v if...
01“.. .22.
Cheat .

 

.2
3):; . .
. .z . on. 4;“.
. .a .
a .

. J

,. Ix 91‘
1 :1 .721:

.352...

 

w :45“. 5.124:
I I Iy ‘0 iv 5.
E: 9.: (1
hunt}. ‘ .
3 -

 

 

x3...
.2 chit... .1...

3 u at < .

. .3. i..\:v.{ld:u..
.- 5... ‘l' .3 ,

 

 

 

 

 

- :z'...
.1

 

 

 

5.2.4.1: q.
.(tl! . 1'
‘1 LA

 

 

2.... ¥vlrk..l . rr.
uw,..':r:tl .12 .3.) . :1: 1.:P

lIBRARY
Michigan State
University

This is to certify that the

dissertation entitled

SEXUAL BEHAVIOR AND MALE MATE CHOICE IN THE
SPOTTED HYENA, CROCUTA CROCUTA

presented by

Micaela Szykman

has been accepted towards fulfillment
of the requirements for

Ph.D. Zoology and

degree in

 

 

Ecology, Evolutionary
Biology and Behavior

  

Major professor

  

10 December 2001
Date

 

MS U is an Afflrmatiw Action/Equal Opportunity Institution 042771

 

 

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DateDuepss-pts

 

SEXUAL BEHAVIOR AND MALE MATE CHOICE IN THE SPOTTED HYENA,
CROCUTA CROCUTA

By

Micaela Szykman

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

2001

ABSTRACT

SEXUAL BEHAVIOR AND MALE MATE CHOICE IN THE SPOTTED HYENA,
CROCUTA CROCUTA

By

Micaela Szykman
Although female animals tend to be choosier than males with respect to their
mate preferences, sexual selection theory predicts that males should also be
selective when female fecundity varies. In this dissertation, l inquired whether
male mate choice occurs in the spotted hyena (Crocuta crocuta). Although male
hyenas do not invest heavily in reproduction in this species, reproductive success
among female spotted hyenas varies greatly on the basis of social rank.
Therefore, if males can assess the reproductive values of females, then sexual
selection theory predicts that male hyenas should be highly selective in their
choice of mates. Here, I looked for evidence of male mate choice in Crocuta by
determining whether males interact differently with potential mating partners, and
whether variation in these interaction patterns is correlated with male
reproductive success.

To assess male choice of females, I calculated association indices for all
male-female pairs, and analyzed rates of affiliative behavior directed by males
toward females, in a group of free-living hyenas in Kenya. Males initiate most
affiliative interactions with females, and males associate most closely with
females that are likeliest to be fertile. High- and middle-ranking males associate
most closely with high-ranking females, and some individual males pursue a

strategy of long-term associations to acquire mates. Using molecular markers to

determine paternity of cubs, I found that sires associate more closely with the
mothers of those cubs than do non-sires, particularly during the last months
before conception. These association data indicate that male spotted hyenas
possess the ability to assess variation in female reproductive value, and they
prefer females likeliest to maximize male reproductive success.

Masculinization of the female genitalia and female social dominance
appear to influence sexual behavior in this species. Courtship behaviors in
Crocuta are unique among mammals and reflect extreme ambivalence on the
part of the male, who appears to be motivated by conflicting desires to approach
the female and flee from her. Copulation provides a unique physical challenge for
the male hyena, such that he must insert his erect penis into the female’s
forward-facing phallus, which remains flaccid during mating.

I also investigated the occurrence and function of male aggression
directed toward females in this species. Male hyenas rarely direct aggression
toward adult females, as females are socially dominant to all immigrant males in
the clan. Most aggression directed by adult immigrant males against females
occurs when coalitions of two or more males attack a single adult female, who
responds by defending herself and fighting back. Daily attack rates vary
significantly with female reproductive state, and the highest rates of male attack
on females are observed when females are most likely to conceive. The adaptive
significance of male aggression against females in this species is unknown, but
my results are most consistent with predictions of an hypothesis suggesting that
male aggression may function to inform the female about some aspect of male

fitness.

Copyright by
MICAELA SZYKMAN
2001

ACKNOWLEDGMENTS

I am grateful to innumerable people who have provided support and guidance
throughout all stages of this dissertation. This work would not have been possible
without the assistance and encouragement of the following people and
organizations.

I am indebted to Kay Holekamp for allowing and entrusting me to work on
the Mara Hyena Project for my dissertation research. I thank the Office of the
President of Kenya for permission to conduct this research. I also thank the
Kenya Wildlife Service, the Narok County Council, and the Senior Warden of the
Masai Mara National Reserve for their cooperation.

I am thankful for funding and logistical support provided by the department
of Zoology and the Zoology office staff, and for fellowships and encouragement
from Don Hall and Pat Resler of the Ecology, Evolutionary Biology and Behavior
program. I thank the Graduate School for awarding me a travel grant and a
dissertation completion fellowship. The Mara Hyena Project has been supported
by NSF grants ENS-8706939, ENS-9021461, IBM-9309805, IBN-9630667, IBN-
9906445 and lBN-O113170.

Life throughout the dissertation process would never have been as fun or
as rewarding without my friends and colleagues in the Holekamp lab. Members
past and present include: Pat Bills, Erin Boydston, Stephanie Dloniak, Anne
Engh, Paula Garrett, Keron Greene, Tyson Harty, Karen Hubbard, Joe Kolowski,

Terri McElhinny, Eva-Maria Muecke, Keith Nelson, Scott Nunes, Carmen

Salsbury, Jaime Tanner, Russ Van Horn, Sofi Wahaj, and Heather Watts. Their
smiling faces, consistent support, invaluable discussions on hyena behavior and
much more have been indispensable, fruitful, and fun.

Several special people have provided me with unconditional friendship
over many years. Lauren Bliss has been my dear friend for almost 20 years. She
sent me more letters than any friend (only my father sent more) while I was away
in the field, and she continues to offer me boundless support. Terri McElhinny
and Pat Bills were my first friends in East Lansing. They have always opened
their home to me and have helped me in an infinite number of ways. Erin
Boydston was a dedicated companion, both in Michigan and in the field, and
played a critical role in helping me to survive El Nino. Anne Engh has been a
faithful, dependable, and trustworthy friend since the day we met, has offered me
never-ending kindness, and has helped me on countless occasions without my
ever asking. Corine Vriesendorp saved me from intense culture shock upon my
return to civilization after two and a half years of living in the bush. With her
friendship, she has provided me with constant amusement and entertainment, an
athletic outlet, a reliable excuse to get outside, and so much more. Furthermore,
the Vriesendorp family was extremely generous by offering me a peaceful refuge
in their house in Colorado for a summer of dissertation writing (and hiking and
rowing). I have been fortunate to have such wonderful and loyal friends.

The following people contributed to data collection and ground support for
the Mara Hyena Project. I thank Nancy Berry, Erin Boydston, Susan Cooper,

Stephanie Dloniak, Martin Durham, Anne Engh, Josh Friedman, Paula Garrett,

vi

lsla Graham, Tyson Harty, Cathy Katona, John and Peris Keshe, Pilot Nairuori,
Keith Nelson, Karen Nutt, Gabe Ording, Moses Sairowa, Joseph Silantoi, Sofi
Wahaj, Kim Weibel, and Brad White for their assistance in the field. I thank Jenny
Scullon for time consuming data extraction. Sofi Wahaj and Jakob Bro-Jorgensen
each generously allowed me to use one of their photographs for Chapter 3.

I am grateful to the following people who contributed to the genetics lab
work. essential for determining paternity among members of our hyena clan:
Peter Arctander, Mike Bruford, Tamsin Burland, Trevor Coote, Anne Engh, Tom
Friedman, Stephan Funk, Susan Jenks, Pieter Kat, Scot Libants, Kim Scribner,
Russ Van Horn, Delia Wasao, and Robert Wayne. I would not have been able to
relate male behavioral choices with male reproductive success without their
efforts.

I am indebted to many people who came to my rescue numerous times in
the field, and who provided great and lasting friendships to keep me sane and
happy and laughing. Big thanks go to Jakob Bro-Jorgensen, Anthony Cheffings,
Elly and Milton Kirkman, Zohar Navon, Alex Obara, Vincent Onyango, Andrew
and Isla Peart, Howard Saunders, Nasha Silantoi, Matt Walpole, John Watkin,
and Kaz Zanella. I cannot thank the Dundas family enough for everything they
have done for me. Barbara, Nigel, Sarah, and Sean Dundas took me in as a
member of their family and provided me (and the Hyena Project) with endless
emotional and logistical support while in Kenya. I extend very special thanks to

Howard Saunders and John Watkin for years of love and support. They offered

vii

me their home and everything that was theirs, and went above and beyond the
bonds of friendship.

Many, many other people have provided me with friendship and support
throughout my graduate career. To my friends too many too list, I appreciate your
friendship.

This dissertation was improved by contributions from members of my
academic committee: Kay Holekamp, Laura Smale, Lyn Clemens, and Fred
Dyer. I thank Don Straney, who took the place of Fred Dyer during my
comprehensive exams. I appreciate the time and consideration offered by Fred
and Lyn. I am grateful to Laura, who was always available to talk, always positive
and very supportive. I am deeply grateful to Kay, my advisor and mentor, who
offered me the incredible opportunity to work with her and the hyenas. She
provided me with immeasurable guidance, several years of financial support, and
made me a strong and careful scientist.

Finally, this great accomplishment would never have been so successful
without the love and support from my extraordinary family. Corradina and
Maurice Szykman are exceptional people and amazing parents. They always
encouraged me to follow my dreams, taught me to pursue what I love and
whatever makes me happy, and gave me the confidence to be good at what I do.
I also gained strength from Simon and Ruth Szykman, and my enchanting niece,
Anna, the brilliant star in the center of my family’s universe. My most heartfelt

thanks go out to all these people that I love.

viii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ...................................................................... v

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

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

CHAPTER 1

GENERAL INTRODUCTION ................................................................... 1

CHAPTER 2

PATTERNS OF ASSOCIATION AMONG MALE AND FEMALE HYENAS

INDICATE MALE MATE CHOICE .......................................................... 16
Introduction .............................................................................. 16
Methods .................................................................................. 18
Results .................................................................................... 26
Discussion ............................................................................... 38

CHAPTER 3

COURTSHIP AND MATING IN FREE-LIVING SPOTTED HYENAS .............. 44
Introduction .............................................................................. 44
Methods .................................................................................. 51
Results .................................................................................... 63
Discussion ............................................................................... 91

CHAPTER 4

RARE MALE AGGRESSION DIRECTED TOWARD FEMALES IN A FEMALE-

DOMINATED SOCIETY ..................................................................... 101
Introduction ............................................................................ 1 01
Methods ................................................................................. 103
Results .................................................................................. 111
Discussion ........................................................................... _... 132

CHAPTER 5

CONCLUSIONS ................................................................................ 142

LITERATURE CITED ......................................................................... 151

LIST OF TABLES

TABLE 3.1. Sample sizes for mating sessions. In the text and in all analyses,
mating sessions refer to sessions in which we observed mounting, including
mounting both with and without intromisslon.

TABLE 3.2. Results of logistic regression evaluating the influence of female
characteristics on male choice behavior.

TABLE 4.1. Data collection periods for analyses presented in Chapter 4.

TABLE 4.2. Locations (A) and contexts (B) in which individual acts of aggressive
behavior were observed between 1988 and 1995. (Note: aggression by males
toward females observed between 1996 and 1999 are considered in the text but
are not included in the last column of this table.)

LIST OF FIGURES

Figure 2.1. Mean (:t SE) intersexual association indices plotted as a function of
female social rank. By convention, the highest rank possible is one. Each point
represents the mean Al for all females who held a particular rank with all males
observed with each female while she held that rank. N=2665 associating pairs
involving 19 females and 40 males associating during all female reproductive
states.

Figure 2.2. Mean association indices for male-female pairs measured across all
female reproductive states and grouped by female social rank. Female
reproductive states are as follows: P=pregnancy; L1, L2, L3=first, second, and
third trimesters, respectively, of the lactation period; O=other states, when adult
females are neither pregnant nor lactating. Rank categories: high=ranks 1-7,
middle=ranks 8-14, Iow=ranks 15 and higher. Sample sizes are as in Figure 2.1.

Figure 2.3. Mean (1 SE) association indices between females in each rank
category and (A) high-ranking males, (B) middle-ranking males, and (C) Iow-
ranking males. Rank categories are as in Figure 2.2 and are the same for both
sexes. Sample sizes are as in Figure 2.1.

Figure 2.4. Mean number of males observed per session with each female
across all female reproductive states. N=4798 sessions for which 19 females
were observed with 40 males. Bars sharing same letters represent means not
significantly different from one another as determined by Tukey’s post-hoc
comparisons.

Figure 2.5. Synchrony among 105 fertile periods of 35 females observed
between 1988 and 1995. Sample sizes represent the number of 2-week intervals
during which females were concurrently fertile.

Figure 2.6. (A) Association indices for eight females with the eight males who
sired (solid bars) their litters conceived during interval ‘0’ compared with Als for
these same females with 15 males present concurrently who did not sire (open
bars) their litters. Interval ‘0’ represents the date of conception i 2 weeks.
Remaining intervals represent month-long periods surrounding the date of
conception. (B) Percent of observation sessions during which sires and non-sires
were found alone with females, calculated as [(the number of sessions in which
the pair was found alone/the total number of sessions in which that male was
observed )*1 00].

Figure 3.1. Mean daily rates (tSE) at which males engaged in (A) approaches,
(B) investigatory behaviors, and (C) courtship behaviors, plotted as a function of
time relative to conception in targeted females. Interval ‘0’ represents the date of
conception :I: 2 weeks. Remaining intervals represent month-long periods

xi

surrounding the date of conception. N=202 male-female pairs involving 20
females and 21 immigrant males observed together during each of 7 month-long
intervals. lnvestigatory behaviors included sniffing, licking, and greeting.
Courtship behaviors included approach-avoid behavior, presenting, bowing
display, and pawing the ground. Note that all Y-axes are on the same scale.

Figure 3.2. Daily rates at which males engaged in (A) approaches, (B)
investigatory behaviors, and (C) courtship behaviors, plotted as a function of
female social rank and age categories. Daily rates were rank transformed to
perform 2-way ANOVA on data that were not normally distributed (see Methods
for details). Transformed rates are plotted here (meaniSE).

Figure 3.3. (A) Times of day at which matings were observed in relation to total
number of hours of observation. (B) Daily mating rate calculated for morning and
evening hours of observation. AM represents 0500-0900 hours, and PM
represents 1600-2000 hours.

Figure 3.4. Mating positions. (A) The male squats behind the female as he
attempts to achieve intromission. (B) The male stands and thrusts after achieving
intromission. (C) The male rests his chin and chest on the female’s back after
ejaculation.

Figure 3.5. Temporal patterning of mounting behavior for a single male (WALD)
mounting a single female (SEIN) when no other hyenas were present. Observers
remained with this pair through minute 50.

Figure 3.6. Temporal patterning of mounting sequences for two males (A=MER,
B=EXC) alternately mounting a single female (GOL) during a single observation
session. Observers remained with these animals through minute 150. MER was
higher-ranking than EXC, and each male sired one of the cubs in the twin litter
produced 110 days after these copulations were observed.

Figure 3.7. Duration of 275 mounts observed during 39 mating sessions,
suggesting a possible functional distinction between short (<60 3) and long
(>1OOS) mounts.

Figure 3.8. Mean (:tSE) hourly rates at which individuals approached, and mean
(tSE) percent of focal animal time during which individuals followed, members of
the opposite sex in two different contexts. Open bars represent adult males and
solid bars represent adult females. The ‘mating' context consists of sessions
during which we observed mounting behavior. The ‘other’ context includes all
other sessions, involving observations of hyenas present at dens, at carcass
sites, and at locations other than dens and carcasses. Numbers over bars
represent number of individuals included in each sample. * P<0.001

xii

Figure 3.9. (A) Total time immigrant males were observed with 20 females, and
(B) number of immigrant males that were observed associating with same
females, during the months surrounding each female’s conception. Letters over
bars represent means significantly different from one another at P<0.05.

Figure 3.10. Association patterns between four females and the males who sired
(open diamonds) their litters (N=3 males; open diamonds in (A) and (B) represent
the same male), and the 17 males present concurrently who did not sire (solid
diamonds) their litters. Estimated day of conception is indicated by the solid
vertical line. Although the number of consecutive sessions during which each
female was observed varied, the time scales in (A)-(D) are all standardized to
include the three months preceding and the first month following conception.

Figure 4.1. Daily rates (meaniSE) at which adult hyenas (N=23 females and 20
immigrant males) directed aggression toward other adults of each sex, controlled
for the number of hours observed during which each adult was found in the
presence of other adults of each sex between 1988 and 1995.

Figure 4.2. Percent of total individual acts of aggression of high intensity that
were directed by male and female hyenas against members of the same and
opposite sex. Sample sizes indicate total number of individual aggressive acts
observed between 1988 and 1995.

Figure 4.3. Percent of total attacks on other adults that occurred in coalitions.
Sample sizes indicate total number of attacks observed between 1998 and 1995.

Figure 4.4. Female responses to male approaches. Sample sizes indicate
numbers of male approaches observed for which female response was recorded.

Figure 4.5. Numbers of individual males participating in 227 attacks on adult
females. Sample sizes indicate numbers of interactions observed between 1988
and 1999 in which adult immigrant males attacked adult Talek females.

Figure 4.6. Example of male coalitionary aggression toward a female. Three
males surround and attack a female, who crouches in the center of this
photograph with her teeth bared. Note the males’ bristled tails, indicating a high
level of excitement.

Figure 4.7. Daily rates (meaniSE) at which adult males in each of three rank
categories baited females. Rank categories high, middle, and low represent male
ranks 1-6, 7-12, and 2 13, representing N=23, 18, and 8 males, respectively.
Bars sharing same letters represent means not significantly different from one
another.

Figure 4.8. Proportion of males observed baiting females per year of male tenure
in the Talek clan. Numbers over bars represent the number of males represented

xiii

in each year of tenure.

Figure 4.9. Daily baiting rates (meaniSE) per female for 22 females in various
reproductive conditions observed between 1988 and 1995. “N” indicates
nulliparous, “P” indicates pregnant, “L1 " represents the first trimester of the
lactation interval, “L2" indicates the second trimester of the lactation interval, “L3"
indicates the third trimester of the lactation interval, and “0” indicates the period
between weaning one litter and conceiving the next. Each female was observed
in all six reproductive states.

Figure 4.10. Observations of baitings relative to the date of the nearest female
conception. The interval labeled “0” represents the period starting 2 weeks prior
to conception and ending 2 weeks after conception. The interval labeled “1" thus
represents the period 2-6 weeks after conception, and so forth. For this analysis,
we only included baitings that took place within one year of a female’s nearest
conception. (N=172 baitings of 27 females)

Images in this dissertation are presented in color.

xiv

Chapter 1

GENERAL INTRODUCTION

This dissertation describes intersexual association patterns and sexual behavior,
and examines male mate choice in a gregarious mammal, the spotted hyena
(Crocuta crocuta). Spotted hyenas possess many unusual behavioral and
morphological traits that make them fascinating subjects for the study of sexual
behavior. For example, female social dominance in spotted hyena groups
creates a social environment that is unique among mammals, and influences
male hyena courtship and mating behavior. Furthermore, high variation among
female hyenas, with respect to their reproductive value, suggests that sexual
selection may act in an unexpected way in this promiscuous species in which
males provide no parental care, and in which dominant females are assumed to
make final decisions regarding mating. In this introductory chapter, I set the stage
for the study of male mate choice in Crocuta and explain why this peculiar animal
is an appropriate subject for the study of this phenomenon.

In 1871, Darwin proposed the concept of sexual selection as a possible
explanation for the existence of intraspecific sex differences in behavior and
morphology, such as dimorphic body size and extravagant male ornaments.
These traits could not be explained by natural selection because they did not
improve survival, but they did appear to Darwin to improve mating success.

Danlvin (1871) identified two components of sexual selection: intrasexual and

intersexual selection. lntrasexual selection involves direct competition for mates
among conspecifics of the same sex. Intersexual selection is mediated by choice
and influences the likelihood that a given individual will be chosen as a mate by
conspecifics of the opposite sex. lntrasexual selection is ubiquitous among
animals and is most often characterized by a struggle between males for the
possession of females (Darwin 1871; Clutton-Brock et al. 1982; Smuts 1987b).
During intersexual selection, females typically have more opportunities than do
males to choose a particular mating partner among many potential mates
(Andersson 1982; Smuts 1987b). Thus in nature most intrasexual selection
involves male-male competition and most intersexual selection involves female
choice of mates.

The differential operation of these two modes of sexual selection in the
two sexes was originally attributed to the highly unequal investments made by
males and females in individual gametes (Parker et al. 1972). Anisogamy, the
unequal size of male and female gametes, sets the stage for sexual conflict. In
many species, females invest far more heavily than do males in the production of
offspring, if for no other reason than because ova, in contrast to sperm, are
generally laden with nutrients for early support of the embryonic offspring (Parker
et al. 1972). As males compete for access to the most fecund females, females
should be very careful in their choice of mating partners (Parker et al. 1972).

To show that these sex-specific differences in behavior result in
differences in the potential rate of reproduction, it is necessary to obtain accurate

data documenting differential reproductive success in males and females. The

potential rate of reproduction by males and females is measured as the
maximum number of independent offspring that parents can produce per unit
time (Clutton-Brock and Vincent 1991 ). In a classic study, Bateman (1948)
observed that reproductive success was more variable in male Drosophila than in
females. He determined that this variation existed because males could increase
their reproductive success by increasing the number of females with which they
mated, whereas females maintained the same reproductive success regardless
of the number of matings they secured. That is, females were not limited by the
number of mates they could attract but by other factors, including the number of
mature eggs they possessed.

Trivers (1972) extended the idea of unequal reproductive investment by
the two sexes beyond gamete production to encompass the contributions made
by males and females in the form of parental care. He defined parental
investment as “any investment by the parent in an individual offspring that
increases the offspring’s chance of surviving (and hence reproductive success)
at the cost of the parent’s ability to invest in other offspring (p.139).” Parental
investment may thus include providing nutrition to the female prior to offspring
production in addition to the more common parental role of feeding and
protecting offspring. Whether sexual selection acts via male-male competition or
female choice of mates depends on limitations imposed by differential parental
investment. Theory predicts that the sex investing more time and energy in
parental care should become a limiting resource. Individuals of the non-limiting

sex will necessarily compete for individuals of the limited sex. The limited sex

then has the opportunity to choose among competing prospective mates. Thus,
the relative investment of each parent determines the extent to which animals
compete for members of the limited sex (Thornhill and Gwynne 1986). Among
mammals, in which obligate maternal care occurs with lactation, parental
Investment is greater for the female than the male, so theory predicts that sexual
selection should operate in mammals via male-male competition and female
choice. In fact, these patterns are typical of many ungulates (e.g. red deer,
Cervus elaphus: CIutton-Brock et al. 1982), carnivores (e.g. sea lions, Otaria
byronia: Campagna and LeBoeuf 1988) and primates (e.g. baboons, Papio
cynocephalus: Altmann et al. 1988).

Examples of extended maternal care abound in the literature and are not
limited to mammals. Many female birds are sole caretakers of their broods and
are well known for their careful selection of males with long tails, colorful
plumage or elaborate bowers (e.g. widowbirds, Euplectes progne: Andersson
1982; bowerbirds, Ptilonorhynchus violaceus: Borgia 1985; finches, Carpodacus
mexicanus: Hill 1990). In species where males provide the majority of parental
care, parental investment theory (Trivers 1972) predicts female-female
competition over mates and male choosiness during mate choice (Petrie 1983;
Oring et al. 1991; Andersson 1994). This theory may be limited in its practical
application, since it fails to predict the circumstances under which competition or
choice by both males and females should occur.

In addition to affecting potential reproductive rates of the two sexes,

differential parental investment in offspring can also have an effect on the

operational sex ratio of a population. The operational sex ratio (OSR) is defined
as the ratio of sexually active males to sexually receptive females in a population
at any given time (Emlen and Oring 1977). Factors affecting OSR include biased
overall sex ratios, reproductive longevity, differential mortality rates, migration
schedules, reproductive synchrony, physiological constraints on reproduction,
and the spatial and temporal distribution of resources in the environment (Emlen
and Oring 1977; Kvarnemo and Ahnesjo 1996). OSR was first presented by
Emlen and Oring (1977) as a concept useful in elucidating the intensity of sexual
selection in a population. They argued that OSR provides “an empirical measure
of the degree of monopolizability of mates,” and OSR can therefore be used in
ecology as a predictor of mating systems. In addition, the greater the degree of
monopolization, the more intense the intrasexual selection will be, and the
greater the resulting variance in reproductive success within the competing sex.
Thus, OSR can be used to predict the intensity of intrasexual competition and the
direction of sexual selection (Gwynne 1981; Kvarnemo et al. 1995; Kvarnemo
and Ahnesjo 1996). For example, one would expect to find intense male-male
competition when operational sex ratios are strongly male biased.

Theoretical developments since 1977 elaborate on the use of operational
sex ratio as a predictor of sex roles in reproductive behavior by including sex
differences in variance in mate quality as an additional factor influencing the
direction of sexual selection (Gwynne 1981; Parker 1983; Petrie 1983; Owens
and Thompson 1994). Specifically, these authors suggest that OSR can be used

to predict the direction of intrasexual selection, or which sex will be competing for

mates, but argue that OSR does not predict extent or direction of mate choice.
These authors propose that high variance in the quality of one sex promotes
increased mate choice behavior in the opposite sex. Mate quality may be
measured in a variety of ways, including measuring traits like female size or by
calculating lifetime reproductive success. For example, female size is strongly
correlated with fecundity in animals of indeterminate growth such as insects
(Gwynne 1981) and fish (C6te and Hunte 1989). Thus, if female body size is a
normally distributed trait within a population, we should expect increased mate
choice by males, who should have the ability to evaluate the reproductive
potential of females, as predicted by body size.

Gibson and Langen (1996) agree that variance in mate quality will
influence mate choice. However, they further suggest that costs of searching for
mates might influence the level of optimal choosiness (for more on search theory
and costs of searching, see Real 1990). For example, male thirteen-lined ground
squirrels, Spermophilis tridecemlineatus, use information on female reproductive
condition to guide their future search activities. That is, males invest more time
searching for females that they anticipate will be in estrus the next day
(Schwagmeyer 1995). Owens and Thompson (1994) incorporate search costs,
including predation risk and physiological constraints on reproduction, into their
model describing the behaviors which should be optimized during mate choice,
specifically the trade-off between reproductive rate and mate quality. As noted
above, the direction of sexual selection can be predicted by different potential

reproductive rates in males and females (Clutton-Brock and Vincent 1991), since

OSRs are biased in favor of the sex with higher reproductive rates. However,
when there is sufficiently greater variation in mate quality among the sex with the
lower potential reproductive rate, the sex with the higher potential mating rate
should be the more choosy sex (Owens and Thompson 1994).

To summarize, competition between males is expected to occur when (1)
male parental investment involves little or nothing more than the initial
contribution of sperm, (2) males have higher potential reproductive rates than do
females, and (3) operational sex ratios are male-biased, especially under
conditions of highly monopolizable resources (food and/or mates). Intersexual
selection via female mate choice is expected when (1) females invest more than
males in parental care, (2) females have lower reproductive rates than do males,
(3) OSR is male-biased and resources are monopolizable, and (4) there is
variation in male mate quality.

Although male-male competition and female mate choice are the most
common expressions of the operation of sexual selection in animals, Darwin
(1871) himself recognized that rare cases occur in nature in which males might
have the opportunity to choose the most attractive females from an array of
mates. Indeed, male mate preferences have been documented in nature in
studies reporting that males prefer older, more experienced mates (Burley and
Moran 1979), younger mates who are likely to be virgins (Forsberg 1987), larger,
more fecund females (Gwynne 1981; Verrell 1985; Verrell 1989; Berglund and
Rosenqvist 1993; Verrell 1995), females who exhibit better parental care (Verrell

1990; Solomon 1993), and females with higher reproductive potential (Berger

1989; Schwagmeyer and Parker 1990; Schwagmeyer1995). In general, male
mate choice is expected when males invest heavily in reproduction, when
females are synchronous in their sexual receptivity, or when there is high
variation in the quality of potential mates (Gwynne 1981; Parker 1983; Owens
and Thompson 1994).

Male investment in reproduction may take the form of increased parental
care, as in some monogamous mammals (e.g. carnivores: Moehlman 1989;
rodents: Solomon 1993; primates: Tardiff 1997) and many species of fishes (e.g.
Gross and Shine 1981; Berglund and Rosenqvist 1993). Alternatively, some male
insects, e.g. Mormon crickets (Anabrus simplex), transfer personal,
nonrenewable nutrients, in the form of a large spermatophore or nuptial gift, to a
female during copulation, at a significant cost to themselves (Gwynne 1981;
Forsberg 1987). In such cases, males are expected to discriminate among
females with respect to their allocation of expensive ejaculate or spermatophore,
and indeed they do (Gwynne 1981 ). High male investment now limits the number
of matings possible for males and generates opportunities for eliciting female-
female competition. Finally, variation in female quality promotes male mate
choice.

Some males prefer mates based on their previous reproductive success
and on their future reproductive potential. In bison (Bison bison), males identify,
and choose to mate with, more fertile females (Berger 1989). Support for male
discrimination in bison is presented by Komers and colleagues (1994) who show

that males are more likely to tend females when the females are close to estrus,

as indicated by low progesterone levels. In feral pigeons (Columba Iivia), both
males and females prefer more experienced mates, but only up to a certain age
when great age appears to become a liability offsetting the benefit of greater
experience (Burley and Moran 1979). These forms of male mate choice require
that males be capable of evaluating females. Tests of the hypothesis that
variation in female quality promotes male mate choice must show that (3)
females vary with respect to their reproductive potential, (b) females of varying
potential are simultaneously available, (c) males do not copulate at random, and
(d) when given simultaneous presentation, males will selectively mate with
females promoting higher fitness, not just with as many females as possible
(Parker 1983; Petrie 1983). In this dissertation, I attempted to test these
predictions to address the hypothesis that variation in female quality promotes
selective mate choice in male spotted hyenas.

Spotted hyenas exhibit many unusual behavioral and morphological traits
that are reversed from mammalian norms. They live in matriarchal, female-
dominated societies in which the females are more aggressive, and
approximately 10% larger, than males (Kruuk 1972; Frank 1986b). In addition,
females exhibit highly masculinized external genitalia complete with
pseudoscrotum and a fully erectile phallus (Matthews 1939; Neaves et al. 1980).
Although spotted hyenas are monomorphic in appearance, and although they
exhibit patterns reversed from mammalian norms with respect to social
dominance, hyenas nevertheless exhibit many of the same sex differences as

those seen in virtually all other mammals. For example, females are responsible

for all parental care, and males take the traditional, active role in courtship (Kruuk
1972; Frank 1986b). Despite several long-term field studies on this species
conducted by excellent scientists (Kruuk 1972; Frank 1986b; Frank 1986a; Mills
1990), little is known either about sexual behavior of hyenas in the wild, or about
features of mate choice in this species.

The present study was initiated after Holekamp and colleagues (1996)
determined that female spotted hyenas do, in fact, vary with respect to their
reproductive potential. Similarly, intrasexual variation in reproductive success is
also associated with social rank in many other gregarious mammals (e.g.
primates: Altmann et al. 1988; ungulates: Clutton-Brock et al. 1988; carnivores:
Creel et al. 1992). Crocuta clans are structured by rigid dominance hierarchies,
and an individual’s position in this hierarchy determines its access to food and
other resources (Kruuk 1972; Tilson and Hamilton 1984; Frank 1986a; Henschel
and Skinner 1987). In this species, female social rank significantly influences
several aspects of female reproductive success. High-ranking female hyenas
begin breeding at younger ages, are more frequently able to support pregnancy
and lactation concurrently, experience shorter intervals between litters, and their
offspring are more likely to survive to adulthood than are those of lower-ranking
females (Frank et al. 19953; Holekamp et al. 1996). Considering the significant
variation observed in female reproductive success in this species, if males can
assess reproductive value of females, then sexual selection theory predicts that
male hyenas should be highly selective in their choice of mates. Male behavior

patterns that increase the probability that a particular male will mate with a

10

certain individual female are indicative of male mate choice (Halliday 1983).
Here, I looked for evidence of male mate choice in Crocuta by determining
whether males interact differently with potential mating partners, and whether
variation in these interaction patterns is correlated with male reproductive
success. My specific goals were (1) to examine predictions of modern sexual
selection theory in a socially role-reversed mammal, the spotted hyena, in which
females vary significantly in their quality as mates, (2) to further elucidate the
mating system of this species, and (3) to describe male-female interactions not
previously reported in the literature, and to attempt to elucidate the functional
significance of a subset of these.

To achieve these goals, I utilized demographic and behavioral data
collected over 11 years by researchers on the Mara Hyena Project, initiated in
1988 by Drs. Kay Holekamp and Laura Smale. Although I personally spent over
two consecutive years in the field between June 1997 and August 1999 collecting
detailed accounts of courtship and sexual behavior, sample sizes for these
behaviors were greatly increased by supplementing my own data with archived
data from several years. I also made use of paternity data painstakingly
generated by many workers from several labs in two countries (Libants et al.
2000; Engh et al. 2002; SM. Funk and AL. Engh, unpublished data). Linking
patterns of male Crocuta behavior and mate choice with male reproductive
success could not have been achieved without these essential paternity data.
Therefore, I have elected to use the term “we” throughout the following three data

chapters to indicate the collaborative nature of this intensive study.

11

The first step I took toward evaluating male mate choice was to examine
basic association patterns among male and female spotted hyenas. My goal in
Chapter 2 of this dissertation was therefore to test the hypotheses that, because
females vary with respect to their quality as mates, (1) male hyenas should
preferentially associate with high-quality females, and (2) male preferences
should additionally be affected by female reproductive state. I predicted that male
hyenas should associate most closely with high-ranking females and with
females most likely to be fertile. Interactions between pairs of adult males and
females were observed intensively in our study population of free-living hyenas in
Kenya during a 7-year period. During this time, multiple females were known to
conceive around the same time, creating opportunities for male choice.
Behavioral data extracted from focal animal surveys (Altmann 1974) indicated
that males initiate and maintain most associations with females. I subsequently
evaluated the strength of male-female associations by calculating association
indices (using the twice-weight index of association: Cairns and Schwager 1987)
for all male-female pairs. Results unambiguously showed that male hyenas
associate most closely with high-ranking females and with females that are
likeliest to be fertile. High- and middle-ranking males associate most closely with
high-ranking females, but low-ranking males associate equally closely with
females in all rank categories, suggesting that males in different rank categories
may be using different reproductive strategies to attract females or obtain them
as mates. Using molecular markers to determine paternity of cubs born during

the study period, I found that sires associate more closely with the mothers of

12

those cubs than do non-sires, particularly during the last months before
conception. These association data indicate that male spotted hyenas do indeed
exhibit selective mate choice, and that they prefer females likeliest to maximize
male reproductive success, thus confirming both hypotheses addressed in this
chapter. This chapter was published recently in Behavioral Ecology and
Sociobiology, so future reference in Chapters 3, 4, and 5 to the association data
reported in Chapter 2 will cite the published manuscript of Szykman et al. (2001).
The association data provide only partial insight into individual male
behavior and decision-making regarding mate choice, so I next examined
courtship and mating behavior in Crocuta in more detail in Chapter 3. The goals
of this chapter were to document male and female sexual behavior, including
courtship and mating, not previously described in the literature, and to evaluate
the hypothesis that male affiliative behaviors reflect mate choice. I predicted that
males would exhibit more affiliative behaviors towards high-ranking females and
toward females closest to their time of known conception. Furthermore, given a
choice among multiple females of various ranks and in various reproductive
states, a male should actively choose to maintain associations with females
likeliest to maximize his reproductive success. In Chapter 3, I first describe
courtship and mating behavior in Crocuta, and then test the predictions outlined
above. I found that males direct most affiliative behaviors toward females
approaching conception, and that male decisions about staying with an individual
female are influenced by her social rank and her reproductive state, measured as

time to conception.

13

Finally, I investigated the occurrence and function of rare male aggression
directed toward females in this species. My specific goals in Chapter 4 were to
describe the conditions under which male attacks on females occur, shed light on
the mechanisms mediating male aggression toward females, and address
possible adaptive functions of this unusual behavior. Male hyenas rarely direct
aggression toward adult females, as females are larger and socially dominant to
all immigrant males in the clan (Kruuk 1972; Frank 1986b; this study). Most
aggression directed by adult males against females occurs when coalitions of
two or more males attack a single adult female, who responds by defending
herself and sometimes fighting back (Kruuk 1972; this study). Kruuk (1972)
described these coalitionary attacks by males against females, and referred to
them as ‘baiting’. Although Kruuk doubted that these attacks were related to
sexual behavior in hyenas, I included analyses of baiting in this dissertation
because my data suggest these attacks may play some role in hyena courtship.
Specifically, daily baiting rates vary significantly with female reproductive state,
and the highest rates of male attacks on females are observed during the period
when she is most likely to conceive. The adaptive significance of male
aggression against females in this species remains unknown, but my results are
most consistent with an hypothesis suggesting that male aggression informs
females about male fitness.

The dissertation concludes with a discussion and integration of the three
data chapters. In summary, sexual behavior in the spotted hyena is unique

among mammals. Masculinization of female genitalia and social dominance of

14

females appear to influence the structure of male courtship and other sexual
behaviors in this species. Furthermore, my data suggest that male hyenas make
choices about with which females to associate, and with whom they should
ultimately mate, apparently based on the males’ ability to assess variation in
female reproductive value. The cues males use to make these assessments
remain unknown, but I will speculate on these and other unanswered questions

in my conclusions in Chapter 5.

15

Chapter 2
ASSOCIATION PATTERNS AMONG MALE AND FEMALE SPOTTED HYENAS
REFLECT MALE MATE CHOICE

INTRODUCTION

Because female animals produce larger gametes and often invest more time and
energy in parental care than do males, limited access to females usually limits
male reproductive success (Parker et al. 1972; Trivers 1972; Andersson 1994).
Females can, therefore, generally be highly selective in their choice of mates.
However, Darwin (1871) recognized that males might also sometimes have
opportunities to choose the most attractive females from an array of mates.
Indeed, in various animal species, male preferences have been documented for
older females with more experience (Burley and Moran 1979), younger females
likely to be virgins (Forsberg 1987), larger and more fecund females (Gwynne
1981; Verrell 1985; Sargent et al. 1986; Verrell 1989; Berglund and Rosenqvist
1993; Verrell 1995), females with higher reproductive potential (Berger 1989;
Schwagmeyer and Parker 1990; Schwagmeyer 1995), and females exhibiting
superior parental care (Verrell 1990; Solomon 1993). In general, male mate
choice is expected when males invest heavily in reproduction or when there is
high variation in the quality of potential mates (Gwynne 1981; Parker 1983;
Owens and Thompson 1994). Here we inquire whether male mate choice occurs

in a long-lived social mammal, the spotted hyena. Although males do not invest

16

heavily in reproduction in this species, females vary greatly in their ability to
produce surviving offspring.

Spotted hyenas are gregarious carnivores that live in social groups called
clans. Each clan contains one to several matrilines of adult females and their
offspring, as well as one to several adult immigrant males. Crocuta clans are
fission-fusion societies in which individuals travel, rest, and forage in subgroups
that can change in composition from day to day, or even hour to hour. Although
associations among clan members shift frequently, clans are nevertheless rigidly
structured by hierarchical rank relationships (Kruuk 1972; Tilson and Hamilton
1984; Frank 1986a; Holekamp and Smale 1990; Holekamp and Smale 1993),
and an individual's social rank determines its priority of access to food (Kruuk
1972; Tilson and Hamilton 1984; Frank 1986a). Female hyenas are philopatric,
but all males disperse to new clans sometime after reaching reproductive
maturity at two years of age (Smale et al. 1997). Adult females are socially
dominant to all adult males not born in the clan (Kruuk 1972; Smale et al. 1993).

Female reproductive success varies enormously with social rank in this
species (Frank et al. 19953; Holekamp et al. 1996). In fact, reproductive skew
among female Crocuta is greater than that observed among females of any other
plural-breeding mammalian species for which comparative data are available
(Holekamp and Smale 2000). High-ranking female hyenas begin breeding at
younger ages, are more frequently able to support pregnancy and lactation
concurrently, experience shorter intervals between litters, and their offspring are

more likely to survive to adulthood than are those of lower-ranking females

17

(Holekamp et al. 1996). Considering the significant variation observed in female
reproductive success in this species, if males can assess reproductive value of
clan females, then sexual selection theory predicts that male hyenas should be
highly selective in their choice of mates, preferring high- over low-ranking
females. Male behavior patterns that serve to increase the chances that a
particular female will mate with a certain individual male are indicative of mate
choice (Halliday 1983; Bercovitch 1991). Here we looked for evidence of male
mate choice in Crocuta by evaluating variation in interactions observed between
males and their potential mating partners, and then inquiring whether variation in
these interaction patterns is correlated with male reproductive success. One
mechanism by which a male might be able to increase his chances of mating
with a particular female is the development of an affiliative relationship with her
(Smuts 1983; Bercovitch 1995). In this paper, we present data addressing the
questions of whether male hyenas prefer to associate with high-ranking females,

and whether male preferences vary with female reproductive state.

METHODS

SII'I || I'lll'

We conducted this study in the Talek area of the Masai Mara National Reserve,
Kenya. This is an area of open, rolling grasslands grazed year round by large
concentrations of several different ungulate species. The subject population was

one large Crocuta clan inhabiting a home range of approximately 65 kmz. We

18

monitored Talek hyenas continuously from June 1988 to June 1995. Throughout
this period, we conducted behavioral observations of Talek hyenas 23 to 31 days
per month, except during April 1991, when observers were only present on 14
days. Between 1988 and 1995, the Talek study clan varied in size from 50 to 70
individuals, consisting on average of 12 adult immigrant males (range: 6-20) and
20 adult females (range: 16-28) and their juvenile offspring. We identified all
hyenas in the Talek clan individually by their unique spot patterns, and we
determined their sex from the dimorphic glans morphology of the erect phallus
(Frank et al. 1990).

We established mother-offspring relations on the basis of regular nursing
associations. We assigned birth dates to litters by estimating cub ages when they
were first observed above ground at natal or communal dens. Cub ages could be
estimated to :l: 7 days based on the pelage, size, and other aspects of cub
appearance and behavior. We determined social ranks of individual hyenas
based on outcomes of several thousand dyadic agonistic interactions during
which one individual exhibited submissive behavior to the other, but regardless of
whether or not the submissive behavior was elicited by aggression from the
social partner (Holekamp and Smale 1990; Smale et al. 1993). Adult females and
adult immigrant males were ranked in separate intrasexual dominance
hierarchies. By convention we assigned the highest-ranking (alpha) individual in
each hierarchy a rank position of one. Only immigrant males present in the clan
for longer than 6 months were included in the male hierarchy. For some analyses

we divided each intrasexual dominance hierarchy into equal thirds: high, middle,

19

and low ranks representing ranks 1-7, 8-14, and 2 15, respectively.

We conducted daily behavioral observations from vehicles between 0530
and 0900 h and between 1700 and 2000 h. During the study period, we observed
hyenas for a total of 4407.8 hours during 15,353 observation sessions. We
initiated an observation session when we first drove up to one or more hyenas
separated from others by at least 200 m. An observation session ended either
when all hyenas moved out of sight, for example into bushes, or when we drove
on to a new location. We identified all individual hyenas present in each session,
and we excluded from analyses any sessions in which one or more unidentified
hyenas were present. We located hyenas while driving daily circuits around the
study clan’s home range, visiting the area’s high points, and scanning with
binoculars to sample all parts of the home range every day for presence of
subgroups of hyenas. We then visited each subgroup to determine its
composition. We also conducted multiple 30-minute focal animal surveys
(Altmann 1974) on each of 15 adult male and 23 adult female hyenas to compare
rates at which males and females initiated and maintained associations with
opposite-sex conspecifics. In each survey we calculated the rate at which the
focal animal approached opposite-sex conspecifics, and we also calculated the
percent total time in each survey the focal animal spent following opposite-sex
conspecifics. An approach was scored when a focal hyena moved to within 1 m
of another individual. The focal animal was considered to be following another
hyena in any observation minute during which It walked behind another individual

moving in the same direction. An hourly approach rate was calculated for each

20

focal animal as [(# of approaches / # minutes observed with opposite-sex
conspecific) * 60]. Percent time spent following was calculated as [(# of min
spent following / # min observed with opposite-sex conspecific) * 100]. We

calculated mean values for each individual observed during multiple surveys.

We calculated an association index (Al) for each male-female pair using the
twice-weight index of association (Cairns and Schwager 1987):

(A 1' B)together/ I (Awithout B) + (Bwithout A) + (A 1' B)together]
where (Awithout 3) represents the number of observation sessions in which female
A was observed but male B was not present, (Bwnhout A) represents the number of
observation sessions in which male B was present but female A was absent, and
[(A + Bhogetherj represents the number of sessions in which both female A and
male B were present. Als were calculated only for females observed throughout
complete reproductive cycles within the study period. A complete reproductive
cycle began at the time of conception of one litter in which at least one cub
survived to weaning, and ended at the conception of the subsequent litter. For
cases in which paternity of cubs was known, we examined association patterns
relative to the date of a female’s conception. In other analyses, we divided each
complete reproductive cycle into five reproductive states as follows.

Pregnancy (P): The duration of pregnancy in Crocuta is 110 days

(Schneider 1926; Kruuk 1972), so we calculated conception dates by subtracting

110 days from estimated birth dates to obtain conception dates that are accurate

21

to :l: 7 days. However, since the extent of variation in gestation length in this
species is not currently known, for analyses in which we focused on male-female
interactions around the time of conception, we considered a female’s “fertile
period” as her date of conception t 1 week.

Lactation (L1, L2, L3): The lactation period began with the birth of a litter
and ended with its weaning. Weaning conflicts and cessation of nursing indicated
when cubs were weaned (Holekamp et al. 1996). Weaning conflicts between
cubs and their mothers are easily observed in this species due to the
characteristic begging behavior and loud whining exhibited by cubs at this time.
We recorded all weaning conflicts in field notes as critical incidents ("all-
occurrence" sampling, Altmann 1974). In determining weaning dates, we
searched all field notes for observations of nursing behavior when mother and
cub were found together. If mother and cub were not found together frequently
after the last observed nursing bout, the weaning date was identified as being
midway between the last nursing bout and the next sighting of mother and cub
together without nursing (Holekamp et al. 1996). All weaning dates used in these
analyses were accurate to within 1 10 days. Lactation periods varied in length
among Talek females from 7 to 21 months (Holekamp et al. 1996; Holekamp and
Smale 2000). Therefore, to better compare association patterns among females
with different lactation period lengths, we divided the total period of lactation for
each female into three intervals of equal length, represented as L1, L2, and L3.

Other (0): Females assigned to this condition had weaned one litter but

had not yet conceived their next litter. Little is known about the estrous cycle of

22

the spotted hyena. Matthews (1939) and Lindeque (1981) suggest that estrous
cycles in this species recur every 14 days. However, there are no apparent
morphological indicators of estrus in female spotted hyenas as there are, for
example, in many cercopithecine primates (reviewed in Melnick and Pearl 1987),
and copulations are rarely observed in Crocuta. Although most females weaned
one litter weeks or months before conceiving another, females of all social ranks
were occasionally observed to conceive while still nursing a previous litter
(Holekamp et al. 1996). Here, females known to be pregnant while still nursing a
previous litter were assigned a reproductive condition of ‘P’ (see above).

The ability of male hyenas to access or monopolize fertile females should
vary with the degree to which estrous cycles are synchronized among clan
females. Therefore we estimated the extent to which female fertile periods
overlapped during the study period, by assuming that such overlap was possible
if a particular conception occurred within 2 weeks of conception in another
female. We considered asynchronous conceptions to be those occurring more
than 2 weeks before or after any other conception known to occur during the
study period.

Most rank positions in the female dominance hierarchy were occupied by
multiple individuals during the course of this study. We therefore calculated
overall mean association indices for all rank positions in the female hierarchy by
summing AIs across all female reproductive states for all females holding each
rank, and dividing by the total number of Als. We then compared mean Als of

male-female dyads among social rank categories and among female

23

reproductive states. Using only associations observed between 1990 and 1995
for a subset of females bearing litters during this period, we also compared
female associations with known sires to those with males known not to have
sired their litters.

In addition to calculating association indices, we also calculated the mean
number of males present with each female and compared these values among
female reproductive states and categories of female social rank. Finally, we were
interested in comparing male-female associations in Crocuta with the male-
female “consortships” observed in other mammals. Therefore we determined the
percent of all sessions in which each male was observed during which he was
found alone with a particular female. To do this, we counted the number of
sessions a pair was seen alone together, without any other hyenas present,
divided that value by the total number of sessions the male was observed, then
multiplied the result by 100. We did this for each month-long interval surrounding

conception of the female’s litter.

i ' ' t' I
We used microsatellite markers identified from Crocuta DNA to determine male
parentage in the Talek clan. Between 1990 and 1995, we collected blood from
185 Talek hyenas. Sampling was conducted after individuals were anaesthetized
with Telazol (2.5 mg/kg) delivered from a COz-powered rifle in a lightweight
plastic syringe. We then immediately extracted DNA from blood using Puregene

kits (Gentra Systems, Inc.), and stored the extracted DNA in liquid nitrogen.

24

We considered all adult males present in the clan at the time of conception
as potential sires. Paternity assignments were based on 13 microsatellite
markers described elsewhere (Libants et al. 2000; SM. Funk and AL. Engh,
unpublished data). Paternity was determined using three approaches. First,
males were excluded as potential fathers if their genotypes were inconsistent
with those of offspring; that is, if each putative father-cub pair did not share an
allele at every locus surveyed. Paternity was also evaluated by testing
significance of pairwise relatedness (rxy: Queller and Goodnight 1989), and by a
maximum-likelihood analysis (program ‘Cervus’) which compared likelihoods of
paternity for all non-excluded males (e.g., most likely fathers, second most likely
fathers, etc., Marshall et al. 1998). A male was only assigned as the sire of a
particular cub if exclusion probability (Chakraborty et al. 1988) exceeded 0.99, if
the calculated r,,y value did not differ significantly from 0.5, and if the maximum
likelihood method assigned paternity at a confidence of 95%. If two of the three
conditions were not satisfied, then we considered paternity unresolved. A more
comprehensive analysis of paternity in the Talek clan is presented by Engh et al.
(2002)

We analyzed Association Index data using 1-way, 2-way and repeated
measures analysis of variance (ANOVA). For analyses in which we compared
Als within a male-female dyad across months, we treated Als generated monthly
by each pair as repeated measures. We evaluated relationships among smaller
subsets of the data that were not normally distributed, using Mann-Whitney U

tests. We calculated correlation coefficients (Spearman’s R) to ascertain whether

25

mean male-female Als varied with female social rank. We presented mean
values as 3(- i SE. We considered differences between groups to be statistically

significant when P<0.05.

RESULTS

EEEII'IIIII Il'll
We calculated overall Als for 40 immigrant males observed together with 19 adult
females. Association indices ranged from 0 to 1, with a mean of 0.062 :I: 0.002.
Social ranks ranged from 1 to 20 for females and from 1 to 24 for males. Overall
mean Als between males and females varied significantly with female social
rank, with males associating more closely with high-ranking than with low-ranking
females (Figure 2.1; Spearman rank correlation: Rs=-0.760, N=19 ranks,
P<0.001).

We then examined the effects of female reproductive state and rank
category on patterns of association. Females were observed over 40 complete
reproductive cycles (52:21 :I: 0.1 cycles per female). We observed individual
females in each reproductive state for 0-844.6 hours: 7p=66.2 :t 14.8, YL1=120.5
:t 22.1, 7L2=84.8 :I: 18.2, 7L3=56.0 :t 12.4, 70:24.5 i 6.8 hours per female.
Maximum Als reached 1.00 in all states except ‘P’ and ‘L1’, which had the lowest
maximum Al values of 0.368 and 0.500, respectively. Males not only associated
most closely with the highest-ranking females when females were grouped into

three rank categories (Figure 2.2; ANOVA: F2,2650=58.325, P<0.001), but also

26

0'100 _ -- R8 = -0.760

 

0.075 —

 

 

 

 

“l

g l.

.2: 0.050 — 1 “ i

j i

< 0.025 .— i i

I

0000 I J l I
' 0 5 10 15 20

Female social rank

Figure 2.1. Mean (:l: SE) intersexual association indices plotted as a
function of female social rank. By convention, the highest rank possible is
one. Each point represents the mean Al for all females who held a
particular rank with all males observed with each female while she held
that rank. N=2665 associating pairs involving 19 females and 40 males
associating during all female reproductive states.

27

Female social rank

 

 

 

 

 

0-15 — I High
El Middle
Low

x ._

a, 0.10

.0

.E

C

.9

+0

.9?

0 .

8

a, 0.05 - V 9'
,5 /

“ a i
7 7
7 7
/ Z
/ 7
/ /
a a
B 1 3 l

0.00 ' ~

P L1 L2 L3 0
Female reproductive state

Figure 2.2. Mean association indices for male-female pairs measured
across all female reproductive states and grouped by female social rank.
Female reproductive states are as follows: P=pregnancy; L1, L2, L3=first,
second, and third trimesters, respectively, of the lactation period; O=other
states, when adult females are neither pregnant nor lactating. Rank
categories: high=ranks 1-7, middle=ranks 8-14, Iow=ranks 15 and higher.
Sample sizes are as in Figure 2.1.

28

with females that were most likely to be fertile, during states ‘L3’ and ’0’ (Figure
2.2; F4,2650=12.959, P<0.001). Moreover, there was a significant Interaction
between female rank and reproductive state, indicating that males associated
most closely with high-ranking females at times when these females were most
likely to be fertile (Figure 2.2; 2-way ANOVA: F8,2650=7.565, P<0.001).

Male social rank also influenced intersexual association patterns in
hyenas. High-ranking males associated more closely with high-ranking females
than did lower-ranking males (Figure 2.3; 2-way ANOVA: F4,2656=5.150,
P<0.001). In fact, both high-ranking males and high-ranking females associated
more closely overall with members of the opposite sex than did lower-ranking
individuals (males: F2,2656=45.458, P<0.001; females: F2,2556=18.447, P<0.001).
Post-hoc tests revealed that both high- and middle-ranking males associated
most closely with high-ranking females (Figures 2.3A and B), while low-ranking
males tended to associate more or less equally with females in all three rank
categories (Figure 2.30; high-ranking males: F2,1341=34.917, P<0.001; middle-
ranking males: F2,1059=15.767, P<0.001; low-ranking males: F2,255=1.291,
P=0.28).

A female Talek hyena may travel alone, with kin, with a single adult male,
or with a sizeable entourage of males. The average number of males present
with an individual female varied with female reproductive state (Figure 2.4;
ANOVA: F4,4793=10.347, P<0.001) but not with female social rank (F2,4795=1.229,

P=0.29). The greatest number of males associated with females during the

29

Figure 2.3. Mean (1: SE) association indices between females in each rank
category and (A) high-ranking males, (B) middle-ranking males, and (C) Iow-
ranking males. Rank categories are as in Figure 2.2 and are the same for both
sexes. Sample sizes are as in Figure 2.1.

30

Association index

Association index

Association index

0.12

0.09

0.06

0.03

0.00

0.12

0.09

0.06

0.03

0.00

0.12

0.09

0.06

0.03

0.00

 

 

High Middle Low

1

 

 

High Middle Low

 

 

 

High Middle Low

Female social rank

31

30'

Number of males / female / session

I

P L1 L2 L3 O

 

 

Female reproductive state

Figure 2.4. Mean number of males observed per session with each female
across all female reproductive states. N=4798 sessions for which 19
females were observed with 40 males. Bars sharing same letters represent
means not significantly different from one another as determined by
Tukey’s post-hoe comparisons.

32

reproductive states ‘L3’ and ‘0’, when females were most likely to conceive, but
also during pregnancy (Tukey’s post-hoc comparisons).

To evaluate the roles played by adults of each sex in initiating
associations, we calculated rates of behavior performed during 346 focal animal
surveys on 15 adult males, and during 127 focal animal surveys on 23 adult
females, for which at least one adult member of the opposite sex was present
(males: 32:23.1 :1: 4.3 focal animal surveys / individual; females:7=5.5 :I: 0.7 focal
animal surveys / individual). Males approached females at significantly higher
rates than vice versa, and in fact, females hardly ever approached males (males:
0.036 d: 0.007 approaches / hour; females: 0.002 :t: 0.001 approaches / hour;
Mann-Whitney U test: U1=16.000, P<0.001). Similarly, females rarely followed
males, and males spent a significantly greater percent of their focal animal time
following females than females spent following males (males: 8.6 :I: 1.7 % of their
time spent following females; females: 0.5 :t: 0.2% of their time spent following
males; Mann-Whitney U test: U1=10.000, P<0.001). Thus, associations between
male and female spotted hyenas appeared to be initiated and maintained almost

exclusively by males.

WWW

Dates of 105 fertile periods were known during the study period from 1988 to
1995 for 35 female hyenas, so we evaluated reproductive synchrony by
measuring overlap among these fertile periods ()-(=2.7 1 0.2 fertile periods per

female). We found that 56.3% of fertile periods occurred when no other Talek

33

female was concurrently fertile (Figure 2.5). Although up to five females were
observed to be fertile concurrently during the study period, this occurred only
once, and on average, only 1.6 1: 0.1 females were fertile concurrently. Thus,
reproductive synchrony among Talek females was generally low throughout the
study period. (Addendum. In August of 1998, there was a 2-week period when
the fertile periods of nine females overlapped. This was highly unusual, and was
possibly due to atypical ecological conditions experienced as a consequence of

an El Nifio event.)

6 . . 'll _ .
We next contrasted intersexual associations involving males who sired cubs with
those of males who failed to sire offspring. We examined Als during month-long
intervals surrounding the conception of a litter for which the father was known via
paternity analysis for a subset of eight females, each of which had a complete
reproductive cycle preceding that conception. Males known to sire the offspring
of these females during these particular reproductive cycles associated
significantly more closely with their mates than did other males (Figure 2.6A; 2-
way repeated measures ANOVA: F1,35=14.064, P<0.001), particularly near the
time of conception (Figure 2.6A; F12,85=1.961, P=0.038). Finally, we evaluated a
male's ability to maintain exclusive access to a particular female by examining
the percent of all observation sessions in which he was observed during which
he was found alone with that female. Sires spent significantly more time alone

with their future mates (Figure 2.6B; 2-way repeated measures ANOVA:

34

% Occurrence

 

 

# Females concurrently fertile

Figure 2.5. Synchrony among 105 fertile periods of 35 females observed
between 1988 and 1995. Sample sizes represent the number of 2-week
intervals during which females were concurrently fertile.

35

Figure 2.6. (A) Association indices for eight females with the eight males who
sired (solid bars) their litters conceived during interval ‘0’ compared with Als for
these same females with 15 males present concurrently who did not sire (open
bars) their litters. Interval ‘0’ represents the date of conception 1: 2 weeks.
Remaining intervals represent month-long periods surrounding the date of
conception. (B) Percent of observation sessions during which sires and non-sires
were found alone with females, calculated as [(the number of sessions in which
the pair was found alone/the total number of sessions in which that male was

observed)*100].

36

Association index

% Sessions pair found alone

0.30 ' a Non-sires Conception
' Sires /
0.24 F

0.18—

0.12”

I

0.06

 

 

IIIII I III
000 5 0 5

' -10 -

Conception

5 _.
D Non-Sires /

' Sires

 

 

.3
l

 

w l'
IIIIIIIIIIIII
-10 -5 0 5

Months surrounding conception

37

F1,36=6.501, P=0.012) than did males failing to sire offspring, particularly during
the months immediately preceding conception (Figure 2.6B; F12,85=1.857,

P=0.05).

DISCUSSION

Our association data indicate that male hyenas appear to assess relative
reproductive value of individual females, and that they may allocate their own
reproductive effort among clan females based on this assessment. Male hyenas
associate most closely, and in the greatest numbers, with females that are most
likely to be fertile. Males of other mammalian species similarly prefer to associate
with females when they are most likely to conceive (e.g. baboons, Papio
cynocephalus: Seyfarth 1978a, b; Smuts 1985; bison,: Komers et al. 1994).
Since neither visual nor acoustic signals emitted by female hyenas appear to
vary with their reproductive condition, males presumably evaluate female
reproductive state using olfactory cues. Although large numbers of males
associated with females during pregnancy as well as during periods of maximal
fertility (Figure 2.4), exclusive associations between females and sires were not
observed after the first month of gestation (Figure 2.6B). Since females and sires
do not remain together until cubs are born, exclusive male-female associations in
Crocuta apparently do not function to reduce infanticide, as occurs in some
primates (e.g. Sterck et al. 1997). Instead male hyenas may be unable to detect

the early stages of pregnancy in females, or they may associate with pregnant

38

females to increase the probability of mating with them in the future. Studies of
various primates show that a male can increase his chances of mating with a
particular female by associating with her, even when she is not fertile (e.g. Smuts
1983; Smuts 1985; Hill 1987; Bercovitch 1995), and this may also hold true in
Crocuta. Male hyenas might also continue to associate with their mates for a
short time after conception to guard them from rival males, as occurs in lions
(Panthera Ieo: Packer and Pusey 1982).

Like spotted hyenas, lions are polygynous carnivores in which individual I.-
males consort with fertile females. Reproduction is highly synchronous within lion
prides, and several females may be in estrus at once (Schaller 1972; Bertram
1975). Although rival males are more likely to be found near a consorting pair
when availability of other estrous females is low, consorting lion pairs are found
alone in approximately fifty percent of sightings (Schaller 1972; Packer and
Pusey 1982). By contrast, consorting hyena pairs are found alone in less than
five percent of sightings (Figure 2.68), so it appears that male hyenas may be far
less successful than male lions at monopolizing access to fertile females. Male
lions often ‘herd’ females away from other pride members (Packer and Pusey
1982), but female Crocuta are dominant to males, so male behavior does not
constrain female movement in this species.

In our study population reproductive synchrony among females is low
(Figure 2.5), as compared to lions, for example, where 50-75% of females in a
pride may be concurrently in estrus (Schaller 1972). Low reproductive synchrony

among female hyenas should theoretically permit male hyenas to associate

39

 

closely with fertile low-ranking females without foregoing mating opportunities
with higher-ranking females. Although we found that males associated with
dominant females to the greatest extent at times when those females were most
likely to be fertile, most males preferred high-ranking females during all stages of
the female reproductive cycle. This supports the hypothesis that males actively
choose among females, and indicates that males attempt to maximize their
chances of mating with females having the greatest potential to produce
surviving offspring, and thus also to maximize male reproductive success.
Similarly, in many primates, males show mating preferences for high-ranking
females (reviewed in Robinson 1982; Silk and Boyd 1983).

In groups of canids and social mongooses, the highest-ranking individuals
of each sex prefer the highest-ranking opposite-sex individual as their resting and
grooming partners, closest associates, and mates (Rasa 1977; Derix and van
Hooff 1995). In the current study, social rank of both male and female hyenas
influenced intersexual patterns of association. Both high— and middle-ranking
males associated most closely with the highest-ranking females, but low-ranking
males failed to exhibit a preference for higher-ranking females. Low-ranking
males may thus either be less adept than high-ranking males at assessing
reproductive value of clan females, or low- and high-ranking males may utilize
different reproductive strategies.

Our comparison of association patterns of sires and non-sires suggests
that, when a male hyena invests time with one prospective mate, this may pay off

in terms of increasing his chances of mating with that female. However, our data

40

also suggest that time spent with one particular female represents a substantial
investment that might compromise the male’s ability to explore alternative
reproductive possibilities. Sires spend more of their time in exclusive isolation
with their future mates than do non-sires, particularly during the last months
before conception. One obvious benefit to associating with a female in isolation is
exclusive access to that female when she is ready to mate. If a male associating
with a particular female can keep her away from rivals, he has a greater chance
of siring that female’s offspring. Immigrant male Crocuta rarely fight among
themselves (Kruuk 1972; Frank 1986b), but the fission-fusion nature of hyena
society may permit males to use physical separation with prospective mates as
an effective alternative to combat, as also occurs in lions (Packer and Pusey
1982). Interestingly, although female hyenas in the current study were highly
attractive to large numbers of males when females were likeliest to conceive
(Figure 2.4), known sires were able to maintain exclusive access to females
relatively often during these periods of maximal fertility (Figure 2.6B).

Close and exclusive intersexual associations were related to paternity in
the current study, but the factors causing these patterns remain unknown. That
is, a male might make himself more attractive to a particular female by spending
more time with her, and thus enhancing his own chances of siring her cubs.
Alternatively, a male might spend more time with a female because she has
already chosen him to sire her cubs. Thus our data do not rule out the possibility
that close intersexual associations in Crocuta reflect mate choice by females as

well as by males. Indeed, female hyenas make exceptionally heavy and

41

prolonged investment in each of their cubs (Hofer and East 1993; Hofer and East
1995; Holekamp et al. 1996), and receive no help from males in rearing young,
suggesting that females should be highly selective in their choice of mates.
Nevertheless, the reproductive performance of high-ranking females is far
superior to that of low-ranking females (Holekamp et al. 1996), generating
extreme variance in female reproductive success, and thus mate quality, as a
function of social rank. Therefore, male Crocuta ought to be choosy, and our data
are consistent with this hypothesis. In fact, we found that male-female
associations in this species are almost exclusively initiated and maintained by
males. In contrast, intersexual consortships in other social carnivores and in
primates frequently involve a COOperative effort by both members of the pair to
maintain proximity (Stern and Smith 1984; Small 1990). In these other mammals
consorting pairs are observed to copulate, and consorting males father most
offspring (Packer and Pusey 1982; Bercovitch 1987; Small 1990). We rarely
observe hyenas mating so we don’t yet know to what extent copulation frequency
in Crocuta can be predicted by male-female association patterns (but see
Chapter 3). However, unusually close intersexual associations are correlated
with paternity in Crocuta, as they are in other gregarious mammals.

Finally, our data allow us to contrast male-female association patterns in
Crocuta with patterns previously documented for female-female and mother-
offspring associations in this same species. Using the same AI calculations as
those utilized here, Holekamp et al. (1997) showed that association patterns

among adult female Crocuta are strongly influenced by both kinship and rank.

42

Adult females associate more closely with their adult female kin than with
unrelated adult females, and high-ranking mother-offspring pairs associate more
tightly than do low-ranking dyads. When we compare absolute values of
association between that earlier study and this one, we find that, on average,
adult males associate with adult females at about the same levels as adult
females associate with other unrelated adult females. Furthermore, males sirlng
litters associate with their mates almost as closely as do mothers with their own
offspring, with Als over twice as high as those of males failing to sire those cubs.
Although male-male association patterns have not yet been examined in this
species, our results suggest that adult males may form their closest associations

with the females with whom they eventually mate.

43

Chapter 3

COURTSHIP AND MATING IN FREE-LIVING SPOTTED HYENAS

INTRODUCTION

Courtship and mating are two aspects of sexual behavior required for the
successful fusion of egg and sperm in many animals, and the peculiar courtship
and mating behaviors of the spotted hyena are the focus of this chapter. The
functions of animal courtship behavior are both numerous and diverse. Courtship
behaviors may function to facilitate recognition by potential mates of species and
sex (Littlejohn and Loftus-Hills 1968; Ptacek 2000), or to synchronize
physiological and behavioral processes necessary for fusion of gametes (Crews
1975; Verrell 1997). Courtship behaviors may also indicate mate quality, and
permit opportunities for mate choice (Clutton-Brock et al. 1982; Bradbury and
Gibson 1983; Karino 1995; Hoikkala et al. 1998). Finally, courtship may help
break down barriers that prevent mates from coming together to copulate,
particularly in solitary or aggressive species (Caro 1994). Courtship and mating
in carnivorous animals can be potentially dangerous, since these behaviors bring
together mates that are aggressive and well armed.

Mating systems in mammalian carnivores range from monogamous to
highly polygynous or promiscuous, and successful mating is achieved with a

variety of reproductive strategies (Schaller 1972; Ewer 1973; Clutton-Brock 1989;

44

Creel and Macdonald 1995). In general, mammalian mating systems are
determined by the dispersion of resources and mates (Emlen and Oring 1977;
Clutton-Brock 1989) and the degree of parental care required to maximize
reproductive success for both sexes (Davies 1991). Male and female sexual
strategies vary as access to, and competition over, mates differ between solitary
and group-living species.

Whether solitary or gregarious, monogamous or polygynous, the
challenges to achieving successful copulation among mammalian carnivores
begin with finding a receptive conspecific and convincing this individual to
consider mating. This latter task may require extensive courtship and moderate
persuasion on the part of the male and various physiological and behavioral
responses on the part of the female. The male may attempt to restrain the female
by grabbing her from behind with his front paws and teeth at the time of
copulation, as observed in most felids and mustelids (Schaller 1972; Ewer 1973;
King 1989; Caro 1994). This behavior may serve to keep the female from biting
the male, to elicit an immobilization response in the female similar to that seen in
cubs when being carried by their mothers (Schaller 1972; Vargas and Anderson
1998), or to help the male orient his body properly for achievement of
intromission (King 1989; Beaver 1992). Some males may injure or even kill
females while trying to mate with them (e.g. elephant seals, Mirouanga elaphus:
Cox and LeBoeuf 1977; LeBoeuf and Mesnick 1990).

Successful copulation requires physiological and behavioral coordination

between males and females. Many females in the felid, mustelid, ursid,

45

herpestid, and procyonid families of carnivores are induced ovulators, and
ovulation is triggered by repeated or lengthy bouts of copulation (Hayssen et al.
1993). For example, mating pairs of lions will copulate every 15-30 minutes for 2-
4 days (Schaller 1972; Bygott et al. 1979; Packer and Pusey 1983). Canids are
not induced ovulators, but mating pairs remain locked together in a ‘copulatory
tie’, which has been reported for all canid species in which copulation has been
observed (Asa and Valdespino 1998). A copulatory tie or look is a mechanical
connection formed between the vagina and the penis which holds the mating pair
together following penile insertion (Dewsbury 1972). In canids, this tie is caused
by a swelling of the bulbis glandis, near the base of the penis (Beach and
LeBoeuf 1967). Unlike the sexual behavior of many canids and felids, which has
been studied extensively, little is known about the coordination of behavioral
events required for successful copulation in hyenids. This chapter aims to
elucidate sexual behavior in a peculiar gregarious member of the carnivore family
Hyaenidae, the spotted hyena.

Spotted hyenas are particularly interesting subjects for the study of sexual
behavior for two reasons. First, spotted hyenas live in female-dominated
societies in which females are larger and more aggressive than males, and
females are socially dominant to adult breeding males (Kruuk 1972; Smale et al.
1993). Second, female Crocuta possess unusually heavily ‘masculinized’
genitalia. The female’s clitoris is elongated and fully erectile, and her labia are
fused and filled with connective tissue, thus resembling male testes (Matthews

1939; Neaves et al. 1980; Hamilton et al. 1986). This unique female morphology

46

makes intromission by the male considerably more difficult than it is in other
mammals, and also makes copulation by force physically impossible. These odd
behavioral and physical characteristics exhibited by female spotted hyenas may
force males to make extra efforts to convince females to mate with them, and
pose several other unusual problems for male hyenas. First, the male must
determine when a female is in estrus, even though it may be dangerous for him
to approach her. Second, the male must attract a female and convince her that
he is a suitable mating partner. Third, the male must insert his erect penis into
the female’s flaccid pseudo-penis to successfully achieve intromission.

Very little is known about courtship and mating in spotted hyenas, and
even less is known about the underlying reproductive physiology, or even about
the length of female estrous cycles. Grimpe (1916) reported that a captive female
came into estrus 3 weeks after parturition when cubs were removed, and then
again 2 weeks later, though no physical or behavioral indicators of estrus were
mentioned. Matthews (1939) and Lindeque (1981) used histological analyses of
the ovaries of wild hyenas to similarly conclude that female hyenas in the wild
cycle approximately every 14 days. In other studies, estrous cycles in captive
Crocuta have been reported as 21-60 days (Schneider 1926) and 45 days
(Golding 1969). The duration of the estrous period, the length of time that a
female hyena is receptive to sexual advances by the male, remains unknown.

Kruuk (1972) was the first to conduct long-term research on spotted
hyenas in the wild. However, during almost four years of detailed observations of

hyenas in the Serengeti National Park and Ngorongoro Conservation Area in

47

Tanzania, he only made 14 observations of mounting behavior, and only five of
these appeared to culminate in intromission (Kruuk 1972). He reported that male
hyenas are always afraid of, and subordinate to, females, and that males show
little or no aggression toward females during mating. Therefore, Kruuk suggested
that the female’s cooperation is essential to the success of the copulation. He
also observed that there appears to be a surprisingly high level of tolerance
among rival males in this species, but he believed that only one of the several
males present with a receptive female is likely to mate with her. Kruuk concluded
that the “pair bond” is weak in Crocuta, and he observed no clear pattern of any
one female associating with a particular male for an extended period near the
time of conception. One goal of the present study was to replicate and elaborate
on Kruuk’s limited findings in regard to mating behavior in Crocuta.

Frank (1986b) described his own observations of male sexual behavior in
wild spotted hyenas in Kenya, and preliminarily outlined the effects of social rank
on reproduction. He divided male courtship into attendance behaviors and active
courtship. Attendance behaviors included following and consorting with a female
by a male. Active courtship behaviors, which were seen only in the context of
male-female interactions, included three types of behaviors. First, the male
behaved submissively by bobbing his head up and down as he slowly
approached the female. Second, the male experienced an approach-avoidance
conflict whereby any movement of the female would cause him to retreat, after
which the male would resume his deliberate approach. Third, the male exhibited

a unique behavior that consisted of extending his foreleg and crossing it over the

48

opposite foreleg (Frank 1986b). Frank maintained that any adult male might
follow, consort with, or occasionally court a female, but that eventually the
highest-ranking (alpha) male will exclude all other males using only mild threats,
gain exclusive access to all receptive females, and mate with them. Frank
concluded that spotted hyenas have a highly polygynous mating system, in which
the alpha male sires a disproportionately large number of cubs (Frank 1986b).

Mills (1990) observed that sexual behavior by male spotted hyenas in the
Kalahari was similar to that described by Kruuk (1972) and Frank (1986b).
Kalahari clans are smaller than clans in East Africa and usually contain only one
immigrant male who performs all observed matings (Mills 1990). Immigrant
Kalahari males investigate, court, and attempt to mate with clan females,
following the characteristic behavior patterns originally described by Kruuk
(1972). Mills further suggested that persistence might be the key to male
reproductive success, rather than intense male-male competition. Like Kruuk
(1972) and Frank (1986b), Mills did not observe any overt aggression among
males competing over direct access to a female. One of our goals in this chapter
was to quantify the time and effort that males must invest to gain mating
opportunities with receptive females.

Two recent reports from our research group indicate that Crocuta are best
described as promiscuous (sensu Clutton-Brock 1989): males mate with multiple
females, and there appears to be no lasting bond between individual males and
females after mating has occurred (Engh et al. 2002; Szykman et al. 2001).

Male-female association patterns indicate that male spotted hyenas prefer

49

females likeliest to maximize their own reproductive success (Szykman et al.
2001). Females have been observed mating with multiple (up to three) males
during a single estrous period (over four consecutive days), and males have
been observed mating with several different females over the course of several
years (Engh et al. 2002). Both observations of mating and paternity tests based
on microsatellite data confirm that males ranked 1-14 in the male hierarchy
copulate with females, and that males ranked 1-11 sire cubs. Some litters
(approximately 20%) are sired by multiple males (Engh et al. 2002).

Successful reproduction requires active participation by both sexes, and
its occurrence is profoundly influenced by female sexual behavior. Female
animals differ in the extent to which they attract mates and in how they respond
to male sexual overtures. After attempting to quantify various characteristics of
female mammals in estrus, Beach (1976) concluded that there are three different
components of female sexuality. Attractivity represents the ‘stimulus value’ of the
female to the male. This includes both behavioral and non-behavioral cues,
which are influenced by ovarian hormones, and which can be measured by the
male’s interest in the female. These cues function to attract males to the female,
and provide males with information about the female’s reproductive status.
Proceptivity reflects the female’s attraction to the male and consists of appetitive
activities directed by the female toward the male in response to stimuli received
from the male. Proceptive behaviors indicate the female’s motivation to mate,
serve to bring the male and female into proximity, and arouse the male (and

perhaps also the female) to facilitate successful copulation. The female’s specific

50

behavioral responses that are ’necessary and sufficient’ for successful copulation
indicate female receptivity, the female’s readiness to allow copulation. Although
female sexual behavior in Crocuta has not previously been investigated with
Beach’s (1976) schema in mind, we will do so here.

We seek to add to existing knowledge about sexual behavior in Crocuta
by presenting data collected during 11 years of observation of a free-ranging
group of spotted hyenas in Kenya. In this chapter, we will describe and quantify
sexual behavior in free-living spotted hyenas, including pre~copulatory affiliative
behaviors that serve to bring the sexes into proximity, as well as the postures
and actions involved in mating itself. We will examine rates of male affiliative
behaviors toward females during the months surrounding conception, report
detailed sequences of events from field observations of courtship and mounting,
and document aspects of female sexual behavior not previously well-described in

the literature.

METHODS

Sll'l lI l'lll'

Spotted hyenas are gregarious carnivores that live in social groups called clans.
Each clan contains one to several matrilines of adult females and their offspring,
as well as one to several immigrant adult males. Clans are rigidly structured by
hierarchical rank relationships (Kruuk 1972; Tilson and Hamilton 1984; Frank

1986b; Holekamp and Smale 1990; Mills 1990; Holekamp and Smale 1993), and

51

an individual's social rank determines its priority of access to food and other
resources (Kruuk 1972; Tilson and Hamilton 1984; Frank 1986b). Adult females
are socially dominant to all adult males not born in the clan (Kruuk 1972; Smale
et al. 1993). Female hyenas are philopatric, but all males disperse to new clans
sometime after reaching reproductive maturity at 2 years of age (Smale et al.
1997). Natal males may remain in their natal clans up to 38 months after puberty
(Smale et al. 1997), thus generating two types of adult males in every clan: adult
natal males and immigrant males. Natal males differ significantly from immigrant
males with respect to intrasexual aggressive behavior, intersexual courtship
behavior, and plasma testosterone profiles (Holekamp and Smale 1998). Natal
males are significantly more aggressive toward other adult males than are
immigrants, are socially dominant to all immigrants, and are the winners of all
fights with immigrants. Despite the fact that they outrank immigrant males, natal
males court females at significantly lower rates than do immigrant males, and
natal males have significantly lower plasma testosterone levels than do
Immigrant males, even when controlling for male age and body size (Holekamp
and Smale 1998).

We conducted this study in the Talek area of the Masai Mara National
Reserve, in southwest Kenya. This is an area of open, rolling grassland grazed
year round by large concentrations of several different ungulate species. The
subject population was one large Crocuta clan inhabiting a home range of
approximately 65 kmz. We monitored Talek hyenas continuously from June, 1988

to August, 1999. Between 1988 and 1999, the Talek study clan varied in size

52

from 45 to 79 individuals, not including transient individuals present in the clan for
less than 6 months. The clan consisted on average of 14 adult immigrant males
(range: 5-16) and 19 adult females (range: 12-26) and their juvenile offspring. We
identified each individual hyena in the Talek clan by its unique spot patterns, and
we determined its sex from the dimorphic glans morphology of its erect phallus
(Frank et al. 1990).

We established mother-offspring relations on the basis of regular nursing
associations. We assigned birth dates to litters by estimating cub ages when they
were first observed above ground at natal or communal dens. Cub ages could be
estimated to 1: 7 days based on their pelage, size, and other aspects of cub
appearance and behavior. We were then able to calculate conception dates by
subtracting the average gestation period of 110 1: 1 days (Schneider 1926; Kruuk
1972; ML. Weldele, pers. comm.) from estimated birthdates. We determined
social ranks of individuals based on outcomes of several thousand dyadic
agonistic interactions during which one individual exhibited appeasement
behavior to the other, but regardless of whether or not the appeasement was
elicited by aggression from the social partner (Holekamp and Smale 1990; Smale
et al. 1993). Adult females and adult immigrant males were ranked in separate
intrasexual dominance hierarchies. Only immigrant males present in the clan for
longer than 6 months were included in the male hierarchy. By convention, we
assigned the highest-ranking (alpha) individual in each hierarchy a rank position

OH.

53

We conducted behavioral observations from vehicles between 0530 and
0930 hours and between 1630 and 2000 hours. During the 11-yr study period,
we observed hyenas for 8166 hours during a total of 39,600 observation
sessions. Adult female and adult male hyenas were present together for 3477
hours during 7260 observation sessions. We initiated an observation session
when we first drove up to one or more hyenas separated from other individuals or
groups by at least 200 m. An observation session ended either when all hyenas
moved out of sight, for example into bushes, or when we drove on to a new
location. We identified all individual hyenas present in each session, and we
excluded from analyses any sessions in which one or more unidentified hyenas
were present. We located hyenas while driving daily circuits around the study
clan’s home range, visiting the area’s high points, and scanning with binoculars
to sample all parts of the home range every day for presence of subgroups of
hyenas. We then visited each subgroup to determine its composition and
identified all individual hyenas present. Note that some of the analyses presented
below utilize a subset of the source data (e.g. data from only two of the 11 total

years).

0 l' llfi'l' [ 'I III'

We wanted to quantify male affiliative behavior toward females, and vice versa,
so we recorded the following behaviors for all individuals as critical incidents

(Altmann 1974) during interactions between males and females.

54

E l [ET I'
Approach. An individual hyena walks to within 1m of another hyena.
Follow. An individual hyena walks behind another hyena within 100m, and

moving in the same direction.

Itfll' I' . I' l'

Sniff. An individual hyena approaches within 1m of another hyena, or its
urine, puts its nose down, and inhales its scent. After sniffing an adult female or
her urine, a male may display flehmen, a behavior exhibited by various mammals
and characterized by an open mouth and retraction of the upper lip, often with
head extended or elevated (Estes 1972).

Lick. An individual hyena makes contact with another hyena with its
tongue.

Greet. Two hyenas stand head-to-tail and sniff each other’s genitals.
When both members of the pair lift their inside hind legs, this is considered a
complete, or symmetrical, greeting. If only one individual lifts his/her hind leg, and
the other individual sniffs without lifting his/her hind leg, this is considered an
asymmetrical greeting (after East et al. 1993).

Countable

Although we were uncertain about the exact functions the following behaviors

fulfilled, we included them here as courtship behaviors because they were only

exhibited by adult male spotted hyenas toward adult females, and they appeared

to indicate a male’s sexual interest in a particular female.

55

Approach-Avoid. A male walks within 10m of a female, who is usually lying
down, and repeatedly alternates taking a few steps toward her, then retreating a
few steps back. It is not necessary for the female to make any motion to prompt
the male’s retreat.

Present. A male stands in front of a female, at right angles to her, with his
legs stiff and outstretched anteriorly and posteriorly, letting the female sniff his
side, ventrum, and genitalia. The male often appears nervous and is very skittish
as the female sniffs him.

Bowing display. A male approaches a female, stops, crosses one foreleg
over the other, and lowers his head. Sometimes he licks one of his extended
forelegs while they are crossed or rubs his face against his foreleg.

Paw ground. A male approaches a female and repeatedly scratches the
ground in front of her with his forepaw.

Mating
The term mating, or mating behavior, as used throughout this study,
encompasses the following three behaviors.

Mount. An adult male, positioned behind a standing female, lifts his
forepaws up on top of the female’s back or hindquarters. We defined a short
mount as any mount lasting less than 60 s and a long mount as any mount
lasting longer than 100 s.

Intromission. A male successfully inserts his penis into the female’s

urogenital canal. Observers can usually determine when intromission has

56

occurred (e.g. see Figure 3.4B). When intromission occurs during a mount, we
refer to it as copulation.

Ejaculation. It is very difficult to determine when ejaculation has occurred
in free-living hyenas, but we adopt here a suggestion proposed by SE. Glickman
and L. Koscia (pers. comm.) based on their observations of captive hyenas. That
is, it appears likely that a male hyena has ejaculated once he has finished
thrusting and remains mounted on the female with his chin and chest resting on
her back for several minutes (see Figure 3.30 below). Therefore, when a
mounted male was observed to lower his head and chest onto the female’s back

in the current study, we interpreted this to mean that ejaculation had occurred.

‘11. ‘ ‘. _-. 91-. ' v.10. ‘ Hr ‘ o u-. ‘ -.| H ‘ P‘Io. '

Largely because copulations are so seldom observed in free-living hyenas, little
is known about female sexual behavior in Crocuta. Here, we measured
attractivity of females using male-female association patterns and male affiliative
behaviors directed at females. We considered approach of males by females,
and following of males by females, as proceptive behavior in this species. We
monitored female receptivity by recording female responses to male affiliative
behavior. Three possible types of responses were recorded: neutral, positive, or
negafive.

Neutral. The female makes no motion in response to a male’s advance.
She may just lie or stand still and allow the male to sniff her, lick her, etc.

Positive. The female approaches or follows the male.

57

Negative. The female makes a threatening gesture towards the male, or
lunges at, chases, or bites him.

For all male-female affiliative interactions observed between June 1997
and August 1999, we calculated daily rates of male affiliative behavior by dividing
the number of times each behavior occurred within a pair by the total number of
hours each male-female pair was observed together during each of 7 month-long
intervals surrounding a female’s known conception, and then multiplying this
number by 24. Together, the two 2-week intervals on either side of the
conception date represented month “zero”. To be included in this analysis, it was
necessary that the male and female members of the pair be observed together
for at least 1 hour during each month-long interval. For these analyses, we
divided male affiliative behaviors into three categories.

A) We analyzed approach behavior alone in its own category.
B) We grouped sniff, lick, and greet into a category designated as
investigatory behaviors.
C) We grouped approach-avoid, present, bowing display, and paw ground
into the category of courtship behaviors.
We then examined rates of these behaviors in relation to female social rank and
female age. Female social rank was divided into three categories: high, middle,
and low rank categories represented ranks 1-7, 8-14, and 2 15, respectively.
Female age was similarly divided into three categories: young, middle, and old

age classes represented ages 2-6, 7-11, and 12-16 years, respectively.

58

We defined a “mating session” as an observation session during which we
observed any mounting behavior. We calculated daily mating rates by dividing
the number of mating sessions observed in the morning (0500-0900 hours) and
evening (1600-2000 hours) sessions by the total number of hours hyenas were
watched during these periods, and then multiplying by 24. We recorded mount
duration as the length of time (in seconds) a male remained mounted atop the
female and inter-mount interval as the length of time that elapsed between long
mounts. Matings were considered fertile (i.e. resulting in conception) when a litter
was observed within 1 month of the predicted birth date. This represents an
extremely liberal criterion for assessment of the fertility of a particular mating. We
calculated a predicted birth date by adding 110 days (average gestation period)
to the date on which we observed copulation or mounting.

We calculated rejection coefficients (RC), in the manner of Beach and
LeBoeuf (1967), to determine whether females treated adult immigrant and adult

natal males differently. RC was calculated as follows:
:-.i-. 01-.- .A ll “1-.....AI1" X100
total # affiliative behaviors exhibited by the male towards the female

Female aggressive responses included all levels of threat, from mild threat,

through chase and intention movement to bite, to actual biting.

El 'III 'I'

We used microsatellite markers identified from Crocuta DNA to determine male

parentage in the Talek clan. Between 1990 and 1999, we collected blood from

59

191 Talek hyenas. Sampling was conducted after individuals were anaesthetized
with Telazol (2.5 mg/kg) delivered from a COg-powered rifle in a lightweight
plastic syringe. We then immediately extracted DNA from blood using Puregene
kits (Gentra Systems, Inc.), and stored the extracted DNA in liquid nitrogen.

We considered all adult males present in the clan at the time of conception
as potential sires. Paternity assignments were based on 12 microsatellite
markers described elsewhere (Libants et al. 2000; SM. Funk and AL. Engh,
unpublished data). Briefly, we employed a maximum-likelihood approach
(Thompson 1975; Meagher 1986) to assign paternity to cubs with known,
genotyped mothers using the program CERVUS (Marshall et al. 1998). A male
was considered the father of a cub when CERVUS assigned him as such with
95% confidence. A more comprehensive analysis of paternity in the Talek clan is

presented by Engh et al. (2002).

I'I'I' ['I _ l'l I'

We next compared the behavior of male and female hyenas in mating vs. non-
mating situations. To reiterate, we defined a mating session as an observation
session during which we observed any mounting behavior. We designated
observation sessions that did not contain mounting behavior, as ‘other’ sessions.
For each mating pair, we calculated the rate at which each animal approached its
partner, and we also calculated the percent total time during each mating session
each animal spent following its partner. An hourly approach rate was calculated

for each individual as [(# of approaches / # minutes observed during mating

60

session) * 60]. Percent time spent following was calculated as [(# of min spent
following / # min observed during mating session) * 100]. We calculated mean
values for individuals if they were observed during multiple mating sessions.
Finally, we compared the rates of approach and following behavior by males and
females during these mating sessions to those calculated in ‘other‘ sessions, as

described above.

I I I _[ I . l'
Using only litters for which sires could be identified by paternity analysis, we
compared sires and non-sires with respect to their association patterns with
females around the time these litters were conceived. Here, we simply recorded
whether each male was present with the female during each consecutive session
in which the female was observed, starting 3 mos before conception and ending
1 mo after conception. Numbers of consecutive sessions with observations
varied considerably among females for this 4-month-Iong period, but this analysis
was useful nonetheless to inquire whether sires appeared to be engaging in any

sort of consortship or mate-guarding behavior.

I I I | . . | . I' .
For data collected between June 1997 and June 1999, we recorded two
additional aspects of male behavior to evaluate male mate choice. First, when a
male sniffed an individual female during an observation session, we

subsequently recorded whether or not that male stayed with the female he had

61

sniffed (sniff-and-stay) until they disappeared out of sight or until observers left
that session. Second, when a male showed a clear preference for one individual
female over others during an observation session, the male was recorded as
choosing that female. A male was determined to be exhibiting a clear preference
(choice) for a particular female when he repeatedly courted and/or followed a
single female throughout a session in which at least one other adult female was
present. Both of these male decision-making behaviors were recorded as
dichotomous values: sniff-and-stay vs. sniff-and-go or choice vs. no choice, and
these were mutually exclusive.

We then determined which female characteristics influenced male
choices. We used logistic regression to determine which independent variables
influenced the value of eachh dichotomous dependent variable (Legendre and
Legendre 1998; Hosmer and Lemeshow 2000). Female predictor, or explanatory,
variables included social rank, time to conception, and age. Because we were
specifically interested in a male’s preference as it was expressed among multiple
females, and because up to 12 females were present in a given session at the
same time, we assigned each female in the session a value for rank, time to
conception, and age, relative to the other females in the session. All three
explanatory variables were simultaneously used in the model and were analyzed
using Logit Estimate model in the Systat 8 statistical software package (Systat

1998)

62

:1 II I I. l' I I
We presented mean values as 7 3: SE. We analyzed data that were not normally
distributed using non-parametric tests. Two-way ANOVAS were conducted on
rank-transformed data to test for interactions, since interaction analyses were not
possible using the Kruskal-Wallis one-way analysis of variance. We considered
differences between groups to be statistically significant when P<0.05. Where we
conducted post-hoc tests, P values were calculated using a Bonferroni

adjustment for multiple comparisons.

RESULTS

BI [ffil'l'll"|l'l I'

We first examined rates at which males directed affiliative behavior toward
females near the time of conception during the 2-year period starting in June
1997 and ending in June 1999. Here, we divided male affiliative behaviors into
three categories: approaches, investigatory behaviors, and courtship behaviors,
as described above in Methods. Data included affiliative interactions observed
within 202 male-female pairs, involving 20 females and 21 immigrant males.
Time relative to conception had a signifiCant influence on male approach
behavior and investigatory behaviors, but only a marginal influence on courtship
behaviors (Figure 3.1; Kruskal Wallis: approach: KW5=21.918, P=0.001;
investigate: KW6=27.844, P<0.001; court: KW6=11.083, P=0.086). Interestingly,

the highest rates of approach and investigatory behaviors were observed in the

63

Figure 3.1. Mean daily rates (:tSE) at which males engaged in (A) approaches,
(B) investigatory behaviors, and (C) courtship behaviors, plotted as a function of
time relative to conception in targeted females. interval ‘0’ represents the date of
conception :1: 2 weeks. Remaining intervals represent month-long periods
surrounding the date of conception. N=202 male-female pairs involving 20
females and 21 immigrant males observed together during each of 7 month-long
intervals. lnvestigatory behaviors included sniffing, licking, and greeting.
Courtship behaviors included approach-avoid behavior, presenting, bowing
display, and pawing the ground. Note that all Y-axes are on the same scale.

64

-54-3-2-1o1
-54-3-2-1o1

 

 

 

20"

>Mu \ mmcwrmoaaxx >3 \ 20328 >550:ng

F
q

C

5

b

0

 

—
5

2 1 1
>3 \ 22223 959500

-5 -4 -3 -2 -1 O 1
Monthly interval surrounding conception

65

month directly preceding the female's fertile period. Although rates at which
males emitted courtship behaviors were highest during the month of conception
and the month preceding conception, the differences apparent in this analyses
did not quite satisfy our criterion for statistical significance.

Rates of male affiliative behavior also varied as a function of female rank
category and female age class (Figure 3.2). Female social rank alone did not
significantly influence male afflliative behavior. However, we found a significant
interaction between female rank category and female age class for male
approaches, investigatory behaviors, and courtship behaviors (2-way ANOVA:
approach: F4,117=4.250, P=0.003; investigate: F4,117=2.407, P=0.05; court:
F4,117=3.346, P=0.012). Males tended to prefer older to younger females in the
middle- and low-ranking categories.

Male hyenas did occasionally exhibit flehmen after sniffing female urine
and sometimes after sniffing various parts of the female body (17 observations of
immigrant males exhibiting flehmen out of 411 male investigatory behaviors in
1997-1999). In the past, male hyenas have only been observed exhibiting
fiehmen after sniffing female urine (Holekamp et al., unpublished data). In this
study, we also observed male hyenas exhibiting flehmen after sniffing a female’s
hind legs, feet, and derriere.

Immigrant males rarely greeted with adult females. When they did, the
male usually approached the female first, stood with his head towards her tail,
and lifted his inside hind leg over her nose. The female then sometimes lifted her

inside hind leg over the male’s nose, and both members of the pair mutually

66

Figure 3.2. Daily rates at which males engaged in (A) approaches, (8)
investigatory behaviors, and (C) courtship behaviors, plotted as a function of
female social rank and age categories. Daily rates were rank transformed to
perform 2-way ANOVA on data that were not normally distributed (see Methods
for details). Transformed rates are plotted here (meantSE).

67

Female age class

' Young

D Middle
Old

 

 

 

  

 

WHMN4.7/////////////7///é
"I

   

 

 

 

 

 

150'

_ _

0 0 O
0 5

1

>3 \ 858059..

100“

E3 \ 9039.3 bofimmzmgc.

50”
O

 

 

 

1507

100‘

>8 \ 20328 aEmtsoo

_
0 0
5

Middle Low

High

Female rank category

68

sniffed each other’s genitalia. During the 5-year period, 1991-1995, we observed
only 6 symmetrical and 6 asymmetrical greetings between immigrant males and
adult females. Both immigrant males and adult females initiated asymmetrical
greetings to members of the opposite sex on three separate occasions, exposing
their genitals to the other hyena to sniff, but the other member of the pair did not
respond by lifting its own hind leg on these occasions. These 12 total greetings
between immigrant males and adult females represented only 1.5% of all male-
female affiliative interactions observed during this 5-year period. All immigrant
males (N=6 different males) that greeted with adult females (N=9 different
females) were ranked either 1 (N=6 greetings), 2 (N=4 greetings), or 3 (N=2
greetings) in the male hierarchy. Ranks of females who greeted with males
ranged from 1 to 26, including the highest and lowest ranking females present in
the clan.

Adult natal males performed relatively few affiliative behaviors toward
adult females. Ten different adult natal males approached, investigated, or
courted 16 different females, representing only 4% of all affiliative interactions
observed between adult males and females from 1997 to 1999. The youngest
age at which natal males were first observed courting or mounting adult females
was 19.7 months, although the average age of all natal males observed courting
and mounting females was 36.4 1: 0.9 mos (N=20 adult natal males). We were
interested to see whether female responses to male affiliative behavior varied
with male status, so we next examined rejection coefficients for natal versus

immigrant males. Rejection coefficients (RC) for immigrant males were

69

significantly higher than were those for natal males (Mann Whitney U: U1
=3042.000, P=0.001), suggesting that females responded more aggressively to
sexual interest expressed by immigrant males than to that exhibited by natal

males.

Wm

Between 1988 and 1999, only 39 observation sessions occurred during which
observers saw male hyenas mounting females. These consisted of 20 ‘complete’
copulations, which included intromission and ejaculation by immigrant males, and
an additional 19 mounting events, with or without intromission, but not including
ejaculation (Table 3.1 ). ‘Mating’ and ‘mating sessions’ thus both refer to the
observation sessions that included both mounting with and without intromission.
Seventeen immigrant males mounted or copulated with 18 adult females, and
five adult natal males mounted four different adult females.

Mating behavior occurred throughout the year, and the monthly
distribution of sessions in which mating was observed reflects the pattern of
known conceptions recorded for the Talek clan, with both mating behavior and
number of known conceptions peaking in August. Most (35/39) mating sessions
occurred during our morning observation hours, between 0530 and 0930 (Figure
3.3A). Three more took place between 1000 and 1230, and one took place at
1850, during evening observation hours. Overall, mating sessions took place at
significantly higher rates during the morning than afternoon observation hours

(Figure 3.38; Mann-Whitney U: U1=25.000,

70

TABLE 3.1. Sample sizes for mating sessions. In the text and in all analyses,
mating sessions refer to sessions in which we observed mounting, including

mounting both with and without intromission.

 

. # Males mount or mate / # Other individuals present/
# Sessrons . .
sessron sess10n

Mounting without clear intromission 19 1 0-8
Total sessions with mounting only 19
Mounting with intromission
(copulation) 18 1 0-8
Momting with intromission

. 2 2 0-1
(copulation)
Total sessions with copulation 20
Total mating sessions (mounting with 39

and without intromission)

71

d—l—Id

ON$O§QON$O§
ALA—J

# of matings observe:

0500
0600
0700
0800
0900'
1000
1100i

3000
2500 »
2000
1500 -
1000 ~
500 l

# of hours hyenas observe

0600 _

0700 _

0500 |
0300 I
0900]
1000]

O
O
‘-
‘-

 

1200?
1400 I

1200‘

1300

1300‘1

O
O

‘—

1 500

1500‘

Time of day

Observation times

1600

1600'

1700 —
1800 —

1900 |

1700

1800

1900

2000‘

2000

l

2100

2100i

Figure 3.3. (A) Times of day at which matings (N=39) were observed in
relation to total number of hours of observation. (B) Daily mating rate
calculated for morning and evening hours of observation. AM represents
0500-0900 hours, and PM represents 1600-2000 hours.

72

P=0.007). It should be emphasized, however, that we rarely followed hyenas
between the hours of 2000 and 0500 hours, so additional matings may have
taken place during these nighttime hours.

Copulations generally involved a predictable sequence of events. The
male first made a series of approaches towards the female, often rushing at her
as if to mount, then veering off at the last second. After a few approaches, the
male then mounted the female from behind, attempting repeatedly to achieve
intromission. This task was apparently made extremely difficult by the female’s
elongated clitoris with a forward facing opening. The male had to squat down and
under the female, so low that his rump was sometimes on the ground, to
manipulate his erect penis into the female’s flaccid phallus (Figure 3.4A). This
typically took several attempts over a period of several minutes. Once
intromission was achieved, the male then assumed the ‘standard’ position
mounted behind the female, stretching the female’s flexible phallus posteriorly
(Figure 3.4B). Some thrusting usually occurred at this time. Finally, the male
remained still (no thrusting) and remained mounted, sometimes licking the
female’s back but never biting her, and ultimately resting his chin and chest on
the female’s back for several minutes (Figure 3.40). We were unable to
determine whether or not a copulatory lock occurs at this time.

Figures 3.5 and 3.6 present examples of typical mating sequences. It can
be seen in both of these figures that one to several short mounts preceded each
long mount. Entire sequences of mounting, thrusting and resting, like those

shown here, were

73

Figure 3.4. Mating positions. (A) The male squats behind the female as he
attempts to achieve intromission. (B) The male stands and thrusts after achieving
intromission. (C) The male rests his chin and chest on the female’s back after
ejaculation. The subadult offspring of the mating female can be seen standing
behind his mother in (C).

74

 

75

Wald mounts Sein

 

 

J l l l

 

0 1O 20 30 40

Minute of observation

Figure 3.5. Temporal patterning of mounting behavior for a single male
(WALD) mounting a single female (SEIN) when no other hyenas were
present. Observers remained with this pair through minute 50.

76

50

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A

8 r1 m m

3

C

3

O

E

a

2 Ll U L.

l I l
O 50 100 150
B

T:

0 l

.23

C

D

O

E

0

X

L“ d i L

l l I

0 50 100 150

Minute of observation

Figure 3.6. Temporal patterning of mounting sequences for two males
(A=MER, B=EXC) alternately mounting a single female (GOL) during a
single observation session. Observers remained with these animals
through minute 150. MER was higher-ranking than EXC, and each male
sired one of the cubs in the twin litter produced 110 days after these
copulations were observed.

77

sometimes repeated several times over the course of several hours, and possibly
longer, as we were unable to determine whether additional copulations took
place prior to or following our observations of mating. However, we never
observed any one mating pair copulating repeatedly on successive days. Figure
3.5 depicts a single male mounting a single female when no other hyenas were
present. We observed this type of situation, when only the mating pair was
present, on 15 occasions, 10 of which involved only one long mount, as
represented in the figure. Figures 3.6A and B depict two males mating with one
female during the same observation session. We observed two males alternate
mating with the same female on two separate occasions, and observed two
males mate with the same female in succession (one after the other but without
repeated alternations) on a third occasion. In 24 of the 39 (62%) mating
sessions, hyenas other than the focal pair were present (Y=2.6 1 0.4 other
hyenas present, range=1-8). These other animals included adult females, adult
males, and subadults, both related and unrelated to the mating female. During
three of these 24 sessions (12.5%), a second adult male attempted to interrupt
the mating pair. Twice, the individual who interrupted was an immigrant male
higher-ranking than the mating male, and the interrupting male succeeded in
gaining access to the female on these occasions. In each of these two cases, the
two immigrant males present alternated mounting the female (as in Figures 3.6A
and B). In one of these sessions, the interrupting, dominant male was only very
mildly aggressive to the subordinate male while the latter was mounted. An

approach by the dominant male, accompanied by a mild threat, was usually

78

enough to persuade the subordinate male to dismount. However, the subordinate
male was not deterred for long: as soon as the dominant male moved away, the
subordinate male mounted the female again.

On only One occasion did we observe any severe aggression between two
males competing over an estrous female. Here again, the subordinate male was
not obviously deterred by the dominant male’s approaches, and he persisted in
mounting the female. As the session proceeded, the inter-male aggression
escalated, with the dominant male repeatedly biting the subordinate male who
was mounted, and eventually the two males had a standing fight. This fight
represented the most intense aggression we ever observed between males
competing over an estrous female. Finally, on a different occasion, an adult natal
male interrupted a mating pair by approaching them, but the natal male made no
attempt to mount the female after interrupting the mating of his rival.

We recorded a total of 462 mounts in the 39 mating sessions observed. Of
those 462 mounts, 275 had an accurate recorded duration to 1: 2 seconds, and
mount duration ranged from 1-1080 seconds (Figure 3.7; 7(=40.0 :l: 6.9 s).
Mounts were clearly divisible into short mounts, ranging in duration from 1-45 s
($5.3 :1: 0.5 s), and long mounts, which lasted longer than 100 s (32:3029 :l:
32.2 s, range=120-1080 5). Long mounts were observed in 20 mating sessions.
Males performed an average of 1.8 :1: 0.3 (range 1-5) long mounts during these
20 sessions. We calculated the number of short mounts performed during

intervals between long mounts, as well as the duration of inter-mount intervals.

79

# of mounts

1 00 I
50

20-

 

 

0 100 200 300 400 500 1000

Mount duration (in seconds)

Figure 3.7. Duration of 275 mounts observed during 39 mating
sessions, suggesting a possible functional distinction between short
(<60 s) and long (>100 s) mounts.

80

Males performed an average of 1.9 t 0.9 (range 0-11) short mounts during the
inter-mount interval. it took males an average time of 24.0 i 6.7 min (range 1-86
min) to achieve a second long mount. it is noteworthy that two of the three
longest inter-mount intervals were likely influenced by the interference of another
male who achieved a long mount of his own during that time. When we
calculated inter-mount interval length with these two samples excluded, we found
that it took males an average time of only 16.8 i 4.9 min to achieve a second
long mount. Calculation of latency to first mount and number of short mounts
preceding first long mounts was problematic since we usually encountered
mating pairs after mounting behavior had apparently already begun. However, it
is worth noting that it took males at least 18.0 :t 3.7 minutes and 12.9 :1: 2.6 short
mounts to achieve their first long mount. This is a minimum estimate based on
the time elapsed between the beginning of the observation session and the time
at which the male achieved his first long mount. Although successful intromission
was not consistently observable, it is likely that males achieved intromission
during all long mounts, because some males were occasionally observed to
intromit even during mounts lasting less than one minute. Males performed more
than one long mount in almost half (8/18) of the mating sessions in which we
observed any long mounts at all. This suggests that multiple long mounts over
the course of several hours or days may be typical of this species.

Only five natal males were observed to mount females. These mounting
sessions involved only short mounts and did not resemble the mounting

sequences performed by immigrant males in mating contexts (as depicted in

81

Figures 3.5 and 3.6). That is, natal males mounted at most a total of seven times,
if at all, during a mounting session, they never remained mounted for longer than
1 5, they never performed any thrusting behavior, and none of the natal males

ever achieved intromission.

WW
Although exact duration of the estrous period is unknown in this species, we
observed one female mating with three different males in a single 60-hour period.
Therefore, we surmise that estrus lasts at least 60 hours in this species. Our
efforts to estimate female cycle length in the field based on female behavior have
been largely unsuccessful. However, we have had the good fortune of observing
five females mate multiple times in the course of several months. The shortest
interval between two observed copulations, when the first copulation was
apparently not fertile, was 23 days. A second interval was 64 days. Both these
intervals involve multiples of roughly 3 weeks, thus suggesting that the non-fertile
estrous cycle in Crocuta lasts approximately 3 weeks, rather than the 2 weeks
commonly cited for this species (Matthews 1939; Lindeque 1981). The youngest
age at which female hyenas were observed copulating during the present study
was 24.2 mos, and the youngest female known to conceive was 25.4 mos.

in total we observed 39 copulations during 11 years of continuous daily
observation. This represents a surprisingly small number of mating observations
considering that, during the same time period, we recorded 158 conceptions

associated with births of litters eventually observed above ground (i=14.4 known

82

conceptions / yr). Furthermore, of the 39 mating observations, only 15 (38.5%)
were associated with later production of known litters, suggesting that females
might often need to mate two to three times, over repeated estrous cycles, to
achieve a fertile conception or successful pregnancy. Our scant observations of
mating would thus further suggest that the 158 litters produced might have
required over 400 copulations, of which we saw at most only ten percent!

The likelihood of conception was apparently not influenced by the details
of male sexual behavior. That is, whether or not a litter was observed 80-140
days after mating could not be predicted based on the mean duration of mounts
per mating session (Mann Whitney U: U1=62.000, P=0.247), the number of short
mounts per session (Mann Whitney U: U1=114.000, P=0.81), the number of long
mounts per session (Mann Whitney U: U1=84.000, P=0.17), or whether or not at

least one long mount was observed during the session (x12=0.869, NS).

In Chapter 2 (Szykman et al. 2001), we used focal animal data to evaluate the
roles played by adults of each sex in initiating associations by calculating rates of
approach and following behavior performed during non-mating sessions when
both males and females were present together. In these non-mating contexts, we
found that males approached and followed females at significantly higher rates
than females approached or followed males (Szykman et al. 2001). Using the
same calculations as those utilized there, we here contrasted male and female

approach and following behavior in ‘mating’ contexts to those same behaviors

83

observed in ‘other’, non-mating contexts. ‘Mating’ contexts refer to sessions
during which we observed mounting behavior. ‘Other’ contexts include all other
sessions, including observations of hyenas present at dents, at carcass sites,
and at locations other than dens and carcasses. There were significant main
effects and significant interaction effects for context by gender for both
approaches and following behavior (Figure 3.8; 2-way ANOVA: approach:
F1,61=17.744, P<0.001; follow: F1,61=13.967, P<0.001). We therefore conducted
post-hoc t-tests on each pair of samples, using the Bonferroni adjustment for
multiple pair-wise comparisons. During mating sessions, males approached
females at significantly higher rates than females approached males (t25=-3.949,
P=0.001), and males also approached females at higher rates than vice versa in
other contexts (t35=-6.220, P<0.001). Both males and females approached
individuals of the opposite sex at significantly higher rates during mating sessions
than in other contexts (male: t30=5.312, P<0.001; female: t31=5.267, P<0.001).
Following behavior showed a dramatically different result. While males spent a
greater proportion of focal animal time following females in non-mating contexts
than vice versa (t36=-5.879, P<0.001), females showed a tendency to spend a
greater proportion of their focal animal time following males than vice versa in
mating sessions. However, this tendency was not statistically significant
(t25=1.780, P=0.087). Females were significantly more likely to follow males
during mating sessions than in other contexts (t34=4.863, P<0.001), whereas
male following behavior did not vary significantly between contexts (t27=0.140,

P=0.890).

84

30*

25*

# Approaches / hr

20‘

15—«

10*

 

I Female _'_
Cl Male ‘

 

 

 

 

10
15

 

 

 

13

 

15

 

 

 

 

 

 

 

2’ i
APP APP
Other Mating

I
l
23
I i _ I
Flw Flw
Other Mating

Figure 3.8. Mean (tSE) hourly rates at which individuals approached,
and mean (iSE) percent of focal animal time during which individuals
followed, members of the opposite sex in two different contexts. Open
bars represent adult males and solid bars represent adult females. The

‘mating’ context consists of sessions during which we observed mounting

behavior. The ‘other’ context includes all other sessions, involving
observations of hyenas present at dens, at carcass sites, and at
locations other than dens and carcasses. Numbers over bars represent
number of individuals included in each sample. * P<0.001

85

530

“25

~15

:10

 

5UlMOIIOJ luads awu %

I I I E I . I.
We next examined temporal patterns in the associations between males and
females, relative to the time of conception. Time to conception was significantly
associated with the number of hours males and females were observed
together (Figure 3.9A; repeated measures ANOVA: N=20 females, F6,114=5.707,
P<0.001). Males spent significanty more time with females during the month
surrounding conception than during any preceding month except the one
immediately preceding, or the one immediately following, conception (Tukey’s
post-hoc comparisons). Furthermore, more individual males were observed
associating with females during the months surrounding conception than at other
times (Figure 3.9B; repeated ANOVA: N=20 females, Fs,114=5.832, P<0.001).
This was not due to females being observed for more hours during the month
around conception, because females were actually observed more during early
lactation than during periods when they were most likely to conceive (see
Chapter 2).

We next analyzed consecutive observation sessions to examine how
consistently males were found with specific females around and preceding the
time of conception. We further inquired whether these patterns differed for sires
and non-sires, as well as whether there was any evidence of mate-guarding by
sires. Previous association data indicated that males known to have sired the
offspring of individual females associated significantly more closely with their
mates than did other males (Szykman et al. 2001). Furthermore, sires spent

significantly more time alone with their future mates than did males failing to sire

86

20—

a a a a
1OL

# Hours female observed with male

 

-5-4-3-2-101

# Males observed with female
0)
l

l

l

-5-4-3-2-101

 

 

Monthly interval surrounding conception

Figure 3.9. (A) Total time immigrant males were observed with 20
females, and (B) number of immigrant males that were observed
associating with same females, during the months surrounding each
female’s conception. Letters over bars represent means significantly
different from one another at P<0.05.

87

offspring, particularly during the months immediately preceding conception
(Szykman et al. 2001). Those analyses used association indices, representing
the proportion of observation sessions in which individual males and females
were seen together, to measure the strength of male-female associations. Here,
we examined the temporal distribution of these associations more precisely.

individual males varied in the extent to which they associated with females
around the time of conception (Figure 3.10). One male in particular (the 5th
ranked male in Figures 3.10A and B) was consistently observed with females
during the 2 to 3 weeks preceding conception, and clearly succeeded in fathering
cubs using this strategy. Other males, of both high and low rank (ranks 2 and 16,
respectively) succeeded in siring offspring without exhibiting a clear pattern of
association with their future mate (Figures 3.100 and D). Thus, male hyenas may
sometimes appear to ‘guard’ their mates, but intensive mate-guarding is clearly
not necessary in this species to ensure that a male will sire the cubs of a
particular female.

We can sometimes utilize information from long-term male associations
with an individual female to infer female receptivity. In the field, we have
successfully predicted dates of birth for several litters by adding 110 days (the
period of gestation) to the last day of a male-female consortship, like those
depicted in Figures 3.10A and B. This is a particularly useful tool, given that the
female hyena provides minimal behavioral signs of estrus, and that we so rarely

observe copulation.

88

Figure 3.10. Association patterns between four females and the males who sired
(open diamonds) their litters (N=3 males; open diamonds in (A) and (B) represent
the same male), and the 17 males present concurrently who did not sire (solid
diamonds) their litters. Estimated day of conception is indicated by the solid
vertical line. Although the number of consecutive sessions during which each
female was observed varied, the time scales in (A)-(D) are all standardized to
include the three months preceding and the first month following conception.

89

Male rank

Male rank

Male rank

Male rank

Amar Q Non-sires

 

 

 

 

 

 

 

 

 

 

 

A OSires
20
15:. o o o o o o
.1 x . z . z .. . 3 g :
10I o :
5 4 000 000 ooooooooooooooooo8’o’ 000 o o
9
ol . .
O 10 20 30 40 5O
Nav
B
20 '1
15 A o o 0 «so: :00 o 0.0 o ’80 ’08
8 “ 8 8‘8 8 .88 ‘8 8 °
10 4 00 o 8! o o 3 e 0
00000003
5 so 0 o 030 000 000 o 000 0000
O T T T8 g T fl . 'T T fl
0 1O 20 30 4O 50 60 70 80 90
C Conception
20 1
15 3
10 7 ‘ ..
5
01 . «3° 3 .
0 10 70
Mp
D
20 -
15 go o o 000 . o 000 o o 9’08 .0
.30 a c ” so: ,, s e z s 8 3°
10 — o o
08 ° .
5-0 009 0000 so 0000». so.
0 T ‘r’ T g 3? T T T l
0 1O 20 30 40 50 60 70

Consecutive sessions female observed

90

E I . [I . I I . I I .
We evaluated male choices for 210 male-female pairs involving 20 male and 27
female hyenas using logistic regression. Male hyenas clearly preferred some
females over others, and decisions made by male hyenas to stay with or pursue
a particular female were significantly influenced by certain female characteristics.
Males who sniffed a particular female during an observation session were
significantly more likely to stay with that female when she was closer to the

time of conception, than to leave the scene or to stay with another female (Table
3.2A). Female social rank and age did not significantly influence males” choices
to stay with females or leave them after sniffing them (Table 3.2A). Given a
choice among several females, male hyenas exhibited the strongest preferences
for, and actively pursued, high-ranking females, but males also preferred females

when they were closer to the time of conception (Table 3.2B).

DISCUSSION

III [fil'l' II. II I I! Hi
Female dominance and masculinized genitalia are among the many unique
attributes of spotted hyenas that make them interesting subjects for the study of
sexual behavior. Although some of the general affiliative and investigatory
behaviors described above for hyenas also occur among other members of the
order Carnivore (e.g. lions: Schaller 1972), many of the courtship behaviors

discussed in this chapter are unique to this species. The approach-avoid

91

 

Table 3.2. Results of logistic regression examining effects of independent
variables (female characteristics) on male choice behavior.

A Sniff-and-stay

 

 

 

 

Female characteristic Estimate S.E. t-ratio p-value
Rank 0064 0.077 -0.823 0.410
Time to conception -0.395 0.085 -4.630 <0.001
Age 0.037 0.074 0.507 0.612
Log likelihood=292.896
N=492
Correct predictions=62.0%

B Choice
Female characteristic Estimate S.E. t-ratio p-value
Rank —0.222 0.087 -2.547 0.011
Tlme to conception -0.162 0.08 -2.011 0.044
Age 0.007 0.078 0.093 0.926

 

 

Log likelihood=303.077
N=459
Correct predictions=55.5%

92

 

behavior, bowing, and pawing the ground, all reflect some degree of motivational
conflict and hesitancy on the part of the male, supporting the notion that male
Crocuta fear females (Kruuk 1972; Holekamp and Smale 2000). It is likely that
immigrant males have adopted these courtship behaviors as a consequence of
their subordinate status, and that these courtship behaviors evolved from
displacement activities motivated by conflicting male desires to approach the
female and flee from her. The female

seems to take little notice of the male’s amorous advances, and she does little to
encourage him to pursue or approach her further.

Yet the male is not deterred. Despite the female’s apparent disinterest, the
male persists. Immigrant male hyenas approached, investigated and courted
females throughout their reproductive cycle, but these behaviors peaked near the
time of conception (Figure 3.1). From these data, we can infer that female
attractivity increases prior to conception and declines thereafter. While males of
all social ranks approached, investigated and courted females, only high-ranking
males were observed to greet with them, confirming what East and colleagues
(1993) and Mills (1990) found in their earlier observations of hyena greeting
ceremonies. Rates of male affiliative behavior also varied as a function of female
rank category and female age class, with males preferring older females in the
lower-ranked categories (Figure 3.2). Female age did not show any interaction
with female rank among high-ranking females. This could be because age has
little influence on female reproductive value in high-ranking females, who begin

breeding at an early age, and cubs in their first litters often survive to

93

reproductive maturity.

Many male mammals exhibit flehmen. A male sniffs or licks some urine
from a female, then he lifts his head with his lips parted, in a posture that may
serve to transport odor to nasopalatine ducts and the vomeronasal system (Dagg
and Taub 1970; Estes 1972; Doving and Trotier 1998). In this study, male
hyenas were observed exhibiting flehmen after sniffing female urine, but also
occasionally after sniffing a female’s hind end. it has been reported that the
vomeronasal organ functions to identify and interpret volatile components of
female urine and vaginal secretions (Doving and Trotier 1998). Male hyenas are
probably taking in female urinary pheromones when exhibiting flehmen after
sniffing the female herself, despite no obvious presence of urine. This is likely
because female hyenas are often observed to roll or lie down in their own urine,
perhaps to cool off or cover themselves in their own tantalizing scent.

Natal male hyenas have been reported to court females at significantly
lower rates than do immigrant males (Holekamp and Smale 1998). This is clearly
not the result of females responding to natal males with higher levels of
aggression. Rejection coefficients calculated in response to male affiliative
behaviors indicate that females respond more aggressively to the advances of
immigrant males than to those of natal males. This is an interesting paradox, as
female hyenas behave most aggressively towards males with whom they will
ultimately mate, and they virtually ignore advances by adult natal males, who
rarely mate within their clan (Engh et al. 2002). Among domestic dogs (Canis

familiaris), male behaviors and female responses are highly variable (Beach and

94

LeBoeuf 1967) as they are in Crocuta. Individual male dogs are acceptable to
females, and individual females are discriminating among males, both to varying
degrees. Beach and LeBoeuf (1967) ultimately found that rejection by the female
did not affect the frequency of male-female interactions or the probability of a
copulatory lock. Similarly, in hyenas, aggressive responses by the female did not

appear to discourage immigrant males.

III' II' If I l"!

Copulations are rarely observed in field studies of Crocuta. The small number of
observations of copulation may be because Crocuta are mating at night, as
suggested by Kruuk (1972) and Frank (1986b). It is also possible that mating
pairs actively try to isolate themselves to avoid interference by other clan
members. This would also make it more difficult for us to find and observe the
mating pair. in fact, on several occasions during mating sessions, the courting
pair vanished into thick bushes after the male had mounted the female once or
twice, but had not yet achieved intromission. Thus, we were unable to continue to
observe the pair.

Tentative approaches by the male, and attempted mounts that are aborted
at the last second, may function to help the male build up the courage to mount
the female, an idea suggested by SE. Glickman (pers. comm.) Also, these
approaches may serve to test the female’s receptivity, as she may respond by

turning to chase the male, or she may stand still ready to accept him. The female

95

standing still and lowering her head appears to be a clear signal to the male that
it is safe to mount her.

The thrusting we occasionally observe prior to intromission may help the
male find the female’s vaginal opening and line the male up properly for
successful intromission, as suggested for the same behavior observed in
domestic dogs (Beach and LeBoeuf 1967) and cats (Fe/is domesticus: Beaver
1992). Repeated mounting without intromission is not commonly reported in
other carnivores. Male hyenas appear to have greater difficulty achieving
intromission than other male carnivores due to the strange genital morphology of
the female hyena, with her long, flexible phallus and urogenital opening facing
anteriorly. The extended time the male spends mounted, not thrusting, and
resting his chest on the female’s back, might represent a post-ejaculation period
serving to facilitate sperm transfer, similar to the copulatory lock in canids (Asa
and Valdespino 1998). Although we were unable to observe in the present study
whether or not a copulatory lock occurred in Crocuta, prolonged intromission by
the male hyena might serve to enhance fertility.

We observed male hyenas perform multiple ejaculations, as inferred from
repeated long mounts in which the male rested his chin on the female’s back,
during a single mating session (e.g. Figure 3.6). This suggests that females may
require multiple matings, both during a single estrous cycle and over successive
estrous cycles, to ensure conception. Fewer than half of the matings we
observed ultimately resulted in the production of a litter, and we have observed

individual females each mating with the same male, and sometimes with a

96

different male, over two estrous periods within several weeks of each other.
Female hyenas might require multiple matings within a given estrous period
because they are induced ovulators. Presence of new corpora lutea in the
ovaries of captive female spotted hyenas not recently exposed to males suggests
that females can ovulate spontaneously, but that does not rule out the possibility
that they are induced ovulators in the context of mating (N. Place and A. Conley,
pers. comm.) In addition, there may be unusual obstacles to conception in
female Crocuta, perhaps resulting from their masculinized morphology and/or

physiology (Frank et al. 1995b; Frank 1997).

Il'l [ 'I' If l l"!

Male hyenas follow and attempt to maintain proximity to females, sometimes
associating with certain individual females as closely as females do with their
own offspring (Szykman et al. 2001). Furthermore, association patterns between
adult males and females clearly have some relevance to sexual behavior in
hyenas, as male hyenas who sire cubs associate more closely with their future
mates than do males who do not sire cubs (Szykman et al. 2001). These
associations in Crocuta are not usually cooperatively maintained by both partners
as commonly occurs in primate consortships (Stern and Smith 1984; Small
1990), nor are they as exclusive. However, we found here that females
approached and followed males at significantly higher rates during mating
sessions than during observations conducted in other contexts (Figure 3.8). This

suggests that female approaching and following behavior directed at immigrant

97

males should appropriately be considered to be proceptive behavior, indicating
the female’s active interest in mating. This subtle behavioral change, may be the
only demonstration of proceptive behavior by the female spotted hyena. The
subtlety of proceptive behavior in female hyenas stands in stark contrast with
other carnivore species in which females commonly exhibit obvious proceptive
behavior. For example, a lioness in estrus repeatedly approaches and rubs her
chosen male, rolls and swishes her tail in front of him, and presents her genital
area for him to sniff, frequently initiating mating (Schaller 1972). Domestic cats
sometimes show dramatically exaggerated forms of these same behaviors,
adding to them loud vocalizations to call for males (Beaver 1992), to which any
owner of an intact female cat can attest. Proceptive behavior by female domestic
dogs may appear more subtle than in cats, but it stiliis very clear; an estrous
female seeks out a male and presents her anogenital region near the male’s
muzzle for his inspection (Beach et al. 1982). Early androgen exposure during
early development in Crocuta (Licht et al. 1992) might possibly have an

attenuating effect on female proceptive behavior in this species.

Wm

Sniffing females often provides male mammals with information about female
reproductive condition via pheromones (reviewed in Bronson 1989). Male hyenas
frequently sniff female conspecifics, presumably to evaluate each female’s
reproductive state. It is therefore very interesting that time to conception emerged

here as the only significant female characteristic influencing a male hyena’s

98

decision to stay with a female after he has sniffed her (Table 3.2A). A male’s
‘choice’, indicated by his persistent following of, or extensive courtship behavior
directed toward, a particular female, may reflect a decision-making process by
the male after which he invests more time in that particular female. The male’s
choice behavior appears to be influenced by the female’s time to conception, and
may be affected even more strongly by female social rank (Table 3.28). We
interpret these data to indicate that male hyenas can discriminate among female
hyenas, and that they may make strategic choices based on a female’s social
rank and her time to conception. Overall, the models tested in the logistic
regression may not have very strong predictive value, but the significant effects
of rank and time to conception variables suggest that these parameters do
indeed influence male behavior. In Chapter 2, we similarly found that males
prefer to associate with high-ranking females and females most likely to be
fertile. It is currently unknown how males evaluate female rank in this species.
Temporal patterns in male-female associations suggest that a strategy of
‘consorting’ with a particular female may be more successful for some males
than for others. Individual males may adopt one of several different reproductive
strategies to maximize their own reproductive success (e.g. Figure 3.10). Our
association data also suggest, in conjunction with our observations of two males
alternating copulations with a single estrous female (Figure 3.6), that mate
guarding is not a strategy commonly utilized by male Crocuta. Interactions
among resident immigrant males are generally characterized by low levels of

aggression, even when access to an estrous female is at stake (Kruuk 1972;

99

 

Frank 1986b). Here we observed only one incident of severe aggression among
males competing over access to estrous females, and most male-male
interactions at such times are surprisingly amicable.

In this study, we have presented several years of detailed observations
from the field to further describe and elucidate spotted hyena sexual behavior.
Courtship among hyenas is unusual among mammals, not merely because it
reflects some degree of hesitancy on the part of the male, but because male
hyenas experience such extreme conflicting desires to approach the female and
flee from her. The iocomotor patterns observed during copulation are unique due
to the female’s masculinized genitalia: the smaller, subordinate male hyena must
insert his erect penis into the female’s forward-facing phallus, a feat that may
take several minutes to achieve. Finally, some male hyenas may use close
associations, or ‘consortships’, to increase their chances of mating with a
particular female. These data, in conjunction with the data presented in Chapter
2, suggest that individual male hyenas may adopt different reproductive

strategies to attract and acquire mates.

100

Chapter 4

RARE MALE AGGRESSION DIRECTED TOWARD FEMALES
IN A FEMALE-DOMINATED SOCIETY

INTRODUCTION

Aggression directed by males toward conspeciflc females occurs in a wide range
of animal taxa (e.g. insects: Stone 1995; birds: McKinney et al. 1983; mammals:
Smuts and Smuts 1993). Male mammals have been observed to behave
aggressively towards females in a variety of contexts. For example, male olive
baboons (Papio cynocephalus) attack females during feeding competition, in
defense of a close female associate or her offspring, and during disputes with
other males (Smuts 1985). In addition, male aggression toward females often
occurs in a sexual context (reviewed in CIutton-Brock and Parker 1995). Males in
mammalian species ranging from feral horses (Equus cabal/us) to hamadryas
baboons (Papio hamadryas) chase and bite females to keep females in their
groups and to discourage females from associating with rival males (Stammback
1987; Linklater et al. 1999). In other baboons and some macaque species, males
behave aggressively toward females to punish persistent refusal to mate (Smuts
1987a). In some species, males may physically force females to copulate with
them, as observed in orangutans (Pongo pygmaeus: Mitani 1985). All of the
species mentioned above exhibit strong sexual dimorphism. That is, males are

larger and better armed than females, and males dominate females. Here we

101

 

document male aggression against females in the spotted hyena, a species in
which females exceed males in body size by approximately 10%, and in which
females are socially dominant to males (Kruuk 1972).

In addition to discerning that female Crocuta are socially dominant to
males, Kruuk (1972) observed that the only circumstances in which adult male
hyenas behave aggressively toward adult females involve what he called
“baiting” behavior. During baiting, the female hyena sits or lies in a crouched
position with her ears flattened back against her head and her teeth bared, while
multiple males surround her with their ears cocked forward and their elevated
tails bristled. The males often rush at the female, and attempt to sniff or bite her,
and she sometimes sustains serious wounds from their attacks (Kruuk 1972).
The female defends herself, often lunging at one or more of the participating
males, and she may inflict serious wounds on them. Kruuk (1972) was puzzled
by the adaptive significance of this odd behavior pattern, but he doubted that
baiting was related to sexual behavior, during which males show fear and behave
extremely submissively to females.

Here we present data suggesting that this rare male aggression directed
toward females is tightly coupled to female reproductive state in spotted hyenas.
First we compare rates at which intrasexual and intersexual aggression occur
among free-living Crocuta. We then examine the circumstances under which
male hyenas behave aggressively toward females, and consider the possible
proximate causes and adaptive function of male aggression directed toward

females in this species.

102

 

METHODS

S I . I . I I I I
Spotted hyenas are gregarious carnivores that live in social groups called clans.
Each clan contains one to several matrilines of adult females and their offspring,
as well as one to several immigrant adult males. Clans are rigidly structured by
hierarchical rank relationships (Kruuk 1972; Tilson and Hamilton 1984; Frank
1986b; Holekamp and Smale 1990; Mills 1990; Holekamp and Smale 1993), and
an individual's social rank determines its priority of access to food and other
resources (Kruuk 1972; Tilson and Hamilton 1984; Frank 1986b). Adult females
are socially dominant to all adult males not born in the clan (Kruuk 1972; Smale
et al. 1993). Female hyenas are philopatric, but all males disperse to new clans
sometime after reaching reproductive maturity at 2 years of age (Smale et al.
1997). Although natal males may remain in their natal clans up to 51 months
after puberty (Smale et al. 1997; Holekamp et al., unpublished data), and
although they occasionally participate with immigrant males in coalitionary
aggression directed at females, such participation is rare (Holekamp and Smale
1998). We therefore focus exclusively here on aggressive interactions between
immigrant males and adult females.

We conducted this study in the Talek area of the Masai Mara National
Reserve, in southwest Kenya. This is an area of open, rolling grasslands grazed

year round by large concentrations of several different ungulate species. The

103

subject population was one large Crocuta clan inhabiting a home range of
approximately 65 km2 (Frank 1986a; Boydston et al. 2001). We identified all
hyenas in the Talek study clan individually by their unique spot patterns, and we
determined their sex from the dimorphic glans morphology of the erect phallus
(Frank et al. 1990). We established mother-offspring relations on the basis of
regular nursing associations. We photographed immigrant males immediately
upon their first sighting in the Talek area. We calculated the duration of a male’s
tenure in the study clan at the time when he performed a behavior of interest by
subtracting his date of arrival from the date on which that behavior was observed.
In this study, we monitored Talek hyenas continuously from June 1988 through

August 1999.

B I . I I I.
Throughout the study period, we recorded all approaches, appeasements, and
aggressive behaviors among hyenas as critical incidents ("all-occurrence
sampling“ of Altmann 1974). We recorded an approach when one individual
moved to within 1 m of another individual. Appeasement behaviors included head
bob, carpal crawl, giggle, squeal, back off, and assumption of the body posture in
which the ears are flattened back against the head, the tail is tucked between the
legs, the body is lowered and bent, and the head is down. Listed in order of
increasing intensity, aggressive behaviors included displace, intention movement
to bite, push, stand over, lunge, rush, snap, chase, and bite (Holekamp and

Smale 1990). Any interaction between two or more hyenas in which we observed

104

one or more of these individual acts of aggression was called an aggressive
interaction, or attack. For some analyses we divided acts of aggression into “low
intensity” and “high intensity” categories. The former included displace, intention
movement to bite, push, stand over, and Iunge. Individual acts of aggression
considered to be of high intensity included rush, snap, chase, and bite. If an
aggressive interaction involved only one aggressor and one recipient, we called it
a dyadic interaction. By contrast, a coalitionary interaction involved two or more
aggressors attacking a single target animal.

We assigned each aggressive interaction a location, the physical locality
where the interaction took place, and a context, the social situation immediately
preceding the attack. Locations included “dens,” “kills,” and “other" places away
from dens or food. “Dens” included both natal and communal dens, and “kills”
included both freshly killed ungulates and old ungulate carcasses. We observed
aggressive interactions among adult hyenas in the following five contexts: during
competition over food, in defense of cubs, when one hyena was being “pesky”
(e.g. persistent sniffing, etc.) to another, during displacement aggression
(“scapegoating”), and in situations where none of these other possibilities
applied, such that the aggression appeared to be entirely unprovoked.

We calculated daily rates at which adult hyenas of each sex emitted and
received aggression for a portion of the study period, between June 1988 and
May 1995. For this analysis, we divided the number of aggressive interactions
involving each individual as either attacker or target by the total number of hours

each individual was observed during that period, and then multiplied this number

105

by 24. In a separate analysis, we calculated the percent of attacks emitted and
received by adults of each sex that occurred in coalitions by summing the
number of attacks occurring in coalition, dividing by the total number of attacks,
then multiplying this by 100.

We determined intrasexual social ranks of individuals in the clan’s
dominance hierarchy from outcomes of several thousand dyadic agonistic
interactions during which one individual exhibited appeasement to the other,
whether or not the appeasement was elicited by aggression from the social
partner, as described by Holekamp & Smale (1990) and Smale et al. (1993). By
convention, the highest-ranking (alpha) individual was assigned a rank position of
one. To evaluate frequency of male aggression towards other males as a
function of social rank category, we calculated rates of male attacks on other
males during an 8.6-month period for which the male hierarchy was completely
stable. When contrasting rates of aggressive behavior by males directed at both
other males and females, we divided male social ranks into three equal rank
categories, high, middle, and low.

For each aggressive interaction in which one or more males attacked a
female, we recorded the identity and social rank of the female target, and the
number of males participating in each attack, as well as their identities and social
ranks. For each attack, we also recorded, whenever possible, the particular acts
of aggressive behaviors emitted by each individual male, the female’s responses
to each individual male (classified simply as aggressive or non-aggressive), and

the location, context, and duration (in minutes) of each attack. For analyses

106

examining baiting rates and individual acts of aggression exhibited by different
individual males toward females, we used a subset of the males (N=32) for which
critical incident data on individual acts of aggression were recorded.

We first determined whether rate or intensity of male aggression varied
with female reproductive condition. To do this, for the date on which each
intrasexual aggressive interaction was observed, we assigned each female to
one of the six reproductive states described below. If we were uncertain about a
female’s reproductive condition when she was the target of male aggression,
then that interaction was excluded from relevant analyses. We defined the six
reproductive states as follows.

Nulliparous (N): Here females were older than 24 months, but had not yet
conceived their first litter. The youngest age at which female Crocuta in this
population first conceived was 25.4 months, and females were observed to
copulate as early as 24.2 months of age (see Chapter 3). Age at first parturition
can readily be established in this species because the female bears an enlarged
clitoris (the pseudopenis) which tears when her first litter is born (Frank and
Glickman 1994), and heals leaving a large patch of pink scar tissue on its
posterior surface. Here the condition of this surface was carefully monitored in all
nulliparous females older than 2 years, during the phallic erections so commonly
exhibited when hyenas greet each other (Kruuk 1972; East et al. 1993). Thus,
even when a female's first litter was lost before her cubs first appeared above
ground, it was possible to reliably determine when (to within 10 days) she had

delivered her first litter.

107

Pregnant (P): The duration of pregnancy in Crocuta is 110 1: 1.0 days
(Schneider 1926; Kruuk 1972; ML Weldele, pers. comm.) We assigned birth
dates to litters by estimating cub ages when we first observed cubs above
ground at natal or communal dens. Cub ages could be estimated to :t 7 days
based on their pelage, size, and other aspects of their appearance and behavior.
We calculated conception dates by subtracting 110 days from estimated birth
dates, thus resulting in conception dates that were accurate to i 7 days.
Although most female hyenas wean one litter weeks or months before conceiving
another, females of all social ranks have occasionally conceived while still
nursing a previous litter (Holekamp et al. 1996). Females known to be pregnant
while still lactating were assigned a reproductive state of ‘P’ (see above).

Lactating (L1, L2, L3): The lactation interval began with the birth of a litter
and ended with its weaning. Weaning conflicts and cessation of nursing indicated
when cubs were weaned. Weaning conflicts between cubs and their mothers are
easily observed in this species due to the characteristic begging behavior and
loud whining exhibited by cubs at this time. We recorded all weaning conflicts in
field notes as critical incidents (Altmann 1974). In determining weaning dates, we
searched all field notes for observations of nursing behavior when mother and
cub were found together. If mother and cub were not found together frequently
after the last observed nursing bout, we identified the weaning date as being
midway between the last nursing bout and the next sighting of mother and cub
together without nursing. However, we only used intervals of 20 days or less in

this analysis, so all weaning dates used here were accurate to within 3: 10 days.

108

Lactation periods vary in length among Talek females from 7 to 21 months
(Holekamp et al. 1996; Holekamp and Smale 2000). Therefore, to compare rates
and intensity of male aggression among females with different lactation period
lengths, we divided the total period of lactation into three periods of equal length
(represented as L1, L2, and L3) for each female.

Other (0): Females assigned to this condition had weaned one litter but
had not yet conceived their next litter. Little is known about the reproductive cycle
of the spotted hyena. Matthews ( 1939) and Lindeque (1981) suggested that
estrous cycles in this species recur every 14 days, but there are no apparent
morphological indicators of estrus in female spotted hyenas as there are, for
example, in many cercopithecine primates (reviewed in Melnick and Pearl 1987),
and copulations are rarely observed in Crocuta.

For each female in each reproductive state, we calculated a daily rate of
male aggression toward that female using the formula: (number of male attacks
directed at each female per reproductive state/total number of hours the female
was observed in that state * 24 hrs). We assigned any adult female who was
observed but not attacked by males while she was in a given reproductive state a
daily rate of zero for that state.

We next examined the occurrence of male aggression relative to the date
of a female’s nearest conception. We calculated time relative to conception by
subtracting the female’s nearest conception date from the date on which she was
observed being attacked by males, and then dividing occurrence of these attacks

into month-long intervals. Together, the two 2-week intervals on either side of the

109

conception date represented month “zero”.

El'lllr'l' Hill I.

We used determination of paternity to compare rates of male aggressive
behavior toward each female Crocuta between males who sired her cubs and
males who did not sire her cubs but who were present with that female during the
period in which her litter was conceived. We used nuclear genetic markers to
determine paternity of cubs born in the Talek clan (Engh et al. 2002). Between
1990 and 1999, we collected blood from 191 Talek hyenas anaesthetized with
Telazol (2.5 mg/kg) delivered from a COz-powered rifle in a lightweight plastic
syringe. We then immediately extracted DNA from whole blood using Puregene
kits (Gentra Systems, Inc.) and stored the extracted DNA in liquid nitrogen.

We considered all adult males present in the clan at the time of conception
as potential sires. Paternity assignments were based on 12 microsatellite
markers described elsewhere (Libants et al. 2000). Briefly, we employed a
maximum-likelihood approach (Thompson 1975; Meagher 1986) to assign
paternity to cubs with known, genotyped mothers using the program CERVUS
(Marshall et al. 1998). A male was considered the father of a cub when CERVUS
assigned him as such with 95% confidence. A more comprehensive analysis of
paternity in the Talek clan is presented by Engh et al. (2002).

Our behavioral data were not normally distributed, so we conducted all
analyses using non-parametric tests. Mean values were presented as Y 1: SE,

and we considered differences between groups to be statistically significant when

110

P<0.05. All statistical tests were two—tailed. Data collection periods varied for

different analyses presented in the text, so please refer to Table 4.1 for details.

RESULTS

5 I'll . .

Adult male hyenas behaved aggressively toward other adults at far lower mean
daily rates than did adult females (Figure 4.1; Mann-Whitney U: U1=409.000,
P<0.001). Whereas males and females directed attacks at adult males at
marginally different rates (Mann-Whitney U: U1=305.500, P=0.07), the aggressive
behavior of males and females toward adult females differed dramatically (Mann-
Whitney U: U1=444.000, P<0.001). Specifically, adult females were frequently
aggressive toward other females, and although females tended to attack other
females at higher rates than they attacked males, this difference was not
statistically significant (Wilcoxon signed-ranks test: Z=1.855, N=23, P=0.07). In
contrast to females, adult males very rarely attacked adult females, and did so at
significantly lower rates than they attacked other males (Wilcoxon signed-ranks
test: Z=-3.621, N=20, P<0.001). In fact, during over 30,000 observer hours
watching Talek hyenas between 1988 and 1995, we observed only 256 individual
acts of aggression in 195 attacks directed at adult Talek females by adult
immigrant males compared to 1207 aggressive acts in 826 attacks on adult
immigrant males by adult females during the same period. The 256 acts of
aggression directed by males toward females involved 30 different adult females,

and 40 different adult immigrant males participated in at least one of these

111

Table 4.1 Data collection periods for analyses presented in Chapter 4.

 

Analysis Location Data collection period
All adult aggression Figures 1 , 2, 3 1988-1995
Approaches and responses Figure 4 1988-1999
Baiting, general Figures 5, 6, 7, 8, 9 1988-1999
Male-male aggression Text p. 124 1992-1993
Baiting, sires vs. non-sires Text p. 130 1991-1999

112

 

020.. C0_mwm.tmmm >:m.Q

[:1 Against males

 

 

 

 

 

 

 

 

 

 

 

 

6 F I Against females
5 _
.9 4 —
‘3
C
.9
8 3 r “r
93
a)
8’
2: 2 ’ I.
'63
O
1 ._
0 i I
By females By males

Figure 4.1. Daily rates (meaniSE) at which adult hyenas (N=23 females and
20 immigrant males) directed aggression toward other adults of each sex,
controlled for the number of hours observed during which each adult was
found in the presence of other adults of each sex between 1988 and 1995.

113

events. On average, each of these 40 males participated in 12.4 :I: 1.8 (range: 1-
51) attacks directed at adult females during his tenure in the Talek clan.

Intensity of aggression varied among intrasexual and intersexual attacks
(Figure 4.2). Male attackers exhibited a significantly greater proportion of low-
intensity aggression than we observed for female attackers, regardless of the sex
of the target animal (x21=94.52, P<0.001). Immigrant males exhibited the lowest
proportion of high intensity aggressive behaviors when attacking adult females
(x21=26.02, P<0.001). Adult females exhibited the greatest proportion of high
intensity aggressive behaviors when they attacked other females, though this
proportion did not differ significantly from that of female attacks on males
(181:1 .73, NS). That is, 60% of individual acts of aggression by females were of
high-intensity, regardless of the sex of the target animal.

In contrast to male-male aggression and aggression directed against
either sex by adult females, most (76%) male attacks upon females occurred in
coalitions (Figure 4.3). The proportion of total attacks on adult males that
occurred in coalitions was similar for male and female aggressors (181:1 .475,
NS), but a significantly larger proportion of attacks on adult females occurred in
coalitions formed by male aggressors than by female aggressors (x21=232.99,
P<0.001). Both males and females were more likely to form coalitions to attack
females than to attack males (male attackers: x21=161.2, P<0.001; female
attackers: x21=34.18, P<0.001).

Interestingly, females were significantly more likely to respond

aggressively to males approaching them in an aggressive manner than when

114

\l
O

13 Against males

 

N
O

.é‘

2 1207 1315 ,
a) l Agalnst females
E 60 1

.c

.9

.c

C —.

.9

8

9 40

ca

8

"5 - 256
In 30 ~

‘6

to

To

3

39

.2

U

.E

°\O

_L
O

 

 

 

 

 

 

 

By females By males

Figure 4.2. Percent of total individual acts of aggression of high intensity
that were directed by male and female hyenas against members of the
same and opposite sex. Sample sizes indicate total number of individual
aggressive acts observed between 1988 and 1995.

115

 

 

 

80 ’ 195
D Against males
70 - lAgainst females
2 60 —
.9
fi 50 ~
0
0
E 40 -
(I)
x
8 30 -
.1:
.I.: 1136
1:5 20 r 200
o 826
o\ 10 ~
0

 

   

 

 

 

 

By females By males

Figure 4.3. Percent of total attacks on other adults that occurred in coalitions.
Sample sizes indicate total number of attacks observed between 1998 and
1995.

116

ma

I3<

males approached females without exhibiting aggression (Figure 4.4; x21=44.05,
P<0.001). In general, females responded aggressively towards males who
approached aggressively, whereas they tended to ignore or investigate males
who approached in a non-aggressive manner. Male hyenas thus increased their

own risk of injury by behaving aggressively towards females.

W

The number of males participating concurrently in attacks on adult females varied
from one to six (Figure 4.5), with a mean of 2.5 1: 0.1 males joining forces per
attack. On average, each male participating in these attacks directed 1.9 :I: 0.1
(range: 1-19) individual acts of aggression at the female per attack. The two most
intense forms of aggression, chasing and biting, were directed at the female in
15.6% of attacks. On average, male attacks on females lasted 3.84 :t 0.49 min
(range 0.5 to 49 min), and attack duration did not vary with the number of males
participating (Kruskal-Wallis: KW5=3.204, NS).

Male aggression toward females was remarkably consistent in
appearance, despite variability in the number of participating males. As males
approached, they typically bristled their manes and tails, and surrounded and
rushed the female, who crouched defensively in their midst with her teeth bared
(Figure 4.6). Because the form of male attacks upon females was so consistent
in appearance, and because it invariably resembled the intersexual aggressive
interactions Kruuk (1972) called “baiting behavior”, we adopted Kruuk’s

terminology. Thus “baiting” will henceforth be used here to refer to an

117

178 117 El Non-aggressive response

 

 

 

 

 

 

 

 

 

 

70 _ lAggressive response
a) 60 "'
U)
8
% 50 ~ 92
(D 60
'- 40 ~
9
to
E 30 ~
0)
LI.
°\o 20 r
10 -
0

Non-aggressive Aggressive

Male approach

Figure 4.4. Female responses to male approaches. Sample sizes indicate
numbers of male approaches observed for which female response was
recorded.

118

40,.

81

35

30

25

% Total attacks on females
N
O

 

 

Number of males participating

Figure 4.5. Numbers of individual males participating in 227 attacks on
adult females. Sample sizes indicate numbers of interactions observed
between 1988 and 1999 in which adult immigrant males attacked adult
Talek females.

119

 

 

Figure 4.6. Example of male coalitionary aggression toward a female.
Three males attack a female, who crouches in the center of this
photograph with her teeth bared. Note the males’ bristled tails,
indicating a high level of excitement.

120

 

unprovoked aggressive attack directed by one to several adult males toward a
single adult female, who responds by defending herself, and sometimes also with
counter-aggression.

During most intrasexual aggressive interactions among natal Talek
hyenas, and aggressive interactions in which females attacked immigrant males,
approach or attack by one or more dominant animals elicited unambiguous
submissive behavior from the targeted lower-ranking individual (Smale et al.
1993). That is, when aggressors approached targets, the aggressors’ heads
were held high, their ears were cocked forward, their manes and tails were
bristled, and they stood their ground at the end of the interaction. During these
ordinary aggressive interactions, subordinates flattened their ears back against
their heads, tucked their tails between their legs, exhibited a facial expression
called a “fear grin”, bobbed their heads up and down, and backed off or fled to
terminate the interaction. By contrast, males engaging in baiting behavior held
their heads high with their ears cocked forward and manes and tails bristled, yet
they often simultaneously bobbed their heads up and down, backed away
spontaneously from the crouching female, and fled when singled out and chased
by the targeted female. Thus, males participating in baiting appeared to
experience conflicting desires to attack and flee. Although females targeted
during baiting crouched to defend vulnerable body parts and flattened their ears
back, they did not otherwise exhibit typical submissive behavior. The baring of
teeth by baited females clearly involved the contraction of a different set of facial

muscles from those contracted during normal “fear grins”. When baited by a lone

121

male, an individual female was readily able to demonstrate her dominant status,
and she never backed off from, or submitted to, the male. However, when
surrounded by a group of males, the female appeared to experience greater
difficulty protecting herself from simultaneous attack by multiple males. Thus,
males acting in groups launched attacks against dominant females, yet these
coalitionary attacks did not result in any change in dyadic rank relationships
between individual males and females.

Males baited females at a variety of locations where many hyenas were
found together. Eleven percent of male attacks upon females occurred at dens,
26% occurred at sites of ungulate kills, and 63% occurred at locations away from
dens or food (Table 4.2A). Whereas the duration of male attacks on females did
not vary with location (Kruskal-Wallis test: KW2=3.269, NS), the number of
participating males was significantly greater at kill scenes than at dens or other
locations (Kruskal-Wallis test: KW2=13.841, P<0.001). This may be due to the
fact that the average number of hyenas present at kill scenes is greater than the
average number of hyenas found at dens or locations away from dens or food
(Kruskal-Wallis test: KW2=535.225, P<0.001). Nevertheless, even when males
outnumbered females at a kill scene, males were never able to usurp food from
feeding females. At 886 fresh ungulate carcasses, at which both immigrant males
and adult females were present during 640 hours of observation, not once was
an adult immigrant male ever observed displacing an adult female from food.

Regardless of the location of these aggressive interactions, all male

baiting of females appeared to be entirely unprovoked (Table 4.2B). No male

122

Table 4.2. Locations (A) and contexts (B) in which individual acts of
aggressive behavior were observed between 1988 and 1995. (Note:
aggression by males toward females observed between 1996 and 1999
are considered in the text but are not included in last column of this table.)

 

 

 

 

Female-Female Female-Male Male-Male Male-Female

A Location Number % Number % Number % Number %
Dens 483 58.5 378 33.3 47 23.7 28 10.8
Kills 307 37.2 391 34.4 102 51.0 67 26.2
Other 36 4.3 367 32.3 51 25.4 161 63.1
Total 826 1136 200 256

B Context Number % Number % Number % Number %
Food 342 54.0 194 43.7 97 60.6 0 0.0
Intervention 16 2.5 1 0.2 0 0.0 0 0.0
Pesky 40 6.3 11 2.5 0 0.0 0 0.0
Scapegoat 42 6.6 19 4.3 25 15.6 0 0.0
Unprovoked 151 23.9 215 48.4 35 21.9 256 100.0

 

123

 

aggression toward females occurred during feeding competition, even when food
was present at the location of the interaction. Furthermore, no baitings occurred
during defensive interactions on behalf of another hyena, during displacement
aggression (“scapegoat” in Table 4.2B), or in response to the female behaving in
a “pesky” fashion toward males. By contrast, male-male aggression occurred in a
variety of different contexts (Table 4.2B), most often during displacement
aggression and feeding competition, where high-ranking males were able to

displace lower-ranking males from desired resources.

WWW
Baiting rates varied significantly as a function of male rank category. High- and
middle-ranking males baited females at higher rates than did lower-ranking
males (Figure 4.7; Kruskal-Wallis test: KW2=28.683, P<0.001). This was not
merely a function of high-ranking males behaving more aggressively in general.
High-ranking males were no more aggressive toward other immigrants than were
lower-ranking males, even when taking into account the number of potential
targets available to males at each rank (Kruskal-Wallis test: KW2=0.202, NS).
Individual males baited females at rates that varied greatly (range: 0.408—
6.360 baitings/day observed/tenure year in clan). There was a 15-fold difference
between the minimum and maximum rates at which males baited females over
their full tenure in the clan. However, due to the high variation among years for
individual males, differences in baiting rates among individual males were not

statistically significant (Kruskal-Wallis test: KW31=40.024; NS). Nevertheless,

124

 

'01
I

.3
O
I

Daily baiting rate

.0
01
I

 

 

0.0

High Middle Low
Male rank category

Figure 4.7. Daily rates (meaniSE) at which adult males in each of three rank

categories baited females. Rank categories high, middle, and low represent
male ranks 1-6, 7-12, and 2 13, representing N=23, 18, and 8 males,
respectively. Bars sharing same letters represent means not significantly
different from one another.

125

individual males varied significantly in the number of individual aggressive acts
directed towards females during each baiting (Kruskal-Wallis test: KW31=65.610,
P<0.001).

Male tenure is a better predictor of male reproductive success than is
social rank (Engh et al. 2002), but here tenure could not predict baiting rates
better than could male social rank. Daily baiting rates did not vary significantly
with the year of male tenure (Kruskal-Wallis test: KW10=15.384; NS). However,
the proportion of males who were observed baiting females did generally
increase with tenure in the clan (Figure 4.8; x21o=20.796, P=0.02). Only 30% of
males in their first year in the clan baited Talek females, whereas 50-70% of
males remaining in the clan for 2 to 9 years were observed baiting females. All
males remaining in the clan for 10 and 11 years were observed to bait females
during their residence in the clan, although only two males persisted in the Talek

clan for such a long period of time.

MW

We observed females of all social ranks being baited. We found no significant
relationship between daily baiting rates and female rank for females in any
particular reproductive state, or overall (Spearman rank correlation: rs=-0.301,
N=18, NS). Furthermore we found no significant relationship between a female’s
social rank and either the number of males participating in her baitings (Kruskal-
Wallis test: KW18=20.311, NS) or the duration of her baitings (Kruskal-Wallis test:

KW15=18.807, NS).

126

1.0 .

0.8 ~
24

’I

0.7 1 50 37

0.6 1

0.5 -
0.4- 53

0.3 ~

Proportion males observed to bait

0.2 1

0.1 I

 

 

0.0 -

Year of tenure

Figure 4.8. Proportion of males observed baiting females per year of
male tenure in the Talek clan. Numbers over bars represent the number
of males represented in each year of tenure.

127

We found females being baited in all six reproductive states. We observed
individual females in each reproductive state for 0-844.6 hours: YN=17.9 t 8.4,
xp=662 :l: 14.8, 21:1205 s 22.1 , 30584.8 1 18.2,7L3=56.0 a: 12.4, 20:24.5 :l:
6.8 hours per female. Neither the number of male participants per baiting
(Kruskal-Wallis test: KW5=6.557, NS), nor baiting duration (Kruskal-Wallis test:
KW15=8.544, NS) varied with female reproductive state. Proportions of high and
low-intensity acts of aggression directed at the female by participating males did
not vary with female reproductive state (x25=3.77, NS). The proportion of baitings
in which the target female responded to male aggression with an aggressive
counterattack also did not vary with female reproductive state (125:3.90, NS).
Furthermore, in 68% of male attacks upon females that were either in the last
trimester of lactation or between weaning one litter and conceiving the next, the
female left her defensive crouch to threaten or chase one or more of the
participating males, whereas in 32% of these interactions her behavior was
entirely defensive. Thus, even when females were most likely to be sexually
receptive, they usually responded to male aggression with aggressive
counterattack.

Despite finding no significant differences among various details of baiting
events and female reproductive state, female reproductive state did significantly
influence the overall daily rate at which females were baited (Figure 4.9; Kruskal-
Wallis test: KW5=19.313, P=0.002). Baiting rates were significantly higher when

females were in either the last third of the lactation interval, or between weaning

128

 

2.0 r -.

 

 

1.5 —-
.9
(0
E
.9
L.
8 1.0 -
,9
9
O)
.9
'6
.0 0.5 —
..>.~
'66
o

 

 

   

0.0
N P L1 L2 L3 0

Female reproductive state

Figure 4.9. Daily baiting rates (mean 1- SE) per female for 22 females in
various reproductive conditions observed between 1988 and 1995. “N”
indicates nulliparous, “P” indicates pregnant, “L1" represents the first
trimester of the lactation interval, “L2" indicates the second trimester of the
lactation interval, “L3“ indicates the third trimester of the lactation interval,
and “0” indicates the period between weaning one litter and conceiving the
next. Each female was observed in all six reproductive states.

129

and conception, than when females were in any other reproductive state. This is
noteworthy because it is while females are in these two particular reproductive
states that they are known to conceive. As mentioned earlier, females are
probably often cycling during the “other” state, and must also often cycle during
their last trimester of lactation, as females are sometimes known to maintain
lactation and pregnancy concurrently. Mean baiting rates did not differ
significantly between the L3 and 0 intervals (Mann-Whitney U test: U1=300.000;
NS).

Most baitings occurred during the last months preceding conception, with
the greatest number of baitings observed during the month in which conception
occurred (Figure 4.10). However, we never observed males mounting females
during, or immediately after, a baiting event. The number of observed baitings
dropped off sharply during the second month post-partum. Overall, males baited
females at significantly higher rates before than after conception (Wilcoxon

signed-ranks test: Z=2.184, N=97, P=0.029).

E ii I I . I I . I _ .
We next contrasted baiting behavior of males who sired cubs with that of males
present concurrently with sires who failed to sire offspring. During the period
1991-1999 for which we have paternity data, we observed 201 baitings. Only 49
(14.4 %) of these baitings were observed within 6 mos of conceptions of litters
with known paternity. When we compared the proportion of baitings in which

sires and non-sires participated during 6-month time blocks before and after

130

 

Conception

 

10r—

Number of baitings observed

 

O A'LANAQBbflbSDAIbI'LAQN’L'bh661$9®¢~¢

Month relative to nearest conception

Figure 4.10. Observations of baitings relative to the date of the nearest female
conception. The interval labeled “0” represents the period starting 2 weeks
prior to conception and ending 2 weeks after conception. The interval labeled
“1” thus represents the period 2-6 weeks after conception, and so forth. For
this analysis, we only included baitings that took place within one year of a
female’s nearest conception. (N=172 baitings of 27 females)

131

conception, we found no significant difference between sires and non-sires
(x21=0.178, NS). Nor did overall rates of baiting behavior differ between sires and
non-sires during this year-long time period (Mann-Whitney U test: U1=613.000,
NS). However, while sires baited females at similar rates both before and after
conception (Wilcoxon signed-ranks test: Z=0.356, N=11, NS), non-sires baited
females at significantly greater rates before than after conception (Wilcoxon
signed-ranks test: Z=2.106, N=86, P=0.035). We found no significant difference
between the mean number of individual acts of aggressive per male per baiting
for males who sired cubs during a female’s reproductive cycle and those males
who did not sire cubs of that particular female at that time (Mann-Whitney U test:
U1=1096.000, NS). Finally, females did not appear to respond differently to
attacks carried out by males who went on to sire cubs with them and those males
who did not sire cubs. That is, the proportion of male attacks to which females
responded aggressively did not differ between males who sired cubs and those
who did not sire cubs during that female’s reproductive cycle (x21=3.342, P=0.07,

NS).

DISCUSSION

WW
Kruuk (1972) first observed that female spotted hyenas appear to be more
aggressive than males, and our results quantitatively confirm that females are

indeed more aggressive than males in this species. Adult immigrant male spotted

132

 

hyenas exhibited intrasexual aggressive behavior at far lower daily rates, and at
lower intensities, than did adult females, and rates of intrasexual aggression
were higher for both sexes than were rates of intersexual aggression. Rates of
male aggression toward females are far lower in Crocuta than in most other
species for which male aggression directed toward females has been quantified
(savanna baboons: Papio cynocephalus: Hausfater 1975; wild horses, Equus sp.:
Berger 1983, Berger, 1986 #338; rhesus macaques, Macaca mulatta: Lindburg
1983; orangutans: Mitani 1985; eg. chimpanzees, Pan troglodytes: Goodall
1986; Bercovitch et al. 1987; elephant seals: LeBoeuf and Mesnick 1990;
Mesnick and Le Boeuf 1991; Connor et al. 1992a; dolphins, Tursiops sp.: Connor
et al. 1992b; Japanese macaques, Macaca fuscata: Soltis et al. 1997). Darwin
(1871) noted that males of most polygynous species are “more pugnacious” than
females, but the reverse is clearly true in the spotted hyena. Male attacks on
females occur rarely in hyenas and usually involve coalitions of several males.
These attacks occur wherever male hyenas encounter adult females, except in
the presence of lions or alien hyenas. In the present study, all male attacks on
females were entirely unprovoked.

High- and middle-ranking males baited females at higher rates than did
lower-ranking males, and the proportion of males who participated in baitings
increased with immigrant male tenure in the Talek clan. Although reproductive
success among male hyenas tends to increase with rank and tenure (Engh et al.
2002), we could detect no differences in rates or intensities of baiting between

males who sired cubs and males who did not sire cubs. However, our failure to

133

detect such differences may be due to small sample sizes.

Females of all social ranks were baited with equal frequency and
comparable intensity. Females were baited throughout the reproductive cycle,
but baiting rates were highest during periods when they were most likely to
conceive litters. Kruuk (1972) believed that baiting in spotted hyenas was
unrelated to sexual behavior because, whereas high levels of intersexual
aggression are observed during baiting, aggression by males toward females is
never observed in association with copulation. Our observations confirm that
male hyenas never attempt to mount females during baiting as, for example, do
male dolphins during coalitionary “herding” of females (Connor et al. 1992b).
However, our analysis of baiting behavior revealed a strong association with
female reproductive state. Baiting rates here were extremely low during
pregnancy and during the first two thirds of the lactation interval. Indeed, daily
baiting rates were highest for females in the last third of lactation and during the
“other” state, stages of their reproductive cycle during which they were most
likely to conceive litters. In addition, the greatest number of baitings was
observed during the month in which conception was known to occur. Similarly,
male aggression toward females in many other mammalian species is most
commonly observed near the time of conception (reviewed in Smuts and Smuts
1993). In fact, Smuts & Smuts further reported that they encountered very few
instances of male aggression toward females in non-sexual contexts in their
survey of the literature on non-primate mammals (see Table 1 in Smuts and

Smuts 1993). Our data lead us to conclude that, despite its unlikely appearance,

134

 

male attacks on females are somehow related to sex in the spotted hyena.

‘ net 11 -_n_ on 011-. ‘ an. ‘ on 01-. Q ‘11- -
Aggressive interactions involving multiple males were often preceded by social
sniffing (Kruuk 1972) in which individual males bristled their manes and tails
excitedly, leaned against each other, and vigorously sniffed the ground. The
function of social sniffing appears to be similar to that of “group ceremonies” in
wolves (Mech 1970) or “social rallies” in African hunting dogs (Estes and
Goddard 1967; Creel and Creel 1995). That is, social sniffing among spotted
hyenas appears to function as a mechanism to elevate and synchronize the
arousal levels experienced by multiple individuals concurrently (Kruuk 1972;
Holekamp et al. 2000). Thus, adult males often worked themselves into a highly
excited state before they approached and threatened a female.
The finding that males bait females during the phases of her reproductive
cycle when she is most likely to conceive suggests that males are able to detect
when a female is in, or is approaching, estrus. Endocrine correlates of estrus
have not been described in the spotted hyena, but plasma levels of estrogen,
progesterone, and testosterone all appear to be much higher around the time of
parturition than around the time of conception in spotted hyenas (Licht et al.
1992; Holekamp et al., unpublished data). Levels of excreted testosterone in
feces also peak late in gestation in this species (Dloniak et al., in prep). Thus,
absolute levels of these steroid hormones in females do not appear to play a

significant role in stimulating baiting behavior by males. However, it remains

135

possible that the relative levels of some combination of steroid hormones,
including androstenedione, might cue males that a female is approaching estrus.

Four hypotheses suggest possible adaptive functions of male attacks
directed toward females in the spotted hyena. First, male hyenas might attack
females to displace them from desired resources. This hypothesis predicts that
baiting should occur in temporal association with feeding competition, and that
males should at least occasionally be able to usurp food from females. Although
26% of observed baitings occurred at sites where ungulates were killed and
devoured, no male attacks against females involved contests over food or any
other resource. Furthermore, since immigrant male hyenas invariably defer to
females during feeding competition (Kruuk 1972; Tilson and Hamilton 1984;
Frank 1986b), we can promptly reject this hypothesis.

A second hypothesis suggests that aggressive behavior might allow males
to test female receptivity (Beach 1976). If males are using aggressive acts to test
female receptivity, then we would predict that responses by females to male
attacks should vary with female reproductive condition, but we found no such
variation, and baiting did not immediately precede the few successful copulations
that we did observe. In many other mammalian species, males assess female
reproductive state via investigation of chemical signals emitted by females (e.g.
Johnston 1983), yet such assessment is generally not associated with attack
behavior directed by the male at the female. Male Crocuta also frequently
engage in a variety of non-aggressive behaviors that appear to give them

information about a female’s current reproductive state (see Chapter 3), such as

136

sniffing her body and recently excreted urine and feces which contain gonadal
steroid hormones and their metabolites (Koretz 1992; SM. Dloniak et al. & S.E.
Glickman et al. unpublished data). Our observations demonstrate that male
approaches are less likely to be met with aggression by the female if the male
approaches non-aggressively. Therefore, risk of injury to both the male and the
female is lower for these non-aggressive interactions. Because males should
incur lower costs when they obtain information on female reproductive state via
interactions with females that involve no aggression, we conclude that this
functional explanation for baiting behavior seems unlikely.

Third, male attacks directed toward females might function in this species
as a form of sexual coercion, or sexual harassment. That is, baiting behavior may
function to increase the probability that a female will mate with a particular male
at a time when she is likely to be fertile, and to decrease the probability that she
will mate with other males at that time. As summarized by Smuts & Smuts
(1993), this hypothesis predicts that male aggression toward females should
occur frequently when females are most likely to conceive new litters, should
occur infrequently when females are in other phases of their reproductive cycles,
and should incur some cost to the female. These predictions are confirmed by
our observations of male aggression toward females among Crocuta.

Clutton-Brock & Parker (1995) suggested three general forms of sexual
coercion: (1) intimidation, where males punish females who refuse to mate and
thus increase the probability of securing future matings, (2) forced copulation,

where physical force is used to achieve intromission, and (3) harassment, where

137

male aggression imposes costs on females that increase the probability of
immediate mating. Intimidation seems unlikely here considering that adult female
Crocuta are approximately 10% larger than males (Matthews 1939; Wildon 1968;
Kruuk 1972; Frank 1986b), and sport weapons that are effectively
indistinguishable from those of males. Furthermore, we found that non-sires
baited females at higher rates before than after conception, the opposite of the
result predicted by the intimidation hypothesis. Forced copulation can most likely
be ruled out altogether in this species in light of the monomorphic external
genitalia of males and females. That is, the female’s clitoris is elongated into a
fully erectile pseudopenis, through which the urogenital canal extends, and
through which the female urinates, copulates, and gives birth (Matthews 1939;
Neaves et al. 1980; Frank et al. 1990). The bizarre genital morphology of the
female makes rape impossible, and intromission by the male requires complete
female cooperation.

Because intimidation and forced copulation seem unlikely in the spotted
hyena, we suggest that male aggression toward females might function mainly to
elevate stress levels in females or otherwise alter female reproductive
endocrinology. Baiting of lactating females might induce stress that reduces milk
production, and thus hasten the end of lactation. Stress is known to depress milk
production in other mammals (e.g. dairy cows, Bos taurus: Ray et al. 1992;
Tancin et al. 1995). Baiting of female hyenas during the early weeks of
pregnancy might stress females when fetuses are highly vulnerable, and thus

induce abortion (Wasser and Barash 1983). Although we currently have no

138

method for detecting abortions in free-living hyenas, we found that 13 of 20
observed copulations failed to result in the birth of a litter, and thus appeared to
be infertile matings. Baiting during the period between weaning one litter and
conceiving the next might either induce early abortions or alter the reproductive
endocrinology of the non-pregnant female to bring her more rapidly into estrus.
We can infer that baiting imposes costs on female Crocuta because females
sometimes sustain wounds during baiting, females often fight back in ways that
risk further injury, baiting appears stressful for female targets, and baiting
occupies time and energy that females might otherwise spend resting, foraging,
or caring for their cubs. However, the fact that sires were observed baiting
females with whom they had recently mated provides evidence against this
hypothesis, as does our observation that sires baited females at rates that did not
differ before and after conception.

Finally, baiting might function to demonstrate some aspect of male fitness
to prospective mates. This hypothesis predicts that aggressive behavior directed
at females should vary among individual males, and that male reproductive
success should vary with some parameter of male aggressive behavior. Although
rates of baiting did not differ statistically significantly among individual males,
there was a 15-fold difference between the minimum and maximum rates at
which any individual male baited females. This, together with our results showing
that individual males did vary significantly with respect to numbers of aggressive
acts they directed toward females during each baiting, suggests that there are

individual differences in baiting behavior among males, which might potentially

139

reflect male fitness. However, males did not appear to gain direct mating
opportunities as a result of baiting. Males who sired the cubs of a target female
did not bait her at greater rates or with greater intensity than did males who failed
to sire cubs of that female. We might not expect to see as strong a relationship in
this species between rates of baiting behavior and success at sirlng cubs as we
do, for example, between association indices and success at sirlng cubs
(Szykman et al. 2001). This is because most baitings involve a group of males,
whereas only one or two males can potentially sire a female’s litter.

The mechanism by which males are demonstrating their fitness to females
might not be through repeated baitings, as measured by baiting rates, but
through individual acts of aggression exhibited during baiting. That is, females
might assess males based on some particular aspect of individual male behavior
during baiting. Therefore, the hypothesis remains viable that males might
demonstrate some aspect of their fitness to females during baiting. Here, sample
sizes of cases in which we were able to reliably record all specific behavioral acts
by all participating males were too small to yield much statistical power, so this
matter remains unresolved.

In spotted hyenas, male aggression directed towards females remains a
fascinating and mysterious phenomenon. The tight association between male
attacks on females and a female’s time to conception provides strong evidence
for some role for male aggression in hyena sexual behavior. None of the
hypotheses presented here satisfactorily explain all aspects of baiting behavior,

although our data are most consistent with the hypothesis that male attacks on

140

 

females play a role in informing the female about some aspect of male fitness.
Whatever the function of baiting, given the larger body size, peculiar genital
morphology, and social dominance of female Crocuta, if male aggression
towards females is related to sexual behavior, it probably functions in conjunction

with female choice as a mode of sexual selection in this species.

141

Chapter 5
CONCLUSIONS

The inspiration for this dissertation was provided by the exciting findings of
Holekamp and colleagues (1996), in which they determined that reproductive
success varies significantly with social rank among free-living female spotted
hyenas. High-ranking female hyenas begin breeding at younger ages, are more
frequently able to support pregnancy and lactation concurrently, experience
shorter intervals between litters, and their offspring are more likely to survive to
adulthood than are those of lower-ranking females. The finding that social rank
affects female reproductive success was not unique: in several gregarious
mammalian species, high-ranking females enjoy greater reproductive success
than do their lower-ranking counterparts (e.g. primates: Altmann et al. 1988;
ungulates: Clutton-Brock et al. 1988; carnivores: Creel et al. 1992). But the rank-
related variation in reproductive output is extreme in Crocuta; the highest-ranking
females in a single clan benefit from a six-fold greater lifetime reproductive
success than do the lowest-ranking females (Holekamp et al. 1996; Holekamp
and Smale 2000). These results motivated me to inquire what the males were
doing with respect to mate choice and sexual behavior in this peculiar society.
When I first embarked on my study of mate choice and sexual behavior in
spotted hyenas, I did not think that the males would be able to express their mate

choices as straightforwardly as occurs in other polygynous animals in which

142

males are known to be selective (Verrell 1990; Berglund and Rosenqvist 1993).
In general, male mammals in polygynous societies tend to be larger and more
aggressive than females, and they usually compete intensely for access to
breeding females (Darwin 1871; Clutton-Brock 1989). By contrast, adult breeding
Crocuta males are smaller than females and socially subordinate to them. In
addition, males generally exhibit only low levels of intrasexual aggression, and
their courtship behavior is fairly subtle (Kruuk 1972; Frank 1986b). However, with
such substantial variation among female hyenas in regard to their quality as
mates, sexual selection theory predicts that male hyenas should be selective in
their choice of mates (Gwynne 1981; Parker 1983; Petrie 1983; Owens and
Thompson 1994). This study thus sought to examine these theoretical
predictions regarding male mate choice in the spotted hyena. I used behavioral
data collected in the field on a free-living population of spotted hyenas to address
these predictions regarding male mate choice, and subsequently used
information obtained from genetic paternity data to show a relationship between
male mate choice behavior and male reproductive success. A second goal was
to further elucidate the mating system of, and sexual behavior in, this species.
Finally, I wanted to describe male-female interactions not previously reported in
the literature, and to attempt to elucidate the functional significance of a subset of
these.

We examined association patterns among male and female spotted
hyenas in Chapter 2, male and female sexual behavior in Chapter 3, and male

aggression towards females in Chapter 4. All of these behaviors were influenced

143

by female reproductive condition and, to varying extents, by male and female
social rank. We found that male hyenas associate most closely, and in the
greatest numbers, with females most likely to conceive (Chapter 2). Males also
direct a greater number of affiliative behaviors toward females as females
approach the time of conception (Chapter 3). Finally, males direct the greatest
number of aggressive attacks toward females, and males attack females at the
highest rates, during the periods when females are most likely, or are known, to
conceive (Chapter 4).

Since morphological characteristics of female hyenas, and behavioral
signals emitted by females, do not appear to vary with their reproductive
condition, males presumably evaluate female reproductive state using olfactory
cues. Such information obtained by males during investigatory behaviors directed
toward females appears to influence male movement patterns. That is, males
appear to use the female’s reproductive condition, measured here as time to
conception, in making decisions regarding whether or not to associate closely
with a particular female at a given time. It is also possible that females exhibit
other subtle signs reflecting their willingness to mate, including increased
tolerance to the presence of males, which males may perceive and use in their
decision-making processes.

Male hyenas were found to associate more closely with high-ranking than
with low-ranking females (Chapter 2). However, effects of female social rank do
not have as strong an influence on male affiliative behaviors as expected

(Chapter 3). Instead, we found a significant interaction between a female’s rank

144

and her age, whereby males appear to direct more affiliative behaviors towards
older, low-ranking females than toward younger ones. Interestingly, female age
did not influence male attention directed toward high-ranking females. Both of
these results are consistent with known variation in life histories between high-
and low-ranking females. High-ranking females begin producing offspring soon
after reaching reproductive maturity at 2 years of age (Holekamp et al. 1996; this
study), whereas low-ranking females begin breeding at significantly older ages
(Holekamp et al. 1996). Intersexual association patterns and data documenting
affiliative behavior indicated that male hyenas can assess the relative
reproductive value of individual females, and that males appear to allocate their
own reproductive effort among clan females based on this assessment.

It is not known by what mechanism male hyenas are able to discriminate
among females of various ranks. Males might be able to assess relative female
ranks directly by observing dyadic agonistic interactions among females. It is
reasonable to believe that males might be using the same mechanisms to
understand rank relationships among females in their new clans that hyena cubs
do when first learning their own social ranks. That is, in addition to participating in
coalitions that reinforce aggression directed at lower-ranking conspecific targets,
young spotted hyenas may learn their positions in the clan’s social hierarchy by
watching their mothers intervene on their behalf in fights with other clan members
(Engh et al. 2000). It is less clear by what mechanism males are associating high
rank with high reproductive value among females. Males might use female rank

as an indirect measure of female reproductive success, much as male insects

145

and amphibians use female size as an indirect measure of female fecundity
(Gwynne 1981; Verrell 1995). Alternatively, males might be able to recognize
relatedness among natal hyenas. Males may discriminate those females who
produce many offspring from those females who only produce a few by
monitoring association patterns among natal animals (kin associate more closely
than do non-kin: Holekamp et al. 1997), or via MHC odortype cues (Yamazaki et
al. 1994; Todrank et al. 1998). Studies of kin recognition in Crocuta are currently
in progress in the Holekamp lab (R.C. Van Horn and SA. Wahaj, unpublished
data).

Female social rank does not appear to influence male attacks on females
(Chapter 4), suggesting that these attacks are sexually motivated behaviors
directed at a female near her time of conception regardless of her long-term
reproductive value. Most males present in the clan for several years bait females,
and males of higher social rank bait females at higher rates than do lower-
ranking males. It is possible that males of different ranks adopt different
strategies to attract mates. This hypothesis is supported by the male-female
association data. Both high- and middle-ranking males associate most closely
with high-ranking females, but low-ranking males associate about equally with
females of all social ranks. Low-ranking males may either be less adept at
assessing the reproductive value of clan females, or they may utilize different
reproductive strategies than those used by high-ranking males, whose dominant
status may often allow them to exclude lower-ranking males from access to high-

value females.

146

The ability of male hyenas to access or monopolize receptive females
should vary with the degree to which estrous cycles are synchronized among
clan females. Data from Chapter 2 suggest that female reproductive synchrony is
relatively low in our hyena study population, and that males should be able to
monopolize females without incurring significant costs resulting from lost mating
opportunities. However, two issues subsequently arise. First, under certain
ecological conditions, up to nine females may be in estrus concurrently (Chapter
2), and thus males are forced to make decisions regarding in which female they
should invest time and effort. Second, I suggest that it is precisely this time and
effort necessary for males to persuade females to mate with them that requires
males to make informed decisions about with whom to associate closely, even if
females are never concurrently in estrus. At least for some males, this strategy of
“consorting” with a particular female may be related to his chances of mating with
her (see temporal patterns of intersexual association in Chapter 3).

The real challenge to the behavioral data presented in this dissertation is
to unambiguously demonstrate that close intersexual associations and increased
affiliative behavior by males directed at females represent mate choice by males,
and that these measures truly reflect increased reproductive success for males.
Evidence from other polygynous mammalian species indicates that consorting
pairs are frequently observed to copulate, and that consorting males father most
offspring (Packer and Pusey 1982; Bercovitch 1987; Small 1990). In this study,
close and relatively exclusive intersexual associations are associated with

paternity in spotted hyenas (Chapter 2), but the direction of the cause-and-effect

147

relationship involved remains unknown. That is, a male might make himself more
attractive to a particular female by spending more time with her, and thus
enhancing his own chances of siring her cubs. Alternatively, a male might spend
more time with a female because she has already chosen him to sire her cubs.
Clearly our data do not rule out the possibility that close intersexual associations
in Crocuta reflect mate choice by females as well as by males.

I believe I have successfully shown that spotted hyenas are indeed
fascinating subjects for the study of sexual behavior, particularly when their
unique behaviors are considered in the context of sexual selection theory. On the
one hand, hyenas represent a typical polygynous mammalian species in which
males provide no parental care for their offspring, and females are expected to
be selective in their choice of mates. However, whereas male-male competition
is predicted in most polygynous species, male aggression toward other males is
surprisingly low in Crocuta (Kruuk 1972, this study; Frank 1986b). Furthermore,
the high variation in female mate quality observed in Crocuta provides male
hyenas with a strong incentive for mate choice. In this study, I sought to elucidate
sexual behavior in spotted hyenas and examine male behavior indicative of mate
choice. Courtship and mating are unique in this species, and male hyena
behavior appears to reflect some degree of choice for high-ranking, high-quality
females.

There are several questions that emerge from this study and require
additional investigation by future workers. Additional data need to be collected to

determine more clearly the function of male baiting behavior. A new approach

148

might be to address the coordinated behavior among immigrant males when
baiting. Determining whether individual males are cooperating with one another
during baiting to increase their inclusive fitness, as might be reflected in levels of
genetic relatedness among participating males, could shed further light on the
adaptive function of this unusual behavior.

A next logical step would be to attempt to elucidate individual reproductive
strategies and mate choice decisions among female spotted hyenas. Watching a
single female allow two males to alternate mounting her, which we saw on two
separate occasions (Chapter 3) gives observers the impression that the female is
not doing much choosing. We even know from genetic paternity analyses that
about 20% of twin litters are sired by multiple males (Engh et al. 2002). However,
it is entirely possible that female hyenas have made their choices long before
copulation occurs. In addition, it remains possible that female Crocuta can
continue making mate-choice decisions after they have copulated with multiple
males via cryptic choice (Eberhard 1996). indeed, the bizarre and tortuous
configuration of the female reproductive tract in this species (Frank et al. 1995b)
suggests that female hyenas might be able to create competitive conditions for
sperm between copulation and the point at which egg and sperm actually meet.
Because females are socially dominant in Crocuta, and they possess
masculinized genitalia that make copulation by force impossible, it is highly likely
that females are running the mating game and making final choices regarding
which males sire their offspring. Further study of female mate choice should

nicely complement analyses of male mate choice in this species. That is,

149

understanding what a female is looking for in a mate, and determining the
mechanisms by which she evaluates males and acts on her decisions, should

yield further insight into male reproductive strategies.

150

LITERATURE CITED

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

Altmann J, Hausfater G, Altmann SA (1988) Determinants of reproductive
success in savannah baboons, Papio cynocephalus. In: Clutton-Brock TH
(ed) Reproductive Success. University of Chicago Press, Chicago, pp
403-418

Andersson M (1982) Female choice selects for extreme tail length in a
widowbird. Nature 299:818-820

Andersson M (1994) Sexual Selection. Princeton University Press, Princeton

Asa CS, Valdespino C (1998) Canid reproductive biology: an integration of
proximate mechanisms and ultimate causes. American Zoologist 38:251-
259

Bateman AJ ( 1948) lntrasexual selection in Drosophila. Heredity 2:349-368

Beach FA (1976) Sexual attractivity, proceptivity, and receptivity in female
mammals. Hormones and Behavior 7:105-138

Beach FA, Dunbar IF, Buehler MG (1982) Sexual characteristics of female dogs
during successive phases of the ovarian cycle. Hormones and Behavior
16:414-442

Beach FA, LeBoeuf BJ (1967) Coital behavior in dogs. I. Preferential mating in

the bitch. Animal behaviour 15:546-558

151

Beaver BV (1992) Feline behavior: A guide for veterinarians. WB Saunders
Company, Philadelphia

Bercovitch FB (1987) Reproductive success in male savanna baboons.
Behavioral Ecology and Sociobiology 21 :163-172

Bercovitch FB (1991) Mate selection, consortship formation, and reproductive
tactics in adult female savanna baboons. Primates 32:437-452

Bercovitch F3 (1995) Female cooperation, consortship maintenance, and male
mating success in savanna baboons. Animal Behaviour 50:137-149

Bercovitch FB, Sladky KK, Roy MM, Goy RW (1987) Intersexual aggression and
male sexual activity in captive rhesus macaques. Aggressive Behavior
132347-358

Berger J (1983) Induced abortion and social factors in wild horses. Nature
303:59-61

Berger J (1989) Female reproductive potential and its apparent evaluation by
male mammals. Journal of Mammalogy 70:347-3358

Berglund A, Rosenqvist G (1993) Selective males and ardent females in
pipefishes. Behavioral Ecology and Sociobiology 32:331-336

Bertram BCR (1975) Social factors influencing reproduction in lions. Journal of
Zoology 177:463-482

Borgia G (1985) Bower quality, number of decorations and mating success of
male satin bowerbirds (Ptilonorhynchus violaceus): An experimental

analysis. Animal Behaviour 32:266-271

152

Boydston EE, Morelli TL, Holekamp KE (2001) Sex differences in territorial
behavior exhibited by the spotted hyena (Hyaenidae, Crocuta crocuta).
Ethology 107:369-385

Bradbury JW, Gibson RM (1983) Leks and mate choice. In: Bateson P (ed) Mate
Choice. Cambridge University Press, Cambridge, pp 109-138

Bronson FH (1989) Mammalian reproductive biology. University of Chicago
Press, Chicago

Burley N, Moran N (1979) The significance of age and reproductive experience in
the mate preferences of feral pigeons, Columbia Iivia. Animal Behaviour
27:686-698

Bygott JD, Bertram BC, Hanby JP (1979) Male lions in coalitions gain
reproductive advantages. Nature 282:839-41

Cairns SJ, Schwager SJ (1987) A comparison of association indices. Animal
Behaviour 35:1454-1469

Campagna C, LeBoeuf BJ (1988) Reproductive behaviour of southern sea lions.
Behaviour 105:233-261

Caro TM (1994) Cheetahs of the Serengeti Plains. University of Chicago Press,
Chicago

Chakraborty R, Meaher T, Smouse P (1988) Parentage analysis with genetic
markers in natural populations. I. The expected proportion of offspring with
unambiguous paternity. Genetics 118:527-536

Clutton-Brock TH (1989) Mammalian mating systems. Proceedings of the Royal

Society of London, Series B 236:339-372

153

 

Clutton-Brock TH, Albon SD, Guinness FE (1982) Red Deer: Behavior and
Ecology of Two Sexes. University of Chicago Press, Chicago

Clutton-Brock TH, Albon SD, Guinness FE (1988) Reproductive success in male
and female red deer. In: Clutton-Brock TH (ed) Reproductive Success.
University of Chicago Press, Chicago, pp 325-343

Clutton-Brock TH, Parker GA (1995) Sexual coercion in animal societies. Animal
Behaviour 49:1345-1365

Clutton-Brock TH, Vincent ACJ (1991) Sexual selection and the potential
reproductive rates of males and females. Nature 351 :58-60

Connor RC, Smolker RA, Richards AF (1992a) Dolphin coalitions and alliances.
In: Harcourt AH, de Waal FBM (eds) Coalitions and Alliances in Humans
and Other Animals. Oxford University Press, Oxford, pp 415-443

Connor RC, Smolker RA, Richards AF (1992b) Two levels of alliance formation
among male bottlenose dolphins (Tursiops sp.) Proceedings of the
National Academy of Science 89:987-990

C6te lM, Hunte W (1989) Male and female mate choice in the redlip blenny: Why
bigger is better. Animal Behaviour 38:78-88

Cox CR, LeBoeuf BJ (1977) Female incitation of male competition: Mechanism in
sexual selection. American Naturalist 111:317-335

Creel S, Creel N, Wildt DE, Monfort SL (1992) Behavioral and endocrine
mechanisms of reproductive suppression in Serengeti dwarf mongooses.

Animal Behaviour 43:231-245

154

Creel S, Macdonald D (1995) Sociality, group-size, and reproductive suppression
among carnivores. Advances in the Study of Behavior 24:203-257

Creel SR, Creel NM (1995) Communal hunting and pack size in African wild dogs
Lycaon pictus. Animal Behaviour 50:1325-1339

Crews D (1975) Psychobiology of reptilian reproduction. Science 189:1059-1065

Dagg Al, Taub A (1970) Flehmen. Mammalia 342686-695

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

Davies NB (1991) Mating systems. In: Krebs JR, Davies NB (eds) Behavioural
Ecology, an Evolutionary Approach, 3rd edn. Blackwell Scientific
Publications, Oxford, pp 263-294

Derix RRWM, van Hooff JARAM (1995) Male and female partner preferences in
a captive wolf pack (Canis lupus): Specificity versus spread of sexual
attention. Behaviour 132:128-148

Dewsbury DA (1972) Patterns of copulatory behavior in male mammals. The
Quarterly Review of Biology 47:1-33

Doving KB, Trotier D (1998) Structure and function of the vomeronasal organ.
Journal of Experimental Biology 201:2913-2925

East ML, Hofer H, Wickler W (1993) The erect penisis a flag of submission in a
female-dominated society: Greetings in Serengeti spotted hyenas.
Behavioral Ecology and Sociobiology 33:355-370

Eberhard WG (1996) Female Control: Sexual Selection by Cryptic Female

Choice. Princeton University Press, Princeton

155

Emlen ST, Oring LW (1977) Ecology, sexual selection and the evolution of
mating systems. Science 197:215-223

Engh AL, Esch K, Smale L, Holekamp KE (2000) Mechanisms of matenal rank
‘inheritance' in the spotted hyaena, Crocuta crocuta. Animal Behaviour
60:323-332

Engh AL, Van Horn RC, Funk SM, Scribner KT, Bruford MW, Szykman M, Smale
L, Holekamp KE (2002) Reproductive skew among males in a female
dominated mammalian society. Behavioral Ecology In press

Estes RD (1972) The role of the vomeronasal organ in mammalian reproduction.
Mammalia 36:315-341

Estes RD, Goddard J (1967) Prey selection and hunting behavior of the African
wild dog. Journal of Wildlife Management 32:52-76

Ewer RF (1973) The Carnivores. Cornell University Press, Ithaca

Forsberg J (1987) A model for male mate discrimination in butterflies. Oikos
49:46-54

Frank LG (1986a) Social organization of the spotted hyaena (Crocuta crocuta). 1.
Demography. Animal Behaviour 34:1500-1509

Frank LG (1986b) Social organization of the spotted hyaena Crocuta crocuta. 2.
Dominance and reproduction. Animal Behaviour 34:1510-1527

Frank LG (1997) Evolution of genital masculinization: Why do female hyaenas

have such a large ’penis'? Trends in Ecology and Evolution 12:58-62

156

Frank LG, Glickman SE (1994) Giving birth through a penile clitoris: Parturition
and dystocia in the spotted hyena (Crocuta crocuta). Journal of Zoology
234:659-665

Frank LG, Glickman SE, Powch l (1990) Sexual dimorphism in the spotted hyena
(Crocuta crocuta). Journal of Zoology 221 :308-313

Frank LG, Holekamp KE, Smale L (1995a) Dominance, demography, and
reproductive success of female spotted hyenas. In: Sinclair ARE, Arcese
P (eds) Serengeti ll: Dynamics, management, and conservation of an
ecosystem. University of Chicago Press, Chicago, pp 364-384

Frank LG, Weldele ML, Glickman SE (1995b) Masculinization costs in hyaenas.
Nature 377:584-585

Gibson RM, Langen TA (1996) How do animals choose their mates? Trends in
Ecology and Evolution 11:468-470

Golding RR (1969) Birth and development of spotted hyaenas Crocuta crocuta at
the University of Ibadan Nigeria. lntemational Zoo Yearbook 9:93-95

Goodall J (1986) The chimpanzees of Gombe: Patterns and Behavior. Harvard
University Press, Cambridge

Grimpe G ( 1916) Hyanologische Studien. Zool Anz 48:49-61

Gross MR, Shine R (1981) Parental care and mode of fertilization in ectothermic
vertebrates. Evolution 352775-793

Gwynne DT (1981) Sexual difference theory: Mormon crickets show role reversal

in mate choice. Science 213:779-780

157

Halliday TR (1983) The study of mate choice. In: Bateson PPG (ed) Mate
Choice. Cambridge University Press, Cambridge, pp 3-32

Hamilton WJ, Tilson RL, Frank LG (1986) Sexual monomorphism in spotted
hyenas, Crocuta crocuta. Ethology 71 :63-73

Hausfater G (1975) Dominance and reproduction in baboons. Contributions in
Primatology 7:1-150

Hayssen V, van Tienhoven A, van Tienhoven A (1993) Adeli's Patterns of
Mammalian Reproduction. Cornell University Press, Ithaca

Henschel JR, Skinner JD (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

Hill DA (1987) Social relationships between adult male and female rhesus
macaques. 1. Sexual consortships. Primates 28:439-456

Hill GE (1990) Female house finches prefer colourful males: Sexual selection for
a condition-dependent trait. Animal Behaviour 40:563-572

Hofer H, East ML (1993) The commuting system of Serengeti spotted hyaenas:
How a predator copes with migratory prey. 3. Attendance and maternal
care. Animal Behaviour 46:575-589

Hofer H, East ML (1995) Virilized sexual genitalia as adaptations of female
spotted hyaenas. Revue Suisse De Zoologie 102:895-906

Hoikkala A, Aspi J, Suvanto L (1998) Male courtship song frequency as an
indicator of male genetic quality in an insect species, Drosophila montana.

Proceedings of the Royal Society London, Series B 265:503-508

158

Holekamp KE, Boydston EE, Smale L (2000) Group travel in social carnivores.
In: Boinski S, Garber PA (eds) On the move: How and why animals travel
in groups. University of Chicago Press, Chicago, pp 587-627

Holekamp KE, Cooper SM, Katona Cl, Berry NA, Frank LG, Smale L (1997)
Patterns of association among female spotted hyenas (Crocuta crocuta).
Journal of Mammalogy 78:55-64

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

Holekamp KE, Smale L (1993) Ontogeny of dominance in free-living spotted
hyaenas: Juvenile rank relations with other immature individuals. Animal
Behaviour 462451-466

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

Holekamp KE, Smale L (2000) Feisty females and meek males. In: Wallen K,
Schneider JE (eds) Reproduction in Context: Social and Environmental
Influences on Reproduction. MIT Press, Cambridge, pp 257-285

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

Hosmer DW, Lemeshow S (2000) Applied Logistic Regression, Second edn.
John Wiley & Sons, Inc., New York

Johnston RE (1983) Chemical signals and reproductive behavior. In:
Vandenbergh JG (ed) Pheromones and Reproduction in Mammals.

Academic Press, New York, pp 3-37

159

Karino K (1995) Male-male competition and female mate choice through
courtship display in the territorial damselfish Stegastes nigricans. Ethology
100:126-138

King C (1989) The Natural History of Weasels and Stoats. Comstock Publishing
Associates, Ithaca

Komers PE, Messier F, Flood PF, Gates CG (1994) Reproductive behavior of
male wood bison in relation to progesterone level in females. Journal of
Mammalogy 75:757-765

Koretz BK (1992) Excretion of testosterone, estradiol, and progesterone in the
feces of spotted hyena (Crocuta crocuta). Undergraduate honors thesis,
University of California, Berkeley

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

Kvarnemo C, Ahnesjo | (1996) The dynamics of operational sex ratios and
competition for mates. Trends in Ecology and Evolution 11:404-408

Kvarnemo C, Forsgren E, Magnhagen C (1995) Effects of sex ratio on intra- and
inter-sexual behaviour in sand gobies. Animal Behaviour 50:1455—1461

LeBoeuf BJ, Mesnick S (1990) Sexual behavior of male northern elephant seals:
I. Lethal injuries to adult females. Behaviour 116:143-162

Legendre P, Legendre L (1998) Numerical Ecology, Second edn. Elsevier,
Amsterdam

Libants S, Olle E, Oswald K, Scribner KT (2000) Microsatellite loci in the spotted

hyena Crocuta crocuta. Molecular Ecology 9:1443-1445

160

Licht P, Frank LG, Pavgi S, Yalcinkaya TM, Siiteri PK, Glickman SE (1992)
Hormonal correlates of masculinization in female spotted hyaenas
(Crocuta crocuta). 2. Maternal and fetal steroids. Journal of Reproduction
and Fertility 95:463-474

Lindburg DG (1983) Mating behavior and estrus in the Indian rhesus monkey. In:
Seth PK (ed) Perspectives in Primate Biology. Today and Tomorrow, New
Delhi, pp 45-61

Lindeque M (1981) Reproduction in the spotted hyena Crocuta crocuta
(Erxleben). MSc thesis, University of Pretoria, Pretoria

Linklater WL, Cameron EZ, Minot EO, Stafford KJ (1999) Stallion harassment
and the mating system of horses. Animal Behaviour 582295-306

Littlejohn MJ, Loftus-Hills JJ (1968) An experimental evaluation of premating
isolation in the Hyla ewingi complex (Anura: Hylidae). Evolution 22:659-
663

Marshall TC, Slate J, Kruuk LEB, Pemberton JM (1998) Statistical confidence for
likelihood-based paternity inference in natural populations. Molecular
Ecology 7:639-655

Matthews LH (1939) Reproduction in the spotted hyena Crocuta crocuta
(Erxleben). Philosophical Transactions Royal Society, Series B 23021-78

McKinney F, Derickson SR, Mineau P (1983) Forced copulation in waterfowl.

Behaviour 86:250-294

161

Meagher TR (1986) Analysis of paternity within a natural population of
Chemaelirium Iuteum. I. Identification of most-likely male parents.
American Naturalist 128:199-215

Mech LD (1970) The Wolf: The Ecology and Behavior of an Endangered
Species. Natural History Press, Garden City

Melnick DJ, Pearl MC (1987) Cercopithecines in multimale groups: Genetic
diversity and population structure. In: Smuts BB, Cheney DL, Seyfarth
RM, Wrangham RW, Struhsaker TT (eds) Primate Societies. University of
Chicago Press, Chicago, pp 318-329

Mesnick SL, Le Boeuf BJ (1991) Sexual behavior of male northern elephant
seals: ll. Female response to potentially injurious encounters. Behaviour
117:262-280

Mills MGL (1990) Kalahari Hyenas: The comparative behavioral ecology of two
species. Chapman and Hall, London

Mitani JC (1985) Mating behaviour of male orangutans in the Kutai Game
Reserve, Indonesia. Animal Behaviour 332392-402

Moehlman PD (1989) lntraspecific variation in canid social systems. In: Gittleman
JL (ed) Carnivore Behavior, Ecology, and Evolution, vol 1. Cornell
University Press, Ithaca, pp 143-163

Neaves WB, Griffin JE, Wilson JD (1980) Sexual dimorphism of the phallus in
spotted hyena (Crocuta crocuta). Journal of Reproduction and Fertility

59:506-51 2

162

Oring LW, Reed JM, Colwell MA, Lank DB, Maxson SJ (1991) Factors regulating
annual mating success and reproductive success in spotted sandpipers
(Actitis macularia). Behavioral Ecology and Sociobiology 28:433-442

Owens IPF, Thompson DBA ( 1994) Sex differences, sex ratios and sex roles.
Proceedings of the Royal Society of London, Series B 258:93-99

Packer C, Pusey AE (1982) Cooperation and competition within coalitions of
male lions: kin selection or game theory. Nature 296:740-742

Packer C, Pusey AE (1983) Male takeovers and female reproductive parameters:
a simulation of oestrus synchrony in lions (Panthera Ieo). Animal
behaviour 31 :334-40

Parker GA (1983) Mate quality and mating decisions. In: Bateson P (ed) Mate
Choice. Cambridge University Press, Cambridge, pp 141-166

Parker GA, Baker RR, Smith VGF (1972) The origin and evolution of gamete
dimorphism and the male-female phenomenon. Journal of Theoretical
Biology 36:529-553

Petrie M (1983) Mate choice in role-reversed species. In: Bateson P (ed) Mate
Choice. Cambridge University Press, Cambridge, pp 167-179

Ptacek MB (2000) The role of mating preferences in shaping interspecific
divergence in mating signals in vertebrates. Behavioural Processes
51 :1 1 1-134

Queller DC, Goodnight KF (1989) Estimating relatedness using genetic markers.

Evolution 432258-275

163

 

Rasa OAE (1977) The ethology and sociology of the dwarf mongoose (Helogale
undulata rufula). Z Tierpsychol 432337-406

Ray DE, Halbach TJ, Armstrong DV (1992) Season and lactation number effects
on mild production of dairy cattle in Arizona. Journal of Dairy Science
75:2976-2983

Real L (1990) Search theory and mate choice: I. Models of single-sex
discrimination. American Naturalist 136:376-404

Robinson JG (1982) lntrasexual competition and mate choice in primates.
American Journal of Primatology Supplement 1:131-144

Sargent RC, Gross MR, den Berghe EP (1986) Male mate choice in fishes.
Animal Behaviour 34:545-550

Schaller GB (1972) The Serengeti Lion: A Study in Predator-Prey Relations.
University of Chicago Press, Chicago

Schneider KM (1926) Uber Hyanenzucht. Die Pelztierzucht 2:1-4, 9-11, 12-14

Schwagmeyer PL (1995) Searching today for tomorrow's mates. Animal
Behaviour 50:759-767

Schwagmeyer PL, Parker GA (1990) Male mate choice as predicted by sperm
competition in thirteen-lined ground squirrels. Nature 348262-64

Seyfarth RM (19783) Social relationships among adult male and female
baboons. l. Behavior during sexual consortships. Behaviour 64

Seyfarth RM (1978b) Social relationships among adult male and female
baboons. ll. Behavior throughout the female reproductive cycle. Behaviour

64

164

Silk JB, Boyd R ( 1983) Cooperation, competition, and mate choice in matrilineal
macaque groups. In: Wasser SK (ed) Social Behavior of Female
Vertebrates. Academic Press, New York, pp 315-347

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

Small MF (1990) Consortships and conceptions in captive rhesus macaques
(Macaca mulatta). Primates 312339-350

Smuts BB ( 1983) Spatial relationships between adult male and female olive
baboons. In: Hinde RA (ed) Primate Social Relationships. Sinauer
Associate, Inc., Sunderland, pp 262-266

Smuts BB (1985) Sex and Friendship in Baboons. Harvard University Press,
Cambridge

Smuts BB (19873) Gender, aggression, and influence. In: Smuts BB, Cheney DL,
Seyfarth RM, Wrangham RW, Struhsaker TT (eds) Primate Societies.
University of Chicago Press, Chicago, pp 400-412

Smuts BB (1987b) Sexual competition and mate choice. In: Smuts BB, Cheney
DL, Seyfarth RM, Wrangham RW, Struhsaker TT (eds) Primate Societies.

University of Chicago Press, Chicago, pp 385-399

165

Smuts BB, Smuts RW (1993) Male aggression and sexual coercion of females in
nonhuman primates and other mammals: Evidence and theoretical
implications. Advances in the Study of Behavior 22:1-63

Solomon NG ( 1993) Body size and social preferences of male and female prairie
voles, Microtus ochrogaster. Animal Behaviour 45:1031-1033

Soltis J, Mitsunaga F, Shimizu K, Yanagihara Y, Nozaki M (1997) Sexual
selection in Japanese macaques: I. Female mate choice or male sexual
coercion? Animal Behaviour 54:725-736

Stammback E (1987) Desert, forest, and montane baboons: Multilevel societies.
In: Smuts BB, Cheney DL, Seyfarth RM, Wrangham RW, Struhsaker TT
(eds) Primate Societies. University of Chicago Press, Chicago, pp 112-
120

Sterck EHM, Watts DP, van Schaik CP (1997) The evolution of female social
relationships in nonhuman primates. Behavioral Ecology and Sociobiology
412291-309

Stern BR, Smith DG (1984) Sexual behaviour and paternity in three captive
groups of rhesus monkeys (Macaca mulatta). Animal Behaviour 32:23-32

Stone GN (1995) Female foraging responses to sxual harassment in the solitary
bee Anthophora plumipes. Animal Behaviour 50:405-412

Systat (1998) Systat 8.0: Statistics. SPSS, Inc., Chicago

Szykman M, Engh AL, Van Horn RC, Funk SM, Scribner KT, Holekamp KE

(2001) Association patterns among male and female spotted hyenas

166

(Crocuta crocuta) reflect male mate choice. Behavioral Ecology and
Sociobiology 50:231-238

Tancin V, Harcek L, Broucek J, Uhrincat M, Mihina S (1995) Effect of suckling
during early lactation and changeover to machine milking on plasma
oxytoci and cortisol levels and milking characteristics in Holstein cows.
Journal of Dairy Research 62:249-256

Tardiff SD (1997) The bioenergetics of prental behavior and the evolution of
alloparental care in marmosets and tamarins. In: Solomon NG, French JA
(eds) Cooperative Breeding in Mammals. Cambridge University Press,
Cambridge, pp 11-33

Thompson EA (1975) The estimation of pairwise relationships. Annals of
HumanGenetics 39:173-188

Thornhill R, Gwynne DT (1986) The evolution of sexual differences in insects.
American Scientist 742382-389

Tilson RL, Hamilton WE (1984) Social dominance and feeding patterns of
spotted hyaenas. Animal Behaviour 28:52-59

Todrank J, Heth G, Johnston RE (1998) Kin recognition in golden hamsters:
evidence for kinship odours. Animal Behaviour 55:377-386

Trivers RL (1972) Parental investment and sexual selection. In: Campbell B (ed)
Sexual Selection and the Descent of Man. Aldine, Chicago, pp 136-179

Vargas A, Anderson SH (1998) Black-footed ferret (Mustela nigripes) behavioral
development: Aboveground activity and juvenile play. Journal of Ethology

16:29-41

167