£3 2... Y0
econ .85.:
. 4 «4.3..

4.949.. x

. a
1 s 145‘.
‘ «$.53

._ . 2‘ 4 a
Eumiwuno. hr. .x
1: .. . fl

:7
5. itfv

. :3 ._
n4”? “W

.1
1.41

.93

Me -

i :13? .1. ..

15...... :3. rvx‘xvui

.. 1.1, .. 2 .

s 2.2095. :rhtf 4
cu . ‘31.!9- v.19}

 

32X .....

,..
. 3.123
. .I . .

W...:t.a..‘.a t ‘ ‘ ‘

30 ....l.. .

V 'v 9A: t

. . ..... lurks} :11.tr.:..
. . . i... .. :X

1‘

 

'

“mars

c303)

 

LIBRARY
Michigan State

 

University

 

 

This is to certify that the
dissertation entitled

SOCIAL AND ECOLOGICAL INFLUENCES ON SURVIVAL
AND REPRODUCTION IN THE SPOTTED HYENA,
CROCUTA CROCUTA

presented by

HEATHER ELIZABETH WATTS

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Department of Zoology and
Program in Ecology,
Evolutionary Biology, and

 

 

 

 

Behavior
% Majorprfirsor’s Signature
Date

MSU is an affirmative-action. equal-opportunity employer

~ .n.—.-—.—.—.-._

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.

 

DAIEDUE

DAIEDUE

DATEDUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRCIDaIeDue.indd-p.1

 

 

SOCIAL AND ECOLOGICAL INFLUENCES ON SURVIVAL AND
REPRODUCTION IN THE SPOTTED HYENA, CROCUTA CROCUTA

By

Heather Elizabeth Watts

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Zoology
Program in Ecology, Evolutionary Biology, and Behavior

2007

ABSTRACT

SOCIAL AND ECOLOGICAL INFLUENCES ON SURVIVAL AND
REPRODUCTION IN THE SPOTTED HYENA, CROCUTA CROCUTA

By
Heather Elizabeth Watts

In this dissertation, I explore variation in survival and reproduction both at
the level of the individual and that of the population, incorporating molecular and
behavioral analyses, in order to elucidate ecological and evolutionary processes
in the spotted hyena (Crocuta crocuta). My approach was to examine variation
within a single population of hyenas in the Masai Mara National Reserve
(hereafter Mara), Kenya, for which longitudinal data are available, and to make
comparisons between this population and a second in Amboseli National Park,
Kenya, for which I collected corresponding cross-sectional data.

Spotted hyenas live in complex social groups that are structured by linear
dominance hierarchies. Contrary to the typical mammalian pattern, however,
females are socially dominant to males. I examined patterns of survival to test
the hypothesis that intense feeding competition, in conjunction with slow
development of the feeding apparatus, favored the evolution of female
dominance in spotted hyenas. As predicted by this hypothesis, weaning was a
particularly challenging life history event for young hyenas, and both social rank
and maternal aid influenced the probability of surviving after weaning.

Despite the important role that top predators play in ecosystems, relatively
little is known about the factors that influence large predator populations

themselves. Therefore, I inquired how social and ecological variables affect

survival and reproduction in spotted hyenas. First, based on longitudinal variation
in the Mara, I found that prey availability, group size, and competition with lions
all influenced birthrates, while competition with lions and rainfall influenced
juvenile survival. I then tested alternative hypotheses to explain variation in
fitness measures between the Amboseli and Mara populations. My results were
most consisted with the hypothesis that interspecific competition with lions was
the primary source of population differences.

Population dynamics can have important influences on population genetic
structure, which may in turn affect population persistence. For example, the
genetic consequences of population bottlenecks are of particular concern to
conservation biologists. While the Mara hyena population has remained large
and stable in recent history, the Amboseli population has undergone a
demographic bottleneck. Despite this bottleneck, however, the Amboseli
population did not exhibit reduced genetic diversity relative to the Mara. Both the
long generation time of hyenas and the frequency of migration may have
influenced the genetic outcome of the Amboseli bottleneck.

Finally, in light of the role that lions play in shaping reproduction and
survival in spotted hyenas, l inquired whether hyenas use avoidance to minimize
potentially costly encounters with lions. Using audio playback experiments, I
demonstrated that hyenas do not consistently use avoidance behavior with lions.
While the relative costs and benefits of an interaction did not influence response
to lion playbacks, I did find consistent individual differences in risk-taking and

vigilance tendencies with respect to lions, independent of social rank.

Copyright by
HEATHER ELIZABETH WATTS

2007

ACKNOWLEDGEMENTS

I owe a great debt of thanks to many people and institutions that
contributed to this dissertation. First and foremost, I would like to thank Dr. Kay
Holekamp, my advisor, who has been an endless source of guidance, support,
and encouragement. Dr. Holekamp provided me with the opportunity to do this
research and has taught me so much along the way. I feel very fortunate to have
been her student. I also appreciate the excellent guidance of my dissertation
committee, Drs. Tom Getty, Kim Scribner, and Laura Smale.

My research has relied heavily upon long-term data collection conducted
in the Masai Mara. These data were collected by: Nancy Berry, Chantal
Beaudoin, Erin Boydston, Susan Cooper, Stephanie Dloniak, Martin Durham,
Anne Engh, Josh Friedman, Paula Garrett, lsla Graham, Tyson Harty, Cathy
Katona, Karen Kapheim, Kay Holekamp, Joe Kolowski, Keith Nelson, Karen Nutt,
Gabe Ording, Laura Smale. Jenn Smith, Micaela Szykman, Jaime Tanner, Kevin
Theis, Russ Van Horn, Sofia Wahaj, Kim Weibel, and Brad White. I would also
like to thank Brent Pav and Wes Dowd for their assistance with data collection in
Amboseli, and Lisa Blankenship and Stephanie Dawes for analyzing videos of
the playback experiments.

I thank Dr. Scribner for opening his lab to me for the genetic analyses.
Russ Van Horn provided valuable assistance with this work and assigned many
genotypes. I am also grateful to Scot Libants and Kristi Filcek for technical

assistance in the Scribner lab.

Many people in Michigan and Kenya provided me with friendship and
assistance, only some of whom are named here. I have been very fortunate to
have a wonderful group of labmates (past and present) in the Holekamp lab and I
thank them all. They have been great companions in the field and in Michigan, as
well as insightful colleagues. In particular, Joe Kolowski and Jaime Tanner have
become very dear friends, and they helped me with this research in countless
ways. Pat Bills has provided technical assistance in the Holekamp lab, and he
and Terri McElhinny have been great friends to me in Michigan. In Kenya, I
thank Wilfred Samoire for all of his hard work to keep camp going. He helped
make my time in Amboseli great. The staff and directors of the Amboseli Baboon
Research Project and the Amboseli Elephant Research Project provided both
assistance and a social outlet in the field. I would like to thank Dr. Susan Alberts
in particular, for providing logistical assistance in Amboseli and for being a
wonderful mentor over the years. I can’t imagine my time in Amboseli without
Bart and Mary Lever-Morrison and Sue Heath; they kept me and our vehicles
going. I would also like to thank John Keshe and Moses Sairowa for taking care
of me when l was in the Mara.

I thank my husband, Wes Dowd, for his love, support, and patience. This
research would not have been possible without his assistance. He worked
tirelessly to help establish the Amboseli field site and provided valuable
comments on the dissertation. My parents, Bob and Debra Watts, and my sister
Lindsay, have encouraged and supported me throughout this process. I thank

them for that.

vi

The Office of the President of Kenya and the Kenya Wildlife Service
provided permission to conduct this research. I also thank the Narok Country
Council and the Wardens and staff of Amboseli National Park and the Masai
Mara National Reserve for their cooperation and assistance.

Finally, I received financial support during my graduate studies from the
National Science Foundation Graduate Research Fellowship Program and
Michigan State University. In particular, I was supported by the Graduate School;
the College of Natural Science; the Program in Ecology, Evolutionary Biology,
and Behavior; and the Department of Zoology. My research has been part of the
Michigan State University Hyena Project, which was funded by grants
IBNO113170, IBN0343381 and IOBOG18022 from the National Science

Foundafion.

vii

TABLE OF CONTENTS

LIST OF TABLES ................................................................................... x

LIST OF FIGURES ................................................................................. xi

GENERAL INTRODUCTION ..................................................................... 1
Overview of Chapters ..................................................................... 4

CHAPTER 1

POST-WEANING MATERNAL CARE AND THE EVOLUTION OF FEMALE

DOMINANCE IN THE SPOTTED HYENA ................................................... 9
Introduction .................................................................................. 9
Methods .................................................................................... 16
Results ..................................................................................... 19
Discussion ................................................................................. 24

CHAPTER 2

ECOLOGICAL DETERMINANTS OF SURVIVAL AND REPRODUCTION IN THE

SPOTTED HYENA ............................................................................... 30
Introduction ................................................................................ 30
Methods ............................................... - ..................................... 32
Results ..................................................................................... 39
Discussion ................................................................................. 47

CHAPTER 3

GENETIC DIVERSITY MAINTAINED IN A POPULATION OF LARGE

CARNIVORES DESPITE A RECENT BOTTLENECK .................................. 53
Introduction ................................................................................ 53
Methods ................................................................ - .................... 57
Results ..................................................................................... 65
Discussion ................................................................................. 70

CHAPTER 4

INTERSPECIFIC COMPETITION INFLUENCES REPRODUCTION IN

SPOTTED HYENAS .................................... . ........................................ 76
Introduction ................................................................................ 76
Methods .................................................................................... 78
Results ..................................................................................... 91
Discussion ................................................................................. 99

CHAPTER 5

RESPONSES OF SPOTTED HYENAS TO LIONS REFLECT CONTEXT-

SPECIFIC INDIVIDUAL DIFFERENCES IN BEHAVIOR ............................. 108
Introduction .............................................................................. 108

viii

Methods .................................................................................. 1 12

Results ................................................................................... 124
Discussion .............................................................................. 133
APPENDICES ................................................................................... 138
APPENDIX A ........................................................................... 139
APPENDIX B ........................................................................... 142
LITERATURE CITED .......................................................................... 146

LIST OF TABLES

Table 2.1. Life table for females born in the study, including survival (P,) and
fertility (F,-) estimates for the demographic matrix model ............................... 42

Table 3.1. Rates of allelic dropout (ADO) and false alleles (FA) for amplification
of fecal DNA. To estimate error rates, a reference genotype was determined
using either a multiple-tubes approach or DNA extracted from a blood sample
from the same individual. Sample sizes for the multiple-tubes approach are
presented as the number of individuals. Matched blood samples were available
for 4 individuals. N/A indicates that the ADO could not be calculated because all
individuals sampled were homozygous for this locus ................................... 66

Table 3.2. Comparison of genetic diversity between Amboseli and Mara
populations. Values are presented as mean :I: 1 SE for the number of alleles (A),
observed and expected heterozygosities (HO and HE). individual heterozygosity

(HI), internal relatedness (IR), and within-clan relatedness (R). Clan sizes are
indicated in parentheses ........................................................................ 68

Table 4.1. Comparison of the two study areas: Amboseli National Park and the
Masai Mara National Reserve ................................................................. 81

Table 4.2. Comparison of spotted hyena study clans in Amboseli National Park
and the Masai Mara National Reserve ...................................................... 83

Table 5.1. Comparison of responses to lion and control playbacks ................ 127

Table A1 Body mass and weaning age of carnivore species. Data are taken
from Gittleman (1986) unless otherwise noted .......................................... 140

Table B.1. PCR reaction conditions for use of Crocuta microsatellites with fecal
DNA. All units are pl ............................................................................ 143

Table B.2. PCR cycles for use of Crocuta microsatellites with fecal DNA........144

Table B.3. Characteristics of microsatellite loci used to type spotted hyenas from
Amboseli National Park and the Masai Mara Game Reserve. The locus-specific
number of alleles, observed frequency of heterozygotes (Ho), and the expected

frequency of heterozygotes (HE) are presented for each population .............. 145

LIST OF FIGURES

Figure 1.1. Mean age at weaning and adult body mass for eight families of
Carnivora. Separate values for the bone-cracking hyaenids and sole non-bone-
cracking hyaenid (i.e., the aardwolf) are also indicated. Species included in
calculations of family means are presented in Appendix A. Data come from
Gittleman (1986) and more recent references (see Appendix A) ..................... 12

Figure 1.2. Monthly mortality rate for females during the months immediately
after den independence (n = 85), weaning (n = 60), and reproductive maturity (n
= 59). The x-axis indicates the number of months following each event. For any
individual, time zero is the date on which that individual left the den permanently,
weaned, or reached 24 months of age (i.e. sexual maturity), respectively. Fates
after den independence could only be followed for 4 months before juveniles
began to wean. The dashed line indicates the mean monthly mortality rate for the
first 12 months of life, when most juveniles are still nursing ........................... 20

Figure 1.3. Survivorship curves for daughters born (a) to high-, mid-, and low-
ranking mothers, (b) to nulliparous and parous mothers, and (c) in singleton, twin,
and triplet litters. Cases in which the proportion of individuals surviving does not
reach zero are due to individuals still alive at the end of the study (i. e., right-
censored data points) ............................................................................ 22

Figure 1.4. Proportion of females surviving after weaning (a) based on social
rank (n=35 high-ranking, n =25 low ranking) and (b) when mother was present (n
= 53) or absent (n = 6). Time zero is assigned for each individual as its date of
weaning ............................................................................................. 23

Figure 2.1. (a) Mean monthly (A) rainfall (mm), (I) lion-hyena interaction rate,
and (0) per capita prey abundance for each year of the study. Means for 1988
are not included since the study did not begin until July of that year. Rate of lion-
hyena interactions is a measure of competition between 2the two species. Per
capita prey abundance IS prey density (prey animals/km2 ) per adult female in the
clan. (b) Estimated human population adjacent to the study area, based on the
number of huts. A polynomial equation was fitted to census data ([3) from
Lamprey and Reid (2004) to generate estimates (0) for all years. Horizontal bar
indicates the period of the current study ................................................... 35

Figure 2.2. Mean (:1) clan size, (0) number of adult females, (A) number of
resident immigrant males, and (0) number of juveniles present in the study clan
for each year between 1988 and 2003. Monthly means 1 SE are shown. Arrows
indicate two clan fission events. Horizontal line indicates canine distemper virus
epizootic ............................................................................................. 40

xi

Figure 2.3. Percent of deaths caused by the major mortality sources for three
spotted hyena populations. Sample sizes for each population are indicated.
Deaths caused by disease are included as illness. Most deaths caused by
hyenas were either siblicide or infanticide. Kruuk (1972) lumped deaths caused
by starvation and illness, together they accounted for 21% of deaths. Asterisks
indicate that those deaths are divided equally between the two mortality sources
here for representation only. Age categories are indicated for the Masai Mara
population only. Individuals less than 2 years of age at death were categorized
as juveniles (hatched bars), while those older than 2 years were categorized as
adults (open bars). Note that a very small percentage of adults died from
starvation, illness, or hyenas in the Masai Mara. Almost all of the deaths due to
illness in the Kalahari were likely due to rabies (Mills 1990) ........................... 43

Figure 2.4. Annual (a) birthrates, (b) recruitment rate and (c) adult mortality rate
from 1989 to 2003. The recruitment rate is the proportion of individuals born in a
given year that survive until 2 years of age. Rates are not given for 1988 since
the study did not begin until July of that year. Recruitment data for individuals
born in 2002 and 2003 were not yet available. Arrows indicate two clan fission
events. Horizontal line indicates canine distemper virus epizootic ................... 45

Figure 3.1. The territories of the study clans in (a) the Masai Mara National
Reserve, and (b) Amboseli National Park. Areas within the park boundaries are
lightly shaded, while clan territories are darkly shaded ................................. 58

Figure 3.2. Allele frequency distribution for Amboseli (solid bars) and Mara
(hatched bars) populations. Distributions for both populations are “L-shaped” as
expected in the absence of a genetic bottleneck ......................................... 69

Figure 3.3. Pairwise R-values among natal animals within the same clan and
between different clans for the Amboseli and Mara populations. Sample sizes

indicate number of R—values. Values are presented as X 1 SE. In both
populations differences within and between clans were significant (P < 0.05)....71

Figure 4.1. The territories of the study clans in (a) the Masai Mara National
Reserve, and (b) Amboseli National Park. Areas within the park boundaries are
lightly shaded, while clan territories are darkly shaded. The Talek clan territory
covered 61 km2, while the Airstrip and OI Tukai territories covered 28 km2 and 26
km2, respectively. The methods used to delineate territories are described by
Boydston et al. (2001) and Kolowski (2007) ............................................... 80

Figure 4.2. Composition of litters born to spotted hyenas in Amboseli National
Park (n = 53 litters) and the Masai Mara National Reserve (n = 55 litters).
Numbers of litters are indicated above each bar. The frequencies of singleton
and twin litters were not significantly different between populations, but we were
unable to test for differences in the frequency of triplet litters ......................... 92

xii

Figure 4.3. Mean 1 SE (a) age at weaning for cubs and (b) inter-birth interval for
spotted hyena populations in Amboseli National Park and the Masai Mara
National Reserve. Sample sizes are indicated above each bar, and represent the
number of cubs (a) and females (b). Asterisks represent significant differences (P
< 0.05) ............................................................................................... 94

Figure 4.4. Mean monthly prey density within the territories of spotted hyena
clans in Amboseli National Park and the Masai Mara National Reserve. Values
represent averages over multiple years ..................................................... 95

Figure 4.5. Mean 1 SE (a) number of spotted hyenas competing at kills and (b)
aggression rates at kills in Amboseli National Park and the Masai Mara National
Reserve. Sample sizes are indicated above each bar, and represent the number
of kills. Asterisks represent significant differences (P < 0.05) ......................... 97

Figure 4.6. Estimated biomass of prey consumed as fresh kills by spotted hyenas
in Amboseli National Park (n = 119 kills) and the Masai Mara National Reserve (n
= 798 kills). Data for the Masai Mara are taken from Cooper et al. (1999).
Species-specific biomass estimates are from Cooper et al. (1999) and Kruuk
(1972) ................................................................................................ 98

Figure 4.7. The frequency with which spotted hyenas were successfully able to
scavenge fresh carcasses from lions and at which spotted hyenas lost their kills
to lions in Amboseli National Park (n = 27 scavenging opportunities, n = 142
hyena kills) and the Masai Mara National Reserve (n = 29 scavenging
opportunities, n = 323 hyena kills). Numbers of carcasses are indicated above
bars. The asterisk represents a significant difference (P < 0.05) ................... 100

Figure 5.1. Local lion density in the two study clan territories, based on utilization
distribution. Darker cells indicate areas of higher lion density. Amboseli National
Park boundary is indicated by the thick line. The two clan territory boundaries are
indicated by thin lines. Locations of each lion playback experiment (N = 21) are
shown as open circles ......................................................................... 120

Figure 5.2. Duration of response to lion (N = 21, filled bars) and control (N = 14,
open bars) playbacks .......................................................................... 125

Figure 5.3. (a) The proportion of time hyenas oriented towards the concealed
speakers, and (b) the rate at which they scanned the surrounding area, before
(filled bars) and after (open bars) sound onset for paired lion and control
playbacks (N = 14 hyenas). Mean values i 1 SE are presented ................... 126

xiii

Figure 5.4. Movement in response to lion (N = 21, filled bars) and control (N =14,
open bars) playbacks. Categories of movement are: approach the sound source,
move away (avoid) from the sound source either immediately or after a brief
initial approach, and no movement (none) ................................................ 129

Figure 5.5. Relationship between responses of 17 hyenas to experimental
playback of lion roars and their behavior when lions were present with them
during daily observation sessions. (a) Proportion of time spent oriented in
response to playback as a function of the proportion of time spent vigilant in the
presence of lions. (b) Movement in response to playback as a function of risk-
taking during interactions with actual lions, measured as the distance an
individual maintained to the nearest lion .................................................. 131

Figure 5.6. Vigilance and risk-taking tendencies during naturally-occurring
interactions with lions are correlated among individual hyenas. Vigilance was
quantified as the proportion of time spent vigilant when lions were present at
daily observation sessions. Risk-taking was quantified as the mean distance that
individuals maintained to lions during these interactions. Values are shown for
each of 17 hyenas. Symbol shape indicates individuals of high (circle), mid
(square), or low (triangle) social rank ...................................................... 132

xiv

GENERAL INTRODUCTION

Understanding patterns of variation in survival and reproduction is
fundamental to answering many important questions in the fields of ecology,
ethology, and evolutionary biology. Examining variation in these fitness
measures can shed light on the dynamics of populations over ecological time
scales, as well as selection pressures shaping phenotypes over evolutionary time
scales. In this dissertation, I explore variation in survival and reproduction both at
the level of the individual and that of the population, incorporating molecular and
behavioral analyses, in order to elucidate ecological and evolutionary processes
in the spotted hyena (Crocuta crocuta).

Spotted hyenas are gregarious carnivores that live in multi-male, multi-
female groups, called clans, that range in size from 8 to 90 individuals (Kruuk
1972; Mills 1990). Whereas the societies of most gregarious carnivores are
composed of relatively small family groups, the societies of spotted hyenas often
contain many unrelated individuals (Van Horn et al. 2004). Thus, spotted hyenas
exhibit a social system quite distinct from those of other carnivores. In fact, the
large, complex clans of spotted hyenas have far more in common with many
primate societies than with the societies of other carnivores (Holekamp et al.
2007)

Clans are structured by strict linear dominance hierarchies in which an
individual’s position, or rank, in the hierarchy is determined by the social rank of
its mother (Frank 1986b; Holekamp & Smale 1991 ). Early in life youngsters of

both sexes ‘inherit’ the social rank of their mother. However, while females

remain in their natal group throughout their lives, virtually all males disperse after
puberty to join a new clan (Smale et al. 1997; East & Hofer 2001; Boydston et al.
2005; Honer et al. 2007). When a male immigrates into a new group, he enters
as the lowest-ranking hyena in the dominance hierarchy; he behaves
submissively to all hyenas he encounters in the new territory (Smale et al. 1993).
This creates a society in which adult females and their cubs are dominant to all
adult male immigrants, and in this respect spotted hyenas differ dramatically from
other carnivores, and from most other mammals as well.

Spotted hyenas are proficient hunters. They feed primarily on ungulates
that they kill themselves, although they also scavenge opportunistically (Kruuk
1972; Cooper et al. 1999). Ungulate carcasses are rich, but ephemeral food
sources, and many hyenas may converge to feed on a single carcass. This leads
to intense feeding competition, in which social rank determines priority of access
to food (Kruuk 1972; Tilson & Hamilton 1984; Frank 1986b). Although feeding
competition within clans can be intense, clan members must cooperate to defend
their territory and the food resources within it from neighboring conspecifics, as
well as to defend carcasses from lions. Feeding competition between lions and
hyenas can be fierce, and lions are a major source of hyena mortality (Kruuk
1972)

Spotted hyenas occupy a wide diversity of habitats including savannas,
woodlands, deserts, and montane forests (e.g., Kruuk 1972; Mills 1990; Sillero-
Zubiri & Gottelli 1992). Interestingly, there is considerable variation among such

populations in both behavioral traits, such as group size and territoriality, and life

history traits, including the age at which cubs are weaned (cf. Kruuk 1972; Mills
1990; Hofer & East 1993a; Holekamp et al. 1996). This variability makes the
spotted hyena an excellent system in which to examine how environmental
factors influence various aspects of behavior and demography. Further, spotted
hyenas are the most abundant large carnivores on the African continent (Cardillo
et al. 2004). Given their abundance and near ubiquity across diverse habitats,
spotted hyenas are likely to be important to a large number of African
ecosystems, and of general conservation interest. I hope that findings from this
dissertation, in particular those focusing on the dynamics of hyena populations,
will be of use in the management of this species.

Unfortunately, the task of examining variation in survival and reproduction
in a long-lived mammal such as the spotted hyena is a difficult one. Most notably,
it requires monitoring many known individuals for extended periods of time and
over scales necessary to capture the variation of interest. One approach is to
examine longitudinal variation in a single population, usually over multiple
generations. This approach is advantageous because it provides complete life
history information for individuals. An alternative approach is to examine variation
among different populations, though usually over shorter time intervals. An
advantage of the latter, cross-sectional comparative approach is that it can
capture greater environmental variation than is typically experienced by any
single population studied longitudinally. Here, I use both of these complementary

approaches with spotted hyenas.

In the Masai Mara National Reserve, Kenya, spotted hyenas have been
monitored since the late 1970’s (e.g. Frank 1986a), and the current study has
been in progress since 1988 (eg. Holekamp et al. 1993). This long-term study
has focused primarily on a single large clan of hyenas, including data on over
400 individuals. These data have been supplemented by observations of other
nearby clans (e.g., Van Horn et al. 2004; Kolowski et al. 2007). In addition to
examining data generated from this longitudinal study, I also collected
corresponding data on a second population of hyenas in Amboseli National Park,
Kenya. In Amboseli, l monitored two clans of hyenas for two years from 2003 to
2005. In addition to collecting detailed behavioral and life history data on
individual hyenas, both studies also monitored relevant ecological measures (eg.
prey availability and rainfall), and involved collection of genetic samples from
known individuals. Using this remarkable dataset I address a diversity of

questions pertaining to variation in fitness in this unique species.

Overview of the Chapters
Spotted hyenas, along with a few species of primates, are rare examples
of mammalian species in which females are socially dominant to males. Despite
great interest in female dominance, the evolutionary forces that generated this
trait have yet to be satisfactorily explained in hyenas or any other species. In
Chapter 1, I examine patterns of survival to test the hypothesis that the intense
feeding competition characteristic of spotted hyenas, in conjunction with slow

development of the feeding apparatus, favored the evolution of female

dominance in this species. As predicted by this hypothesis, I found that weaning
was a particularly challenging life history event for young hyenas, and both social
rank and maternal aid influenced the probability of surviving after weaning.
Whereas these results suggest that constrained juvenile development led to an
extreme form of maternal behavior (i.e. female dominance) in spotted hyenas, I
also discuss how constrained development may shape maternal behavior in
other mammals.

While Chapter 1 examines evolutionary forces shaping social
organization, Chapter 2 investigates how variation in the social and physical
environment affects survival and reproduction in spotted hyenas. Despite the
important role that top predators play in ecosystems (e.g., Estes et al. 1998;
Crooks & Soulé 1999; Wilmers & Getz 2005), relatively little is known about the
factors that influence large predator populations themselves. Utilizing longitudinal
data from the Masai Mara population, I first document substantial variation in
annual survival rates and birthrates. I then test alternative hypotheses to explain
this variation. Prey abundance, group size, and competition with lions all
influenced birthrates, while competition with lions and rainfall influenced juvenile
survival. Further, disease appears to have played a relatively minor role in the
dynamics of this population. This is one of very few studies to examine
simultaneously the roles of these various factors in the dynamics of a large
carnivore population. In contrast to the traditional view that large carnivore
populations are limited primarily by the availability of prey species, my findings

highlight the importance of interspecific competition between carnivores in

shaping the dynamics of these populations. Furthermore, my results suggest that
a key benefit of sociality in spotted hyenas may be the ability to cope with
competition from other carnivores. This contrasts with factors such as group
hunting and protection against infanticide, which have more commonly been
invoked to explain the evolution of sociality in large carnivores.

Population dynamics can have important influences on population genetic
structure, which may in turn affect population persistence. As habitat loss and
fragmentation continue to increase worldwide, the viability of small populations is
of particular concern (Lande 1988; Frankham 2005). From a genetic perspective,
small populations are expected to experience inbreeding and loss of genetic
diversity, which may, in turn, lead to reduced fitness and limited evolutionary
potential (Lynch 1996; Lacy 1997; Keller & Waller 2002; Reed & Frankham
2003). Indeed, even if populations that were once small are able to recover
demographically, the genetic legacies of past bottlenecks may still threaten their
persistence. However, studies examining whether demographic bottlenecks
result in the loss of genetic diversity in natural populations have produced mixed
results (e.g., Hoelzel et al. 1993; Bouzat et al. 1998; Hailer et al. 2006; Kaeuffer
et al. 2007). In Chapter 3, I compare population genetic structure between the
Amboseli and the Masai Mara populations, which have experienced quite
different recent population histories. While the Masai Mara population has
remained large and stable, the Amboseli population has recently undergone a
demographic bottleneck. Despite this bottleneck, however, the Amboseli

population did not exhibit reduced genetic diversity relative to the Masai Mara

population. Further, patterns of relatedness within and between clans were
similar in both populations. I conclude this chapter by considering factors such as
generation time and migration (i.e., dispersal), which may have influenced the
genetic outcome of the Amboseli bottleneck. While the results of this chapter
have important implications for the management of spotted hyena populations,
these results also inform my interpretation of results from Chapter 4. Specifically,
the finding that genetic variation was similar between populations eliminates
differences in genetic diversity as an explanation for population differences
documented in Chapter 4.

I begin Chapter 4 by describing variation in fitness measures between the
Amboseli and Masai Mara hyena populations. Specifically, I found that Amboseli
females reproduced at higher rates than females in the Masai Mara. Shorter
inter-birth intervals in Amboseli were due to shorter lactation periods in that
population relative to the Masai Mara population. Next, I test alternative
hypotheses to explain the observed difference in fitness between populations.
These hypotheses focus on prey availability, intraspecific competition both within
and between clans, and interspecific competition with lions. I did not find strong
support for either prey availability or intraspecific competition as explanations for
the observed fitness differences. My results were most consistent with the
hypothesis that differences are due to interspecific competition with lions,
particularly competition over food.

The findings of Chapters 2 and 4 indicate that competition with lions plays

an important role in shaping reproduction and survival in spotted hyenas.

Therefore, in Chapter 5, I inquire whether hyenas use avoidance to minimize
potentially costly encounters with lions, and whether they adjust their behavioral
responses to lions based on the potential costs and benefits of an interaction. I
also examine the role of behavioral syndromes in the responses of hyenas to
lions. Using a series of playback experiments, I demonstrate that hyenas do not
consistently use avoidance behavior with lions. While the relative costs and
benefits of an interaction did not appear to influence response to lion playbacks, I
did find consistent individual differences in risk-taking and vigilance tendencies
with respect to lions, independent of social rank.

Throughout the remainder of this dissertation I use the term “we” instead
of “I”. This reflects the truly collaborative nature of the work presented here, as

well as the fact that all chapters were prepared in manuscript form.

CHAPTER 1

POST-WEANING MATERNAL CARE AND THE EVOLUTION OF FEMALE
DOMINANCE IN THE SPOTTED HYENA

INTRODUCTION

Spotted hyena (Crocuta crocuta) societies are remarkable among
mammals because they are characterized by female social dominance. Along
with a few species of primates (e.g. some lemurs), they are rare exceptions to
the typical mammalian pattern of male dominance. In comparison to adult males,
adult female Crocuta are larger (Matthews 1939; Kruuk 1972; Mills 1990), more
aggressive (Szykman et al. 2003), and socially dominant (Kruuk 1972; Tilson &
Hamilton 1984; Frank 1986b). Further, sex-reversed dominance in Crocuta is
part of a larger syndrome of female “masculinization”, in which females also
exhibit highly virilized external genitalia. Females have an elongated clitoris
which forms a fully erectile pseudopenis (Matthews 1939). It remains unknown
whether female genital masculinization is an adaptive trait in Crocuta (e.g. East
et al. 1993; Muller & Wrangham 2002), or merely a non-adaptive by-product of
selection for increased aggression (either female or neonatal) in this species
(eg. Gould 1981; East et al. 1993; Frank 1997). In either case, elucidating the
selective pressures leading to female social dominance in Crocuta represents a
key step towards understanding the entire syndrome of female masculinization.

Spotted hyenas are gregarious carnivores that live in large social groups,
called clans, which are structured by linear dominance hierarchies. Crocuta feed
primarily on fresh ungulate carcasses, which are rich, but highly ephemeral, food

sources. Feeding priority is determined by social rank within the clan (Tilson &

Hamilton 1984; Frank 1986b); however, a single carcass is often too big to be
monopolized by one individual. As multiple hyenas feed simultaneously, feeding
competition becomes intense, and an entire carcass may be consumed in
minutes (Kruuk 1972; Frank 1986b). Therefore, the ability to feed rapidly is also a
critical determinant of the quantity and quality of food that an individual can
consume.

Frank (1986b) hypothesized that this highly competitive feeding
environment typical of Crocuta clans selected for large, aggressive females who
were better able to secure food resources for themselves and their dependent
young. This hypothesis, hereafter called the “Feeding Competition” hypothesis,
predicts that feeding competition in Crocuta should be more intense than in
species that do not exhibit female dominance. However, it is not clear that this is
so; other gregarious carnivore species such as wolves (Canis lupus) and wild
dogs (Lycaon pictus) experience similarly intense feeding competition. The
maximum food consumption rate of individual Crocuta, 1250 g/min (Kruuk 1972),
is comparable to those observed in these other carnivores: 1100 g/min for wolves
(Wilmers & Stahler 2002), and 1340 g/min in wild dogs (S. Creel, personal
communication). Further, feeding group size is surprisingly similar among
hyenas, wolves and wild dogs, which despite much larger variation in social
group size, typically feed in group sizes of 8, 9-10, and 10 animals, respectively
(Holekamp et al. 1997; Schmidt & Mech 1997; Creel & Creel 2002; Smith et al. In
review). All three species feed on medium- to large-sized ungulates. Despite the

fact that all three of these large carnivores experience intense feeding

10

competition, female dominance has evolved only in Crocuta, suggesting that
feeding competition alone may not be sufficient to explain the evolution of female
dominance in this species.

Unlike wolves and wild dogs, adult Crocuta have the ability to crack open
bones using powerful skulls and specialized teeth, traits they share with their
closest extant relatives (Koepfli et al. 2006), brown hyenas (Parahyaena
brunnea) and striped hyenas (Hyaena hyaena), as well as with their immediate
ancestors, which were specialized for carrion-feeding and scavenging (Werdelin
& Solounias 1991; Lewis & Werdelin 2000). The most notable cranio—dental
adaptations of the bone-cracking hyenas are robust pre-molars with specialized
enamel, a vaulted forehead, a large sagittal crest, and massive zygomatic
arches; these skull features provide attachment sites for powerful jaw muscles
(Werdelin 1989; Werdelin & Solounias 1991; Joeckel 1998). The dual ability to
hunt and scavenge enables adult Crocuta to exploit an extraordinarily wide array
of food resources. However, the ontogenetic development of this complex and
robust feeding apparatus appears to be a slow and prolonged process that
imposes severe limitations on the feeding ability of juveniles.

All of the bone-cracking hyenas (Crocuta crocuta, Hyaena hyaena, and
Parahyaena brunnea) wean their young at particularly late ages for their body
sizes, relative to other carnivores, including the only extant member of the family
Hyaenidae (the aardwolf, Proteles cn'stata) that lacks specialized adaptations for
bone-cracking (Figure 1.1). Crocuta cubs typically nurse until 12 to 16 months of

age (Kruuk 1972; Hofer & East 1995; Holekamp et al. 1996), and reach sexual

11

—L .3
N -h

.1
0

Age at weaning (mo)

 

0 J 1 1 1 L 1 1 J 1 1 1

0 40 80 120 160 200
Adult body mass (kg)

 

DOODIOCIO

Canidae
Ursidae
Procyonidae
Mustelidae
VIverridae
Hyaenidae
Felidae
Herpestidae

Bone-cracking
hyenas
Aardwolf

Figure 1.1. Mean age at weaning and adult body mass for eight families of
Carnivora. Separate values for the bone-cracking hyaenids and sole non-bone-
cracking hyaenid (i.e., the aardwolf) are also indicated. Species included in
calculations of family means are presented in Appendix A. Data come from
Gittleman (1986) and more recent references (see Appendix A).

12

maturity at 24 months of age. However, they do not reach adult feeding
competence until after 30 months of age (Binder 8 Van Valkenburgh 2000; J.
Tanner & K. Holekamp unpublished data). Consequently, the period between
weaning and the attainment of adult feeding capabilities is likely to be extremely
difficult for young hyenas attempting to feed in groups containing older hyenas
able to feed more efficiently.

We suggest that constraints imposed by development of the massive and
complex feeding apparatus required for bone-cracking, in conjunction with
intense feeding competition, better explain the evolution of female dominance in
Crocuta than feeding competition alone. Hereafter we refer to this hypothesis as
the “Competition & Constraint hypothesis”. By constraint we mean “a mechanism
or process that limits the evolutionary response of a character or set of
characters to external selection acting during a focal life stage” (Schwenk &
Wagner 2004). Although intense feeding competition should select for more rapid
development of the feeding apparatus, the Competition & Constraint hypothesis
proposes that features of the Crocuta skull may be limited in their response to
such selection during the period of skull development, due to the prolonged
ontogenetic processes required to produce a skull and musculature capable of
generating the large bite forces necessary for cracking bone. To compensate for
this developmental constraint in their cubs, we suggest that selection may have
favored large, aggressive female Crocuta who are better able to displace
conspecifics from food resources, and who are thereby able to secure access to

food for their disadvantaged, immature offspring.

13

The Competition & Constraint hypothesis leads to a number of testable
predictions regarding (1) life history patterns within Crocuta, and (2) patterns of
skull development and ontogenetic change in feeding performance among bone-
cracking hyenas and other carnivores. Here, we test three life history predictions
that follow from our hypothesis, while results regarding skull development and
feeding performance are presented elsewhere.

Whereas the Feeding Competition hypothesis focuses on the benefits of
female dominance accruing to lactating females and their nursing young, the
Competition & Constraint hypothesis predicts that the benefits of female
dominance extend beyond weaning. Specifically, if constrained development of
the feeding apparatus has lead to an extended period of juvenile dependence on
the mother that lasts until skull development is finally complete, then young
hyenas feeding in groups should be severely handicapped relative to older
individuals, and weaning should therefore be a particularly challenging life history
event forjuvenile hyenas. The Competition & Constraint hypothesis also predicts
that weaning should be particularly challenging in Crocuta relative to other
carnivores, but comparative data with which to test this prediction are quite
limited.

Here, we examine mortality rates in Crocuta after weaning and compare
them with mortality during the lactation interval, and during the periods following
two other life history events that are potentially dangerous: den independence
(White 2005) and first parturition (Frank et al. 1995b). Crocuta cubs reside in

dens for the first 8 to 12 months of life (Hofer & East 1993a; Boydston et al.

14

 

2005). Permanent departure from the den (den independence) may be a
challenging life history event as cubs are exposed at this time to unfamiliar
environments and new dangers. First parturition may be a dangerous event for
young females because of the unusual female genital morphology of this
species. In female Crocuta, the birth canal is particularly long and passes
through the pseudopenis, which must tear during first parturition to allow passage
of the fetus. Consequently, dystocia (difficult parturition) may be both common
and fatal to primiparous females (Frank & Glickman 1994). Frank et al. (1995b)
estimate that over 8% of primiparous females die during first parturition as a
result of these complications.

Both the Feeding Competition and the Competition & Constraint
hypotheses predict that females who are better able to aggressively displace
conspecifics from food resources will be more successful at raising young. As
social rank in this species is a measure of an individual’s ability to displace
conspecifics from food resources (Frank 1986b; Henschel & Skinner 1987), here
we evaluate whether social rank influences offspring survival as predicted.
Although White (2005) found no effect of maternal rank on cub survival in
Crocuta studied for four years, analysis of long-term data may yield a different
result. Furthermore, according to the Competition & Constraint hypothesis, but
not the Feeding Competition hypothesis, the effect of social rank on survival
should be particularly evident after weaning. Therefore, we compare survivorship

after weaning between the offspring of high- and low-ranking females.

15

 

Finally, not only do female spotted hyenas routinely assist their young
offspring in securing access to food resources (Frank 1986b; Holekamp & Smale
1990; Engh et al. 2000), but they are also observed to aid weaned offspring
(Kruuk 1972). If the ability of mothers to aid their young beyond weaning is
important for cub survival in the spotted hyena, as suggested by the Competition
& Constraint hypothesis, then maternal presence should increase the probability
of survival after weaning. This hypothesis also predicts that maternal support
beyond weaning should be more important in spotted hyenas than in other
carnivores, but unfortunately the comparative data necessary to test this
prediction are currently unavailable.

To date, there are no satisfactory functional explanations for the evolution
of female dominance in any mammalian species. Utilizing a comprehensive,
long-term data set on known individuals, we test the life history predictions of the
Competition & Constraint hypothesis as an explanation for the evolution of
female dominance in Crocuta. These data allow us to begin to elucidate the suite

of selective pressures leading to this unusual trait.

METHODS
We tested the Competition & Constraint hypothesis using data collected
between July 1st 1988 and December 31512003 from members of one large
Crocuta clan in the Masai Mara National Reserve, Kenya. Individual hyenas were
identified by unique spot patterns and sex was determined based on penile

morphology (Frank et al. 1990). Hyenas were observed daily between 0530 —

16

 

0900 hours and between 1700 — 2000 hours. Agonistic interactions, recorded
using all-occurrence sampling (Altmann 1974), were used to determine social
ranks of adult females (see Engh et al. 2000). Adult females were assigned both
a relative rank and rank category, depending on the analysis performed. A
female’s relative rank was the proportion of adult females over which she was
dominant, with the highest-ranking female having a rank of 1 and the lowest a
rank of 0. When sample size was sufficient, three rank categories (high, mid, and
low) were designated by dividing the hierarchy into equal thirds, but when sample
size was smaller, rank was categorized as either high or low. Maternal rank was

assigned as the rank held by a cub’s mother when that cub was born.

Determination of births, death, and other life events

Most adult females in the study clan were fitted with radio collars such that
they could be relocated daily for us to monitor their reproductive status and
observe their young cubs. Natal and communal den sites, where Crocuta cubs
reside, were visited regularly to monitor births and the development of cubs.
When cubs were first observed above ground, their ages were estimated to
within :I:7 days (as in Holekamp et al. 1996). In this population most cubs become
independent of the communal den at 8 to 9 months of age; a cub was considered
independent of the den when it was found more than 200m from the den on at
least 4 consecutive occasions (Boydston et al. 2005). Juveniles typically wean at
12 to 14 months of age in this population, but weaning age ranges from 7 to 24

months (Holekamp et al. 1996). Weaning age was determined based on weaning

17

 

conflicts and cessation of nursing as described by Holekamp et al. (1996); these
data were available only from 1988 to 2000. Nulliparous females were
distinguished from parous females by the permanent scarring of the pseudopenis
that occurs at first parturition (Frank & Glickman 1994). In this population, males
disperse from their natal clan after 2 years of age, but females rarely disperse
(Frank et al. 1995a). Therefore, disappearances of all females were attributed to
death. Once males reach 2 years of age, it is often difficult to determine whether
a missing male has dispersed or died. Because we were interested in survival

beyond 2 years of age, only females were used in analyses presented here.

Data Analysis

All statistical methods used in analyses of survivorship were drawn from
Lee (2003). Specifically, survivorship curves were created using the Kaplan—
Meier method. To examine the overall effect of maternal rank on survival we
used a Cox proportional hazard model, which included relative rank as a
continuous variable. Because litter size and maternal parity may also influence
survival, these factors were also included in the model. Litter size has been
shown to influence survival in Crocuta during the first two years of life, with
individuals from twin litters surviving better than singletons (Wahaj et al. 2007).
We included maternal parity as a factor because breeding experience has been
shown to influence offspring survival in other large mammals (Hastings & Testa
1998; McMahon & Bradshaw 2004; Robbins et al. 2006), and because

primiparous Crocuta may be particularly ill-equipped to rear young. Specifically,

18

 

female Crocuta first reproduce as early as 2.5 years of age in this population
(Holekamp et al. 1996), when their own feeding and hunting capabilities may not
yet be fully developed. To compare cub survivorship during the 14 months after
weaning, young females were categorized based on maternal rank or maternal
presence, and curves were compared using Cox’s F-test. Individuals still alive at
the end of the study period were included as right-censored data points in
survival analyses. STATISTICA 6.1 (StatSoft 2002) was used for all statistical
analyses. Mean values are presented :I: standard error. Differences between

groups were considered significant when P < 0.05.

RESULTS

Monthly mortality rates increased after weaning, and were higher after
weaning (maximum 0.13) than after den independence (maximum 0.06) or
sexual maturity (maximum 0.03; Figure 1.2). Mean monthly mortality for the first
12 months of life, when offspring are still nursing, was 0.053 :I: 0.008 (indicated
as horizontal line on Figure 1.2; range 0.011 to 0.093). We observed an increase
in mortality associated with den independence, but surprisingly, not with first
parturition. Mortality could only be monitored for 4 months after den
independence because juveniles often then began to wean. Since we cannot
detect conceptions in this species until after parturition, examination of mortality
rates a posteriori could potentially underestimate mortality during first parturition.
Therefore, it was necessary to examine mortality during the entire interval from

sexual maturity to 3.5 years of age, which encompasses the age interval when

19

0.18 r
0.16 -
0.14 .
0.12
0.10

0.08

Mortality rate

0.06 -

 

0.04 -
0.02 - E
0.00 PL 4% L + ‘ ' ‘ ' :

1234 2468101214 2468101214161820

Den indp. Weaning Sexual maturity
Months after

 

 

 

Figure 1.2. Monthly mortality rate for females during the months immediately
after den independence (n = 85), weaning (n = 60), and reproductive maturity (n
= 59). The x-axis indicates the number of months following each event. For any
individual, time zero is the date on which that individual left the den permanently,
weaned, or reached 24 months of age (i.e. sexual maturity), respectively. Fates
after den independence could only be followed for 4 months before juveniles
began to wean. The dashed line indicates the mean monthly mortality rate during
the first 12 months of life, when most juveniles are still nursing.

20

most females first reproduce (i.e., 2.5 to 3.5 years of age, Holekamp et al.
1996).Thus, the mortality rate during this period probably overestimates mortality
associated with first parturition as it includes all mortalities, even those due to
causes other than dystocia.

Overall, female survivorship was significantly influenced by maternal rank
(Cox proportional hazard: whole model x2 = 15.13, n = 164 females, P = 0.002;
maternal rank, P = 0.005). Daughters of high- and mid-ranking females
experienced higher survivorship than did their low-ranking counterparts (Figure
1.3a). Maternal parity also significantly influenced survival (P = 0.004; Figure
1.3b), but litter size did not (P = 0.21; Figure 1.3c). Cubs born to parous females
survived better than those born to nulliparous females. Further, when we looked
more specifically at survival after weaning, maternal rank was important. The
daughters of high-ranking females were more likely to survive after weaning than

were daughters of low-ranking females (Figure 1.4; Cox’s F1222 = 3.29, n = 60

females, P = 0.008).

Survival after weaning was greater for juvenile females whose mothers
were present in the population, than for those whose mothers died within 6
months of weaning (Figure 1.4b; Cox’s F243 = 3.68, n = 59 females, P = 0.031 ).
In all cases considered in this analysis, mothers were still alive at the time of
weaning. Thus, even though cubs stop relying on maternal milk as a food
resource when they are weaned, they clearly continue to rely on maternal aid for

survival.

21

 

(a) 1.0

.0
on

Proportion surviving

A
C-
V
.3
o

.0
on

Proportion surviving

0.8

0.6

.o
.5

Proportion surviving
O
'm

.0
o

 
  

0.6 ’

— High-ranking (n = 75)
-------- Mid-ranking (n = 42)
- - - Low-ranking (n = 47)

 

 

0.6

O
N

 

Age in years
— Parous (n = 133)
-------- Nulliparous (n = 30)

 

 

 

0 2 4 6 8 10 12 14

Age in years
— Singleton (n = 44)
g ........ Twin (n = 112)
i; - -- Triplet (n = 3)

fi‘r

Y I j j
------

 

 

 

4 6 8 10 12 14
Ageinyears

Figure 1.3. Survivorship curves for daughters born (a) to high-, mid-, and low-
ranking mothers, (b) to nulliparous and parous mothers, and (c) in singleton, twin,
and triplet litters. Cases in which the proportion of individuals surviving does not
reach zero are due to individuals still alive at the end of the study (i.e., right-

censored data points).

22

(a)

(b)

Proportion surviving

Proportion surviving

1.0

0.8 -

1.0

0.8 -

  
 
 

 

+ High-ranking
----A~ Low-ranking

"‘1 ----- A: ----A ----- A: ----A---A----A ----- A

 

  
 

 

4 6 8 10 12 14
Months after weaning

+ Mother alive
----Ar Mother dead

 

4 6 8 10 12 14
Months after weaning

Figure 1.4. Proportion of females surviving after weaning (a) based on social
rank (n = 35 high-ranking, n = 25 low-ranking) and (b) when mother was present
(n = 53) or absent (n = 6). Time zero is assigned for each individual as its date of

23

 

DISCUSSION

Our results demonstrate the maternal support can strongly influence
offspring survival beyond weaning in Crocuta, a species which displays both
intense competition over food and delayed development of a massive and
complex feeding apparatus. As predicted by the Competition & Constraint
hypothesis, but not the Feeding Competition hypothesis, weaning was
particularly challenging relative to other life history events in the lives of young
hyenas. Maternal support, particularly the ability of mothers to aggressively
displace conspecifics from food (i.e., social rank), exerts a strong influence on
offspring survival during this difficult time. These findings are strongly consistent
with the hypothesis that female dominance has evolved in Crocuta as a result,
not only of intense feeding competition, but also constrained developmental of
the feeding apparatus.

Consistent with the Competition & Constraint hypothesis, as well as with
the Feeding Competition hypothesis, we found that social rank had a significant
effect on female survivorship overall. Slow or prolonged development of the
feeding apparatus may contribute to reduced survivorship observed among the
daughters of nulliparous females, if these mothers have not yet themselves
achieved adult feeding or hunting capabilities. Alternatively, the effect of parity on
survival may result because nulliparous females lack important experience in
rearing young.

Although female dominance appears to have evolved as a mechanism to

compensate for the handicap experienced by young Crocuta during competitive

24

feeding, slow development of the feeding apparatus continues to pose a
formidable challenge to young hyenas. This is particularly evident among low-
ranking hyenas, for whom weaning is especially challenging. Indeed, despite the
tendency for low-ranking hyenas to wean at older ages than high-ranking hyenas
(Holekamp et al. 1996), survivorship is still lower among low-ranking hyenas.
Although the analyses presented here include only females, we expect that the
challenges posed by weaning and the effects of maternal status on survival
would be similar for male hyenas, as long as they remain in their natal clan.
Previous studies have found no difference in survivorship between males and
females in the first 2 years of life (Frank et al. 1995a; White 2005; see also
Chapter 2 of this dissertation). Once males begin to disperse, however, they
likely face even greater energetic challenges without the aid of their mothers than
do philopatric females.

In contrast to weaning, first parturition was not a period of high mortality
for females in this study. This result is surprising in light of previous work
suggesting that female Crocuta experience unusually high rates of mortality the
first time they give birth (Frank et al. 1995b). Frequent observations of dystocia
among nulliparous females in captivity (Frank & Glickman 1994) led Frank et al.
(1995b) to suggest that genital masculinization in female Crocuta is costly due to
increased risk of mortality to both mother and cub(s) at parturition. However, we
found no evidence of increased maternal mortality associated with first
parturition. In fact, mean monthly mortality for females between 2.5 and 3.5 years

of age, when most females first reproduce, was 0.0033 :t 0.0022, notably lower

25

 

than during surrounding age intervals. Our findings still leave open the possibility
that female genital masculinization involves other potential costs such as
increased risk of fetal mortality at parturition, difficulty in successfully conceiving
young, and problems maintaining pregnancies.

Although weaning is clearly challenging for young spotted hyenas, this is
probably also true to some extent for most mammalian carnivores. We suggest,
however, that weaning may be particularly challenging for young Crocuta who
are handicapped during intense feeding competition. Although additional data will
be needed to compare mortality after weaning among carnivore species,
available data do not indicate increased mortality after weaning in species other
than Crocuta. Mortality rates around weaning have been examined in lions
(Panthera Ieo), cheetah (Acinonyxjubatus) and southern elephant seals
(Mirounga Ieonine). In all three species, mortality rates are lower following
weaning than in the period prior to weaning (Packer et al. 1988; Caro 1994;
Pistorius et al. 2001 ). Though very preliminary, these comparisons are consistent
with the idea that weaning may be particularly challenging for Crocuta.

It also remains to be seen whether maternal support following weaning is
of unique importance in Crocuta relative to other taxa. In two carnivore families,
the Otariidae (fur seals and sea lions) and the Phocidae (true seals), maternal
care usually ends entirely at weaning (Schulz & Bowen 2005). However, young
of many other carnivores remain with their mothers for an extended period
following weaning, and thus may potentially benefit from maternal aid during this

time. Whereas many young carnivores are likely to rely on their mothers’ hunting

26

 

 

skills for food after they are weaned, Crocuta may be even more severely
handicapped because they lack not only the hunting skills needed to capture
prey, but also the necessary feeding capabilities to permit them to ingest food as
quickly as adults once a kill has been made.

Crocuta may represent an extreme example of constrained juvenile
development favoring the evolution of robust maternal behavior. However, we
would expect to find similar patterns of exaggerated maternal behavior in other
species with constrained development and parental care provided solely by the
mother. Examples of extended maternal care among placental mammals include
bats, walruses, and primates; in each of these groups lactation is particularly long
relative to female body mass (Hayssen 1993; Schulz & Bowen 2005). These may
all be cases in which developmental constraint hasselected for enhanced
maternal behavior. Juvenile development in bats may be constrained by
substantial post-natal development of the wing morphology that occurs before
young are capable of flight and independent foraging (Hughes et al. 1995;
Elangovan et al. 2002; Elangovan et al. 2004). Interestingly, bats tend to wean at
the same age at which they become volant (Hughes et al. 1995; Elangovan et al.
2002). Feeding performance of juvenile walruses may be limited by slow
development of the musculature or skull morphology necessary to consume
mollusks via suction-feeding (Schulz & Bowen 2005). Finally, primates are
characterized by relatively large brains, which are energetically expensive and
slow to develop (reviewed in Barton 2006). Although the evolutionary

relationships between brain size and life history in mammals remain unresolved,

27

the energetic demands required for brain growth may have favored extended
juvenile dependence among primates in general. Brain size does not seem to
explain the evolution of female dominance in some lemur species, however,
which are characterized by relatively small brains among primates (Kappeler
1996; Barton 2006). Thus, while each of these examples suggests that
developmental constraints may be important in the evolution of maternal
behavior, we emphasize that it was likely the unique combination of constrained
development in conjunction with extreme competition in Crocuta that lead not
simply to robust maternal behavior in general, but to female dominance in
particular.

Is there any evidence that a similar combination of intense competition
and developmental constraint may have lead to the evolution of female
dominance in lemurs? Most hypotheses suggested to explain the evolution of
female dominance in lemurs have focused on the harsh, seasonal fluctuations in
food availability as a key selective pressure for the evolution of this trait
(reviewed in Radespiel & Zimmermann 2001). Thus, while feeding competition
may be important in lemurs, we are not aware of any developmental constraints
that may have selected for dominant females as a compensatory mechanism in
these species. However, we suggest that detailed examinations of potential
developmental constraints in lemurs may be a profitable avenue of future
research. On the other hand, it is also possible that different evolutionary forces
than those proposed for Crocuta have lead to the evolution of female dominance

in lemurs.

28

The rarity of female-dominated societies among mammals suggests that
rather exceptional conditions must be present in order for this unusual trait to
evolve. Our findings suggest that the challenges posed by intense competition
amongst conspecifics over food, in combination with an extended period of
offspring dependence due to protracted development of the feeding apparatus,

selected for female dominance in spotted hyenas.

29

 

CHAPTER 2

ECOLOGICAL DETERMINANTS OF SURVIVAL AND REPRODUCTION IN THE
SPOTTED HYENA

INTRODUCTION

Predators play a key role in the structuring and dynamics of terrestrial and
aquatic ecosystems. In many communities, changes in predator populations
have strong effects on species at lower trophic levels. Predator-mediated trophic
cascades can lead to increases in herbivore populations, which in turn can cause
habitat loss or modification (Estes et al. 1998; Beschta 2003; Hebblewhite et al.
2005). Loss of large predators can result in mesopredator releases and
subsequent loss of diversity in bird, reptile, and rodent species (Crooks & Soulé
1999; Henke & Bryant 1999). Further, declines in predator species can cause
shifts in patterns of intraguild predation, altering dynamics at lower trophic levels
(Estes et al. 1998; Springer et al. 2003). More recently, Wilmers and Getz (2005)
suggested that the presence of top predator populations may even buffer their
ecosystems from the effects of climate change.

Mammalian carnivores occupy the role of top predators in many
ecosystems, yet the factors influencing the dynamics of carnivore populations
themselves are often poorly understood. In general, populations may be
influenced by bottom-up forces (resources; White 1978) and/or top-down forces
(natural enemies; Hairston et al. 1960). Bottom-up population control via prey
abundance has been suggested for a variety of carnivores (reviewed in Fuller &
Sievert 2001). Reduced food abundance can cause carnivore populations to

decline through starvation, increased susceptibility to disease, and/or increased

30

 

risk of intra- and interspecific killing (Schaller 1972; Mech 1977; Funk et al.
2001). Reduced food abundance can also affect populations by reducing energy
for reproduction (Boertje & Stephenson 1992; Creel & Creel 2002). Top-down
forces, including disease and anthropogenic disturbance, have also been
implicated in the dynamics of carnivore populations. Disease outbreaks can
dramatically increase mortality rates in carnivores, leading to population declines
(Young 1994; Roelke-Parker et al. 1996). Increased anthropogenic disturbance
can reduce carnivore abundance as a result of direct killing of carnivores by
humans (Woodroffe & Ginsberg 1998) as well as through indirect effects such as
disruption of behavior leading to reduced foraging efficiency (Wielgus 8. Bunnell
1994; Kerley et al. 2002; Boydston et al. 2003b). Interspecific competition among
carnivores (including intraguild predation, Holt & Polis 1997) has also been
suggested to influence predator populations (Laurenson 1995). The effects of
such competition may be complex. lnterspecific competition may reduce access
to food resources via exploitation and/or interference competition (bottom-up;
Creel & Creel 2002), but it may also include interspecific killing (top-down;
Palomares & Caro 1999). Further, there is also potential for nonlethal predator
effects, if risk of intraguild predation induces costly behavioral changes in a
subordinate predator.

Few studies have examined the influence of all of these factors
simultaneously (Kissui & Packer 2004). Here we describe long-term demographic
patterns in a population of free-living spotted hyenas (Crocuta crocuta) under

continuous observation since 1988, and assess the effects of per capita food

31

 

 

abundance, interspecific competition, anthropogenic disturbance, and disease on
two key determinants of population dynamics: reproduction and survival. Spotted
hyenas are the most abundant large carnivore in sub-Saharan Africa (Cardillo et
al. 2004) and occupy a wide diversity of habitats including deserts, montane
forests, woodlands, and savannas (Mills & Hofer 1998). Consequently, spotted
hyena populations are potentially important to a large number of African

ecosystems.

METHODS

The study was conducted in the Talek area of the Masai Mara National
Reserve (hereafter Mara), Kenya. Spotted hyenas are gregarious carnivores that
live in social groups called clans. Here, one large clan was observed between
July 1St 1988 and December 31312003. Individual hyenas were identified by
unique spot patterns, and sexed based on penile morphology (Frank et al. 1990).
Observations were made during two daily data-collection periods, between 0530
- 0900 hours and between 1700 — 2000 hours. During each data-collection
period, the Talek area was searched by vehicle, and an observation session was
initiated each time one or more hyenas was located. Observation sessions lasted
from 5 minutes to several hours and ended when observers left that individual or
group.

Female spotted hyenas are philopatric, while males disperse at adulthood
(Frank 1986a). Consequently, spotted hyena clans are composed of adult

females, their young offspring, and immigrant males. To assess clan

32

composition, females over 3 years old were considered adults, as was any
younger female that had already conceived her first litter. Males were considered
adults at 2 years of age. Resident natal males were adults born in the study clan
who had not yet dispersed. Resident immigrant males had emigrated from other
clans and been present in the study clan for at least 6 months. Juveniles were all
hyenas other than adults. Mean monthly clan size (i.e. the total number of
juveniles, adult females, and both natal and immigrant adult males present) was

calculated for each year of the study.

Determination of births, deaths, and other life history events

Spotted hyena cubs reside in dens until they are at least 8 to 9 months of
age (Boydston et al. 2005). Here den sites were visited regularly throughout the
study to monitor births and development of cubs. Ages of cubs were estimated to
within :I:7 days when they were first observed above ground (as in Holekamp et
al. 1996). A cub was considered independent of the den when it was found more
than 200m from the den on at least 4 consecutive occasions (Boydston et al.
2005)

Most male spotted hyenas disperse from their natal clan after 2 years of
age, but females rarely disperse (Frank et al. 1995a; East et al. 2003; Boydston
et al. 2005; Honer et al. 2007). Therefore, disappearances of all females, and
males less than 2 years old, were attributed to death in the current study, except
in cases of clan fission. It was often difficult to determine the fate of dispersing

males, so only females were used in analyses of adult mortality.

33

 

Ecological variables

Spotted hyenas prey primarily on ungulates they kill themselves, though
they frequently compete with lions (Panthera leo) for food at kills (Kruuk 1972).
The Talek area is comprised of rolling grassland grazed year round by resident
ungulates; these are joined for 3 to 4 months each year by large migratory herds.
To monitor prey abundance in Talek, biweekly counts were conducted between
0800 — 1000 hours of all ungulates within 100 m of two 4 km transect lines in
different areas of the Talek home range; one additional 4 km transect was added
in 2001. Transect counts were used to generate monthly estimates of prey
density, which ranged from 21.3 to 1917.5 animals/kmz, with a mean of 277.3 1
21.0 animals/km2 (n = 363 counts). Per capita prey density, estimated by dividing
prey density by the number of adult females present in the clan, was used as the
measure of prey abundance (Figure 2.1a). This per capita measure was selected
in order to assess the effects of both the abundance of prey animals and the
intensity of intraspecific competition for food resources.

Although increasing group size may lead to increasing food competition,
group size might also affect reproduction and survival independently of
intraspecific food competition. Increasing group size may confer benefits such as
reduced risk of predation or improved defense of resources, but it might also
increase rates of disease transmission within groups. Therefore, clan size was
included in our analyses in order to examine potential effects of intraspecific

interactions on survival and reproduction, in addition to effects of direct

34

Rainfall/Per capita prey density

Number of huts

.3
O
O

(D
O

O)
O

40-

20>

 

A l A I

 

 

O 1 1 1 1 L 1 1 4 1 L 1 1
1988 1990 1992 1994 1996 1998 2000 2002

2400 .

2000 r

1600 >

1200 >

800 ~

400 ~

Year

000°..fi..

60.00....

 

~ 0.20
1 0.18

0.16

40.14

0.12

-0.10

0.08

- 0.06

0.04
0.02
0.00

 

1970 19130 1990

Year

1950 1 960 2000

35

 

Rate of lion-hyena interactions

Figure 2.1. (a) Mean monthly (A) rainfall (mm), (I) lion-hyena interaction rate,
and (0) per capita prey abundance for each year of the study. Means for 1988
are not included since the study did not begin until July of that year. Rate of lion-
hyena interactions is a measure of competition between the two species. Per
capita prey abundance is prey density (prey animals/kmz) per adult female in the
clan. (b) Estimated human population adjacent to the study area, based on the
number of huts. A polynomial equation was fitted to census data (a) from
Lamprey and Reid (2004) to generate estimates (0) for all years. Horizontal bar
indicates the period of the current study.

competition for food. To analyze the effect of clan size on reproduction, juveniles
were excluded in the calculation of clan size.

The Talek study area is located on the reserve edge, adjacent to a
growing human population (Boydston et al. 2003b). Human census data for the
area are available from periodic surveys conducted in 1950, 1961, 1967, 1974,
1983, 1999, and 2002 (Lamprey & Reid 2004). These surveys provide population
estimates based on the total number of huts in the area. To obtain estimates of
population size for the current study period, these census data were used to fit a
polynomial equation describing population growth between 1950 and 2002. This
equation was then used to interpolate and extrapolate population estimates for
each year of this study (1988-2003; Figure 2.1 b). Based on these estimates, the
population grew from 1225 huts in 1988 to 2245 huts in 2003. This measure of
population size is expected to reflect overall levels of anthropogenic disturbance
in the study area.

Rainfall was recorded daily within the Talek home range; monthly rainfall
varied from 0 to 336 mm, with a mean of 89.5 :1: 5.1 mm. Mean monthly rainfall
was calculated for each year of the study (Figure 2.1a). Rainfall may influence
survival and reproduction through effects on disease dynamics (Altizer et al.
2006), flooding of hyena dens (Frank et al. 1995a), effects on the behavior of
other carnivores (Durant et al. 2004), effects on prey animals, or effects on rates
of human-carnivore conflict. In the study area, rainfall is strongly and positively
correlated with rates of livestock depredation (Kolowski & Holekamp 2006), a

pattern which has also been found in other areas of Kenya (Patterson et al. 2004;

36

Woodroffe & Frank 2005). Since local pastoralists often kill hyenas in response to
livestock depredation, rainfall may be an important predictor of mortality rates.
Rainfall may also influence food abundance, but this variation should be reflected
in our measure of prey abundance. In order to test whether a negative
relationship between rainfall and juvenile survival might be due to den flooding,
the relationship between the ecological predictors and juvenile survival (i.e.
recruitment to 2 years of age) was also analyzed separately for the period after
den independence, when den flooding no longer poses a risk to youngsters.
Lions are the primary competitors of spotted hyenas. The two species
have a high degree of diet overlap (Kruuk 1972; Hayward 2006). Further, lions
often steal food from hyenas, and represent a major source of hyena mortality
(Kruuk 1972; Mills 1990; Frank et al. 1995a). Therefore, the presence or
absence of lions with hyenas was recorded in each observation session. The
degree of competition with sympatric lions was estimated annually using the
mean monthly rate of lion-hyena interactions, calculated as the number of
observation sessions at which lions were present with hyenas during each
month, divided by the number of data-collection periods during that month
(Figure 2.1a). This measure controls for variation in intensity of observation

effort.

Data Analysis

Birthrates were calculated as the total number of hyena cubs born during

the year of interest, divided by the mean number of adult females present in the

37

clan during that year. Adult mortality was calculated as the total number of adult
female deaths during a given year, expressed as a proportion of all adult female
hyenas present at the beginning of that year. Recruitment was calculated as the
number of juvenile hyenas that survived to reach 2 years of age from the cohort
of individuals born in a given year, expressed as a proportion of all individuals
born in that cohort. This measure of recruitment reflects variation in juvenile
survival, and does not reflect variation in birthrates.

STATISTICA 6.1 (StatSoft 2002) was used for statistical analyses. The
Kaplan-Meier method was used to estimate age-specific survivorship for all
individuals born during the study. Individuals still alive at the end of the study
were included as right-censored data. Life table data are only available until 12
years of age since few females survive beyond this age. Mean values are
presented :1: 1 standard error. Multiple regression was used to examine the
effects of the ecological variables on annual reproduction (i.e. birthrate), adult
mortality and recruitment. One-year time intervals were selected for analysis in
order to minimize variance due to regular seasonal fluctuations in the
environment, and to minimize pseudoreplication due to temporal sampling. Lion-
hyena interaction rate, the measure of lion competition, was not normally
distributed, so these data were log-transformed to values that fit a normal
distribution.

An age-structured Leslie matrix model was constructed to estimate
population growth, based on vital rates observed over the entire study. This

estimate of population growth was selected because it was not influenced by

38

observed changes in clan size due to fission events. Annual survival probabilities
(P,) and fertilities (F,-) were estimated for females following Caswell (2001). Since
spotted hyenas breed year-round (Lindeque & Skinner 1982; Holekamp et al.

1999), P,- and F,- were estimated based on a birth-flow population as:

P' = l(i)+1(i+1)
' I(i-1)+l(i)

 

(Eqn.1.1)

 

FI=1(0.5)[’"‘+1;”"‘+1) (Eqn. 1.2)

where I,- is the survivorship to age i (given in Table 2.1) and m,- is the average
number of female offspring produced by a female of age i. A 1:1 offspring sex
ratio was assumed for the population. However, spotted hyenas do have a
seasonal peak in breeding (Holekamp et al. 1999). Therefore a population growth
estimate, based on a birth-pulse population with post-breeding censuses, was
also generated for comparison. Population growth (A) was calculated using

MATLAB (MathWorks 2001) as the dominant eigenvalue for the matrix.

RESULTS
Demographic patterns
Clan size during the study ranged from 27 to 79 hyenas with a mean of
57.5 1 0.8 (Figure 2.2). The clan underwent two fission events. The first occurred
during a 7-month period between late 1989 and early 1990 (Holekamp et al.
1993). The second occurred gradually over a period of years, and was complete
by late 2001 (K. Holekamp unpublished data). All animals leaving the clan during

both fissions were subsequently observed elsewhere. Onset of both fission

39

80 -
70- I _. I
60 -
50 i
40 r

30 -

Number of individuals

20 -

 

10 -

 

0 l l l l 1 L 4 I L l l l l l J_ l

1988 1990 1992 1994 1996 1998 2000 2002
Year

Figure 2.2. Mean (III) clan size, (0) number of adult females, (A) number of
resident immigrant males, and (0) number of juveniles present in the study clan
for each year between 1988 and 2003. Monthly means :1: SE are shown. Arrows
indicate two clan fission events. Horizontal line indicates canine distemper virus
epizootic.

40

events coincided with peaks in numbers of juveniles and overall clan size (Figure
2.2).

Reproductive output did not vary significantly with maternal age once
females reached reproductive maturity (Table 2.1; n = 34 females, KruskaI-Wallis
T= 8.60, df = 7, P > 0.28). Therefore, data for all mature females were pooled for
subsequent analyses of birthrates. Mortality in the first two years of life was 63%
and declined thereafter (Table 2.1). This drop in mortality at 2 years coincides
with the age at which spotted hyenas begin to reach reproductive maturity, so
mortality was subsequently examined separately for hyenas less than 2 years old
(juveniles) and those older than 2 years (adults). Since survivorship did not vary
with sex in the first 2 years of life (n = 329; Gehan’s Wilcoxon Test = -0.183, P =
0.86), males and females were grouped to examine juvenile mortality.

It was possible to determine the cause of 73 deaths (Figure 2.3). The
greatest mortality source was lions, accounting for 27% of deaths with known
causes. Humans and starvation of cubs after the death of the mother were each
responsible for 18% of deaths. Other important mortality sources were illness
(11%), infanticide (8%), siblicide (5%), and den flooding (4%). Whereas juvenile
mortalities (n = 49) were well represented in all source categories, adult
mortalities (n = 24) were caused almost exclusively by lions and humans, with

less than 2% in each other category (Figure 2.3).

41

nomodxm $9.229: do 82:5: E mcozom: L8 Esooom £58 Emu nmeomcmo...

 

 

 

Ed med F N :d Ed d 3 9.:

Ed 8d ed m «Pd 2d F 3 2-2

mad vmd mm; d Ed dd N d. E d 3

odd odd dd d Ed 8d F 02 dd

odd dd 3 3 2d 9d N 9: ME

med mdd mm; on Fwd dmd m ddm we

end «dd odd mm Ed 2d 4 03 dd

«ad «dd 2: mm dmd 3d d m. 3. 3

Ned «dd 43 t. and 3d N dd. 3

Rd 8d 8d 2 end ddd m. dB 3

d ddd d d Rd dud 4N odd 3

d 8d d 0 mod med 3 mam: 7d

228.. 3. .2.
F: 35:3 an: fizz—am Lon m:_>_>._:m 3a.. comonxo S memo»
3245.5 zozéEm 32%.... uses a c2883. 3:33.: 9.59 u a 5 8.4

.5er XENE

oEaEmoEou 05 8.. wofiEzwm EV bate.— ucm an: _m>_>5m 9622: Spam 9: 5 Son mmfiEfi Lo.— oBE 3: .PN 038.

42

60 _ Masai Mara (this study, n = 73)
“i
201 WW

 

 

 

60 _ Kalahari (Mills 1990, n = 28)

F

 

 

 

Percent of deaths
8

 

   

60I Ngorongoro & Serengeti (Kruuk 1972, n = 24)
4o .
20 ******* .
0
Lions Humans Starvation Illness Hyenas

Figure 2.3. Percent of deaths caused by the major modality sources for three
spotted hyena populations. Sample sizes for each population are indicated.
Deaths caused by disease are included as illness. Most deaths caused by
hyenas were either siblicide or infanticide. Kruuk (1972) lumped deaths caused
by starvation and illness, together they accounted for 21% of deaths. Asterisks
indicate that those deaths are divided equally between the two mortality sources
here for representation only. Age categories are indicated for the Masai Mara
population only. Individuals less than 2 years of age at death were categorized
as juveniles (hatched bars), while those older than 2 years were categorized as
adults (open bars). Note that a very small percentage of adults died from
starvation, illness, or hyenas in the Masai Mara. Almost all of the deaths due to
illness in the Kalahari were likely due to rabies (Mills 1990).

43

Based on the demographic matrix model, the birth-flow population growth
rate (A) was 1.01. This indicates that the population size was stable. The birth-

pulse model estimate of population growth was the same (A = 1.00).

Ecological influences

Annual birthrates (Figure 2.4a) were significantly and positively correlated
with clan size (values expressed as the standardized regression coefficient :1 SE;
b = 0.584 :I: 0.202, t= 2.90, P = 0.018) and prey abundance (b = 0.597 1: 0.208, t
= 2.87, P = 0.018), while they were negatively correlated with lion competition (b
= -0.571 1 0.204, t= -2.80, P = 0.021 ). Neither rainfall (P > 0.22), nor human
population size (P > 0.41) had a significant effect on birthrates. Overall, the
predictor variables explained 51% of the variation in annual birthrates (F59 =
3.91, P = 0.037).

Annual recruitment (Figure 2.4b) of juveniles to 2 years of age was
significantly and negatively correlated with rainfall (b = -0.566 :1: 0.197, t= -2.88,
P = 0.024) and lion competition (b = -0.548 1: 0.190, t= -2.88, P = 0.024), but not
with clan size (P > 0.19), prey abundance (P > 0.27), or human population size
(P > 0.64). Overall, the predictor variables explained 68% of variation in annual
recruitment (F51 = 6.02, P = 0.018). The results for recruitment between den
independence and 2 years of age were very similar to those of the previous
analysis (overall model: Rzadj = 0.70, F5] = 6.62, P = 0.014); recruitment of den-
independent cubs was significantly negatively correlated with rainfall (b = -0.808,

P = 0.004) and lion competition (b = -0.464, P = 0.039), but not with clan size,

44

Recruitment

 

Adult
mortality rate

 

0.0 ‘ ‘
1988 1990 1992 1994 1996 1998 2000 2002

Figure 2.4. Annual (a) birthrates, (b) recruitment rate and (c) adult mortality rate
from 1989 to 2003. The recruitment rate is the proportion of individuals born in a
given year that survive until 2 years of age. Rates are not given for 1988 since
the study did not begin until July of that year. Recruitment data for individuals
born in 2002 and 2003 were not yet available. Arrows indicate two clan fission
events. Horizontal line indicates canine distemper virus epizootic.

45

prey abundance, or human population size. The effects of all significant
ecological variables on recruitment were similar to each other in absolute
magnitude, as were effects on birthrates. Using the ecological predictor
variables, we were unable to fit an overall model to explain a significant amount
of variation in annual adult mortality rates (Figure 2.4c; F53 = 0.88, P > 0.53).

The relationship between rainfall and human-carnivore conflict in this
population may result from seasonal changes in prey distribution with rainfall.
While rainfall and prey abundance were not correlated among years (r = 0.21, P
= 0.45, n = 15 years), there was a trend towards a negative correlation between
quarterly rainfall and prey density (r = -0.22, P = 0.09, n = 62 quarters).

There was no evidence of high adult mortality (Figure 2.40) coincident with
epidemics of either canine distemper virus (CDV),which infected hyenas and
other carnivores in the ecosystem in late 1994 and early 1995 (Roelke-Parker et
al. 1996), or rabies that infected wild dogs (Lycaon pictus) in late 1989 (Kat et al.
1995). However, recruitment of juveniles was lowest for the cohort born in 1994
(Figure 2.4b). Between July 1994 and March 1995 three juvenile hyenas died
from illness. Although post-mortem tests were not performed to confirm CDV
infection, two exhibited symptoms consistent with CDV. These three deaths
account for 50% of all deaths due to illness during the entire study (Figure 2.3).
Monthly mortality rates during the CDV epizootic (Jul 1994 to Jun 1995) were
compared with mortality rates in months before and after the epizootic (Jan-
Jun1993 and Jul-Dec 1996), but there was no significant difference in mortality

between CDV and non-CDV periods for either juveniles (Mann-Whitney U test; U

46

= 64.0, P > 0.6, n = 12 per group) or adults (Mann-Whitney U test; U = 71.0, P >

0.9, n = 12 per group).

DISCUSSION

Both top-down and bottom-up forces influenced spotted hyena
demography. Per capita food abundance, which is based on both prey
abundance and hyena abundance (i.e. intraspecific competition), had a positive
effect on birthrates. This is consistent with other studies indicating the importance
of food abundance for reproduction in spotted hyenas. Cooper (1993) and
Holekamp et al. (1999) found that seasonal variation in reproduction corresponds
to variation in local prey abundance. Moreover, variation in social rank, which
determines priority of access to food, is strongly correlated with individual
variation in reproductive rate (Holekamp et al. 1996; Hofer & East 2003).

Food abundance did not influence survival. Although nursing cubs were
observed to starve to death after the loss of their mother, only one adult has
been observed to die as a result of starvation. The Mara is relatively prey-rich,
with a year-round resident ungulate population (Ogutu & Dublin 2002).
Consequently, prey abundance may rarely reach levels low enough to cause
hyenas to starve. In other populations with lower prey abundance or greater
fluctuations in prey abundance (e.g., Mills 1990; Hofer & East 1993a), food
abundance may have a greater effect on survival either through starvation,
increased susceptibility to disease, or increased competition with other

carnivores.

47

Clan size had a positive effect on birthrates, suggesting that group living,
and living in large groups in particular, confers significant benefits. Lion density in
the Mara is high (Ogutu & Dublin 2002), and this might favor large hyena groups
in order to protect cubs and/or to acquire and defend carcasses. The positive
effect of clan size on birthrates, but not recruitment, suggests that the advantage
of large group size is in access to food resources, rather than protection of cubs.
The relatively high abundance of prey animals in the Mara may also be important
in supporting large hyena groups. Indeed, hyena density in the Mara is among
the highest reported for any population (Trinkel et al. 2006), and large clan size is
likely to facilitate the defense of food resources from conspecifics in the
ecosystem. Clan size ranges widely among spotted hyena populations from 10 to
80 hyenas (Kruuk 1972; Mills 1990). This indicates that the factors favoring large
social groups in the Mara are not ubiquitous across Africa, but we expect that lion
populations and prey density are important determinants of clan size continent-
wide.

Rainfall was negatively correlated with recruitment. This effect cannot be
explained solely by den flooding during periods of heavy rain. Such events are
quite rare (Frank et al. 1995a; this study), and the negative relationship between
rainfall and recruitment remains even for older cubs that no longer reside at
dens. Rainfall could have a negative effect on recruitment due to increased rates
of disease infection, though disease does not seem to be of primary importance
in this population. It seems most likely that rainfall influenced recruitment as a

result of deliberate killing of hyenas by local pastoralists, in response to livestock

48

depredation. Juveniles are vulnerable to direct killing by humans, as well as to
starvation if their mothers are killed. Livestock depredation is thought to increase
during periods of high rainfall because abundance of natural prey species
fluctuates with rainfall; when natural prey are scarce conflicts increase (SabenIval
et al. 1994; Polisar et al. 2003; Woodroffe & Frank 2005). In Talek, wider
availability of green vegetation during periods of rain leads to a more dispersed
distribution of prey.

Using a general estimate of anthropogenic disturbance, human population
size, we found no effect on measures of survival or reproduction. However, this
estimate of population size fails to capture changes in human behavior or smaller
scale patterns of disturbance that may be affecting the hyena population. Given
that the local human population is known to influence both daily activity patterns
and space use in this hyena population (Boydston et al. 2003b; Kolowski et al.
2007), it will be important to determine whether these behavioral changes are
buffering the hyena population from more severe (i.e. demographic) effects of
human disturbance, or whether they signal demographic changes that we have
yet to detect.

Evidence indicates that disease has played a relatively minor role in the
population dynamics of spotted hyenas during this study. Disease was not a
major source of mortality in this population, and the two epizootics that occurred
in this ecosystem during the study period had no noticeable impact on adult
mortality rates. Although the CDV epizootic may have influenced juvenile

survival, the effect was not statistically significant. The magnitude of any effect of

49

the CDV and rabies epizootics on the spotted hyena population was small in
comparison to their respective effects on the lion and wild dog populations in this
ecosystem. The CDV epizootic killed lions in all age classes and resulted in a
loss of approximately 30% of the population (Roelke-Parker et al. 1996), and the
rabies epizootic killed roughly one third of the local wild dogs (21 dogs; Kat et al.
1995). Indeed, the finding that disease has had a minor role in the dynamics of
this spotted hyena population stands in contrast to the predominant influence that
disease has had on lion and wild dog populations (Woodroffe & Ginsberg 1999;
Kissui & Packer 2004). This discrepancy may reflect a general difference in
disease resistance between hyenas and other carnivores. Greater disease
resistance may have evolved in spotted hyenas and other hyaenids because of
their heavy reliance on carrion and scavenging; a similar hypothesis has been
suggested for scavenging birds (Blount et al. 2003). However, data from other
hyena species will be needed to test this hypothesis. Counter to the prediction of
this hypothesis, results from other studies suggest the effects of disease may
vary among spotted hyena populations (Figure 2.3). While disease-related
mortalities are relatively rare in the Serengeti (Hofer & East 1995; East et al.
2001), 43% of known mortalities in the Kalahari were caused by rabies (Mills
1990). Interestingly, the two rabies outbreaks that infected Kalahari spotted
hyenas were not known to kill any brown hyenas (Hyaena brunnea) in the area
(Mills 1990).

Although lions and humans were the major causes of mortality for adult

spotted hyenas in the current study (Figure 2.3), we were unable to explain

50

 

variation in adult mortality rates using these and other ecological predictors. ~
Perhaps adults are less susceptible than juveniles to negative effects of
ecological variation apparent during this study. It is also possible that both
competition with lions and human disturbance are influencing adult mortality, but
not in an easily predictable manor. For example, killing of hyenas by local
pastoralists can occur in clumped events such as mass poisonings (Holekamp et
al. 1993). Similarly, being killed by a lion may be a chance event for an adult
hyena, since adults regularly come into close contact with lions without dying.
Such random events can potentially have significant and rapid effects on hyena
demographics (Holekamp et al. 1993), and we hypothesize that stochastic events
contribute greatly to adult mortality in this species.

Interspecific competition with lions was the only ecological factor found to
influence both birthrates and recruitment. While the effect of interspecific
competition on birthrates is likely due to competition for food (either interference
or exploitation), the effect on recruitment could be due indirectly to feeding
competition and/or directly to intraguild predation on juveniles. Although further
research will be needed to assess the potential effects of feeding competition on
survival, lions were the single leading cause of hyena mortality in this study and
in at least three other hyena populations (Ngorongoro, Serengeti & Etosha:
Kruuk 1972; Trinkel & Kastberger 2005). This suggests that direct killing is an
important mode of competition between these species. Further, it is possible that

lions influence hyenas populations by inducing behavioral changes in hyenas

51

(i.e. nonlethal predator effects), though such effects have not yet been
documented.

lnterspeciflc competition between carnivores can be particularly intense
relative to competition within other guilds, as carnivores have both morphological
and behavioral adaptations for killing (Palomares & Caro 1999; Creel et al.
2001). Our study adds to a growing literature (lions, Cooper 1991; cheetahs,
Laurenson 1995; wild dogs, Creel & Creel 2002; cheetahs, Durant et al. 2004)
suggesting that interspecific competition may be more important than previously
recognized in the dynamics of large carnivores populations.

While both top-down (e.g. lion predation and humans) and bottom-up
forces (e.g. food competition) influenced spotted hyena demography here, the
relative importance of each varied among demographic measures. Survival was
most strongly influenced by top-down forces, whereas reproduction was
influenced by bottom-up forces. These data support the growing consensus that
both top-down and bottom-up mechanisms are important in population dynamics
and community structure (eg. Hunter & Price 1992; Menge 2000), and our
results highlight the need for greater understanding of the simultaneous roles of
both mechanisms (Munch et al. 2005; Borer et al. 2006). Examining the
contributions of both top-down and bottom-up forces to the underlying
demographic processes affecting population growth (i.e. survival, reproduction

and migration) should yield important insights into their respective roles.

52

 

CHAPTER 3

GENETIC DIVERSITY MAINTAINED IN A POPULATION OF LARGE
CARNIVORES DESPITE A RECENT BOTTLENECK

INTRODUCTION

The maintenance of genetic diversity has become a prominent
management objective in conservation biology, yet the ecological and
demographic processes affecting population levels of genetic diversity in wild
populations, and the consequences of losses of genetic diversity remain poorly
understood. Increasing habitat loss and fragmentation, due largely to
anthropogenic influences, have resulted in size reductions in many plant and
animal populations. Whereas the persistence of small populations is immediately
threatened by non-genetic factors such as demographic or environmental
stochasticity (Lande 1988), increasing attention has focused on genetic factors
associated with population viability. Small populations are expected to
experience inbreeding and the loss of genetic variation due to genetic drift that
may, in turn, lead to reduced population fitness and limited evolutionary potential
(Lynch 1996; Lacy 1997; Keller & Waller 2002; Reed & Frankham 2003).

Many conservation efforts focus on recovery of small populations using
reintroductions, habitat restoration, dispersal corridors, and other approaches.
Although numerous populations have rebounded due to recovery efforts (e.g.,
reintroduction of Canis lupus in Yellowstone, Bangs et al. 1998; recovery of
Haliaeetus albicilla in Europe, Hailer et al. 2006), optimism has been tempered
by concerns regarding the genetic legacies of small ancestral populations (i.e.

population bottlenecks, including founder events). Considerable theoretical (e.g.

53

Chakraborty & Nei 1977) and empirical (e.g. Leberg 1992; England et al. 2003)
attention has been directed to estimating generational declines in heterozygosity
and allelic diversity due to bottlenecks in population size. Unfortunately, a clear
understanding of how population bottlenecks influence the long-term viability of
populations is lacking. In particular, two questions remain unanswered. First,
what is the relative importance of levels of genetic diversity to the persistence of
small populations in the wild (Lande 1988; Caro 8 Laurenson 1994; Soulé & Mills
1998; Spielman et al. 2004; Frankham 2005)? Second, and most relevant to the
current study, what is the relationship between demographic bottlenecks and
genetic diversity in wild populations? Here, we use the term demographic
bottleneck to refer to a decline in population census size. Although theory
predicts that genetic diversity will be lost when effective population size becomes
small (e.g. Nei et al. 1975), field studies following population declines have
produced mixed results.

Numerous studies have found significant losses of genetic variation
following a demographic bottleneck (Panthera leo, Packer et al. 1991; Mirounga
angustirostrus, Hoelzel et al. 1993; Tympanuchus cupido, Bouzat et al. 1998;
Loxodonta africana, Whitehouse & Harley 2001). Meanwhile, other populations
show high levels of genetic diversity despite demographic bottlenecks (Cam's
lupus, Forbes & Boyd 1996; Terrapene ornate, Kuo & Janzen 2004; Cam's lupus,
Aspi et al. 2006; Haliaeetus albicilla, Hailer et al. 2006; Falco peregrinus, Brown
et al. 2007; Ovis aries, Kaeuffer et al. 2007). These latter results support the

notion that genetic bottlenecks have been overemphasized in conservation

54

biology (e.g. Lande 1988; Dinerstein & McCracken 1990). However, the varied
results from field studies speak to the complexity of the relationship between
population history and genetic diversity. At this juncture, more studies from both
population recoveries and reintroductions are necessary to help elucidate the
circumstances under which genetic variation is lost or maintained in small
populations, and the extent to which reduction in population size predicts local
extinction.

Large carnivores appear particularly vulnerable to ecological and
demographic changes that can impact population levels of genetic diversity
because of low densities, tendencies to range over large areas, and the threat
they pose to humans and livestock (Woodroffe & Ginsberg 1998; Woodroffe
2000; Mills et al. 2001). Indeed, large carnivore populations around the world are
declining and experiencing local extinctions, due largely to the direct and indirect
effects of human activity (Woodroffe & Ginsberg 1998; Woodroffe 2001).
Consequently, understanding the factors that allow small populations of large
carnivores to persist is of particular conservation importance.

The present study examines the population genetic outcome of a recent
demographic bottleneck in a population of free-living spotted hyenas (Crocuta
crocuta). Although spotted hyenas are relatively abundant across Africa, they
have undergone local extinctions in many areas (Mills & Hofer 1998). In what is
now Amboseli National Park, in southern Kenya, the spotted hyena population
recently rebounded from a relatively small population just a few decades ago.

Amboseli is a small national park, with an area of only 390 km2, comprised

55

 

primarily of semi-arid savanna. It is now largely surrounded by habitat dedicated
to agriculture and pastoralism (Campbell et al. 2005), although dispersal
corridors clearly persist, as indicated by occasional sightings in the park of
certain rare carnivores (e.g. African wild dogs, Lycaon pictus; H. Watts & W.
Dowd unpublished data).

Early reports from this part of Kenya make clear that spotted hyenas once
abounded in the area that is now Amboseli National Park (Thompson 1885).
However, in the early 1970’s, the Amboseli hyena population experienced a
major decline, possibly due to disease (Faith & Behrensmeyer 2006) or local
anthropogenic activity. Subsequently, hyena sightings were infrequent (Faith &
Behrensmeyer 2006) throughout the 1970’s and 1980’s, and the park’s hyena
population was estimated at just 50 animals (C. Moss personal communication).
The period spanning the 1970’s and 1980’s coincided with growth in the local
pastoral population and expansion of agricultural practices, leading to habitat
loss and fragmentation in the ecosystem (Campbell et al. 2003; Campbell et al.
2005), changes that were likely detrimental to many animal populations. Then, in
the 1990’s, the hyena population underwent a rapid expansion (Faith &
Behrensmeyer 2006), which were likely related to two other changes that
occurred in Amboseli in the same period: a dramatic reduction in the lion
population (a major competitor with hyenas) due to human persecution
(Behrensmeyer 1993; Chardonnet 2002), and an increase in the abundance of
prey species such as wildebeest and zebra due to the expansion of grassland

habitat (Western 1989). Regardless of the cause of the marked population

56

 

increase, by the early 2000’s the hyena population in Amboseli was estimated to
be between 300-400 animals (H. Watts & K. Holekamp, unpublished data).
Here, we examine the genetic consequences of this demographic
bottleneck by comparing genetic diversity and patterns of relatedness in the
Amboseli hyena population with a second hyena population known not to have
undergone a demographic bottleneck. Further, we discuss factors that may have

influenced the population genetic outcome of the Amboseli bottleneck.

METHODS

Study populations and sampling

Spotted hyenas live in large social groups called clans, which defend
group territories (Kruuk 1972). Whereas female spotted hyenas are philopatric,
male hyenas disperse from their natal clan once they reach adulthood (Boydston
et al. 2005). Consequently, hyena clans are typically composed of natal females,
their offspring, and immigrant males. As part of a larger study on the behavioral
ecology of spotted hyenas in Amboseli National Park, we monitored hyenas
belonging to two clans, the Airstrip and OI Tukai clans, from July 2003 to July
2005 (Figure 3.1a). During this period we collected fecal and tissue samples for
genetic analysis from 80 individually recognized hyenas. For males present in
the study clans when research commenced, we were able to distinguish
immigrant males from natal males who had not yet dispersed based on
pronounced behavioral differences, particularly during agonistic encounters

(Holekamp & Smale 1998).

57

 

 

      

 

 

1’) ~ 1 ° .. . .
i. Masai Mara. 00%
National' Resenter ”
,2
an
i " 1 o 12 aztlIoIfiaw 1 o 1 2 3 Kilometers
\ , ‘ . . 1.:- v” II:-

 

 

Figure 3.1 . The territories of the study clans in (a) the Masai Mara National
Reserve, and (b) Amboseli National Park. Areas within the park boundaries are
lightly shaded, while clan territories are darkly shaded.

58

In order to evaluate levels of genetic diversity in the Amboseli population,
it was important to have a large reference population with known historical
demographic and genetic data for comparison (Bouzat 2000; Matocq &
Villablanca 2001). Spotted hyenas in the Masai Mara Game Reserve, located
approximately 250 km west of Amboseli, have been monitored continuously
since 1979 (e.g. Frank et al. 1995a). The Masai Mara (hereafter Mara) is a large
reserve, characterized by rolling savanna habitat, which covers an area of 1500
km2. In contrast to Amboseli, the Mara hyena population is known to have
remained relatively large and stable since the late 1970’s. Blood and tissue
samples were collected from 72 individuals in this population as described by
Engh et al. (2002). To ensure similarity in sampling intensity and timing of sample
collection between Amboseli and Mara populations, samples were taken from
known individuals belonging to two intensively studied Mara clans between July
2002 and July 2004 (Figure 3.1b). Study clan sizes were similar in both
populations (see Table 3.2 in Results) and covered similar geographic sampling
areas. The Amboseli clans defended adjacent territories covering 28.0 km2
(Airstrip clan) and 26.4 km2 (OI Tukai clan), while the Mara clans were separated
by a single territory; these clans defended territories covering 24.7 km2 (Talek

West clan) and 31.0 km2 (Mara River clan; Kolowski et al. 2007).

Microsatellite genotyping

Most DNA samples from the Amboseli population were obtained from

feces, all of which were collected from known hyenas within 10 minutes of

59

defecation. Samples were then either (i) extracted shortly after collection (2 — 6
hrs), or (ii) immediately frozen or stored in >95% ethanol for later extraction
(usually 18 — 36 hrs, some samples stored up to 28 months). Fecal DNA was
extracted using a QIAamp DNA Stool Mini Kit (Qiagen) following the
manufacturer’s protocol for an approximately 19 sample. Rather than originating
from feces, most DNA samples from the Mara population originated from blood
collected from anaesthetized animals (for details see Engh et al. 2002). DNA was
extracted from blood samples shortly after collection using a Puregene kit
(Gentra Systems). For tissue samples from both populations, DNA was extracted
using a DNeasy Blood & Tissue Kit (Qiagen).

We amplified 10 microsatellite loci (CCr01, CCr04, CCr07, CCr11, CCr12,
CCr13, CCr14, CCr15, CCr16, CCr17) previously described for this species
(Libants et al. 2000; Engh 2002; Van Horn et al. 2004). Reaction conditions used
for DNA extracted from blood and tissue followed Libants et al. (2000) and Engh
(2002). For DNA extracted from feces, these conditions were modified as follows
(i) reactions contained either 1 pl (CCr11), 3pl (CCr07, CCr15), or 5 pl (CCr01,
CCr04, CCr12, CCr13, CCr14, CCr16, CCr17) of DNA extract, (ii) 0.25 pg of
bovine serum albumin (BSA) was added to reactions, and (iii) 1U ChromaTaq
DNA polymerase (Denville Scientific) was used for each reaction. Additionally,
the number of PCR cycles was increased to 40 for amplification of all loci.
Complete details of PCR reaction conditions are provided in Appendix B.

To genotype individuals using DNA extracted from fecal samples, a

modified multi-stage, multiple-tubes approach (Navidi et al. 1992; Taberlet et al.

60

1996) was used. First, a single positive amplification was carried out for one
extract from each individual. For individuals scored as heterozygous from this
first amplification, no further amplifications were performed. In cases where only
one allele was scored (homozygotes) on the first amplification, a second
amplification was performed. If the second amplification produced a
heterozygous genotype sharing one allele with the initial amplification, the
individual was scored as heterozygous. If the second amplification was
homozygous for the same allele as in the first amplification, the individual was
scored as homozygous for that allele. Cases in which the second amplification
was homozygous for a different allele than the first amplification were scored as
heterozygous for those two alleles. In cases where more than two alleles were
observed in either replicate, the sample was presumed to have been

contaminated by more than one individual and was discarded.

Reliability of genotypes obtained from fecal DNA

The use of fecal DNA with microsatellite markers typically results in a
higher frequency of genotyping errors than with other source materials such as
blood or tissue, which generally yield a higher quality and quantity of DNA (Bayes
et al. 2000). Since the source material for DNA differed between our two study
populations, it was necessary to assess the reliability of genotypes generated
from fecal source material. To assess error rates for genotypes derived from
fecal samples, we randomly selected a subset of individuals for each locus. This

subset of individuals was genotyped from fecal DNA using a traditional multiple-

61

tubes approach (Navidi et al. 1992; Taberlet et al. 1996). Our goal was to sample
approximately 10% of individuals (i.e., 8 or 9 individuals per locus). Genotypes
obtained using the traditional multiple-tubes approach were then used to
estimate error rates. Initially 3 replicate positive PCRs were performed. If the
initial set of PCRs resulted in at least 2 identical heterozygous genotypes for an
individual, that individual was classified as heterozygous for that genotype. If 2 or
more of the initial PCRs were scored as homozygous genotypes, then an
additional 4 replicate PCRs were conducted, for a total of 7 independent replicate
positive PCRs. Individuals with identical homozygous genotypes for at least 6
replicate PCRs were classified as homozygous. An allele that appeared only
once in the 7 replicates was classified as a false allele. Individuals for which 2
different alleles were each scored in 2 or more replicates were classified as
heterozygotes.

The traditional multiple-tubes approach generated a consensus genotype
for each individual. All positive amplifications performed for an individual were
then used to estimate error rates following Broquet and Petit (2004). Specifically,
the rate of allelic dropout was calculated as the number of amplifications in which
an allele was lost as a proportion of all amplifications of heterozygous individuals
(based on consensus genotype) at a given locus. The probability of a false allele
was determined as the number of amplifications resulting in one or more false
alleles expressed as a proportion of all amplifications for a given locus.

Additionally, for four individuals from the Mara population, DNA samples

derived from both blood and feces were available. For these individuals,

62

 

genotypes obtained based on 1 to 3 amplifications of fecal DNA per individual
were compared to genotypes obtained using blood samples. Results from blood
samples are presented in Table 3.2, but were not used to calculate overall error

rates.

Measures of genetic diversity and relatedness
We used the program CERVUS 3.0 (Kalinowski et al. 2007) to estimate
the number of alleles (A), allele frequencies, observed and expected

heterozygosities (HO and HE), and to test for deviations from Hardy-Weinberg
equilibrium. Genetic diversity was compared between populations using loci-
based measures (A, HO, and HE) and two measures to quantify individual genetic
diversity: individual heterozygosity (HI) and internal relatedness (IR). H. is the
proportion of heterozygous loci per individual. Due to differences in allele
frequencies between the two populations, we were unable to standardize HI to

account for missing data (Coltman et al. 1999; Lukas et al. 2004). Nevertheless,
since missing data were similarly distributed in both populations, they should not
have biased this measure. Internal relatedness (Amos et al. 2001) was

calculated as:

2 _ .
i2%-§—fii (Eqn. 3.1)

where N is the total number of loci, H is the number of loci that are homozygous,

and f,- is the frequency of the rth allele in the genotype across all loci. IR is

analogous to Queller and Goodnight’s (1989) R, but reflects genetic correlation

63

 

between a pair of alleles at a locus. Thus, it can be thought of as reflecting the
degree of relatedness between an individual’s parents. More negative IR values

indicate greater genetic diversity and suggest that an individual is more

“outbred”. Only individuals genotyped at 5 or more loci (37:1: SE = 9.5 :t 0.1 loci)
were included in these analyses. The Mann-Whitney U-test was used to test for
differences in genetic diversity between populations. Unless otherwise indicated,
statistical analyses were performed in Statistica (StatSoft 2002). Differences
between groups were considered significant when P < 0.05. Mean values are
presented :I: 1 standard error (SE).

The program BOTTLENECK (Cornuet & Luikart 1996) was used to test for
a genetic signature of a population bottleneck. Genetic drift at small population
sizes is expected to lead to a loss of rare alleles and a transient excess of
heterozygosity. A Wilcoxon signed-rank test was used to test for a heterozygote
excess, and the allele frequency distribution was examined for a lack of low
frequency alleles (‘mode-shift’ test). A two-phased mutation (TPM) model was
used with 95% probability of single-step mutations, as recommended for
microsatellites (Piry et al. 1999). Bottleneck tests were performed for each
population as a whole and for each clan individually because clan demographic
histories may differ.

In order to asses whether patterns of relatedness differed between
populations, pairwise relatedness values (R) were estimated for individuals
sampled from each population using program RELATEDNESS 5.0 (Queller &

Goodnight 1989). Since the inclusion of related animals to estimate population

64

allele frequencies can bias R (Queller 8 Goodnight 1989), only adults were used
to estimate allele frequencies. Analyses were conducted separately for each
population. While immigrant males were included in estimates of population allele
frequencies, pairwise estimates of relatedness were only generated for natal
animals. Only individuals genotyped at 8 or more loci were included in
relatedness analyses. Since pairwise relatedness values are not independent,
two-sample randomization tests (9999 iterations) implemented in POPTOOLS
(Hood 2006) were used to test for differences in mean relatedness between

groups.

RESULTS
Reliability of genotypes
For fecal DNA, across all loci the rate of allelic dropout was 0.027 and the
probability of a false allele was 0.010 (Table 3.1). Error rates varied between loci,
though performance does appear to reflect allele size (range 95 to 199 bp).
Agreement between genotypes generated from both feces and blood was 100%

(39/39 genotypes, 4 individuals).

Genetic diversity 8 test for bottleneck

All loci were in Hardy-Weinberg equilibrium in both populations, except for
CCr16 in the Mara population (x2 = 9.68, df = 1, P < 0.05), which exhibited a
deficiency of heterozygotes (see Appendix B for details of polymorphism at each

locus). When the entire available Mara dataset was examined

65

Table 3.1. Rates of allelic dropout (ADO) and false alleles (FA) for amplification
of fecal DNA. To estimate error rates, a reference genotype was determined
using either a multiple-tubes approach or DNA extracted from a blood sample
from the same individual. Sample sizes for the multiple-tubes approach are
presented as the number of individuals. Matched blood samples were available
for 4 individuals. N/A indicates that the ADO could not be calculated because all
individuals sampled were homozygous for this locus.

 

 

 

 

Multiple-tubes Blood samples

Locus ADO FA N ADO FA
CCr01 N/A 0 9 0 0
CCr04 0.023 0.020 13 0 0
CCr07 0 0 9 0 0
CCr1 1 0 0 9 0 0
CCr12 0 0.036 8 0 0
CCr13 0 0 9 0 0
CCr14 0.211 0.029 6 0 0
CCr15 0.042 0 8 0 0
CCr16 0 0 9 0 0
CCr17 0 0 6 0 0
Overall* 0.027 0.010 3? =86

 

*Calculated following Broquet and Petit (2004)

66

(samples collected between 1990-2005, Van Horn et al. 2004), however, all loci
were in Hardy-Weinberg equilibrium.

We found no evidence for lower genetic variation in the Amboseli
population relative to the Mara population. Neither the mean number of alleles

(Mann-Whitney U = 46.5, n1 = n2 = 10, P = 0.79) nor observed heterozygosity (U
= 46.0, n1 = n2 = 10, P = 0.76) differed significantly between the Amboseli and

Mara populations (Table 3.2). Similarly, H. did not vary significantly between
populations (U = 2667.5, n1 = 80, n2 = 72, P = 0.43). There was a trend for IR to
be higher in the Mara (U = 2406.0, n1 = 80, n2 = 72, P = 0.08).

Consistent with these results, we found no genetic signature of a
population bottleneck in Amboseli. There was no excess of heterozygosity (whole
population: P = 0.5, OI Tukai clan: P = 0.5, Airstrip clan: P = 0.3), nor was their
any evidence of a mode-shift in the allele frequency distribution in Amboseli as a
whole, or in either clan (Figure 3.2). Similarly in the Mara, we found no excess of
heterozygosity (whole population: P = 0.4, Talek West clan: P = 0.5, Mara River
clan: P = 0.7), and no evidence of a mode-shift in allele frequency distributions

(Figure 3.2).

Relatedness

Mean R—values for known mother-offspring pairs (Mara: 0.45 1: 0.04, n =
27 pairs; Amboseli: 0.42 :I: 0.04, n = 32 pairs) were only slightly lower than the
expected value of 0.5. In both populations relatedness was significantly higher

between individuals in the same clan, than between individuals from different

67

.Coowv 2950.3. E0: 928 moum 5.0 992 :o Ema

.9: n 2v _oo_ mmocom 85:28 So mo:_m> 55 $286.: 5.25.2

 

 

 

 

 

mmdd a did Rod a Ed Bed 4 Rmd Edd a Sad mddd a Edd md a 3. mm azmmww
Sod a Red dmdd « Edd owed a Bed 95d .4 ddmd god a mood ad a 3. d4 58%“me
Rod H Ed Edd H mdmd owed a Bed dddd « ddmd dd H 3. «A eds.
Sod H dBd Rod 4 add. Rod a mmmd Edd a «Ed Edd a ommd B a 04 mm Mmhmw
:dd 4 Sod dmdd 1" Rod. Rod 4 Ndmd dmdd n mvmd dddd H mdmd ed a we 9. “mm“
owed « dedd- Edd H Edd Bod a dvmd Edd n tmd ed u 3. dd 88:2.
m m. f .. m: .. o: ..< z cowwwmmm

 

.3355ch E 86065 Em woum :30 .fit 80:859. :m_o-:_£_>> cam Am; $96999 BEBE
éIV €823.92 E3365 .AmI new OIV onmogNBoHoc vogooaxo new 320on *3 mo_o__m do 52:3: 9: 5.. mm
P H :moE mm omucomoa 9m mo:_m> .mcozmSQoa Ems. new __mwonE< 5928 £296 ozocom .6 comtmanO Nd 25m...

68

05'

- Amboseli
Masai Mara

04

9
w

Frequency

 

 

o . i . .
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Allele frequency class

Figure 3.2. Allele frequency distribution for Amboseli (solid bars) and Mara
(hatched bars) populations. Distributions for both populations are “L-shaped" as
expected in the absence of a genetic bottleneck.

69

clans (Figure 3.3; Amboseli, n1 = 1041, n2 = 1039 pairs, P < 0.0001; Mara, n1 =
786, n2 = 592 pairs, P < 0.0001 ). Within-clan relatedness differed significantly
between clans in both populations (Amboseli, n1 = 666, n2 = 375, P < 0.0001;
Mara, n1 = 666, n2 = 120 pairs, P < 0.0001), with higher relatedness in the

smaller clans (Table 3.2).

DISCUSSION
Levels of genetic diversity were similar between the Amboseli and Mara
populations, despite the recent demographic bottleneck in Amboseli. Levels of

heterozygosity (HO) reported here for spotted hyenas (0.56 to 0.60) are on the

low end of the range of values reported for robust populations of other large
carnivores. Mean heterozygosity was estimated as 0.61 to 0.79 for brown bears
(Ursus actos, Paetkau et al. 1998), 0.42 to 0.56 for cougars (Puma concolor,
Anderson et al. 2004 and references therein), 0.77 for leopards (Panthera
pardus, Spong et al. 2000), 0.75 for lions (Panthera Ieo, Spong et al. 2002), and
0.53 to 0.71 for wolves (Canis lupus, Roy et al. 1994; Weckworth et al. 2005;
Pilot et al. 2006).

In addition, we failed to detect genetic evidence of a population bottleneck
in the Amboseli population. However, Cornuet and Luikart’s (1996) test for
heterozygote excess assumes no migration and no population substructure.
Migration could introduce rare alleles into a population, masking any

heterozygote excess caused by a bottleneck (Cornuet 8 Luikart 1996). Although

70

0.101
1041

+ 786
0.05 - r +

 

 

 

a)

:5

E

0': 0.001 y

a)

.‘2

E

IS.“ '0-05' 1039 ' 5:2

-0.‘IO

Within Between Within Between
Amboseli Masai Mara

Figure 3.3. Pairwise R—values among natal animals within the same clan and
between different clans for the Amboseli and Mara populations. Sample sizes

indicate number of R-values. Values are presented as X :1: SE. In both
populations differences within and between clans were significant (P < 0.05).

71

human development surrounding some areas of the park might limit gene flow
into Amboseli, we do not believe that this population is isolated. Indeed, spotted
hyenas are capable of moving over large distances (Hofer 8 East 19930) and
almost all male disperse from their natal clan at adulthood (Mills 1990; East 8
Hofer 2001; Boydston et al. 2005; Hdner et al. 2007). Therefore, it is likely that
the Amboseli population receives migrants, and that this might hide any genetic
signature of a bottleneck. A number of empirical studies of other species have
also failed to detect genetic signatures of a bottleneck, despite known
demographic bottlenecks and in some cases even despite significant losses in
genetic diversity (Whitehouse 8 Harley 2001; Aspi et al. 2006; Hailer et al. 2006;
Brown et al. 2007).

The maintenance of genetic variation despite a demographic bottleneck
has been reported in numerous other studies (see Introduction), and a variety of
factors have been suggested to explain these observations. First, species with
long generation times should be better able to maintain high levels of genetic
variation during a bottleneck, than shorter-lived species (Kuo 8 Janzen 2004;
Hailer et al. 2006). Second, population census size may underestimate effective
population size, which would lead to an overestimate of the severity of a
bottleneck. Populations that are not completely isolated may maintain enough
gene flow (increasing effective population size) to retain a large degree of genetic
variation (Burns et al. 2004). Several studies have shown that just a few migrants
into a small population may be sufficient to maintain or restore the level of

genetic variation (Keller et al. 2001; Vila et al. 2003; Hogg et al. 2006). Moreover,

72

 

I"—""'"‘ W

theoretical models developed by Chesser and colleagues (reviewed by Sugg et
al. 1996) demonstrate that socially-structured populations (e.g. non-random
mating, sex-biased dispersal), such as those of many gregarious species, have
larger effective population sizes and can maintain greater genetic diversity than
same-size populations lacking social structure. Finally, strong selection for
genetic variation may act to stem the loss of genetic diversity and counteract the
effects of genetic drift (Kaeuffer et al. 2007).

We estimated a generation time of 5.7 years for spotted hyenas (based on
the life table in Chapter 2, Table 2.1 of this dissertation). If we assume that the
Amboseli population size was greatly reduced in size for approximately 20 years,
this would correspond with 3.5 generations. This suggests that the Amboseli
bottleneck was not severe in duration with respect to generation time in this
species.

While the potential for selection to maintain or promote genetic diversity in
spotted hyenas remains unknown, it is likely that migration into the Amboseli
population has occurred. The trend towards more negative IR values in
Amboseli, which indicate more outbred individuals, is consistent with migration
into the population. Such migration may have lead to a larger effective population
size during the bottleneck than was indicated by census population estimates,
contributing to the maintenance of genetic variation. Further, the social
structuring of hyena populations by clan structure, female philopatry, and multiple
paternity (Engh et al. 2002) should have also facilitated the maintenance of

genetic diversity (Chesser 1991; Sugg 8 Chesser 1994). Compared to some

73

other carnivores, spotted hyenas may be better able to disperse across potential
barriers, including areas of human development, due to their behavioral plasticity.
Hyenas can utilize a wide range of habitats and food resources, and have been
found to quickly modify their behavior in response to human activity (Mills 8
Hofer 1998; Boydston et al. 2003b; Hayward 2006; Kolowski 2007). Such
plasticity may make spotted hyena populations highly resilient in the face of
changing landscapes, particularly relative to the resilience shown by other
mammalian carnivores.

Not only were patterns of genetic diversity similar in Amboseli and the
Mara, but so too were patterns of relatedness. This similarity in relatedness
structure between populations indicates that the current Amboseli population is
not recently descended from a group of closely related individuals, such as a
single matriline (Van Horn et al. 2004). Our microsatellite based estimates of
relatedness were only slightly below expectation for known mother-offspring
pairs, as in a previous study of Mara hyenas (Van Horn et al. 2004). As found by
Van Horn et al. (2004), relatedness was higher among clan-mates than among
animals born in different clans. The magnitude of our Amboseli and Mara R-
values (within-clan R = 0.05 and 0.07), however, was considerably greater than
that of those estimated in the previous study (within-clan R = 0.01). Although we
did find that within-clan relatedness differed significantly among clans, all values
were at least moderately higher than those found by Van Horn et al. (2004). It is
unlikely that this difference is due to biases in estimates of R between studies, as

values for mother-offspring pairs were very similar. Our within-clan estimates of

74

 

R suggest that the opportunity for kin selection to shape group-level cooperation
in spotted hyenas may be greater than previously thought, though we emphasize
that average within-clan relatedness is still quite low.

Clan size may be an important influence on within-clan relatedness in
spotted hyenas. In both our study populations, relatedness was higher in the
smaller clan than in the larger clan. While data from additional clans would be
needed for a rigorous test of this hypothesis in spotted hyenas, this relationship
between group size and relatedness has also been found in other gregarious
mammals (Spong et al. 2002; Lukas et al. 2005).

Populations that have experienced demographic bottlenecks are
vulnerable to genetic effects, demographic stochasticity, and/or environmental
stochasticity (Soulé 8 Mills 1998). In order to successfully manage wild
populations in smaller and increasingly fragmented habitats, it is essential to
develop approaches to minimize the adverse effects of potential bottlenecks.
Examination of historical bottlenecks to better understand the variation in
potential outcomes and the factors that may influence those outcomes should
provide valuable insight necessary to develop conservation strategies. In the
case of the spotted hyena described here, our results suggest that the behavioral
plasticity and the predominance of male dispersal in this species may make it
less vulnerable to losses of genetic diversity. Further, our findings emphasize the
importance of management efforts aimed at facilitating natural dispersal patterns,

for the conservation of spotted hyena populations.

75

CHAPTER 4

INTERSPECIFIC COMPETITION INFLUENCES REPRODUCTION IN
SPOTTED HYENAS

INTRODUCTION

Understanding sources of intraspecific variation in fitness is critical to
elucidating both population dynamics and evolutionary processes. For long-lived
species such as many large mammals, however, obtaining sufficient data to
examine variation in fitness is difficult. Despite the challenges, studies examining
longitudinal variation within populations have found that environmental variables,
including those associated with food, predators, conspecifics, and climate, can
have strong effects on measures of fitness obtained from large mammals
(Cheney et al. 1988; Clutton-Brock et al. 1988; Packer et al. 1988; Kelly et al.
1998; Forchhammer et al. 2001; Altmann 8 Alberts 2003; Durant et al. 2004). For
example, longitudinal studies have revealed that female reproductive success
and juvenile recruitment in cheetahs (Acinonyxjubatus) are negatively correlated
with the presence of lions (Panthera leo), a competitor and predator (Kelly et al.
1998; Durant et al. 2004), and that reproductive rates among female baboons
(Papio cynocephalus) decline as group size increases (Altmann 8 Alberts 2003).

An alternative approach to examining longitudinal variation within a single
population is to examine variation among populations (e.g. Hanby et al. 1995).
Studies comparing populations usually offer the advantage of capturing a wider
range of environmental variability (both qualitative and quantitative) than is

typically experienced by any single population. However, comparative studies of

76

multiple populations simultaneously are often expensive and logistically
challenging. Furthermore, when performed by multiple investigators, studies of
different populations often involve use of contrasting methods, and this
compromises performance of many types of analysis.

Clearly, longitudinal and cross-sectional comparative studies represent
two complementary approaches, each offering different advantages and
disadvantages. Here we compare cross-sectional data from two populations of
spotted hyenas (Crocuta crocuta) in southern Kenya; these data were collected
using identical methods developed for longitudinal monitoring of this species. Our
goal in comparing these spotted hyena populations was to examine sources of
variation in fitness in this large gregarious carnivore. Spotted hyenas occupy a
wide range of habitats, including tropical savanna, swamps, montane forests and
harsh deserts (e.g., Kruuk 1972; Mills 1990; Sillero-Zubiri 8 Gottelli 1992); they
also exhibit considerable behavioral variation among these habitats. Spotted
hyenas live in social groups, called clans, that vary in size among populations
from 8 to 90 individuals (Kruuk 1972; Mills 1990). Whereas spotted hyenas in all
populations occupy group territories, the permanence of these territories, the
vigor of territorial defense, and the degree to which hyenas forage outside of their
territories all vary greatly among populations (Mills 1990; Hofer 8 East 1993b;
Boydston et al. 2003a; Trinkel et al. 2004).

Life history parameters also appear to vary among populations. All spotted
hyena cubs reside in dens for the first months of life, but the age at which cubs

become independent from the den varies among populations from 9 to 15

77

months (Mills 1990; Hofer 8 East 1993a; Boydston et al. 2005). Similarly,
reported values for the mean age at which cubs are weaned range from 12 to 18
months among populations (Kruuk 1972; Hofer 8 East 1993a; Holekamp et al.
1996). These population differences suggest that environmental variation may
have important influences on fitness in this species.

We initiated the population comparison described in this study because it
offered an opportunity to focus on effects of interspecific competition on
measures of fitness in spotted hyenas. That is, it appeared that the two study
areas differed greatly with respect to lion density; one area supported a large,
stable lion population (Ogutu 8 Dublin 2002; Ogutu et al. 2005), whereas lions
had been extirpated from the other study area just over a decade earlier
(Chardonnet 2002). However, we were also interested in assessing the effects
of other ecological variables that might promote variation in fitness measures
among hyena populations. Our purpose here, then, is to first compare fitness
measures that contribute to lifetime reproductive success between the two study
populations. We then test alternative ecological hypotheses, including
interspecific competition with lions, to explain observed differences between

populations.

METHODS
Study animals
Spotted hyenas are proficient hunters that feed primarily on ungulates

they kill themselves (Kruuk 1972; Holekamp et al. 1997). Ungulate carcasses

78

represent rich, but ephemeral food sources. Consequently, competition among
clan-mates for access to a kill can be intense. Despite competition within a clan,
clan members must cooperate to defend a group territory and the food resources
within it from neighboring hyenas. Moreover, hyenas also compete vigorously
with lions over access to kills (Kruuk 1972). Although lions typically steal
carcasses from hyenas, the reverse is also observed. Hyenas can acquire food
from lions by aggressively displacing lions from a kill (“active scavenging”), as
well as by passively scavenging portions of carcasses discarded by lions. The
outcome of competitive interactions between lions and hyenas depends on the
relative numbers of lions and hyenas, as well as on whether or not adult male
lions are present (Kruuk 1972; Cooper 1991; H6ner et al. 2002). Lions also pose
a threat to hyenas as predators; they are a major source of mortality for spotted
hyenas (Kruuk 1972; Hofer 8 East 1995; Trinkel 8 Kastberger 2005).

Female spotted hyenas breed throughout the year (Lindeque 8 Skinner
1982; Holekamp et al. 1999), bearing litters of one, two, or rarely three cubs.
Whereas males disperse from their natal clans beginning at 2 years of age,
females remain in the natal clan throughout their lives (Boydston et al. 2005).
Young females begin to reproduce between 2.5 and 6 years of age (Holekamp et

al. 1996; Hofer 8 East 2003).
Study populations

The two study populations inhabit Amboseli National Park and the Masai

Mara National Reserve, respectively (Figure 4.1, Table 4.1 ). Amboseli is

79

 

(a)

 

 

 

   
   
   

  

Amboseli
National Park

w.,- Jfl‘v: r", .. ,
r, > ~31 " ..
.‘ I - 1-5... '. .. .‘1';
if," -. .,.,« .
. \f 1...,...,. .- __ ‘./ .I"
I “ 1*";
1., 1‘
r. . . ‘;
Vi“ "u." it"
. {KT-A"
rt ~
4» . , .7 , 7 .
,- “-71,“... .1.‘ ”.3 w. ..
M . M T

National Reserve
. .. . ..mq

 

 

1 01 2I0lomotm
II-ZI

 

 

 

Figure 4.1. The territories of the study clans in (a) the Masai Mara National
Reserve, and (b) Amboseli National Park. Areas within the park boundaries are
lightly shaded, while clan territories are darkly shaded. The Talek clan territory
covered 61 km2, while the Airstrip and OI Tukai territories covered 28 km2 and 26

km2, respectively. The methods used to delineate territories are described by
Boydston et al. (2001) and Kolowski (2007).

80

Table 4.1. Comparison of the two study areas: Amboseli National Park and the
Masai Mara National Reserve.

 

 

Population: Amboseli Masai Mara
Location 2°40’S, 37°15’E 1°30 S, 35°20’E
Mean monthly 23.0 mm 84.9 mm
rainfall (range) (0 to 98.5 mm) (4 to 231.5 mm)
Mean daily Low: 15.8 °C Low: 13.8 °C
temperature1 High: 31.8 °C High: 28.3 °C
Habitat types - 1) Grassland 1) Grassland
ranked in order of 2) Scrub/Bush 2) Scrub/Bush
abundance2 3) Woodland 3) Woodland

Mean prey density

Mean per capita
prey density
Lion density3
Mean lion pride
size (range)3
Hyena density4

Mean body
condition of adult
female hyenas5

90.5 :I: 13.7 animals/km2

3.5 :I: 0.5
animals/kmzlfemale

0.079 to 0.135 lions/km2

16 lions
(11 — 24 lions)

1.65 hyenas/km2
2.60 i 0.06

234.0 1: 32.1 animals/km2

15.3 :I: 2.3
animals/kmzlfemale

0.439 lions/km2

26 lions
(12 - 48 lions)

0.95 hyenas/km2
2.27 i: 0.06

 

1Mara temperatures are from Kolowski (2007) for 2003.
2Habitat types are from Western (2007) for Amboseli and Kolowski et al. (2007)

for the Masai Mara.

3Data on Mara lion population are from Ogutu 8 Dublin (2002).
4Based on mean clan sizes and territory sizes in Table 4.2.

5Body condition reflects recent food intake, and is scored on a scale from 1
(thinnest) to 4 (fattest). Mara data are from 2003-2005.

81

comprised primarily of semi-arid savanna dotted with permanent swamps. The
area supports large populations of ungulates, including wildebeest
(Connochaetes taurinus), zebra (Equus burchellr), Thomson’s gazelle (Eudorcas
thomsonir), and buffalo (Syncerus caffer), which move in and out of the park
seasonally. Other than from tourist vehicles, the Amboseli hyenas have very little
exposure to anthropogenic activity. Two clans of spotted hyenas in Amboseli
National Park were monitored continuously between July 2003 and July 2005,
encompassing 4 clan-years of observation. The OI Tukai clan contained an
average of 38.7 hyenas and defended a territory of 26.4 km2. The Airstrip clan
contained an average of 51.0 hyenas with a territory of 28.0 km2 (Table 4.2).
The Masai Mara (hereafter Mara) is characterized by rolling savanna
habitat with both permanent and seasonal watercOurses. It is grazed year round
by large concentrations of ungulates including Thomson’s gazelle, topi
(Damiscilus lunatus) and impala (Aepyceros melampus), which are joined by
large migratory herds of wildebeest and zebra for 3 to 4 months each year,
between June and October. Spotted hyenas in the Mara have been the subject of
long-term study for over two decades (e.g. Frank et al. 1995a). Here we
examined data collected between July 1988 and June 1992 on the Talek clan,
which contained an average of 56.6 hyenas during this period, and defended a
territory of 61 km2 (Table 4.2). As for Amboseli, the Mara study period
encompassed 4 clan-years of observation, though on a single clan. Since the

mid-1990’s, increasing anthropogenic disturbance in the Talek study area has

82

Table 4.2. Comparison of spotted hyena study clans in Amboseli National Park
and the Masai Mara National Reserve.

 

 

 

 

Population: Amboseli Masai Mara
Clans: Airstrip OI Tukai Talek
Territory size 28.0 km2 26.4 km2 61.0 km2
Mean clan size 51. 0 38.7 56.6
(range) (42 — 64) (32 — 48) (48 — 68)
Mean # adult 13.1 :1: 0.2 12.7 :1: 0.3 16.7 :t 0.5
females

Mean # immigrant 9.5 :1: 0.4 5.6 :1: 0.2 10.1 :I: 0.3
males

Mean # juveniles 19.7 :I: 0.4 15.4 :I: 0.9 23.9 :I: 0.8

 

83

lead to reduced observability of these hyenas and altered their use of space and
temporal patterning of activity (Boydston et al. 2003b; Kolowski et al. 2007).
Specifically, in addition to tour vehicles, Talek hyenas are now exposed daily to
intensive human activity associated with livestock grazing. Consequently,
although Talek hyenas were also being observed intensively from 2003 to 2005,
we considered the 1988-1992 Mara study period to be most appropriate for

comparison with Amboseli.

Behavioral observations

All methods of data collection were exactly the same between populations,
unless otherwise noted. Individual hyenas were identified by unique spot
patterns, and sexed on the basis of penile morphology (Frank et al. 1990).
Observations were made during two daily data-collection periods, one between
0530 — 0900 hours and one between 1700 - 2000 hours. During a data-collection
period, the study area was searched by observers in vehicles, and an
observation session was initiated each time one or more hyenas was located.
Observation sessions lasted from 5 minutes to several hours and ended when
observers left that individual or group. During all observation sessions,
aggressive interactions (described by Holekamp 8 Smale 1990) were recorded

using all-occurrence sampling (Altmann 1974).

84

Body condition

At each observation session, we scored the body condition of all adult
females on a scale from 1 (thinnest) to 4 (fattest). This measure reflects recent
food intake by hyenas and is similar to belly scores used for other carnivores
(cheetahs, Caro 1994; lions, Pusey 8 Packer 1994). We then calculated mean
scores for each adult female that was present for at least 1 year of the study
(Amboseli, n = 19 females; Mara, n = 12 females), based on her body condition
at four time points each year of the study. That is, each female was assigned a
body condition score in the first session during which she was observed each
January, April, July, and October of the study. Since body conditions were not
scored systematically in the Mara between 1988 and 1992, we used data

collected concurrently with the Amboseli study (2003-2005) for comparison here.

Fitness measures

Den sites were visited regularly by observers to monitor births and the
development of cubs. The ages of all hyenas born during the study were
estimated to within 17 days based on size, pelage and other aspects of
appearance and behavior of cubs when they were first observed above ground
(Holekamp et al. 1996). Cubs were typically first observed above ground at
approximately one month of age. Litter size was also determined when cubs first
appeared above ground, based on the number of cubs observed. We made no
assumptions about mortality occurring before cubs first appeared above ground;

however, some females kept their cubs so well hidden that they could not be

85

observed until the cubs were older than 3 months of age, and we did not assign
litter sizes for such litters. Weaning ages were determined based on the
occurrence of weaning conflicts and cessation of nursing, as described by
Holekamp et al. (1996). Inter-birth interval was the period in months between
consecutive parturitions when at least one member of the first litter survived to
weaning. For some females, data were available for multiple inter-birth intervals
(range: 1 to 3 intervals per female). Therefore, a mean inter-birth interval was
calculated for each female. The inter-birth interval is composed of both the period
of lactation, and the recovery period between the end of lactation and onset of
the next pregnancy. Here the recovery period was defined as the time between
weaning of one litter and conception of the subsequent litter. We determined
conception dates by subtracting the gestation period (110 days) from the birth
date of a litter. For females for which data were available on multiple recovery
periods (range: 1 to 3 periods per female), a mean value was calculated. We
quantified survivorship for the first year of life, the period during which mortality is
highest in this species (Frank et al. 1995a; Hofer 8 East 1995). All cubs that
disappeared before 1 year of age were considered to have died.

We estimated lifetime reproductive success for females from each
population as: L-(n/I)-S, where L is reproductive lifespan, n is litter size, i is inter-
birth interval, and Sis offspring survival to 1 year. Both male and female offspring
were included in this calculation. Based on long-term data from the Mara
population, mean lifespan for females that reach reproductive maturity was 8.2

years (n = 76 females). Since the average age at which females first reproduce is

86

3.5 years (Holekamp et al. 1996; Hofer 8 East 2003), we subtracted 3.5 from 8.2
to let L = 4.7 years for both populations. For all other parameters, we used the

mean value obtained for each population.

Prey abundance

In order to monitor the abundance of prey species in both study areas,
biweekly counts were conducted of ungulate species along road transects
distributed throughout each clan territory. In the Mara, censuses were conducted
along two 4 km transects in the Talek territory. In Amboseli, we employed 4
transects covering a total of 7 km in the DI Tukai territory and 3 transects
covering a total of 7 km in the Airstrip territory. An additional 3 km transect was
added in the Airstrip territory in May 2004. Transects were run between 0800 and
1100 hours, counting all ungulate species within 100 m of each transect line.
These counts were then used to estimate mean monthly prey densities in each

study area.

Intraspeciflc competition within clans

Following Holekamp et al. (1993), we assessed feeding competition with
conspecifics at ungulate kills using two methods. First, we calculated the mean
number of hyenas at kills in each study area. Second, we calculated rates of
aggression at kills, expressed as the number of aggressive interactions per
individual hyena present per hour observed. We also used an additional measure

to scale prey abundance to the number of hyenas in each study area. That is, for

87

each population we calculated monthly per capita prey density as the number of
prey animals per adult female hyena per km2. These values were then used to
generate a monthly average for the entire study period.

Hyenas prey upon species ranging in size from < 1kg (e.g. birds and
invertebrates) to over 500 kg (e.g. giraffe and buffalo), and carcass size might
potentially influence competitive interactions among feeding hyenas. Therefore,
we examined whether there was any difference in the size of prey that hyenas
consumed in the two populations. Each time one or more hyenas were observed
consuming portions of a fresh carcass (i.e., killed within 6 hours and excluding
cases in which the prey animal was known to have died of causes other than
hyena predation) we identified the species being consumed (details in Cooper et
al. 1999). Biomass estimates for each prey species were taken from Cooper et
al. (1999) and Kruuk (1972). Although this method is likely to underestimate the
occurrence of small prey items (Cooper et al. 1999), data from both populations
should be similarly biased. Data from the Mara population were taken from
Cooper et al. (1999, Table 1), which includes data for June 1988 — May 1995.
Cases in which the prey species could not be identified were omitted (n = 21

Amboseli, n = 51 Mara cases omitted).

lntraspecific competition between clans
While most competitive interactions occur between clan-mates, hyenas
also engage in direct competition with hyenas from neighboring clans. Clan wars

are a conspicuous and intense form of inter-clan competition, in which a group of

88

clan members engage in coordinated attacks against a group of neighboring
hyenas (Kruuk 1972; Hofer 8 East 1993b; Boydston et al. 2001). Clan wars most
frequently occur when a kill is made near a territory boundary, with clans fighting
over possession of the kill. Here, we compared the frequency of clan wars in the
two populations. In both populations, study clans had multiple neighboring clans

with which clans wars could potentially occur.

lnterspecific competition with lions

Between July 2003 and July 2005, we searched daily for lions within
Amboseli National Park. Efforts to locate lions were greatest in and around
territories of the two study clans, but also extended throughout the park.
Individual lions were identified using whisker patterns and other unique markings
(Pennycuick 8 Rudnai 1970). Lion cubs were aged and categorized as 0-1 year
or >1 year following Schaller (1972). We identified a total of 61 lions in Amboseli
during the study. Most lions belonged to one of three resident prides (pride size:
11 to 24 lions), but there were also several nomadic males. Lion density was
estimated using the total number of identified lions in the park (total area 390
kmz) during six-month intervals, thus accounting for births, deaths, and dispersal
beyond the park.

Ogutu and Dublin (2002) used the same approach to identify lions within
the Masai Mara National Reserve concurrent with our study. Between
September 1990 and July 1991, they identified 348 lions within the 792 km2

region of the reserve in which our study area is located. They determined that

89

these lions belonged to 12 prides with an average pride size of 26 lions. We used
their estimates of lion density for comparison with Amboseli.

We also examined two measures of interspecific interference competition
over food in each population. First, we quantified the frequency with which hyena
kills were stolen by lions. Fresh carcasses (or portions thereof) were considered
to be hyena kills if hyenas were in possession of the carcass when first observed
and the prey animal was not known to have died of a cause other than hyena
predation. Cases in which lions were in possession of the carcass when first
observed, but in which it was clear based on wounds to the prey animal and/or
the distribution of blood on the Iion(s) and hyena(s) that the kill had been made
by hyenas, were also considered to be hyena kills. A kill was categorized as
stolen if lions gained possession of the kill and proceeded to feed, even if the
hyenas later regained possession of the carcass remains. Second, we quantified
the frequency with which hyenas scavenged fresh carcasses (or portions thereof)
from lions. All fresh carcasses in the possession of lions, at which hyenas were
also present, were considered potential scavenging opportunities for hyenas.
Only carcasses scavenged by hyenas, from which they could have obtained
significant nutrition or energy (i.e., scraps of skin or small quantities of bone were
excluded), were counted as successfully scavenged. Carcasses obtained from

both active and passive scavenging were included here.

90

Statistical analyses

Survivorship in the first year of life was compared between populations
using Gehan’s generalized Wilcoxon test, which is designed for use with survival
data. Other comparisons between populations were made using t-tests, and with
Mann-Whitney U-tests when the assumptions of the parametric test were not
met. The Kolmogorov-Smirnov two-sample test was used instead of the Mann-
Whitney U-test when there were many tied ranks. Statistical tests were
performed in Statistica 6.1 (StatSoft 2002). Mean values are presented 1: 1

standard error (SE) throughout.

RESULTS
Fitness Measures

Litter size did not differ significantly between populations (D = 0.08, n1 =

53, n2 = 55, P > 0.10), although litters on average were slightly larger in the

Amboseli sample (f = 1.68 :1: 0.08 cubs) than in the Mara sample (Y = 1.56 i:
0.07 cubs). Litter sizes may be slightly elevated in Amboseli due to the
occurrence of triplet litters (n = 4, 7.5% all litters), which were not observed
during the study period in the Mara (Figure 4.2). When singleton and twin litters
were examined, litter composition was not significantly different between

populations (Pearson x2: x21 = 0.006, P = 0.94). We could not compare litter

composition statistically when triplet litters were included because the expected

frequencies were too small.

91

 

 

nnnnnn
FFFFF

92

Cubs were weaned at significantly younger ages in Amboseli than in the
Mara (Figure 4.3a; U = 273.0, n1 = 25, n2 = 38, P = 0.005). Mean age at weaning
in Amboseli was 10.8 (1: 0.33) months, compared with 12.5 (t 0.54) in the Mara.

Inter-birth intervals were significantly shorter in Amboseli than in the Mara
(Figure 4.3b; t= -2.27, n1 = 13, n2 = 14, P = 0.03). Mean inter-birth interval was
15.7 (:I: 1.0) months in Amboseli, and 19.3 (:1: 1.3) months in the Mara. The
difference in inter-birth intervals between populations was due to differences in
the length of the lactation period (Figure 4.3a), as there was no difference in the
recovery period between populations (U = 68.0, n1 = 11, n2 = 14, P = 0.62).

Survivorship of cubs to 1 year of age did not differ significantly between
the two populations (Gehan’s Wilcoxon test = 1.39, n1 = 51, n2 = 60, P = 0.16).
The proportion of cubs surviving to 1 year was 0.59 in Amboseli, compared with
0.63 in the Mara.

Estimated lifetime reproductive success for females was 3.6 offspring in
Amboseli and 2.9 offspring in the Mara. Thus, lifetime reproductive success was

estimated to be 24% higher for females in Amboseli than in the Mara.

Ecological predictors

If these observed population differences in fitness measures were due to
variation in prey abundance, we expected that prey density would be higher on
average in Amboseli than in the Mara. In contrast, however, mean monthly prey
density was significantly greater in the Mara than in Amboseli (Table 4.1, Figure

4.4; U= 193.0, n1 = 24, n2 = 43, P < 0.0001).

93

.3 ._\ ..L
C) m o N A

Age at weaning (mo)

A

 

 

Amboseli Mara
24 -

N
O

—L
a)

Inter-birth interval (mo)
00 B

 

 

Amboseli Mara

Figure 4.3. Mean 1- SE (a) age at weaning for cubs and (b) inter-birth interval for
spotted hyena populations in Amboseli National Park and the Masai Mara
National Reserve. Sample sizes are indicated above each bar, and represent the
number of cubs (a) and females (b). Asterisks represent significant differences (P
< 0.05).

94

700 '

—O— Amboseli
A -o- Mara
6001
NE ” ‘
x I ‘,
To 500- ~,
.5. x
C ,’ \‘
£3 400 0 x
2‘ ‘1
"2
<0 300 -
'0
>
(D
‘5, 200 -
C
8
5 100

 

 

 

J

Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun
Month

Figure 4.4. Mean monthly prey density within the territories of spotted hyena
clans in Amboseli National Park and the Masai Mara National Reserve. Values
represent averages over multiple years.

95

If fitness differences between populations were due to varying degrees of
intraspecific interference competition between clans, then between-clan
competition should be greater in the Mara than in Amboseli. Contrary to this
prediction, clan wars occurred more frequently in Amboseli, with 2.5 clan wars
observed per year, than in the Mara, where only 0.8 clan wars were observed per
year.

The observed differences in fitness measures between populations may
have been due to variation in intraspecific competition within clans, irrespective
of differences in prey density. If this is the case, then within-clan competition
should be greater in the Mara than in Amboseli. As predicted by this hypothesis,
the number of hyenas at kills was larger in the Mara than in Amboseli (Figure

4.5a; U = 16657.0, n1 = 150, n2 = 292 kills, P < 0.0001), and the individual
biomass of prey consumed by hyenas was smaller in the Mara than in Amboseli
(Figure 4.6; D = 0.33, n1 = 119, n2 = 798, P = 0.001). The mean biomass of prey
was 187 i 9 kg in Amboseli compared with 133 :l: 3 kg in the Mara. Further, the

mean body condition of adult females was significantly better (i.e. fatter) in

Amboseli than in the Mara (Table 4.1; t= -3.82, n1 = 19, n2 = 12, P = 0.0007),

indicating greater food intake in Amboseli. Contrary to other predictions of this
hypothesis, however, aggression rates at kills were actually higher in Amboseli

than in the Mara (Figure 4.5b; U = 11418.0, n1 = 100, n2 = 264 kills, P = 0.047),

and per capita prey density was significantly higher in the Mara than in Amboseli

(Table 4.1; u = -5.59, n1 = 24, n2 = 42, P < 0.0001).

96

Number of hyenas

 

 

Amboseli Mara

2.5 '(b)

2.0- 100

Aggressions/hyena/hr

0.5

 

 

0.0

Amboseli Mara

Figure 4.5. Mean :l: SE (a) number of spotted hyenas competing at kills and (b)
aggression rates at kills in Amboseli National Park and the Masai Mara National
Reserve. Sample sizes are indicated above each bar, and represent the number
of kills. Asterisks represent significant differences (P < 0.05).

97

    

 

% % % o/ % / o/

.
0
0

0

0000000
666666

251 -500

201 -250

Biomass of individual prey (kg)

51 -200

0
cm
0

am

mew

Ywa

heD

S.

W%bc
filmw
cafe
pnop.
sorS)
Value.”
.03ng
kmnpk
% m}
fiasmK
Sfiawm
aMwala
d a(\./
SMDLQ
mteaQ
UUUUU
snoeflx
nabr.
O)ae.ll
C 0.3
yBdot
emmoe
r acr
pemmm
(Awfldnuoo
S=|:.H®
wnmn
(aem
mk .KO
rrrrr
dPfimm
6' aa
tmbas
aopre
.mfimam
flaaMm
ENSIfi.
..IH .88
emMae
4ofiMs
ebkes
.md 7Tb
F.m__:mb

98

Finally, if population differences in fitness measures were due to varying
intensities of interspecific competition with lions, then such competition should be
greater in the Mara than in Amboseli. Lion density in the Mara was 0.439
lions/km2 for lions of all ages, and 0.352 lions/km2 for lions >1 year of age
(Ogutu 8 Dublin 2002). By contrast, Amboseli lion density ranged from only
0.079 to 0.135 lions/km2 for lions of all ages. For lions >1 year of age, lion
density ranged from 0.061 to 0.094 lions/kmz. Thus, lion density was 3 to 6 times
greater in the Mara than in Amboseli. Our finding that body condition was
significantly better in Amboseli (Table 4.1) than in the Mara is also consistent
with lower levels of interspecific competition in Amboseli. Although the frequency
at which lions stole hyena kills did not differ between populations (Figure 4.7;

Yates corrected x2: x21 = 2.04, P = 0.15), hyenas were able to scavenge fresh

carcasses from lions more frequently in Amboseli than in the Mara (Figure 4.7;

Yates corrected x2: x21 = 7.18, P = 0.007). The enhanced ability of Amboseli

hyenas to scavenge from lions is also consistent with the hypothesis that
population differences in fitness are due to varying intensities of competition with

lions.

DISCUSSION
Female spotted hyenas in Amboseli reproduced at significantly higher
rates (i.e., had shorter inter-birth intervals) than did females in the Mara. This
effect appeared to be due largely to shorter lactation periods in Amboseli, where

cubs weaned almost 2 months earlier, on average. We did not find evidence of

99

70% '

11- W Mara
- Amboseli

60% . 16
50% '
40%

30%

Percent occurrence

20%~ ’

10% *

 

 

0% , V///////////////////%
Scavenged from lions Stolen by lions

Figure 4.7. The frequency with which spotted hyenas were successfully able to
scavenge fresh carcasses from lions and at which spotted hyenas lost their kills
to lions in Amboseli National Park (n = 27 scavenging opportunities, n = 142
hyena kills) and the Masai Mara National Reserve (n = 29 scavenging
opportunities, n = 323 hyena kills). Numbers of carcasses are indicated above
bars. The asterisk represents a significant difference (P < 0.05)

100

any trade-offs for this higher reproductive rate in terms of smaller litter sizes,
longer recovery periods, or reduced offspring survival. It is possible that females
in Amboseli have shorter reproductive Iifespans than Mara females, but we do
not currently have sufficient data for such a comparison and with fewer mortality
threats in Amboseli (i.e. fewer lions and anthropogenic disturbances) this would
certainly not be expected. If reproductive lifespan is equivalent in the two
populations, as we assume, then we estimate that females in Amboseli
experienced 24% greater lifetime reproductive success than Mara females.
However, if generation time is shorter in Amboseli than in the Mara, as
suggested by population differences in inter-birth interval, then this measure
actually underestimates fitness differences between the two populations. In
general, our assessment that lifetime reproductive success is greater in Amboseli
than the Mara is consistent with the recent demographic histories of these two
populations. Whereas hyena population size in the Mara has been stable (Watts
8 Holekamp In review), the Amboseli population has undergone a dramatic
expansion since the late 1980’s (Faith 8 Behrensmeyer 2006).

Although litter sizes did not differ significantly between populations, the
frequency of triplet litters in Amboseli was remarkable. No triplet litters were
observed in the Mara during the study period examined here, and in over 15
years of observation in the Mara only 2 triplet litters (1.2% of all litters) have been
observed (Wahaj et al. 2007). Further, in the Mara all cubs born in triplet litters
died before reaching 4 months of age (Wahaj et al. 2007), whereas in Amboseli,

more than 40% of triplet cubs survived beyond weaning (H. Watts 8 K. Holekamp

101

unpublished data). As with the Mara population, triplet litters have been reported
only rarely or not at all in other field studies (0 of 20 litters, Mills 1990; 1 of 32
litters, Golla et al. 1999; 5 of 735 litters, Hofer 8 East 2007). In contrast, the
frequency of triplet litters in Amboseli was comparable to that observed in
captivity, where hyenas are well-fed and 12% of litters are triplets (Wahaj et al.
2007). We hypothesize that the conditions permitting enhanced reproductive
rates in Amboseli may also support the production of triplet litters.

Our results did not support the hypothesis that differences in fitness
between populations were due to differences in prey abundance. Contrary to the
prediction of this hypothesis, prey density was actually higher in the Mara than in
Amboseli, yet the Amboseli population exhibited higher reproductive rates.
Although prey density exhibited greater seasonal fluctuations in the Mara, it
remained higher than in Amboseli in almost all months of the year (Figure 4.4).
It is important to note that our findings do not negate the potential importance of
prey abundance with respect to understanding differences in both behavior and
fitness among other spotted hyena populations. For example, patterns of prey
abundance may explain variation in land-use strategies observed among hyena
populations (Trinkel et al. 2006).

We also failed to find support for the hypothesis that population
differences in fitness measures were due to varying levels of intraspecific
competition between clans. Clan Wars actually occurred more frequently in
Amboseli than in the Mara, suggesting that interference competition between

clans is more intense in Amboseli. Although the Amboseli population has

102

undergone a recent expansion, land adjacent to the OI Tukai and Airstrip clans
was occupied by neighboring clans throughout the study period (H. Watts 8 K.
Holekamp unpublished data). Consequently, it appears that all suitable habitat
was occupied by hyenas in Amboseli as it was also known to be in the Mara
(Masai Mara: Holekamp et al. 1993; Boydston et al. 2001).

Our results did not provide strong support for the hypothesis that
population differences were due to reduced intraspecific competition within-clans
in Amboseli. Large antelope like buffalo were consumed relatively often in
Amboseli, where hyenas rarely fed on small antelope like Thomson’s gazelles
that are such a large part of the diet of Mara hyenas (Cooper et al. 1999).
Therefore, kills were of larger mean biomass in Amboseli than in the Mara.
Furthermore kills in Amboseli were attended by fewer hyenas than in the Mara,
and body condition was better in Amboseli than in the Mara. These three
findings are consistent with an hypothesis invoking intra-clan competition to
explain apparent differences in fitness between clans. However, per capita food
abundance was lower in Amboseli than the Mara, overall hyena density was
nearly twice as high in Amboseli as the Mara (Table 4.1), and rates of aggression
at kills were higher in Amboseli than the Mara; these measures indicate that
intraspecific competition was more intense in Amboseli than in the Mara.

Our results were most strongly consistent with the hypothesis that fitness
differences between Amboseli and Mara hyenas were due to differences in
interspecific competition with lions. Lion density was considerably higher in the

Mara than in Amboseli. Furthermore, our estimates of lion and hyena densities in

103

both populations indicate that the lionzhyena ratio was 1:2 in the Mara and 1:12
(or more) in Amboseli. Based on these ratios, we would expect that hyenas in the
Mara would have been more likely to lose kills to lions and less likely to gain kills
from lions, whereas hyenas in Amboseli may have been better able to defend
kills from lions and even to acquire kills made by lions. Although, we found no
difference in the rate at which hyenas lost kills to lions in the two populations, we
did find that Amboseli hyenas were better able to acquire food from lions. The
enhanced ability of Amboseli hyenas to scavenge from lions could be due not
only to more successful active scavenging, but may also reflect differences in the
ability of hyenas to scavenge passively. Higher lion density and larger pride sizes
in the Mara may mean that lions in this population consume a larger proportion of
each carcass before abandoning it, thereby leaving less to be consumed by
scavenging hyenas.

The fact that hyenas in Amboseli lost their kills to lions with the same
frequency as hyenas in the Mara, despite living at lower lion density, may reflect
differences between the populations in prey biomass. As Amboseli hyenas
typically preyed upon larger species than Mara hyenas, kills in Amboseli likely
took longer to consume and would therefore have provided a greater opportunity
for detection and acquisition by lions. Similarly, larger kills may also facilitate
detection and recruitment of clan-mates by neighboring clans for clan wars. This
could explain the more frequent occurrence of clan wars in Amboseli relative to

the Mara.

104

Lions can potentially influence hyena populations through direct killing,
competition for food, and/or behavioral effects (i.e. inducing costly antipredator
behavior). There was no evidence for inter-population differences in juvenile
mortality, suggesting that increased lion density did not result in higher rates of
predation on juvenile hyenas. However, differences between populations in the
ability of hyenas to scavenge from lions, suggest that lions influenced spotted
hyena fitness via competition for food resources. Further, higher body condition
scores in Amboseli than in the Mara indicate that food intake was higher in
Amboseli despite availability of more prey in the Mara. These body condition
scores are based on recent data from the Talek clan, which continues to live with
high lion density (Ogutu et al. 2005), though clan size is now slightly smaller
(range: 47-55, Kolowski et al. 2007) than it was during the 1988-1992 study
period. Assuming that recent body condition reflects body condition during the
1988-1992 study period, these results are also consistent with an effect of lions
on hyenas via competition for food. Finally, behavioral effects of lions on hyena
populations have yet to be documented. Most research in this area has focused
on avoidance of lions by spotted hyenas, which does not appear to be common
in these or other populations (Mills 8 Gorman 1997; Boydston et al. 2003b;
Chapter 5 of this dissertation).

There are at least two reasons, why we expect the effects of competition
with lions to be particularly important in these two hyena populations. First, prey
densities in the Mara and Amboseli are relatively high compared to other hyena

populations (Trinkel et al. 2006). Creel (2001) has pointed out that carcasses are

105

a more valuable resource than live prey among carnivores, and, therefore, that
competition is likely to be most intense for carcasses. Moreover, Creel suggests
that competition over carcasses may be further intensified at high prey densities
due to higher corresponding carnivore densities. Second, the intensity of
interspecific competition among carnivores is expected to be influenced by the
ability of competitors to detect kills (Creel 2001). Both Amboseli and the Mara are
characterized by open habitat, which allows the detection of kills over long
distances. Together, high prey density and open habitat set the stage for
interspecific competition among carnivores to have considerable impacts on
these populations.

These results add to findings from other studies indicating that competition
with lions is an important influence shaping spotted hyena populations. In
Tanzania’s Ngorongoro Crater for example, Honer et al. (2005) suggest that
reduced competition with lions, in conjunction with high prey abundance, led to a
recent increase in the hyena population. Although the results from the current
study indicate that lions influence hyena populations by affecting rates of
reproduction, other studies have implicated lions in hyena survival. Lions were
the leading source of mortality in 4 of the 5 hyena populations examined to date,
accounting for as much as 70% of all hyena deaths (Kruuk 1972; Mills 1990;
Trinkel 8 Kastberger 2005; Watts 8 Holekamp In review). Moreover, a
longitudinal analysis of over 15 years of data from the Mara hyena population
found that both annual birthrates and juvenile survival were negatively correlated

with lion competition, assessed by rates of all lion-hyena interactions (Watts 8

106

Holekamp In review). Having demonstrated an influence of lions on spotted
hyenas population dynamics, we predict that competition with lions may also be a

strong selective pressure shaping the behavior of spotted hyenas.

107

CHAPTER 5

RESPONSES OF SPOTTED HYENAS TO LIONS REFLECT CONTEXT-
SPECIFIC INDIVIDUAL DIFFERENCES IN BEHAVIOR

INTRODUCTION

lnterspecific competition plays a pivotal role in the structuring and
dynamics of communities. Competition among mammalian carnivores can be
especially intense, and such competition may have cascading effects on lower
trophic levels (Palomares et al. 1995; Rogers 8 Caro 1998; Crooks 8 Soulé
1999; Elmhagen 8 Rushton 2007). Not only do carnivores compete for food, but
they also have both morphological and behavioral adaptations for killing, leading
to widespread intraguild predation (Palomares 8 Caro 1999; Creel et al. 2001).
Consequently, carnivore populations can be adversely affected by competition
with dominant species within their local guild (Laurenson 1995; Creel 8 Creel
1996; White 8 Garrott 1997; Crooks 8 Soulé 1999; Vucetich 8 Creel 1999;
Tannerfeldt et al. 2002; Helldin et al. 2006).

One strategy employed by numerous carnivores to mitigate the
deleterious consequences of interference competition involves shifting patterns
of space use to avoid dominant species (Creel et al. 2001, and references
therein). However, avoidance behavior may impose significant energetic costs;
for instance, the subordinate species may be relegated to areas with lower prey
abundance (Creel 8 Creel 2002). Not surprisingly, therefore, there are also cases
in which spatial avoidance is not observed among carnivore species that
compete intensively (White et al. 1994). The absence of avoidance behavior may

be due to high costs of avoidance, use of alternative coping behaviors to

108

minimize adverse effects of a dominant competitor, or both. Whereas studies
examining interspecific competition often focus on the average behavioral
responses to competitors, there may also be considerable variation in responses
both within and among individuals. Responses to competitors should theoretically
be influenced by the relative costs and benefits of a potential interaction, as well
as by aspects of an individual’s “personality”. Animal personalities, or “behavioral
syndromes”, have been found to influence patterns of behavior across a wide
range of taxa (reviewed in Gosling 2001; Sih et al. 2004b), but field studies
remain rare, and to our knowledge there is only one such study in a wild
carnivores (Heinsohn et al. 1996).

In light of these considerations, here we examined aspects of interspecific
avoidance between spotted hyenas (Crocuta crocuta) and lions (Panthera Ieo),
the two most common large carnivores occurring throughout most of Africa.
These two species are vigorous competitors, and they have a high degree of
dietary overlap (Hayward 2006). Lions, the larger and more powerful competitors,
frequently steal ungulate kills from spotted hyenas (Kruuk 1972; Schaller 1972),
and they also represent a major source of hyena mortality (Kruuk 1972; Hofer 8
East 1995). However, the adverse effects of lions on spotted hyenas may be
mitigated, in part, by the ability of hyenas to scavenge food from lions. Spotted
hyenas are highly efficient scavengers, as well as hunters, and they can utilize
parts of carcasses that lions discard, including bones. Further, if spotted hyenas
gather in sufficient numbers, they are capable of stealing kills from lions (Kruuk

1972; Cooper 1991). Indeed, competitive interactions between lions and hyenas

109

are quite complex, and the outcome of any particular interaction is influenced by
several factors including the relative numbers of lions and hyenas, whether an
adult male lion is present, and the motivational state characterizing members of
both species at the time of the interaction (Kruuk 1972; Cooper 1991).

Lion and hyena densities are positively correlated at study sites across
Africa (Creel 8 Creel 1996; Creel 8 Creel 2002), suggesting a lack of avoidance
behavior over large geographic scales. This pattern is thought to reflect the
positive association of both lions and hyenas with resident prey species (Creel 8
Creel 2002), and suggests that the benefits to hyenas of living in close proximity
to prey generally outweigh the costs of living with lions. However, analyses over
such large scales cannot detect adjustments in space use that individual hyenas
might make on smaller spatial scales in response to lions.

Because interactions with lions may have both costs (i.e., loss of food,
injury, or death) and benefits (i.e., scavenging opportunities) for individual
hyenas, we conducted playback experiments on free-living spotted hyenas to
determine whether individuals use avoidance to minimize potentially costly
encounters with lions. Hyenas were observed to vary in their behavioral response
to the perceived presence of a lion. Therefore, we also inquired whether this
variation reflected flexible behavioral adjustments based on the trade-offs
associated with a potential lion encounter, or consistent individual differences in
behavior (i.e. behavioral syndromes).

We evaluated three factors likely to influence the potential costs and

benefits to an individual hyena of interacting with one or more lions: sex of the

110

lion, local lion density, and hyena social rank. If hyenas adjust their behavioral
responses to lions based on trade-offs between costs and benefits, then we
expected one or more of these factors to influence the response of hyenas to lion
playbacks. Adult male lions pose a greater physical threat to hyenas than do
females or juveniles (Cooper 1991). Further, hyenas are rarely able to
aggressively displace male lions from food, even when they are greatly
outnumbered by the hyenas (Cooper 1991; H6ner et al. 2002). Consequently,
hyenas should respond more strongly to male than female lions.

We expected local lion density to influence both hyena vigilance and
avoidance behavior. Vigilance behavior is typically influenced by predation risk
(e.g. Bednekoff 8 Ritter 1994; Burger 8 Gochfeld 1994; Hunter 8 Skinner 1998;
Laundre et al. 2001), and avoidance behavior can‘only be effective if an animal
has the ability to move to an area where it is less likely to encounter the dominant
species. Therefore, we expected hyenas to react more strongly to lion playbacks
when they are in areas of high lion density than in areas of low lion density.

Social rank within the linear dominance hierarchy of a hyena clan
determines an individual’s priority of access to food resources. Consequently,
lower—ranking individuals have fewer feeding opportunities and are likely to
experience greater energetic stress than higher-ranking animals (Tilson 8
Hamilton 1984; Frank 1986b; Hofer 8 East 1993a). As such, potential
scavenging opportunities should be more valuable to lower-ranking individuals,

and the costs of avoiding lions greater, than to high-ranking hyenas. Therefore,

111

we expected that low-ranking hyenas would be less likely to avoid the perceived
presence of a lion than would high-ranking hyenas.

Individual differences in behavior might also influence responses to lions.
Therefore, we assessed whether responses of hyenas to lion playbacks reflected
consistent individual differences in behavior by comparing an individual’s
behavior during the playback experiment with its behavior in the presence of
lions. Consistent patterns of behavior by individual hyenas in both situations
would indicate an effect of behavioral syndromes. We also inquired whether such
syndromes were context-specific to lion interactions or more generally expressed
in other contexts as well. Because distinct personality traits have been observed
among captive spotted hyenas, including boldness, assertiveness, fearfulness,
vigilance, and nervousness (Gosling 1998), it seemed reasonable to expect that
behavioral syndromes in free-living spotted hyenas might influence how

individuals behave in response to lions.

METHODS
Study Population 8 Behavioral Observation
The study was conducted in Amboseli National Park, in southern Kenya.
This is a semi-arid area, dominated by open savannah habitat (Altmann et al.
2002; Western 2007). The area supports large herbivore populations, which
migrate in and out of the park seasonally. We monitored two clans of spotted
hyenas and the sympatric lion population continuously between July 2003 and

July 2005. All hyenas were identified individually by their unique spot patterns,

112

and were sexed based on the morphology of the erect phallus (Frank et al.
1990). All lions were individually identified using whisker patterns and other
unique markings (Pennycuick 8 Rudnai 1970). The study area was searched
daily between 0530-0900 h and between 1700-2000 h by observers in a vehicle.
An observation session was initiated each time one or more hyenas or lions were
located. The geographic location of each observation session was recorded
using a GPS unit. Upon initiation of an observation session, and every 15
minutes thereafter, scan sampling (Altmann 1974) was used to record the identity
and activity of each hyena or lion present, and presence and type of any food
item. When lion and hyenas were present at the same session, scans were
conducted every 10 minutes, and the distance of each hyena to the nearest lion
was recorded. At the start of an observation session the body condition of each
hyena was scored on a scale from 1 (thinnest) to 4 (fattest); similar methods
have been used to assess body condition in both cheetahs and lions (Caro 1994;
Pusey 8 Packer 1994). Observation sessions lasted from 5 minutes to several
hours, and ended when observers left that individual or group.

Agonistic interactions between hyenas were recorded using all-occurrence
sampling at observation sessions (Altmann 1974). To determine individual social
ranks, a linear dominance hierarchy was constructed for each study clan based
on outcomes of all dyadic agonistic interactions observed among adult clan
members, following the methods of Holekamp and Smale (1990). Individuals
were then assigned to one of three rank categories (high, mid, or low),

determined by dividing the dominance hierarchy of each clan into equal thirds.

113

Playback Experiments

In order to quantify the response of spotted hyenas to the perceived
presence of a lion, audio-playback experiments were conducted between
December 2004 and July 2005. We obtained 14 high-quality lion recordings from
the Borror Laboratory of Bioacoustics (Ohio State University) and Dr. Craig
Packer (University of Minnesota). Each recording consisted of one bout of
roaring from a single adult lion (7 males, 5 females) and 2 recordings of 2 adult
lions roaring (one bout per individual). All recordings were obtained from lion
populations outside the study area, and therefore were considered to be
uniformly unknown to the focal hyenas. A roar bout was broadcast to a focal
hyena using a Creative Nomad Jukebox 3 connected to Fender Passport P-150
loudspeakers concealed in a vehicle 80 m to 130 m (Mean 104.9 1: 2.4 m SE)
from the hyena. Sound pressures of recordings were matched to natural levels
for playbacks. The speakers were separated by 1 m and oriented at
approximately 135 degrees to the hyena on the opposite side of the vehicle such
that they were not in view. Playbacks were not performed in windy conditions that
might interfere with sound transmission. Recordings were selected randomly for
each playback, except that a given lion recording was played to no more than
two focal hyenas. Playbacks were performed around dawn and dusk (0630 —
0900 hrs and 1730 — 1900 hrs, respectively) when lions are naturally heard to

roar. The focal hyena was in open habitat, isolated from all other individuals by at

114

least 400 m and was more than 400 m from any active den. The geographic
location of the playback experiment was recorded using a GPS unit.

At each playback, one observer filmed the focal hyena using a Sony DCR-
TRV33 digital video camera for 3 minutes before sound onset and for at least 3
minutes after sound onset. If the focal hyena was still reacting to the playback
after 3 minutes, filming continued until the hyena ceased responding, the hyena
went out of sight, or 1 hour had elapsed since sound onset (in the case of one
playback), whichever came first. A hyena was defined as no longer responding to
the playback when it resumed its pre-playback behavior (in most cases this was
resting with head down) or was no longer regularly orienting and scanning. A
second observer determined the focal hyena’s distance to the speakers using a
Bushnell Yardage Pro Sport laser rangefinder at the start of the playback
experiment, and subsequently every 5 - 10 seconds. All hyenas were resting
when a playback was initiated due to constraints of the experimental setup and to
ensure uniformity across experiments. To limit variation in response among
hyenas due to hunger state, playbacks were only performed when focal hyenas
were scored in the two most common and intermediate body conditions (scores 2
and 3).

For a control playback treatment we used 14 high-quality recordings of
baboon contest ‘wahoo’ vocalizations emitted by adult male baboons, obtained
from Dr. Anne Engh (University of Pennsylvania). Contest wahoos were chosen
as control sounds because they are similar in sound frequencies and loudness to

lion roars, but should be neither threatening nor attractive to hyenas. Wahoo

115

recordings were selected for similarity in regard to approximate duration and
general sound structure to the lion recordings. Six of the wahoo recordings were
edited (repeated to create a longer wahoo bout, cut short, and/or extraneous
vocalizations by other baboons removed) using Praat software (Boersma 8
Weenink 2004) in order to create recordings more similar in duration to the lion
recordings. Each wahoo recording was used only once for a playback. Control
playbacks were conducted in the same fashion as the lion playbacks.

Lion playbacks were performed on 21 adult hyenas (13 females, 8 males).
Fourteen of these focal hyenas were also the subject of a control playback (8
females, 6 males). All male subjects were immigrant males who had been
resident in their clan for at least 6 months. To prevent habituation, paired
playbacks to focal hyenas were separated by at least 3 weeks, except for one
hyena for which playbacks were separated by 2 weeks. Mean (1: SE) duration
between playbacks was 62.8 :1: 15.1 days. The frequency with which playbacks
were conducted was well below the naturally occurring rate at which either lion
roars or baboon wahoos are heard in the study area. In two playback
experiments the focal hyena was presented with a recording of two lions roaring.
Since we could not detect any difference in response between these playbacks
and those with a single lion roar, we lumped them for all analyses, except when

examining effects of lion sex.

116

Analysis of Playback Experiments

In the field, all playbacks were conducted and filmed by H.E.W.
Videotapes were then scored at Michigan State University by two different
trained observers (L.M.B. 8 S.E.D.) who were not present when playbacks were
conducted. For each playback, scores for each behavioral measure were
averaged over the two observers except response duration and distance moved
after sound onset, which were scored separately by H.E.W., based on
observations at the time of the experiments. lnterobserver reliability was high
(Pearson correlation: mean r = 0.96, range 0.89 to 1.00) for all measures scored
from videotapes.

From the videotape of each playback the following behavioral measures
were extracted and recorded separately for the 3 minutes before and after sound
onset: rate of visual scanning of the surrounding area, proportion of time spent
orienting toward the hidden speakers, and proportion of time spent traveling. To
calculate the rate of scanning, we defined a scan as a movement of the subject’s
face of at least 90 degrees from the midsagittal plane. A hyena was scored as
‘orienting’ when its head was elevated off the ground and it was facing in the
direction of the speakers. Although hyenas would briefly scan their surroundings,
most vigilance behavior exhibited by hyenas after playbacks consisted of
orienting towards the speakers. Additionally, we calculated the following
measures after sound onset: latency to orient, latency to travel, response

duration, and distance moved. Distance moved was measured as the change in

117

distance (relative to the speakers) from the focal hyena’s position at sound onset
to its position 3 minutes later. For analyses examining variation in responses to
lion playbacks, we also recorded the distance moved between sound onset and
the time at which the hyena ceased responding to the playback. Hyenas that
failed to orient or travel were automatically assigned a response latency of 3
minutes. Response duration was the time from the hyena’s initial response to
when the hyena was no longer responding to the playback or the hyena went out
of sight. If the hyena showed no response to the playback, response duration
was set at 0 seconds. Overall movement in response to the playback was
categorized as either: 1) approaching the sound source (i.e. the speakers), 2)
avoiding the sound source by moving away from it, either immediately or after a
brief initial movement towards it, 3) neutral movement laterally which neither
increased nor decreased the focal hyena’s distance from the sound, or 4) no
movement. Neutral movement was not observed in response to any playback
and this category was therefore discarded. Sample sizes vary slightly (as noted
in Results) for some measures, because they could not be quantified for all
playbacks.

For all measures collected before and after sound onset, the difference in
behavior between these two 3-mintue periods was used for analyses. That is, the
value before sound onset was treated as the baseline for each measures and

subtracted from the value after sound onset.

118

Quantifying Local Lion Density

Territory boundaries were identified for each study clan based on both
territorial behavior and observed locations of adult female clan members.
Following Kolowski (2007), an initial Minimum Convex Polygon (MCP) was
created for each clan based on recorded locations of dens and territorial
behaviors. Territorial behaviors included aggressive encounters with neighboring
clans, defecation at latrine sites, and border patrols (described in Kruuk 1972;
Henschel 8 Skinner 1991; Boydston et al. 2001). Next, a 95% fixed-kernel home
range contour (Worton 1989; Powell 2000) was created for each clan using
observed locations for adult female clan members between July 2003 and July
2005 (Airstrip clan, N = 632 location; OI Tukai clan, N = 1269 locations),
excluding locations at dens. Each MCP was then expanded to match the 95%
contour, where the contour extended beyond the MCP. The final clan territories
were 28.0 km2 and 26.4 km2 in size, for Airstrip and OI Tukai clans, respectively
(see Figure 5.1).

The geographic locations of all lions 18-24 months of age or older
observed in the study area were recorded between July 2003 and July 2005.
These observations (N = 668 locations) were used to generate a fixed-kernel
utilization distribution (UD). Rather than estimating lion population density per se,
this method provided an estimate of intensity of use by local lions for each 1 km x
1 km grid cell in our study area, based on the observed density of lion locations
(Figure 5.1). Grid cell values ranged from 1 (highest density) to 100 (lowest

density). Since the mean lion density values differed between the two clan

119

Figure 5.1. Local lion density in the two study clan territories, based on utilization
distribution. Darker cells indicate areas of higher lion density. Amboseli National
Park boundary is indicated by the thick line. The two clan territory boundaries are
indicated by thin lines. Locations of each lion playback experiment (N = 21) are
shown as open circles.

 

>77 '— A T T —T———T_T —— ~' “ a_' —7
'

I g Amboseli National Park
I N ‘ i‘,

I , o O \x) // TN... ._

i , ./ \'\v '1. .,

: z’ mm: - .

I ‘1 0 ” "\ Territory "

l I O /' ‘\ 5* . ~ I,

I _ .1 J" \= v! ,1; JV). "\\\x‘

l D j? i ~ 0 ,I ‘ I _ x. "OlTukai‘7
' '5'? ' I V/ .... Territory

' P ' ' fl. 0 ‘ \T — l"

. 4:5 x

 

 

120

territories (3?: SE = 94.7 : 6.7 and 76.6 1; 27.6 for Airstrip and OI Tukai clans,
respectively), lion use was categorized separately for each clan. Playback
locations that occurred in grid cell with values less than the territory mean were
categorized as high local density (smaller values indicate higher density), while
those with values greater than the mean were categorized as low density (Figure
5.1).

To avoid temporal autocorrelation in sighting locations for lions and
hyenas, repeated locations for any given individual of either species were
separated by at least 6 hours, and no more than 2 locations per individual per
day were included. A period of 6 hours was conservatively selected to allow
sufficient time for either species to cross their entire territory (White 8 Garrott
1990). ArcView GIS 3.2 (Environmental Systems Research Institute, Redlands,
CA) and the Animal Movement Analyst (Hooge 8 Eichenlaub 2000) program
extension were used to perform spatial analyses. Smoothing parameters for UDs
were based on ad hoc calculations. The Grid Tools program extension (Jenness

2006) was used to calculate mean grid cell values for each clan territory.

Quantifying Behavioral Syndromes

We identified two candidate behavioral dimensions that we expected to
influence responses to lions, and that we could quantify in our study subjects:
risk-taking and vigilance. We predicted a hyena’s inclination to approach or avoid
in response to a played back lion roar would be influenced by its risk-taking

tendencies. Similarly, we expected that the amount of time a hyena spent

121

orienting towards a played back lion roar would reflect how its tendency to be
vigilant varied more generally. It is relevant to note that vigilance does not
necessarily reflect fearfulness; in captive hyenas vigilance and fearfulness reflect
different components of personality (Gosling 1998).

We assessed the risk-taking and vigilance tendencies of each focal hyena
outside of the experimental playback situations using scan sampling data
collected during behavioral observation sessions in which no playback
experiment was conducted. We quantified vigilance behavior in three contexts
(sensu Sih et al. 2004a): when lions were present in the same observation
session as the focal hyena , when the focal animal was present at communal
dens, and a “baseline” context which included all sessions away from dens, when
no lions, alien hyenas or food were present. These three contexts were selected
because sufficient behavioral observations were available for each to perform
statistical analyses, and because behavior in these contexts should not be
unduly influenced by social rank. We considered the playback experiment and
observation sessions when lions were present to represent two different
situations (sensu Sih et al. 2004a) within the broader context of interactions with
lions.

For observation sessions, vigilance was quantified as the proportion of
scans by observers in which a hyena was engaged in vigilance behavior.
Vigilance behavior was defined as being stationary in any body position with
head elevated. When lions were present, other behaviors that involved alert

orientation towards a lion, such as approaching a lion or vocalizing in response to

122

a lion, were also included as vigilance behaviors. To quantify behavior at
communal dens, only sessions at active communal dens were included where no
lions or food were present. Risk-taking behavior was quantified at sessions with
lions present as the mean distance of a particular hyena to the nearest lion. Risk-
taking could only be quantified in the context of interacting with lions, since we
lacked an appropriate behavioral measure of risk-taking in the other contexts.
Only individuals with a minimum of 10 scans at sessions where lions were

present, and 20 scans at den and “baseline” sessions, were included for analysis

(Y3: 1SE scans per individual: 32 1: 3 at lion sessions, 112 1: 18 at den sessions,
56 :1: 4 at “baseline” sessions). Vigilance data other than during sound playbacks
were collected from May 1St 2004 to July 11th 2005, whereas data on distance to

lions were collected between September 1St 2003 and July 11th 2005.

Statistical Analyses

Responses to lion and control playbacks were compared for those (N =
14) subjects receiving both treatments using paired ttests. As necessary, some
behavioral measures were log transformed to normality before analysis.
Measures that could not be transformed to normality were analyzed using
Wilcoxon signed-ranks tests. All tests were two-tailed (d = 0.05).

To examine variation in responses to lion playbacks in greater detail, a
subset of behavioral measures were selected for further analysis. MANOVA was
used to assess the effect of the three predictor variables (lion sex, local lion

density, and hyena social rank) on hyena response to lion roars. Three measures

123

of response to playback were used in the MANOVA: proportion of time oriented,
scan rate, response duration. Logistic regression was used to assess effects of
the predictor variables on avoidance behavior, using a binomial response
variable (avoidance or no avoidance). Relationships between behavior in
response to playbacks and behavior in the presence of lions, as well as in other
contexts, were analyzed using correlation and multinomial logistic regression. For
this analysis, movement in response to playbacks was treated as a multinomial
response variable (approached, no movement, avoided). STATISTICA 6.1
(StatSoft 2002) was used for statistical analyses. Differences between groups
were considered significant when P < 0.05. Mean values are presented 1: 1
standard error.
RESULTS

Lion and Control Playbacks

Following playbacks of lion roars, only 8 of 21 focal hyenas had ceased
responding by 3 minutes after sound onset (Figure 5.2). By contrast, 13 of 14
focal hyenas had ceased responding by 3 minutes after sound onset during
control playbacks. Thus response patterns differed significantly between

experimental and control playbacks (Yates corrected x21 = 8.34, P = 0.004)

After a lion roar, hyenas moved further from the speakers, spent more
time oriented towards the sound (Figure 5.3a), and continued responding longer
than after a control sound (Table 5.1). There was also a trend for hyenas to
visually scan the surrounding area more frequently after a lion roar than after a

control sound (Figure 5.3b). There was no difference in the proportion of time

124

 

.
I
.
U
.
i
-
..
,.
1
a
4 'o
5..
O.)
o
.7
>5
7‘,
.. ,,
-7.
1
J"
v ‘9‘
-
‘30
.-
l -.
,L
. .
Q
_.
..
_.
..
.,
.-
'!
O.
..‘.
-.
..
. J’
1,5,
..
fic-
7.1?
‘
1“.
no
..
u .‘g
< ”C-
.3,
Q.
U-
..‘-
:F-
.I"
. av
_.
t
.

- Lion
[:I Control

Number of occurrences

 

 

 

6 9 12 15 30 45 60
Duration of response (min)

Figure 5.2. Duration of response to lion (N = 21, filled bars) and control (N = 14,
open bars) playbacks.

125

0.6 -

(a)

0.5-

 

I——I

0.4 ~
0.3 1

0.2 ~ _L

0.1-

 

Proportion of time oriented

 

 

 

 

 

 

 

0.0 . .1
Lion Control

1.2 - (b) T
1.0-

 

0.8 —

 

0.6 -

0.4 , .L

0.2 I

 

Scans per minute

 

 

 

 

 

0.0

 

 

Lion Control

Figure 5.3. (a) The proportion of time hyenas oriented towards concealed
speakers, and (b) the rate at which they scanned the surrounding area, before
(filled bars) and after (open bars) sound onset for paired lion and control
playbacks (N = 14 hyenas). Mean values :1: 1 SE are presented.

126

Table 5.1. Comparison of responses to lion and control playbacks.

 

Variable :Y— |ion :f contro| N Test P
statistic

 

Response duration (8)3.d 268.2 28.9 14 t= 4.92 P 0.0003

Proportion of time 0.36 0.19 14 t= 2,24 b 0.043
oriented a

Distance at 3 min (m) 12.2 -4.4 14 T = 4 c 0.049
Scan rate (scans/min) 0.57 0.21 14 t: 2.05 b 0.061

"9663467161"t‘IBRé“”"” .. .. .. _. . - 0. .15.- .. __ ._ .0. .03“ - _ -14AMT;_§6._W..WO_.1..T6.- .._.

traveling

Latency to orient (s)3 5.8 7.5 12 t: .040 b 0.700
Latency to move (s) 150.9 172.4 13 T = 3,5 0 0.281
Rate of movement (m/s) 0.04 0.02 14 T = 14 C 0.575

 

3 Variable was log transformed to normality for statistical analysis

9 Paired t-test

C Wilcoxon signed-ranks test

d Only response duration was quantified beyond the first 3 minutes after sound onset.

127

hyenas spent traveling, the latency to orient, or the latency to move after each
type of playback. There was, however, significantly less variation in latency to

orient after a lion than a control playback (Levene’s test: F124 = 9.73, P = 0.005),

with all hyenas orienting relatively quickly to the lion roar. Hyenas were more
likely to move (either approach or avoid the sound source) after a lion roar than
after a control sound (Figure 5.4. McNemar test for significance changes: x2 =
5.14, P = 0.023). Hyenas were only observed to avoid in response to lion

playbacks, not control playbacks (Figure 5.4).

Individual Differences in Responses to Lion Playbacks

Individual hyenas varied greatly in their responses to lion roars. Although
all hyenas oriented towards the roar within 20 seconds of sound onset (X = 6.58

i 1.42 s), response duration ranged from 42.5 seconds to over 1 hour (f = 9.46
:I: 3.05 min). In the first 3 minutes after the playback, hyenas spent as little as
1.8% and as much as 82.3% of their time oriented in the direction of the roar, and
from 0 to 73.6% of their time traveling. Some hyenas were observed to approach
in response to the lion roars, some avoided the sound source, but most did not
move at all (Figure 5.4). The average total distance traveled after a lion roar was
122.4 :1: 73.2 rn, but the maximum was 1.5 km. No hyena was observed to

vocalize during any playback experiment.

128

100 .
- Lion (N = 21)
' E3 Control (N = 14)

80-

 

60:

 

40

20 .
0L .
Approach None Avoid
Movement in response to playback

Percent occurrence of response

 

 

   

 

Figure 5.4. Movement in response to lion (N = 21, filled bars) and control (N =14,
open bars) playbacks. Categories of movement are: approach the sound source,
move away (avoid) from the sound source either immediately or after a brief
initial approach, and no movement (none).

129

Do Responses Reflect Trade-Offs Between Potential Costs and Benefits to
Listeners?
Response measures (scan rate, proportion of time oriented, response

duration) did not vary significantly with lion sex (MANOVA: F3] = 0.358, P =
0.79), local lion density (F3’7 = 0.728, P = 0.57), or the social rank of the hyena
subject (F644 = 0.478, P = 0.81 ), nor were there any significant second order

interactions. Similarly, avoidance after a playback was not predicted by lion sex

(Logistic regression: X2 = 0.04, P = 0.85), local lion density (x2 = 2.96, P =
1 1

0.09), or social rank (X21 = 4.86, P = 0.09).

Do Responses Reflect Behavioral Syndromes in Listeners?

The proportion of time hyenas spent orienting towards the roars after
playback was positively correlated with vigilance behavior in the presence of
lions during daily behavioral observation sessions (Figure 5.5a; Pearson
correlation: r = 0.696, N = 17, P = 0.002), but was not correlated with vigilance
behavior in the absence of lions either at dens (r= 0.045, N = 19, P = 0.86) or
elsewhere (“baseline” context: r = 0.113, N = 21, P = 0.63). A hyena’s movement
in response to lion playbacks was predicted by their risk-taking behavior in the
presence of lions, quantified as mean distance to the nearest lion (Figure 5.5b;

Logistic regression: X22 = 6.49, P = 0.04). Across individuals, risk-taking

(distance to nearest lion) and vigilance behavior in the presence of lions were
correlated (Figure 5.6; Pearson correlation: r = -0.627, N = 17, P = 0.007). Those

hyenas that were most risk-prone were also most vigilant. Neither behavioral

130

_s
o
\

(a)

.0 .0 .0
:1 o: oo
o
o
o
o
0

Proportion of time oriented to playback
S
O
0 Q)

 

¥

0.0 0.2 0.4 0.6 0.8

Low High
Vigilance in presence of lions

.0
O
O
I‘
r

O)
01

' (b)

Risk-averse
01 0)
0'1 0

Mean distance to nearest lion (m)
0'1
0

Risk-taking
A
O

 

 

approach none avoid
Movement in response to playback

Figure 5.5. Relationship between responses of 17 hyenas to experimental
playback of lion roars and their behavior when lions were present with them
during daily observation sessions. (3) Proportion of time spent oriented in
response to playback as a function of the proportion of time spent vigilant in the
presence of lions. (b) Movement in response to playback as a function of risk-
taking during interactions with actual lions, measured as the distance an
individual maintained to the nearest lion.

131

.0
oo

 

 

6»
2 ':E
.9 0 6
“5 ' A
g o
a) “‘N.‘ A
33 0 4
'5. \“A\ o D
.E A A \“59.
o
g A d“‘~-~ U
.12 0.2— o A‘ ~~~~~ \
.9 3
> 0 ~
_J A
0.0 . - . 4 :
30 40 50 60 70
Risk-taking Risk-averse

Mean distance to nearest lion (m)

Figure 5.6. Vigilance and risk-taking tendencies during naturally-occurring
interactions with lions are correlated among individual hyenas. Vigilance was
quantified as the proportion of time spent vigilant when lions were present at
daily observation sessions. Risk-taking was quantified as the mean distance that
individuals maintained to lions during these interactions. Values are shown for
each of 17 hyenas. Symbol shape indicates individuals of high (circle), mid
(square), or low (triangle) social rank.

132

dimension was influenced by social rank (Figure 5.6; ANOVA: vigilance, F214 =

0.629, P = 0.55; distance to lions F2114 = 2.17, P = 0.15).

DISCUSSION

Do Hyenas Avoid Lions Based on Immediate Costs and Benefits?
Spotted hyenas clearly distinguished between the playbacks of lions and those of
baboons. Hyenas responded to lion roars with heightened vigilance levels and
movement away from the sound. Although hyenas did move significantly further
away from the lion roars than from the baboon wahoos, this result may have
been driven largely by a few individuals that exhibited particularly strong lion
avoidance behavior. In examining all the lion playbacks, only 29% of individuals
moved away from the lion roar, while some individuals approached in response.
This indicates that hyenas do avoid lions under some circumstances, but do not
do so systematically. Our failure to find consistently strong avoidance of lions by
hyenas is not surprising in light of the scavenging opportunities lions provide for
hyenas. In contrast, similar audio playback experiments examining avoidance of
lions by cheetahs and wild dogs found much more consistent avoidance behavior
by both species (Durant 2000; Creel 8 Creel 2002). In contrast to spotted
hyenas, neither cheetahs nor wild dogs routinely scavenge from lions (Caro
1994; Creel 8 Creel 2002).

A relationship similar to that between lions and hyenas exists between

wolves (Canis lupus) and coyotes (Canis latrans) in North America. That is,

133

wolves are known to kill coyotes but also represent a source of scavenged food
for coyotes (Switalski 2003; Wilmers et al. 2003). Although individual variation
among coyotes in response to wolves has not been formally studied, patterns of
avoidance behavior by coyotes appear to vary among populations (Fuller 8 Keith
1981; Paquet 1991; Arjo 8 Pletscher 1999). This variation is likely due to
differences in prey availability that influence prospects for scavenging; when prey
are abundant wolves do not consume entire carcasses, thus creating plentiful
scavenging opportunities for coyotes, and the two carnivores overlap in space
(Paquet 1992). It remains to be seen if prey availability influences patterns of lion
avoidance by hyenas. However, spotted hyenas and lions have a greater degree
of dietary overlap than do coyotes and wolves (Thurber et al. 1992; Hayward
2006), and this similarity in diet may constrain the ability of hyenas to avoid lions

under any circumstances.

Individual Differences in Responses to Lion Playbacks

The high degree of variation among individual hyenas in their responses
to lion roars was remarkable. Hyenas varied not only in regard to whether or not
they avoided lion roars, but also in the overall strength of their response. For
example, the degree of vigilance, quantified as the proportion of time a hyena
was oriented to the sound source, varied considerably among individuals.

The variation we observed here in response to lion roars could not be
explained by differences in the potential costs and benefits to an individual hyena

of interacting with a lion, as indicated by lion sex, local lion density, and hyena

134

social rank. Rather, observed differences in responses to lion roars were
consistent with context-specific behavioral syndromes. Both risk-taking and
vigilance tendencies were consistent between playback experiments and
occasions when the same hyenas monitored in playback experiments were
actually found in the presence of lions. Based on the behavior of individuals at
communal dens and elsewhere (in the “baseline” context), these syndromes
appear to be specific to the context of interactions with lions. However, other
contexts in which risk-taking and vigilance may be important, such as feeding
with groups of conspecifics, have yet to be explored. Whereas a number of other
studies have similarly found syndromes to be context-specific (e.g. Coleman 8
Wilson 1998; Reale et al. 2000), others have found strong correlations across
contexts (e.g. Hessing et al. 1993; Malmkvist 8 Hansen 2002; Johnson 8 Sih
2005)

Interestingly, risk-taking and vigilance with respect to lions appear to be
strongly correlated in free-living spotted hyenas. The absence of individuals
expressing a risk-taking and non-vigilant phenotype, as well as individuals with a
risk-averse and vigilant phenotype, may occur because these are both
maladaptive phenotypes. We suggest that non-vigilant, risk-taking individuals are
likely to suffer high costs such as injury and death, whereas vigilant and risk-
averse individuals are likely to reap few benefits from lions while also engaging in
costly and potentially unnecessary vigilance and/or avoidance behavior. Thus,
the observed continuum of risk-taking - vigilance phenotypes (Figure 5.6) may

represent multiple strategies with comparable outcomes for hyenas. Risk-takers

135

'must be vigilant; they potentially gain windfalls by scavenging from lions but must
expend energy to carefully monitor their surroundings. Meanwhile, non-risk
takers can afford to be less vigilant, trading reduced probability of obtaining food
via scavenging for reduced vigilance costs.

The potential costs and benefits to a hyena of interacting with a lion are
likely to vary based on a number of factors. Environmental factors such as the
relative densities of hyenas and lions, local prey availability, and the age and sex
composition of the local lion population, as well as factors intrinsic to a particular
hyena, including its age, nutritional state, and reproductive status, may all
influence perceived costs and benefits. Consequently, the trade-offs involved in
such interactions are likely to vary through time. Although our results suggest
that hyenas do not adjust their behavior in response to these tradeoffs, it is also
possible that the tradeoffs examined in this experiment did not vary sufficiently in
magnitude to influence the responses of hyenas. Whereas this experiment
focused on the responses of lone hyenas to a single set of lion roars, future
experiments might manipulate the relative numbers of hyenas and lions, provide
additional acoustic cues indicating whether or not roaring lions have made a kill,
or test hyenas when they are energetically stressed. Such additional experiments
will be necessary to make a more definitive assessment of whether hyenas
adjust their behavior with respect to lions as costs and benefits vary. It certainly
seems reasonable to expect that, within groups of hyenas sharing a particular
behavioral syndrome, individuals should also adjust their behavior based on

tradeoffs.

136

We have yet to determine the extent to which individual differences in
behavior of spotted hyenas are genetically based and/or influenced by
experience. Studies from other species have found a genetic component to such
individual differences (Boissy 1995; Bouchard 8 Loehlin 2001; Drent et al. 2003;
van Oers et al. 2004), though experience can also be important (Stamps 2003).

In spotted hyenas, selection is likely to favor risk-taking behavior when the
costs of interacting with lions are low and the benefits are high. For example,
when a lion population has a heavily female biased sex-ratio (e.g. Cooper 1991),
risk-taking hyenas are likely to benefit nutritionally from their willingness to
engage in contests with lions over food. In contrast, when the costs are high, as
when many male lions are present in the local population, risk-averse behavior
may increase hyena survivorship and therefore be favored. If these individual
differences in behavior are genetically based, and the trade-offs associated with
lion interactions vary in time, this could potentially generate fluctuating selection
pressures that would maintain the observed diversity in behavioral types

(Dingemanse et al. 2004).

137

APPENDICES

138

APPENDIX A

139

Table A1. Body mass and weaning age of carnivore species. Data are taken
from Gittleman (1986) unless otherwise noted.

 

Adult body Weaning

 

140

Species mass (kg) age (mo) References
Canidae
Canis latrans 1 0.6 3.2
Canis lupus 33.1 1 .1
Canis mesomelas 7.7 2.0
Cerdocyon thous 6 3.0
Lycaon pictus 22 2.5
Otocyon mega/otis 3.9 3.4
Vulpes lagopus 3.2 0.7
Vulpes vulpes 4.1 1.8
Vulpes zerda 1.5 2.2
F elidae
Acinonyx jubatus 58.8 3.6
Caracal caracal 1 1.59 4.0
Fells chaus 6.65 3.3
Fells silvestris 4.67 2.8
Leopardus geoffroyi 2.2 2.1
Leopardus pardalis 11.88 . 6.0 Murray 8 Gardner
(1997)
Lynx lynx 19.3 3.7
Lynx rufus 6.2 2.0
Panthera leo 155.8 7.0 Pusey 8 Packer (1994)
Panthera onca 86.2 3.8
Panthera pardus 52.4 4.6
Panthera tigris 161 5.4
Pn'onailurus 5.5 0.8
bengalensis
Pn'onailurus viverrinus 8.8 1.7
Herpestidae
Cynictus penicillata 0.06 1.4
Herpestes javincus 0.78 1.0
Sun'cata suricata 0.73 1.8
Hyaenidae
Crocuta crocuta 52 12.8
Hyaena hyaena 26.8 11.0 Kruuk (1976)
Parahyaena brunnea 43.9 11.8
Proteles cristata 8.34 3.5 Koehler 8 Richardson

(1990)

Table A1 (cont’d)

 

Adult body Weaning

 

Species mass (kg) age (mo) References
Mustelidae
Enhydra lutn’s 28.3 6.0 Estes (1980)
Monson et al. (2000)
Gulo gulo 11.6 2.3
lctonyx striatus 0.77 1.8
Lontra canadensis 8.2 3.0
Lutra lutra 8.9 3.7
Lutrogale perspicillata 8.78 4.1
Martes americana 0.87 1.5
Martes pennanti 3.75 1.5
Martes zibellina 1 .18 1 .6
Meles meles 11.6 3.1
Mustela altaica 0.19 1.8
Mustela frenata 2.33 1.0
Mustela lutreola 0.59 2.3
Mustela nivalis 0.08 1.0
Mustela slbirica 0.57 1.8
Poecilogale albinucha 0.3 2.5
Taxidea taxus 4.05 1.4
Procyonidae
Bassariscus astutus 0.87 3.9
Procyon Iotor 6.4 3.9
Ursidae
Ailuropoda 1 35 5.9
melanoleuca
Ursus americanus 110.5 5.5
Ursus arctos 298.5 23.9
Ursus maritimus 365 28.0 DeMaster 8 Stirling
(1981)
Ursus thibetanus 103.8 3.9
Viverridae .
Arctictis binturong 13 1.8
Civettictis clvetta 12.02 4.6 Ray (1995)
Genetta genetta 1.9 4.0 Lariviere 8 Calzada
(2001)
Genetta tigrina 2.06 1.8
Hemigalus derbyanus 0.83 2.3

 

141

APPENDIX B

142

 

m m m m m m P m m m

F

v v v. v m v v V v v

P _. F _.
mN 0N mN 9N

F F F r F
m. 9N mN 0N mN 0N
9N 0N 9N mN m. mN 9N 9N 0N 9N

m m m m m m m m m

mN mN ON ON 9N 0N 0N 9N 0N 9N

h :00 0 :00 m :00 3.00 m :00 N200 _._LOO ROCCO VEOO _.OEOO

<20
0920

of
owmcoEzoa
cm...

.oEtd
n_._.z“U

<wm

cotzm

 

..1 06 BE: :6. .<20 .88 53> moEEmmocoE 8390 do mm: .8 982628 cozomo. mun. fim 038.

143

 

 

8.96 ds o8 9: 6F 0.8 68 8 .2 0.8 .86 me .2 0.8 Eco
8.96 ds 56 F 6F 0.8 .56 F 6.. 0.8 .56 F .2 0.8 200
8.96 ds o8 me 6.. 0.8 .86 11.4 6F 0.8 68 me .6 0.8 250
8.96 cs 86 me .6 0.8 .86 me .6 0.8 .86 8 6.. 08 «too
866 d9 86 8 6F 0.3 .56 N 6F 0.8 .86 9. 6F 0.8 m :8
8.96 d: ooo 8 6F 0.8 .86 me .9 0.8 8.0. m... .9 0.8 ”8.96 NV 86 8 6F 0.8 68 8 6.. 08 ~50
8.96 ds 56 F .9 0.8 .56 F 6F 0.8 .56 F .9 0.3 :50
8.96 di 56 F .6 0.8 .56 F 6F 0.8 .56 F 6F 0.8 850
8.96 ds 56 F 6.. 0.8 .56 F 6F 0.8 .56 F .2 0.3 350
8.96 d5 56 F 6.. 0.8 .56 F 6F 0.8 .56 F 6.. 0.8 5.8
:_E N .5». 0:5 um £25.53ch _NEE 91:30:06— mm_o>0 ..mEtn—

 

.(ZD _momh— 2:; w¢u___®umw0._0_e 925080 50 Ow: ..Oh m®_0>0 10n— .N.@ 9th

144

Table 8.3. Characteristics of microsatellite loci used to type spotted hyenas from
Amboseli National Park and the Masai Mara Game Reserve. The locus-specific
number of alleles, observed frequency of heterozygotes (Ho). and the expected

frequency of heterozygotes (HE) are presented for each population.

 

Amboseli population

 

Masai Mara population

 

 

 

Locus Alleles HO HE Alleles HO HE

CCr01 3 0.123 0.118 4 0.437 0.448
CCr04 8 0.797 0.767 7 0.569 0.580
CCr07 4 0.400 0.405 6 0.508 0.488
CCr1 1 2 0.413 0.407 2 0.507 0.452
CCr12 3 0.709 0.637 4 0.714 0.705
CCr13 5 0.475 0.464 4 0.514 0.572
CCr14 7 0.816 0.802 6 0.906 0.830
CCr15 6 0.750 0.690 6 0.809 0.781
CCr16 2 0.455 0.437 2 0.265 0.441
CCr17 6 0.831 0.751 7 0.721 0.700

 

145

 

LITERATURE CITED

Altizer, S., A. Dobson, P. Hosseini, P. Hudson, M. Pascual & P. Rohani (2006)
Seasonality and the dynamics of infectious diseases. Ecology Letters, 9, 467-
484.

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

Altmann, J. & S. C. Alberts (2003) Variability in reproductive success viewed
from a life-history perspective in baboons. American Journal of Human
Biology, 15, 401-409.

Altmann, J., S. C. Alberts, S. A. Altmann & S. B. Roy (2002) Dramatic change in
local climate patterns in the Amboseli Basin, Kenya. African Journal of
Ecology, 40, 248-251.

Amos, W., J. Worthington Wilmer, K. Fullard, T. M. Burg, J. P. Croxall, D. Bloch
& T. Coulson (2001) The influence of parental relatedness on reproductive
success. Proceedings of the Royal Society of London Series B-Biological
Sciences, 268, 2021-2027.

Anderson, C. R., F. G. Lindzey & D. B. McDonald (2004) Genetic structure of
cougar populations across the Wyoming Basin: Metapopulation or
megapopulation. Journal of Mammalogy, 85, 1207-1214.

Arjo, W. M. & D. H. Pletscher (1999) Behavioral responses of coyotes to wolf
recolonization in northwestern Montana. Canadian Journal of Zoology, 77,
1919-1927.

Aspi, J., E. Roininen, M. Ruokonen, l. Kojola & C. Vila (2006) Genetic diversity,
population structure, effective population size and demographic history of the
Finnish wolf population. Molecular Ecology, 15, 1561 -1576.

Bangs, E. E., S. H. Fritts, J. A. Fontaine, D. W. Smith, K. M. Murphy, C. M. Mack
& C. C. Niemeyer (1998) Status of gray wolf restoration in Montana, Idaho,
and Wyoming. Wildlife Society Bulletin, 26, 785-798.

Barton, R. A. (2006) Primate brain evolution: Integrating comparative,
neurophysiological, and ethological data. Evolutionary Anthropology, 15, 224-
236.

Bayes, M. K., K. L. Smith, S. C. Alberts, J. Altmann & M. W. Bruford (2000)
Testing the reliability of microsatellite typing from faecal DNA in the savannah
baboon. Conservation Genetics, 1, 173-176.

Bednekoff, P. A. & R. Ritter (1994) Vigilance in Nxai Pan springbok, Antidorcas
marsupialis. Behaviour, 129, 1-11.

146

Behrensmeyer, A. K. (1993) The bones of Amboseli: the taphonomic record of
ecological change in Amboseli Park, Kenya. Research and Exploration, 9,
402-421.

Beschta, R. L. (2003) Cottonwoods, elk, and wolves in the Lamar Valley of
Yellowstone National Park. Ecological Applications, 13, 1295-1309.

Binder, W. J. & B. Van Valkenburgh (2000) Development of bite strength and
feeding behaviour in juvenile spotted hyenas (Crocuta crocuta). Journal of
Zoology, 252, 273-283.

Blount, J. D., D. C. Houston, A. P. Moller & J. Wright (2003) Do individual
branches of immune defense correlate? A comparative case study of
scavenging and non-scavenging birds. Oikos, 102, 340-350.

Boersma, P. & D. Weenink, (2004) Praat: doing phonetics by computer (Version
4.2.30). http://wwwmraatorqfl

Boertje, R. D. & R. O. Stephenson (1992) Effects of ungulate availability on wolf
reproductive potential in Alaska. Canadian Journal of Zoology. 70, 2441-
2443.

Boissy, A. (1995) Fear and fearfulness in animals. The Quarterly Review of
Biology. 70, 165-191. '

Borer, E. T., B. S. Halpern & E. W. Seabloom (2006) Asymmetry in community
regulation: effects of predators and productivity. Ecology, 87, 2813-2820.

Bouchard, T. J. J. & J. C. Loehlin (2001) Genes, evolution, and personality.
Behavior Genetics, 31, 243-273.

Bouzat, J. (2000) The importance of control populations for the identification and
management of genetic diversity. Genetica, 110, 109-115.

Bouzat, J. L., H. H. Chang, H. A. Lewin, R. L. Westemeier, J. D. Brawn & K. N.
Paige (1998) Genetic evaluation of a demographic bottleneck in the greater
prairie chicken. Conservation Biology, 12, 836-843.

Boydston, E. E., K. M. Kapheim, M. Szykman & K. E. Holekamp (2003a)
Individual variation in space use by female spotted hyenas. Journal of
Mammalogy, 84, 1006-1018.

Boydston, E. E., K. M. Kapheim, R. C. Van Horn, L. Smale & K. E. Holekamp
(2005) Sexually dimorphic patterns of space use throughout ontogeny in the
spotted hyena (Crocuta crocuta). Journal of Zoology, London, 267, 271-281.

147

 

Boydston, E. E., K. M. Kapheim, H. E. Watts, M. Szykman & K. E. Holekamp ‘
(2003b) Altered behavior in a large African carnivore associated with
increased human activity. Animal Conservation, 6, 207-219.

Boydston, E. E., T. L. Morelli & K. E. Holekamp (2001) Sex differences in
territorial behavior exhibited by the spotted hyena (Hyaenidae, Crocuta
crocuta). Ethology, 107, 369-385.

Broquet, T. & E. Petit (2004) Quantifying genotyping errors in noninvasive
population genetics. Molecular Ecology, 13, 3601-3608.

Brown, J. W., P. J. Van Coeverden De Groot, T. P. Birt, G. Seutin, P. T. Boag &
V. L. Friesen (2007) Appraisal of the consequences of the DDT-induced
bottleneck on the level and geographic distribution of neutral genetic variation
in Canadian peregrine falcons, Falco peregn'nus. Molecular Ecology, 16, 327-
343.

Burger, J. & M. Gochfeld (1994) Vigilance in African mammals: Differences
among mothers, other females, and males. Behaviour, 131, 153-169.

Burns, E. L., M. D. B. Eldridge & B. A. Houlden (2004) Microsatellite variation
and population structure in a declining Australian Hylid Litoria aurea.
Molecular Ecology, 13, 1745-1757.

Campbell, 0., D. Lusch, T. Smucker & E. Wangui (2005) Multiple methods in the
study of driving forces of land use and land cover change: A case study of
SE Kajiado District, Kenya. Human Ecology, 33, 763-794.

Campbell, D. J., H. Gichohi, R. Reid, A. Mwangi, L. Chege & T. Sawin (2003)
Interactions between people and wildlife in Southeast Kajiado District, Kenya.
LUCID Working Paper: 19. Nairobi: International Livestock Research
Insitutue.

Cardillo, M., A. Purvis, W. Sechrest, J. L. Gittleman, J. Bielby & G. M. Mace
(2004) Human population density and extinction risk in the world's
carnivores. PLoS Biology, 2, e197.

Caro, T. M. (1994) Cheetahs of the Serengeti Plains. Chicago: University of
Chicago Press.

Caro, T. M. & M. K. Laurenson (1994) Ecological and genetic factors in
conservation: a cautionary tale. Science, 263, 485-486.

Caswell, H. (2001) Matrix Population Models: Construction, Analysis, and
Interpretation. Sunderland, MA: Sinauer Associates.

Chakraborty, R. & M. Nei (1977) Bottleneck effects on average heterozygosity
and genetic distance with stepwise mutation model. Evolution, 31, 347-356.

148

Chardonnet, P., (2002) Conservation of the African Lion: Contributions to a
Status Survey. International Foundation for the Conservation of Wildlife,
Paris, France.

Cheney, D. L., R. M. Seyfarth, S. J. Andelman & P. C. Lee, (1988) Reproductive
success in vervet monkeys. In: Reproductive success: 384-402. T. H.
Clutton-Brock (Ed.). University of Chicago Press, Chicago.

Chesser, R. K. (1991) Gene diversity and female philopatry. Genetics, 127, 437-
447.

Clutton-Brock, T. H., S. D. Albon & F. E. Guinness, (1988) Reproductive success
in male and female red deer. In: Reproductive Success: 325-343. T. H.
CIutton-Brock (Ed.). University of Chicago Press, Chicago.

Coleman, K. & D. S. Wilson (1998) Shyness and boldness in pumpkinseed
sunfish: individual differences are context-specific. Animal Behaviour, 56,
927-936.

Coltman, D. W., J. G. Pilkington, J. A. Smith & J. M. Pemberton (1999) Parasite-
mediated selection against inbred Soay sheep in a free-living, island
population. Evolution, 53, 1259-1267.

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., K. E. Holekamp & L. Smale (1999) A seasonal feast: long-term
analysis of feeding behaviour in the spotted hyaena Crocuta crocuta. African
Journal of Ecology, 37, 149-160.

Cornuet, J.-M. & G. Luikart (1996) Description and power analysis of two tests for
detecting recent population bottlenecks from allele frequency data. Genetics,
144, 2001-2014.

Creel, S. (2001) Four factors modifying the effect of competition on carnivore
population dynamics as illustrated by African wild dogs. Conservation
Biology, 15, 271-274.

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

Creel, S. & N. M. Creel (2002) The African Wild Dog: Behavior, Ecology, and
Conservation. Princeton, N.J.: Princeton University Press.

149

Creel, S., G. Spong & N. Creel, (2001) lnterspecific competition and the
population biology of extinction-prone carnivores. ln: Carnivore Conservation:
35-60. J. L. Gittleman, S. M. Funk, D. Macdonald & R. K. Wayne (Eds.).
Cambridge University Press, Cambridge.

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

DeMaster, D. P. & l. Stirling (1981) Ursus maritimus. Mammalian Species, 145,
1-7.

Dinerstein, E. & G. F. McCracken (1990) Endangered greater one-horned
rhinoceros carry high levels of genetic variation. Conservation Biology, 4,
417-422.

Dingemanse, N. J., C. Both, P. J. Drent & J. M. Tinbergen (2004) Fitness
consequences of avian personalities in a fluctuating environment.
Proceedings of the Royal Society of London Series B-Biological Sciences,
271 , 847-852.

Drent, P. J., K. van Oers & A. J. van Noordwijk (2003) Realized heritability of
personalities in the great tit (Parus major). Proceedings of the Royal Society
of London Series B-Biological Sciences, 270, 45-51.

Durant, S. M. (2000) Living with the enemy: avoidance of hyenas and lions by
cheetahs in the Serengeti. Behavioral Ecology, 11, 624-632.

Durant, S. M., M. Kelly & T. M. Caro (2004) Factors affecting life and death in
Serengeti cheetahs: environment, age, and sociality. Behavioral Ecology, 15,
11-22.

East, M. L., T. Burke, K. Wilhelm, C. Greig 8. H. Hofer (2003) Sexual conflicts in
spotted hyenas: male and female mating tactics and their reproductive
outcome with respect to age, social status and tenure. Proceedings of the
Royal Society of London Series B-Biological Sciences, 270, 1247-1254.

East, M. L. & H. Hofer (2001) Male spotted hyenas (Crocuta crocuta) queue for
status in social groups dominated by females. Behavioral Ecology, 12, 558-
568.

East, M. L., H. Hofer, J. H. Cox, U. Wulle, H. Wiik & C. Pitra (2001) Regular
exposure to rabies virus and lack of symptomatic disease in Serengeti

spotted hyenas. Proceedings of the National Academy of Science, 98,
15026-15031 .

East, M. L., H. Hofer & W. Wickler (1993) The erect 'penis' as a flag of
submission in a female-dominated society: greetings in Serengeti spotted
hyenas. Behavioral Ecology and Sociobiology, 33, 355-370.

150

Elangovan, V., H. Raghuram, E. Yuvana Satya Priya & G. Marimuthu (2002)
Postnatal growth, age estimation and development of foraging behaviour in

the fulvous fruit bat Rousettus Ieschenaulti. Journal of Biosciences, 27, 695-
702.

Elangovan, V., H. Raghuram, E. Yuvana Satya Priya & G. Marimuthu (2004)
Wing morphology and flight performance in Rousettus Ieschenaulti. Journal
of Mammalogy, 85, 806-812.

Elmhagen, B. & S. P. Rushton (2007) Trophic control of mesopredators in
terrestrial ecosystems: top-down or bottom-up? Ecology Letters, 10, 197-206.

Engh, A. L., (2002) The causes and consequences of social rank in the spotted
hyena, Crocuta crocuta. PhD dissertation. Michigan State University, East
Lansing.

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

Engh, A. L., S. M. Funk, R. C. Van Horn, K. T. Scribner, M. W. Bruford, S.
Libants, M. Szykman, L. Smale& K. E. Holekamp (2002) Reproductive skew
among males in a female-dominated mammalian society. Behavioral
Ecology, 13, 193- 200.

England, P. R., G. H. R. Osler, L. M. Woodworth, M. E. Montgomery, D. A.
Briscoe & R. F rankham (2003) Effects of intense versus diffuse population
bottlenecks on microsatellite genetic diversity and evolutionary potential.
Conservation Genetics, 4, 595-604.

Estes, J. A. (1980) Enhydra Iutn's. Mammalian Species, 133, 1-8.

Estes, J. A., M. T. Tinker, T. M. Williams & D. F. Doak (1998) Killer whale
predation on sea otters linking oceanic and nearshore ecosystems. Science,
282, 473-476.

Faith, J. T. & A. K. Behrensmeyer (2006) Changing patterns of carnivore
modification in a landscape bone assemblage, Amboseli Park, Kenya.
Journal of Archaeological Science, 33, 1718-1733.

Forbes, S. H. & D. K. Boyd (1996) Genetic variation of naturally colonizing
wolves in the central Rocky Mountains. Conservation Biology, 10, 1082-
1090.

Forchhammer, M. C., T. H. CIutton-Brock, J. Lindstrom & S. D. Albon (2001)
Climate and population density induce long-term cohort variation in a
northern ungulate. Journal of Animal Ecology, 70, 721-729.

151

 

Frank, L. G. (1986a) Social organization of the spotted hyaena (Crocuta crocuta). a
I. Demography. Animal Behaviour, 34, 1500-1509.

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

Frank, L. G. (1997) Evolution of genital masculinization: Why do females have
such a large 'penis'? Trends in Ecology and Evolution, 12, 58-62.

Frank, L. G. & S. E. Glickman (1994) Giving birth through a penile clitoris:
parturition and dystocia in the spotted hyaena (Crocuta crocuta). Journal of
Zoology, London, 234, 659-665.

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

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

Frank, L. G., M. L. Weldele & S. E. Glickman (1995b) Masculinization costs in
hyaenas. Nature, 377, 584-585.

Frankham, R. (2005) Genetics and extinction. Biological Conservation, 126, 131-
140.

Fuller, T. K. & L. B. Keith (1981) Non-overlapping ranges of coyotes and wolves
in northeastern Alberta. Journal of Mammalogy, 62, 403-405.

Fuller, T. K. & P. R. Sievert, (2001) Carnivore demography and the
consequences of changes in prey availability. In: Carnivore Conservation:
163-178. J. L. Gittleman, S. M. Funk, D. Macdonald & R. K. Wayne (Eds.).
Cambridge University Press, Cambridge.

Funk, S. M., C. V. Fiorello, S. Cleaveland & M. E. Gompper, (2001) The role of
disease in carnivore ecology and conservation. In: Carnivore Conservation.
J. L. Gittleman, S. M. Funk, D. Macdonald & R. K. Wayne (Eds.). Cambridge
University Press, Cambridge.

Gittleman, J. L. (1986) Carnivore life history patterns: Allometric, phylogenetic,
and ecological associations. American Naturalist, 127, 744-771.

Golla, W., H. Hofer & M. L. East (1999) Within-litter sibling aggression in spotted
hyaenas: effect of maternal nursing, sex and age. Animal Behaviour, 58, 715-
726.

152

Gosling, S. D. (1998) Personality dimensions in spotted hyenas (Crocuta
crocuta). Journal of Comparative Psychology, 112, 107-118.

Gosling, S. D. (2001) From mice to men: What can we learn about personality
from animal research? Psychological Bulletin, 127, 45-86.

Gould, S. J. (1981) Hyena myths and realities. Natural History, 90, 16-24.

Hailer, F., B. Helander, A. Folkestad, S. Ganusevich, S. Garstad, P. Hauff, C.
Koren, T. Nygard, V. Volke, C. Vila & H. Ellegren (2006) Bottlenecked but
long-lived: high genetic diversity retained in white-tailed eagles upon
recovery from population decline. Biology Letters, 2, 316-319.

Hairston, N. G., F. E. Smith 8 L. B. Slobodkin (1960) Community structure,
population control, and competition. American Naturalist, 94, 421-425.

Hanby, J. P., J. D. Bygott & C. Packer, (1995) Ecology, demography, and
behavior of lions in two contrasting habitats: Ngorongoro Crater and the
Serengeti Plains. In: Serengeti ll: Dynamics, Management, and Conservation
of an Ecosystem: 315-331. A. R. E. Sinclair & P. Arcese (Eds.). University of
Chicago Press, Chicago.

Hastings, K. K. & J. W. Testa (1998) Maternal and birth colony effects on survival
of Weddell seal offspring from McMurdo Sound, Antarctica. Journal of Animal
Ecology, 67, 722-740.

Hayssen, V. (1993) Empirical and theoretical constraints on the evolution of
lactation. Journal of Dairy Science, 76, 3213-3233.

Hayward, M. W. (2006) Prey preferences of the spotted hyaena (Crocuta
crocuta) and degree of dietary overlap with the lion (Panthera [90). Journal of
Zoology, 270, 606-614.

Hebblewhite, M., C. A. White, C. G. Nietvelt, J. A. McKenzie, T. E. Hurd, J. M.
F ryxell, S. E. Bayley & P. Paquet (2005) Human activity mediates a trophic
cascade caused by wolves. Ecology, 86, 2135-2144.

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

Helldin, J. 0., O. Liberg & G. Gloersen (2006) Lynx (Lynx lynx) killing red foxes
(Vulpes vulpes) in boreal Sweden - frequency and population effects. Journal
of Zoology, 270, 657-663.

Henke, S. E. & F. C. Bryant (1999) Effects of coyote removal on the faunal

community in the western Texas. Journal of Wildlife Management, 63, 1066-
1081.

153

 

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

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

Hessing, M. J. C., A. M. Hagelso, J. A. M. van Beek, P. R. Wiepkema, W. G. P.
Schouten & R. Krukow (1993) Individual behavioural characteristics in pigs.
Applied Animal Behaviour Science, 37, 285-295.

Hoelzel, A. R., J. Halley, S. J. O'Brien, C. Campagna, T. Arnborm, B. Le Boeuf,
K. Ralls & G. A. Dover (1993) Elephant seal genetic variation and the use of
simulation models to investigate historical population bottlenecks. Journal of
Heredity, 84, 443-449.

Hofer, H. & M. East, (1995) Population dynamics, population size, and the
commuting system in Serengeti spotted hyenas. In: Serengeti ll: Dynamics,
Management, and Conservation of an Ecosystem: 332-363. A. R. E. Sinclair
& P. Arcese (Eds.). University of Chicago Press, Chicago.

Hofer, H. & M. East (2007) Siblicide in Serengeti spotted hyenas: a long-term
study of maternal input and cub survival. Behavioral Ecology and
Sociobiology, In press.

Hofer, H. & M. L. East (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. & M. L. East (1993b) 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. & M. L. East (2003) Behavioral processes and costs of co-existence in
female spotted hyenas: a life history perspective. Evolutionary Ecology, 17,
315-331.

Hogg, J., S. Forbes, B. Steele & G. Luikart (2006) Genetic rescue of an insular
population of large mammals. Proceedings of the Royal Society of London
Series B-Biological Sciences, 273, 1491-1499.

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

Holekamp, K. E., S. T. Sakai & B. L. Lundrigan (2007) Social intelligence in the
spotted hyena (Crocuta crocuta). Philosophical Transactions of the Royal
Society of London B, 362, 523-538.

154

 

Holekamp, K. E. & L. Smale (1990) Provisioning and food sharing by lactating
spotted hyenas, Crocuta crocuta (Mammalia: Hyaenidae). Ethology, 86, 191-
202.

Holekamp, K. E. & L. Smale (1991) Dominance acquisition during mammalian
social development: the "inheritance" of maternal rank. American Zoologist,
31, 306-317.

Holekamp, K. E. & L. Smale (1998) Dispersal status influences hormones and
behavior in the male spotted hyena. Hormones and Behavior, 33, 205-216.

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

Holekamp, K. E., L. Smale & M. Szykman (1996) Rank and reproduction in the
female spotted hyaena. Journal of Reproduction and Fertility, 108, 229-237.

Holekamp, K. E., M. Szykman, E. E. Boydston & L. Smale (1999) 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.

Holt, R. D. & G. A. Polls (1997) A theoretical framework for intraguild predation.
American Naturalist, 149, 745-764.

Htiner, O. P., B. Wachter, M. L. East & H. Hofer (2002) The response of spotted
hyaenas to long-term changes in prey populations: functional response and
interspecific kleptoparasitism. Journal of Animal Ecology, 71, 236-246.

Honer, O. P., B. Wachter, M. L. East, V. A. Runyoro & H. Hofer (2005) The effect
of prey abundance and foraging tactics on the population dynamics of a
social, territorial carnivore, the spotted hyena. Oikos, 108, 544-554.

Héner, O. P., B. Wachter, M. L. East, W. J. Streich, K. Wilhelm, T. Burke & H.
Hofer (2007) Female mate-choice drives the evolution of male-biased
dispersal in a social mammal. Nature, 448, 798-801.

Hood, 6., (2006) POPTOOLS 2.7.5. http://www.cse.csiro.au/poptools.

Hooge, P. N. & B. Eichenlaub, (2000) Animal movement extension to Arcview.
ver. 2.0. US. Geological Survey, Anchorage, AK, USA.

Hughes, P. M., J. M. V. Rayner & G. Jones (1995) Ontogeny of 'true' flight and

other aspects of growth in the bat Pipistrelllus pipistrellus. Journal of Zoology,
London, 235, 291 -31 8.

155

Hunter, L. T. B. & J. D. Skinner (1998) Vigilance behaviour in African ungulates:
The role of predation pressure. Behaviour, 135, 195-211.

Hunter, M. D. & P. W. Price (1992) Playing chutes and ladders: heterogeneity
and the relative roles of bottom up and top down forces in natural
communities. Ecology, 73, 724-732.

Jenness, J. (2006) Grid Tools (Jenness Enterprises) v. 1.7 (grid_tools_jen.avx)
extension for ArcView 3.x. Jenness Enterprises. Available at:
http://www.iennessent.com/arcview/gu'id tools.htm.

Joeckel, R. M. (1998) Unique frontal sinuses in fossil and living Hyaenidae
(Mammalia, Carnivora): description and interpretation. Journal of Vertebrate
Paleontology, 18, 627-639.

Johnson, J. C. & A. Sih (2005) Precopulatory sexual cannibalism in fishing
spiders (Dolomedes triton): a role for behavioral syndromes. Behavioral
Ecology and Sociobiology, 58, 390-396.

Kaeuffer, R., D. Coltman, J. Chapuis, D. Pontier & D. Réale (2007) Unexpected
heterozygosity in an island mouflon population founded by a single pair of
individuals. Proceedings of the Royal Society of London Series B-Biological
Sciences, 274, 527-533.

Kalinowski, S. T., M. L. Taper & T. C. Marshall (2007) Revising how the
computer program CERVUS accommodates genotyping error increases
success in paternity assignment. Molecular Ecology, 16, 1099-1106.

Kappeler, P. M. (1996) Causes and consequences of life-history variation among
stepsirhine primates. American Naturalist, 148, 868-891.

Kat, P. W., K. A. Alexander, J. S. Smith & L. Munson (1995) Rabies and African
wild dogs in Kenya. Proceedings of the Royal Society of London Series B-
Biological Sciences, 262, 229-233.

Keller, L. F., K. J. Jeffery, P. Arcese, M. A. Beaumont, W. M. Hochachka, J. N.
M. Smith & M. W. Bruford (2001) Immigration and the ephemerality of a
natural population bottleneck: evidence from molecular markers. Proceedings
of the Royal Society of London Series B-Biological Sciences, 268, 1387-
1394.

Keller, L. F. & D. M. Waller (2002) Inbreeding effects in wild populations. Trends
in Ecology and Evolution, 17, 230-241.

Kelly, M. J., M. K. Laurenson, C. D. FitzGibbon, D. A. Collins, S. M. Durant, G. M.
Frame, B. C. R. Bertram & T. M. Caro (1998) Demography of the Serengeti
cheetah (Acinonyx jubatus) population: the first 25 years. Journal of Zoology,
London, 244, 473-488.

156

Kerley, L. L., J. M. Goodrich, D. G. Miquelle, E. N. Smirnov, H. B. Quigley & M.
G. Hornocker (2002) Effects of roads and human disturbance on Amur tigers.
Conservation Biology, 16, 97-108.

Kissui, B. M. & C. Packer (2004) Top-down population regulation of a top
predator: lions in the Ngorongoro Crater. Proceedings of the Royal Society of
London Series B-Biological Sciences, 271, 1867-1874.

Koehler, C. E. & P. R. K. Richardson (1990) Proteles cn'status. Mammalian
Species, 363, 1-6.

Koepfli, K.-P., S. M. Jenks, E. Eizirik, T. Zahirpour, B. V. Valkenburgh & R. K.
Wayne (2006) Molecular systematics of the Hyaenidae: Relationships of a
relictual lineage resolved by a molecular supermatrix. Molecular
Phylogenetics and Evolution, 38, 603-620.

Kolowski, J. M., (2007) Anthropogenic influences on the behavior of large
carnivores in the northern Serengeti ecosystem. PhD dissertation. Michigan
State University, East Lansing, MI.

Kolowski, J. M. & K. E. Holekamp (2006) Spatial, temporal, and physical
characteristics of livestock depredations by large carnivores along a Kenyan
reserve border. Biological Conservation, 128, 529-541.

Kolowski, J. M., D. Katan, K. R. Theis & K. E. Holekamp (2007) Daily patterns of
activity in the spotted hyena. Journal of Mammalogy, 88, 1017-1028.

Kruuk, H. (1972) The Spotted Hyena: A Study of Predation and Social behavior.
Chicago: University of Chicago Press.

Kruuk, H. (1976) Feeding and social behaviour of the striped hyaena (Hyaena
vulgaris Desmarest). East African Wildlife Journal, 14, 91-111.

Kuo, C.-H. & F. J. Janzen (2004) Genetic effects of a persistent bottleneck on a
natural population of ornate box turtles (Terrapene omata). Conservation
Genetics, 5, 425-437.

Lacy, R. C. (1997) Importance of genetic variation to the viability of mammalian
populations. Journal of Mammalogy, 78, 320-335.

Lamprey, R. H. & R. S. Reid (2004) Expansion of human settlement in Kenya's
Maasai Mara: what future for pastoralism and wildlife? Journal of
Biogeography, 31, 997-1032.

Lande, R. (1988) Genetics and demography in biological conservation. Science,
241, 1455-1460.

Lariviere, S. & J. Calzada (2001) Genetta genetta. Mammalian Species, 680, 1-6.

157

 

 

Laundre, J. W., L. Hernandez 8 K. B. Altendorf (2001) Wolves, elk, and bison: *2
reestablishing the "landscape of fear" in Yellowstone National Park, USA.
Canadian Journal of Zoology, 79, 1401 -1409.

Laurenson, M. K., (1995) Implications of high offspring mortality for cheetah
population dynamics. In: Serengeti ll: Dynamics, Management, and
Conservation of an Ecosystem: 385-399. A. R. E. Sinclair 8 P. Arcese (Eds.).
University of Chicago Press, Chicago.

Leberg, P. L. (1992) Effects of population bottlenecks on genetic diversity as
measured by allozyme electrophoresis. Evolution, 46, 477-494.

Lee, E. T. 8 J. W. Wang (2003) Statistical methods for survival data analysis,
Third edn. Hoboken, N.J.: John Wiley 8 Sons.

Lewis, M. E. 8 L. Werdelin (2000) The evolution of spotted hyenas (Crocuta
crocuta). IUCN Hyaena Specialist Group Newsletter, 7, 34-36.

Libants, S., E. Olle, K. Oswald 8 K. T. Scribner (2000) Microsatellite loci in the
spotted hyena Crocuta crocuta. Molecular Ecology, 9, 1433-1449.

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

Lukas, D., B. J. Bradley, A. M. Nsubuga, D. Doran-Sheehy, M. M. Robbins 8 L.
Vigilant (2004) Major histocompatibility complex and microsatellite variation
in two populations of wild gorillas. Molecular Ecology, 13, 3389-3402.

Lukas, D., V. Reynolds, C. Boesch 8 L. Vigilant (2005) To what extent does living
in a group mean living with kin? Molecular Ecology, 14, 2181-2196.

Lynch, M., (1996) A quantitative-genetic perspective on conservation issues. In:
Conservation Genetics: Case Histories from Nature: 471-501. J. C. Avise 8
J. L. Hamrick (Eds.). Chapman 8 Hall, New York.

Malmkvist, J. 8 S. W. Hansen (2002) Generalization of fear in farm mink, Mustela
vison, genetically selected for behaviour towards humans. Animal Behaviour,
64, 487-501.

MathWorks, (2001) MATLAB, version 6.1. MathWorks, Natick, MA.

Matocq, M. D. 8 F. X. Villablanca (2001) Low genetic diversity in an endangered
species: recent or historic pattern? Biological Conservation, 98, 61-68.

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

158

 

McMahon, C. R. 8 C. J. A. Bradshaw (2004) Harem choice and breeding 9‘
experience of female southern elephant seals influence offspring survival.
Behavioral Ecology and Sociobiology, 55, 349-362.

Mech, D. L. (1977) Productivity, mortality, and population trends in wolves in
northeastern Minnesota. Journal of Mammalogy, 58, 559-574.

Menge, B. A. (2000) Top-down and bottom-up community regulation in marine
rocky intertidal habitats. Journal of Experimental Marine Biology and Ecology,
250, 257-289. ~

Mills, M. G. L. (1990) Kalahari Hyaenas: Comparative Behavioral Ecology of Two
Species. London: Unwin Hyman.

Mills, M. G. L., S. Freitag 8 A. S. Van Jaarsveld, (2001) Geographic priorities for
carnivore conservation in Africa. In: Carnivore Conservation: 467-483. J. L.
Gittleman, S. M. Funk, D. Macdonald 8 R. K. Wayne (Eds.). Cambridge
University Press, Cambridge.

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

Mills, M. G. L. 8 H. Hofer, (1998) Hyaenas: Status Survey and Conservation
Action Plan. IUCN, Gland, Switzerland.

Monson, D. H., J. A. Estes, J. L. Bodkin 8 D. B. Siniff (2000) Life history plasticity
and population regulation in sea otters. Oikos, 90, 457-468.

Muller, M. N. 8 R. Wrangham (2002) Sexual mimicry in hyenas. Quarterly
Review of Biology, 77, 3-16.

Munch, S. B., M. L. Snover, G. M. Watters 8 M. Mangel (2005) A unified
treatment of top-down and bottom-up control of reproduction in populations.
Ecology Letters, 8, 691-695.

Murray, J. L. 8 G. L. Gardner (1997) Leopardus pardalis. Mammalian Species, 1-
10.

Navidi, W., N. Arnheim 8 M. S. Waterman (1992) A multiple-tubes approach for
accurate genotyping of very small DNA samples by using PCR: Statistical
considerations. American Journal of Human Genetics, 50, 347-359.

Nei, M., T. Maruyama 8 R. Chakraborty (1975) The bottleneck effect and genetic
variability in populations. Evolution, 29, 1-10.

159

Ogutu, J. O., N. Bhola 8 R. Reid (2005) The effects of pastoralism and protection
on the density and distribution of carnivores and their prey in the Mara
ecosystem of Kenya. Journal of Zoology, London, 265, 281-293.

Ogutu, J. O. 8 H. T. Dublin (2002) Demography of lions in relation to prey and
habitat in the Maasai Mara National Reserve, Kenya. African Journal of
Ecology, 40, 120-129.

Packer, C., L. Herbst, A. E. Pusey, J. D. Bygott, J. P. Hanby, S. J. Cairns 8 M. B.
Mulder, (1988) Reproductive success of lions. In: Reproductive success:
363-383. T. H. CIutton-Brock (Ed.). University of Chicago Press, Chicago.

Packer, C., A. E. Pusey, H. Rowley, D. A. Gilbert, J. Martenson 8 S. J. O'Brien
(1991) Case study of a population bottleneck: lions of the Ngorongoro Crater.
Conservation Biology, 5, 219-230.

Paetkau, D., L. P. Waits, P. L. Clarkson, L. Craighead, E. Vyse, R. Ward 8 C.
Strobeck (1998) Variation in genetic diversity across the range of North
American brown bears. Conservation Biology, 12, 418-429.

Palomares, F. 8 T. M. Caro (1999) lnterspeciflc killing among mammalian
carnivores. The American Naturalist, 153, 492-507.

Palomares, F., P. Gaona, P. Ferreras 8 M. Delibes (1995) Positive effects on
game species of top predators by controlling smaller predator populations:
An example with lynx, mongooses, and rabbits. Conservation Biology, 9,
295-305.

Paquet, P. C. (1991) Winter spatial relationships of wolves and coyotes in Riding
Mountain National Park, Manitoba. Journal of Mammalogy, 72, 397-401.

Paquet, P. C. (1992) Prey use strategies of sympatric wolves and coyotes in
Riding Mountain National Park, Manitoba. Journal of Mammalogy, 73, 337-
343.

Patterson, B. D., S. M. Kasiki, E. Selempo 8 R. W. Kays (2004) Livestock
predation by lions (Panthera Ieo) and other carnivores on ranches
neighboring Tsavo National Park, Kenya. Biological Conservation, 119, 507-
516.

Pennycuick, C. J. 8 J. Rudnai (1970) A method of identifying individual lions,
Panthera Ieo, with an analysis of the reliability of identification. Journal of
Zoology, 160, 497-508.

Pilot, M., W. Jedrzejewski, W. Branicki, V. E. Sidorovich, B. Jedrzejewska, K.

Stachura 8 S. M. Funk (2006) Ecological factors influence population genetic
structure of European grey wolves. Molecular Ecology, 15, 4533-4553.

160

 

Piry, S., G. Luikart 8 J.-M. Cornuet (1999) BOTTLENECK: A computer program
for detecting recent reductions in the effective population size using allele
frequency data. Journal of Heredity, 90, 502-503.

Pistorius, P. A., M. N. Bester, S. P. Kirkman 8 F. E. Taylor (2001) Pup mortality
in southern elephant seals at Marion Island. Polar Biology, 24, 828-831.

Polisar, J., I. Maxit, D. Scognamillo, L. Farrell, M. E. Sunquist 8 J. F. Eisenberg
(2003) Jaguars, pumas, their prey base, and cattle ranching: ecological
interpretations of a management problem. Biological Conservation, 109, 297-
310.

Powell, R. A., (2000) Animal home ranges and territories and home range
estimators. In: Research Techniques in Animal Ecology: Controversies and
Consequences: 65-110. L. Boitani 8 T. K. Fuller (Eds.). Columbia University
Press, New York.

Pusey, A. E. 8 C. Packer (1994) Non-offspring nursing in social carnivores:
minimizing the costs. Behavioral Ecology, 5, 362-374.

Queller, D. C. 8 K. F. Goodnight (1989) Estimating relatedness using genetic
markers. Evolution, 43, 258-275.

Radespiel, U. 8 E. Zimmermann (2001) Female dominance in captive gray
mouse lemurs (Microcebus mun'nus). American Journal of Primatology, 54,
181-192.

Ray, J. C. (1995) Civettictis civetta. Mammalian Species, 488, 1-7.

Reale, D., B. Y. Gallant, M. Leblanc 8 M. Festa-Bianchet (2000) Consistency of
temperament in bighorn ewes and correlates with behaviour and life history.
Animal Behaviour, 60.

Reed, D. H. 8 R. F rankham (2003) Correlation between fitness and genetic
diversity. Conservation Biology, 17, 230-237.

Robbins, A. M., M. M. Robbins, N. Gerald-Steklis 8 H. D. Steklis (2006) Age-
related patterns of reproductive success among female mountain gorillas.
American Journal of Physical Anthropology, 131, 511-521.

Roelke-Parker, M. E., L. Munson, C. Packer, R. Kock, S. Cleaveland, M.
Carpenter, S. J. O'Brian, A. Pospischil, R. Hoffman-Lehmann, H. Lutz, G. L.
M. Mwamengele, M. N. Mgasa, G. A. Machange, B. A. Summers 8 M. J. G.
Appel (1996) A canine distemper virus epidemic in Serengeti lions (Panthera
Ieo). Nature, 379, 441-445.

161

 

Rogers, C. M. 8 M. J. Caro (1998) Song sparrows, top carnivores and nest
predation: a test of the mesopredator release hypothesis. Oecologia, 116,
227-233.

Roy, M. S., E. Geffen, D. Smith, E. A. Ostrander 8 R. K. Wayne (1994) Patterns
of differentiation and hybridization in North American wolflike canids,

revealed by analysis of microsatellite loci. Molecular Biology and Evolution,
11, 553-570.

Sabenlval, V. K., J. P. Gibbs, R. Chellam 8 A. J. T. Johnsingh (1994) Lion-human
conflict in the Gir Forest, India. Conservation Biology, 8, 501-507.

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

Schmidt, P. A. 8 L. D. Mech (1997) Wolf pack size and food acquisition. The
American Naturalist, 150, 513-517.

Schulz, T. M. 8 W. D. Bowen (2005) The evolution of lactation strategies in
pinnipeds: A phylogenetic analysis. Ecological Monographs, 75, 159-177.

Schwenk, K. 8 G. P. Wagner, (2004) The relativism of constraints on phenotypic
evolution. In: Phenotypic Integration: Studying the Ecology and Evolution of
Complex Phenotypes: 390-408. M. Pigliucci 8 K. Preston (Eds.). Oxford
University Press, Oxford.

Sih, A., A. Bell 8 J. C. Johnson (2004a) Behavioral syndromes: an ecological and
evolutionary overview. Trends in Ecology and Evolution, 19, 372-378.

Sih, A., A. M. Bell, J. C. Johnson 8 R. E. Ziemba (2004b) Behavioral syndromes:
An integrative overview. Quarterly Review of Biology, 79, 241-277.

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

Smale, L., L. G. Frank 8 K. E. Holekamp (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., S. Nunes 8 K. E. Holekamp (1997) Sexually dimorphic dispersal in
mammals: patterns, causes, and consequences. Advances in the Study of
Behavior, 26, 181-250.

Smith, J. E., J. M. Kolowski, K. E. Graham, S. E. Dawes 8 K. E. Holekamp (In
review) Social and ecological determinants of fission-fusion dynamics in the
spotted hyaena. Animal Behaviour.

Soulé, M. E. 8 L. S. Mills (1998) No need to isolate genetics. Science, 282, 1658-
1659.

162

Spielman, D., B. W. Brook 8 R. Frankham (2004) Most species are not driven to
extinction before genetic factors impact them. Proceedings of the National
Academy of Science, 101, 15261-15264.

Spong, G., M. Johansson 8 M. Bjorklund (2000) High genetic variation in
leopards indicates large and long-term stable effective population size.
Molecular Ecology, 9, 1773-1782.

Spong, G., J. Stone, S. Creel 8 M. Bj6rklund (2002) Genetic structure of lions
(Panthera Ieo L.) in the Selous Game Reserve: implications for the evolution
of sociality. Journal of Evolutionary Biology, 15, 945-953.

Springer, A. M., J. A. Estes, G. B. van Vliet, T. M. Williams, D. F. Doak, E. M.
Danner, K. A. Forney 8 B. Pfister (2003) Sequential megafaunal collapse in
the North Pacific Ocean: An ongoing legacy of industrial whaling?
Proceedings of the National Academy of Science, 100, 12223-12228.

Stamps, J. (2003) Behavioural processes affecting development: Tinbergen's
fourth question comes of age. Animal Behaviour, 66, 1-13.

StatSoft, (2002) STATISTICA (data analysis software system). StatSoft, Tulsa,
OK.

Sugg, D. W. 8 R. K. Chesser (1994) Effective population sizes with multiple
paternity. Genetics, 137, 1 147-1 155.

Sugg, D. W., R. K. Chesser, F. S. Dobson 8 J. L. Hoogland (1996) Population
genetics meets behavioral ecology. Trends in Ecology and Evolution, 11,
338-342.

Switalski, T. A. (2003) Coyote foraging ecology and vigilance in response to gray
wolf reintroduction in Yellowstone National Park. Canadian Journal of
Zoology, 81, 985-993.

Szykman, M., A. L. Engh, R. C. Van Horn, E. E. Boydston, K. T. Scribner 8 K. E.
Holekamp (2003) Rare male aggression directed toward females in a female-
dominated society: baiting behavior in the spotted hyena. Aggressive
Behavior, 29, 457-474.

Taberlet, P., S. Griffin, B. Goossens, S. Questiau, V. Manceau, N. Escaravage,
L. P. Waits 8 J. Bouvet (1996) Reliable genotyping of samples with very low
DNA quantities using PCR. Nucleic Acids Research, 24, 3189-3194.

Tannerfeldt, M., B. Elmhagen 8 A. Angerbjérn (2002) Exclusion by interference

competition? The relationship between red and arctic foxes. Oecologia, 132,
213-220.

163

 

Thompson, J. (1885) Through Masai Land. London: Low, Marston, Searle, 8
Rivington.

Thurber, J. M., R. O. Peterson, J. D. Woolington 8 J. A. Vucetich (1992) Coyote
coexistence with wolves on the Kenai Peninsula, Alaska. Canadian Journal
of Zoology, 70, 2494-2498.

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

Trinkel, M., P. H. Fleischmann 8 G. Kastberger (2006) Comparison of land-use
strategies of spotted hyenas (Crocuta crocuta, Erxleben) in different
ecosystems. African Journal of Ecology, 44, 537-539.

Trinkel, M., P. H. Fleischmann, A. F. Steindorfer 8 G. Kastberger (2004) Spotted
hyenas (Crocuta crocuta) follow migratory prey: Seasonal expansion of a
clan territory in Etosha, Namibia. Journal of Zoology, London, 264, 125-133.

Trinkel, M. 8 G. Kastberger (2005) Competitive interactions between spotted
hyenas and lions in the Etosha National Park, Namibia. African Journal of
Ecology, 43, 220-224.

Van Horn, R. C., A. L. Engh, K. T. Scribner, S. M. Funk 8 K. E. Holekamp (2004)
Behavioral structuring of relatedness in the spotted hyena (Crocuta crocuta)

suggests direct fitness benefits of clan-level cooperation. Molecular Ecology,
13, 449-458.

van Oers, K., P. J. Drent, G. de Jong 8 A. J. van Noordwijk (2004) Additive and
nonadditive genetic variation in avian personality traits. Heredity, 93, 496-
503.

Vila, C., A.-K. Sundqvist, O. Flagstad, J. Seddon, S. Bj6rnerfeldt, I. Kojola, A.
Casulli, H. Sand, P. Wabakken 8 H. Ellegren (2003) Rescue of a severely
bottlenecked wolf (Canis lupus) population by a single immigrant.
Proceedings of the Royal Society of London Series B-Biological Sciences,
270, 91-97.

Vucetich, J. A. 8 S. Creel (1999) Ecological interactions, social organization, and
extinction risk in African wild dogs. Conservation Biology, 13, 1172-1182.

Wahaj, S. A., N. J. Place, M. L. Weldele, S. E. Glickman 8 K. E. Holekamp
(2007) Siblicide in the spotted hyena: analysis with ultrasonic examination of
wild and captive individuals. Behavioral Ecology, doi:
10.1093/beheco/arm076.

Watts, H. E. 8 K. E. Holekamp (In review) Ecological determinants of survival
and reproduction in the spotted hyena. Journal of Animal Ecology.

164

Weckworth, B. V., S. Talbot, G. K. Sage, D. K. Person 8 J. Cook (2005) A Signal
for Independent Coastal and Continental histories among North American
wolves. Molecular Ecology, 14, 917-931.

Werdelin, L. (1989) Constraint and adaptation in the bone-cracking canid
Osteoborus (Mammalia: Canidae). Paleobiology, 15, 387-401.

Werdelin, L. 8 N. Solounias (1991) The Hyaenidae: taxonomy, systematics and
evolution. Fossils and Strata, 30, 1-104.

Western, D. (1989) The ecological role of elephants in Africa. Pachyderrn, 12,
42—45.

Western, D. (2007) A half a century of habitat change in Amboseli National Park,
Kenya. African Journal of EcoIOQY, doi:10.1 1 11/j.1365-2028.2006.00710.x.

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

White, P. A. (2005) Maternal rank is not correlated with cub survival in the
spotted hyena, Crocuta crocuta. Behavioral Ecology, 16, 606-613.

White, P. J. 8 R. A. Garrott (1997) Factors regulating kit fox populations.
Canadian Journal of Zoology, 75, 1982-1988.

White, P. J., K. Ralls 8 R. A. Garrott (1994) Coyote - kit fox interactions as
revealed by telemetry. Canadian Journal of Zoology, 72, 1831-1836.

White, T. C. R. (1978) The importance of a relative shortage of food in animal
ecology. Oecologia, 33, 71-86.

Whitehouse, A. M. 8 E. H. Harley (2001) Post-bottleneck genetic diversity of
elephant populations in South Africa, revealed using microsatellite analysis.
Molecular Ecology, 10, 2139-2149.

Wielgus, R. B. 8 F. L. Bunnell (1994) Dynamics of a small, hunted brown bear
Ursus arctos population in Southwestern Alberta, Canada. Biological
Conservation, 67, 161-166.

Wilmers, C. C., R. L. Crabtree, D. W. Smith, K. M. Murphy 8 W. M. Getz (2003)
Trophic facilitation by introduced top predators: grey wolf subsidies to

scavengers in Yellowstone National Park. Journal of Animal Ecology, 72,
909-916.

Wilmers, C. C. 8 W. M. Getz (2005) Gray wolves as climate change buffers in
Yellowstone. PLoS Biology, 3, e92.

165

 

Wilmers, C. C. 8 D. R. Stahler (2002) Constraints on active-consumption rates in
gray wolves, coyotes, and grizzly bears. Canadian Journal of Zoology, 80,
1256-1261.

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

Woodroffe, R., (2001) Strategies for carnivore conservation: lessons from
contemporary extinctions. In: Carnivore Conservation: 61-92. J. L. Gittleman,
S. M. Funk, D. W. Macdonald 8 R. K. Wayne (Eds.). Cambridge University
Press, Cambridge.

Woodroffe, R. 8 L. G. Frank (2005) Lethal control of African lions (Panthera Ieo):
local and regional population impacts. Animal Conservation, 8, 91-98.

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

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

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

Young, T. P. (1994) Natural die-offs of large mammals: implications for
conservation. Conservation Biology, 8, 410-418.

166

     

lljlljljjjllljjljljljlllljjlljll