territories and adopted a system of temporal and Spatial segregation
during a severe dry season. She concluded that "the change may
reflect long-term population effects or seasonal food shortage."
Social structure, and Spatial utilization, then, may vary across
both environmental gradients and time, and may even vary within groups
over variable environmental conditions.

Anderson (1970), in a consideration of ecological structure and
gene flow in Small mammal populations, divided potentially-suitable
habitats into colonization habitats and survival habitats. In the
former, populations capitalized on abundant, temporally discrete
resources; the the latter, segments of the population survived in
areas which were relatively salubrious at all times. Time—space
segregation afforded the modus operandi for exploitation of resources
in the colonization habitat, while strict territoriality proved to

be the energetically feasible alternative in the stable survival habitat.

12h

In Rodman, tamarins occupying lowland areas, which provide a
relatively stable food resource throughout the dry season, may be
thought of as occupying survival habitats. Their well—defined
territories, relatively stable group compositions, and higher rates
of infant survival support this contention. Those tamarins occupying
upland sites utilize a seasonal resource, and may be thought of as
inhabiting colonization habitats. The observed segregation of groups
in space and time, the apparent inability of these groups to establish
and defend a geographically—fixed territory, and the low rates of
infant survival are in agreement with Anderson's (1970) observations
on colonizing animals.

At this point, it is necessary to note that a fundamental
difference exists between the situation in Anderson's (1970) study
and that involving 3. 0. geoffroyi in Panama. Anderson (1970) found
it necessary to invoke group selection in explaining the genetic
differences found between survival and colonization components of
the house mouse, Mus, since the colonizing element did not contribute
genetic information to the survival populations (i.e. there was no
direct genetic feedback). As indicated earlier, the colonization
element of S. 0. geoffroyi does have a feedback mechanism - immigration
to survival habitats and the rise to sexual dominance by individuals
from colonization habitats. Thus the S. 0. geoffroyi social system
allows the transmission of genetic information in the traditional
Darwinian manner.

Anderson (1970) characterized the function of the survival component
of the population as "maintaining the ongoing population in both the

ecological and genetical senses", and the function of the colonization

125

component as "a test of the full range of ecological and genetic
variation under all of the accessible variations of the fundamental
niche of the species", while at the same time providing for the
dissapation of excess animals. It is Suggested that populations of

S. 0. geoffroyi follow this pattern, and that the accrued advantages
are responsible for their success under environmental conditions which

are less than equitable.

Results: Reproduction

A distinct birth peak was evident from late April through early
June. Eighteen of the twenty-six groups observed with carried young
were seen during this time period. Of the remaining eight, five were
seen in March, and single groups with carried young were seen in
February, July, and October. Six of the seven well—developed
foetuses from the collected sample were from the months of February,
March, and April.

Reproductive activity was not, however, limited to the period
preceding the annual birth peak. Pregnant females were found in
the collected sample from September through May, newly-born young
were observed in June and July, and one infant was collected that
had been born in August. Moreover, two females collected in April
and two females collected in May were found to be in the early stages
of pregnancy. This observation, together with the presence of
well-developed foetuses in a female collected in September, point
to the possibility of a potentially bimodal birth peak, with young
being born in April and May, and from September through October. One

social group containing infants born in August, and another set of

126

twins which had been born earlier, presumably in March, were actually
observed. It appears, however, that the reproductive potential
evident from the pregnancies in April and May is rarely realized,

for the Sightings of infants outside the March to July season of
parturition are rare.

Some evidence also exists for a postpartum estrous, or at least
a period of estrous relatively soon after parturition. Three lactating
females collected during this study were also pregnant and in two
of the three, these embryos were well-developed. Also, as indicated
above, one group (with one reproductive female) contained two sets
of twins which were separated in age by only five or Six months.

If there is a potentially bimodal birth—peak, as the data suggest,
the effect of lactation anestrous, if it exists at all, must be
relatively Slight.

A trimestral comparison of somatic and reproductive tract
measurements for mature males is presented in Table 17. No differences
in the Size or weight of the reproductive organs were observed.
However, body weight differed markedly over trimesters (F = 3.8h3; 2,
52 d.f., p = .028). Post-data comparison of the trimestral means
for body weight using the Student-NeWman—Keulstuna;(Sokal and Rohlf,
1969) indicates that mature males weighed significantly more during
the July to October trimester than during either of the other tri-
mesters (p < .05). The average weights of mature males did not
differ between the November-February trimester and the March-June
trimester (p > .05). Gametogenic tissue was evident in the histo-
logical examinations of all mature males collected from April to
December. Testes of males collected from January to April were not

examined.

127

.309o300I3360 .m 066 .0660I6o362 .0 .63663900I309E0>oz .3 ”36600 6309006330

.06 63 6960303 6336093
.66 63 636060360606 360633 33.36

.0 3.0 3003606 op 3603030
H

 

 

636. 36o.o 66 33.6663 63.6366 H3630: HHHmme 33636
366. 63o.o 66 63.3 o3.6 3360033 63H60H 36636
006. 360.0 00 00.3 36.6 63633 039009 99030
003. 000.0 00 00.3 00.33 930603 033003 36030
336. 366.o 36 36.6333 63.3336 H3630: mHHmmH 3603
663. 336.o 36 66.3 63.6 Hoamaa 636003 3303
000. 000.3 30 00.3 00.6 63633 633003 9003
633. 636.6 36 66.3 63.33 3H6603 mHHmaH H603
o36. 3o3.3 63 63.o 36.6 sOHHHoaoo 66636 oHssaasasm
366. 636.3 63 6o.6 6o.o6 neoHs 66636 6366663666
000. 333.0 03 60.0 30.30 930603 06630 0396063030
366. o63.o 66 66 o3 63.366 63H6603 3606
000. 030.0 00 00.00 00.303 0990303 mwom
OHHnHHaam 3 Ho oHHaHHaam 3 a 36 .m

03333969030 006603036030

 

3.0303006333 30>o 036080360606

mo 6063360600 6 9333 «wmoxxkomm .o .m 0366 035966 mo 636060366606 90633 0>3poswo3m03 036 o3H6Eom .63 03969

128

The comparison of somatic and reproductive measurements for
mature females over trimesters is found in Table 18. Again, no
significant differences in reproductive tract measurements were
observed among trimesters. However, two measurements, body weight
and the width of the suprapubic gland, did differ over trimesters
(F = 2.h6h; 2, 37 d.f., p = .099; and F = 3.752; 2, 39 d.f., p = 0.32,
respectively). A post—data comparison of means as described above
indicated that mature females were significantly heavier during the
July to October trimester (p < .05). Weights did not differ between
the other two trimesters (p > .05). The mean width of the suprapubic
gland was significantly greater in the March-June trimester (p < .01).
No significant differences in suprapubic gland width existed between
the other two trimesters (p > .05).

Reproductive females appear to be older than nonreproductive
females as evidenced by the criteria of a higher frequency of scarring,
especially on the ears and the face, and the higher evidence of
damaged canine teeth (7/22 vs. l/26, x2 = 5.60h, p < .025, l d.f., and
8/22 vs. 3/26, x2 = h.h33, p < .05, l d.f., respectively). The evidence
of a highly developed suprapubic gland was also more prevalent in
reproductive females (.62 = 11.112, p < .001, l d.f.).

The mean number of newborn infants per group (per female) was
2.00 : .hO S.D. (n = 26). Although social groups usually contained at
least two adult females, only one mature female per group was observed
to bear young. The number of reproductive females in the collected
sample did not differ significantly from the expected number, based
on the expectation of one reproductive female per average group size

of 6.93 tamarins (x2 = 1.820, p > .1, l d.f.).

129

.88 63 096080360608 360633 .08 63 960303 .00063030 0036806 96660036

.08 63 0960303 636>0
.88 63 096080360608 360633 3303
.00063030 0036806 96660036 .08630 63 960303 60cm
.309O900I6360 .0 066 .066wa6036z .0 .63663906130980>oz .3

"66600 0309008336

300810

 

 

363. 006.0 00 30.0 00.0 069600 063090
366. 666.3 66 63.6 06.6 660.63: 063096
663. 636.3 36 33.6366 66.6066 6H3630> 0:30H6
663. 666.0 33 66.666 63.666 H3030... .696 H308
000. 000.3 33 00.0 03.0 6906039 636>o 96030
366. 366.0 33 63.3 33.6 3H03: 330>o H3636
360. 000.0 33 03.3 00.6 690606 636>o 96030
663. 666.0 63 63.663 66.666 H3630: 330>o H003
063. 366.0 63 66.6 66.6 mH60033 330>o H003
333. 666.0 63 66.6 63.6 3H03> 330>o H303
630. 000.0 03 00.6 30.6 690603 636>o 9603
003. 006.0 03 03.0 60.0 6039306oo 06630 0396663660
660. 666.6 63 66.3 66.66 3H03: 03036 0333603660
663. 663.6 63 63.6 63.03 3H6803 08036 0336603630
603. 600.0 63 00.03 00.000 0690603 6000
660. 363.6 03 66.36 06.666 6H3030: 3006
03H03H0H0 3 00 03H03H0H0 3 a 30 .m

3H33330po36 003003630030

 

3

.0309008339 30>o 096080360608 60
6O033668oo 6 6933 «mmofifi6000 .o .0 036806 036968 mo 096080360608 90639 0>39060o3603 066 03968o0

.03 0396B

130

The mortality rate for infant tamarins is quite high -- about
50 percent or more during the first six months of life. Most groups
are successful in raising only one infant. It also appears that the
site of the group's home range, eSpecially under severe climatic
conditions, may affect survival. In this study the lowland group
and the surrounding three lowland groups were all successful in
raising both of their twins until they were about 3 months of age.
Two groups lost an infant at h months of age. The others survived
until at least 6 months of age. In contrast, Red group, which
occupied an intermediately moist range, lost its two young within
two months after birth, and in the seasonally dry upland habitat
surrounding the home range of the upland group, two groups lost
their twin infants within the first two months after birth, and two

groups apparently did not produce young.

Discussion: Reproduction

Abundant evidence exists for a definite S. 0. geoffroyi breeding
season in central Panama. Wislocki (1930b, 1939) found small embryos
in the uteri of females in January and early February. Given a
gestation period of about lhO-lSO days (Epple, 1970a), these tamarins
would have been born in April, May, and June. Enders (1935) observed
that a peak season of parturition occurred during these months, and
Moynihan (1970) observed that large numbers of infants appeared
"quite suddenly" in the Panama City market in May. Fleming (1973)
stated, on the basis of Moynihan's (1970) work, that S. 0. geoffroyi
was seasonally monoestrous, with a breeding peak in January and a
peak season of parturition in the early wet season. While the dates

are accurate, Moynihan (1970) did not suggest that this tamarin is

131

seasonally monoestrous, although the data themselves are suggestive.
Information gathered during this study also offer support for the
existence of a birth peak. Epple (1970a), in her study of captive

S. 0. geoffroyi, found that births occurred in March, April, May,

and August. Her observations suggest that these tamarins are seasonally
polyestrous or that a postpartum estrous, or a period of estrous
beginning shortly after parturition, occurs. She also observed
indications of postpartum estrous in two of three full—term deliveries.
Hampton and Hampton (1965) working with the closely related S. 0.
oedipus under captive conditions, observed birth peaks in spring and
autumn.

With such a body of evidence pointing to a seasonal birth peak,
the presence of pregnant females throughout the year, and the data
indicating the possibility of a second birth peak sometime in August
or September, are rather surprising. The existence of the primary
birth peak in April-June is easily understood since it is correlated
with the flush of insects and fruits which occurs at the beginning
of the wet season. However, pregnancies leading to the birth of
young at other times of the year appear, on the surface, to be wasted
reproductive effort, since young born during these less favorable
times of the year would be unlikely to survive the rigors of food
scarcity in the late wet season and the dry season. Such a strategy
would, however, be advantageous during exceptionally favorable years
and in the colonization of new habitats. The adaptive advantages
of extra birth-peak young under these circumstances may outweigh
their apparently maladaptive disadvantages in seasonal habitats at

carrying capacity.

132

The actual fate of most young which are conceived outside the
November-February breeding peak leading to the March-June birth peak
is unknown. Their scarcity in groups observed after the birth peak
implies that either very few are conceived or that those which are
conceived are either resorbed, aborted, or must die shortly after
birth. In light of the frequency of these extra breeding season
pregnancies seen in the collected sample, the latter possibilities
are the more likely of the alternatives.

The scarcity of newborn young in August and September is also
surprising when one considers the small embryos found in mature
females collected during April and May. In August and September
insect foods and many of the smaller fruits are fairly abundant. Thus
the failure of these pregnancies to contribute young to the population
must not be due to inadequate diet for either mother or offspring;
but must rather be due to other factors, the majority of which are
probably density dependent and behavioral in nature.

The mechanism through which tamarin reproduction is attuned to
the seasonally variable environment is unknown. However, one could
state unequivocably that the reproductive synchrony responsible for
the April-June birth peak is achieved through the perception of
proximal environmental stimuli. These must be highly correlated with
the ultimate factors that govern the success of reproduction over
long periods of time. As Lancaster and Lee (1965) point out, it
has been repeatedly demonstrated that "internal rhythm by itself
cannot maintain a regular periodicity of birth seasons, but must be
triggered and synchronized by stimuli external to the animals." For

the sake of convenience the discussion of a possible mechanism for

133

achieving the seasonal effect will be divided into two segments: 1)
The various proximal environmental cues which exist, and the probability
of their perception by the tamarins, and 2) the actual mechanics of
reproduction as suggested by the analysis of reproductive tracts and
additional data from the literature.

In a discussion of the factors affecting the physiology of
reproduction, Amoroso and Marshall (1960) wrote, "Of the external,
or environmental factors, the most important have been found to be
light, temperature, and food supply." Changes in humidity and
precipitation which presage seasonal peaks of food abundance may be
added to this list (8.9. Koford, 1965; DuMond and Hutchinson, 1967;
Rosenblum, 1968; Vandenbergh and Vessey, 1968; Rudran, 1973; Harrison
and Dukelow, 1973). Fleming (1973) discusses the correlative
significance of these factors in the reproduction of Panamanian
mammals. In the discussion which follows, each of these factors
and their probable effects on the reproduction of S. 0. geoffroyi,
will be discussed separately.

As Fleming (1973) indicates, differences in photoperiod at this
latitude are slight -- about 63 minutes' difference exists between
the longest and shortest days of the year. Because of the barely
perceptible changes in day length which occur over the year, he
discounts changes in the length of photoperiod as proximal environ—
mental stimuli affecting reproduction in Panamanian mammals. Koford
(1965), in his study of the exotic rhesus monkeys (Macaca mulatta) on
Cayo Santiago Island, which lies far to the north of Panama and
experiences daylight changes of about one-half minute per day, also
discounts light as a factor affecting the reproduction of these tropical

primates.

13h

Spinage (1973), in discussing the role of photoperiodism in
the reproduction of tropical ungulates, stated that daylight cannot
be so easily discarded as a potential environmental cue which triggers
reproduction. He observed that while slight changes in daylength
might indeed by imperceptible, the sudden shift in the timing of
sunrise and sunset which occur at the summer and winter solstices
might not be. He observed that even on the equator, where daylength
is constant to within two minutes throughout the year, a regular shift
of sunset and sunrise times of about 25 minutes occurs over a two
month period at the solstice. He also found that several tropical
ungulates appeared to utilize photoperiod, rather than the traditionally
accepted stimuli of food and water availability and forage quality, as
the proximal environmental stimulus in synchronizing reproductive
activity.

The sunrise—sunset shift is of greater magnitude in Panama —-
about AT minutes over two months' time (Figure 8). A comparison of
the reproductive data in Fleming (1973) for Panamanian mammals with
sunset-sunrise times shows that the breeding seasons of many of these
mammals coincoide with the shift in sunrise-sunset times. The peak
effective S. 0. geoffroyi breeding activity, based on the parturition
peak minus the lhO-lSO day gestation period (Epple, 1970a), coincides
with this shift and the month immediately following it.

This correlation may be fortuitous -- it may be that the climatic
changes from wet season to dry season may be the proximal cue, or
that reproduction is attuned to quantitative or qualitative changes
in the diet which accompany this seasonal change. Perhaps many of

the plants, which are themselves synchronized for reproduction in the

Figure 8.

Shifts in the time of sunrise and sunset at Balboa,
Panama Canal Zone.

Figure 8.

(hours)

time

0554

0550 ‘

0004 ‘

0609 ‘

0514

0024 ‘

0020

0034

0044
1754

1001 i

1011 ‘

1021

1020

1031 ‘

1030 ‘

1041 ‘

L

L

135

sunrise

/peak breeding
activity

 

 

SLHWSHBt

vpeakv breTadingv activity

Y

 

3:4

LA

136

dry season (Janzen, 1967), use this shift as an environmental cue
for the initiation of their reproductive efforts, and the tamarins
might use the condition of the plants as their environmental cue.
Plants, whose peculiar estrogens may influence mammal reproduction
(e.g. Negus and Pinter, 1966), may respond to photoperiodic stimuli
which are too subtle to be sensed by mammals.

Temperature may be dismissed as a factor which influences the
timing of reproduction. While the mean monthly temperature does reach
a low in September and October, and begins to climb during the time
that peak tamarin breeding occurs, the differences in mean temperature
are so slight as to be imperceptible, and the variation among days
of the month, and over years, make temperature changes, at best, a
highly undependable cue.

The popularly accepted theory regarding the onset of breeding
activity in central Panama is that reproductive activity is triggered
by the drop in precipitation and humidity which mark the beginning
of the dry season (see Figure 1). For many mammals (see Fleming, 1973),
the correlation between the onset of the dry season and the onset
of breeding activity is a good one. This correlation, however, does
not hold for S. 0. geoffroyi. While Moynihan (1970), on the basis of
his observations and Wislocki's (1930b, 1939) data, concluded that
most copulations, or effective copulations occurred in January (the
first full month of the dry season), his observations, and his
interpretation of Wislocki's data, seem in error. Given the lhO-ISO
day gestation period, young conceived on the first day of January would
not be born until the 20th of May at the earliest, and most of the

young would not be born until June, after the April-May birth peak

137

which Moynihan (1970) observed. Tamarins born in April and May, and
the advanced foetuses in Wislocki's (1930b, 1932, 1939) January and
February collections, were conceived in November and December. Those
born in March had to have been conceived in October, or even earlier.
Wislocki (1939) did, however, encounter pregnant females in his June
sample, and these foetuses were, as he suggested, conceived in January
and February. From this segment of his sample he incorrectly inferred
that the peak breeding season occurred in January and February. This
statement was accepted by Moynihan, apparently without reference to
Wislocki's other data, or his own good data. The matings which lead
to the April-June birth peak, rather than beginning with the onset

of the dry season, begin in the wet months of late October, November,
and early December, and extend through the beginning of the dry season
in late December, January, and February. It appears that the stimulus
or stimuli which release the physiological changes and associated
behaviors leading to the birth peak antedate the coming of the dry
season by at least a month.

Thus, the stimulus, or stimuli, which control the timing of
tamarin reproduction, remain unknown. The changes in precipitation
and humidity which occur in the transition from the wet to dry season
occur after the peak breeding season is well underway, and thus do not
trigger tamarin breeding activity, as had been supposed (e.g. Moynihan,
1970). Photoperiodic changes, the presence or absence of certain plant
foods, the possible existence of plant estrogens which might effect
tamarin breeding behavior, and perhaps the low levels of available
insects and fruit during the last two months of the wet season and
during the dry season, are potential environmental cues which merit

further investigation.

138

Clark (1972) and Chapman (1972) discuss the phenomenon of
seasonal breeding in mammals. As the latter author indicates,
mammals can be classified into two groups on the basis of the
reproductive cycle exhibited by the males: 1) Those species that
are fecund and breeding throughout the year, and 2) those species
which are fecund during some seasons and sexually quiescent in others.
The seasonally breeding species may be further differentiated into
two categories —- those which exhibit morphological changes in the
testes, whether macroscopic or histological, and those that do not.

Most primate Species inhabiting the climatically seasonal Middle
American tropics show indications of seasonal breeding e.g., Aotus
trivirgatus (Wislocki, 1930b); Cebus capucinus (Oppenheimer, 1967);
Ateles fhsciceps and A. geoffroyi (Eisenberg, 1973); Saimiri sciureus
(DuMond and Hutchinson, 1967); and Saguinus oedtpus (Moynihan, 1970,
and this study). Detailed reproductive data are available only for
the Saimiri.

Male Saimiri undergo an annual testicular cycle which is
accompanied by morphological changes attributable to changes in
testosterone levels (DuMond and Hutchinson, 1967). Breeding is highly
correlated with low levels of precipitation and humidity (DuMond and
Hutchinson, 1967; DuMond, 1968; Rosenblum, 1972). Harrison and
Dukelow (1973) found that female Saimiri on an ovulation—inducing
hormone regime showed optimal breeding response only at low humidities.
Thus the Saimiri, which are small insectivore-frugivores occupying a
niche which is very similar to that of Saguinus (Eisenberg et al., 1972),
offer 21 classic example of reproductive cyclicity accompanied by

readily visible morphological changes.

139

In contrast, no apparent seasonal changes in reproductive tract
morphology or measurements are evident in the seasonally reproductive
S. 0. geoffroyi. Testis size remains relatively constant, and the
testes appear to produce sperm throughout the year. However, as
Chapman (1972) points out, the presence of spermatozoa in either the
testes themselves or the epididymides does not necessarily indicate
fertility. Copulations which Moynihan (1970) observed during every
month of the year in his captive, wild—caught S. 0. geoffroyi appear
to be largely ineffective during the non-breeding season(s). The
reproductive tracts of nonpregnant females also show no seasonal
effects.

It is possible that the cyclicity of the system revolves around
the reproductive condition of the reproductive females (as noted
earlier, only about one-half of the mature females are actually
reproductive at one time). Unfortunately, it was impossible to
determine whether or not this group exhibited seasonal cyclicity.

The advantages of maintaining constant reproductive tract
parameters and retaining the potential of producing two litters per
year, over exhibiting seasonal changes in reproductive tract size
and morphology and being limited to one litter per year, can only be
surmised. It seems likely that the former strategy would be the
most advantageous in situations where the animals live in a habitat
that is temporally short-lived and is also influenced by predictable
seasonal fluctuations. Using this strategy, animals invading new,
temporally discrete habitats (e.g. scattered areas of second growth
in a seasonal tropical forest) might reproduce more rapidly than

they would were they bound to a strictly seasonal cycle. This system

lhO

would also allow the more rapid replacement of individuals lost to
mortality. The latter system would be more advantageous where the
animals occupy a more or less permanent, yet seasonal, habitat type
and are subject to constant, and relatively lower, mortality rates.
The latter system would minimize the energy output for reproduction
by limiting this activity to the time of year when reproductive
return would be the greatest.

The significantly-greater body weights of both males and females
in the third trimester (July-October) appear to be related to the
general overall increase in individual condition due to the presence
of abundant food during the first half of the trimester and the
preceding trimester. These figures may also reflect, in part,
the maturation and growth of Juveniles which entered the adult
population in the first trimester as sexually mature, but smaller,
adolescents. The general increase in condition is noticeable, however,
even in the older animals. This phenomenon appears to be strictly
somatic in nature and unrelated to the "fatted male" condition which
accompanies reproduction in Saimiri (see DuMond and Hutchinson, 1967).

The significance of the wider suprapubic gland in females during
the peak season of parturition (second trimester, March-June) is open
to speculation. Pregnant females of the common marmoset (Callithrix
jacchus) have been observed to scent mark more frequently during the
two to three month period preceding parturition (H. Box, pers. comm.).
Perhaps the wider gland might reflect increased marking activity.

It seems probable that, as Epple (1972) suggests, scent markings plays
an important role in communicating reproductive information to other

group members. Furthermore, it is not implausible that scent marking

1111

by reproductive females in areas of territorial overlap might serve
to synchronize reproduction in the population as a whole. The use
of pheromones in establishing reproductive synchrony is well known
(e.g. Michael and Keverne, 1968; Bruce, H. M., 1970).

Social groups were observed to contain only one reproductive
female, although the average group contained at least one more
sexually-mature female. This phenomenon was also observed by Epple
(1970a), who found that only one female, the alpha female, reproduced
in captive groups of S. 0. geoffroyi and Callithrix jacchus. While
the mechanism underlying this phenomenon has not been elucidated
for either callitrichid, current research on the reproduction of
C. jacchus indicates that dominant behavior on the part of one female
may lead to a loss of reproductive cyclicity in the subordinate
female(s). J. P. Hearn (pers. comm.) has observed that when two
normal, cycling C. jacchus females are placed together, one will
dominate the other, and the subordinate animal will stop cycling. It
is probable that this behaviorally-induced blockage of the endocrine
system also occurs in the Panamanian tamarin.

The possible significance of this phenomenon will not be
discussed herein. It is worth noting, however, that the actual
reproductive potential of mature, free—ranging tamarin females is
lower than one would expect on the basis of their reproductive
performance in the laboratory, where they may bear two litters of
multiple young per year (6.9. Deinhardt and Deinhardt, 1966). The
importance of individual females to the reproductive effort, and the
lower actual reproductive potential per female due to behavioral

constraints, must be considered in any scheme to "harvest" these primates.

1&2

The high rates of infant mortality evident in populations of
the Panamanian tamarin (Moynihan, 1970, and this study) might be
expected given this tamarin's small size, its many predators, and
an adult population which probably approaches carrying capacity.
Within this framework, it also appears that infant mortality may
also vary according to habitat type. The limited data comparing the
reproductive success of upland vs. lowland groups seem to indicate
that upland groups are less successful in raising young than are
lowland groups. This information dovetails well with the data
presented elsewhere in this work on the differences in group composi-
tion, group stability, and the use of time and space found between
upland and lowland groups. The apparent differences in reproductive
success between groups from these two habitats is predicted in
Anderson's (1970) comparison of survival and colonization habitats.
As indicated in the other sections dealing with this concept, it
must be emphasized that the data are only suggestive, and are

inadequate for making definite statements about this apparent phenomenon.

Results and Discussion: Parasites

Helminths: The helminth parasites of S. 0. geoffroyi have been
discussed in detail by Thatcher and Porter (1968). The following
information regarding parasites found in the collected sample of 131
tamarins is presented in order to provide additional information
pertaining to the incidence of certain parasites in this primate, and
to include two forms which are not mentioned in the account by

the above authors.

113

Two nematodes, Trypanoxyuris callithricis and Subulura jacchi were
the most common intestinal helminths encountered. Eighteen tamarins
(1h percent) were infected with from 1-19 (§'= 6.h) S. jacchi. This
infection rate is substantially higher than the 9 percent (7/161) rate
recorded by Thatcher and Porter (1968). Only ten (8 percent) were
infected with T. callithricis; compared to 52/161 (32 percent) in the
sample collected by Thatcher and Porter (1968). One to ten
T. callithricis were observed per infection (§'= A.2).

Seven tamarins (5 percent) were infected with an acanthocephalan,
Prosthenorchis eZegans. Thatcher and Porter (1968) recorded a similar
infection rate (A percent). However, L. D. Hendricks (pers. comm.)
found 18 P. eZegans infections in a sample of 31 tamarins, an
incredibly high infection rate of 58 percent. This rate is comparable
with the 57 percent infection rate observed by Porter (1972) in 37
Saguinus oedipus oedipus. Infection rate by this important parasite,
then, may vary widely. It is possible, however, that the high rates
of infection recorded by Hendricks and Porter may result from their
animals being maintained in captivity for a time before their
examination. Epple (1970a) observed that captive tamarins may become
infected by eating cockroaches, an intermediate host of P. elegans,
found in most monkey facilities.

Two parasites not recorded by Thatcher and Porter (1968) were
obtained from the collection. The first, a cestode, Atriotaenia sp.,
was recovered from the small intestines of two tamarins. The second,
a pentastomid larva, Porocephalus sp., was recovered from the lungs of

two individuals.

lhh

Filarial worms (Dipetalonema marmosetae) were discovered in the
scapular and thoracic areas of h tamarins. Since these were discovered
incidently, the actual incidence of infection, as noted by Thatcher
and Porter (1968) is probably much higher.

Ectoparasites: Fairchild et al. (1968) state that ticks are
seldom found on nonhuman primates in Panama. They found only two
ticks on S. 0. geoffroyi - one Rhipicephalus sanguineus and one nymph
of the genus Amblyomma.

Six of the tamarins in the collection harboured one tick each;
two were attached to one monkey. All the ticks, which were immature
stages of Amblyomna, were attached to the nasal septum. When engorged,
they completely occluded the nasal aperture. It appears that only
those ticks which adhere to the nasal septum survive the allogrooming
and autogrooming behaviors of the tamarins. Ticks, together with
large quantities of body hair, were frequently recovered from the
stomachs of tamarins, particularly during the dry season.

Two nasal mites were recovered from two monkeys. These have
been identified as tritonymphs of an unknown species of the
Audycoptidae. Further identification is pending.

No attempt was made to collect chiggers from S. 0. geoffroyi.
Brennan and Yunker (1968) report that three Species, Eutrombicula
alfreddugesi, E. geoldii, and Pseudbschoengastia bulbifera infest
Panamanian tamarins.

Blood parasites: Blood samples from lh8 tamarins were examined
for blood parasites through direct microscopic examinations of thick
and thin blood smears. Trypanosomes were observed in the blood of

122 (88.9 percent) of the tamarins; microfilaria were found in the

th

blood of 108 (73 percent) of the animals examined. Ninety—three
(62.8 percent) of the tamarins were determined to have mixed micro-
filarial and trypanosomal infections. In the total sample, 138

(93.2 percent) of the tamarins harbored at least one of the two types
of blood parasites. These infection rates are in close agreement
with those found in 908 tamarins from Panama and Colon provinces of
Panama by Sousa et al. (197h).

Three types of trypanosomes were identified in the sample:
Trypanosomu minasense, T. rangeli, and T. cruzi. Direct blood
examination revealed that 78 (52.7 percent) of the tamarins were
infected with T. minasense, 37 (25 percent) were infected with
T. rangeli, and 2 (l.h percent) were infected with T. cruzi. One
animal harbored a mixed T. cruzi and T. minasense infection; two
others harbored mixed infections of T. rangeli and T. mindsense.
Nine animals were infected with trypanosomes which could not be
identified to Species.

The above results, however, are probably incomplete. In a test
of the blood of 15 tamarins using the techniques of direct blood
examinations, animal inoculation, hemoculture, and xenodiagnosis
(described in detail in Sousa et al., l97h), direct blood examinations
proved to be 100 percent effective in revealing the presence of
T. minasense, but only about 60 percent effective for T. rangeli
and completely ineffective for T. cruzi. It thus appears likely
that the incidence of infections by T. rangeli and T. cruzi are
much higher.

Trypanosomes were present even in very young tamarins. One

individual of one month or less of age was infected with both

11:6

trypanosomes and microfilaria. Two infants of two months of age,
and two of three months of age, harbored trypanosomes. (The ages of
infants were determined using data on dental ontogeny in Levy et al.,
1972). These infections may have been acquired through triatomine
vectors; although Lushbaugh et a1. (1969), indicates that Chagas'
disease (T. crusi) can be transmitted intraplacentally in marmosets.
It seems unlikely, however, that the high incidence of
trypanosomal infections in Panamanian tamarins can be attributed to
intraplacental transmission. The apparent avoidance of Hemipteran
prey by tamarins also renders the possibility of infection by
ingestion unlikely. A more probable explanation is that a commensal
relationship exists between S. oedipus and Rhodnius paelescens, the
principal triatomine vector of trypanosomiasis in Panama. As
mentioned previously, the accumulated debris, or nest pads, found at
the customary tamarin roosting sites may serve as refugia for
triatomine bugs, particularly during the dry season. The potential

verification of this possibility deserves further investigation.

SUMMARY

Estimates of Population Density

A modified version of the strip census method of population
estimation was used in an attempt to assess the number of tamarins
inhabiting dense, second growth vegetation on the Pacific slope of
the Panama Canal Zone. This census method, which is commonly and
uncritically used by primatologists, proved unsatisfactory under the
existing environmental conditions. The combination of low levels of
visibility within the forest and the large size of tamarin home
ranges militated against the probability of observer-group contact.
Daily estimates of population density were highly variable; on most
days, one would conclude that there were no tamarins in the area,
while on others estimates as high as 200 tamarins per square kilometer
were obtained. When plotted, the distribution of the number of
animals seen over censuses was decidedly skewed. Attempts to
normalize the distribution and to assess limits of precision for
individual estimates by increasing the number of censuses taken or
by lengthening the census transects would be impractical under the
limitations imposed by the variable tamarin activity pattern and the
monetary and temporal considerations present in most field studies.
It is suggested that the strip census method be reserved for species
living under conditions in which the probabilities of animal-observer

contact are much higher.

1&7

198

A second method of density estimation, the mapping of tamarin
home ranges and the extrapolation of density estimates from those
home ranges to larger areas is more promising, but not devoid of
shortcomings. The following factors must be considered in the
application of this census technique: 1) It is physically difficult
to map tamarin home ranges in the dense vegetation favored by the
animal; thus the number of estimates from which extrapolations can
be made may be inadequate; 2) the degree of exclusivity in home
range usage varies with the group and the habitat; thus one must
carefully document home range overlap before assigning density values
to the home range; 3) immigration and emigration are common phenomena;
one must be certain that the movements of the marked animal followed
are representative of group movements; and h) animal density within
habitats may vary seasonally; thus, estimates of ecological density
must be adjusted according to season.

Neither of the above methods is sufficient for the formulation
of absolute density estimates. The former method Should not be used
to assess the population density of tamarins living in densely—foliated
environments; the latter method, if rigorously and critically applied,

will yield valid estimates of population density.

Analysis of the Tamarin Diet

The diet of S. 0. geoffroyi is diverse and exhibits great
seasonal variability. Fruits, flowers, or buds of 108 plant species
were used by tamarins during the study. Of these plants, 69 were

identified to family, 5h to the generic level, and 26 to species.

1h9

Eight orders of the Class Insecta were also represented in the diet,
as were members of the Class Arachnida. Vertebrate prey utilized
by tamarins included two Species of lizards and the eggs of birds.

The animal portion of the diet was found to be substantial -
between 36 and 6h percent of the volume, depending on the season.
While these figures may be inflated due to the behavioral considerations
discussed in the text, they do indicate the importance of animal
protein in the dietary requirements of this species. The shift from
a preponderance of plant foods during the dry season to insect prey
during the early wet season, when both fruit and insects are at their
yearly peak of abundance, seems to denote a preference for insects
when they are readily available.

Orthopterans, primarily large grasshoppers, were the most
important insect prey, comprising over 70 percent (by volume) of
the diet's animal component. Lepidopterans (primarily larvae) and
Coleopterans together comprised over 20 percent of the remainder.
Vertebrate prey items, i.e. lizards and bird eggs, are probably
relished, but appeared to be relatively unimportant prey items.

Small fruits (< 1.5 cm) were the mainstay of the plant component
of this tamarin's diet. The presence of various fruit types in the
diet reflect annual and biennial phenological cycles within the plant
community. The implications of seasonal differences in fruit
utilization are discussed in the text.

The scarcity of both insects and fruits during the dry season
and the latter part of the wet season 1&5 reflected in the loss of
fat reserves and weight by the tamarins. Striking differences in

the local availability of food elicit marked changes in tamarin behavior.

150

Groups inhabiting food-poor upland areas both devote a greater
percentage of their time budgets to foraging and increase their
spatial range. Seminomadic wanderings occur, with the groups
travelling far beyond the boundaries of their wet season home range.

Territorial disputes with neighboring groups also become more frequent.

The Use of Time and Space

The activity pattern of this tamarin may be unique in the
Platyrrhini. Unlike other members of the suborder, this primate
begins its daily activity comparatively late in the morning and
generally terminates its activity well before sunset. Thus it does
not utilize a one to three hour period in which it might forage.

The significance of this activity pattern, together with the
assumption of apparent torpor on the part of roosting animals, has
been discussed in relation to predation pressures, competition, and
the optimal procurement of energy.

Levels of activity were not constant throughout the day, but
rather followed a definite pattern. During the early morning hours,
the tamarins were active in foraging, travel, and territorial defense.
Activity slackened during the heat of the day, when the tamarins
alternately loafed and foraged in localized, heavily-shaded areas.
Mid-afternoon marked an increase in travel and foraging activity.
Activity, in the form of both foraging and travel, increased
dramatically just prior to entering the roost tree. The day's
activity was often climaxed with rapid, silent, and direct travel

of 200 meters or more to a roost tree.

151

Travels within the home range appeared to be influenced by a
number of factors, including: 1) The availability of food resources,
2) the existence of areas of overlap with other groups, 3) the
availability of suitable arboreal runways and associated escape cover,
A) the location of suitable roosts and rain shelters, 5) the weather,
and 6) the time of day with attendant thermal considerations.

Areas of low brush, vine-enshrouded trees (particularly along
forest edges), and areas of vine-covered trees interspersed with
emergent palms with favored habitats for foraging and travel. Areas
of Open forest, whether on xeric or mesic sites, sparsely-forested
openings, and areas of grass were avoided. The canopy of the gallery
forest, with its associated vines, was used in foraging, travelling,
loafing, and roosting.

One tree Species, Anacardium excelsum, the wild cashew, was
particularly important to tamarins in this area. Anacardium excelsum
is the predominant tree in the gallery forest, where it exceeds 20 m
in height. Most of the wet season roosts occurred in tangles of vines
in the tops of A. excelsum; these tangles of vines were also used as
rain shelters. The nondeciduous leaves provided protection from the
sun at all times of the year and also served as dry season refugia for
the tamarins' insect prey. Finally, the A. excelsum fruit and the
Swollen capsule above the fruit were important food sources during
the late dry season when other fruits were scarce. The movements
of tamarins throughout the year were affected by the location of
A. excelsum.

The location of areas of group overlap also influenced patterns
of movement within the home range. Most areas of overlap were visited

daily; intergroup altercations often occurred at these sites.

152

The pattern of home range usage employed by the lowland groups
in this study differed radically from the usage pattern employed by
the upland groups. The home ranges of lowland groups were defended
as exclusive territories; in contrast, the home ranges of upland
groups overlapped extensively and were not defended as exclusive
areas. The difference in utilization patterns may result from
behavioral responses to seasonal differences in the carrying capacity
of the home ranges. An argument for this viewpoint may be found in

the text.

Group Composition and Stability

The size of S. oedipus social groups observed was small, averaging
six to seven animals per group. The male/female sex ratio appeared
to be 1:1. The social groups observed did not appear to be family
groups, but rather groups consisting of an adult, presumably alpha
female, an adult, presumably alpha male, young of the year, and a
complement of transient, subordinate animals (some sexually mature)
of both sexes.

Group composition was not stable. One reproductive female and
one resident, presumably reproductive male, formed the core of the
social groups. Infants and juveniles less than one year of age were
also sedentary. The remainder of the group changed frequently through
emigration, immigration, and mortality. Of these three mechanisms,
emigration and immigration were responsible for most alterations in
group composition. Forty-one immigrations and emigrations (considering
emigration and subsequent immigration by the same individual as

separate events), fourteen disappearances and one death were recorded

153

while observing five groups over a period of lO-lS months per group.
While emigration to adjacent groups was common, one emigration of

6.5 km was recorded, and several tamarins travelled more than 2 km

from their group of origin. Both mature and immature tamarins of

both sexes immigrated and emigrated according to the proportion of
their sex-age class in the population. Groups inhabiting upland

sites appeared to be less stable than groups occupying lowland sites.
The probable evolutionary and behavioral significance of the apparently
bisexual dominance system, high rates of group interchange, and

differences in group stability are discussed in the text.

Reproduction

Most tamarin births occurred from March through June, with a
distinct birth peak from late April to early June. Reproductive
activity, however, occurred throughout the year. Pregnant females
and/or newly born young were encountered in every month of the year.
A number of early pregnancies in April and May, and the presence of
well-developed foetuses in a female collected in September point to
the possibility of a potentially bimodal birth peak. However, the
potential for extra-birth peak reproduction is probably not realized,
since sightings of infants outside the March-June birth period were rare.

These observations were supported by the morphological examination
of male and female reproductive tracts. No significant differences
were apparent in the size or appearance of reproductive tracts
collected throughout the year. Histological examination of male
testes indicated that gametogenic tissue was present from April to

January. Testes of males collected from January to April were not

151i

examined, but gametogenesis must occur during that time to account
for young born in May and June. Thus while the potential for
reproduction is present throughout the year, few young are born,

or survive long enough to be observed, outside the March—June period
of parturition.

A number of external stimuli which might influence the timing
of reproduction have been examined. Changes in photoperiod and
temperature were discarded as possible factors since changes in these
factors are almost certainly too small to be perceptible. Changes
in humidity and rainfall between the wet and dry seasons were also
discarded since peak tamarin breeding activity occurs prior to this
event. Shifts in the time of sunrise and sunset which coincide with
the initiation of breeding activity are suggested as possible external
cues, as are plant estrogens.

The data on reproduction also indicate that only one female
per social group is reproductively active. This phenomenon is

discussed in light of past observations and current research.

Parasites

The incidence of parasites in collected tamarins, and the
significance of these parasite loads are briefly discussed with

reference to other research and aSpects of tamarin ecology.

BIBLIOGRAPHY

BIBLIOGRAPHY

Aldrich-Blake, F. P. G. 1970. Problems of social structure in forest
monkeys. Pages 79—101 in J. H. Crook, ed. Social behavior in
birds and mammals. Academic Press, New York.

Amoroso, E. C., and F. H. A. Marshall. 1960. External factors in
sexual periodicity. Pages 707-831 in A. S. Parkes, ed. Marshall's
physiology of reproduction. Vol. 1, Part 2. Longman's, London.

Altmann, S. A. l97h. Baboons, space, time, and energy. Amer. Zool.
1h: 221-2A8.

Anderson, Paul K. 1970. Ecological structure and gene flow in small
mammals. Symp. Zool. Soc. Lond., No. 26: 299-325.

Baldwin, J. D., and J. Baldwin. 1972. The ecology and behavior of
squirrel monkeys (Shimiri oerstedi) in a natural forest in western
Panama. Folia primat. 18: 161-18h.

Bates, B. C. 1970. Territorial behavior in primates: a review of
recent field studies. Primates ll: 27l-28h.

Benirschke, K., J. M. Anderson, and L. E. Brownhill. 1962. Marrow
chimerism in marmosets. Science 138: 513—515.

Bernstein, I. S. 196A. Role of the dominant male rhesus in response
to external challenges to the group. J. Comp. Physiol. Psychol.
57: ADA-A06.

1966. Analysis of a key role in a capuchin (Cebus
albifrons) group. Tulane Stud. 2001. 13(2): h9-5A.

 

. 1970. Primate status hierarchies. Pages 71-110 in
L. A. Rosenblum, ed. Primate behavior, developments in field
and laboratory research. Vol. I. Academic Press, New York.

 

, and T. P. Gordon. l97h. The function of aggression in
primate societies. Amer. Sci. 62: BOA-311.

 

Brander, R. B., and W. W. Cochran. 1969. Radio-location telemetry.
Pages 95-103 in R. H. Giles, ed. Wildlife management techniques.
The Wildlife Society, Washington, D. C.

Brennan, J. M., and C. E. Yunker. 1968. The chiggers of Panama. Pages

221-271 in R. L. Wenzel and V. J. Tipton, eds. Ectoparasites of
Panama. Field Mus. Nat. Hist., Chicago, Illinois.

155

156

Brown, J. L., and G. H. Orians. 1970. Spacing patterns in mobile
animals. Ann. Rev. Ecol. Syst. 1: 239-262.

Bruce, H. M. 1970. Pheromones. Br. Med. Bull. 26: 10-13.

Burt, W. H. l9h3. Territoriality and home range concepts as applied
to mammals. J. Mammal. 2h: 3h6-352.

Carpenter, C. R. 193A. A field study of the behavior and social
relations of howling monkeys (Alouatta palliata). Comp. Psych.
Monogr. 16(5): 1-212.

. 1952. Social behavior of non-human primates in Structure
et physiologie des societies animals, Colloques Internationaux du
Centre National de la Recherche Scientifique 3A: 227-2h6.

 

l95h. Tentative generalizations on the grouping behavior
of non-human primates. Human Biology 26(3): 269-276.

 

Chapman, D. I. 1972. Seasonal changes in the gonads and accessory
glands of male mammals. Mammal Rev. 1(7/8): 231—2u8.

Chapman, F. M. 1929. My tropical air castle. Nature studies in
Panama. Appleton, New York. A16 pp.

Chivers, D. J. 1969. On the daily behavior and Spacing of howling
monkey groups. Folia Primat. 10: h8-102.

Clark, H. C., L. H. Dunn, and J. Benavides. 1931. Experimental
transmission to man of a relapsing fever spirochete in a wild
monkey of Panama -- Leonticebus geoffroyi (Pucheran). Amer. J.
Trop. Med. 2(h): 2h3-257.

Clark, J. R. 1972. Seasonal breeding in female mammals. Mammal Rev.

1(7/8): 217-230.

Coimbra-Filho, A. F., and R. A. Mittermeier. 1973. Distribution and
ecology of the genus Leontopithecus Lesson, 18h0, in Brazil.
Primates 1A(l): h7-66.

Crook, J. H., and J. S. Gartlan. 1966. Evolution of primate societies.
Nature 210: 1200-1203.

Deinhardt, F., and J. Dienhardt. 1966. The use of Platyrrhine monkeys
in medical research. in Some recent developments in comparative
medicine (R. N. T-W-Fiennes, ed.), pp. 127—152. Symp. 2001. Soc.
Lond. No. 17, London, Academic Press.

Deinhardt, F., A. W. Holmes, R. B. Capps, and H. Popper. 1967. Studies
on the transmission of human viral hepatitus to marmoset monkeys:
I. Transmission of disease, serial passages, and description of
liver lesions. J. Exp. Med. 125(k): 673-688.

157

. 1970. Nutritional requirements of marmosets. in R. S.
Harris, ed. Feeding and nutrition of non-human primates.
Academic Press, New York.

 

DeMoor, P. P., and F. E. Steffens. 1972. The movements of vervet
monkeys (Cercopithecus aethiops) within their ranges as revealed
by radio-tracking. J. Anim. Ecol. Al(3): 677-687.

Drake, K. J. 1972. Initial experiences with Ketamine anesthesia in
non-human primates. Lab. Primate Newsl. 11(2): 18-27.

Dreizen, S., L. A. Friedman, B. M. Levy, and S. Bernick. 1972. Studies
on the biology of the periodontium of marmosets: XIV. The effect
of experimentally produced ostoemyelitis on the marmoset
periodontium. J. Periodontol. A3(6): 3A7-351.

DuMond, F. V., and T. C. Hutchinson. 1967. Squirrel monkey reproduction:
the 'fatted' male phenomenon and seasonal spermatogenesis.
Science 158: 1067-1070.

. 1968. The squirrel monkey in a seminatural environment.
Pages 88-1A5 in L. A. Rosenblum and R. Cooper, eds. Chap. A.
Academic Press, New York.

 

Dunn, F. L. 1963. Acanthocephalans and cestodes of South American
monkeys and marmosets. J. Parasitol. A9(5): 717-722.

Eisenberg, J. F., N. A. Muchenhirn, and R. Rudran. 1972. The relation
between ecology and social structure in primates. Science 176:

863-87A.

. 1973. Reproduction in two species of spider monkeys,
AteZes fhsciceps and Ateles geoffroyi. J. Mammal. 5A: 955-957.

 

, and R. W. Thorington, Jr. 1973. A preliminary analysis
of a neotropical mammal fauna. Biotropica 5(3): 150—161.

 

Eisenmann, E. 1955. The species of Middle American birds. Transactions
of the Linnaean Society of New York 7: 1-128.

Elliot, D. G. 1912. Description of a new species of Oedipomidas. Bull.
Amer. Mus. Nat. Hist. 31: 137.

Enders, R. K. 1930. Notes on some mammals from Barro Colorado Island,
Canal Zone. J. Mammal. 11: 280-292.

1935. Mammalian life histories from Barro Colorado Island.
Bull. Mus. Comp. Zool. 78: 385-502.

 

Epple, G. 1967a. Soziale Kommunikation bei Callithrix jacchus Erxleben,
1777. Pages 2A7-25A in D. Starck, R. Schneider, and H. J. Kuhn,
eds. Int. Congr. Primatol. Soc. I. (Prog. in Primatol.) Stuttgart;
Gustav Fischer.

158

1967b. Comparative investigations on the sexual and

 

social behavior of marmosets (Hapalidae). Folia Primat. 7(1):

37-65.

, and R. Lorenz. 1967. Vorkommen, morphologie und funktion

 

der Sternaldruse bei der Platyrrhini. Folia Primat. 7: 98-126.

1968. Comparative studies on vocalization in marmoset

 

monkeys (Hapalidae). Folia Primat. 8: l—AO.

1970a. Maintenance, breeding, and development of
marmoset monkeys (Callitrichidae) in captivity. Folia Primat.

12(1): 56—76.

 

1970b. Quantitative studies on scent marking in the
marmoset (Callithrix jacchus). Folia Primat. l3: A8-62.

 

1972. Social communication by olfactory signals in
marmosets. Int. Zoo. Yb. 12: 36-A2.

 

Fairchild, G. B., G. M. Kohls, and V. J. Tipton. 1968. The ticks
of Panama. Pages 167-219 in R. L. Wenzel and V. J. Tipton, eds.
Ectoparasites of Panama. Field Mus. Nat. Hist., Chicago, Ill.

Fleming, T. H. 1971. Population ecology of three Species of
Neotropical rodents. Univ. of Michigan. Mus. of 201. Misc.
Pub. No. 1A3, 73 pp.

. 1973. The reproductive cycles of three Species of
opossums and other mammals in the Panama Canal Zone. J.

Mammal. 5A(2): A39-A55.

 

Fooden, J. 196A. Stomach contents and gastrointestinal proportions
in wild-shot Guinian monkeys. Amer. J. Phys. Anthrop. 22: 227-
231.

Gartlan, J. S. 1968. Structure and function in primate society.
Folia Primat. 8: 89-120.

Hampton, J. K., Jr. 196A. Laboratory requirements and observations
of Oedipomidus oedipus. Amer. J. Phys. Anthrop. 22: 239-2AA.

, and S. H. Hampton. 1965. Marmosets (Hapalidae)
(Oedipomidus oedipus): breeding seasons, twinning, and sex
of offspring. Science 150: 915-917.

 

, S. H. Hampton, and B. T. Landwehr. 1966. Observations
on a successful breeding colony of the marmoset, Oedipomidus
oedfipus. Folia Primat. A: 265—287.

 

Hampton, S. H., and S. K. Hampton. 1967. Rearing marmosets from
birth by artificial laboratory techniques. Lab. Anim. Care 17:
1-10.

159

Handley, C. 0., Jr. 1966. Checklist of the mammals of Panama. Pages
753-795 in R. L. Wenzel and V. J. Tipton, eds. Ectoparasites of
Panama. Field Mus. Nat. Hist., Chicago, Ill.

Harrison, R. H., and W. R. Dukelow. 1973. Seasonal adaptations of
laboratory maintained squirrel monkeys (Saimiri sciureus). J.
Med. Primat. 2: 277-283.

Hayne, D. W. 19A9. An examination of the strip census method for
estimating animal populations. J. Wildl. Manage. 13: lA5-157.

Hays, W. L. 1963. Statistics for psychologists. Holt, Rinehart,
and Winston, New York. 719 pp.

Herrer, A., V. E. Thatcher, and C. M. Johnson. 1966. Natural infections
of Leishmania and trypanosomes demonstrated by skin culture. J.

Parasitol. 52(5): 95h—957.

Hershkovitz, P. 19A9. Mammals of northern Colombia: preliminary
report number A: monkeys (primates) with taxonomic revision of
some forms. Proc. U. S. Nat. Mus. 98: 323-A27.

1966. Taxonomic notes on the tamarins, genus Saguinus,
(Callitrichidae, Primates), with descriptions of four new forms.
Folia Primat. A: 381-395.

 

1969. The evolution of mammals on southern continents:
VI. The recent mammals of the Neotropical region. A zoogeographic
and ecological review. Quart. Rev. Biol. AA, 1.

 

. 1970. Dental and peridontal diseases and abnormalities
in wild-caught marmosets (Primates, Callitrichidae). Amer. J.
Phys. Anthrop. 32(3): 377-39A.

 

1972. Notes on New World monkeys. Int. Zoo. Yb. 12:

 

3-12.

Hill, W. C. O. 1957. Primates. Comparative anatomy and taxonomy.
Vol. III Pithecoidea. Univ. Press, Edinburgh.

Hladik, A., et C. M. Hladik. 1969. Rapports trophiques entre
vegetacion et primates dans la foret de Barro Colorado (Panama).
Terre Vie 1: 25-117.

Hladik, C. M. 1967. Surface relative du tractus digestif de quelques
primates, morphologie de villosites intestinales et correlations
avec le regime alimentaire. Mammalia 31: 120-1A7.

1970. Les singes de Nouveau monde. Science et Nature,
No. 102: 1A-23.

 

, A. Hladik, J. Bousset, et al. 1971. Le regime alimentaire
des primates delJinede Barro Colorado (Panama). Folia Primat. 16:
85-122.

 

16o

Holdridge, L. R., and G. Budowski. 1956. Report of an ecological
survey of the Republic of Panama. Caribb. For. 17: 92-110.

1967. Life zone ecology. Tropical Science Center, San
Jose, Costa Rica. 206 pp.

 

Hunt, R. D., L. V. Melendez, N. W. King, at al. 1970. Morphology of
a disease with features of malignant lymphoma in marmosets and owl

monkeys innoculated with Herpesvirus saimiri. J. Natl. Cancer
Inst. AA: AA7—A65.

Janzen, D. H. 1967. Synchronization of sexual reproduction of trees
within the dry season in Central America. Evolution 21: 620-630.

, and T. W. Schoener. 1968. Differences in insect abundance
and diversity between wetter and drier sites during a tropical dry
season. Ecology A9: 98-110.

 

1973. Sweep samples of tropical insects: description of
study Sites with data on species abundance and size distributions.

Ecology 5A(3): 659-708.

 

Jolly, Alison. 1972. Troop continuity and troop spacing in Propithecus
verreauxi and Lemur catta at Berenty (Madagascar). Folia Primat.
17(5): 335-362.

Jones, C. 1972. Natural diets of wild primates. Pages 58-77 in
R. N. T.-W. Fiennes, ed. Pathology of Simian primates, Part 1.
Karger, Basel.

Kaufmann, J. H. 1962. Ecology and social behavior of the coati (Nasua
narica) on Barro Colorado Island, Panama. Univ. Calif. Pub. in
Zool., Vol. 60, No. 3: 95-222.

Kinzey, W. G. 1972. Canine teeth of the monkey, Callicebus moloch:
lack of sexual dimorphism. Primates 13(A): 365-369.

Koford, C. B. 1965. Population dynamics of rhesus monkeys on Cayo
Santiago. Pages 160-17A in L. DeVore, ed. Primate behavior--
field studies in monkeys and apes. Holt, Rinehart, and Winston,
New York.

Korschgen, L. J. 1969. Procedures for food-habits analysis. Pages
233-250 in R. H. Giles, ed. Wildlife management techniques.
The Wildlife Society, Washington, D. C.

Kourany, M., and J. A. Porter. 1969. A survey of enteropathogenic
bacteria in Panamanian primates. Lab. Amer. Care 19(3): 336-3A1.

Krebs, C. J. 1972. Ecology: the experimental analysis of distribution
and abundance. Harper and Row, New York. 69A pp.

161

Lancaster, J. B., and R. B. Lee. The annual reproductive cycle in
monkeys and apes. Pages A86-513 in I. DeVore, ed. Primate
behavior--fie1d studies in monkeys and apes. Holt, Rinehart,
and Winston, New York.

Levy, B. M., S. H. Hampton, and J. K. Hampton, Jr. 1972. Some aspects
of marmoset biology. Int. Zoo. Yb. 12: 51-56.

Lushbaugh, C. C., G. Humason, and N. Gengozian. 1969. Intra-uterine
death from congenital Chagas' disease in laboratory-bred marmosets
(Saguinus fhscicollis Zagonatus). Amer. J. Trop. Med. Hyg. 18(5):
662-665.

MacArthur, R., and R. Levins. 196A. Competition, habitat selection,
and character displacement in a patchy environment. Proc. Nat.
Acad. Sci. 51: 1207-1210.

Marler, P. 1965. Communications in monkeys and apes. Pages 5AA-58A
in I. DeVore, ed. Primate behavior—-fie1d studies of monkeys
and apes. Holt, Rinehart, and Winston, New York.

Martin, A. C., R. H. Gensch, and C. P. Brown. 19A6. Alternative methods
in upland gamebird food analysis. J. Wildl. Manage. 10: 8-12.

Mason, W. A. 1966. Social organization of the South American monkey,
CaZZicebus maloch. A preliminary report. Tulane Stud. Zool. 13,
NO. 1: 23—28.

1968. Use of space by CbZZicebus groups. Pages 200-216
in P. T. Jay, ed. Primates: studies in adaptation and variability.
Holt, Rinehart, and Winston, New York.

 

Mendez, E. 1970. Los principales mamiferos silvestres de Panama.
Private edition. 283 pp.

Michael, R. P., and E. B. Keverne. 1968. Pheromones in communication
of sexual status in primates. Nature, Lond. 218: 7A6-7A9.

Mittermeier, R. A. 1973. Group activity and population dynamics of
the howler monkey on Barro Colorado Island. Primates 1A(1): 1-19.

Montgomery, G. G., W. W. Cochran, and M. E. Sundquist. 1973. Radio-
locating arboreal vertebrates in tropical forest. J. Wildl.

Manage. 37(3): A26-A28.

Mosby, H. s., I. M. Cowan, and L. Karstad. 1969. Collection and
preservation of biological materials. Pages 259-275 in R. H.
Giles, ed. Wildlife management techniques. The Wildlife Society,
Washington, D. C.

Moynihan, M. 196A. Some behavior patterns of Platyrrhine monkeys. I.

The night monkey (Aotus trivirgatus). Smithsonian Misc. Coll.
lA6(5): l-8A.

162

. 1967. Comparative aspects of communication in New World
primates. Pages 236-266 in D. Morris, ed. Primate ethology.
Weidonfeld and Nicolson, London.

 

. 1970. Some patterns of Platyrrhine monkeys. II. Saguinus
geoffroyi and some other tamarins. Smithsonian Contrib. Zool.

28: 1-77.

 

Negus, N. C., and A. J. Pinter. 1966. Reproductive responses of
.Microtus montanus to plants and plant extracts in the diet. J.
Mammal. A7(A): 596-601.

Nelson, T. W. 1975. Quantitative observations of feeding behavior in
Saguinus geoffroyi (Callitrichidae, Primates). Primates 16(2):
223-226.

Neu, C. W., C. R. Byers, and J. M. Peek. 197A. A technique for
analysis of utilization-availability data. J. Wildl. Manage.

38(3): SAl-SAS.

Noble, 0. K. 1939. The role of dominance in the social life of birds.
Auk 56: 263-273.

Odum, E. P. 1971. Fundamentals of ecology. W. F. Saunders and Co.,
Philadelphia, 676 pp.

Oppenheimer, J. R. 1969. Behavior and ecology of the white-faced
monkey Cebus capucinus, on Barro Colorado Island, C. Z. Univ.
microfilms, Univ. Michigan, Ann Arbor.

Patterson, R. L., A. Koren, and R. L. Northrup. 1973. Experimental

rubella-virus infection of marmosets (Saguinus spp.). Lab. Anim.
Sci. 23: 68.

Pianka, E. R. 1970. On r and k-selection. Amer. Nat. 10A: 592-597.

Porter, J. A., Jr. 1972. Parasites of marmosets. Lab. Anim. Sci.
22: 503-506.

Rosenblum, L. A. 1968. Some aSpects of female reproductive physiology
in the squirrel monkey. Pages 1A7-169 in L. A. Rosenblum and
R. W. Cooper, eds. The squirrel monkey. Chap. 5. Academic
Press, New York.

1972. Reproduction of squirrel monkeys in the laboratory.
Pages 130-1A3 in Breeding primates. (Karger, Basel).

 

Rudran, R. 1973. The reproductive cycles of two subspecies of purple-
faced langurs (Presbytis senex) with relation to environmental
factors. Folia Primat. l9: Al-60.

Sandler, J., R. S. Davidson, W. E. Greene, and R. D. Holzschuh. 1966.
Effects of punishment intensity on instrumental avoidance behavior.
J. Comp. Physiol. Psychol. 61(2): 212-216.

163

Schoener, T. W. 1968. Sizes of feeding territories among birds.
Ecology A9: 123-1A1.

Shadle, A. R., E. A. Mirand, and J. T. Grace, Jr. 1965. Breeding
responses in tamarins. Lab. Amin. Care 15: 1-10.

Smythe, N. 1970a. Ecology and behavior of the agouti (Dasyprocta
punctata) and related species on Barro Colorado Island, Panama.
Unpublished doctoral dissertation, Univ. of Maryland. 201 pp.

1970b. Relationships between fruiting seasons and seed
dispersal methods in a Neotropical forest. Am. Nat. 10A: 25-35.

 

Sokal, R. R., and F. J. Rohlf. 1969. Biometry. W. H. Freeman and
Co., San Francisco. 776 pp.

Sousa, 0. E., R. H. Rossan, and D. C. Baerg. 197A. The prevalence of
trypanosomes and microfilaria in Panamanian non-human primates.
J. Trop. Med. Hyg. 20: A05-A10.

Southwick, C. H., Chairman. 1975. Non-human primates: Usage and
availability for biomedical programs. Natl. Acad. Sci., Washington,
D. C. 122 pp.

Spinage, C. A. 1973. The role of photoperiodism in the seasonal
breeding of tropical African ungulates. Mammal. Rev. 3(3): 71-8A.

Struhsaker, T. T. 1967a. Behavior of vervet monkeys and other
Cercopithecus. Science 156: 1197-1203.

1967b. Ecology of vervet monkeys (Cercopithecus aethiops)
in the Masai-Amboseli Game Reserve, Kenya. Ecology A8(6): 89l-90A.

 

Telford, S. R., Jr., A. Herrer, and H. A. Christensen. 1972. Enzootic
cutaneous leishmaniasis in eastern Panama: III. Ecological factors
relating to the mammalian hosts. Ann. Trop. Med. Parasitol. 66(2):

173-179.

Thatcher, V. E., and J. A. Porter, Jr. 1968. Some helminth parasites
of Panamanian primates. Trans. Amer. Microscopical Soc. 87(2):

186-196.

Thorington, R. W., Jr. 1967. Feeding and activity of Cebus and
Saimiri in a Columbian forest. Pages 180-18A in D. Starck, R.
Schneider, and H. J. Kuhn, eds. First Cong. of Int. Primat. Soc.
(Prog. in Primatol.). Stuttgart; Gustav Fischer Verlag.

1968a. Observations of squirrel monkeys in a Colombian
forest. Pages 69-85 in L. A. Rosenblum and R. W. Cooper, eds.
The squirrel monkey. Academic Press, New York.

 

1968b. Observations of the tamarin, Saguinus midas.
Folia Primat. 9: 95-98.

 

16A

. 1970. Feeding behavior of non-human primates in the wild.
Pages 15-27 in R. S. Harris, ed. Feeding and nutrition of non-
human primates. Academic Press, New York.

 

1971. Strip census-—a feasibility study on Barro Colorado
Island. Unpublished Smithsonian report.

 

1972. Censusing wild populations of South American monkeys.
Pages 26—32 in 11. International symposium on health aspects of
the International Movement 9£_Animals. Pan American Health
Organization. WHO. 1972. Sci. Publ. No. 235.

 

 

Vandenbergh, J. G., and S. H. Vessey. 1968. Seasonal breeding of
free-ranging rhesus monkeys and related ecological factors. J.
Reprod. Fert. 15: 71-79.

Weber, E. 1972. Breeding cotton-headed tamarins at Melbourne Zoo.
Int. Zoo. Yb. 12: A9-50.

Wislocki, G. B. 1930a. A study of the scent glands in marmosets,
especially Oedipomidas geoffroyi. J. Mammal. 11: A75-A83.

. 1930b. On a series of placental stages of a platyrrhine
monkey (Ateles geoffroyi) with some remarks upon age, sex, and
breeding period in Platyrrhines. Contributions to Embryology
No. 133, Carnegie Inst., Wash. Pub. No. AlA: 173-192.

 

. 1932. Placentation in the marmoset (Oedipomidas geoffroyi)
with remarks on twinning in monkeys. Anat. Rec. 52: 381-399.

 

. 1939. Observations on twinning in marmosets. Amer. J.

Anat. 6A: AA5-A83.

 

Young, M. D. 1970. Natural and induced malarias in Western Hemisphere
monkeys. Lab. Anim. Care 20(2): 361-367.