.. :3. _ . _ . £13.53

. first .
v Wumwuw; mm“
. «mm

53?...

.1. :3...
z . .

3.:

.

. y,
(I 9
Mummy.»

Afar
.0 1.5. .

.Vr . . 3

V... V V V
1293: .. V . .
.V 2 x V . . . V . . . V , . .«P 03
,. _ . . $5.“ ..3.. 4:...
z... . V V 3...... .31....
. .
1

..
.:

3 .5:
If...

Rina}; V . . , . l .\
. V In: )
. 2? .... ,
. .5 r. .

I h
.33”. V . .
V LkmnflfiQ s. _ .3H
. V .. 193%, (333$:
s

 

.12....
. V

. v . 3.1..

3. 1

 

 

f 3)
3.: .

 

kw: . A144»
:... . Lamannzflwv u
.3.w§.m§.mf

» i w. I

a: if .s..u4!~....a ..
I
.r
a

\V

 

1%.”. 1.} . .
. . 1:,

. ,{V
. tour . nfiu...
t... :3 , g .
1 13:, , i x
. .34.}. . . .v ‘\3.«-.r
V.. 4» . . 3. f.
'V V '1‘-

 

 

r Iiwvlyksr v.

 

THr'rns

70H

This is to certify that the

dissertation entitled

SOCIO-ECOLOGICAL DETERMINANTS OF SPACE UTILIZATION
IN THE SPOTTED HYENA, CROCUTA CROCUTA

 

presented by

ERIN ELIZABETH BOYDSTON

has been accepted towards fulfillment
of the requirements for

Ph .D . degree in Zoolggy

7%/ &%9
/ Ma joi’firofessorfl

Date March 16, 2001

MS U is an Affirmative Action/[q ual Opportunity Institution 0- 12771

 

 

LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

1-)!)

JUN 1h34 @092;

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c/CIRC/Dateompss-ms

 

SOCIO-ECOLOGICAL DETERMINANTS OF SPACE UTILIZATION IN THE
SPOTTED HYENA, CROCUTA CROCUTA

By

Erin Elizabeth Boydston

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

DOCTOR OF PHILOSOPHY

Department of Zoology
2001

ABSTRACT

SOClO-ECOLOGICAL DETERMINANTS OF SPACE UTILIZATION IN THE
SPOTTED HYENA, CROCUTA CROCUTA

By

Erin Elizabeth Boydston

Large mammalian carnivores generally have large space requirements
and are extremely sensitive to habitat fragmentation. Consequently, large
carnivores tend to decline in areas of increasing human density, and disappear
from small parks and reserves. Within a population, individuals with different
patterns of space utilization may be subject to different selection pressures and
mortality risks from humans. Understanding the determinants of space utilization
patterns may help us ameliorate effects on wildlife of expanding human
populations and changing land-use patterns. Here I investigated space utilization
in the spotted hyena (Crocuta crocuta) in order to elucidate sources of variation
in space utilization among individuals of this species large carnivore species.
Integrating field data from one large Crocuta social group, or clan, with maps and
remotely-sensed images into a Geographic Information System (GIS), I tested
predictions of hypotheses suggesting that space utilization patterns in Crocuta
are affected by age, sex, social rank, reproductive state, and ecological
variables, including lions, ungulate prey, vegetation type, and human activity.

I determined the borders of the study clan’s territory from locations of
group territorial activities, such as clan wars and border patrols. Female clan

members led most of these territorial activities. Most alien hyenas encountered

within the clan’s territory were males. The probability of attack on alien hyenas by
residents varied with sex of intruders and residents.

l mapped den sites used by the study clan, and calculated frequency and
distance of den moves. Hyenas used dens throughout their territory for rearing
cubs, and on average, moved their clan’s communal den approximately every
four weeks. Site selection of dens varied with female social rank.

Male and female hyenas showed remarkably similar patterns of space
utilization from 1 mo of age through the age of weaning. A sex difference in
space utilization emerged at reproductive maturity but before males dispersed;
males were found farther from the center of the territory than were females.

Space utilization patterns among adult female hyenas varied with
reproductive state and rank. Whether or not they had dependent cubs residing at
the clan’s communal den, high-ranking females tended be found in the vicinity of
the communal den. Low-ranking females with cubs at the den were also found
near the den, but these females ranged more widely when they did not have den-
dwelling cubs. When unconstrained by the needs of den-dwelling cubs, low-
ranking females may have used more remote areas in order to avoid competition
with conspecifics. Surprisingly, all hyenas avoided a central short grass plain that
contained the highest prey abundance in the clan's territory, and instead they
utilized areas of lower prey concentration and heavier vegetation cover. Long-
terrn data suggested that these hyenas have altered their space utilization

patterns in response to increasing human activity.

ACKNOWLEDGMENTS

The efforts and kindness of many people made this work possible.

I thank the Office of the President of Kenya for permission to conduct this
research. I also thank the Kenya Wildlife Service, the Narok County Council, and
the Senior Warden of the Masai‘Mara National Reserve for their cooperation.

Field research was supported by NSF grants lBN9309805, lBN9630667,
and lBN9906445. Funds from the Department of Zoology and Ecology,
Evolutionary Biology, and Behavior Program contributed to field and lab supplies,
and the College of Natural Science provided summer support.

A number of people assisted in the field. I wish to thank the people with
whom I had the privilege to work: Anne Engh, lsla Graham, Karen Nutt, Laura
Sams, Brad White, and in particular Micaela Szykman for weathering the
excitement and occasional difficulties of El Niho with me. I wish to also thank
many other people whose assistance in the field made this work possible: Nancy
Berry, Susan Cooper, Stephan Dloniak, Martin Durham, Josh Friedman, Paula
Garrett, Tyson Harty, Cathy Katona, Gabe Ording, Kim Weibel.

I thank Russ Kruska, Onyango Okello, Robin Reid, and Cathy Wilson at
ILRI in Nairobi for providing advice and geographical information.

I extremely grateful to Barbara and Nigel Dundas for first introducing me to

Africa, and for their tremendous support in Nairobi and in the Mara.

In Michigan, many other family members offered their support. My parents
Robert and Carolyn Boydston have encouraged my interest in wildlife for over
thirty years and made it possible for me to pursue my dreams.

I appreciate the support and help of the community of graduate students in
the Holekamp Lab: Stephanie Dloniak, Anne Engh, Keron Greene, Terri
McElhinny, Eva Maria Muecke, Scott Nunes, Micaela Szykman, Russ van Horn,
and Sofi Wahaj. Other lab associates Rick Berg, Pat Bills, Karen Hubbard, and
Carmen Salsbury also provided technical help and encouragement.

I thank my East Lansing housemates Laurie Boger, Beth Capaldi, Jennifer
Rosinski, Emily Lyons, Julie Stepanek, and Cindy Wei their friendship and
tolerance. For helpful suggestions and discussions, friendship, and constant
encouragement, I am particularly grateful to Puja Batra, Beth Capaldi, and Carlos
Lopez Gonzalez.

I thank my guidance committee Guy Bush, Fred Dyer, Kay Holekamp, and
Laura Smale for their support at all stages and insightful comments on earlier
drafts, and I thank Don Straney who stepped in during the absence of one
member. I extend special thanks to Laura Smale for her role in data collection,
for helpful suggestions during all phases of this research, and for her scientific
mentorship. I am grateful to my adviser Kay Holekamp for her guidance and
incredible foresight in the development of this research and its applications to
conservation. I wish to thank her for her great scientific leadership and for

sharing the fascinating world of hyenas.

Undergraduates Leigh Chirgwin, Karen Kapheim, Toni Lyn Morelli, and
Elaine Wolfe worked directly on this project in the lab. I am very grateful to them
for their hard work, initiative, and also patience in dealing with the sometimes
tedious aspects of data extraction and digitizing.

I wish to thank Gay Bradshaw at Oregon State University for first
introducing me to the concept and usefulness of 3 Geographic Information
System (GIS) for my research, for GIS training in her lab, and advice in planning
my field research. I am in debt to Scott Bergen at Oregon State University for his
time and assistance with GIS analyses.

I thank David Skole and Walter Chomentowski of the Basic Science and
Remote Sensing Initiative (BSRSI) Laboratory at Michigan State University for
generous access to computer equipment and assistance with mapping that
launched the GIS component of this research. Under Walter Chomentowski’s
guidance, Sara Smith helped digitize the study site from air photos. I extend a
special thanks to Chris Barber at BRSRI for assistance with digitizing and spatial
analyses, for being a patient teacher, and for sharing his extensive knowledge of
GIS and remote sensing. I thank the BSRSI support staff and researchers for
their time and technical support.

From the Computational Ecology and Visualization Lab at Michigan State
University, I thank Stuart Gage, Manuel Colunga, Bryan Pijanowski, Amos
Ziegler for training and access to computer equipment when I first started
Ieaming to use GIS technology, and for continuing to be a helpful resource

throughout this process.

vi

Many people at Michigan State University were kind enough to share their
expertise and answer questions when I knocked on their doors: Bilal Butt, David
Campbell, Mark Cochrane, Bryan Epperson, Catherine Lindell, Barb Lundrigan,
Dave Lusch, Jiaguo Qi, Alan Tessier, Scott Winterstien, and Ned Walker, who
identified the ticks that I collected. For their help and patience with a variety of

tasks, I also thank the Zoology and EEBB support staff.

vii

TABLE OF CONTENTS

LIST OF TABLES .................................................................................................. x
LIST OF FIGURES ............................................................................................... xi
CHAPTER 1:

INTRODUCTION 1
Space utilization ......................................................................................... 3
GIS technology and applications ............................................................... 5
Overview of chapters ................................................................................. 8

CHAPTER 2:

SEX DIFFERENCES IN TERRITORIAL BEHAVIOR ......................................... 13
Introduction .............................................................................................. 13
Methods ................................................................................................... 16
Results .................................................................................................... 22
Discussion ............................................................................................... 32

CHAPTER 3:

DENNING BEHAVIOR ....................................................................................... 39
Introduction .............................................................................................. 39
Methods ................................................................................................... 42
Results ..................................................................................................... 45
Discussion ............................................................................................... 55

CHAPTER 4:

ONTOGENY OF SPACE UTILIZATION ............................................................. 64
Introduction .............................................................................................. 64
Methods ................................................................................................... 65
Results ..................................................................................................... 70
Discussion ............................................................................................... 89

CHAPTER 5:

SOCIAL AND REPRODUCTIVE DETERMINANTS OF SPACE

UTILIZATION PATTERNS IN ADULT FEMALES .............................................. 92
Introduction .............................................................................................. 92
Methods ................................................................................................... 93
Results ................................................................................................... 103
Discussion .............................................................................................. 1 15

viii

CHAPTER 6:
ECOLOGICAL DETERMINANTS OF SPACE UTILIZATION AND THE
CONSEQUENCES OF EDGE EFFECTS ON SPOTTED HYENA BEHAVIOR.120

Introduction ............................................................................................. 1 20
Methods ................................................................................................. 126
Results ................................................................................................... 135
Discussion .............................................................................................. 1 55
LITERATURE CITED ........................................................................................ 160

LIST OF TABLES

Table 5-1. UD calculations for each individual for 1996-1998 ........................... 108

Table 6-1. Distances of hyenas to dense vegetation cover ............................... 149

Chapter 1

INTRODUCTION

Large mammalian carnivores are important in the maintenance of
biodiversity and serve as keystone predators in many ecosystems (Hofer & East,
1995; Terborgh et al., 1999; Terborgh & Soule, 1999). Despite their ecological
importance, however, populations of carnivores around the world are declining
and disappearing (Stuart et al., 1985; Schonewald-Cox et al., 1991;
Wikramanayake et al., 1998; Woodroffe 8. Ginsberg, 1998; Terborgh & Soule,
1999). In addition to the predicted increase in herbivore populations and
instability in prey population dynamics caused by the loss of top predators
(Terborgh & Soule, 1999), the decline of large carnivores reduces the resilience
and functioning of ecosystems by breaking other ecological links between
species (Soule & Noss, 1998; Berger, 1999; Crooks & Soule, 1999). Recent
research has emphasized the need to understand individual behavior of large
carnivores in order to conserve them and their roles in ecosystems in the face of
burgeoning human populations and widespread habitat fragmentation (e.g.,
Creel, 1998; Berger, 1999; Caro, 2000). Space utilization by carnivores is one
area in which there is a direct link between behavior and conservation, but in
which more research is needed in order to aid conservation efforts (Caro, 1999).

Large carnivores typically have large space requirements (Ewer, 1973;

Gittleman & Harvey, 1982), and their wide-ranging behavior is contributing to

their declining numbers (Woodroffe & Ginsberg, 1998; 2000). Large carnivores
appear to be extremely sensitive to habitat fragmentation, and they are strongly
affected by “edge effects” (Woodroffe & Ginsberg, 1998; 2000). Local people
often perceive predators as threats to their livestock and personal safety, and the
interactions between humans and carnivores along the edges of protected areas
can lead to high predator mortality and local extinctions of carnivore populations
(Woodroffe & Ginsberg, 1998; 2000). Understanding the determinants of
carnivore ranging patterns can potentially help us reduce the risk of predator
mortality resulting from conflicts with humans. Currently, however, there is a
shortage of empirical data on ranging behavior in fragmented landscapes, data
documenting how ranging behavior naturally varies within and between years,
and also data documenting how behavior varies between disturbed and
undisturbed habitats (Caro, 1999). Such data are critical in allowing us to identify
and predict how carnivores will respond to Iong-tenn environmental changes and
increasing human pressures. Here I sought to help fill this void by contributing to
the fields of animal behavior and conservation biology a study of space utilization
patterns in one species of large carnivore, the spotted hyena (Crocuta crocuta) .
Although spotted hyenas are not yet listed as endangered or threatened,
they nevertheless offer rich opportunities to test predictions of hypotheses
suggesting determinants of individual space utilization patterns, and to apply
knowledge of behavior to wildlife conservation efforts. Distributed widely across
sub-Saharan Africa and across a spectrum of habitats, spotted hyenas are the

most abundant large carnivores in Africa, and they are keystone predators in

many ecosystems (Kruuk, 1972; Hofer & East, 1995). A better understanding of
their behavioral ecology and responses to anthropogenic disturbance should be
useful to those concerned with conservation of African ecosystems and
biodiversity, or with improving reserve design and management practices.
Furthermore, compared with other large carnivores, Crocuta are extraordinarily
flexible in their behavior and ecology. For example, they breed year round, they ‘
can be active at any time of day, and they are opportunistic feeders with catholic
tastes. Therefore their responses to long-term environmental changes should
represent very conservative indicators of how other top predators are likely to

respond to such changes (Arcese & Sinclair, 1997).

SPACE UTILIZA TION

Studies of animal space utilization seek to answer questions about “why
an animal is in a particular place at a particular time” (Sanderson, 1966; Bekoff &
Mech, 1984). The term “space utilization” is often used synonymously with, or
alongside, other terms including “space use” or ”use of space” (Gascon & Miller,
1982; Thery, 1992; Kessel & Brent, 1996), “movement patterns” (Sanderson,
1966; Bekoff & Mech, 1984), and “spacing” (Brown & Orians, 1970). These terms
have broad and rarely explicit definitions, and they can be related to a variety of
topics. Sanderson (1966), for example, uses “mammal movements” to
encompass “activity, home range, migration, immigration, emigration, and
movements associated with behavior and territory.” Space utilization provides an

important link between an animal and its social and physical environment that

helps us better understand its social interactions and ecology. However, in order
to answer the question of “why” an animal uses a particular area, we have to first
show where that animal spends its time.

Several different aspects of space utilization have been explored
extensively in the behavioral ecology literature. Perhaps the greatest amount of
past attention has focused on various methods for calculating the home ranges
of individual animals, or entire groups of animals (Burt, 1943; Bekoff & Mech,
1984; Andreassen et al., 1993; Powell, 2000). A home range is usually defined
as the area an animal (or group) needs to complete its daily activities over a
period of time (Burt, 1943; Powell, 2000). Home range therefore refers simply to
the area an animal uses. A territory, by contrast, is defined as the area an
individual (or group) defends (Burt, 1943). Documenting the size, shape, and
location of home ranges or territories represent common strategies that
researchers use for elucidating patterns of space utilization by animals (Hayne,
1949; Fritts 8 Mech, 1981; Bekoff & Mech, 1984; Fuller & Kat, 1990; Harris et al.,
1990; Andreassen et al., 1993; Andreka et al., 1999; Perrin et al., 2000).
Additional useful information can be gathered by documenting distances at which
animals are found from particular landscape features.

My purpose here was to use both traditional and newly-developed
techniques to elucidate patterns of space utilization in the spotted hyena.
Specifically, I examined territorial behavior, patterns of den usage, ontogenetic
change in space utilization, and the social and ecological variables influencing

where adult female Crocuta spend their time. That is, I tested predictions of

hypotheses suggesting that space utilization patterns in adult female Crocuta are
affected by social rank, reproductive condition, concurrent distributions of lions
(Panthera Ieo) and ungulate prey, or by human activity, particularly pastoralists
and their livestock. The study population was one large hyena clan, the Talek
clan, occupying a territory at the edge of Kenya’s Masai Mara National Reserve,
in the northeastern part of the Serengeti-Mara ecosystem. The Talek clan of
hyenas has been intensively studied by Dr. Kay Holekamp and Dr. Laura Smale
since early 1988, and an extensive long-term database on the behavior of these
hyenas and on their environment has been collected. I collected new field data
on these hyenas from 1996-1998, and integrated these field data with remotely-
sensed data using new computer technologies. I also utilized field notes collected
prior to 1996 to study aspects of relatively infrequent behaviors such as den
moves and group territorial activities, that would not have been possible to study
in the course of just a few years. Also from observations in previously collected
field notes, I was able to document space utilization patterns for hyenas in this

same clan a decade earlier, and compare these earlier patterns to current ones.

GIS TECHNOLOGYAND APPLICA TIONS

I applied Geographic lnfonnation System (GIS) technology to study the
social and ecological determinants of Crocuta space-utilization patterns. While
the use of GIS technology is growing rapidly in many fields, it is still relatively rare
in studies of animal behavior. However, this powerful technology enables

biologists to map the behavior of individuals onto real-world landscapes, and one

important application of this technology is to ask questions about space utilization
and ranging behavior in changing environments.

At the simplest level, GIS is a computerized mapping technology that
permits users to overlay multiple digitized maps on top of each other, much like a
perfectly aligned stack of transparencies. Because each digitized map in any
such “stack” may depict a different habitat feature or location information
obtained from a different array of individual animals, GIS represents a useful tool
for testing hypotheses suggesting specific relationships among spatial variables.
The main tools used in GIS technology are data acquired via remote sensing,
Global Positioning System (GPS) technology, and GIS software.

The two most widely used forms of remotely sensed data are satellite
imagery and aerial photography. However, these images may exist without '
geographic units until the user provides such units in a process called geo-
rectification. To geo-rectify satellite images, GPS locations called “ground
control” or “ground truth” points can be collected at locations in the field that are
visible even from outer space, such as bridges, dams, or specific topographic
features. Any digital satellite image or air photo can then be calibrated using
these ground control points, such that the entire image contains coordinate
information. GPS links data collected on the ground with digital images on a
computer screen. A GPS unit measures its location through communication with
particular satellites in orbit around the earth, and returns this information to the
user in real-world geographic coordinates such as latitude and longitude. Finally

GIS software such as ArcView or Arclnfo can be used to combine and analyze

multiple layers of spatial data such as satellite imagery, topographic maps, and
field data that include location information. Once all the data are in a common
coordinate system, GIS technology allows each new layer of spatial data to be
precisely superimposed on top of previous layers. For example, maps showing
the home ranges of specific individual hyenas can be laid down on top of
digitized topographic, vegetation, or road maps to assess habitat use by hyenas
in relation to particular landscape features.

None of these sorts of GIS layers existed for the Talek clan of hyenas
prior to the project presented here. To initiate the GIS analysis of variables
affecting hyena space utilization patterns, from 1996 to 1998, observers in the
field regularly recorded the geographic coordinates at which 13 radio-collared
adult Talek females were found multiple times per week. These coordinates
comprised the primary data set for my study of the space utilization patterns
exhibited by Talek females (Chapters 5 and 6). The location data for each female
were collected with a GPS unit in the field, or they were digitized directly onto a
digital air photo map of the study site using ERDAS software. The digital air
photo map was a composite image made from thirty overlapping air photos that
were taken in 1991 by the Kenya Wildlife Service and Department of Rangeland
Science and Remote Sensing. I scanned the air photos into a computer, and
geo-rectified them in UTM meters with GPS data from the field, in order to create
a digital base map of the study site on which trees, stream crossings, and other
landmarks were clearly visible. From this base map, I digitized creeks, roads,

locations of villages, and borders of the territory defended by the Talek clan. A

digital habitat map showing open grassland and areas of denser vegetation cover
such as bushes and trees was also made from the air photos. In the field from
1996 to 1998, I also conducted weekly census counts of wild ungulates and
domestic livestock in each part of the Talek clan’s territory, and recorded the
locations of all lion sightings. All of these different data types, along with various
remotely—sensed images of the area depicting vegetation cover types and water
availability, were then integrated and analyzed using Arclnfo, ArcView, and
ERDAS Imagine software. Some of these data layers were specific only to
certain time periods, but many of the GIS layers can be used in future to launch
new studies of hyena space utilization in this area, and they will also permit a

spatial component to be added to future studies of behavior.

OVERVIEW OF CHAPTERS

Each of the following five chapters (Chapters 2 through 6) presents results
not shown previously for any hyena population, and thus contributes new
information to our understanding of hyena behavior. The chapters document
variation in hyena space utilization patterns, and answer some of the questions
about why individual hyenas exhibit certain space utilization patterns. However,
the view of the hyena’s spatial world presented here is hardly comprehensive,
and the results raise more questions than they answer.

Chapter 2 contributes to our understanding of sex differences in behavior,
showing that differential investment by male and female hyenas in territorial .

defense reflects the differing life histories of male and female hyenas. In its

exploration of defense of the Talek clan’s group territory, Chapter 2 sets the
stage for all subsequent parts of the dissertation. This chapter defines the
geography of the Talek clan’s territory, showing the borders that l determined
from locations of group territorial activities recorded by numerous trained
observers between 1988 and 1998. Being able to draw the boundaries of the
Talek clan’s defended territory based on direct observations of territorial
behaviors and scent-marking sites represents a substantial improvement here
over many earlier studies of space utilization in mammals that were more difficult
to observe such that territoriality could usually only be indirectly inferred from
locations of animals. Definition of the Talek clan group territory permitted many
different types of analyses of space utilization not possible in other studies. Maps
of the Talek clan territory appear several times in this dissertation, and the
subsequent chapters, Chapters 3 through 6, all focus on movements and
locations of Talek hyenas, or their dens, in reference to these defended borders.
Chapter 3 provides new descriptive information on denning habits in
Crocuta, such as locations of den sites and frequency of den moves, based on
observations made over 10 years on the Talek clan. This chapter shows which
communal dens the Talek clan hyenas used most often and for the longest
durations during this period, although it leaves unanswered questions of why
communal den moves occurred and why hyenas denned at particular sites.
Chapter 3 shows variation among adult females In den site selection that might
contribute to variation in female reproductive success. This chapter highlights the

biological importance of dens in this species. This is important for subsequent

chapters, in which the distance to the communal den is one of ways I selected to
estimate variation in space utilization patterns (Chapters 4 and 5).

In Chapter 4, I explore how space utilization varies with age and sex of
hyenas. An individual’s pattern of space utilization may vary tremendously
throughout its life: young dependent mammals may stay close to their mothers or
to a den site, dispersing individuals often cover large distances, and reproductive
adults may establish territories or travel in search of mates. Ontogenetic patterns
of space utilization have received little attention in the mammalian literature.
Despite dramatic sex differences in dispersal behavior in mammals, we know
little about the emergence of sex differences in space utilization behavior prior to
dispersal. Whereas in Chapter 2 I examine differences in territorial behaviors
between adult females and immigrant males, in Chapter 4 I compare space
utilization patterns of females and natal male hyenas from birth through
adulthood. My finding that male and female hyenas show remarkably similar
space utilization patterns throughout much of their early ontogeny thus
represents an important contribution to our understanding of development,
movement, and dispersal in this species.

In Chapter 5, I show that space utilization patterns among adult female
hyenas vary with their rank and reproductive state. Social rank strongly
influences space utilization, with high-ranking females tending to be found closer
to the group’s communal den than are low-ranking females. Females exhibit
different space utilization patterns depending on whether or not they have cubs

residing at the communal den. Females with den-dwelling cubs use smaller

10

areas than do other females. Low-ranking females without dependent cubs range
most widely when prey resources are scarce, using areas farther from the
communal den to avoid competition with high-ranking females. This study thus
shows how space utilization patterns can vary among individuals within the same
social group. The assumption that the space utilization pattern of a group
represents that of all its individual members may be true for some group-living
species, (e.g., Lycaon pictus: Burrows, 1995), but important individual variation
may be overlooked in other species, particularly those with fission-fusion
societies as are common in many carnivores (Kruuk & Macdonald, 1985; Packer
et al., 1990; Holekamp et al., 2000). The variation in space utilization
documented here has important conservation implications, because individuals
may be subject to different selection pressures and risks as a result of their
differential utilization of the landscape.

Chapter 6 was originally planned to be an exploration of the ecological
determinants of the space utilization patterns of spotted hyenas. However, the
patterns apparent in my field data collected during the 1996-98 time period at first
appeared to make no biological sense. Specifically, hyenas were almost never
found during this period in the center of their territory where prey were most
abundant. Only by taking an historical look at hyena behavior and environmental
changes in this study system could these unexpected patterns of space
utilization be interpreted. This chapter thus makes the most direct contribution to

conservation of any chapter in this dissertation, by documenting temporal

11

changes in the space utilization patterns of hyenas due to recent anthropogenic

disturbance in the Mara ecosystem.

12

Chapter 2

SEX DIFFERENCES IN TERRITORIAL BEHAVIOR

INTRODUCTION

Defense of group territories occurs in many gregarious mammalian
carnivores including lions (Panthera leo: Schaller, 1972; Packer, 1986), dwarf
mongooses (Helogale parvu/a: Rood, 1983; Rasa, 1987), suricates (Suricata
sun'catta: Doolan & Macdonald, 1996), canids (Mech, 1970; Bekoff 8 Wells,
1986), brown hyenas (Hyaena brunnei: Owens & Owens, 1979b; Mills, 1990),
and spotted hyenas (Crocuta crocuta: Kruuk, 1972; Mills, 1990; Henschel &
Skinner, 1991). In these carnivores, territories of neighboring groups generally
overlap little or not at all, group members often scent-mark to advertise territory
boundaries, and resident animals frequently direct aggressive behavior toward
alien conspecifics found within the territory (e.g., wolves, Canis lupus: Mech,
1970; Kruuk, 1972; lions: Schaller, 1972; red foxes, Vulpes vulpes: Macdonald,
1981; coyotes, Canis Iatrans: Bowen, 1982; spotted hyenas: Mills, 1990;
Henschel & Skinner, 1991).

Investment in territorial behavior by individual members of animal groups
should vary with the potential costs or benefits accruing to those individuals from
advertisement and defense of the territory (e.g., Milinski & Parker, 1991; Packer
& Pusey, 1997). Because strategies for maximizing reproductive success are

sexually dimorphic in most animals (Darwin, 1871; Bateman, 1948; Trivers, 1972;

13

Emlen & Oring, 1977; CIutton-Brock, 1988) the costs and benefits of territorial
behavior in mammalian carnivores should theoretically differ for males and
females. These costs and benefits should also vary as a function of the sex of
alien animals encountered within the home territory, and behavior consistent with
these predictions has been observed to date in a number of carnivores. For
example, lions, dwarf mongooses, suricates, and many mustelids defend their
territories most vigorously against same-sex intruders (Powell, 1979; Rood,
1983; McComb et al., 1994; Grinnell et al., 1995; Heinsohn & Packer, 1995;
Doolan 8. Macdonald, 1996). Among domestic dogs (Canis domesticus), group
members are more aggressive toward strangers of the same sex than toward
opposite-sex individuals (King, 1954). Fights between opposite-sex pairs of
brown hyenas from different social groups are rare, whereas same-sex fights are
both more common and more intense (Mills, 1983; 1990). Henschel & Skinner
(1991) reported that female spotted hyenas in their study population in southern
Africa engaged more frequently in territorial scent-marking than did males, and
that females invested more time in territorial activities than did males. Our broad
objective here was to replicate and extend these findings in a population of
spotted hyenas in eastern Africa.

Spotted hyenas live in complex stable social groups called clans (Kruuk,
1972). Each clan contains multiple adult females, their offspring, and one to
several resident immigrant adult males. Female hyenas generally spend their
entire lives in their natal clans, while all males disperse shortly after puberty

(Henschel & Skinner, 1987; Smale et al., 1997; Holekamp & Smale, 1998b).

14

Males that successfully immigrate are subordinate to all natal females and their
offspring in the new clan (Tilson & Hamilton, 1984; Smale et al., 1993). Thus,
adult females are socially dominant to adult immigrant males, and immigrant
males invariably defer to all natal animals during competition over food
(Holekamp & Smale, 1998b).

Although all adult female clan members breed, the reproductive success
(R8) of female Crocuta varies directly with their priority of access to food (Frank,
1986b; Frank et al., 1995; Holekamp et al., 1996; Holekamp et al., 1999b). In
eastern Africa, the diet of spotted hyenas consists mainly of medium and large-
bodied ungulates that they hunt themselves (Kruuk, 1972). Hyenas usually hunt
within the boundaries of their clan territories, and defense of territory borders is
often associated with competition over food (Kruuk, 1972; Henschel & Skinner,
1991). RS of male Crocuta is most strongly affected by whether or not males can
manage to join new clans after dispersing from their natal groups, and by the
duration of their tenure in those new clans (Engh et al., unpubl. data). These
different determinants of RS in males and females should generate
corresponding sex differences with respect to perception of potential threats to
critical resources, to the value hyenas of each sex assign to the group territory,
and hence, to the risks males and females are willing to take in its defense

My specific goals here were to document the territorial behavior exhibited
by members of one large clan of spotted hyenas, examine leadership roles
during group interactions with alien hyenas, and compare hourly rates at which

resident females and immigrant males engaged in territorial marking and

15

aggressive behavior toward intruders. The forms of territorial behavior that were
monitored included patrols along territorial borders, clashes or “wars” between
neighboring clans, and other encounters between resident and alien hyenas
within the territory of the study clan. I hypothesized that since female hyenas are
philopatric and require exclusive access to food available within the territory, they
should stand to lose more than males from alien intrusions into their territory
(e.g., Henschel & Skinner, 1991; Heinsohn & Packer, 1995). I therefore expected
that female Crocuta would engage more frequently than males in territorial
marking and defense and that they would take more risks when confronting
intruders. I further expected resident females to behave more aggressively than
resident immigrant males toward intruders, particularly alien female intruders who
would not show male-like deference to resident females during competition for
access to food resources critical for successful reproduction. Because resident
males should attempt to maximize their own mating opportunities, I hypothesized
that they would play a larger role than females in defending the clan’s territory
against intruding males who might attempt to join the clan and thus compete with

resident males for access to females.

METHODS

Study population.— The study animals comprised one large clan of
spotted hyenas inhabiting the Talek region of the Masai Mara National Reserve,
Kenya. This area is characterized by open rolling grassland broken by seasonal

creek beds, and it is grazed year round by large concentrations of several

16

different ungulate species (Frank, 1986a). Between 25 May 1988 and 25 May
1998, observers monitored Talek hyenas on 3120 d. Each Talek hyena was
individually identified by its unique spots and other natural marks, and sexed
based on the dimorphic glans morphology of the erect phallus (Frank et al.,
1990). Reproductive states of Talek females were known, and ages of all hyenas
born in the study clan since 1988 were known to within i 7 d, as described
previously (Holekamp & Smale, 1993; Holekamp et al., 1996). Natal animals
were considered to be adults at 3 years of age, and younger animals were
considered to be subadults. All hyenas born in Talek were considered to be
resident animals, as were those adult males who had emigrated from other clans
but who had been present in the Talek area and tolerated by residents there for
at least 30 d. Although subadults occasionally participated in territorial behaviors,
I focused here exclusively on behavior exhibited by resident adult females and
resident immigrant males. Adult natal males rarely participated in territorial
activities, so all adult males used in all data analyses below were immigrants. All
adult hyenas observed in the Talek area that had never previously been seen
there were called “alien intruders.”

Behavioral observations— Most behavioral data were collected from 0530
to 0900 h, and from 1700 to 1930 h, but these were supplemented with
observations made near mid-day and at night using night vision binoculars. On
any given day of observation, either one or two research vehicles were deployed
in the Talek area and environs, with one or two observers per vehicle.

Throughout the study period, each time observers came across one or more

17

hyenas separated from other hyenas by at least 200 m, observers initiated an
observation session. The observation session ended when observers left that
individual or group. Duration of observation sessions ranged from 5 min to
several hours. During every observation session, all aggressive, appeasement,
and scent-marking behaviors (Kruuk, 1972) were recorded as critical incidents
(all-occurrence sampling, Altmann, 1974). All clan wars, other appearances of
alien hyenas in the Talek area, and interactions between Talek hyenas and
intruders were also recorded.

When groups of intruders are detected near territorial boundaries, resident
hyenas may initiate a cooperative attack to expel them (Kruuk, 1972; Hofer &
East, 1993c; Holekamp et al., 1993). These clan wars are characterized by
coordinated rushes and attacks by both sets of participants, as well as by
frequent recruitment vocalizations to call allies to the scene (e.g., Kruuk, 1972;
East 8. Hofer, 1991). Although actual physical contact between opponents is rare
during clan wars, serious injury or death sometimes results from severe biting
(Lawick & van Lawick-Goodall, 1971; Kruuk, 1972; Henschel & Skinner, 1991).
Clan wars here were identified, and distinguished from other encounters with
alien hyenas, when multiple rushes and attacks were exhibited by both sets of
participants. Whenever permitted by observation conditions during clan wars,
observers recorded the identities of all individual Talek hyenas participating.
Observers also recorded the identities of Talek hyenas in the front lines of an

attack as well as those in the rear. Observers divided participating Talek hyenas

18

into those seen in a front line in at least one attack per clan war and those never
observed in a front line at all during a particular clan war.

Similarly, observers recorded information on participation and group
leadership during each border patrol observed. During a border patrol, several
hyenas maintain body postures indicating great excitement as they move along,
or to, a territorial boundary where they engage in high rates of scent-marking
behavior and socially facilitated defecation in “latrine” areas (Kruuk, 1972; Mills,
1984; Henschel & Skinner, 1991; Sillero-Zubiri & Gottelli, 1992). Here the leader
of a border patrol was defined simply as the individual at the front of a moving
queue of hyenas. Multiple individuals were often observed in the lead during a
single border patrol. l divided participating hyenas into those seen in the lead at
least once and those never observed to lead at all during a particular border
patrol. Records from border patrols also included information on whether or not
each participating hyena engaged at least once in scent-marking via deposition
of “paste” from anal scent glands (called “pasting” by Kruuk, 1972) or defecation
in latrine areas along borders of the Talek iclan’s territory. A latrine can be
identified by the presence of several piles of feces deposited by multiple hyenas
at the same spot. Hyena feces often turn white after drying, making latrines
conspicuous for many weeks after deposition (Kruuk, 1972). Here I calculated
the proportion of individuals of each sex present at each border patrol that either
scent-marked or defecated at least once before the group began to disband.

Furthermore, in a subset of observed border patrols, all occurrences of pasting

19

behavior by all participating hyenas were recorded, and that I could therefore
compare individual pasting rates for male and female participants.

In addition to large groups of intruders found near territorial boundaries
during clan wars, lone alien hyenas or small groups of aliens may occur deep
within another clan’s territory. I recorded all behavioral interactions between
Talek hyenas and intruders during every observation session in which at least
one intruder and one Talek resident were present simultaneously. In these
observation sessions, it was usually possible to record identities of all Talek
hyenas present. Observation sessions in which Talek residents encountered
aliens were included in analyses of aggression below only when all Talek hyenas
present could be identified, and also when all alien hyenas present could be
sexed.

In each session where resident and alien hyenas were present
concurrently, I assumed there was the potential for each resident to direct
aggression toward each alien. These opportunities for aggression were called
encounters. The total number of encounters per session was calculated as the
product of the number of Talek residents present times the number of alien
hyenas present in that session. Thus a session in which five Talek hyenas met
up with two aliens resulted in ten encounters, each of which might possibly have
involved aggression. When both aliens and residents occurred together, for each
Talek hyena present I calculated the hourly rate at which it directed aggression
toward aliens of each sex. To correct for the number of potential alien targets,

hourly aggression rates were divided by the number of male or female aliens

20

present. Aggressive behaviors directed at aliens were called attacks and
included threats, displacements, chases, and bites (Kruuk, 1972). Submissive
behaviors included flattening the ears back against the head, head-bobbing,
backing away or retreating, grinning, and giggling (Kruuk, 1972). Submissive
behavior not preceded by aggressive behavior emitted by another hyena was
called “unsolicited appeasement.”

Spatial data and statistical analyses.— Precise geographic locations of all
territorial behaviors and all encounters with intruders were recorded by reference
to landmarks or by latitude and longitude determined with a Magellan Meridian
XL Global Positioning System (GPS) unit. From aerial photographs (1220,000
scale) of the study area, I generated a detailed digitized map of the northeastern
portion of the Masai Mara National Reserve and geo-referenced this map to the
GPS field data. On this digital map, I placed locations of all clan wars and group
scent-marking sites and defined the borders of the Talek clan’s territory as the
lines best incorporating these points. Thirteen adult Talek females wearing radio-
collars were tracked multiple times each week during 1996-1998, and their
locations were also plotted on the digitized map. Approximate boundaries of
territories occupied by neighboring clans, other than boundaries shared with the
Talek clan territory, were determined by recording locations of clan wars elicited
by the sound playbacks conducted by Ogutu & Dublin (1998).

Chi-square tests were used to compare participation in border patrols and
clan were by resident males and females. I also used chi-square tests to

compare proportions of encounters in which resident animals either attacked

21

intruders or exhibited no aggression toward them. Hourly rates of dyadic
aggression against aliens of each sex were calculated for all Talek hyenas.
Aggression rate data were not normally distributed, so these were analyzed with
the Kruskal-Wallis test for variance among groups and Dunn’s test for multiple
comparisons. Differences between groups were considered significant when p <

0.05.

RESULTS

During the study period, field researchers observed Talek hyenas engage
in 31 clan wars and 35 border patrols involving intensive scent-marking via
pasting or socially facilitated defecation at latrine sites (Figure 2-1a). Borders of
the area marked and defended by the Talek clan, as inferred from the data in Fig
1a, are drawn in Figure 2-1 b, along with locations of 133 interactions between
Talek hyenas and aliens that could not be classified as clan wars, and 160
sightings of aliens alone. The Talek River, which formed the northern boundary
of the Masai Mara National Reserve in this area, also formed the northern border
of the Talek clan’s territory. A tributary of the Talek River formed much of the
eastern border, and a large seasonal water course called OI Keju Rongai formed
much of the southern border of the Talek territory. The western border was not
marked by a water course. The area advertised and defended by the Talek clan
thus covered approximately 61 kmz. On average, 94.4 i 1.9% of all 3130 sites at
which 13 radio-collared Talek females were found during 1996-1998 fell within

the Talek clan territorial boundaries (Figure 2-1c). Finally, territories of

22

.88: 538 a 3.60 .3 BESS 8°59: mam: 852.2% em; SE9 5.23

we saw a: on as 85.58 SE2 no 8:85.82 285. 85% mean Susan: .3 335%
mweoueewu no mwewneaon Qwezxoaaw we“ 2 eo=w_we e_ BEES xw_w._. we“ no 5:82 05 ©5265,.

wawmwe mice we“ Anv ”89-83 5232 nwxowz-o_nwe new; mw_wE£ xw_w.r mp 22:; .w meonwos 85
A3 ӣ52me xw_w._. .3 nwunwnwne: mew=w Co 35:36 om: new 52:5. xw_w._. 9: Eng; mwewE ew__w new
9:02me xw_w.r ewwzewn Amewz, 5.0 0:620wa Ewuenooew m9 ho meonwog me_e_w.eoo .wp .9“. e_ wuwn
E0: note: mwewnenon _we2E9 xw_w._. By 6.0wa ewneon new mew; 5.0 no meonwos new n2w£n£
$228552, £5, 12.22 on. Co 5an Ewumwwetoe we“ 3 meesoem meE anszweow ._.-N 2:9...

 

was”. 38. .0

Ex: x. o

  
  

‘U'

 
   

\ I

.
lm§ww .1 .moEoo

<_2<Nz<._.

     

 

3:02me :22.
fi— meewnezooew mew=< . m
zen mew=< o

 

 

o .w
I
:0 I n.
D.
D
o Immanul Eu 0 I D /
D D D. <
20wa .wneom I o a <

 

2w; $6 0

 

 

 

 

 

23

neighboring hyena clans bordered the territory of the Talek clan in all compass
directions, and clan wars were observed between Talek hyenas and members of
all neighboring clans (Figure 2-1d).

In 9 clan wars, observers were able to record the identities of all
participants as well as all occurrences of agonistic behaviors, and observers
were also able to ascertain which individuals were in the front lines of each
attack. Both adult females and males participated in all 9 of these clan wars. Of
all Talek females in these clan wars, 38 (75%) were observed at least once in the
front line of an attack whereas only 10 of 27 (37%) adult males were observed in
front lines. Thus females were significantly more likely to lead attacks than were
males (12: 10.28; d.f. = 1; P < 0.01; Figure 2-2). Of all adult females participating
in clan wars, 36% were pregnant, 58% were lactating, and 6% were neither
pregnant nor lactating.

In the 23 border patrols for which observers knew the identities of all Talek
hyenas participating, only adult females participated in three patrols, only adult
males participated in two patrols, and 18 patrols involved participants of both
sexes. All mixed-sex border patrols were always led by females. Of all females
that participated in border patrols, 40 (42%) were observed in the lead at least
once during a border patrol, whereas only 7 (15%) of 47 male participants were
seen leading a patrol. Thus, as with clan wars, border patrols were significantly
more likely to be led by females than by males (12: 10.42; d.f. = 1; P < 0.01;
Figure 2-2). There were no sex differences in the proportions of individuals

present during border patrols who either pasted along the way (12 = 0.42; d.f. = 1;

24

70 t" I Females
6O _ El Males
5O " 96

(D
O
I

% Hyenas seen leading
_. N 4:.

O O

l I

DO

 

 

 

 

 

 

 

Border patrols Clan wars

Figure 2-2. Percent of resident immigrant males and resident
females present during group territorial behaviors that were seen
leading border patrols or clan wars. Sample sizes indicate
cumulative totals of hyenas present in all border patrols and clan
wars included in this analysis.

25

P > 0.10) or defecated in latrine areas (78’ = 0.70; d.f. = 1; P > 0.10). However,
females tended to paste during border patrols at slightly higher hourly rates than
did males (T = -1.95; d.f. = 11; P = 0.07; Figure 2-3).

Excluding clan wars, observers found 458 alien hyenas in the Talek
territory during 293 different observation sessions. Mean intruder group size was
1.6 i 0.1 individuals (n = 293). Talek hyenas were present with aliens in only 133
of those 293 observation sessions (Figure 2-1 b). Thus Talek hyenas may have
detected fewer than half of all known intrusions by one or more aliens. Mean
intruder group size when aliens were detected by Talek residents was 1.4 i 0.1
individuals (n = 133; Figure 2-4). An ungulate carcass was present during 42
sessions with intruders, and no food was present during 91 sessions. All Talek
hyenas were individually identified and sexes of all aliens were determined in 107
of 133 sessions with intruders, totaling 46.5 observation hours. Only 13% of
sexed alien intruders were females. Intruder groups contained only males in 93
sessions, only females in 10 sessions, and mixed-sex groups in 4 sessions.
During these 107 sessions, observers recorded 384 encounters between
individual Talek adults and aliens, including 202 encounters involving 55 resident
females, and 182 encounters involving 47 resident immigrant males.

Talek residents attacked aliens, or aliens directed unsolicited
appeasement behavior toward residents, in 68 sessions. In 34 sessions, neither
residents nor intruders showed aggression or submission. In five additional
sessions, Talek hyenas exhibited submissive behavior to aliens, but showed no

aggression, nor did aliens show submission to residents. In each of these five

26

12

1.5'

1.0'

0.5"

 

 

 

0.0

# Pastes / individual / border patrol

Females Males
Sex

Figure 2-3. Rates of pasting behavior during border patrols exhibited by
resident Talek males and females. Sample sizes indicate numbers of
border patrols.

27

% Alien groups observed
—-¥ N O) A 01 O) \I CD
0 O O O O O O O

O

   

 

F118 98
B Aliens only
- l Aliens encountering
_ Talek residents
_ 6 3
_ n =1. 4 l:

 

 

 

12345678-12

Number of hyenas in alien groups

Figure 2-4. Sizes of 293 alien groups observed within the Talek clan
territory. Alien hyenas met Talek hyenas in 133 cases. Numbers of
groups are indicated above bars.

28

sessions, only one resident hyena was present (n = 4 Talek males; 1 Talek
female), and in four cases that individual was outnumbered by aliens.

Resident Talek females were significantly more likely to attack alien
females than alien males ([3 = 11.42; d.f. = 1; P < 0.001; Figure 2-5). Although
resident females appeared more likely than did resident males to attack intruding
females, this difference was not statistically significant ([2 = 2.91; d.f. = 1; P =
0.088). Similarly, although resident males appeared somewhat more likely to
attack alien males than alien females, this difference was not statistically
significant (,1? = 0.735; d.f. = 1; P > 0.10). Resident males were significantly more
likely than were resident females to attack alien males (1’2 = 9.25; d.f. = 1; P <
0.005).

Mean hourly rates of aggression directed by individual Talek hyenas
toward aliens are shown in Figure 2-6. The individuals exhibiting the lowest rates
of aggression were resident Talek males confronting alien females. Rates of
aggression varied among groups (Kruskal-Wallis test statistic = 7.816; d.f. = 3; P
= 0.05; Figure 2-6) and were significantly higher when resident immigrant males
encountered alien males than when they encountered alien females (Dunn’s test
statistic Q = 5.64; P < 0.01; Figure 2-6). In contrast, rates of aggression directed
by Talek females toward aliens did not vary with intruder sex (Dunn’s test statistic
Q = 0.171; P > 0.05; Figure 2-6). There was no significant difference between
attack rates that resident females and resident-males directed toward alien males

(Dunn’s test statistic Q = 1.586; P > 0.05; Figure 2-6). However, resident females

29

50 ' I'— l Alien females
E Allen males

 

165

 

 

184

% Residents attacking

10" I
I

Females Males

 

 

 

 

 

 

 

 

Sex of attacking residents

Figure 2-5. Percent of resident Talek males and females that attacked
alien males or females at least once during an encounter. Sample sizes
indicate total number of encounters (resident-alien pairs).

30

 

 

 

 

 

 

 

 

2'0 I Alien females g3
E; U Alien males 1'
L“ 1.5- * ——
8 15 40 -I
S I!

\ 1.0-

(D

x

0

£9

E 0.5-

:u: 14

0.0 '

 

 

 

 

 

 

Females Males
Sex of attacking residents

Figure 2-6. Hourly rates of attack on alien males and females by resident
Talek females and resident immigrant Talek males. Sample sizes indicate
numbers of individual Talek hyenas.

31

attacked alien females at higher rates than did resident males (Dunn’s test

statistic Q = 5.033; P < 0.01; Figure 2-6).

DISCUSSION

Intrusion pressure and defense of borders.— The territorial behavior
exhibited by spotted hyenas varies among study populations. Vigorous and
frequent territorial behavior in Crocuta appears to occur where hyena intrusion
pressure is intense or when specific carcasses are contested (Kruuk, 1972;
Henschel & Skinner, 1991). In areas characterized by very low hyena density,
clans occupy exclusive ranges but there is little evidence of intrusion pressure or
contested carcasses in border areas, and both clan wars and border patrols tend
to be rare (Tilson & Henschel, 1986). Frank (1986b; 19863) monitored the Talek
clan from 1979 to 1983 but observed only three clan wars and rarely saw border
patrols or found latrines. Frank concluded that Mara hyena clans occupy ranges
with diffuse limits separated by wide areas of “no-man’s-land.” By contrast, the
1988-1998 data presented here suggest that the boundaries of the Talek territory
are well-defined (Figure 2-1), that the areas Frank (1986b; 19863) described as
no-man’s-Iand are occupied by members of neighboring clans, and that territories
of Mara hyenas form an uninterrupted mosaic in suitable habitat as do those of
Crocuta inhabiting other areas of eastern Africa (Kruuk, 1972; Hofer & East,
1993b). The differences between my results and those obtained by Frank
(1986b; 1986a) can probably best be explained by differential observation efforts:

Frank and his assistants observed Talek hyenas for 517 d whereas we watched

32

them for 3120 d. Rates at which clan wars were seen in both studies were in fact
quite similar when corrected for total time spent observing Talek hyenas.

In the current study, alien hyenas were observed frequently in the Talek
area, and intruders generally met hostile receptions from the Talek residents
detecting their presence. Most of the alien hyenas that could be sexed in the
Talek territory were males, a finding consistent with the earlier observation that
Talek males, but not females, regularly make exploratory excursions into the
territories of neighboring clans in preparation for dispersal (Smale et al., 1997;
Holekamp & Smale, 1998b). If males originating in neighboring clans behave like
Talek males, then many male intruders in the current study were probably visiting
Talek to assess prospects for immigration into the study clan. Only 13% of the
alien intruders into the Talek territory were females, which stands in marked
contrast to the finding by Henschel & Skinner (1991) that 71 - 89% of intrusions
into their clan in southern Africa involved females. Although female Crocuta have
occasionally been known to emigrate from their natal clans, successful female
immigration into an existing clan has never been observed. Instead, dispersing
females either form a new clan or lead nomadic lives (Mills, 1990; Holekamp et
al., 1993). Thus heavy intrusion pressure from females in some populations is
more likely to be associated with foraging than with dispersal (e.g., Hofer & East,
1993c), and in the Talek area, xenophobic behavior toward aliens may help
minimize intrusion pressure by foragers from neighboring clans.

Sexually dimorphic territorial behavior.— I observed marked differences

between adult immigrant male and adult female hyenas with respect to

33

leadership during cooperative territorial activities, marking of territorial
boundaries, and aggression toward intruders. Females were more likely than
males to lead border patrols and clan wars, and resident females tended to
scent—mark along territorial boundaries at higher hourly rates than did adult
males. Thus my results support and extend the earlier findings of Henschel &
Skinner (1991), who found that resident females scent-marked more frequently
than males and invested more time in territorial activities than did males. Our
leadership data also suggest that female Crocuta are willing to assume more
risks than males during territorial defense. Interestingly, the fact that Talek border
patrol groups sometimes contained only resident males demonstrates that group
territorial defense is not strictly the domain of females. Nevertheless, the fact that
most cooperative territorial activities are led by female spotted hyenas sets this
species apart from wolves, jackals, dingos, and other gregarious canids in which
males generally take the lead in territorial defense (reviewed in Holekamp et al.,
2000)

The sexually dimorphic territorial behavior of Talek hyenas was not limited
to leadership, but also expressed itself in probabilities and rates of attack that
varied based on the sex of the resident and that of the intruder. During
encounters between residents and aliens inside the Talek territory, residents
were more likely to attack same than opposite-sex intruders. Furthermore,
although hourly rates of female aggression did not vary significantly with intruder
sex, resident males directed higher hourly rates of aggression toward alien males

than toward alien females. Thus spotted hyenas exhibit sexually dimorphic

34

territorial behavior conforming to predictions of game theoretical models
suggesting that individual group members should engage in territorial defense in
proportion to both their need for resources available in the defended space and
the severity of the threat to those resources posed by specific types of intruders
(e.g., Hammerstein, 1981; Temeles, 1989; 1990). In this regard, spotted hyenas
also resemble lions, canids and other carnivores (McComb et al., 1994; Grinnell
et al., 1995; Heinsohn et al., 1996).

Sex-specific selection pressures.— The sexually dimorphic territorial
behavior observed in the present study suggests that ultimate explanations for
territoriality may differ for male and female Crocuta, as they do in lions (Packer et
al., 1990; Grinnell et al., 1995; Heinsohn & Packer, 1995). Possible explanations
for territorial behavior in hyenas include defense of resources such as food or
potential mates (e.g., Kruuk & Macdonald, 1985) and offspring defense (e.g.,
Wolff 8. Peterson, 1998). Since resident immigrant male hyenas that have not yet
fathered any cubs in Talek may attack alien male intruders (Smale et al., 1997),
the offspring defense hypothesis seems a less likely explanation for male
territoriality than does resource defense. Furthermore, access to females limits
male reproductive success in this species (Engh et al., unpubl. data), and the fact
that rates of aggression by resident males were significantly higher when they
encountered alien males than females (Figure 2-6) suggests that the most likely
explanation for male aggression toward male intruders is defense of potential

mates.

35

Females are the limiting sex in this species, so female territoriality is more
likely explained by hypotheses suggesting defense of offspring or food resources
than one suggesting defense of mates. Wolff 8 Peterson (1998) proposed that
female territoriality against other females evolved to protect immobile altricial
young from female infanticide. Because hyenas cubs reside at dens until 8 - 12
months of age (Hofer 8 East, 19933; Holekamp 8 Smale, 19983), female
territoriality might promote defense of immobile young. lnfanticide has been
documented by conspecifics of both sexes in Crocuta (e.g., Kruuk, 1972; Hofer 8
East, 1995), and offspring defense has been shown to be a powerful selective
force promoting group living in other carnivores (Owens 8 Owens, 1984; Packer
et al., 1990), so similar selective forces may have shaped territorial behavior by
female Crocuta.

In lions, lnfanticide by adult males accounts for 27% of all cub mortality in
the first year of life (Pusey 8 Packer, 1994). Lions and spotted hyenas occur
sympatrically in many African ecosystems, utilize virtually identical food
resources, and both live in multi-male, multi-female groups with fission-fusion
structures and male-biased dispersal (Packer, 1986; Pusey 8 Packer, 1987;
Packer et al., 1990; Grinnell et al., 1995). Female lions are most gregarious when
they have dependent young and they cooperatively defend young against
infanticidal males (Packer et al., 1990; Pusey 8 Packer, 1994). Male lions are
approximately 50% larger than females, and this extreme sexual dimorphism
probably prevents lone females from confronting males with aggressive defense

of young

36

In contrast to lions, male spotted hyenas are roughly 10% smaller than
females (Matthews, 1939; KruUk, 1972; Mills, 1990). Thus, although infanticidal
behavior may be directed at hyena cubs by conspecifics of either sex (Hofer 8
East, 1995), lone female hyenas can effectively defend their offspring against
either male or female intruders detected near their cubs. Clan wars did not occur
near dens. Females participate in territorial defense regardless of their
reproductive state, and to date only resident Talek hyenas have been observed
directing aggressive behavior (including infanticidal aggression) toward cubs of
Talek females. Thus it seems unlikely that female Crocuta cooperatively defend
their group territories exclusively to protect their offspring from conspecific attack,
although this may certainly be a beneficial consequence of territorial behavior
Serengeti hyenas ignore intruders in transit through their territories, but attack
commuters feeding at kills and engage in wars with neighboring clans (Hofer 8
East, 1993c). In the present study I found that Talek females attack with equal
frequency all intruders posing a potential threat to the food supply within their
territory. Females may be more likely to attack intruding females than males
because the former potentially represent a more severe long-term threat to the
food supply. That is, if resident animals reveal any weakness in their territorial
defense, alien females may recruit allies in their home clans and attempt to take
over the area, as has been observed during clan fission events by Mills (1990)
and Holekamp et al. (1993). Defense of food resources has previously been
proposed as the primary function of territorialin in spotted hyenas (Kruuk, 1972;

Henschel 8 Skinner, 1991). Indeed, it appears that natural selection has favored

37

female Crocuta to maintain boundaries of a territory supporting enough herbivore

prey to feed themselves and their young throughout the year.

38

Chapter 3

DENNING BEHAVIOR

INTRODUCTION

In the fission-fusion societies of spotted hyenas, dens have important
protective and social functions (Kruuk, 1972; Mills, 1983; East at al., 1989;
Cooper, 1993; Hofer 8 East, 19933; Holekamp 8 Smale, 1998a). The primary
protective benefit of dens is as a shelter for cubs. Hyena den holes have
constricted entrances too narrow for adult hyenas or other large carnivores to
enter (Kruuk, 1972; Hofer 8 East, 19933; Cooper et al., in prep), and cubs can
escape from approaching predators such as lions by fleeing underground (East
et al., 1989; Holekamp 8 Smale, 1998a). Because many group members
regularly visit dens, dens are important in the social development of cubs. Cubs
interact with each other and with adults there, and dens also serve as meeting
points for adults that othenlvise spend much of their time alone (Holekamp et al.,
2000). Furthermore, within the fission-fusion society of a hyena social group,
dens are important locations for the assembly of subgroups, such as hunting
parties and territorial border patrols (Kruuk, 1972).

Cubs of spotted hyenas, as well as brown hyenas (Hyaena brunnea), live
at dens longer than cubs of any other carnivores (Owens 8 Owens, 19793; Mills,

1990). Two types of dens, natal and communal dens, have been described for

39

spotted hyenas (Kruuk, 1972; East et al., 1989; Mills, 1990; Holekamp 8 Smale,
19983). Most Crocuta cubs start life at an isolated natal den used by only one
mother for shelter of a litter consisting of one or two cubs (East et al., 1989;
Holekamp et al., 1999b). Hyena natal dens usually have only one or two
underground entrances (East et al., 1989) and were originally excavated by
aardvarks or warthogs, although the dens may be modified by hyenas upon
occupation (Kruuk, 1972). Ned-natal cubs spend much of their time inside the
den, and rarely emerge unless their mothers are present (Kruuk, 1972; East et
al., 1989; Holekamp 8 Smale, 1998a). Thus a cub spends the first weeks of life
interacting primarily with its mother and its Iittennate when one is present.
Within the territory of any hyena social group, or “clan” (Kruuk, 1972),
there is usually one communal den utilized concurrently by several litters of
similar ages (Kruuk, 1972; Mills, 1990) ranging up to 15 mo of age (Mills, 1990).
Communal dens often have multiple entrances leading to a network of tunnels
(Kruuk, 1972; Mills, 1990), and as with natal dens, spotted hyenas modify
existing dens but do not excavate them (Kruuk, 1972). Mothers usually transfer
their cubs to the clan's communal den 1-5 weeks after birth (Kruuk, 1972; East et
al., 1989). With this move from the isolation of the natal den, young cubs enter a
complex new social environment in which they are exposed to hyenas of various
ages, social ranks, and relatedness. Mothers return often to the den to nurse
their dependent offspring, and other adult females, adult males, and subadults
also visit frequently. At the communal den, young cubs begin to learn their social

ranks in the clan’s linear dominance hierarchy, and older clan members learn the

40

identities and ranks of new cubs born into the clan (Holekamp 8 Smale, 1993;
Smale et al., 1993; Holekamp 8 Smale, 19983; Engh et al., 2000).

The move to the communal den appears crucial for a cub’s social
integration into the clan, but in order for 3 cub to make a successful debut at the
communal den, its mother must safely transfer the cub from the natal to the
communal den (East et al., 1989; Holekamp et al., 2000). Cubs are vulnerable to
above-ground dangers during the move, when mothers shepherd or carefully
carry them one at a time (East et al., 1989; Holekamp 8 Smale, 1998a). A
shorter distance over which to move cubs between natal dens and communal
dens might therefore be expected to decrease mortality risk during den moves,
and females might benefit from using natal dens located close to the communal
den. Because high-ranking females are more likely to occur in the vicinity of the
clan’s communal den than are low-ranking females (Chapter 5), high-ranking
females may be able to use natal dens closer to the communal den than can Iow-
ranking females.

- My objectives here were to describe patterns of communal den use by
hyenas belonging to one well-studied clan during a 10-year period, and to
examine locations of natal dens relative to those of communal dens. I present
descriptions of communal den locations, and also describe distances of den
moves, periods of den use, and frequency of re-use of dens. I consider reasons
that hyenas might initiate communal den moves, and specifically I examine the
hypothesis that ectoparasite load at dens might prompt hyenas to vacate a den. I

then examine placement of natal dens with respect to the communal den to

41

determine whether there are rank-related differences in locations of isolated natal
dens and in the distances that females move their cubs when transferring them

to the communal den.

METHODS

The study clan, which usually contains 70 to 80 hyenas, defends a
territory encompassing 62 km2 (Boydston et al., in press) in the Talek region of
the Masai Mara National Reserve, in the northeastern part of the Serengeti-Mara
ecosystem. This is an area of open rolling grassland interspersed with seasonal
creek beds lined with bushes. Large concentrations of several resident ungulate
species graze this area year round, and these are joined for three or four months
each year by large migratory herds of wildebeest (Connochaetes taun'nus) and
zebra (Equus burchelll) from the southern part of the Serengeti.

Each Talek clan hyena was individually identified by its unique spots and
other natural marks, and sexed based on the dimorphic glans morphology of the
erect phallus (Frank et al., 1990). Social ranks of all clan members were known
based on their positions in a matrix of outcomes in dyadic agonistic interactions
(Smale et al., 1993). Maternal kin relationships were known for all natal Talek
clan animals, as described previously (e.g., Holekamp et al., 1993). Talek
females usually give birth to 1 or 2 cubs, and cubs reside at dens until about 8
mo of age. Cub birth dates were estimated to :I: 7 days based on cub pelage,

size, and other aspects of appearance when first seen above ground.

42

The use of dens by Talek hyenas was monitored from June 1988 to
January 1998. Here a single den included all the tunnel entrances that hyenas
were seen using for a continuous period of time within 3 200 m radius. Precise
geographic locations of all dens that Talek hyenas occupied were recorded by
reference to landmarks or by latitude and longitude determined with a Magellan
Meridian XL Global Positioning System (GPS) unit. From aerial photographs
(1 :20,000 scale) of the study area, I generated a detailed digitized map of the
northeastern portion of the Masai Mara National Reserve and geo-referenced
this map to our GPS data. On this digital map, I placed locations of all dens.

A communal den (CD) was identified as a den shared by three or more
females for rearing cubs of any ages. A CD was said to be “active” during the
time that litters of cubs resided at the site. A den, once utilized by Talek hyenas,
was considered to be “inactive” when it was unoccupied by cubs. I defined a
natal den (ND) as a den other than the active CD that was used by one or two
mothers for rearing neo—natal cubs. East et al. (1989) used the term “birth den” to
describe a natal den, but here the term “birth den” describes the den where a
litter was believed to have been born, whether this den was an isolated natal or
the communal den.

Litters were included in the study only if they were found at a natal den
within 30 days of birth or if observers saw them within 2 wk of birth at the
communal den. The den where a litter was first observed was assumed to be the
birth den. Based on observations of mothers at dens and other locations, and on

observations of cubs, l determined the dates that females moved in and out of

43

dens. Most den moves were not observed, and the move date was recorded as
the date half-way between the last sighting of a female or her cubs at the old den
and the first sighting of the female or her cubs at the new den. However, all dates
of den moves were known within :1: 11 days. Using the GIS software ArcView,
linear distances were measured between all birth dens and the active CD, and
between all consecutive dens that a female used for rearing a litter until moving
the litter to the CD. Also, the number of times that a female moved her litter was
determined by counting the number of different dens she used between leaving
the birth den and moving her litter to the CD. For example, for a female that gave
birth at a private ND and moved her cubs after 4 wk to the CD, the number of
moves was one. In this example, the distance of the birth den to the CD would
equal that of the total distance moved since the female used only one ND. By
contrast, for a female that gave birth at a private ND and moved her cubs after a
couple of weeks to another ND, and then moved her litter to the CD, the total
number of moves would be two. The total distance moved would be the sum of
the distance of the first move from the birth den to the second ND plus the
distance of the move from the second ND to the CD.

In order to examine ectoparasites in hyena dens, I collected dirt from the
entrances to dens, from July 1997 to January 1998. Samples of between 100—
200 g dirt were collected, usually at midday when adult hyenas were not present
at dens. Dirt was scraped into a plastic container from different spots around the
mouth of the den up to about a half meter into the den tunnel, and from different

entrances if hyenas were using more than one den entrance concurrently at a

particular site. Dirt samples were not taken if the ground was still wet from a
recent rain. Dirt samples were weighed and carefully searched for the presence
of ticks and any other invertebrates visible to the naked eye, and some ticks were
preserved in ethanol for identification. The number of ticks was counted and
expressed as the number per 100 g of dirt.

Statistics were calculated using SYSTAT 8.0. All mean data are presented
plus or minus their SE. To compare the number of ticks in active CDs vs. inactive
CDs, I used the Mann Whitney U test for non-normal data. To examine variation
in the number of ticks depending on time since hyenas had vacated the dens, I
used the Kruskal-Wallis non-parametric test statistic. Where rank was the
independent variable, I used Spearman’s rank correlation coefficient for
regression analyses to examine the effect of rank on the dependent variable of
distance moved and numbers of moves. In these regression analyses, data for all
litters were averaged for each individual female in order to control for different
numbers of litters produced per female. Pearson’s correlation coefficient was

used for 3 comparison between continuous variables.

RESULTS

Communal dens.— Hyenas used a total of 57 different dens for communal
denning between 1988 and 1998, and an additional 41 dens were used only as
natal dens (Figure 3-1). All dens were within 1 km of the bed of a seasonal creek

or river, and 67% of dens were within 250 m of a seasonal creek or river. Thus

45

 

I Communal dens
0 Natal dens

 

 

A
N

 

 

 

Figure 3-1. Communal dens (CDs, filled squares, n = 57) and natal dens
(NDs, open circles, n = 41) used by Talek hyenas from 1988 to 1998.
Some 003 were also used as NDs, but the NOS shown here were never
used for communal denning.

46

most dens were situated close to a water course which could provide shade and
shelter.

Roughly half (49%) of the 57 different CD locations were used only once
during the study period, but 29 dens were repeatedly used for communal denning
(Figure 3—23). On average, hyenas maintained a CD at a particular site for a
continuous period of 1.02 :I: 0.11 mo (n = 132 CD uses). Several communal dens
were only temporary or transitional, as hyenas had 22 CDs that they occupied for
a week or less before moving again. Talek hyenas maintained only 4 CDs
continuously for more than 4 mo, and the longest period of continuous use of one
den as 3 CD was 8.1 mo (Figure 3-2b). Den holes persisted for years, and Talek
hyenas reused dens with varying frequency. Of sites that were used more than
once, the average time elapsed between consecutive uses as a CD was 13.4 i
2.0 mo (n = 73 CD uses at 29 different dens). The maximum period between
consecutive uses of a den as a CD was 65 mo, with the first occupation of this
site observed in 1992 and the second in 1997. Two dens were used in both the
first and last years of the study period, as well as several times in between.
Despite the abundance of dens in the study site, only a few dens which were
occupied several times and over periods of several months stood out as
particularly popular sites for Talek clan 005 (Figure 3-3).

Den moves.— Distances between CDs that were used consecutively were
generally short, averaging only 1.5 :l: 0.1 km (n = 131 moves). Ten moves
involved relocating 3 to 5 km from the previous CD. The maximum distance of 3

den move was 6.2 km, and no other den moves involved distances greater than

47

50

40

30

% Dens

20

10

 

 

2 4 6 8 10 12
# Times den used as CD

80 r
94
7o - B

60
50
40
30
20
10

% Communal den uses

 

 

1 2 3 4 5 6 7 8 9
Duration of CD use (mo)

Figure 3-2. (a) Number of times that Talek hyenas repeatedly used
particular CDs from 1988 to 1998. Numbers over bars indicate the
number of dens in each category (total = 57 dens). (b) Maximum
number of months that a given CD remained active for a continuous
period. Sample sizes over bars indicate number of times CDs were in
use (n = 192 CD uses).

48

 

 

Talek River

' ° . ' . 2
69 2 . .3. . g 20 "
o o 2 0
g o 7 20 3 2 9’5 5 30 2
o O 1 O 7 6
O . Q C
o . 2 a
2 ° .
2 3
it 1 g
Total # mo dens ' -
used as CDS # times den used
/
- < 1 mo 3
o 2 -4 mo 0\ A
9 5 ' 8 ”'0 Dot size indicates 0 1 2 Km
® 9 _ 15 mo #mo den used E N

 

 

Figure 3-3. The 57 sites that were used for communal denning by Talek hyenas
between 1988 and 1998. The size of each dot reflects the total amount of time
that a den was used during the entire study period. Numbers labeling dens
indicate the number of times each site was reused as a communal den. Dens
that have no number labels were used only once.

49

5 km. The timing of CD moves varied little from month to month (Figure 3-4), with
5 to 12% of moves (n = 131) occurring in any given month of the year. Although
den moves did not vary significantly with season, the months with the fewest den
moves (Feb.-Apr.; Figure 3-4) coincided with the months during which Talek
hyenas experience an annual trough in the number of births (Feb-May:
Holekamp et al., 1999b). Furthermore, November was the month with the most
den moves (12%), and November was also the month with the highest mean
number of births (Holekamp et al., 1999b). Moves between CDs were observed
on 15 occasions between 1989-1995. Some of these den moves occurred over
the course of a few days with different groups of females moving their cubs at
different times. In 73% of these moves, the highest ranking female with cubs at
the den was among the first to move her cubs.

The number of ticks per 100 g of dirt collected from dens varied
significantly with the length of time since hyenas had vacated the dens (Kruskal-
Wallis H = 8.157, P = 0.043, d.f. = 3; Figure 3-5). Ticks did not appear to be
associated with hyena dens while hyenas occupied those dens (Figure 3-5).
Dens currently in use by hyenas had very few ticks (0.3 :I: 0.3 ticks/ 100 g dirt; n =
8 samples from 4 dens), whereas dens that hyenas had vacated for at least one
month had significantly more (Mann-Whitney U = 40.5, P = 0.013, d.f. = 1), with
19.0 :t 8.5 ticks/ 100 g (n = 14 samples from 11 different dens). Only three dens
were sampled both while the den was in use as a CD, and again within 3 days of
the hyenas vacating the site. Samples from two of those three dens contained no

ticks at all, either before or after hyenas moved out. The third den had no ticks 2

5O

.1.
.h
l

16

A A
o N
l r
F3
K3

% CD moves

 

l

I I

l

I l

 

O N uh O} on
I

1234567 89101112
Month

Figure 3-4. Percent of total CD moves that Talek hyenas made
during each month of the year (n = 131 moves). Sample sizes

over bars indicate the number of moves observed during each

month.

51

 

 

 

 

 

 

 

 

 

40 r .9
"-6
c»
8 30 -
B V__
x
o
'1: 20 '
41:
3:? 10 .. 3
2 10 ' +
8
0 A l I I J
active <1 mo 2-4 mo >6mo
CD

Time since hyenas vacated den

Figure 3-5. The mean number of ticks (with SE bars) per 100 g of dirt collected
from dens varied significantly with the length of time since hyenas had vacated
the dens. Sample sizes indicate the number of dirt samples collected for each
time period. A total of 14 different dens, including 2 that were only used as NDs,
are represented here.

52

wk before hyenas moved out, but a day after hyenas moved out, there were 16
ticks/100 9. Thus, in one case out of three, there was there evidence that an
increase in ectoparasites population density might have contributed to the
initiation of a den move, but in the other two cases, other factors presumably
prompted the den move.

Most ticks found in dens were soft-bodied ticks, Omithodorus moubata,
that were several mm in body length up to nearly 1 cm in body length. Ticks that
could not be positively identified as O. moubata were found in only three
samples. These other soft-bodied ticks were < 1 mm in body length and were
sometimes seen clinging to the ventral side of large 0. moubata, and may have
been larval stages of O. moubata. Although hard-bodied ticks are often found on
immobilized Talek hyenas, O. moubata have never been found on these hyenas
(Holekamp et al., unpubl. data), and they instead appear to specialize on
warthogs as hosts (Walton, 1962).

I observed fleas in only 2 of 14 dens, neither of which was occupied by
hyenas at the time I took the sample. One of these two dens had only 3 few
fleas, but the other appeared to have hundreds hopping around on the substrate,
and this also happened to be the den with the greatest density of ticks of all
those sampled (113 per 100 g of dirt). At the time of sampling, some of the dens
not currently occupied by hyenas were clearly occupied by other animals,
including warthogs and banded mongoose. However, some dens that were
overgrown with vegetation around the holes and had spider webs across tunnel

entrances were probably not in use by any mammals at the time of sampling.

53

Birth dens.— Birth den locations were observed for 79 litters born to 39
different mothers (2.0 :1: 0.3 litters per female). Sixteen litters (20%) were reared
at the CD from birth, and 63 litters (80%) were born at isolated NDs. Of dens
found, forty-one were used only as N03 and never as CDs during the study
period (Figure 3-1). There were few known cases of repeated uses of NOS (n = 3
of 41 sites that were used only as NDs) in contrast to the re-use of 003.
However, 16 litters were born at the active CD and an additional 20 litters used
inactive CD sites as their natal dens.

Isolated NDs were located, on average, 2.1 :l: 0.2 km (n = 63 litters) from
the active CD. Including the 16 litters born at the CD, mothers selected birth dens
that were situated an average of 1.7 :t 0.2 km (n = 39 mothers) from the active
CD when cubs were born. This distance was correlated with social rank, and was
greater for low-ranking females than for high-ranking females (Spearrnan's rank
correlation rs = 0.32, P = 0.05, n = 39; Figure 3-63). Mothers sometimes moved
their cubs to a second ND before moving them to the CD, and the CD location
sometimes changed between the time that a female gave birth and the date on
which she moved her cubs to the CD. Thus the total distance that a female
moved her cubs between birth and the date on which cubs were moved to the
CD did not always equal the distance of the birth den to the CD. Therefore I also
examined the total cumulative distance that a female moved her cubs, before
they arrived at the CD for 37 of 39 mothers for which this information was
available. Females actually moved their young cubs, on average, a distance of

1.6 1: 0.2 km from birth until arrival at the CD. Just as the distance of birth dens to

the CDs was greater for low-ranking females (Figure 3—63), the total distances
over which mothers moved their cubs were greater for low-ranking than high-
ranking females (rS = 0.41, P = 0.02, n = 37; Figure 3-6b). However, the two
different measures of distance (the distance of birth dens to the CD when cubs
were born and the total distance that cubs were moved from birth until reaching
the CO) were strongly correlated (Pearson’s r = 0.79). In addition to moving their
small cubs over greater distances, low-ranking females were more likely to make
intermediate ND moves before transferring cubs to the CD while high-ranking
females were more likely to rear their litters from birth at or near the CD from
birth (rs= 0.39, P = 0.03, n = 37; Figure 3-6c).

Most N05 (89%) were occupied by only one mother at a time, but among
the 63 litters born at NDs, there were three cases of pairs of females sharing ND
for rearing their litters. In one of the cases, the females were known to be sisters,
but in the other two cases, the females appeared to be unrelated (Holekamp et

al. unpubl. data).

DISCUSSION

Talek hyenas occupied dens throughout much of their defended territory
during the study period. Most dens of spotted hyenas inhabiting the Kalahari
were located in or near river beds (Mills, 1990), as was also the case for Talek
hyenas. Although there appeared to be an abundance of dens available for
hyenas to use as N08 or CDs, many of which persisted in the environment for

long periods of time, hyenas repeatedly reused only a few dens. These preferred

55

 

 

Distance of birth den to CD (km)

 

0 10 20 30
Social rank

Figure 3-6. (3) Distances of birth dens to the active CD vs. social
rank. Distances for multiple litters from the same female were
averaged, and thus each dot represents one mother but possibly
more than 1 litter.

56

N 00 #- 01 O) \l
I

 

Distance moved to CD (km)

 

 

O

0 10 20 30
Social rank

Figure 3-6 cont.- (b) Distance that females actually moved their cubs
from the birth den to the CD, averaged for all litters for each female.
Distances moved were not equal to distances of birth dens to the CD,
because some females used intermediate dens before moving to the
CD, and sometimes the CD location changed between the time that a
female gave birth and when she moved her litter to the CD.

57

# Den moves

 

0.0 -~ '
0 1O

 

I I

20 30
Social rank

Figure 3—6 cont. - (c) The number of den moves a female made
between giving birth and transferring her cubs to the CD, averaged for
all litters for each female. If cubs resided at the CD from birth, then
the number of moves was 0; if cubs were born at an ND and then
transferred directly to the CD, the number of moves was 1.

58

dens were usually close to other dens that hyenas did not reuse, and presumably
had certain characteristics other than their locations that made the dens
particularly suitable for occupation.

There was great variation in the length of occupation of any given den for
communal denning, ranging from a day to several months, and there was no
strong seasonal pattern to the timing of CD moves. Spotted hyenas in the
Kalahari moved their communal den on average every 45 days (Mills, 1990).
Despite enormous differences between the Serengeti-Mara and Kalahari
ecosystems, Talek hyenas moved their communal den only slightly more often
than did Kalahari hyenas. Talek clan den moves, both among natal dens and
from natal to communal dens, were generally over distances of less than 2 km.
Kruuk (1972) observed clear seasonal differences in den placement for hyenas in
the Serengeti, where the two areas in which most dens were found in the wet
and dry seasons were over 20 km apart. Since seasonal and spatial variation in
prey abundance and distribution are much more striking in the Serengeti than in
the Mara, Serengeti hyenas may be able to decrease their commuting distance
to hunting grounds (Hofer 8 East, 19930; a; b) by moving their dens closer to
prey aggregations (Kruuk, 1972). However, in the Talek area where resident prey
are relatively abundant and tend to concentrate year round in the center of the
Talek clan territory (see Chapter 6), seasonal movements of dens would be
unlikely to increase hyena foraging efficiency.

Hyena den moves may occur in response to disturbances at the den (e.g.,

by lions or intruding hyenas), but in most cases the proximate causes of den

59

moves are unknown. In coyotes, reasons for most den moves were also unclear
(Bekoff 8 Wells, 1982), but human disturbance was implicated in 20% of
observed moves of coyote dens (Bekoff 8 Wells, 1982) and this also caused
wolves to abandon dens early (Ballard et al., 1987). Hyena den moves clearly
occur in response to disturbance by humans (Mills, 1990), when a cub dies at the
den (Holekamp 8 Smale unpubl. data), or when the den floods (Holekamp 8
Smale unpubl. data). A build up of fleas may also cause spotted hyenas to move
their den (Mills, 1990). Pilot data reported here on ticks indicated that an
ectoparasite load may sometimes be associated with a den move, but that
ectoparasites appear not to be an important causal agent involved in the initiation
of most den moves in the Talek clan. Hyenas may follow a simple “rule of thumb”
(Dyer, 1994) dictating that they remain at a den as long as there is no serious
disturbance. However, what the tolerance of hyena mothers is for disturbances is
unclear. That there is no clear pattern to den moves may result from an adaptive
strategy that minimizes den disturbances by potentially infanticidal intruding
hyenas or other predators, or it may be the pattern that simply results from the
great behavioral plasticity shown by hyenas in their responses to disturbances.
Although factors mediating the initiation of den moves are poorly
understood, it appears that higher-ranking females are most likely to lead den
moves (present study, Holekamp et al., 2000), and lower-ranking females may
then need to move their cubs in order to keep them in contact with their peer
cohort. Since den moves often occur over a period of a few days, late-moving

females may be moving to keep up with the clan rather than in response to the

60

same factors that prompted the initiating females to move their cubs. Because
den moves occurred less often during months with low birth rates and more often
in months with higher birth rates (Holekamp et al., 1999b), the number of
mothers using a den or the cohort size of den-dwelling cubs might be another
factor influencing the probability that hyenas will move their den.

Although only a handful of dens emerged as “popular” CDs for Talek
hyenas during the 10 year study, many dens were used for N08, transit stops, or
short-term CDs. Talek hyenas often made short den hops to dens where they
keep their cubs for only a few days. This may indicate that the choice of new den
placement is not a coordinated decision made prior to moving by all the mothers
with cubs at the current CD. Perhaps hyena mothers assess suitability of dens
after rather than before they move their cubs.

Most cubs were born at private natal dens that were within 2 km of the
CD. Only 20% of litters studied here were reared from birth at the CD, and this
percent probably overestimates the number of litters born at the CD, since not all
NDs were found. In a small sample, East et al. (1989) found that 30% of 20 litters
in the Serengeti were born at the CD instead of private N05, and so the pattern
of ND use by hyenas appears similar in both the Mara and Serengeti. Low-
ranking females reared their cubs farther from the CD, moved them more often,
and moved them over greater distances before reaching the CD than did high-
ranking females. Although I could not associate cub mortality closely enough in
time with den moves to examine rank effects of ND moves on cub survival, it is

likely that having to move cubs over greater distances represents one factor

61

contributing to the lower reproductive success of low-ranking females (Frank et
al., 1995; Holekamp et al., 1996).

Den moves occur frequently enough that most den-dwelling cubs probably
experience many den moves, and virtually no cubs would fail to experience at
least one den move before leaving the den altogether. Living at dens appears to
be necessary for a cub’s appropriate social development and integration Into the
clan, and den moves thus probably represent one of the costs of sociality in this
gregarious carnivore. Living at different dens might also help familiarize cubs with
their clan’s territory, and this potential benefit might compensate for some of the
costs associated with movement of dens.

A private ND may be more critical to the success of rearing cubs for low-
ranking females than for high-ranking females. Low-ranking females may benefit
from reduced competition with conspeciflcs or lower infanticidal aggression, by
having NDs that are more isolated at a time when cubs are particularly
vulnerable. Harassment from conspecifics may be less for low-ranking females
when dens are situated farther from the CD. When females have young cubs,
they remain almost constantly at the den and may not hunt for the first week after
parturition (Henschel 8 Skinner, 1990). Mothers also appear more reluctant to
leave the den in the face of danger than they are at later stages of cub
development.

The occasional sharing of natal dens should be more likely among close
relatives than distant or non-relatives, due to the potential benefits of enhanced

inclusive fitness (Hamilton, 19643; Trivers, 1974). Talek females, however, were

62

closely related in only one of three cases of ND-sharing observed here. In order
for there to be a chance that females will share a natal den, females must exhibit
a high degree of reproductive synchrony, since cubs are transferred to the CD
within weeks of birth (Kruuk, 1972; East at al., 1989). Due to their low
reproductive rates (Frank et al., 1995; Holekamp et al., 1996), low-ranking
females in particular may rarely have the option of sharing 3 ND with a relative.
ND-sharing may be rare among high-ranking females, because high-ranking
females are more likely to rear their cubs from birth at the CD. High-ranking
females can successfully displace lower-ranking mothers from the CD entrance
and therefore have unrestricted access to their cubs which appears critical during
the first weeks. Cubs of low-ranking females may even be temporarily trapped
underground when a high-ranking female nurses her cubs in the den entrance
and thus blocks the exit of other cubs. Furthermore, since high-ranking females
produce cubs at higher rates than do low-ranking females (Holekamp et al.,
1 996), high-ranking females are also more likely to be rearing cubs at the CD at
the same time as relatives who represent important allies. The use of natal and

communal dens by spotted hyenas reflects the costs and benefits of den-sharing

that vary with female social rank.

63

Chapter 4
ONTOGENY OF SPACE UTILIZATION

IN TRODUC TI ON

A dispersing mammal leaves a familiar physical and social environment
(Waser 8 Jones, 1983; Smale et al., 1997), and moves to a new environment
that is unfamiliar in both these respects. Because dispersers may travel long
distances across new landscapes (Holekamp, 1984; Fuller et al., 1992; Thomson
et al., 1992), the basic space utilization patterns of dispersers and non-dispersers
differ markedly. In most mammalian species, dispersal is sex-biased with males
more likely to disperse (Lidicker, 1975), or more likely to disperse greater
distances than their female counterparts (Greenwood, 1980; Holekamp, 1986;
Packer 8 Pusey, 1987). Dispersal thus represents an important sex difference in
the life histories of male and female conspeciflcs (Greenwood, 1980; Smale et
al., 1997). However, little is known about sex differences in space utilization
patterns prior to dispersal. Do sex differences in space utilization patterns arise
during ontogeny prior to dispersal? My objective here was to address this
Cluozestion in the spotted hyena.

Spotted hyenas live in large social groups, called “clans” (Kruuk, 1972)
that defend group territories (Kruuk, 1972; Henschel 8 Skinner, 1991; Boydston
9‘ 3L. in press). Like many gregarious mammals (Greenwood, 1980; Mills, 1990;

Smale et al., 1997), male spotted hyenas, but not females, disperse from their

natal groups (Henschel 8 Skinner, 1987; Holekamp 8 Smale, 1998b; Holekamp
8 Smale, 19983). As they grow up in their natal clans, young Crocuta first
experience a spatial world that is very tightly restricted, and then later learn to
navigate the much wider spatial environment used by adults. Crocuta cubs spend
the first several months of life confined to the vicinity of a den (see Chapter 3:
Dens). Cubs rarely appear above ground during the first few weeks of life (East
et al., 1989; Holekamp 8 Smale, 19983), but later at the communal den, they are
frequently above ground. During the period of den residency, a cub’s knowledge
of its physical environment is primarily limited to the immediate vicinity of the
communal den. Cubs become independent of the communal den around 8 mo of
age, and in this second phase of life, young hyenas travel alone or with small
groups of conspecifics around the territory defended by the adult clan members.
Focusing on patterns of space utilization in the spotted hyena from birth through

reproductive maturity, I attempted to determine how patterns of space utilization

vary with sex and age throughout early ontogeny.

METHODS

Study site, study group, and field observations.— The study clan, which
usually contains 70 to 80 hyenas, defends a territory encompassing 61.5 km2
(BOYdston et al., in press) in the Talek region of the Masai Mara National
Reserve, in the northeastern part of the Serengeti-Mara ecosystem. This is an

area of open rolling grassland interspersed with seasonal creek beds.

65

Each Talek hyena was individually identified by its unique spots and other
natural marks, and sexed based on the dimorphic glans morphology of the erect
phallus (Frank et al., 1990). Social ranks of all clan members were known based
on their positions in a matrix of outcomes in dyadic agonistic interactions (Smale
et al., 1993). Maternal kin relationships were known for all natal Talek clan
animals, as described previously (e.g., Holekamp et al., 1993).

Talek females usually give birth to 1 or 2 cubs, and cubs reside at dens
until they are about 8 to 9 mo of age. Here cub birth dates were estimated to 1: 7
days based on cub pelage, size, and other aspects of appearance when they
were first seen above ground. Weaning in the Talek population occurs, on
average, when cubs are 13.4 mo old (Holekamp 8 Smale, 1993; Holekamp et al.,
1996), but weaning age may range from 7 - 21 mo (Holekamp et al., 1996;
Szykman et al., In review). Weaning conflicts and cessation of weaning indicated
when cubs were weaned. lf mother and cub were not found frequently together,
weaning dates were estimated to :10 days as the midway point between the last
nursing bout and the next sighting of mother and cub together without nursing.
Males were considered to be reproductively mature at 24 mo of age and females
were considered to be reproductively mature at 36 mo of age (Holekamp 8
Smale, 19983). When a male was no longer found in the Talek clan territory, he
was considered to have dispersed.

Regular hours of observation were 0600 — 0900 and 1700 — 2000. An
observation session was initiated when we found one or more hyenas separated

from other conspecifics by at least 200 m. An observation session ended when

66

observers left that individual or group. Duration of observation sessions ranged
from 5 min to several hours. The geographic location of an observation session
was recorded in reference to local landmarks (e.g., 150 m south of Porcupine
Tree), and sometimes also as geographic coordinates using a hand-held
Magellan Global Positioning System. Date, time, context (den, fresh kill, old
ungulate carcass, or other), and the identities and activities of all hyenas present
were also recorded for every observation session. During each observation
session, all aggressive, appeasement, and scent-marking behaviors (Kruuk,
1972) were recorded as critical incidents (all-occurrence sampling, Altmann
1974). Scan sampling (Altmann, 1974) was also conducted throughout all
observation sessions at 30 min intervals, and focal animal surveys (FAS)
(Altmann, 1974) were conducted on selected individuals during some
observation sessions. During Scan and FAS samples at the communal den,
distances at which individuals were observed from the entrance of the den were
estimated and recorded in meters.

Den dependence: birth through 8 mo.— During the period of den
dependence, cubs stayed underground when they were not on the surface
interacting with clan members. A cub was considered independent of the
communal den when it was observed in four consecutive sessions more than 200
m from the communal den. The date of each cub’s independence from the den
was then identified as the first of those 4 observation sessions. Den

independence usually occurred soon after 8 mo of age, and thus all analyses

67

involving the period of den dependence were conducted with data collected
through 8 mo of age.

Distances to the communal den during the period of den dependence
were examined for individuals from ten litters born between March 1988 and May
1990. Each litter contained one male and one female cub. Maternal ranks for the
pairs ranged from one (offspring of the alpha female) to 16. Maternal ranks of
cubs were assigned as their mothers’ position in the female dominance
hierarchy, and all subjects were also grouped into one of two rank categories.
High-ranking cubs had maternal ranks less than ten, and low-ranking cubs had
ranks greater than or equal to ten.

Data describing distances at which cubs were observed from the den
entrance were collected for all 20 subject animals until juveniles demonstrated
independence from the communal den. Two types of distance data were
obtained from scan sessions and FAS. ln scans at sessions centered around the
communal den, the distance of each hyena to the main den entrance was
recorded, and an average distance to the den for each the individual was
calculated for each session. During FAS, all movements of the focal subject were
recorded, and from this detailed information, a single maximum distance of the
focal hyena to the den during that FAS was obtained.

Den independence: 8 to 36 mo.— Location data from nine of the ten litters
in the preceding section, plus 2 more litters born in 1994-95, were analyzed from
8 mo of age until 36 mo of age, death, or dispersal. Locations of these animals

during observation sessions were plotted on maps of the Talek Clan study area

68

and then digitized into a computer GIS database. For each point in space at
which each subject was found, I calculated the distance (in km) from the current
active communal den and also from the geographic center of the territory using
ArcView GIS software.

Locations of adults: 1991-92.— Geographic locations for 11 Talek adult
females, 10 adult natal males older than 36 mo, and 14 resident immigrant males
were digitized from field data collected during 1991-92. Of these hyenas, 3 adult
females, 7 natal males, and 6 immigrant males were fitted with radio-collars after
they reached 26 mo of age. The radio-collared animals were tracked from the
ground by car and from the air by hot air balloon. Only one geographic location
per animal per morning or evening was included in the dataset presented here. If
an animal was located more than once during morning or evening observation
hours, only the first observation for that individual was included in this dataset.

Statistical analyses.— All dependent variables were measures of distance,
such as distance to the den or to the center of the territory. Independent
variables were sex, rank or rank group, and age. In examining effects of age,
sex, and rank on distances, I used ANOVA. To compare distances of males and
females at particular ages, I used t-tests with a Bonferonni correction for multiple
tests. Because some individuals were seen more frequently than others, mean
distances from either the territory center or communal den at which individuals
were found each month were calculated for use in analyses to control for unequal
numbers of observations per individual. ANOVAs and all statistical analyses were

performed using SYSTAT software.

69

RESULTS

In an ANOVA examining the effects of age, sex, and rank group (high- vs.
low-ranking hyenas), mean distances at which cubs less than 9 mo old were
observed from the communal den varied significantly only with age (F = 36.862,
P < 0.001, d.f. = 6, 112; Figure 4-1). There were no significant interaction effects
among the independent variables. The maximum distance at which these cubs
were observed from the den also increased significantly with age (F = 9.587, P <
0.001, d.f. = 6, 100). There were no sex differences in measures of either mean
distance (F = 0.014, P = 0.907, d.f. = 1, 112; Figure 4-1) or maximum distance to
the communal den (F = 0.589, P = 0.445, d.f. = 1, 100). Neither mean (F = 0.009,
P = 0.926, d.f. = 1, 112) nor maximum distance to the communal den (F = 0.012,
P = 0.911, d.f. = 1, 100) differed significantly between high- and low-ranking
cubs.

Locations at which cubs from eleven litters were found were recorded after
these cubs became independent of the communal den. I first examined litters for
sex differences in space utilization after cubs became independent of the
communal den by comparing males and females within litters only for the time
that both the male and female cub were alive and present in the Talek clan. For
eight of the eleven litters, both the male and female cubs survived for at least 7
mo after becoming independent of the den. Distances at which these 16 cubs
were found from the communal den revealed no differences between male and

female Iittermates even after they became independent of the communal den

70

N
O
1

0 Females (n = 10)
I Males (n = 10) //

_\
01
I

 

_x
O
l
/
x

01
I

Mean dist to den (m)

 

I I I l J I I

0%
123456789
Age(mo)

Figure 4-1. During the period that cubs resided at the communal den,
mean distance at which cubs were observed to the communal den
increased with age for female cubs (dashed line, open circles) and
their male littermates (solid line, open squares). Distance did not vary
with sex of the cubs. Bars indicate the SE of the mean distance. For
each of 10 female cubs, there were 13 i 4 observations per mo, and
for each of 10 male cubs, there were 14 i 5 observations per mo (n =
2,278 total observations).

71

(Figure 4-2). Littermates continued to show strikingly similar patterns of space
utilization even after weaning (Figure 4-2), when the pair ceased having reliance
on a common food source as a stimulus for remaining in close proximity to each
other. In only one pair (Figure 4-2h) were distances to the den measured for both
individuals until the male dispersed from the Talek clan (at age 35 mo), although
several pairs were followed until both littermates reached reproductive maturity.
Between age 9 and 36 mo, cubs were found an average distance of 1.41 :t 0.12
km (n = 8 litters) from the communal den, but this distance from the den did not
vary with maternal rank of litters (F = 0.643, P = 0.423, d.f. = 1, 2597; Figure 4-3).
The patterns observed in Figure 4-2 suggested that perhaps littermates
were more likely than randomly selected cubs to occur together in the same
places. Therefore, I next examined the eleven litters for sex differences using all
surviving cubs, regardless of whether particular littermates still survived. When all
cubs at all possible ages from these litters were included in the sample, an
ANOVA examining the effects of age and sex on distance to the communal den
revealed a significant interaction between age and sex (F = 7.725, P = 0.006, d.f.
= 1, 487; Figure 4-4). Age had a highly significant effect on distance at which
animals were found from the clan’s communal den (F = 245.3, P < 0.001, d.f. = 1,
487; Figure 4-4) but sex alone did not (F = 2.117, P = 0.146, d.f. = 1, 487; Figure
4-4). In the first several months after cubs became independent of the communal
den, males and females did not differ in the distances at which they were found
from the den. However, after 24 mo, males were found increasingly farther from

the den than females (Figure 4-4).

72

Figure 4-2. Mean distances (km) from the Talek clan’s communal den at which
8 pairs of littermates were found, from after the time that cubs became
independent of the communal den until one littermate in the pair died or
disappeared. Females are shown as dashed lines with open circles. Males are
shown as solid lines with filled squares. Weaning ages are indicated with
arrows. An asterisk (*) denotes the death of an individual. Graphs a-e are
presented from highest to lowest maternal rank (MR). 3) Cubs bb and bud, MR
= 1. b) Cubs sx and chy, MR = 5.

73

 

 

Distance to den (km)
0 —3 N 00 h 01

 

 

 

 

 

 

      

 

  

 

 

8 Wean, 18 38
10”” Age of bb 8 bud (mo)
£3 B 5293’: 1 0 SM) '
5 i - chy(m) n i
'o 2" I‘, I '
.9
(D
.‘L’
C)
O L l
8 18 28 38

Age of ex 8 chy (mo)

74

 

Esp . . .
33

8 2.5- C

‘C

.9

a, 2.0-

0

E

E 1'5- (1IEVean)ed,

0 mo

 

 

 

 

1 .0
20 25 30 35 40

 

 
 
  
 
    

  

 

 

E2.0 I I I I I I

6?, Wean, *

ac: D

'0 1.5r -
.9

8

g 1.0- o rum (1) -
17.5 I choc (m)

D l L I l l

 

0.5 '
8 9101112131415

Age of rum 8 choc (mo)

Figure 4-2 cont.- 0) Cubs mp and dd, maternal rank (MR) = 5. d) Cubs
rum and choc, MR = 8.

75

 

 

 

 

 

  

’E‘ 3 . . .
:16 .
ac) E
"o 2 - O -
.9. . o
Q) o
g 1 ' I ' 5 "
£9 a a I o gan (f)
.9 I Ween, I fro (m)
D 13mo

O I l I

8 13 18 23 28
Age of gan 8 fro (mo)

E4 . . r . . .
5, ,z0
a:J 3 .. F / .J
n
.9. -
an 2 '
2 o amar (f)
g 1- ‘3’! I kon (m) -
0 3:63;, I l J l I l

 

 

 

0
18 20 22 24 26 28 30 32
Age of amar 8 kon (mo)

Figure 4-2 cont.- 3) Cubs gan and fro, maternal rank (MR) = 9; the gap
in observations was due to clan fission event (Holekamp et al., 1993). f)
Cubs amar and kon, MR = 9. The gap indicates a period when amar
was not seen.

76

 

00

N

.A

 

Distance to den (km)

 

Wean,
19 mo

 

13 18 23 28
Age ofje & mu (mo)

O
(D

 

 

 

 

 

 

E 7 a a
E, 6 O rv (f) '-
g 5 I eh (m) "
B 4 Wean, —i
G) 3 15*mo -
o
g 2 fix [6‘ IR ‘1 i) '
g 1 a I ‘
O-
8 18 28 38
Age of rv & sb (mo)

Figure 4-2 cont.- 9) Cubs je and mu, maternal rank (MR) = 10. h) Cubs
rv and SD, MR =12.

77

N N
O 01

l I
I-O—l

_\
01
l
I-o-i

_\
O
I
I-o-I
I-o-I

Distance to den (km) / litter

 

5 1O 15
Maternal rank of litter

 

.0
001

Figure 4-3. Distances from the Talek clan’s communal den at which
cubs were found from age 9 to 36 mo were unrelated to the maternal
rank of each pair. Each dot is the average of all observations for a pair
of cubs (n = 324 i 57 observations per litter) between ages 9 to 36 mo.
Observations for individuals did not necessarily span the entire age
interval.

78

 

 

 

 

% 5', A. Females Q 5' B. Males
GE, 0 . .. g 0 0
£4. . €4I . . .0.
E . Q O 5’ .. ~
:3' ” ' :3' ‘0' . co 3.
a) o '0 .° .8 ~ 0
U n .o .flQO ... 02 : .
g2 on... a"; 0 E;
s1 .61
‘6': 5!.)
a D

0 0

0 0

 

 

Figure 44. Mean distances from the Talek clan’s communal den at which
a) female and b) male hyenas were found between the time they left the
communal den and either death or disappearance. Each dot represents
the mean distance of all observations for one cub during a particular
month. Lines are Loess curves smoothed through the data.

79

In an ANOVA examining the effects of age and sex on distance to the
center of the territory, age was a significant factor (F = 4.664, P = 0.032, d.f. = 1,
333; Figure 4-5), but sex was not significant (F = 2.527, P = 0.113, d.f. = 1, 333;
Figure 4-5). However, the interaction between age and sex was significant (F =
5.000, P = 0.026, d.f. = 1, 333; Figure 4-5).

Figure 4-6 shows distances at which male and female cubs were found
from the communal den (Figure 4-6a) or center of the territory (Figure 4-6b) from
24 mo - 36 mo of age, regardless of whether the opposite-sex littermate was still
surviving. Thus in these graphs (Figure 4-6a, 4-6b), females and males in the
samples are not necessarily siblings. The distances from the communal den
observed for male and female hyenas began to diverge after 2 yrs of age (Figure
4-6a), when males were considered reproductively mature and females were
usually in their last year before starting to breed. The interaction between age
and sex was significant in an ANOVA examining the effects of age and sex on
distance to the den (F = 2.340, P = 0.008, d.f. = 11, 1010; Figure 4—63), and thus
Bonferoni post-hoc tests for multiple comparisons were made to determine at
which specific ages males and females differed in their distances to the
communal den. There were no significant sex differences before animals
reached age 30 mo, but at ages 30, 32, 33 and 36 mo, males were significantly
farther from the communal den than were females (Figure 4-6a). Distance to the
center of the territory also varied significantly in an ANOVA examining the effects
of age and sex (F = 2.538, P = 0.004, d.f. = 11, 1010; Figure 4-6b). Bonferoni

post-hoc tests for multiple comparisons revealed a sex difference in distance

80

A. Females

U1

A

AN

Distance to center (km) / female
00

I
I
I

O
_\
O

20
Age (mo)

01

A

(A)

N

A

 

 

Mean distance to center/ male

I

O

B. Males

O I. ‘ g
' 0
.. II
0 0.. ...:..: ..
’ o
o

10 20 30 40
Age (mo)

Figure 4-5. Mean distances from the geographic center of the Talek clan’s
territory at which a) female and b) male hyenas were found between the
time they left the communal den and either death or disappearance. Each
dot represents the mean distance of all observations for one cub during a
particular month. Lines are Loess curves smoothed through the data.

_ 0 Females .

 

 

 

 

 

E 3'5 I Natal males
as , .
5 _ A .
U 2.5
2 r
g 1 5 /'\ I \ ’0' it
" \
cu \ ’ I
2» P T ‘
D
0.5 . u u . .
24 28 32 36
A Age (mo)
E, 3.5- ,
B B *
“E . .
3 2.5- ,
2 I 1 \ '-
3 4 43\\ ’l \I‘ I \
C 1.5‘ 44 ' 4,3 ' I 3.4 \ II
.9 4,4 ' 4.4 4'5 4'SV 3.2 3.2
.‘L’ 3.4
D
0.5 ' I f 1 l ' ' 1 j
24 28 32 36
Age (mo)

Figure 4-6. Mean distances (i SE) at which young females
and males were found from a) the communal den and b) the
geographic center of the Talek clan’s territory. Individuals
here are included regardless of fate of littermate. Females
are shown as dashed lines with open circles. Males are solid
lines with filled squares. Sample size numbers under error
bars indicate the number of females and males,
respectively, at each month of development.

82

from the center of the territory first at age 32 mo and subsequently also at ages
33, 34, and 36 mo. These graphs suggest that a sex difference in space
utilization first emerged around 30 mo of age. This difference may have resulted
from movements toward the periphery of the natal territory and forays into other
hyena clan territories that males begin taking before permanently dispersing
(Smale et al., 1997). Because the number of individuals in these samples was
small, increasing the sample sizes should yield a more robust result and might
more clearly indicate the age at which sex differences emerge.

ln'Figure 4-7, cub ages were grouped into five blocks of several months
for a comparison across broader developmental stages from birth through 36 mo.
In an ANOVA examining the effects of age, sex, and their interaction, distance
from the communal den varied significantly with both age (F = 91.310, P < 0.001,
d.f. = 4, 481; Figure 4-7) and sex (F = 8.872, P < 0.005, d.f. = 1,481; Figure 4-7).
There was also a significant interaction between age and sex (F = 2.979, P =
0.019, d.f. = 4, 481; Figure 4-7). Not surprisingly, the largest increase in distance
to the den appeared when cubs “graduated” from the den, once cubs were no
longer restricted to use of a limited area around the den. Interestingly, the space
utilization of females as measured by their distance to the den appeared to
change little after approximately 16 mo (but see also Chapter 5). By contrast,
male space utilization patterns did continue to change, and distances from the
den at which males were found continued to increase with age, until death or

dispersal.

83

I Females 6

N
01
I

D Natal males

 

 

 

N
O
I
N
(D

 

_\
01

1.0

Mean distance to den (km)

 

 

 

 

 

 

 

_J
0-8 9-16 17-24 25-32 33-36

Age group (mo)

Figure 4-7. Mean distances (km) to the communal den grouped into age
intervals. Numbers over bars are sample sizes for the numbers of females and
males. Not every individual observed within a time period was necessarily
observed during every month of that time period.

A Crocuta clan contains three classes of reproductively mature adults:
adult females, adult natal males, and adult immigrant males. All three classes of
adults were present in the Talek clan in 1991-92, and their space utilization
patterns are compared in Figure 4-8. The mean number of locations : SE per
individual that were digitized for adults in each class in 1991-92 were: 20 i 4
locations per female for 10 females (Figure 4-83), 40 i 9 locations per natal male
for 10 adult natal males (Figure 4-8b), 35 i 7 locations per immigrant male for 14
adult immigrant males (Figure 4-80). While 92% of observations for both females
(n = 219 locations)1 and immigrant males (n = 532 locations) were within the
Talek clan’s territorial boundaries, only 65% of locations for adult natal males (n =
413 locations) were situated inside the Talek territory. Of these 10 natal males,
seven were radio-collared and tracked intensively. Locations before or after
dispersal for 6 of the 7 radio-collared adult natal males are shown in Figure 4-9.
After dispersal, Talek natal males were found on average 6.81 .t 0.25 km (n = 3
males; Figure 4-9b) from the center of the Talek clan territory whereas before
dispersal these same males were found only 2.72 i 0.24 km (n = 3 males; Figure
4-9a) from the center. Thus as expected, natal males were found significantly
farther from the center of the Talek clan territory after than before they dispersed

(t-test, T= 11.89, P < 0.001, n = 3).

 

' Adult female space utilization patterns based on field data from 1996-98 are discussed in detail
in Chapters 5 and 6. Here data from 1991-92 are presented for direct comparison to data for
immigrant and natal males.

85

M—i

.fl

 

A Adult females

 

 

 

: C . C
“vs" .'
O
’0 . o. .
——— . v
. O
B 0 Adult natal males
00 “Jul-EM . . C Q
o (1% o s. . .
0 mg ‘ I .
yo 0 ":0. ' :v00
0 O '
o c «D
=3»:- .-:.-- :t‘i’
u’h '° , .
- - - v-r: -
J“. ... O
o pre-dispersal '4‘:
° post-dispersal .

 

 

 

 

 

 

Figure 48 Locations of adult hyenas relative to borders of
the Talek clan territory (in gray) and the Mara Reserve
(black line) during 1991-92. a) 219 locations for 10 adult
females. b) 413 locations before dispersal (filled circles, n =
302) and after dispersal (open circles, n = 111); 10 males
are represented. c) 532 locations for 14 immigrants male
members of the Talek clan that were born in clans other
than Talek.

86

A. Before dispersal

 

MU
Dispersed 62 mo
(N= 85 locations)

 

- 30-34 mo
- 35-39 mo
. 40-44 mo
. 45-49 mo
SCAR

Dispersed 56 mo
(N= 93 locations)

.- 30-34mo
- 35-38mo
. 39-42mo
0 43-46mo

 

 

  
  

CHY
Dispersed 45 mo
(N= 78 locations)

- 20-24 mo
- 25-29 mo
. 30-34 mo
' 35-39 mo

 

 

 

Figure 4-9. Locations to which adult natal male hyenas were
tracked during 1991-92. The Talek clan territory is in gray and

the border of the Mara Reserve is the black line along the north
edge of the Talek clan territory. Sizes of dots represent ages of
individuals when observations were made. a) Locations at which

3 males were found during the months before they dispersed.

87

B. After dispersal

 

 
 
 
 

SHER
Dispersed 42 mo
(N= 51 locations)

48-52mo
- 53-57mo
. 58-61 mo
62-66mo

 

   
 

PB
Dispersed 37 mo
(N = 39 locations)

- 47-51 mo
- 52-55mo
56-59mo
60-63mo

 

 

Q
Dispersed 33 mo
(N = 31 locations)

31 - 34 mo
- 35-38 mo
. 39-42 mo

0 43-46m0

 

 

Figure 4-9 cont. - b) Locations at which 3 males were found,
primarily after these males had dispersed. Q appeared to move
to the periphery of the Talek clan territory around the time he
dispersed.

88

 

DISCUSSION

Space utilization patterns, as measured by distances at which cubs were
observed from the communal den, varied significantly with age during the first 8
mo of life. During observation sessions, cubs moved increasingly greater
distances away from the den entrance with each month of age during their
development. Similarly in another carnivore, the coyote (Canis Iatrans), distances
that pups moved from the den increased linearly with age through 7 mo of age
(Harrison et al., 1991). Like coyotes, hyena cubs gradually enlarged their spatial
world around the communal den as they matured.

Although striking early sex differences in some behaviors such as
mounting, are evident in young spotted hyenas living at communal dens
(Holekamp & Smale, 1998a), I found no sex differences in the distances that
males and females moved from the den entrance during the first 8 mo of life.
Much of a cub’s time above ground at the communal den is spent socializing with
clan members (Holekamp & Smale, 1998a), and the locations of conspecifics
around the den probably influence the movements of cubs. Distances traveled
during this period from the den entrance are also undoubtedly affected by the
limited locomotor abilities of cubs and the time needed to escape from potential
danger. Movements that took den-dwelling cubs relatively far away from the den
entrance were generally very brief excursions, often accompanied by their
mothers. During the period of den dependence, cubs are quick to flee into the
den for protection if frightened (Holekamp & Smale, 1998a), and so they may

have remained generally close to the den not only because they are not yet

89

familiar with the larger spatial world, but also for quick access to a safe
underground haven. There may be trade-offs between exploring and being able
to return quickly the den. In general, selection pressures on the space utilization
patterns of male and female cubs while living at the communal den should
theoretically be quite similar.

After 8 mo of age, cubs became independent of the communal den, and
they then began traveling around the group territory. A cub’s average distance to
the den increased by an order of magnitude at den independence. Distances did
not change suddenly at weaning, as might have been expected if the movements
of cubs were strongly influenced by their dependency on their mother. Although I
observed variation among litters in the distance at which they were found from
the den, males and females within litters exhibited similar space-use patterns 1
during this phase of development.

After females became independent of the communal den, their distances
to the den did not continue to increase. Thus soon after female cubs left the
communal den, they seemed to be exhibiting patterns of space utilization
resembling those of adult females, at least with respect to distance to the
communal den. I would expect then that the space utilization patterns ofjuvenile
females are likely to vary with their social rank, as was true for adult females who
did not have dependent cubs (Chapter 5). Juvenile hyenas face foraging
challenges during the long time between weaning and becoming proficient adult
hunters (Holekamp et al., 1997b), and low-ranking juveniles in particular have

difficulty acquiring food. Therefore, it is likely that low-ranking juvenile females

90

must range widely in search of food. Future studies of juvenile development and
space utilization could reveal whether rank is one of the primary determinants of
space utilization patterns in young females.

It appears that my inability to observe emergence of a sex difference
within matched pairs of siblings was due to the fact that littermates associate
extremely closely. This is presumably to assist each other in vigilance against
predators and in cooperative acquisition and defense of food (Smale et al.,
1995). Earlier data suggest that a sex difference in space utilization emerges as
early as 24 mo, which represents the age of reproductive maturity in males. At
this age, natal males and females begin to differ in the number of weeks that they
are absent from the Talek clan territory (Smale et al., 1997; Holekamp 8 Smale,
1998a), with males being absent quite frequently while females are seldom
absent. Emergence of a difference in space utilization between males and
females during ontogeny appears to coincide with reproductive maturity of males,

and by 30 mo, the sex difference in space utilization is quite distinct.

91

Chapter 5
SOCIAL AND REPRODUCTIVE DETERMINANTS OF SPACE UTILIZATION

PATTERNS IN ADULT FEMALES

INTRODUCTION

Like many other group-living mammals, spotted hyenas (Crocuta crocuta)
live in fission-fusion societies (Kruuk, 1972; Holekamp et al., 1997a; Holekamp et
al., 2000). In spotted hyena social groups, called clans (Kruuk, 1972), individuals
travel, rest, and forage alone or in subgroups (Holekamp et al., 2000). Subgroup
composition typically changes several times during the course of a single day,
and individuals are frequently also found alone (Holekamp et al., 1997a;
Holekamp et al., 2000). Hyenas assemble for myriad reasons ranging from
cooperative activities, such as territorial defense (Henschel 8 Skinner, 1991;
Boydston et al., in press), to intense direct competition over ungulate kills where
priority of access is determined by social rank (Tilson 8 Hamilton, 1984; Frank,
1986b). Hyenas also frequently congregate at the clan’s communal den, which
serves as the social center of the clan’s territory. Females with young cubs visit
the den regularly to nurse their dependent offspring, and subadults, adult males,
and adult females without den-dwelling cubs also frequently visit the communal

den. Thus, in the fission-fusion societies of spotted hyenas, patterns of space

92

utilization are likely to vary among individuals with social and reproductive
parameters.

My objectives in this study were to document variation in space utilization
patterns among adult female members of one large hyena clan, and to examine
reproductive condition and social rank as potential sources of this variation. I
expected females with den-dwelling cubs to contract their home ranges and act
much like central place foragers, whereas I expected the movements of females
having no den-dwelling cubs to be relatively unconstrained by current location of
the communal den. Because high-ranking females breed at far higher rates than
those of low-ranking females, they are more likely to have cubs at the den at any
given time than are low-ranking females. Therefore, to determine whether social
rank has effects on space utilization beyond those mediated directly by
differential reproductive rates, I examined effects of social rank on space

utilization by females while controlling for reproductive state.

METHODS

Study animal and site.— Distributed widely across sub-Saharan Africa and
across a spectrum of habitat types, spotted hyenas function as keystone
predators in many different ecosystems (Kruuk, 1972). Density, clan size, and
territory area vary tremendously across the range of this species. Clans
containing only ten individuals occupy territories of 1000 km2 in the Kalahari
(Mills, 1987), while clans containing up to 80 members defend 30 - 40 km2

territories in the Ngorongoro Crater (Kruuk, 1970; 1972). The social organization

93

of spotted hyenas remains remarkably constant, however, across varying
population densities and diverse habitats. Each clan contains multiple adult
females, their offspring, and one to several immigrant males who join the clan as
adults. Female hyenas generally spend their entire lives in their natal clans,
whereas all males disperse shortly after puberty (Frank, 1986b; Henschel 8
Skinner, 1987; Smale et al., 1993; Smale et al., 1997; Holekamp 8 Smale,
1998a). Social relationships within a clan are organized on the basis of a linear
dominance hierarchy, with immigrant males subordinate to all natal females and
their young (Kruuk, 1972; Tilson 8 Hamilton, 1984; Frank, 1986b; Smale et al.,
1993; Smale et al., 1997; Holekamp 8 Smale, 1998a).

An individual’s position in the clan’s hierarchy determines its priority of
access to food (Tilson 8 Hamilton, 1984; Frank, 1986b; Mills, 1990). Access to
food, in turn, profoundly affects the reproductive performance of adult females
(Frank, 1986b; Holekamp et al., 1996; Holekamp et al., 1999b). Although all adult
female clan members breed, they do so at rates that increase dramatically with
social rank (Frank et al., 1995; Holekamp et al., 1996). Breeding in most
populations occurs throughout the year, although some populations show
seasonal birth peaks or troughs (Lindeque 8 Skinner, 1982; Cooper, 1993;
Holekamp et al., 1999b). Females give birth to one or two cubs per litter (Kruuk,
1972; Mills, 1990; Holekamp et al., 1996), and cubs are reared at dens for
approximately the first eight months of life. Generally there is one primary
communal den per clan, though sometimes there are two such dens in use

concurrently.

94

The study clan, which usually contains 70 to 80 hyenas, defends a
territory encompassing 62 km2 (Boydston et al., in press) in the Talek region of
the Masai Mara National Reserve, in the northeastern part of the Serengeti-Mara
ecosystem. This is an area of open rolling grasslands interspersed with seasonal
creek beds. Large concentrations of several resident ungulate species graze this
area year round, and these are joined for three or four months each year by large
migratory herds of wildebeest (Connochaetes taun'nus) and zebra (Equus
burchellr) from the southern part of the Serengeti. Boundaries of the territory
marked and defended by the Talek clan were known from observations of group
territorial behaviors including border patrols and clan wars (Boydston et al., in
press).

Collection of spatial data.— Between July 1996 and April 1998, l
documented space utilization patterns of adult female members of the Talek clan.
Researchers were present in the study site on over 90% of days during the 21
mo study period. Researchers conducted regular daily observations between
0600 - 0900 and 1700 — 2000 hours, but observations were also conducted at
other times of day.

At any given time, approximately half the adult females in the clan were
fitted with radio-collars (Telonics lnc., Mesa, A2) with signals in the 150-151 MHz
range. Overall, 13 collared adult females were included in this study, spanning a
range of social ranks from rank 2 (highest) to rank 26 (lowest). Each time a radio-
collared female was found, her geographic location was recorded in reference to

local landmarks (e.g., 150 m south of Porcupine Tree), or as geographic

95

coordinates in Universal Transverse Mercator meters (UTMs) using a hand-held
Magellan Global Positioning System (GPS). Date, time, context (den, fresh kill,
old ungulate carcass, or other), and other data were also recorded.

Radio-tracking was the primary method used to locate the 13 female
hyenas. Researchers radio-tracked the collared females from one or two
vehicles, equipped with scanning receiver units. As researchers drove around the
study area, the frequencies of all collars were continuously scanned on the
receiver. When a signal was heard with an omni-directional antenna, that
particular frequency was then followed using a uni-directional antenna, and the
location of its wearer pinpointed by either visually sighting the animal or localizing
the signal to an area less than 200 m2 when the hyena was not visible, as for
example when she was concealed by bushes. Since animals could often be
observed directly or their signals localized to small areas, spatial accuracy of
locations was high, and estimation of locations through triangulation, which
generally yields less precise spatial information (White 8 Garrott, 1990) was
often unnecessary. Each week, researchers attempted to acquire at least three
radio-tracking locations per female. The clan’s entire territory was driven at least
once every two days, and special efforts were made to radio-track any collared
females not found during the preceding few days.

Because of the open habitat, Talek hyenas could also often be found
without radio-tracking (Holekamp et al., 1997a; Holekamp et al., 1997b;
Holekamp et al., 1999a). Researchers recorded locations of radio-collared

females that were found without radio-tracking them. These sightings were called

96

“opportunistic” locations and were used to supplement radio-tracked data. One
method of acquiring opportunistic locations was to make visual scans of the
study site with binoculars from high points. It was possible to see hyenas up to
1.5 km away without binoculars, and hyenas could be seen up to 3 or 4 km away
with the aid of binoculars. Other methods of acquiring opportunistic locations
included following vultures flying to kills where hyenas might be present, and
following the sounds of hyena vocalizations to the place from which these sounds
were emanating. Observations of collared hyenas encountered while radio-
tracking other hyenas were also considered opportunistic locations.

Depending on whether or not radio-tracking equipment was used to find
and follow signals, each location for a female was assigned to one of two
categories: radio-tracked or opportunistic. Combining radio-tracked locations with
other sightings of focal individuals is common in studies of carnivore movements
and ranging patterns (coyotes: Bekoff 8 Wells, 1986; feral dogs: Daniels 8
Bekoff, 1989; wild dogs: Fuller 8 Kat, 1990; brown hyenas: Mills, 1990; Andreka
et al., 1999), and here all tracked and opportunistic locations were similarly
combined in all analyses.

To avoid autocorrelation among sequential locations, any location for a
particular female that was not separated by at least one hour from the previous
location for the same female was excluded from the data set, whether the
location was identified with or without tracking. One hour allowed sufficient time
for hyenas to cross the Talek territory and was thus sufficient for statistical

independence of observations (White 8 Garrott, 1990). Furthermore, all tracked

97

and opportunistic locations that were within 200 m of the communal den were
also excluded to avoid biasing locations of females toward the site of the
communal den, which observers usually visited at least twice daily.

Assignment of social rank, age, and reproductive state.— All individual
members of the Talek study clan were recognizable by their unique spots. Sex
was determined by the dimorphic glans morphology of the erect phallus (Frank et
al., 1990). Social ranks of all clan members were known based on their positions
in a matrix of outcomes of dyadic agonistic interactions (Smale et al., 1993).
Maternal kin relationships were known for all natal Talek clan animals, as
described previously (Holekamp et al., 1993).

Females were considered to be adults when they conceived their first
litter, or at 3 years of age, whichever came first. Talek females usually give birth
at private natal dens, and transfer their cubs to the clan’s communal den after 2-5
weeks. Cubs then reside at the communal den until they are approximately 8
months of age. Cub birth dates were estimated here to :t 7 days based on cub
pelage, size, and other aspects of appearance when cubs were first observed
about ground at natal or communal dens. Conception dates were calculated by
subtracting the length of the gestation period, 110 days (Schneider, 1926; Kruuk,
1972), from cub birth dates. Like birthdates, conception dates were thus known
to :l: 7 days. Weaning in the Talek population occurs, on average, when cubs are
13.4 mo old (Holekamp 8 Smale, 1993; Holekamp et al., 1996) but weaning age
ranges from 7 - 21 mo (Holekamp et al., 1996; Szykman et al., In review).

Weaning conflicts and cessation of nursing indicated when cubs were weaned. In

98

determining weaning dates, all field notes were searched for observations of
nursing behavior when mother and cubs were found together. If mother and cub
were not found together frequently after the last observed nursing bout, the
weaning date was identified as being midway between the last nursing bout and
the next sighting of a mother and cub together without nursing. However, only
intervals of 20 days or less were used in this analysis, so all weaning dates used
here were accurate to within 1 10 days.

For each day on which a female’s location could be pinpointed, she was
assigned to one of four possible reproductive states. Pregnancy (P)
encompassed the 110 days from conception to parturition. After parturition,
females were considered to have entered the first phase of lactation (L1). During
this period, litters resided at dens, usually a communal den shared by 2 or more
litters. L1 ended when cubs became independent of the communal den or died.
Prior to den-independence, cubs were regularly seen at the communal den.
When a cub was found at least 200 m away from the den on four or more
consecutive occasions, that cub was considered to be ‘free-roaming’ and
independent of the communal den. Females that continued nursing free-roaming
cubs were classified as being in the second phase of lactation (L2) which ended
at weaning. Females that were neither pregnant nor lactating were assigned a
reproductive state of “other” (O). Nulliparous adult females were also assigned to
this “0” state. Three females became pregnant while still in the second phase of
lactation. At such times, their reproductive states were assigned as L2. For some

analyses, females were further classified as either having den-dwelling cubs (L1)

99

or having no dependent cubs currently residing at the communal den (L2, 0, and
P).

Prey availability.— Food resources available to Talek hyenas were
monitored by counting all prey animals found within 100 m of 18 km of transect
lines once per week. Prey availability was then averaged for bi-weekly intervals.
For the total study period, the half of the bi-weekly census intervals in which the
highest numbers of prey animals were counted were designated as intervals
when prey were abundant (> 900 prey items counted). The other half of the
census intervals were considered intervals when prey were scarce (< 900 prey
items counted).

GIS analyses and statistical methods.— GPS coordinates were recorded
for about 40 landmarks distributed throughout the study site, and these were
used to create a digital photo-mosaic map for use in a Geographic Information
System (GIS) as a base map. Air photos taken in 1991 by Kenya Wildlife Service
and Kenya Rangelands and Ecological Monitoring Unit were scanned into a
computer at 1 m resolution. Using ERDAS/Imagine software, the separate photos
were edge-matched and geo-rectified with the GPS landmark data in UTMs to
create a digital map of the study site on which trees, stream crossings, and other
landmarks were clearly visible. The boundaries of the Talek clan’s territory were
digitized as a separate overlay coverage (using methods described by Boydston
et al., in press), as were the locations of all dens used by the hyenas during the

study period. Finally, locations of all collared females were digitized so that each

100

point in space at which each female was tracked or opportunistically found was
associated with UTM coordinates.

To document space utilization patterns, I calculated linear distances and
home range estimates using ArcView GIS software with the Animal Movement
Extension (Hooge 8 Eichenlaub, 1997). For each location for each female, I
obtained the straight-line distance from that point to the communal den, and also
the distance between that point and the nearest territorial boundary or “edge.”
When two communal dens were in use concurrently, the straight-line “distance to
the den” was measured between the female’s location and the midpoint between
the two dens. The distance to the nearest territorial boundary (“distance to the
edge”) was used as a measure of a female’s tendency to be peripherally located.
If a female was found outside the Talek territory, her distance to the nearest
boundary was assigned a negative value. In comparing distance to the den or
edge for females in different reproductive states, I used mean distances per
female per sampling interval. I calculated means for only those females for which
at least 30 locations had been recorded in a given reproductive state, in order to
eliminate means with high variances due to small sample sizes.

Sizes of areas used by individuals were first estimated for each female
based on all recorded locations, except those observations of females at the
communal den. I calculated Minimum Convex Polygons (MCPs) (Hayne, 1949)
and fixed kernel utilization distributions (UD) (Worton, 1989; Powell, 2000) with
50 and 95% probabilities, in order to present the results of both the traditional

and widely-used MCP method and the newer fixed kernel UD method. Because

101

MCPs rely on only the outermost points at which an individual was observed,
they do not require independence among observations (Swihart 8 Slade, 1985).
However, MCPs do not show intensity of use (Harris et al., 1990) which UDs
reveal. The number of observations needed to estimate home range size with
MCPs has been suggested as 18 for lizards (Sceloporus virgatus: Rose, 1982),
25-35 for Muntjac (Muntiacus reevesi: Harris et al., 1990), 40-1 50 for coyotes
(Canis Iatrans: Bowen, 1982; Messier 8 Barrette, 1982), 29 for mountain lions
(Puma concolor. Pierce et al., 1999) and 35-120 for wolves (Canis lupus: Fritts 8
Mech, 1981 ). Although the number of observations required to calculate fixed
kernel UDs has been suggested for only a few species, fixed kernel UDs do not
require sample sizes as large as those needed for home range estimates based
on MCPs (Seaman et al., 1999; Powell, 2000).

The sizes of areas used by individual females during each reproductive
State (with and without den-dwelling cubs) were estimated from 95% UDs only.
UDs were calculated only for females in particular reproductive states for which
sample sizes of more than 50 locations had been recorded. These UDs were
calculated from data that were collected during continuous periods of at least 3
mo but no more than 4 mo, an amount of time that was usually sufficient time for
collection of at least 50 locations for an individual female. When more than one
UD could be calculated for a female for a particular reproductive state because
she cycled repeatedly during the study, the mean UD size was reported for that

female in that state.

102

SYSTAT 8.0 was used for all statistical software. Distance to the den,
distance to the territory edge, and size of UDs were dependent variables in all
analyses. Rank, reproductive state, and the current number of available prey
were the independent variables. Because I calculated mean values of distance to
den or territory edge for females for use in analyses, sample sizes equaled the
number of females. These mean values were normally distributed. Fixed kernel
UDs were transformed to a log scale for analyses, since UDs were measures of
area. In analyses in which UD size was the dependent variable, sample sizes
were equal to the number of females involved in each analysis. Differences
among females in various reproductive states were assessed with ANOVA, with
sample sizes equal to the numbers of females. Paired t-tests were used for
comparing utilization by the same females under two different conditions. Where
female social rank was the independent variable in linear regression analyses,

Spearman’s rank correlation coefficients were used.

RESULTS

General patterns of space utilization.— The Talek clan’s territory (Figure
5-1a) was located in the northeast part of the Masai Mara National Reserve,
sharing its northern border with the edge of the Reserve. Twenty-one different
communal dens were used by the Talek clan during the study period. Each
communal den was occupied, on average, for 1.5 :l: 0.4 mo (n = 21 dens). During
8 mo in 1996-98, there was only one active communal den site, and during the

remaining time, there were two active communal dens. Most communal dens

103

 

 

 

KENYA
Masai Mara
National Reserve ~
TANZANIA ‘
Talek Clan
Territory
0 1o 20 Km
3 Communal dens
1 996-1 998
0 ., Talek River
0 O . o ..
o o o F
”o

 

as": “x; \

 

 

 

Figure 5-1. a) The location of the territory defended by the
Talek clan (in gray) within the Masai Mara National
Reserve in southwest Kenya. b) Talek clan territory with
water courses and locations of the 17 communal dens
hyenas used during the study period in 1996-98 (three
sites were used more than once each).

104

(Figure 5—1 b) that the Talek clan used during the study period were in the
northwest quadrant of the clan’s territory, except during December 1997 - March
1998 when active communal dens were located concurrently in the eastern and
western parts of the territory.

To illustrate differences between alternative methods for estimating home
ranges for an individual, Figure 5-2 presents representative spatial data obtained
for two of the 13 monitored Talek females throughout the 20-mo study period.
Specifically, Figure 5-2 compares U0 and MCP range calculations for one high-
ranking female (Figure 5-2a) and one low-ranking female (Figure 5-2b). The
MCPs for both females more closely matched the defended borders and size of
the Talek clan’s territory than did UDs. However, MCP calculations did not reveal
the differences between these individuals that were apparent with UD
calculations. Even though all locations falling within 200 m of an active den were
excluded from analyses, areas of core use (50% UDs) reflected intensive use by
both females of areas around communal den sites. Areas of core use were not
always contiguous, and some females had two 50% UDs that were kilometers
apart (Figure 5-2b). In addition to having two areas of core use, the low-ranking
female (Figure 5-2b) had a 50% U0 that was twice as large as the 50% U0 of
the high-ranking female (Figure 5-2a). The size of the 95% UD of the low-ranking
female (Figure 5-2b) was also larger and extended farther east than the 95% UD
of the high-ranking female (Figure 5-2a). By contrast, the MCPs of the two
females were almost equal in size. Thus UDs shed more light on patterns of

space utilization in this species than did MCPs.

105

 

 

 

 

 

 

_ Talek Clan Territory I 50% Utility Distribution

I I I I MCP
\‘ . . . . .
o Hyena location m 95% Utility Distribution

Figure 5-2. All locations and the resulting MCPs and UDs for the full 21-
mo study period for a) one high-ranking female (rank = 9, n = 529
locations) and b) one low-ranking female (rank = 21, n = 476 locations).
The sizes of the MCPs and UDs are given in Table 1.

106

Ten females were radio-collared for at least 17 mo of the study, and three
additional females had radio-collars for 8-9 mo. Excluding sightings at the
communal den and within an hour of a previous sighting of the same female, a
total of 4,838 locations were collected for the 13 females. Most (64 :I: 2%) of the
locations for each female were radio-tracked locations. The mean 1 SEM size of
95% UDs based on all locations for the 10 females was 30.2 i 2.2 km2 (Table 5-
1). On average, 90 :t 2% of each UD fell inside the clan’s territorial boundaries
(Table 5—1). On average, only 44 i- 3% of the entire territory was covered by the
95% U0 of Talek females (Table 5-1), and all individual UDs together covered
only 69% of the 62 km2 of the territory. Thus, approximately 30% of the territory
was not regularly used by any one of these females during the study period. The
95% UDs overlapped on average 76 i 2% (Table 5—1). Percent overlap
decreased significantly with social rank, with UDs of low-ranking females
overlapping least with those of other females (Spearman’s rank correlation rs= -
0.80, P = 0.01, n =10).

Effects of reproductive state, rank, and prey availability.— Distance to the
communal den varied significantly across the four reproductive states L1, L2, O,
and P (ANOVA: F = 3.40, P = 0.03, d.f. = 3, 37; Figure 5-3). Non-lactating
females (0, P) were significantly farther from the den than were lactating females
with den-dwelling cubs (L1) (Figure 5-3). The difference in distance was not
significant between the two phases of lactation (L1 and L2), although the value
for L2 was intermediate between L1 and the two non-lactating states. While cubs

of females in L2 no longer resided at the communal den, these cubs often

107

Table 5-1. UD calculations for each individual over the entire study
period. Numbers in parentheses are results for three females

collared for only 8-9 mo; these were excluded from calculations of
means. All other females (n = 10) were collared for at least 17 mo.

 

% of Mean %
tenitory overlap of
covered by 95% U0 with
95% UD other females

°/o 0f 95%
U0 inside
tenitory

Social 100% MCP 50% U0 95% U0
rank (kmz) (kmz) (kmz)

 

 

(2) (33.7) (4.4) (36.9) (90) (54) (66)
6 45.3 1.5 17.1 96 26 37
7 63.6 2.9 26.9 93 39 32
9 62.7 2.3 25.6 96 40 33
11 46.7 3.1 27.3 99 44 32
13 60.5 2.5 36.1 86 51 73
14 39.0 7.1 34.7 39 50 73
15 55.6 6.4 37.9 36 53 66
20 63.5 4.6 37.7 90 55 70
(21) (23.9) (5.2) (31.6) (96) (49) (70)
24 59.2 4.5 34.2 37 43 74
26 76.0 1.9 25.1 32 33 70
(26) (69.6) (16.9) (62.5) (57) (58L (35)
Mean 62.2 4 30.2 90 44 76
SE 4.1 0.6 2.2 2 3 2

 

108

.ks
1
F5

 

00
I

n=11

+

N
I

Mean distance to den (km)

 

 

 

 

 

 

 

l I I I

L1 L2 O P
Reproductive state

 

 

 

Figure 5—3. Mean distances 1: SE to the communal den
during each of four reproductive states (L1 lactation
with den-dwelling cubs, L2 lactation with free-roaming
cubs, P pregnant, O other). Numbers over bars
indicate numbers of females monitored in each state.
(*) indicates significant difference from L1 (Tukey’s
post-hoc test). Total numbers of locations represented
in each reproductive state were as follows: 1,977
observations for females in state L1; 1,039 for state
L2; 1,017 for state 0; and 1,145 for state P.

109

remained in the vicinity of the den, gradually increasing their travel distances
away from the communal den (as described in Chapter 3). Thus, mothers in L2
returned relatively frequently to the general vicinity of the den, or tended to
remain in that area.

In an ANOVA examining the effects of reproductive state (with or without
den-dwelling cubs) and social rank (high or low-ranking) on the log of the size of
95% UDs, the log-transformed values of the UDs varied significantly with
reproductive state (F = 21.294, P < 0.001, d.f. = 1, 18; Figure 5-4), with the actual
size of 95% UDs increasing from a mean of 18.58 :I: 2.58 km2 when females had
den-dwelling cubs to 36.35 :I: 2.36 km2 when females has no cubs at the den.
The log-transformed UD sizes also varied significantly between and low-ranking
females (F = 5.077, P = 0.037, d.f. = 1, 18; Figure 5-4), with a mean 95% U0 size
of 23.14 :t 2.47 km2 for the high-ranking females while the mean size of 95% UDs
for low-ranking females was 31.79 :I: 2.47 kmz. Examining UD size within the
group of low-ranking, log UDs were significantly larger when females did not
have cubs at the communal den than when females had den-dependent cubs
(paired t—test, t= 5.69, P = 0.005, d.f. = 4). Among high-ranking females, UDs
were larger when females did not have den-dwelling cubs than when females
had den-dependent cubs, but this difference was not significant (paired t-test, T =
2.592, P = 0.061, d.f. = 4).

Examining the relationship between UD size and ranks of individual
females, I found that the log of UD size was unrelated to social rank when

females had den-dwelling cubs (rs = 0.15, F = 1.35, P = 0.28, n = 10). However,

110

_\
CD

" Rank
I High (ranks 2-13) 5
III Low (ranks 14-26)

_x
CD
I

 

Log 95% up (ka)
3%

_\
N
'

 

 

 

 

 

 

 

1.0
Cubs No cubs

Figure 5-4. Females were divided into two groups according to rank:
females in the top half of the dominance hierarchy were assigned to
the “high” rank group, and females in the lower half were assigned
to the “low” rank group. The 95% UDs were significantly smaller
when females had den-dwelling cubs (state L1) than when they did
not (states L, P, O) and were smaller for high-ranking females than
low-ranking females. Sample sizes over bars indicate numbers of
females in each treatment.

111

when females had no den-dwelling cubs, the log of UD size increased
significantly with rank (rs = 0.73, P = 0.03, n = 10; Figure 5-5).

Space utilization patterns of low-ranking females were more strongly
influenced by reproductive state and prey abundance than were those of high-
ranking females (Figure 5-6). Even though all locations falling within 200 m of an
active den were excluded from all analyses, females with den-dwelling cubs were
generally found in the vicinity of the active communal den, regardless of rank or
prey availability (Figure 5-6a). Distance to the communal den during L1 was
unrelated to rank during periods of both prey abundance (rS = 0.67, P = 0.35, n =
10; Figure 5-6a) and prey scarcity (rs = 0.15, P = 0.89, n = 10; Figure 5-6a).
Similarly, distance to the nearest territorial edge did not vary with rank or prey
availability for females in L1 (prey abundance rs = -0.35, P = 0.61, n = 10; prey
scarcity rs = 0.37, P = 0.51, n = 10; Figure 5-6c). Both distance measures were,
however, correlated With social rank when females had no den-dwelling cubs
(states L2, O, or P), and this was true during periods of both prey abundance and
prey scarcity (Figure 5-6b and 5-9b). Low-ranking females without den-dwelling
cubs were found significantly farther from the communal den than were high-
ranking females when prey were abundant (rs = 0.64, P < 0.01, n = 12) and also
during periods of prey scarcity (rs = 0.68, P < 0.01, n = 12; Figure 5-6b). Low-
ranking females having no den-dwelling cubs were also found closer to borders
than were high-ranking females when prey were abundant (rs = -0.87, P < 0.01, n
= 12; Figure 5-6c) as well as during periods of prey scarcity (rs = -0.83, P < 0.01,

n = 12; Figure 5-6d). When prey were scarce, low-ranking females were found

112

1.8"
1.7
1.6
1.5
1.4
1.3
1.2"
1.1'

1.0 I l J
0 10 20 30

Social rank

 

Log 95% up (ka)

 

 

Figure 5-5. Sizes of 95% UDs were correlated with
social rank among females with no den-dwelling
cubs.

113

 

0 Prey abundant (—-)
A ‘ Prey scaroe(—) 6' B

CD
I

 

 

 

U1
I

Cubs at den

.h
I
0

(JD
I
I

Distance to den (km)

N
f
\
*1
i
l

 

 

 

 

 

 

Distance to edge (km)

 

 

 

 

O I L I O J l I
0 10 20 30 0 10 20 30

Social rank

Figure 56 Mean distance at which each female was found from the communal
den and from the nearest edge of the clan territory. a) Distance to the
communal den when females had den-dwelling cubs (state L1 ). b) Distance to
the communal den when females had no den-dwelling cubs (states L2, P, and
O). c) Distance to the nearest territorial edge when females had den-dwelling
cubs (state L1 ). d) Distance to the nearest territorial edge when females had no
den-dwelling cubs (state L2, P, and O).

114

even farther from the den than they were when prey were abundant (paired t-
test, T = 2.68, P = 0.02, n = 12; Figure 5-6b), but the relationship between
distance to the edge of the territory and social rank did not change with the

abundance of prey (paired t-test, T = 0.36, P = 0.72, n = 12; Figure 5-6d).

DISCUSSION

Female members of the Talek clan used surprisingly little of their territory
on a regular basis. Territories of mammalian carnivores are usually described as
being the same size as, or smaller than, home ranges (Ewer, 1973; Powell,
2000). However, if the 95% UD is used to define a home range, then the territory
defended by the entire group was actually larger than individual home ranges
measured for Crocuta in the present study and also in an earlier study of this
species by Henschel and Skinner (1991). Even the overlapping UDs for ten Talek
females covered only 69% of the territory. Maintaining a territory larger than the
area regularly used by an individual may be unique to fission-fusion societies in
which the economics of territorial defense presumably differfrom those proposed
for groups that associate tightly.

Traditional MCPs more accurately reflected the size and shape of the
Talek clan’s group territory than did UDs, but only UDs revealed variation in
overlap among female home ranges and individual variation in areas that were
most intensively used. The lowest mean percent overlap for the UD of any Talek
female was 66% in the present study (Table 5-1), similar to the minimum 65%

overlap that Henschel and Skinner (1987) found between pairs of clan members

115

in Kruger National Park. In the current study, overlap with other females was
greater for high-ranking females than for low-ranking females, as would be
predicted from patterns of association in hyenas in which animals of all ranks
prefer to associate with higher-ranking females (Holekamp et al., 1997a), except
when hunting and feeding (Holekamp et al., 1997b).

The space utilization patterns of Talek females were influenced by female
reproductive state, prey abundance within the clan’s territory, and social rank.
Female hyenas with den-dwelling cubs had smaller home ranges than did other
females, and were generally found closer to the clan’s communal den. When
females had no den-dwelling cubs, home range size and distance to the
communal den increased among lower-ranking females. Low-ranking females
without den-dependent cubs ranged most widely, particularly during periods of
prey scarcity. When their movements were not constrained by the needs of den-
dwelling cubs, low-ranking females may have been using areas farther from the
communal den to avoid competition over food with high-ranking females. Space
utilization patterns of high-ranking females were less strongly affected by
reproductive state and prey abundance than were those of low-ranking females.
High-ranking females were consistently found closer to the communal den, and
although the den was sometimes located near the edge of the territory, high-
ranking females were more likely to be found deeper inside the territory than
were low-ranking females.

High-ranking females with smaller home ranges may be able to conserve

energy by traveling less (Mills, 1990). High-ranking females may also benefit

116

from being more centrally located than low-ranking females, and from rarely
venturing outside of the territory. Talek females who enter the territories of
neighboring clans risk attack from resident animals if they are detected and
recognized as intruders (Boydston et al., in press). Similarly, Talek females who
are absent from the center of the Talek clan’s territory for several months may be
targets of severe aggression from resident hyenas, and drop in social rank, upon
return (Holekamp et al., 1993). The large UDs of low-ranking females may reflect
a strategy that attempts to balance the costs of competition with higher-ranking
females in the vicinity of the communal den against the costs of prolonged
absence, or a trade-off between better resources in the center of the home range
and increased competition for these resources.

Most terrestrial carnivores bear altricial young that reside at dens or
creshes for some time before offspring begin traveling around the home range
independently or with conspeciflcs, and the contraction of home ranges during
denning observed here in female Crocuta also occurs in other carnivore species,
including caracals (Avenant & Nel, 1998), wild dogs (Schaller, 1972; Burrows,
1995), Cheetahs (Durant, 1998), arctic foxes (Landa et al., 1998), and black
bears (Hirsch et al., 1999). However, only in spotted hyenas does it appear that
the same group territory is maintained even when individuals contract their
ranges, although this may also be true for other carnivores living in fission-fusion
societies and defending group territories, such as lions (Packer et al., 1990).
Also, because adult female hyenas participate in group territorial activities

regardless of reproductive state (Boydston et al., in press), the contracted 95%

117

UD does not mean that females with den-dwelling cubs are not venturing
occasionally to the border of the territory.

The point distributions and home range sizes calculated here based on
data from females are probably conservative estimates of space utilization
patterns for hyenas in the Serengeti-Mara ecosystem and elsewhere in Africa.
While dispersing, male hyenas move over larger areas than females and may
traverse up to 4 or 5 clan territories (Smale et al., 1997). Long-distance
commuting is not a regular feature of hyena behavior in the Mara as it is in the
Serengeti, where hyenas frequently make long-distance movements away from
their group territories to forage in areas of high prey abundance (Hofer & East,
1993c; a; b). Where hyenas occupy very large territories, or where hyenas exhibit
commuting behavior, individual home ranges should be much larger than those
of Talek females.

Woodroffe and Ginsberg (1998, 2000) found that vulnerability to edge
effects is a significant contributor to local extinction of mammalian predators
inhabiting protected areas, and that female home range size can be used to
predict the critical reserve size needed to sustain carnivore populations. Because
large carnivores tend to be wide-ranging, individuals in protected areas may
frequently encounter reserve boundaries where they face higher risks of mortality
due to human activity (Woodroffe & Ginsberg, 1998; Woodroffe & Ginsberg,
1999). For gregarious carnivores, the viability of the social group may be

contingent upon variation in ranging behavior among group members. The

118

widest-ranging individuals in a social group are likely to be at the greatest risk for
edge effects.

Given the position of the Talek clan’s territory at the edge of the Mara
Reserve, and the fact that collared females were tracked outside of the territory
on some occasions, all Talek hyenas may be at risk for mortality due to edge
effects. In this case, those at greatest risk are likely to be low-ranking females
without den-dwelling cubs, particularly during periods of prey scarcity. In the
Serengeti, lactating females tend to most frequently travel the longest distances
and appear as a result to be more likely than other females to be victims of
snares (Hofer et al., 1993). Because the widest-ranging individuals are unlikely to
be a random subset of a social group, edge effects might increase inbreeding
potential and decrease effective population size (Ne) (Parker & Waite, 1997;
Creel, 1998). A better understanding of individual variation in space utilization
patterns, and the mechanisms by which edge effects can lead to extinction,
should aid in planning for the protection of wide-ranging carnivores around the

world.

119

Chapter 6
ECOLOGICAL DETERMINANTS OF SPACE UTILIZATION AND THE

CONSEQUENCES OF EDGE EFFECTS ON SPOTTED HYENA BEHAVIOR

INTRODUCTION

Spotted hyenas and the Serengeti-Mara ecosystem— Spotted hyenas
(Crocuta crocuta) are widely distributed across sub-Saharan Africa and represent
keystone predators in many ecosystems (Kruuk, 1972; Hofer & East, 1995).
Hyena densities, sizes of clans, and areas covered by clan territories vary greatly
across Africa as a result of the distribution and abundance of prey (Kruuk, 1972;
Mills, 1987; Hofer & East, 1995; Mills & Gorman, 1997). In the Kalahari desert,
hyena clans with only ten individuals occupy territories greater than 1000 km2
(Mills, 1987), whereas in the Ngorongoro Crater in Tanzania, clans with up to 80
members hold group territories encompassing less than 40 km2 (Kruuk, 1970;
1972).

Spotted hyena densities are relatively high in the Masai Mara National
Reserve (MMNR) in Kenya, the northernmost portion of the vast Serengeti-Mara
ecosystem. The MMNR receives more rainfall than the Serengeti and, unlike
much of the Serengeti, supports both year-round resident herbivore populations

and high densities of carnivores. Hans Kruuk’s (1972) landmark study comparing

120

the behavior and ecology of Crocuta inhabiting Ngorongoro with those of Crocuta
residing in the Serengeti revealed that variation in the distribution and abundance
of prey promotes differences between hyena populations. Since 1972, many
other studies of Crocuta have been conducted in a wide variety of African
habitats (Whateley & Brooks, 1978; Skinner & van Aarde, 1980; Cooper, 1989;
Sillero-Zubiri 8. Gottelli, 1992), but within these populations and habitats, we
know little about the ecological variables influencing variation among individual
hyenas with respect to their patterns of space utilization.

Understanding the determinants of space utilization is important to
successful conservation planning for large carnivores. Carnivores have recently
been shown to be particularly susceptible to “edge effects” (Woodroffe &
Ginsberg, 1998; 2000) that can drive local populations to extinction in fragmented
areas. The “edge” is the line between areas of contrasting land use, such as
where protected wildland meets cattle ranch. Predators inhabiting protected
areas are likely to have home ranges that encompass long portions of such an
edge, and individuals may frequently range into areas of human activity where
they are vulnerable to direct persecution and indirect mortality risks from humans
(Woodroffe & Ginsberg, 1998; 2000). An understanding of the determinants of
carnivore space utilization will enable us to better predict how carnivores will
respond to changes in their environments and help us prevent or ameliorate
effects on wildlife of expanding human populations and changing land-use

patterns.

121

The goal of this study was to elucidate the ecological determinants of
ranging behavior within one clan of hyenas occupying a territory at the edge of
the MMNR (Figure 6-1). Here I sought to answer two main questions about
individual use of space within the territory of one large clan. First, what are the
important ecological determinants of hyena space utilization patterns at this
localized scale? To examine this first question, I examined the effects on hyena
dispersion patterns of three features of the hyena environment that vary in space
and time: lions, prey, and human activity, in particular the grazing of domestic
livestock. The second main question was asked from an historical view point and
inquired whether space utilization patterns of hyenas in this clan have remained
stable since 1988, shortly after the current boundaries of the MMNR were
established (Talbot & Olindo, 1990; Norton-Griffiths, 1995). Here, we were
interested in the possibility that ‘edge effects’ were operating in the Talek Clan.

.Lions.— Lions (Panthera Ieo) and hyenas are close competitors, that
largely utilize the same prey base (Kruuk, 1972; Cooper, 1991). In general, the
abundance of lions and hyenas tends to be tightly correlated in parks throughout
Africa (Creel & Creel, 1996), and both species are generally most abundant in
areas of highest prey density (Mills & Gorman, 1997). Since the occurrence of
lions and hyenas is positively correlated in space on a regional scale, this pattern
might also be apparent on a finer scale, and both large carnivore species might
frequent the same patches within a given area. However, lions are a significant

source of mortality to hyenas (Kruuk, 1972), and one potential mechanism for the

122

.mEmmom .mcozmz Ems. 53.2 $5 2 conga. E 399 cc EQEQ :90 x98. 9:. .Tm 9:9”.

 

 

{we .95sz somcmcom

 

:5. com: <_Z<Nz<._.
mammom .mcozmz
Ems. Emms.
562W— x0_m._.
H
Ex 3 h o <>zm¥

 

 

123

 

successful coexistence of these two carnivores might be for hyenas to avoid lions
within their territories.

Pastoralists and livestock.— Local people often perceive predators as
threats to their livestock and personal safety, so peaceful coexistence between
human and carnivores is rarely possible outside of reserves (Mills, 1991;
Woodroffe & Ginsberg, 1998). Where spotted hyenas range into areas of human
habitation, they are often persecuted, as are other large carnivores. Hyenas that
range in or near the MMNR in areas utilized by pastoralists are sometimes
speared, poisoned, or injured or killed in snares. Because cattle herds in the
Mara are accompanied by one or more herders armed with spears and clubs,
and because hyenas are wary of people on foot, it seemed reasonable to expect
here that hyenas might avoid areas being heavily grazed by cattle.

Ungulate prey.— I expected a strong positive correlation between the
occurrence in space of individual hyenas and their ungulate prey. In general,
hyena population density is strongly correlated with prey density (Kruuk, 1972;
Mills & Gorman, 1997). Hyenas should theoretically seek to minimize energy
spent during foraging, and I therefore expected that they would seek to minimize
travel distances (Mills, 1990), and also that they would frequent parts of their
territory containing the most abundant prey to maximize the likelihood of finding
prey to hunt.

Historical comparison: 1988-90 vs. 1996-98.— Spotted hyenas are
extremely flexible in their behavior and ecology, and their responses to long-term

environmental changes should therefore represent conservative indicators of

124

how top predators in general may respond to such changes (Arcese & Sinclair,
1997). The Talek clan of hyenas has been continuously and intensively studied
since 1988, and long-term records on ungulate prey, rainfall, lions, livestock,
demography, and behavioral observations of Talek hyenas allow ecosystem
changes to be associated with changes in behavior. Even though land use in the
Mara Reserve was designated for wildlife tourism only (Norton-Griffiths, 1995),
the Reserve, which is unfenced, is accessible to pastoralists for grazing livestock.
Thus, the proximity of the Talek clan’s territory to the northern boundary of the
Mara Reserve means this clan may be at risk for edge effects operating on both
sides of the Reserve boundary, due to Talek hyenas ranging outside their
territory (Chapter 5) into areas of human settlement and due to humans utilizing
areas inside the Reserve.

Therefore, the spatial data on female hyenas during 1996-98 were
compared to spatial data from 1988-90 which was the earliest time period from
which comparable data could be analyzed. Prior to 1984, the boundary of the
Reserve was north of the Talek River that the Talek hyenas defend as their
territorial border (Chapter 2), and the Talek clan’s territory would thus have been
situated fairly deep in the Reserve (Frank, 1986a; Talbot & Olindo, 1990; Norton-
Griffiths, 1995). Therefore, 1988 represents relatively undisturbed conditions only
4 years after the Talek River was established in 1984 as the Reserve edge. By
comparing spatial data obtained from the Talek clan between 1988-90 with those
observed between 1996-98, I examined the space utilization behavior of Talek

hyenas to inquire whether edge effects were currently affecting the Talek clan.

125

Given the expectation stated above, that hyenas would avoid areas grazed by
livestock, here I expected that hyenas in 1988-90 would not avoid those
particular areas since grazing of livestock south of the Talek River was rare prior
to 1990. Thus, I expected that hyenas in 1996-98 would exhibit different patterns
of space utilization as a result of anthropogenic disturbance than those observed

among Talek hyenas in 1988-90.

METHODS

Study site.— Between July 1996 and April 1998, I documented space
utilization patterns of 13 adult female members of one large spotted hyena clan
in Kenya. The study clan, which usually contains 70 to 80 hyenas, defends a
territory encompassing 62 km2 with the territorial borders as depicted in
Figure 6-1 (see also Chapters 2 and 3) in the Talek region of the Masai Mara
National Reserve, Kenya. This is an area of open rolling grassland interspersed
with seasonal creek beds. Large concentrations of several resident ungulate
species graze this area year round. The most numerous resident species are
Thompson’s gazelles (Gazelle thomsonil), topi (Damaliscus korn'gum), and
impala (Aepyceros malampus). These resident species are joined for three or
four months each year by large migratory herds of wildebeest (Connochaetes
taurinus) and zebra (Equus burchelll) from the southern part of the Serengeti.
Territorial borders for the Talek clan of hyenas were known from observations of
group territorial behaviors, including border patrols and clan wars (Boydston et

al., in press).

126

Here I used the terms “borders” to refer to the defended borders of the
Talek clan territory, while “boundaries” referred to the officially designated
boundaries of the Mara Reserve. The Talek clan “territory” referred to the
defended borders shown in Figure 6-1. I used the term “edge” to specify the line
separating a protected and an unprotected area. Specifically in this study, “edge”
referred to either that part of the border of the Talek clan territory that was the
same as the Mara Reserve boundary (the Talek River), or the term referred more
generally to any boundary of the Mara Reserve.

Hyenas: field data.— The methods utilized here in the collection of spatial
data on hyenas were described in Chapter 5 in the section Methods: Collection of
spatial data on hyenas.

Hyenas: GIS.— GPS coordinates were recorded for about 40 landmarks
distributed throughout the study site, and these were used to create a digital
photo-mosaic map for use in a Geographic Information System (GIS) as a base
map. Air photos taken in 1991 by Kenya Wildlife Service and Kenya Rangelands
and Ecological Monitoring Unit were scanned into a computer at 1 m resolution.
Using ERDAS/Imagine software, the separate photos were edge-matched and
geo-rectified with the GPS landmark data in UTMs to create a digital map of the
study site on which trees, stream crossings, and other landmarks were clearly
visible. The borders of the Talek clan’s territory were digitized as a separate
overlay coverage (Boydston et al., in press), as were the locations of all dens

used by the hyenas during the study period. Finally, locations of hyenas were

127

digitized so that each point in space at which each individual was tracked or
opportunistically found was associated with UTM coordinates.

To document space utilization patterns, I calculated fixed kernel 95%
utilization distribution (UD) contours and also UD grids (Worton, 1989; Powell,
2000), using ArcView GIS software with the Animal Movement Extension (Hooge
& Eichenlaub, 1997). Both 3 UD contour map and its corresponding UD grid
could be calculated from the same set of locations (see Figure 6-3 for an
example). In Chapter 5, I presented only UD contours and did not discuss the
corresponding UD grids. The 95% contours showed the extent of the area that
hyenas utilized with a 95% probability. UD grids were grids of equal-sized cells,
in which each cell had a value that indicated the probability of use by a hyena of
that particular cell. In other words, each cell had a value associated with the
probability that a hyena was likely to be found there. Grid cell values from the
fixed kernel calculation were based on the relative densities of locations where
an individual hyena was found. For example, cells with large numbers of
locations in them, and also in adjacent cells, would have high probabilities of
utilization by hyenas while cells with few observations of hyenas in them and in
neighboring cells would have low probabilities. One 95% UD contour and a
corresponding UD grid were calculated for each female in each time interval (for
time intervals, see section below, Time intervals and spatial scales). Individual
space utilization patterns were averaged across females to represent the space

utilization of the clan.

128

Lions: field data.— Locations at which lions were opportunistically sighted
in and around the Talek clan territory were recorded, and group composition
(sexes and age classes) of lions were also noted at each sighting. Individual
identifying marks such as ear notches and whisker spot patterns were used to
identify members of the resident pride.

Lions: GIS.— Locations for lions observed in the Talek clan’s territory
were digitized as a layer in the GIS database. Utilization distribution (UD) grids
showing the areas of highest probability of use were calculated with the Animal
Movement extension for all lions, as described above for hyenas. For the
calculation of these UD maps, each sighting of a lion or a group of lions was
counted as one location.

Livestock and pastoralists: field data.— Two methods were used to
identify areas that pastoralists utilized for grazing domestic livestock and to
document numbers of livestock present in the Talek clan territory. The first
method was to run a transect line along the south side of the Talek River to
provide a conservative estimate of total livestock numbers inside the Talek clan
territory based on the numbers seen entering in the mornings and leaving in the
late afternoons. This transect was run one to three mornings each week between
0830 and 1000 and one to three evenings each week between 1630 and 1800.
These were the times when pastoralists were most likely to be leading their herds
into or out of the Reserve. In the morning, the transect was driven from west to
east, and in the afternoon, it was driven from east to west. Herds located within 2

km of the Talek River could be seen, and location, type of herd (cows, goats, or

129

sheep), and size were recorded for each herd seen. Herds containing between
100 and 500 individuals were estimated to the nearest 50 animals, and herds
greater than 500 animals were estimated to the nearest hundred.

The second method was a comprehensive effort to locate all livestock
herds within the clan’s territory during the middle of the day, by which time the
herds had traveled their maximum distance into the Reserve. This survey was
initiated in May 1997 and conducted once per week through mid-March 1998. A
circuit was driven through the entire Talek clan territory and scans were
conducted from highpoints in order to locate all livestock herds within the
mmmw.

Locations of pastoralist settlements outside the Mara Reserve were
digitized. Settlements to the north of the Talek clan territory (i.e., along the north
side of the Talek River that formed the northern border of the Talek clan territory
and also bounded the Reserve in this area) and east of the Talek clan territory
were digitized, as described below (see section on Historical data: comparisons
of 1988-90 with 1996-98).

Livestock and pastoralists: GIS.— Raster (i.e., pixel or grid) maps
representing the relative extent to which pastoralists utilized different areas (grid
cells) for grazing domestic livestock were created through surface interpolation
using the GIS software ArcView with Spatial Analyst. In the GIS, a “surface” of
relative grazing intensity was interpolated from the comprehensive survey data

for each survey date, incorporating both locations and sizes of herds. Data from

130

the riverside transect were used estimate numbers of cattle grazing daily and to
identify the areas through which most herds traveled.

Ungulate prey: field data.— Food availability was monitored by counting
all ungulate prey animals found within 100 m of transect lines as described
previously (Holekamp et al., 1993; Holekamp et al., 1996). Here we expanded
the previous system of 8 km total of transects censused once every two weeks to
include a total of 18 km of transects (shown in Figure 6-2) censused once per
week. The numbers and species of all ungulates observed within 100 m on either
side of the transect line were recorded per km. From September 1, 1996 through
November 30, 1997, the 18 km of transects were counted four to five times per
month. From December 1, 1997 through March 22, 1998 due to the heavy El
Niflo rains, the transects were counted approximately every ten days, for an
average of 2 or 3 times per month.

Ungulates: GIS.— Raster (i.e., pixel or grid) maps of ungulate densities
were created through surface interpolation using the GIS software ArcView with
Spatial Analyst. Each 1 km of transect was digitized as three points, a midpoint
and a point 400 m to each side of the midpoint. Then for each transect day, each
point was assigned the attribute of one-third the total number of ungulates
counted on that 1 km transect on that day. A “surface” of prey densities was then
interpolated from these values in the GIS with an inverse-density-weight function.
The surface interpolation function created a new raster map of relative ungulate

abundance for each census date.

131

 

Ungulate prey
transects

 
    
 

Census years _
— 1996-98 only 2 Km
1988-90 &1996-98 E

 

 

 

Figure 6-2. All 18 km of transects used to estimate numbers of ungulates
each week during 1996-98. (a) Two 4km transects (striped lines) were

also conducted in 1988-90, every 2 weeks. (b) Most of the vegetation in
the Talek clan’s territory was open grass, with some moderate and dense
cover primarily along water courses. Areas of long grass and short grass
inside the territory are labeled, and a circle surrounds the main short
grass area. The unlabeled area to the east was heterogeneous habitat of
grass, bushes, and woodlands.

132

 

Vegetation map.— A vegetation map with 50 m resolution was digitized
from 1:20,000 air photos as a GIS layer (Figure 6-2b). Each 50 m2 grid cell was
assigned to one of three classes of vegetation. “Grass” areas were open areas of
short or long grass with less than 10% of the area (per 50 m2) covered by bushes
or trees. “Moderate” cover included areas that were 10-50% bushes or trees (per
50 m2). Areas of “dense” cover areas were more than 50% bushes or trees (per
50 m2).

Time intervals and spatial scales.— To evaluate the relationships between
the independent variables and hyena space utilization patterns, I divided the
1996-98 time period into five “seasonal” segments of 3-5 mo each, based on
changes in prey abundance, such as the influx of the wildebeest migration.
Division of the study period into these shorter periods allowed me to control for
variation in hyena UDs due to den moves and seasonal conditions (e.g., when
the total available prey was low vs. high). For each statistical test, results were
first calculated within each of the five “seasonal” periods to determine whether
trends were the same regardless of periodic changes in abundance.

Because of the different methods for estimating densities and distributions
of hyenas, lions, cattle, and ungulate prey and the different errors associated with
each, I decreased the spatial resolution of some GIS layers, such as hyena
utilization, and used a coarse spatial scale that encompassed the error of all
layers for analyzing the relationship of hyenas and the independent variables.
Cells of 500 X 500 m were assumed to be a conservative size for spatial

comparisons that would encompass the spatial errors for the different variables

133

while still giving sufficient spatial resolution, and only truly robust patterns should
appear at this scale. All spatial analyses were restricted to the area within the
borders of the Talek clan territory.

Historical data: comparisons of 1988-90 with 1996-98.— Locations for 13
Talek females observed between September 1, 1988 and April 1, 1990 were
digitized from field notes collected by K. Holekamp and L. Smale. These
locations were digitized as a data layer of points in the GIS database, using the
same methods as described in Chapter 5 for the 1996-98 data. Female hyenas
were not radio-collared in 1988-90. Therefore, when comparing the 1988-90 data
to 1996-98 data, all radio-tracked locations from 1996-98 were excluded, and
only opportunistic sightings were included.

Two ungulate census transects of 4 km each, one in short grass and one
in long grass, were conducted in 1988-90 and also in 1996-98 (Figure 6-2a), and
data from these two transects could be directly compared across time. Rainfall
measurements for one location within the Talek clan’s territory were also
recorded daily throughout both time periods.

Locations of Masai settlements, including villages called “manyattas” and
livestock corrals called “bomas,” were known from air photos in 1991 and a field
survey in 2000. Manyattas were small villages composed of a few huts in a circle
surrounded by a thick fence of cut Acacia branches. Other structures, called
bomas, were fences only and contained no huts, but were situated near a house
or manyatta. Inside the manyattas and bomas, livestock were regularly corralled

overnight. In the mornings, Masai men or boys herded livestock out of the

134

 

manyattas or bomas for grazing. Locations of settlements outside the Reserve in
1991 (there were no Masai settlements inside the Reserve at that time) to the
north and east of the Talek clan territory were digitized from air photos. Locations
were also digitized from a field survey of settlements in October 2000 in the
same areas as those visible on the 1991 photos. The 1991 data were assumed
to represent 1988-90 Masai village locations, and the 2000 data were assumed
to be a good estimate of 1996-98 locations, although, in both cases, data may
slightly overestimate the numbers of villages actually present in each time period.

Any other historic data needed for comparisons of ecological conditions
between 1988-90 and 1996-98 were taken from the literature.

Some images in this chapter are presented in color.

RESULTS

Hyena space utilization pattems: 1996-98.— Talek hyenas were found
throughout their territory and occasionally also outside the defended borders of
their territory (Figure 6-3a). The utilization distribution grid (Figure 6-3b) and 50%
UD contour (Figure 630), averaged from the space utilization patterns of all 13
collared females in 1996-98, showed that two particular parts of the territory were
intensively used: one ‘core area’ in the west and one in the east. Much of the
defended territory fell outside the 95% UD contour. (Space utilization patterns are
discussed in detail in Chapter 5.)

Lions: 1996-98.— Lions were frequently seen throughout the study period

in all parts of the Talek hyena clan’s territory during 1996-98. From July 14, 1996

135

 

9.
O . '
o . 0 O
o 0 0 o
. 0 V o
g “a“. o.
. o’J.-°-,-'.‘---' - ". -
' I
O 'i

 

 

 

 

 

 

Figure 6-3. Three different representations of the
aggregate space utilization pattern of 13 female hyenas
studied in 1996-98. (a) All locations (n = 4,838) where the
hyenas were found. (b) A fixed kernel UD grid. The
darkest cells are those in which hyenas were most likely
to be found. (c) UD contours. The entire green area is the
95% UD and the small interior shapes are the 50% U0.

136

through March 25, 1998, lions were seen 299 times in the Talek Clan territory
(Figure 6-4). On average, 14.2 :I: 1.5 individual lions or groups were seen per
month with an average group size of 3.7 :l: 0.2 lions per sighting. More than half
of these sightings (n = 164) were of lions from the resident pride. This pride
appeared to utilize an area overlapping with much of the Talek clan territory,
primarily the eastern two-thirds (Figure 64) Most sightings of alien or
unidentified lions were in the western portion of the Talek clan’s territory. Alien
lions may have been nomadic or they may have been members of other lion
pfides.

The resident lion pride consisted of three adult males and 4 adult females
in July 1996. The reigning coalition of males appeared to have arrived by May
1996. The four females gave birth to eleven cubs around October 1996. One of
the eleven cubs disappeared in September 1997, but the remaining ten were still
alive and still with the pride throughout the study (and even into summer 2000).
One of the three adult males was speared by Masai pastoralists in the Talek
River bed in February 1997. Pride composition did not change during the
remainder of the study period.

There was no regular use of the Talek clan territory by any other pride, but
alien lions were frequently seen. Some alien lions were repeatedly seen over the
course of a few weeks but then were never seen again. Lions were always
present in the hyena clan territory, but the resident pride appeared to have been
absent for 3 months from September 1, 1997 through November 30, 1997 when

there were no sightings of any pride members.

137

.mmémmp mezzo EQEE :30 x23 or: E notomno So; mac: 533 am 8mm u 5 96:80.. To 659”.

 

 

 

 

138

Hyenas and lions had positively correlated spatial distributions, with
hyenas being more likely to be found in areas with higher probability of use by
lions (r = 0.46, P < 0.001; Figure 6-5). Thus the spatial overlap between these
two competing species documented at coarse spatial scales (e.g., Mills &
Gorman, 1997) also appears to occur at finer spatial scales as well.

Livestock and pastoralists: 1996-98.— Herders brought livestock into the
park almost daily throughout the study period, and there was never a span of
more than a few days when cattle did not enter the Talek clan’s territory.
Throughout the 1996-98 study period, pastoralists usually brought herds into the
Reserve in the morning between 0800-1000, lining up with their herds to cross
the Talek River at a few particular points where the river banks were less steep
than elsewhere (Figure 6-6a). Goats and sheep, generally herded by young
Masai boys, stayed close to the Talek River (within 3 km). Herds of cattle,
accompanied by Masai boys or men, then grazed several kilometers into the
Reserve from their points of entry. Pastoralists usually herded their livestock out
of the Reserve by nightfall. Cattle were grazed in the park on 90% of days when
observers were present in the study area. The main reason for no cattle entering
the park on 10% of the days was a high water level in the Talek River, which
prevented herders and cattle from crossing. On fewer than 5 days out of the
entire study period (n = 620 d) was a lack of grazing due to enforcement of anti-
grazing policies by game rangers.

Including days when livestock did not enter the Reserve, an estimated

daily average of 1,580 :1: 140 cows, goats, and sheep were grazed each day in

139

 

 

O C
0 10 20 30 40 50 60 70
I Lion probability of use per grid cell

yena probability of use per grid cell

Figure 6-5. UD grid probabilities for Talek hyenas were positively correlated
with those of lions in 1996-98.

140

 

 

 

/ 0 C8
0

\
“' A Cattle crossings
*- X Cow travel routes

0 Villages&corrals
" _ low Cattle grazing
" [3 intensity

hi
N Water courses 1996-98 B

 

 

 

‘

Talek clan
K 95% UD

 

 

 

 

Figure 6-6. (a) Locations of villages or corrals and cattle crossings along the
northern and eastern borders of the Talek clan territory and (b) GIS grid showing
cattle grazing intensity in relation to the Talek clan's 95% UD contour.

141

the Talek clan’s territory (Figure 67). Livestock numbers exceeded 3,000 on at
least 20% of days and ranged up to 6,000 animals per day. Cattle were grazed
throughout the Talek clan’s 95% UD. but the largest area of intense cattle
grazing was not a part of this 95% UD. Thus hyenas were seldom found in the
largest area of intense cattle grazing (Figure 6-6b).

Ungulate prey: 1996-98.— Like lions and cattle, ungulate prey were
unevenly distributed across Talek clan’s territory (Figure 6-8). At any given time,
most of the prey were concentrated on a central short grass plain. The
distribution grid of prey was heavily skewed with only a few grid cells having very
high values. In order to avoid statistical relationships weighted towards these few
cells containing extreme values, all grid cells were ranked by their prey
abundance value and then aggregated into 10 ‘quantiles’ with an equal number
of cells per quantile. Each quantile thus represented 10% of the total area of the
clan’s territory, with the bottom quantile containing cells with the least abundant
prey and the top quantile containing the greatest abundance (Figure 6-9a).
Quantiles 1 to 6 each contained less than 10% of the total prey. In other words,
60% of the clan’s territory contained a lower percent of prey than would have
been expected based on the size of the area (Figure 6-9a). By contrast, the 10%
of the grid cells with the most prey, the 10th quantile, contained a
disproportionately large amount of prey (Figure 6-9a).

There was significant variation in hyena UD probabilities associated with
varying levels of prey abundance (F = 14.44, P < 0.001, d.f. = 9; Figure 6%).

Hyenas were significantly more likely to be found in quantiles of relatively high

142

5000 "

4000 '

 

3000 " '

 

 

2000

 

'I

.l-
I . I J J .
1 1 1 1I it“
®5fi§> 1,9 1.9 $9,231 $0 .193 $9

1 000

 

'1
‘
'_7

Mean daily livestock estimate

 

     

 

2-week intervals
(Month-Year labeled for first half of month)

Figure 67. Estimated number of livestock animals grazing daily in the
Talek clan’s territory fluctuated over time, but was usually over 1,000
animals.

143

.mE: swam Em H 85:95 65 Co “88 umsflcoo

.85 522.65. 3.25 m__oo 7.9.80 och .0638 :9: 2 3695 SEE So. So: .oocmocsnm
35 22:95 Co w_o>o_ 623.9 E3632 35 Lo moumcm E m=oo 2.0 .8cmvcznm >65 9:355
25 m 8 623.9 E35 c: 52:8 a: $3 :90 0.22. 8-89 05 Co c0382 9:. .m-o 95mm

 

 

EXN w o F

I‘ll

   

no?“
cotioi

0: 3mm
:30 x22.
wmémaw

 

 

 

144

Figure 6-9. (a) Histogram indicating that most prey were concentrated in a small
portion of the Talek clan territory. Each “quantile” was 10% of the total area of
the Talek clan territory (Le, 10% of the grid cells comprising the territory). On
the y-axis is the estimated percent of prey per quantile, ranked from the quantile
with the least to the most. (b) Talek hyenas were relatively unlikely to be found
in grid cells that had disproportionately little prey. Stars indicate quantiles of
prey abundance in which hyenas were significantly more likely to be found than
in other quantiles.

145

 

 

 

 

 

 

 

 

 

 

 

 

 

 

30'
A
5, 7'1
520'
7.9
e ,7
o\°10- _11
O nflllll . . .
2 4 6 8 1O

Prey abundance quantiles

 

increasing prey abundance
*

Ill
10-+Ir 1 +

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Mean hyena probability per 500n‘?

2 4 6 8 10

Prey abundance quantiles
%

increasing prey abundance

 

146

prey abundance (Tukey’s post-hoc multiple comparisons; Figure 6-9b), but they
were unlikely to be found in the area of highest prey abundance even though this
area (the 10th quantile) held almost a quarter of the total estimated prey (Figure
6-9a). This result struck me as very peculiar, and it was in fact this unexpected
result that prompted me to compare hyena space utilization in 1996-98 with that
observed in 1988-90 when the Michigan State University Mara hyena project
began.

Historical comparison: 1988-90 vs. 1996-98.— The historical comparison
of data sets that were collected 8 yr apart revealed that hyena space utilization
patterns in the Talek clan have changed markedly in recent years (Figure 6-10;
Table 6-1). In 1988-90, the mean sizes of 50% UDs and 95% UDs for the
females were 2.3 t 1.1 km2 and 17.7 :I: 5.7 km2, respectively. Mean sizes of both
50% and 95% UDs were significantly larger for females in 1996-98 (based only
on non-radio-tracked data; t-test on log 50% UDs: T= 2.101, P = 0.046, d.f. = 24;
t-test on log 95% UDs: T= 3.482, P = 0.003, d.f. = 24). 1996-98, the mean sizes
of 50% U05 and 95% UDs for the females were 3.5 :l: 1.6 km2 and 28.4 t 6.3
kmz, respectively. Furthermore, females in 1996-98 were found, on average,
farther from the active communal den than were females in 1988-90 (t-test: T =
5.696, P < 0.001, d.f. = 24). The most striking difference in the behavior of
females was that Talek hyenas regularly utilized the prey-rich central plain in
1988-90 (Figure 6-10, Figure 6-13), but they appeared to avoid this area in 1996-

98, when they also tended to range more widely (Figure 6-8, Figure 6-13) .

147

.3226 8:5 8-82

van $326 :33 86me c_ mo_mE& x22. 9 Lo:— mcoamoo. o=m_c:toaao .96 9:9“.

 

 

waémg O

oméwmr O
mcozmoou

 

 

 

148

Table 61. Average distances of hyena locations to dense vegetation and the
percent of non-grass vegetation inside the 95% UD contours showed that
Talek hyenas in 1996-98 were found closer to dense cover, and that they
utilized an area encompassing a higher percent of moderate and dense cover.

 

t-test,
n = 13 hyenas

378114 280:l:17 T=4.5,P<0.001

1 988-90 1 996-98

 

Distance (m) to dense
cover per hyena

% Locations < 200 m to 30 i 2% 47 :1: 4% T = 3.5 P < 0.005
dense cover per hyena ’

Amount of moderate
or dense cover 10% 17%
inside 95% UD

149

Although the vegetation in this area has changed little since the 1980s, the
use of it by hyenas has changed significantly in recent years. Table 6-1 shows
that hyenas in 1996-98 were seen significantly closer to dense cover than were
hyenas in 1988-90, and that the shape of the 1996-98 95% UD for the Talek clan
encompassed a greater percent of moderate and dense cover than did the 95%
UD in 1988-90. In other words, Talek hyenas in 1996-98 were using bushier
areas and avoiding very open areas, relative to the areas used by Talek hyenas
in 1988-90.

Comparisons of ecological conditions in 1988-90 with those observed in
1996-98 showed no significant changes in lions, prey abundance or distribution,
or rainfall that could explain the space utilization changes of the hyenas. Lions
were present in this area in 1988-90 (Ogutu & Dublin, 1998), with the resident
pride having a territory that overlapped with the territory of the Talek clan (Ogutu,
unpubl. data). Numbers of lions were most likely either the same in 1988-90 and
1996-98, or lower in 1996-98, due to a severe canine distemper outbreak in 1994
that reduced lion numbers throughout the Serengeti-Mara ecosystem (Kock et
al., 2000; Packer et al., 2000).

Numbers of ungulates counted on short grass and long grass transects
(shown in Figure 6-2a) were very similar between 1988-90 and 1996-98 with
respect to both abundance and seasonal patterns in the two habitat types (Figure
6-11). Also, rainfall measurements showed the same seasonal patterns of rainfall

and similar amounts of rain in the two time periods (Figure 6-12).

150

.h

0

O
1

Years A

WC} ...... 1 988-90
-O-1 996-98

(JO

0

O
I

_x

O

O
I

 

Prey animals in long grass
N
O
O
I

 

 

O

500 "
400 ’
300 '

 

Prey animals in short grass

 

 

O llllllllllll
2 4 6 8 1012

Month

Figure 611. Monthly averages of bi-weekly prey counts for
the (a) long grass and (b) short grass transects (see Figure
6-2a). Data shown are from September 1, 1988 to April 1,
1990 (dashed line) and September 1, 1996 to April 1, 1998
(solid line).

151

 

 

 

 

 

(I)

”$320 A Years

$ ....9 ..... 1988-90
Q 15 +1996-98
E
£10
in

.E 50‘

cu

CE

0

32

8101'

3 9'

0'1 8-

E 7'

E 6'

“5 5'

% 32

D

=l=l: 2"

1 I . I I I . I I

2 4 6 8 10 12
Month

Figure 6-12. Monthly averages of (a) rainfall per 2-week
interval and (b) number of days with rain per 2-week interval.
Data shown are from September 1, 1988 to April 1, 1990
(dashed line) and September 1, 1996 to April 1, 1998 (solid
line).

152

Only one significant type of change in this area could be identified here
that was coincident with changes observed between 1988-90 and 1996-98 in
hyena space utilization patterns: the increased settlement of humans along the
Reserve edge and pastoralist reliance on the Reserve for grazing pasture (Figure
6-13). In 1988-90, Masai did not graze livestock in the Reserve with the
exception of occasional grazing of small herds (under 100 animals) beside the
Talek River (Holekamp, pers. comm.). Grazing in the Reserve in 1988-90
occurred, at most, on 30-35% of days observed (Holekamp & Smale, 1992) in
contrast to 90% of days observed in 1996-98. Numbers of Masai settlements
doubled along the Talek River and the eastern border of the Talek clan territory
and MMNR between 1991 (n = 36) and 2000 (n = 72). The settlement patterns
also changed, such that villages were more tightly clustered in 2000 than in
1991. In particular, the density and concentration of villages located to the north
of the Talek clan’s territory close to the Talek River increased; villages north of
the Talek clan territory that were within 1 km of the Talek River increased almost
fourfold, from 11 in 1991 to 39 in 2000 (Figure 6-13). The particularly dense
concentration of villages in the middle of the northern border of the Talek clan
territory seemed to be responsible for the pattern of dense livestock grazing in
the center of the Talek clan’s territory (Figure 6-13b). Pastoralists from this area
in the late 19905 used the closest river crossings to enter the Reserve and then
led their herds south onto the nutritious short grass plain, which was also the
prey-rich area utilized by hyenas in 1988-90 but avoided in 1996-98 (Figure 6-10,
13).

153

 

95% UD/

 

95% U0

 

 

 

 

Figure 6-13. Locations of pastoralist villages and corrals and the 95%
U0 contours based on opportunistic locations (i.e., not radio-tracked)
for Talek hyenas in (a) 1988-90 and (b) 1996-98, with also the areas

of highest cattle grazing shown for 1996-98 (see also Figure 6-7b).

154

DISCUSSION

Ecological determinants of hyena space utilization pattems: 1996-98.—
Space utilization patterns exhibited by members of the Talek clan in 1996-98
revealed that these hyenas did not utilize their territory in accordance with
optimal foraging theory predictions that hyenas should attempt to maximize their
energy intake (Krebs & Kacelnik, 1991). I expected that hyenas would forage in
areas of highest prey density in order to decrease their search time and increase
their encounter rate with potential prey items. Although, hyenas were unlikely to
be found in areas of low prey abundance, they were just as unlikely to be found
in the area of greatest prey abundance. The patch containing the most abundant
herbivores was centrally located and held almost a quarter of the total prey. This
prey-rich central plain appeared to be an area of potential high-energy
profitability for a predator and should have been an attractive foraging site to
Talek hyenas. Furthermore, this area was centrally located in the Talek clan
territory, and Talek hyenas were investing energy and risking injury to defend it
against intruding conspecifics (Chapter 2). However, hyenas rarely even passed
through this area. They were instead often found to the east and west of it, and
their space utilization patterns further suggested that they were actively avoiding
this site by traveling around it when crossing their territory (Figure 6-8). Talek
hyenas in 1996-98 did not necessarily avoid all the areas that were heavily
grazed by livestock, but they did avoid the largest area of intense cattle grazing,

which was the prey-rich central plain.

155

There was no evidence that hyenas avoided certain areas due to the
presence of lions. Hyenas and lions had positively correlated spatial patterns, but
lions did utilize the central plain where hyenas were seldom found. Despite the
overlap in areas of use, interactions between hyenas and lions were uncommon
during regular hours of observation, but the few interactions witnessed were
intense. During this study, hyenas and lions were seen feeding together on the
same large carcasses, hyenas were seen mobbing and taking kills from lions,
and lions were seen taking kills from hyenas. Examination of hyena carcasses
revealed that at least four Talek hyenas were killed by lions during 1996-98 (one
adult female “DJ”, one adult male “MIKE”, one 7-month old cub “SHA”, and one
juvenile female “LOP”). Despite the mortality threat to hyenas posed by lions,
Talek hyenas did not avoid areas utilized by lions. Instead, the two species of
large carnivores utilized largely the same areas due to similar ecological
requirements.

Change in hyena space utilization: 1988-90 vs. 1996-98.— The defended
borders of the Talek clan territory have remained remarkably stable since 1988
(Boydston et al., in press), and the demography of the clan has remained
similarly stable to date (Holekamp, unpubl. data). However, clan members
exhibited dramatically different space utilization patterns in 1988-90 and 1996-98.
Comparisons of 1996-98 hyena utilization distributions with those observed in
1988-90 suggest that increasing human activity within the clan’s territory has
altered hyena space-use patterns. At no time in 1996-98 did Talek hyenas exhibit

optimal space utilization with respect to the distribution of prey. The avoidance of

156

the prey-rich central plain appeared to be a recent phenomenon coincident with
increased concentration of human settlements along the MMNR boundary in this
area.

Between 1988-90 and 1996-96, no significant changes were observed in
rainfall patterns or ungulate distributions, so neither of these variables could
account for the observed change in hyena space utilization. Ungulate counts
conducted during both time periods revealed no change in the abundance of
ungulates in either long grass or short grass habitat. The only noticeable change
in the Mara ecosystem documented here was the increase in human density
along the Reserve boundary and the increased reliance on the Reserve for
livestock grazing. In 1996-98, cattle grazed heavily in the interior of the territory
defended by the Talek clan. As cattle moved in and out of the Reserve across
the Talek River, the entire area along the river was generally heavily utilized by
pastoralists either for grazing or for transit to other areas. Clearly hyenas utilized
some areas grazed by cattle in 1996-98. Of the areas that hyenas continued to
share with pastoralists, many contained relatively dense vegetation cover, that is
bushes and trees, that hyenas may have used to hide from cattle herders.
Hyenas may have also avoided cattle by altering their use of time as well as
space. Since pastoralists tended to graze their cattle during daylight hours, Talek
hyenas have apparently shifted their activity patterns between 1988-90 and
1996-98 to be more active at night than in the daylight hours when cattle were

present. Indeed preliminary data collected at call-ins in summer 2000 suggest

157

that the biological rhythms of Talek hyenas have shifted considerably in the past
decade (Holekamp et al., unpubl. data).

Several recent workers have emphasized a need to understand individual
behavior of large mammalian carnivores in order to conserve them in the face of
burgeoning human populations and widespread habitat fragmentation (Creel,
1998; Berger, 1999; Caro, 1999). Woodroffe and Ginsberg (1998; 1999; 2000)
and Woodroffe (2000) demonstrated that edge effects have decimated carnivore
populations around the world. However, there is a lack of information to help
identify the warning signs of demographic decline and ameliorate the effects of
fragmentation. In particular, we lack empirical data on how ranging behavior may
change in fragmented landscapes and over time (Caro, 1999). This study
however provides an example of such consequences for one species of large
carnivore. The long-term data on the Talek clan and the location of the Talek
clan’s territory at the edge of the Mara Reserve offer a unique opportunity to
explore effects of increasing anthropogenic disturbance on behavior of a well-
studied carnivore population. Woodroffe and Ginsberg (1998, 2000) focus on
edge effects operating when carnivores range into areas of human activity, but
they assume that this occurs primarily when predators move from a protected
area into areas of human settlement. The behavioral changes in the Talek
hyenas documented here appeared to be a consequence of edge effects
occurring inside the Reserve, due to erosion of the Reserve boundary. Currently
Talek hyenas appear to be under temporal and spatial constraints that may soon

force them beyond the limits of their behavioral and ecological plasticity. The

158

edge effects observed here in the current study have not yet caused major
demographic changes in the study population, but the behavioral changes
documented may well represent warning signs of environmental degradation in

this important ecosystem.

159

LITERATURE CITED

Altmann, J. 1974. Observational study of behavior: sampling methods.
Behaviour, 49, 227-267.

Andreassen, H. P., Ims, R. A., Stenseth, N. C. & Yoccoz, N. G. 1993.
Investigating space use by means of radiotelemetry and other methods: a
methodological guide. In: The Biology of Lemmings (Ed. by Stenseth, N.
C. & lms, R. A.), pp. 590-618. London: Academic Press.

Andreka, G., Linn, l. J., Perrin, M. R. & Maddock, A. H. 1999. Range use by the
wild dog in the HIuhluwe-Umfolozi Park, South Africa. South African
Journal of Wildlife Research, 29, 1-9.

Arcese, P. & Sinclair, A. R. E. 1997. The role of protected areas as ecological
baselines. Joumal of Wildlife Management, 61, 587-602.

Avenant, N. L. & Nel, J. A. J. 1998. Home-range use, activity, and density of
caracal in relation to prey density. African Journal of Ecology, 36, 347-359.

Ballard, W. B., Whitman, J. S. & Gardner, C. L. 1987. Ecology of an exploited
wolf population in south-central Alaska. Wildlife Monographs, 98, 1-54.

Bateman, A. J. 1948. lntra-sexual selection in Drosophila. Heredity, 2, 349-368.

Bekoff, M. & Mech, L. D. 1984. Simulation analyses of space use: Home range
estimates, variability, and sample size. Symposia of the Zoological Society
of London, 1 6, 32-36.

Bekoff, M. & Wells, M. C. 1982. Behavioral ecology of coyotes: Social
organization, rearing patterns, space use, and resource defense.
Zeitschn‘ft fur Tierpsychologie, 60, 281 -305.

Bekoff, M. & Wells, M. C. 1986. Social ecology and behavior of coyotes.
Advances in the Study of Behavior, 16, 251 -338.

Berger, J. 1999. Anthropogenic extinction of top carnivores and interspecific
animal behaviour: Implications of the rapid decoupling of a web involving
wolves, bears, moose and ravens. Proceedings of the Royal Society of
London 8, 266, 2261-2267.

160

Bowen, W. D. 1982. Home range and spatial organization of coyotes in Jasper
National Park, Alberta. Journal of Wildlife Management, 46, 201-216.

Boydston, E. E., Morelli, T. L. & Holekamp, K. E. in press. Sex differences in
territorial behavior exhibited by the spotted hyena (Hyaenidae, Crocuta
crocuta). Ethology.

Brown, J. L. & Orians, G. H. 1970. Spacing patterns in mobile animals. Annual
Review of Ecology and Systematics, 1 , 239-262.

Burrows, R. 1995. Demographic changes and social consequences in wild dogs,
1964-1992. In: Serengeti Il : dynamics, management, and conservation of
an ecosystem (Ed. by Sinclair, A. R. E. & Arcese, P.), pp. 400-420:
University of Chicago Press.

Burt, W. H. 1943. Territoriality and home range concepts as applied to mammals.
Journal of Mammalogy, 23, 346-352.

Caro, T. 1999. The behaviour-conservation interface. Trends in Ecology and
Evolution, 14, 266-269.

Caro, T. M. 2000. Controversy over behaviour and genetics in cheetah
conservation. In: Behaviour and Conservation (Ed. by Gosling, L. M. &
Sutherland, W. J.). Cambridge: Cambridge University Press.

Clutton-Brock, T. H. 1988. Reproductive Success. Chicago: University of
Chicago Press.

Cooper, S. M. 1989. Clan sizes of spotted hyaenas in the Savuti Region of the
Chobe National Park, Botswana. Botswana Notes and Records, 21, 121-
131.

Cooper, S. M. 1991. Optimal hunting group size: the need for lions to defend
their kills against loss to spotted hyaenas. African Journal of Ecology, 29,
130-136.

Cooper, S. M. 1993. Denning behavior of spotted hyaenas (Crocuta crocuta) in
Botswana. African Journal of Ecology, 31, 178-180.

Cooper, S. M., Boydston, E. E., Lansing, S. & Holekamp, K. E. in prep. Denning
and provisioning behaviour of the spotted hyaena in relation to the
accumulation of bones at dens. In: Hyaenids on the Circum-
Mediterranean: A Taphonomic Perspective (Ed. by Fosse, P. & Horwitz, L.
K.).

161

Creel, S. 1998. Social organization and effective population size in carnivores. In:
Behavioral ecology and conservation biology (Ed. by Caro, T.), pp. 246-
264. Oxford: Oxford University Press.

Creel, S. 8. Creel, N. M. 1996. Limitation of African wild dogs by competition with
larger carnivores. Conservation Biology, 10, 526-538.

Crooks, K. R. & Soule, M. 1999. Mesopredator release and avifaunal extinctions
in a fragmented system. Nature, 400, 563-566.

Daniels, T. J. & Bekoff, M. 1989. Spatial and temporal resource use by feral and
abandoned dogs. Ethology, 81, 300-312.

Darwin, C. 1871. The Descent of Man and Selection in Relation to Sex. London:
John Murray.

Doolan, S. P. & Macdonald, D. W. 1996. Dispersal and extraterritorial
prospecting by slender-tailed meerkats (Suricata suricatta). Journal of
Zoology. London, 239, 697-716.

Durant, S. M. 1998. Competition refuges and coexistence: an example from
Serengeti carnivores. Journal of Animal Ecology, 67, 370-386.

Dyer, F. C. 1994. Spatial cognition and navigation in insects. In: Behavioral
Mechanisms in Evolutionary Ecology (Ed. by Real, L. A.), pp. 66-98.
Chicago: University of Chicago Press.

East, M., Hofer, H. & Turk, A. 1989. Functions of birth dens in spotted hyaenas
(Crocuta crocuta). Journal of Zoology, London, 219, 690-697.

East, M. L. & Hofer, H. 1991. Loud-calling in a female-dominated mammalian
society: ll: Behavioural contexts and functions of whooping of spotted
hyaenas, Crocuta crocuta. Animal Behaviour, 42, 651-669.

Emlen, S. T. & Oring, L. W. 1977. Ecology. sexual selection, and the evolution of
mating systems. Science, 197, 215-223.

Engh, A., Esch, K., Smale, L. & Holekamp, K. E. 2000. Mechanisms of maternal
rank 'inheritance' in the spotted hyaena. Animal Behaviour, 60, 323-332.

Ewer, R. F. 1973. The Carnivores. Ithaca: Cornell University Press.

Frank, L. G. 1986a. Social organization of the spotted hyaena (Crocuta crocuta).
I. Demography. Animal Behaviour, 34, 1500-1 509. .

162

Frank, L. G. 1986b. Social organization of the spotted hyaena (Crocuta crocuta).
ll. Dominance and reproduction. Animal Behaviour, 34, 1510-1527.

Frank, L. G., Glickman, S. E. & Powch, I. 1990. Sexual dimorphism in the spotted
hyaena (Crocuta crocuta). Journal of Zoology, London, 221 , 308-313.

Frank, L. G., Holekamp, K. E. & Smale, L. 1995. Dominance, demography, and
reproductive success of female spotted hyenas. In: Serengeti ll :
dynamics, management, and conservation of an ecosystem (Ed. by
Sinclair, A. R. E. & Arcese, P.), pp. 364-384. Chicago: University of
Chicago Press.

Fritts, S. H. & Mech, L. D. 1981. Dynamics, movements, and feeding ecology of a
newly protected wolf population in northwestern Minnesota. Wildlife
Monographs, 80, 1-79.

Fuller, T. K. & Kat, P. W. 1990. Movements, activity, and prey relationships of
African wild dogs (Lycaon pictus) near Aitong, southwestern Kenya.
African Journal of EcologY. 28, 330-350.

Fuller, T. K., Mills, M. G. L., Bomer, M., Laurenson, M. K. & Kat, P. W. 1992.
Long distance dispersal by African wild dogs in East and South Africa.
Journal of African Zoology, 106, 535-537.

Gascon, D. & Miller, R. A. 1982. Space utilization in a community of temperate
reef fishes inhabiting small experimental artificial reefs. Canadian Journal
of Zoology, 60, 798-806.

Gittleman, J. L. & Harvey, P. H. 1982. Carnivore home-range size, metabolic
needs and ecology. Behavioral Ecology and Sociobiology, 10, 57-63.

Greenwood, P. J. 1980. Mating systems, philopatry and dispersal in birds and
mammals. Animal Behaviour, 28, 1140-1162.

Grinnell, J., Packer, C. & Pusey, A. E. 1995. Cooperation in male lions: kinship,
reciprocity or mutualism? Animal Behaviour, 49, 95-105.

Hamilton, W. D. 1964a. The genetical evolution of social behavior. Joumal of
Theoretical Biology. 7, 1-16.

Hammerstein, P. 1981. The role of assymetries in animal contests. Animal
Behaviour, 29, 1 93-205.

Harris, 8., Cresswell, W. J., Forde, P. G., Trewhella, W. J., Woollard, T. & Wray,
S. 1990. Home-range analysis using radio-tracking data- a review of

163

problems and techniques particularly as applied to the study of mammals.
Mammal Review, 20, 97-123.

Harrison, D. J., Harrison, J. A. 8 O'Donoghue, M. 1991. Predispersal movements
of coyote (Canis latrans) pups in Eastern Maine. Journal of Mammalogy,
72, 752-763.

Hayne, D. W. 1949. Calculation of size of home range. Journal of Mammalogy,
30, 1-18.

Heinsohn, R. & Packer, C. 1995. Complex cooperative strategies in group-
territorial African lions. Science, 269, 1260-1262.

Heinsohn, R., Packer, C. 8. Pusey, A. E. 1996. Development of cooperative
territoriality in juvenile lions. Proceedings of the Royal Society of London
Series B, 263, 475-479.

Henschel, J. R. & Skinner, J. D. 1987. Social relationships and dispersal patterns
in a clan of spotted hyaenas Crocuta crocuta in the Kruger National Park.
South African Tydskr. Dierk., 22, 18-24.

Henschel, J. R. & Skinner, J. D. 1990. Parturition and early maternal care of
spotted hyaenas Crocuta crocuta. Joumal of Zoology, London, 222, 702-
704.

Henschel, J. R. & Skinner, J. D. 1991. Territorial behaviour by a clan of spotted
hyaenas Crocuta crocuta. Ethology, 88, 223-235.

Hirsch, G., Bender, L. C. & Haufler, J. B. 1999. Black bear, Ursus americanus,
movements and home ranges on Drummond Island, Michigan. Canadian
Field-Naturalist, 1 13, 221-225.

Hofer, H. & East, M. L. 1993a. The commuting system of Serengeti spotted
hyaenas: how a predator copes with migratory prey. III. Attendance and
maternal care. Animal Behaviour, 46, 575-589.

Hofer, H. & East, M. L. 1993b. The commuting system of Serengeti spotted
hyaenas: how a predator copes with migratory prey. l. Social organization.
Animal Behaviour, 46, 547-557.

Hofer, H. & East, M. L. 1993c. The commuting system of Serengeti spotted
hyaenas: how a predator copes with migratory prey. ll. Intrusion pressure
and commuter's space use. Animal Behaviour, 46, 559-574.

Hofer, H. & East, M. L. 1995. Population dynamics, population size, and the
commuting system in Serengeti spotted hyenas. In: Serengeti II:

164

Dynamics, Management, and Conservation of an Ecosystem (Ed. by
Sinclair, A. R. E. 8 Arcese, P.), pp. 332-363. Chicago: University of
Chicago Press.

Hofer, H., East, M. L. 8 Campbell, K. L. l. 1993. Snares, commuting hyaenas and
migratory herbivores: humans as predators in the Serengeti. Symposia of
the Zoological Society of London, 65, 347-366.

Holekamp, K. E. 1984. Dispersal in ground-dwelling sciurids. In: Biology of
Ground-Dwelling Squirrels (Ed. by Murie, J. O. 8 Michener, G. R.), pp.
297-320. Lincoln, Nebraska: University of Nebraska Press.

Holekamp, K. E. 1986. Proximal causes of natal dispersal in Belding's ground
squirrels (Spermophilus beldingr). Ecological Monographs, 56, 365-391.

Holekamp, K. E., Boydston, E. E. 8 Smale, L. 2000. Group movements in social
carnivores. In: Group Movements (Ed. by Boinski, S. 8 Garber, P.), pp.
587-627.

Holekamp, K. E., Boydston, E. E., Szykman, M., Graham, I, Nutt, K. J., Birch, S.,
Piskiel, A. 8 Singh, M. 1999a. Vocal recognition in the spotted hyena and
its possible implications regarding the evolution of intelligence. Animal
Behaviour, 58, 383-395.

Holekamp, K. E., Cooper, S. M., Katona, C. l., Berry, N. A., Frank, L. G. 8 Smale,
L. 1997a. Patterns of association among female spotted hyenas (Crocuta
crocuta). Journal of Mammalogy, 78, 55-74.

Holekamp, K. E., Ogutu, J., Frank, L. G., Dublin, H. T. 8 Smale, L. 1993. Fission
of a spotted hyena clan: Consequences of female absenteeism and
causes of female emigration. Ethology. 93, 285-299.

Holekamp, K. E. 8 Smale, L. 1992. Human-hyaena relations in and around the
Masai Mara National Reserve, Kenya. I. U. C. N. Hyaena Specialist Group
Newsletter, 5, 19-20.

Holekamp, K. E. 8 Smale, L. 1993. Ontogeny of dominance in free-living spotted
hyaenas: Juvenile rank relations with other immature individuals. Animal
Behaviour, 46, 451-466.

Holekamp, K. E. 8 Smale, L. 1998a. Behavioral development in the spotted
hyena. BioScience, 48, 997-1005.

Holekamp, K. E. 8 Smale, L. 1998b. Dispersal status influences hormones and

behavior in the male spotted hyena. Hormones and Behavior, 33, 205-
216.

165

Holekamp, K. E., Smale, L., Berg, R. 8 Cooper, S. M. 1997b. Hunting rates and
hunting success in the spotted hyena (Crocuta crocuta). Journal of
Zoology. 242, 1-15.

Holekamp, K. E., Smale, L. 8 Szykman, M. 1996. Rank and reproduction in
female spotted hyenas. Journal of Reproduction and Fertility, 108, 229-
237.

Holekamp, K. E., Szykman, M., Boydston, E. E. 8 Smale, L. 1999b. Association
of seasonal reproductive patterns with changing food availability in an
equatorial carnivore, the spotted hyaena (Crocuta crocuta). Journal of
Reproduction and Fertility, 116, 87-93.

Hooge, P. N. 8 Eichenlaub, B. 1997. Animal movement extension to arcview.
ver. 1.1. Anchorage, AK, USA: US. Geological Survey.

Kessel, A. L. 8 Brent, L. 1996. Space utilization by captive-Dom baboons (Papio
sp.) before and after provision of structural enrichment. Animal Welfare, 5,
37-44.

King, J. A. 1954. Closed social groups among domestic dogs. Proceedings of the
American Philosophical Society held at Philadelphia for promoting useful
knowledge, 98, 327-336.

Kock, R., Chalmers, W. S. K., Mwanzia, J., Chillingworth, C., Wambua, J.,
Coleman, P. G. 8 Baxendale, W. 2000. Canine distemper antibodies in
lions of the Masai Mara. Veterinary Record, 142, 662-665.

Krebs, J. W. 8 Kacelnik, A. 1991. Decision-making. In: Behavioural Ecology (Ed.
by Krebs, J. W. 8 Davies, N. 8.), pp. 105-136. London: Blackwell
Scientific.

Kruuk, H. 1970. Interactions between populations of spotted hyaenas (Crocuta
crocuta Erxleben) and their prey species. In: Animal Populations in
Relation to Their Food Resources (Ed. by Watson, A.), pp. 359-374.
Oxford: Blackwell Scientific Publications.

Kruuk, H. 1972. The Spotted Hyaena: A Study of Predation and Social Behavior.
Chicago: Chicago University Press.

Kruuk, H. 8 Macdonald, D. 1985. Group territories of carnivores: empires and
enclaves. In: Behavioral Ecology: Ecological Consequences of Adaptive
Behavior (Ed. by Sibly, R. M. 8 Smith, R. H.), pp. 521-536. Oxford:
Blackwell.

166

Landa, A., Strand, 0., Linell, J. D. C. 8 Skogland, T. 1998. Home range sizes
and altitude selection for arctic foxes and wolverines in an alpine
environment. Canadian Journal of Zoology, 76, 448-457.

Lawick, H. v. 8 van Lawick-Goodall, J. 1971. Innocent killers. London: Collins.

Lidicker, W. Z. 1975. The role of dispersal in the demography of small mammals.
In: Small Mammals: Their Productivity and Population Dynamics (Ed. by
Golley, F. B., Petrusewicz, K. 8 Ryszkowski, L.), pp. 103-128. London:
Cambridge University Press.

Lindeque, M. 8 Skinner, J. D. 1982. A seasonal breeding in the spotted hyaena
(Crocuta crocuta, Erxleben), in southern Africa. African Journal of
Ecology, 20, 271-278.

Macdonald, D. W. 1981. Resource dispersion and the social organization of the
red fox, Vulpes vulpes. In: Proceedings of the Wondwide Furbearers
Conference (Ed. by Chapman, J. A. 8 Ureley, 0.), pp. 918-949. Frostburg,
MD.

Matthews, L. H. 1939. Reproduction in the spotted hyaena, Crocuta crocuta
(Erxleben). Philosophical Transactions of the Royal Society of London B,
230, 1-78.

McComb, K., Packer, C. 8 Pusey, A. E. 1994. Roaring and numerical
assessment in contests between groups of female lions, Panthera Ieo.
Animal Behaviour, 47, 379-387.

Mech, L. D. 1970. The Wolf: Ecology and Behavior of an Endangered Species.
Minneapolis: University of Minnesota Press.

Messier, F. 8 Barrette, C. 1982. The social system of the coyote (Canis Iatrans)
in a forested habitat. Canadian Journal of Zoology. 60, 1743-1753.

Milinski, M. 8 Parker, G. A. 1991. Competition for resources. In: Behavioural
Ecology: An Evolutionary Approach (Ed. by Krebs, J. W. 8 Davies, N. 8.),
pp. 137-168. Oxford: Blackwell Scientific.

Mills, M. G. L. 1983. Mating and denning behavior of the brown hyaena Hyaena
brunnea and comparisons with other Hyaenidae. Zeitschn'ft fur
Tierpsychologie, 63, 331-342.

Mills, M. G. L. 1984. The comparative behavioural ecology of the brown hyaena

Hyaena brunnea and the spotted hyaena Crocuta crocuta in the southern
Kalahari. Koedoe, 27, 237-247.

167

Mills, M. G. L. 1987. Behavioural adaptations of brown and spotted hyaenas in
the southern Kalahari. South African Journal of Science, 83, 595-598.

Mills, M. G. L. 1990. Kalahari hyaenas: Comparative behavioral ecology of two
species. London: Unwin Hyman.

Mills, M. G. L. 1991. Conservation management of large carnivores in Africa.
Koedoe, 34, 81-90.

Mills, M. G. L. 8 Gorman, M. L. 1997. Factors affecting the density and
distribution of wild dogs in the Kruger National Park. Conservation Biology,
11, 1397-1406.

Norton-Griffiths, M. 1995. Economic incentives to develop the rangelands of the
Serengeti: Implications for wildlife conservation. ln: Serengeti II:
Dynamics, Management, and Conservation of an Ecosystem (Ed. by
Sinclair, A. R. E. 8 Arcese, P.), pp. 588-604. Chicago: University of
Chicago Press.

Ogutu, J. O. 8 Dublin, H. T. 1998. The response of lions and spotted hyaenas to
sound playbacks as a technique for estimating population size. African
Journal of EcologY, 36, 83-95.

Owens, D. 8 Owens, M. 1979a. Communal denning and clan associations in
brown hyenas (Hyaena brunnea, Thunberg) of the central Kalahari Desert.
African Journal of Ecology . 17, 35-44.

Owens, D. 8 Owens, M. 1979b. Notes on social organization and behavior in
brown hyenas (Hyaena brunnea). Journal of Mammalogy. 60, 405-408.

Owens, D. D. 8 Owens, M. J. 1984. Helping behaviour in brown hyenas. Nature,
308, 843-845.

Packer, C. 1986. The Ecology of Sociality in Felids. In: Ecological Aspects of
Social Evolution (Ed. by Rubenstein, D. I. 8 Wrangham, R. W.), pp. 429-
451. Princeton: Princeton University Press.

Packer, C., Altizer, S., Appel, M., Brown, E., Martenson, J., O'Brien, S. J.,
RoeIke-Parker, M., Hofman-Lehmann, R. 8 Lutz, H. 2000. Viruses of the
Serengeti: patterns of infection and mortality in African lions. Journal of
Animal Ecology, 68, 1161-1 178.

Packer, C. 8 Pusey, A. E. 1987. Dispersal and philopatry. In: Primate Societies

(Ed. by Smuts, B. B., Cheney, D. L. 8 Seyfarth, R. M.), pp. 250-266.
Chicago and London: University of Chicago Press.

168

Packer, C. 8 Pusey, A. E. 1997. Divided we fall: cooperation among lions.
Scientific American, May, 52-60.

Packer, C., Scheel, D. 8 Pusey, A. E. 1990. Why lions form groups: food is not
enough. American Naturalist, 136, 1-1 9.

Parker, P. G. 8 Waite, T. A. 1997. Mating systems, effective population size, and
conservation of natural populations. In: Behavioral approaches to
conservation in the wild (Ed. by Clemmons, J. R. 8 Buchholz, R.), pp. 243-
261. Cambridge: Cambridge University Press.

Perrin, M. R., D'Inzillo Carranza, I. 8 Linn, I. J. 2000. Use of space by the
spotted-necked otter in the KwaZqu-Natal Drakensberg, South Africa.
South African Journal of Wildlife Research, 30, 15-21.

Pierce, B. M., Bleich, V. C., Wehausen, J. D. 8 Bowyer, T. 1999. Migratory
patterns of mountain lions: implications for social regulation and
conservation. Journal of Mammalogy, 80, 986-992.

Powell, R. A. 1979. Mustelid spacing patterns: variations on a theme by Mustela.
Zeitschrifl fur Tierpsychologie, 50, 153-165.

Powell, R. A. 2000. Animal home ranges and territories and home range
estimators. ln: Research techniques in animal ecology: controversies and
consequences (Ed. by Boitani, L. 8 Fuller, T. K.), pp. 65-110. New York:
Columbia University Press.

Pusey, A. E. 8 Packer, C. 1987. Dispersal and philopatry. In: Primate societies
(Ed. by Smuts, B. 8., Cheney, D. L., Seyfarth, R. M., Wrangham, R. W. 8
Struhsaker, T. T.), pp. 250-266. Chicago: University of Chicago Press.

Pusey, A. E. 8 Packer, C. 1994. lnfanticide in lions. In: lnfanticide and Parental
Care (Ed. by Parrnigiani, S. 8 vom Saal, F. 8.), pp. 277-299. Chur,
Switzerland: Harwood Academic Publishers.

Rasa, O. A. E. 1987. The dwarf mongoose: a study of behavior and social
structure in relation to ecology in a small carnivore. Advances in the Study
of Behavior, 17, 121-163.

Rood, J. P. 1983. The social system of the dwarf mongoose. In: Advances in the
Study of Mammalian Behavior, pp. 454-488: Special Publication of the
American Society of Mammalogists.

Rose, 8. 1982. Lizard home ranges: methodology and functions. Journal of
Herpetology, 16, 253-269.

169

Sanderson, G. C. 1966. The study of mammal movementsu A review. Journal of
Wildlife Management, 30, 215-235.

Schaller, G. B. 1972. The Serengeti Lion. Chicago: University of Chicago Press.

Schneider, K. M. 1926. Uber Hyanenzucht. Die Peltztierzucht, 2, 1-4, 9-11, 12-
14.

Schonewald-Cox, C., Azari, R. 8 Blume, S. 1991. Scale, variable density, and
conservation planning for mammalian carnivores. Conservation BiologI’. 5,
491-495.

Seaman, D. E., Millspaugh, J. J., Kemohan, B. J., Brundige, G. C., Raedeke, K.
J. 8 Gitzen, R. A. 1999. Effects of sample size on kernel home range
estimates. Journal of Wildlife Management, 63, 739-747.

Sillero-Zubiri, C. 8 Gottelli, D. 1992. Population ecology of spotted hyaena in an
equatorial mountain forest. African Journal of Ecology. 30, 292-300.

Skinner, J. D. 8 van Aarde, R. J. 1980. The distribution and ecology of the brown
hyaena Hyaena brunnea and spotted hyaena Crocuta crocuta in the
central Namib Desert. Madoqua, 12, 231-239.

Smale, L., Frank, L. G. 8 Holekamp, K. E. 1993. Ontogeny of dominance in free-
living spotted hyaenas: Juvenile rank relations with adult females and
immigrant males. Animal Behaviour, 46, 467-477.

Smale, L., Holekamp, K. E., Weldele, M., Frank, L. G. 8 Glickman, S. E. 1995.
Competition and cooperation between littermates in the spotted hyaena,
Crocuta crocuta. Animal Behaviour, 50, 671 -682.

Smale, L., Nunes, S. 8 Holekamp, K. E. 1997. Sexually dimorphic dispersal in
mammals: Patterns, causes, and consequences. In: Advances in the
Study of Behavior (Ed. by Slater, P. J. B., Rosenblatt, J. S., Milinski, M. 8
Snowdon, C. T.), pp. 181-250: Academic Press.

Soule, M. 8 Noss, R. 1998. Rewilding and biodiversity: complementary goals for
continental conservation. Wild Earth, 8, 18-28.

Stuart, C. T., Macdonald, I. A. W. 8 Mills, M. G. L. 1985. History, current status
and conservation of large mammalian predators in Cape Province,
Republic of South Africa. Biological Conservation, 31, 7-19.

Swihart, R. K. 8 Slade, N. A. 1985. Testing for independence of observations in
animal movements. Ecology. 66, 1176-1184.

170

Szykman, M., Engh, A. L., Boydston, E. E. 8 Holekamp, K. E. In review. Rare
male aggression toward females in the spotted hyaena. Animal Behaviour.

Talbot, L. 8 Olindo, P. 1990. The Maasai Mara and Amboseli Reserves. In:
Living with Wildlife: Wildlife Resource Management with Local
Participation in Africa (Ed. by Kiss, A.), pp. 67-74. Washington, DC: The
World Bank.

Temeles, E. J. 1989. The effect of prey consumption on territorial defense by
harriers: Differential responses to neighbors versus floaters. Behavioral
Ecology and Sociobiology, 24, 239-244.

Temeles, E. J. 1990. Northern harriers on feeding territories respond more
aggressively to neighbors than to floaters. Behavioral Ecology and
Sociobiology. 26, 57-63.

Terborgh, J., Estes, J. A., Paquet, P., Ralls, K., Boyd-Heger, D., Miller, B. J. 8
Noss, R. F. 1999. The role of top carnivores in regulating terrestrial
ecosystems. In: Continental Conservation: Scientific Foundations of
Regional Reserve Networks, pp. 39-64. Washington, DC: Island Press.

Terborgh, J. 8 Soule, M. 1999. Why we need megareserves: Large-scale
networks and how to design them. Wild Earth, 9, 66-72.

Thery, M. 1992. The evolution of Ieks through female choice: Differential
clustering and space utilization in six sympatric manakins. Behavioral
Ecology and Sociobiology, 30, 227-237.

Thomson, P. C., Rose, K. 8 Kok, N. E. 1992. The behavioural ecology of dingoes
in north-westem Australia: VI. Temporary extraterritorial movements and
dispersal. Wildlife Research, 19, 585-595.

Tilson, R. T. 8 Hamilton, W. J. l. 1984. Social dominance and feeding patterns of
spotted hyaenas. Animal Behaviour, 32, 715-724.

Tilson, R. T. 8 Henschel, J. R. 1986. Spatial arrangement of spotted hyaena
groups in a desert environment, Namibia. African Journal of EcologY. 24,
173-1 80.

Trivers, R. L. 1972. Parental investment and sexual selection. In: Sexual
Selection and the Descent of Man, pp. 139-179. Chicago: Aldine.

Trivers, R. L. 1974. Parent-offspring conflict. Amen'can Zoologist, 14, 249-264.

171

 

 

Walton, G. A. 1962. The Omithodorus moubata superspecies problem in relation
to human relapsing fever epidemiology. Symposia of the Zoological
Society of London, 6, 83-156.

Waser, P. M. 8 Jones, W. T. 1983. Natal philopatry among solitary mammals.
Quarterly Review of BiologY. 58, 355-390.

Whateley, A. 8 Brooks, P. M. 1978. Numbers and movements of spotted
hyaenas in Hluhluwe Game Reserve. Lammergeyer, 26, 44-52.

White, G. C. 8 Garrott, R. A. 1990. Analysis of Wildlife Radio-Tracking Data. San
Diego, California: Academic Press, Inc.

Wikramanayake, E. D., Dinerstein, E., Robinson, J. G., Karanth, U., Rabinowitz,
A., Olson, D., Mathew, T., Hedao, P., Conner, M., Hemley, G. 8 Bolze, D.
1998. An ecology-based method for defining priorities for large mammal
conservation: The tiger as case study. Conservation Biology, 12, 865-878.

Wolff, J. O. 8 Peterson, J. A. 1998. An offspring-defense hypothesis for
territoriality in female mammals. Ethology, Ecology, and Evolution, 10,
227-239.

Woodroffe, R. 2000. Predators and people: using human densities to interpret
declines of large carnivores. Animal Conservation, 3, 165-173.

Woodroffe, R. 8 Ginsberg, J. R. 1998. Edge effects and the extinction of
populations inside protected areas. Science, 280, 2126-2128.

Woodroffe, R. 8 Ginsberg, J. R. 1999. Conserving the African wild dog Lycaon
pictus. l. Diagnosing and treating causes of decline. Oryx, 33, 132-142.

Woodroffe, R. 8 Ginsberg, J. R. 2000. Ranging behaviour and vulnerability to
extinction in carnivores. In: Behaviour and Conservation (Ed. by Gosling,
L. M. 8 Sutherland, W. J.). Cambridge: Cambridge University Press.

Worton, B. J. 1989. Kernel methods for estimating the utilization distribution in
home-range studies. Ecology, 70, 164-168.

172

IIIIIIIIIIIIIIIIIIIIIIIIIIIIII

lulllIIIIIIIIIIIIIMIIIIIIIIIII