MECHANISMS MEDIATING LIFE HISTORY TRAITS IN THE SPOTTED HYENA (CROCUTA
CROCUTA)
By
Nora Shannon Lewin

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Integrative Biology – Doctor of Philosophy
Ecology, Evolutionary Biology and Behavior – Dual Major
2017

ABSTRACT
MECHANISMS MEDIATING LIFE HISTORY TRAITS IN THE SPOTTED HYENA (CROCUTA
CROCUTA)
By
Nora Shannon Lewin
My dissertation focuses on understanding the mechanisms underpinning growth,
reproduction, and survival in the spotted hyena. Following a general introductory chapter, my
dissertation is composed of four independent research chapters. I begin with Chapter 2 in which
my colleagues and I document a positive linear relationship between social dominance rank and
telomere length. We also report significant variability in telomere length of high-ranking females
among different social groups, suggesting that both social dominance rank and group
membership influence this important biomarker of aging. In Chapter 3, we describe the role of
juvenile concentrations of the hormone, insulin-like growth-factor -1 (IGF-1), in predicting tradeoffs between early-life growth and later-life reproduction and survival among female hyenas. In
Chapter 4, I explore IGF-1 as a potential mechanism of female-biased sexual size dimorphism
by documenting sex-biased concentrations, sensitivities, and adaptive values of IGF-1 during
the early postnatal period. Finally, in Chapter 5, I describe that age-related improvement and
senescence in reproductive performance varies with social dominance rank among female
hyenas. Cumulatively, my dissertation is an exploration of how physiological mechanisms may
be used to understand social, physiological, and evolutionary forces operating in a free-living
social carnivore. My work offers a unique contribution to the field of life-history evolution and
furthers our understanding of the mechanisms that give rise to it.

I dedicate this thesis to my late father, John R. Lewin.

iii

ACKNOWLEDGEMENTS

This dissertation could not have been possible without the guidance, support, and
insightful contributions from my wonderful committee members, Laura Smale, Marc Breedlove,
Tom Getty, and my major advisor, Kay Holekamp. Kay is a touchstone of academic rigor and
excellence, and I can’t imagine having accomplished all that I have without her generosity,
guidance, and invaluable contributions to my growth and development. She is as intellectually
fierce as she is generous and hospitable, and is an inspiration to all that are fortunate to meet
her.
This research was supported by National Science Foundation (NSF) grants
IOS1121474, and DEB1353110 and National Institutes of Health grant 1R01GM105042 to Kay
Holekamp. I was supported by a NSF Graduate Research Fellowship and grants from the
Michigan State University (MSU) College of Natural Sciences, the MSU John C. Shaver
Research Fellowship in Zoology, the MSU Council of Graduate Students, the interdisciplinary
program in Ecology, Evolutionary Biology, and Behavior at MSU, the Society for Integrative and
Comparative Biology, and the Animal Behavior Society. I also thank the Kenyan National
Council for Science, Technology, and Innovation for permission to conduct this research, and
the Kenya Wildlife Service, Narok County Council and the Senior Warden of the Masai Mara
National Reserve for assistance.
This dissertation literally could not have been possible without the contributions from all
past and present research assistants and graduate students who have worked tirelessly for the
MSU Mara Hyena Project since its inception in 1988. The data exist because of you and your
hard work, and I can’t thank you enough. In particular, I thank Pat Bills, Andy Booms, David
Green, Julia Greenberg, Katy Califf, Sheri Congdon, Hadley Couraud, Leslie Curren, Andy
Flies, Timothy Ikime, Lily Johnson-Ulrich, Sarah Jones, Joseph Kamaamia, Wilson Kilong,

iv

Noemie Lamon, Zach Laubach, Agathe Laurence, Kenna Lehmann, Michelle Lindson, Kevin
McCormick, Tracy Montgomery, Philimon Naiguran, Jenna Parker, Benson Pion, Kate Shaw, Eli
Strauss, Eli Swanson, Julie Turner, and Dee White. Asante sana!
I would like to thank my collaborators and acknowledge others whose suggestions
greatly improved the manuscripts of, and the science behind, my first two chapters. For Chapter
One, I thank Mark Haussmann, Lisa Treidel, Ned Place, Kim Scriber, and Janette Kanefsky.
Also big thanks to Melissa Slane for generously housing me during my trips to Lewisburg, PA.
For Chapter Two, I thank Eli Swanson, Barry Williams, Tom Getty, Ben Dantzer, and Michael
Cant and others from the 2016 Early-life workshop in England.
I would like to thank other exceptional faculty and staff in the wonderful MSU community
for their support and kindness over the years. I thank Richard Hill for his warmth, friendship, and
intellectual inspiration over the years - and just across the hall! A big terima kasih to Pam
Rasmussen and Gabe Ording for letting me join an unforgettable “Wild Borneo” experience. I
thank Dr. Inge Steglitz, Joy Potter Milano, and Rique Campa for their enthusiasm and guidance
in the FAST program. I thank Mike Presocki, Sue Platte, Teri Roache, Lisa Craft, and Julia
Ahmed for their administration assistance and help getting all my forms in on time!
My time in Michigan would have been a lot rougher had it not been for some truly
amazing friends. In particular, I thank Béné Bachelot, Briana Brady, Rachael Eaton, Aaron
Florn, Dustin Kincaid, Agathe Laurence, Kenna Lehmann, Tracy Montgomery, Genevieve Pang,
Eli Strauss, and Sean Williams for countless study sessions, snack breaks, board game nights,
potlucks, and yoga classes. It goes to show that you can really live anywhere as long as you
have great people to share it with.
To my favorite Michigan native, best friend, and husband, Austin Dreyer: I met you on
my first day of graduate school, and, well, we knew it. Thank you for everything.
Lastly, and most importantly, I thank my twin sister, Betsy, my mom, Katie, and my late
father, John, for their endless support throughout all of my adventures, especially this crazy

v

Ph.D. I could not have been able to cross this finish line without your love. Thank you.

vi

TABLE OF CONTENTS

LIST OF TABLES ...................................................................................................................... ix
LIST OF FIGURES ................................................................................................................... xii
CHAPTER 1: GENERAL INTRODUCTION ............................................................................... 1
General Introduction .............................................................................................................. 2
Summary of Chapter 2 ................................................................................................... 3
Summary of Chapter 3 ................................................................................................... 4
Summary of Chapter 4 ................................................................................................... 4
Summary of Chapter 5 ................................................................................................... 5
CHAPTER 2: SOCIOECOLOGIAL VARIABLES PREDICT TELOMERE LENGTH IN
WILD SPOTTED HYENAS......................................................................................................... 7
Abstract.................................................................................................................................. 8
Introduction ........................................................................................................................... 8
Materials and Methods ......................................................................................................... 9
Study species ................................................................................................................. 9
Data collection ...............................................................................................................10
Telomere analysis .........................................................................................................11
Statistics.........................................................................................................................12
Results .................................................................................................................................12
Discussion ............................................................................................................................16
APPENDIX ...........................................................................................................................18
CHAPTER 3: JUVENILE CONCENTRATIONS OF IGF-1 PREDICT LIFE-HISTORY TRADEOFFS IN A WILD MAMMAL ......................................................................................................20
Introduction ..........................................................................................................................21
Materials and Methods ........................................................................................................23
Relevant biology of spotted hyenas ...............................................................................24
Study site, data collection, and determination of life-history and fitness traits ................25
Measuring IGF-1 concentrations in plasma ...................................................................27
Statistical analyses .........................................................................................................28
Results .................................................................................................................................31
Predictors of juvenile IGF-1 concentrations ...................................................................31
Juvenile body size .........................................................................................................31
Age at first parturition ....................................................................................................31
Survival to reproductive maturity ..................................................................................31
Survival after reproductive maturity ...............................................................................31
Fitness measures .......................................................................................................32
Discussion ............................................................................................................................35
APPENDIX ...........................................................................................................................40
CHAPTER 4: INSULIN-LIKE GROWTH FACTOR-1 (IGF-1) AS A POTENTIAL MEDATOR OF
FEMALE-BIASED SEXUAL SIZE DIMORPHISM IN WILD FEMALE SPOTTED HYENAS .......55
Introduction ..........................................................................................................................56
Materials and Methods ........................................................................................................60
Field study site, subjects, and quantification of independent variables ..........................60

vii

Quantification of plasma IGF-1 concentrations ..............................................................62
Data analysis.................................................................................................................62
Results .................................................................................................................................63
Variation in IGF-1 concentrations with age and sex .......................................................63
Sex-specific sensitivity of IGF-1 concentrations among males and females ..................66
Male survival to reproductive maturity ...........................................................................69
Discussion ............................................................................................................................71
APPENDIX ...........................................................................................................................76
CHAPTER 5: AGE-RELATED REPRODUCTIVE SUCCESS VARIES WITH DOMINANCE
RANK AMONG FEMALE SPOTTED HYENAS .........................................................................82
Introduction ..........................................................................................................................83
Materials and Methods ........................................................................................................85
Female age at last parturition and lifespan ....................................................................86
Lifetime reproductive success (LRS) .............................................................................86
Age of cub at weaning ...................................................................................................86
Inter-birth interval (IBI) ...................................................................................................87
Cub survival...................................................................................................................87
Statistical analyses ........................................................................................................87
Results .................................................................................................................................88
Lifetime reproductive success (LRS) .............................................................................90
Age of cub at weaning - singleton ..................................................................................92
Age of cub at weaning - twin litter ..................................................................................92
Inter-birth interval (IBI) after singleton litter ....................................................................92
Inter-birth interval (IBI) after twin litter ............................................................................93
Cub survival...................................................................................................................95
Discussion ............................................................................................................................95
APPENDIX .........................................................................................................................101
LITERATURE CITED ..............................................................................................................106

viii

LIST OF TABLES

Table 2.1. (a) Results from multiple regression model predicting telomere length among highranking females only (n = 30 telomere samples from 26 hyenas). The Talek clan is set as the
reference category. Bolded P-values indicate statistically significant differences at an alpha =
0.05
..........................................................................................13
Table A.1. Information on repeated samples from the same individuals ...................................19
Table A.2. Information on the demography and terriroties of sampled clans .............................19
Table A.3. Results from a simple linear regression of offspring adult telomere length on
maternal telomere length (df = 7) ..........................................................................................19
Table B.1. A commercially available enzyme immunoassay kit was validated for spotted hyenas
using the ‘spike and recovery’ method following kit protocol .................................................... 41
Table B.2. Concentrations of juvenile IGF-1 concentrations in sampled female hyenas were not
significantly influenced by maternal dominance rank, age at sampling, or prey density.
Significant effects are highlighted in boldface at a significance level of α = 0.05 .......................41
Table B.3. Juveniles with high concentrations of IGF-1 were heavier, but not taller, than
juveniles with low concentrations of IGF-1. Juveniles born to high-ranking mothers were both
heavier and taller than those born to low-ranking mothers. Mean mass ± SE = 22.88 ± 0.74 kg;
mean shoulder height ± SE = 60.39 ± 0.60 cm. Significant effects are highlighted in boldface at
a significance level of α = 0.05
..........................................................................................42
Table B.4a. Comparison of two multiple linear regression models predicting age at first
parturition using Akaike’s Information Criteria corrected for small sample sizes (AICc). Models
are comparable if ∆AICc<2.0. Only individuals with both IGF-1 and mass measurements were
included in models for comparison (n = 37) ...............................................................................43
Table B.4b. Female hyenas were significantly younger at their first parturition when they
exhibited higher juvenile concentrations of IGF-1, were born to higher-ranking mothers, and
experienced higher local prey density during the 6 months before parturition than other females.
Mean age at first parturition among our samples ± SE = 3.67 ± 0.13 yrs. Models correspond to
those in Table 3.4a except that all available data points were used in Model 1 (1 individual with
IGF-1 but no juvenile mass measurement). Significant effects are highlighted in boldface at a
significance level of α = 0.05
..........................................................................................44
Table B.5. Results from two separate Cox proportional hazards models predicting survival to
reproductive maturity. Heavier juvenile female hyenas experienced significantly greater survival
to reproductive maturity than did smaller females. Model 2 was the best-supported model based
on likelihood ratio tests. One individual had no mass recorded so was not included in Model 2,
in which mass was a predictor variable; however, this individual did have an IGF-1 value so was
included in Model 1. Significant effects are highlighted in boldface at a significance level of α =
0.05
..........................................................................................45

ix

Table B.6. Results from two separate Cox proportional hazards models predicting survival past
reproductive maturity. Individuals with higher juvenile concentrations of IGF-1 were more likely
to survive than those with lower juvenile concentrations. Model 1 was the best-supported model
based on likelihood ratio tests. One individual had no mass recorded so was not included in
Model 2, in which mass was a predictor variable; however, this individual did have an IGF-1
value so was included in Model 1. Significant effects are highlighted in boldface at a significance
level of α = 0.05
..........................................................................................46
Table B.7a. Comparison of three multiple linear regression models predicting annual
reproductive success using Akaike’s Information Criteria corrected for small sample sizes
(AICc). Models are comparable if ∆AICc<2.0. Females were only included in models for
comparison if they survived past 3 years of age, had known death dates, had successfully
weaned at least one offspring, and had complete IGF-1, mass, and age at first parturition
measurements (n = 25).
..........................................................................................47
Table B.7b. Results from three linear models to determine a female’s annual reproductive
success. Models correspond to those in Table 3.7a except that all available data points for each
model were used below. Significant effects are highlighted in boldface at a significance level of
α = 0.05
..........................................................................................48
Table B.8a. Comparison of four multiple linear regression models predicting lifetime
reproductive success using Akaike’s Information Criteria corrected for small sample sizes
(AICc). Models are comparable if ∆AICc<2.0. Females were only included in models for
comparison if they survived past 3 years of age, had known death dates, and had complete
IGF-1, mass, and age at first parturition measurements (n = 29)...............................................49
Table B.8b. Results from three linear models to determine a female’s lifetime reproductive
success. All available data points for each model were used below. Significant effects are
highlighted in boldface at a significance level of α = 0.05. .........................................................50
Table 4.1. Results from (a) multiple linear regression models to estimate influence of socioecological factors on IGF-1 concentrations among juvenile males, ages 0.5-1.0 years (n = 48).
(b) Contrasts using least-squares means for intra-litter dominance rank. P-values indicating
statistically significant differences at α = 0.05 are bolded. .........................................................67
Table 4.2. Results from (a) multiple linear regression models to estimate influence of socioecological factors on IGF-1 concentrations among juvenile males, ages 0.5-1.0 years (n = 48).
(b) Contrasts using least-squares means for intra-litter dominance rank. P-values indicating
statistically significant differences at α = 0.05 are bolded. .........................................................68
Table C.1. Distribution of samples, excluding 12 samples from 9 pregnant females (mean IGF-1
± s.e. (ng/ml) = 176.89 ± 15.98). With the sexes pooled over the growth period of 1-4 years of
age, females displayed significantly higher IGF-1 concentrations than males (W = 4151, p =
0.0007).
..........................................................................................77
Table C.2. Results from a multiple linear regression model estimating influence of socioecological factors on relative juvenile mass among 43 juvenile males (ages 0.5-1.0 years).
Mean mass ± SE = 18.90 ± 0.77 kg. Significant effects are highlighted in boldface at a
significance level of α = 0.05.
..........................................................................................77

x

Table C.3. Results from a multiple linear regression model estimating influence of socioecological factors on relative juvenile mass among 69 juvenile females (ages 0.5-1.0 years).
Mean mass ± SE = 22.88 ± 0.74 kg. Significant effects are highlighted in boldface at a
significance level of α = 0.05.
..........................................................................................78
Table C.4. Results from a Cox proportional hazards model predicting male survival to
reproductive maturity. Males of larger relative juvenile mass between 0.5 and 1.0 years did not
survive proportionally better or worse than smaller males. Model 2 was the best-supported
model based on log likelihood with a higher value indicating a better fitting model ....................79
Table C.5. Results from two separate Cox proportional hazards models predicting female
survival to reproductive maturity. Heavier juvenile female hyenas experienced significantly
greater survival to reproductive maturity than did smaller females. Model 2 was the bestsupported model based on likelihood ratio tests. One individual had no mass recorded so was
not included in Model 2, in which mass was a predictor variable; however, this individual did
have an IGF-1 value so was included in Model 1. Significant effects are highlighted in boldface
at a significance level of α = 0.05 ..........................................................................................80
Table D.1. Summary table of 426 litters born to 113 unique female hyenas using over 27 years
of longitudinal data. The % of total litters born with at least one cub surviving to weaning is
rounded to the nearest integer value. This % was only calculated for age bins where ≥ 5 litters
were recorded. Maternal rank was statistically analyzed as a continuous variable but shown
here as a categorical variable for illustrative purposes only .....................................................102
Table D.2. Effects of the random effects of female identity (ID) and offspring birth year in five
separate linear mixed-effects models measuring reproductive performance. Random effects that
explained a significant portion of the residual variance in each model are highlighted in boldface
at a significance level of α = 0.05 ........................................................................................103
Table D.3. Results from models that best predicted reproductive performance in 5 traits.
Significant effects are highlighted in boldface at a significance level of α = 0.05 .....................104
Table D.4. Results from the best model predicting cub survival to 2 years. Significant effects are
highlighted in boldface at a significance level of α = 0.05 ........................................................104

xi

LIST OF FIGURES

Figure 2.1. Mean telomere length in relation to relative dominance rank in adult hyenas.
Regression line shown is from a linear mixed-effects model with all clans included. A dominance
rank of 1 represents the highest-ranking animal within a clan at the time of sampling, and the
individual with a dominance rank of -1 represents the lowest-ranking. Symbol pattern denotes
the clan to which each individual belonged. Circles: females; triangles: males ..........................14
Figure 2.2. Mean telomere length among high-ranking females from 5 clans. Box plots show
group mean (line within box) ± s.d. (box) and range (whiskers) .................................................15
Figure 3.1. Female hyenas exhibiting higher concentrations of IGF-1 as juveniles were (A)
heavier as juveniles and (B) gave birth to their first litter at younger ages. Females with higher
juvenile plasma IGF-1 concentrations were (A) relatively heavy for their age (t64 = 3.80, P =
0.0006) and (B) gave birth to their first litters at younger ages (t34 = -2.15; P = 0.039) than did
those with lower juvenile plasma IGF-1 concentrations. Regression lines are from the linear
regression models
..........................................................................................33
Figure 3.2. Survival curves to reproductive maturity for juvenile females of varying relative
mass. Female hyenas that were relatively heavy as juveniles survived to reproductive maturity
at higher rates than those that were relatively light as juveniles (likelihood ratio test: n= 61, χ2 =
6.17, df = 1, P = 0.01). Right-censored tick marks (+) indicate females that survived past 3 years
of age (n = 39). Relative mass was statistically analyzed as a continuous variable
(Supplementary Table 5), but partitioned into tertiles for illustrative purposes only ...................34
Figure 3.3. Survival curves after reproductive maturity for females exhibiting varying
concentrations of IGF-1 as juveniles. Female hyenas exhibiting higher concentrations of IGF-1
as juveniles experienced significantly lower rates of survival after reproductive maturity than did
hyenas exhibiting lower juvenile IGF-1 concentrations (likelihood ratio test: n= 39, χ2 = 7.88, df
= 2, P = 0.02). Tick marks (+) indicate females alive at the end of the study period (n = 6). IGF-1
was statistically analyzed as a continuous variable (Table 3.6), but classified into tertiles for
illustrative purposes only
..........................................................................................35
Figure B.1. Morphological traits measured during routine immobilization of study animals.
Shoulder height used for this study indicated by measure “c” above. Reproduced from:
Swanson, E.M., Dworkin, I. & Holekamp, K.E. (2011) Lifetime selection on a hypoallometric size
trait in the spotted hyena. Proceedings of the Royal Society B: Biological Sciences, 278, 3277–
85
..........................................................................................51
Figure B.2. IGF-1 standards (open circles; bottom x-axis) and serial dilutions of a pooled
sample of hyena plasma (filled circles; inset x-axis) plotted against optical density values (yaxis). Samples for this study were diluted to 1:20......................................................................52
Figure B.3. Female spotted hyenas that were younger at first parturition experienced higher
annual reproductive success than others (N = 26 females that survived past 3 years of age,
successfully weaned at least one cub, and had known death dates; r2 = 0.57; t24 = -5.61; P <

xii

0.0001). A female’s annual reproductive success was calculated as the number of her weaned
cubs divided by her reproductive lifespan. Both female and male offspring are included ...........53
Figure B.4. Female spotted hyenas that lived longer weaned more cubs than those with
relatively short lifespans (N = 35 females that survived past 3 years of age and with known
death dates; r2 = 0.76; t33 = 10.33, P <0.0001). Data points include five females that did not
give birth to any offspring and four females that failed to wean any offspring. Both female and
male offspring are included
..........................................................................................54
Figure 4.1. Patterns of circulating concentrations of IGF-1 across the lifespan in male (filled
circles) and female (open circles) spotted hyenas. Inset is magnified between 1-4 years of age
to better visualize the dimorphic growth period..........................................................................65
Figure 4.2. Juvenile mass did not predict survival to reproductive maturity in (A) males, but did
so in (B) females. Relative mass was statistically analyzed as a continuous variable (Tables 4.4
& 4.5), but partitioned into tertiles for illustrative purposes only .................................................70
Figure C.1. Plasma IGF-1 concentrations were significantly higher in females while pregnant
(solid circles) than while lactating (open circles). Two-tailed paired t-test: t8 = 4.53, P = 0.002.
Lines represent individual female subjects (n = 9 females). ......................................................81
Figure 5.1. . The number of litters with at least one cub surviving to weaning varied with binned
maternal age (rounded down to the nearest integer value) and maternal rank (n =253 litters born
to 83 unique female hyenas across 27 years). The total number of cubs that survived born to
high-ranking females was 136, with 74 born to mid-ranking females, and 71 born to low-ranking
females. Female rank was analyzed as a continuous variable, but is shown here as a
categorical variable for illustrative purposes only. For example, the sole female in the 1-year
category gave birth when she was 1.88 years ...........................................................................89
Figure 5.2. Lifetime reproductive success (LRS) of female spotted hyenas that had given birth
to at least one litter and that had full reproductive histories with known death dates (high-rank: n
= 17, mid-rank: n = 15, low-rank: n = 14). LRS was measured as the total number of cubs
weaned; both male and female offspring are included. Shading denotes binned female rank
(black = high, gray = mid, white = low), which was statistically analyzed as a continuous variable
but shown here as a categorical variable for illustrative purposes only ......................................91
Figure 5.3. (A) Inter-birth intervals following successful weaning of singleton litters (n=72 litters
born to 39 females; mean inter-birth interval ± s.e. for high-ranking females: 14.88 ± 0.69
months, for mid-ranking females: 18.59 ± 0.62 months, and for low-ranking females: 17.79 ±
0.84 months). (B) Inter-birth intervals following successful weaning of twin litters (n=53 litters
born to 28 females). Shading denotes female rank (black = high, gray = mid, white = low).
Female rank was statistically analyzed as a continuous variable but shown here as a categorical
variable for illustrative purposes only ........................................................................................94
Figure D.1. Female age at the birth date of her last known litter regardless of whether or not it
was successfully weaned. (A) Females still alive by the end of the study period (high-rank: n =
14, mid-rank: n = 8, low-rank: n = 10), with age at the end of the study period on the y-axis and
age at most recent parturition on the x-axis. (B) Deceased females with known death dates that
had given birth to at least one litter (high-rank: n = 34, mid-rank: n = 29, low-rank: n = 30), with
lifespan on the y-axis and age at last recorded parturition on the x-axis. Shading denotes binned

xiii

female rank (black = high, gray = mid, white = low). A point lying on the solid reference line
(intercept = 0, slope = 1) indicates that the female died the same year as she last gave birth.
........................................................................................105

xiv

CHAPTER 1: GENERAL INTRODUCTION

1

General Introduction
Life-history theory seeks to explain the diverse patterns in reproduction and survival
exhibited by organisms. The increased documentation of natural history patterns observed
within and among species has led to the construction of evolutionary theoretical frameworks to
explain common “strategies.” These strategies are often conceptualized along a fast-slow
continuum with “fast-living” strategies marked by faster development, higher initial reproductive
rates, and shorter lifespans, and “slow-living” strategies described by slower development, lower
reproductive rates, and longer lifespans (Western & Ssemakula 1982; Promislow & Harvey
1990; Stearns 1992). Fundamentally, each strategy is governed by the optimization of energy,
with the balance of resources tipped towards maximizing fitness. In species organized by linear
dominance hierarchies, an individual’s social rank largely determines its growth and
reproductive fitness relative to its peers. While high-ranking individuals appear to exhibit a
“faster” pace of life and low-ranking individuals a “slower” pace of life, little research has
explored if and/or how social rank may contribute towards the evolution of life-history strategies
exhibited within a population or species.
In this dissertation, I have attempted to describe the evolutionary and physiological
mechanisms underpinning life history variation in a social mammal, the spotted hyena (Crocuta
crocuta). Within a social group, called a clan, high-ranking females grow faster, wean and start
reproducing at earlier ages, and exhibit higher reproductive rates than do lower-ranking
females. Interestingly, high-ranking females do not appear to suffer from shorter lifespans,
which would be expected by a “fast-living” strategy. My research has focused on understanding
the mediation of this apparent contradiction with the hope that by integrating the seemingly
disparate worlds of evolutionary biology and biogerontology (the study of aging mechanisms),
we may better understand how mechanisms of life histories have evolved across social and
environmental gradients.

2

Summary of Chapter 2
Presumably, faster growth and higher reproductive rates are linked to shorter lifespans
through the erosion of physiological functions from accumulating cellular “wear and tear.” Much
like a bank account with a fixed amount of funds and increasing withdraws, the “disposable
soma” theory predicts that the degradation of physiological functionality (senescence) is a
realized consequence of when an organism’s energetic cache is no longer sustainable
(Kirkwood & Holliday 1979; Kirkwood & Rose 1991). By occupying the top positions in a clan,
high-ranking spotted hyenas enjoy priority of access to key resources and receive far fewer acts
of aggression. My collaborators and I hypothesized that these social benefits may offset any
accumulation of cellular stress typically associated with faster growth and higher reproductive
rates. We also hypothesized that this relationship may vary among clans that experience
different ecological conditions.
We tested these hypotheses by measuring telomere length among adult hyenas
occupying various social rank positions among five different clans in the Masai Mara National
Reserve of southwestern Kenya. Telomeres are highly conserved, noncoding, repetitive DNA
sequences that confer chromosomal stability and buffer against degradation of coding
sequences (Greider & Blackburn 1985; Harley, Futcher & Greider 1990). When telomeres
shorten to a critical length a regulatory cascade is unleashed often culminating in DNA damage
(“wear and tear”) and replicative senescence (Monaghan & Haussmann 2006). Thus, shorter
telomere lengths have been viewed as useful biomarkers of accumulated cellular stress. We
found a strikingly positive linear relationship between telomere length and social dominance
rank, with high-ranking females exhibiting the longest telomeres and low-ranking males
exhibiting the shortest (Lewin et al. 2015). We also found variable telomere lengths in highranking females among the different clans independent of local food abundance, suggesting
that other socioecological variables may be at play (Reznick et al. 2004). Our results confirmed

3

our hypothesis that high-ranking females do not appear to incur large degrees of cellular
damage despite high energetic withdraws made for growth and reproduction.
Summary of Chapter 3
The theory of antagonistic pleiotropy predicts that a shared mechanism may have
opposing effects on different life-history traits, such that a beneficial trait earlier in life may have
an inextricable cost associated with it later in life (Williams 1957; Rose 1985). High-ranking
female hyenas seem to reap many early-life benefits (e.g. fast growth, early age at first
reproduction) without incurring many later-life costs (e.g. short telomere lengths, short lifespans)
so we investigated whether or not they are subject to pleiotropic mechanisms. The insulin-like
growth-factor-1 (IGF-1) is a metabolic hormone that is hypothesized as a pleiotropic agent in
life-history variation due to its well-documented effects in laboratory model organisms (Dantzer
& Swanson 2012) and some wild vertebrates species (Sparkman, Vleck & Bronikowski 2009).
Because IGF-1 confers its first benefit as growth we expected concentrations of IGF-1 during
this period to predict later-life costs.
We found that juvenile concentrations of IGF-1 predicted heavier juvenile mass, which in
turn predicted greater survival to reproductive maturity. However, independent of mass, higher
juvenile concentrations of IGF-1 predicted earlier age at first parturition and reduced longevity in
adulthood. Our results highlight the importance of early postnatal development as a
determination point in mammals, and suggests that concentrations of IGF-1 during this sensitive
period can be used to predict important later-life trade-offs between growth, reproductive fitness,
and lifespan in wild, long-lived animals.
Summary of Chapter 4
In the previous chapter, we found higher concentrations of IGF-1 among heavier juvenile
female hyenas. The strong relationship between mass and IGF-1 led us to investigate the role
of IGF-1 as a potential mediator of female-biased sexual size dimorphism (SSD) among spotted
hyenas. Here, we had three goals. The first was to describe intra- and inter-sexual variation in

4

circulating IGF-1 concentrations across the lifespan. If IGF-1 mediates the emergence of SSD in
spotted hyenas, then we should expect sex differences in circulating concentrations starting
before the dimorphic growth period between 1-4 years of age (Swanson et al. 2013). In many
wild vertebrate species, sex-biased sensitivity to extrinsic variables during postnatal
development is believed to contribute largely to the realized difference, or degree, of SSD
observed in adulthood (Post et al. 1999; Toïgo, Gaillard & Michallet 1999; Badyaev,
Whittingham & Hill 2001; Wilkin & Sheldon 2009). Thus, our second goal was to identify whether
sex-biased sensitivities to various socio-ecological variables affected juvenile IGF-1
concentrations in both sexes. Theoretically, SSD may arise when the selective optimum for
growth or size differs between the sexes (Ralls 1977; Lande 1980; Price 1984). As such, our
third goal was to investigate whether sex-biased fitness effects existed among spotted hyenas.
Our results showed that females displayed higher concentrations of IGF-1 than did males,
starting before and lasting beyond the dimorphic growth period. We also found that
socioecological variables predicted variation in IGF-1 among juvenile males but not among
juvenile females. Lastly, we found that juvenile mass predicted survival to reproductive maturity
among females but not among males, suggesting a selective force by which SSD may be
maintained in this species.
Summary of Chapter 5
In many hierarchical societies, high-ranking individuals enjoy greater reproductive fitness
than low-ranking individuals. Previous work by Holekamp et al. (1996) found that adult, highranking female hyenas also benefit from higher social status as evidenced by more successful
performance on a number of reproductive traits. In this chapter, we sought to describe patterns
of rank-related reproductive performance over the lifespan. We specifically tested two nonmutually exclusive hypotheses used to predict age-related improvement and decline in
reproductive performance using 27 years of individual-based, longitudinal data from the MSU
Mara Hyena Project, started by Drs. Kay Holekamp and Laura Smale. We found that female

5

hyenas exhibit both reproductive improvement and senescence and that the patterns vary with
social rank. Higher-ranking females started reproducing at younger ages and displayed higher
reproductive success than lower-ranking females. However, these females typically did not live
as long as lower-ranking females, suggesting a potential life-history trade-off. Some lowerranking females lived exceptionally long lives, which afforded them a level of reproductive
fitness comparable to that of high-ranking females. Interestingly, alpha females seemed
unaffected by any potential trade-offs. Specifically, two alphas females started reproducing at
young ages with high reproductive success and exhibited longer than average lifespans. The
findings of this chapter offer a unique and rarely documented perspective on how social
dominance rank may influence population dynamics.

6

CHAPTER 2: SOCIOECOLOGICAL VARIABLES PREDICT TELOMERE LENGTH IN WILD
SPOTTED HYENAS
Nora Lewin, Lisa A. Treidel, Kay E. Holekamp, Ned J. Place, and Mark F. Haussmann
Biology Letters, 2015, 11(2), 20140991

7

Abstract
Telomeres are regarded as important biomarkers of ageing and serve as useful tools in
revealing how stress acts at the cellular level. However, the effects of social and ecological
factors on telomere length remain poorly understood, particularly in free-ranging mammals.
Here, we investigated the influences of within-group dominance rank and group membership on
telomere length in wild adult spotted hyenas (Crocuta crocuta). We found large effects of both
factors; high-ranking hyenas exhibited significantly greater mean telomere length than did
subordinate animals, and group membership significantly predicted mean telomere length within
high-ranking females. We further inquired whether prey availability mediates the observed effect
of group membership on telomere length, but this hypothesis was not supported. Interestingly,
adult telomere length was not predicted by age. Our work shows for the first time, to the best of
our knowledge, the effects of social rank on telomere length in a wild mammal and enhances
our understanding of how social and ecological variables may contribute to organismal
senescence.
Introduction
Social and ecological factors can have important effects on animal health, yet the
proximate mechanisms mediating such effects are still being determined. Within the past
decade, telomere length has been identified as a candidate mechanism through which individual
fitness and life expectancy may vary (Monaghan & Haussmann 2006). Telomeres are highly
conserved, noncoding, repetitive DNA sequences that cap the ends of eukaryotic
chromosomes, confer chromosomal stability, and buffer against degradation of coding
sequences following semiconservative DNA replication (Harley et al. 1990). When telomeres
shorten to a critical length, a regulatory cascade is triggered, potentially culminating in DNA
damage and cell death (Harley et al. 1990). Hence, telomeres may figure importantly in
organismal ageing and the mediation of life history variation (Monaghan & Haussmann 2006).

8

Although telomeres tend to shorten with age (Harley et al. 1990; Hall et al. 2004;
Monaghan & Haussmann 2006), the rate at which this happens may be influenced by the quality
of an organism’s environment (Monaghan & Haussmann 2006; Angelier et al. 2013; Mizutani et
al. 2013). In animal societies structured by linear dominance hierarchies, social rank can act as
a filter through which ecological factors affect the individual. An individual’s dominance status
often determines its priority of access to key resources and has important downstream effects
on health (Sapolsky 2005). Previous work in despotic animal societies has shown large effects
of social rank on stress physiology (Cavigelli & Chaudhry 2012), but we know nothing about its
influence on telomere length. Here, we tested predictions of an hypothesis suggesting that
telomere length varies with socioecological conditions experienced by adult free-living spotted
hyenas (Crocuta crocuta). Specifically, we inquired whether leucocyte telomere length varies
with dominance status within or among hyena groups when individuals experience more or less
challenging ecological conditions.
Materials and Methods
Study species
Spotted hyenas are large-bodied, long-lived carnivores that live in groups called clans,
which may contain over 100 members (Holekamp et al. 2012). Spotted hyenas prey mainly on
large and medium-sized ungulates they kill themselves (Kruuk 1972). Clan-mates defend a
common group territory (Kruuk 1972). Clans are structured by rigid linear dominance hierarchies
in which an individual’s rank position determines its priority of access to food, thereby
influencing virtually every aspect of its biology (Holekamp et al. 2012). In this species,
dominance rank is acquired via associative learning at a young age and maintained through
low-intensity, ritualized aggressive behaviors (Smale, Frank & Holekamp 1993). Female spotted
hyenas are the more aggressive sex, and are socially dominant to all adult males not born in the
clan. Females are philopatric, whereas males typically disperse from their natal clans after
puberty, which occurs at around 24 months of age. Males disperse most commonly at 2-5 years

9

of age (Holekamp & Smale 1998; Van Horn, McElhinny & Holekamp 2003). Males may take
several months to become successfully integrated into a new clan; immigrant males are
subordinate to all natal animals and to immigrant males with longer tenure in the new clan
(Holekamp & Smale 1998; East & Hofer 2001). It is rare for males to remain within their natal
clans long after sexual maturity, so we were only able to obtain four telomere samples from
three adult natal males.
Data collection
We measured telomere length in 64 whole blood samples drawn from 54 adult hyenas.
For 9 individuals, we had repeated samples taken at different ages during adulthood (Table
A.1). Ages at sampling ranged from 2.5 to 13 years, based on known birthdates (± 7 days) or
tooth wear data (± 6 months) (Van Horn et al. 2003). Adult hyenas were members of 5 social
groups, called clans, sampled between 1999 and 2013 in the Masai Mara National Reserve,
Kenya (Table A.2). All clans contained multiple matrilines comprised of females and their
descendants, as well as multiple immigrant males. Among both natal females from different
matrilines and male immigrants, within-clan relatedness is effectively zero (Van Horn et al.
2004).
The habitat occupied by all 5 clans was primarily open, rolling grassland with permanent
water and year-round prey availability. Local prey density was calculated as the mean number
of ungulates counted within 100 meters of multiple 4-km transect lines run through each territory
at biweekly intervals for the 2 years prior to collection of each blood sample. Local prey density
was available for only 1 year prior to collection for one high-ranking female from the Fig Tree
clan and two high-ranking females from the Mara River clan. Per capita prey abundance was
calculated as local prey density divided by the mean number of adults present within the
respective clan during the 2 years prior to collection of each blood sample. Adults included all
natal females, males, and immigrant males. Within each clan, dominance rank matrices were
constructed based on outcomes of dyadic agonistic interactions (Smale et al. 1993). We

10

standardized rank to account for variation in clan size; a relative rank of 1 was the highestranking animal in the clan, and the individual with a relative rank of -1 was the lowest-ranking.
High-ranking females were considered those with relative ranks of 1 to 0, and low-ranking
females were considered those with relative ranks of -0.01 to -1 at the time of darting. Hyenas
were immobilized using Telazol (Zoetis, South Bend, IN; 6.5 mg/kg) administered via a
lightweight plastic dart fired from a CO2-powered rifle (Telinject Inc., Saugus, CA). Blood
samples were collected from the jugular vein into EDTA-treated vacutainers and promptly frozen
in liquid nitrogen until they could be shipped on dry ice to Michigan State University where they
were stored at -80°C.
Telomere analysis
Genomic DNA was extracted using DNeasy Blood and Tissue Kits (Qiagen, Valencia,
CA) and kept frozen at -80°C until telomere analysis. Once shipped on dry ice to Bucknell
University, DNA was digested overnight with HinfI and RsaI, and separated using pulsed field
gel electrophoresis (3 V cm-1, 0.5-7.0 s switch times, 14°C) for 21 hours, followed by in-gel
hybridization at 37°C overnight with a radioactive-labelled telomere-specific oligo (CCCTAA)4.
Gels were then placed on a phosphorscreen (Amersham Biosciences, Buckinghamshire, UK),
which was scanned on a Storm 540 Variable Mode Imager (Amersham Biosciences). Position
and strength of the radioactive signal were determined by densitometry using IMAGEQUANT
5.03v and IMAGEJ 1.42q. Mean telomere length was calculated using the formula: L = Σ(ODi x
Li)/ Σ(ODi) where ODi is the densitometry output at position i and Li is the length (in basepairs) of
the DNA at position i. For all samples, DNA integrity was evaluated empirically by a nucleic acid
stain following agarose gel electrophoresis. DNA was deemed suitable for TRF length analysis if
it was characterized as “a single compact crown-shaped band that migrates in parallel with the
other samples on the gel” (Kimura et al. 2010). Two control hyena samples were run twice in
each gel to determine intra- (2.79%) and inter-assay variability (3.16%).

11

Statistics
We used a Shapiro-Wilk normality test to confirm that the data were normally distributed
(W = 0.99, p = 0.79). We then ran a linear mixed effects model (LMM) of restricted maximum
likelihood to assess the effects of dominance rank and age on adult mean telomere length using
the lme4 package (Bates et al. 2014) in R v.3.1.1 (R Development Core Team 2014). We
included random effects of clan membership, sampling year, and ID because we sampled
among 5 different clans, across multiple years, and had repeated measures on 9 hyenas. Clan
membership and ID, but not sampling year, significantly improved the fit of the model based on
likelihood ratio tests (P < 0.05) and were included in the final LMM. We further tested for effects
of clan membership, prey density, and age on telomere length using only samples from highranking females. This allowed us to control for dominance rank while retaining a large enough
sample size for analysis. Neither random effect of ID nor sampling year improved the model so
we employed a multiple linear regression for our analysis. Pairwise comparisons among clans
were calculated by least square means and Satterwaithe’s approximation for degrees of
freedom (West & Welch 2014). To identify whether there was an age effect on telomere length,
we ran a LMM with ID as a random effect for individuals that were sampled repeatedly as
adults. We employed a simple regression model to investigate whether maternal telomere
length predicted telomere length in adult offspring. Here mothers and offspring with multiple
samples were assigned a single mean adult telomere length.
Results
Among all adults, relative dominance rank significantly predicted telomere length (F1, 34.9
= 8.43, P = 0.006; Figure 2.1), but age did not (F1, 53.5 = 0.76, P = 0.39). The random effects for
hyena ID and clan explained 26.8% and 24.6%, of the residual variation, respectively, and
significantly improved the fit of the model (ID: χ2 = 4.83, df =1, P = 0.028; clan: χ2 = 4.24, df = 1,
P = 0.039). Among high-ranking females (n= 30 telomere samples from 26 hyenas), only clan
membership significantly predicted telomere length (Table 2.1; Figure 2.2). For 9 individuals

12

with repeated sampling, the random effect of ID explained 58.7% of the residual variation and
improved the fit of the model (χ2 = 4.71, df =1, P = 0.030), but there was no significant effect of
age on telomere length (F1, 13.7 = 1.57, P = 0.20). We also found no relationship between
telomere lengths in mothers and their adult offspring (R2 = 0.11, P = 0.39; Table A.3).

Table 2.1. (a) Results from multiple regression model predicting telomere length among highranking females only (n = 30 telomere samples from 26 hyenas). The Talek clan is set as the
reference category. Bolded P-values indicate statistically significant differences at an alpha =
0.05.
(a)
fixed effect

estimate

s.e.

t-value

P-value

intercept

13.91

0.67

20.86

<0.001

Fig Tree

0.15

0.88

0.17

0.863

Mara River

-2.27

0.79

-2.87

0.009

Serena N.

-2.36

0.93

-2.55

0.018

Serena S.

-1.77

0.66

-2.68

0.013

prey per hyena

0.26

0.18

1.43

0.167

age

-0.03

0.08

-0.46

0.647

13

16

Mean telomere length (kb)

15

Sex
f
m

14

Clan
Talek
Fig Tree
Mara River
Serena North
Serena South

13

12

11
-1.0

-0.5

0.0

0.5

1.0

Standardized dominance rank
Figure 2.1. Mean telomere length in relation to relative dominance rank in adult hyenas.
Regression line shown is from a linear mixed-effects model with all clans included. A dominance
rank of 1 represents the highest-ranking animal within a clan at the time of sampling, and the
individual with a dominance rank of -1 represents the lowest-ranking. Symbol pattern denotes
the clan to which each individual belonged. Circles: females; triangles: males.

14

16

●
●
●

●

15

Mean telomere length (kb)

●
●

●
●
● ●
●

●
●●
●
●

●
●
●

●
●

●

14

13

●

●

●
●

●
●

12

●
●

11
Serena North

Serena South

Mara River

Fig Tree

Talek

Figure 2.2. Mean telomere length among high-ranking females from 5 clans. Box plots show
group mean (line within box) ± s.d. (box) and range (whiskers).

15

Discussion
In hyena societies, high social rank ensures superior food access at kills; this forces
many low-ranking animals to forage near territory edges or commute great distances in search
of uncontested food (Hofer & East 1993a; Boydston et al. 2005). Low-ranking female and male
hyenas may thus incur high maintenance costs, similar to birds subjected to environmental
challenges (Hall et al. 2004; Angelier et al. 2013; Mizutani et al. 2013). Metabolic stress may
cause high concentrations of reactive oxidative species and stress hormones, which are known
correlates of telomere shortening (Haussmann & Marchetto 2010). In spotted hyenas in
Tanzania, dominance rank predicted circulating glucocorticoid concentrations among nonlactating females, with lower concentrations found in high- than low-ranking females (Goymann
et al. 2001). Our finding that high-ranking hyenas display longer telomeres than low-ranking
hyenas is consistent with earlier work, and further suggests that social subordination has
important consequences at the cellular level.
Age was not a significant predictor of telomere length in any of our analyses and
appears not to affect telomere length strongly among adult hyenas. Although counterintuitive,
this is consistent with findings in long-lived birds, where the rate of telomere shortening is
notably slower during adulthood than early development (Hall et al. 2004; Monaghan &
Haussmann 2006). Hyenas from high-ranking matrilines might be born with longer telomeres
than those from low-ranking matrilines, but we could not test this, nor did we find any evidence
that maternal telomere length predicts offspring telomere length (Table A.3). Increasing
evidence shows that pre- and post-natal stress can have long-lasting effects on adult telomere
length (Hall et al. 2004; Haussmann et al. 2012; Herborn et al. 2014). Spotted hyenas
experience the highest mortality rates during the period immediately after weaning (Watts et al.
2009), and this may coincide with a high rate of telomere shortening. We speculate that the
rank-related variation observed in adult telomere length may have early developmental origins.

16

As a random effect, clan membership explained a large proportion of the residual
variance in telomere length among adult hyenas. We investigated the clan effect further by
modeling it as a fixed effect where it significantly predicted telomere length among high-ranking
females. Within-clan relatedness among spotted hyenas averages around zero (Van Horn et al.
2003), allowing us to rule out clan-specific genetic effects. Therefore, we tested the hypothesis
that telomere length might be influenced by local prey availability, which varies among clan
territories (Table 2.1). However, this was not a significant predictor within high-ranking females
(Table 2.1), implying that other factors that vary among clans are likely in play. Our analysis also
revealed pronounced individual variation in telomere length even though sampled females all
occupied top positions in their respective dominance hierarchies. Thus, there appears to be
considerable inter inter- and intraclan variation in the relationship between dominance rank and
telomere length. This is the first study to investigate the influence of social dominance on
telomere length in a non-human species, and we encourage future experimental testing of this
relationship in other species.

17

APPENDIX

18

Supplementary Material for Chapter 2
Table A.1. Information on repeated samples from the same individuals.

1.
2.
3.
4.
5.
6.
7.
8.
9.

Hyena ID

Clan

Age1 (yr)

ali
eco
guci
mrph
bail
hex
nav
ldv
oak

Talek
Mara River
Talek
Talek
Talek
Talek
Talek
Talek
Talek

3.93
5.02
3.43
2.72
7.91
4.29
6.45
3.50
2.52

Age2 (yr)
9.24
5.32
6.51
3.72
11.28
7.42
9.71
7.57
7.84

Age3 (yr)
13.04
-

Table A.2. Information on the demography and territories of sampled clans.
Talek

Serena
North

Mara River

Serena
South

Fig Tree

years sampled

1999-2013

2012-2013

2002-2009

2012-2013

2006-2009

mean territory
size

57.3 km2

50 km2

31 km2

28 km2

52.91 km2

mean prey
density per
adult hyena

2.46
animals/
hyena

6.48
animals/
hyena

5.22
animals/
hyena

4.23
animals/
hyena

4.88
animals/
hyena

Table A.3. Results from a simple linear regression of offspring adult telomere length on
maternal telomere length (df = 7).

intercept
maternal telomere length

estimate
11.71
0.17

s.e.
2.82
0.19

19

t-value
4.16
0.93

p-value
0.004
0.39

R2
0.11

CHAPTER 3: JUVENILE CONCENTRATIONS OF IGF-1 PREDICT LIFE-HISTORY TRADEOFFS IN A WILD MAMMAL
Nora Lewin, Eli M. Swanson, Barry L. Williams, and Kay E. Holekamp
Functional Ecology, 2016, 31(4), 894-902.

20

Introduction
Life history theory posits that realized phenotypic outcomes represent compromises
among fitness costs and benefits (Stearns 1992). Trade-offs result when an increase in fitness
due to one trait occurs concomitantly with negative changes in fitness associated with another
trait (Stearns 1989, 1992; Promislow & Harvey 1990; Kirkwood & Rose 1991; Zera & Harshman
2001). Selection can act at different times during the lifespan so fitness costs of a trait need not
accrue at the same time as its benefits, resulting in age-structured trade-offs. Such trade-offs
are hypothesized to contribute importantly to the diversity of observed life histories (Promislow &
Harvey 1990; Kirkwood & Rose 1991; Stearns 1992; Metcalfe & Monaghan 2003; Roff &
Fairbairn 2007). Despite great variation in life histories, some trade-offs appear to occur very
frequently in nature and across taxa (Western & Ssemakula 1982; Stearns 1983; Promislow &
Harvey 1990; Bielby et al. 2007; Swanson & Dantzer 2014; Lemaître et al. 2015). For instance,
life-history strategies are often conceptualized along a fast-slow continuum. Initial costs and
delayed benefits (e.g. slow development, low reproductive rate, and long lifespan) characterize
a “slow” pace of life, whereas the converse, initial benefits and delayed costs, characterizes a
fast pace of life. In fluctuating or competitive environments where no single life history is
optimal, both fast and slow paces may occur within a population (Descamps et al. 2006; Nussey
et al. 2008b; Sparkman et al. 2009; Schuett et al. 2015; Huchard et al. 2016). However, it
remains largely unclear how and when an organism’s life history is determined.
In the past two decades, there has been considerable interest in identifying the
physiological mechanisms mediating life-history variation (Zera & Harshman 2001; Ricklefs &
Wikelski 2002; Flatt & Heyland 2011; Dantzer & Swanson 2012; Swanson & Dantzer 2014).
Endocrine pathways are hypothesized to facilitate life-history evolution through their ability to
affect multiple traits concurrently while also remaining relatively plastic in their responses to
immediate environmental cues, such as food abundance or social pressures (Ketterson & Nolan
Jr 1992; Dufty, Clobert & Møller 2002; Ricklefs & Wikelski 2002; Flatt & Heyland 2011).

21

Regardless of their intrinsic trade-offs, all life histories theoretically optimize the allocation of
limited energy resources. Therefore metabolic hormones may play significant intermediary roles
by signaling appropriate distribution of energy in response to current environmental conditions.
This mechanism should theoretically operate during a sensitive period in early in life, when the
organism passes through a phase of development critical to the determination of resultant life
histories (Ketterson & Nolan Jr 1992; Lindstrom 1999; English et al. 2016). Thus, early hormone
concentrations may serve as important agents of evolution by organizing the development of a
life history best suited for the predicted adult environment.
The polypeptide hormone, insulin-like growth factor-1 (IGF-1), has been proposed as
one candidate mechanism for shaping life-history traits due to its critical importance during
development and its pleiotropic effects throughout the lifespan (Tatar, Bartke & Antebi 2003;
Kenyon 2005; Dantzer & Swanson 2012; Bartke, Sun & Longo 2013; Swanson & Dantzer
2014). IGF-1 and its associated pathways are well characterized in both vertebrate and
invertebrate taxa. In vertebrates, the liver primarily produces IGF-1, but almost all tissues
synthesize IGF-1 as well, enabling IGF-1 to have both endocrine and paracrine effects (Bartke
et al. 2013). Work with laboratory animal models has shown that high circulating IGF-1
concentrations are largely beneficial during postnatal development, and closely associated with
neurogenesis, somatic growth, reproduction, and wound-healing (Kenyon 2005; Yuan et al.
2012; Bartke et al. 2013). However, high concentrations during early postnatal development are
also linked with reduced later-life survival relative to animals with lower circulating IGF-1 early in
life, suggesting both the possibility of a trade-off and the importance of this particular period of
development (Sonntag et al. 1999; Yuan et al. 2009, 2012; Panici et al. 2010; Svensson et al.
2011; Junnila et al. 2013).
IGF-1 has been hypothesized to mediate variation in life-history partitioning among wild
mammalian species (Swanson & Dantzer 2014), but its role in life-history variation within such
species remains largely unknown (Dantzer & Swanson 2012). Work with free-living populations

22

should allow us to elucidate the fitness consequences of naturally occurring variation in early
IGF-1 concentrations, enhance our ability to identify potential targets of selection, and permit
identification of sensitive windows during which endocrine markers may predict later-life traits.
Here, we tested predictions of the hypothesis that juvenile IGF-1 concentrations mediate lifehistory trade-offs by using 26 years of detailed life-history data from wild female spotted hyenas
(Crocuta crocuta) in Kenya. If IGF-1 mediates life-history trade-offs within species, then we
would expect high juvenile IGF-1 concentrations to correlate positively with early benefits such
as large juvenile body size, early age at first parturition, and enhanced survival to reproductive
maturity, but also with the later-life cost of reduced lifespan after reproductive maturity (Dantzer
& Swanson 2012). IGF-1 might influence life histories either directly or indirectly via intermediate
effects on juvenile body size (Metcalfe & Monaghan 2003; Dmitriew 2011; Swanson & Dantzer
2014), so we also explored the independent effect of juvenile size on life-history traits. Finally,
we explored the fitness consequences of contrasting life-history trajectories, focusing on annual
reproductive success and the total number of cubs weaned by each female during her lifespan.
If high juvenile concentrations of IGF-1 are consistent with a faster pace of life, then we would
expect individuals exhibiting high concentrations early in life to have high annual reproductive
success, but lower total reproductive success due to shortened lifespan. We investigated both
direct and indirect influences of IGF-1 on reproductive fitness traits (Metcalfe & Monaghan 2003;
Dmitriew 2011; Swanson & Dantzer 2014). Here indirect effects would be those mediated by
IGF-1 effects on body size, and direct effects would be IGF-1 effects that occur independently of
body size.
Materials and Methods
We measured plasma IGF-1 concentrations at 5.5-12.1 months of age in 70 juvenile
female hyenas, 57 of which were followed until death, and inquired whether these juvenile IGF-1
concentrations predicted later trade-offs in the life histories of sampled individuals. We used 5.5
months as the lower limit of age range because this is the youngest age at which we can safely

23

immobilize juvenile spotted hyenas. We sought to limit variation in juvenile food acquisition by
only looking at the period of maternal provisioning via lactation, so we used an upper age limit of
12.1 months, which is just before the average age of weaning in our study population (13.5
months, Holekamp et al. 1996). Social dominance plays a critical role in hyena society, and it
can have profound, long-lasting direct and indirect effects on offspring fitness (Holekamp et al.
1996; Hofer & East 2003; Watts et al. 2009; Swanson, Dworkin & Holekamp 2011). Therefore,
we included maternal dominance rank as a covariate in all our analyses. Furthermore, available
food resources can directly influence the timing of first parturition in this species (Holekamp et
al. 1996) so we also included local prey density as a covariate in our statistical models.
Relevant biology of spotted hyenas
Spotted hyenas are large-bodied carnivores with protracted early development; they do
not reach full size until 30-36 months of age (Swanson et al. 2013). They live in groups, called
clans, structured by linear dominance hierarchies (Kruuk 1972; Tilson & Hamilton 1984; Frank
1986). An individual’s social dominance rank determines its priority of access to food at kills
(Kruuk 1972; Tilson & Hamilton 1984; Frank 1986), and has profound fitness consequences
(Holekamp et al. 1996; Hofer & East 2003; Watts et al. 2009). Here dominance rank matrices
were constructed based on outcomes of dyadic agonistic interactions (Tilson & Hamilton 1984;
Frank 1986). We standardized dominance rank to account for variation in clan size at time of
immobilization; the highest-ranking animal in the clan has a relative rank of 1, while the lowestranking individual has a relative rank of -1. Because young hyenas learn their dominance ranks
from their mothers and other group mates, but do not assume their proper positions in the clan’s
hierarchy until they are approximately 18 months of age (Holekamp & Smale 1993; Smale et al.
1993; Engh et al. 2000), we assigned each juvenile hyena its mother’s dominance rank at the
time of immobilization.
Spotted hyenas typically give birth to small litters containing only one or two cubs, which
they nurse for up to 24 months. Female hyenas are philopatric and remain in the natal clan

24

throughout their lives, whereas virtually all males disperse after puberty (Smale, Nunes &
Holekamp 1997; Höner et al. 2010; Davidian et al. 2016). Age at first parturition in our study
population is generally between three and five years of age (Holekamp et al. 1996) so we used
three years as the age of reproductive maturity in our survival analyses. Female spotted hyenas
can live at least 24 years in the wild; our oldest female currently alive was born on October 7,
1992. For determination of life history-traits in female spotted hyenas, see electronic
supplementary material.
Study site, data collection, and determination of life-history and fitness traits
Our subjects were part of a long-term field study of spotted hyenas within the Masai
Mara National Reserve, Kenya. Behavioral data, demographic data, body size data and blood
samples were collected between May 1988 and January 2015 from members of the Talek clan.
Three samples with their associated data came from the Mara River clan, which has been
observed since 2001. Ages of hyenas in their natal clans are known to ± 7 days based on
methods described earlier (Holekamp et al. 1996). Juveniles in this study ranged from 5.5 to
12.1 mo. old (mean = 9.3 mo.) at the time of immobilization and blood sample collection.
Hyenas younger than 5.5 months old could not be safely immobilized. As part of the ongoing
study, juvenile hyenas are routinely anaesthetized with Telazol (Zoetis, South Bend, IN; 6.5
mg/kg) administered in a plastic dart fired from a CO2-powered rifle (Telinject Inc., Saugus, CA).
All immobilizations were conducted early in the morning, and all were carried out according to
guidelines specified by the American Society of Mammalogists and approved by the Institutional
Animal Care and Use Committee at Michigan State University (MSU; AUF #05/14-087-00).
Following Telazol administration, hyenas were weighed and measured for four cranial and nine
post-cranial linear morphological measurements (Swanson et al. 2011) (see also Figure B.1).
Blood samples were taken from the jugular vein in heparinized vacutainer tubes, centrifuged,
and plasma was drawn off, aliquoted, then promptly frozen in liquid nitrogen until it could be
shipped on dry ice to MSU, where samples were stored at -80ºC. The 5.5-12.1 mo. age interval

25

during which we sampled young female hyenas represents a phase of rapid growth for both size
traits measured here, mass and shoulder height. However, the ages differ at which these traits
reach maturity, defined as the age in months at which the predicted size for the trait equals 95%
of the asymptotic value, or ‘adult size.’ Shoulder height reaches maturity at 21.03 mo. of age,
whereas mass reaches maturity at 45.28 mo. of age (Swanson et al. 2013).
We calculated average local prey density as the mean number of ungulates counted
within 100 meters of multiple 4-km transect lines run through each clan’s territory at biweekly
intervals during the six months prior to collection of each blood sample (Tables B.2-B.3) or
during the six months prior to the birth of a female’s first litter (Table B.4).
We determined age at first parturition was based on the initial tearing of the female’s
pseudopenis, indicated by the presentation of pink scar tissue on its posterior surface (Frank &
Glickman 1994). This tearing reliably indicates that first parturition has occurred regardless of
whether or not the first litter born is ever seen by observers when it appears above ground.
Female hyenas breed throughout their lives so we measured reproductive success only
for females who survived past three years of age and had known death dates. We used the
number of cubs a female weaned divided by the number of years between her third birthday and
the birth of her last litter as a female’s average annual reproductive success. We used the total
number of cubs a female weaned during her lifetime as her overall reproductive success.
Because female spotted hyenas are philopatric, and because we were observing our
study clans daily, we recorded each female’s date last seen as her date of death. With six
exceptions, we included samples from females disappearing before January 2014 to allow
ample opportunity for reappearance. The six exceptions were females still alive at the end of the
study period.
Of the 35 females that survived to or past three years of age and had known death
dates, five females failed to bear any offspring and four females failed to raise offspring to
weaning so were not included in the analysis of annual reproductive success (Table B.7; Figure

26

B.3). All 35 females were included in the analysis of number of cubs weaned (Table B.8; Figure
B.4), including the five females that did not give birth to any offspring and four other females that
failed to wean any offspring. We performed model comparison using Akaike’s Information
Criteria corrected for small sample sizes (AICc) (Burnham & Anderson 2002). Models are
comparable if ∆AICc<2.0.
Measuring IGF-1 concentrations in plasma
We measured IGF-1 concentrations in plasma collected from hyenas immobilized
between 1990 and 2014. To quantify plasma IGF-1 levels, we used a commercially available
enzyme immunoassay kit (EIA; ALPCO Diagnostics, Salem, NH). We validated this assay for
spotted hyenas using the ‘spike and recovery’ method following kit protocol. Briefly, 25μl of kit
sample buffer (0 ng/ml IGF-1) were added to 25μl plasma aliquots from three different
individuals. The measured IGF-1 concentration served as the baseline concentration. Separate
25μl plasma aliquots from the same three hyena plasma samples were then spiked with 25μl of
50 ng/ml IGF-1 (kit standard E) and measured for IGF-1 concentrations. The measured IGF-1
concentration served as the observed value. The expected value for each hyena sample was
calculated as the sample’s baseline concentration plus the spiked concentration of 50 ng/ml.
Percent recovery was calculated as the [observed value/expected value x 100] (Table B.1). Our
average percent recovery of 116.1% fell within the kit’s acceptable range of 70-130%. Hyena
plasma samples were diluted to 1:20 so measured concentrations fell within the kit standard
curve (Figure B.2). All samples were run in duplicate and randomly assigned to plates (n = 11
plates). Mean intra- and inter-assay coefficients of variation for the IGF-1 immunoassays were
3.06% and 10.5%, respectively.
Plasma samples were kept frozen at -80°C, but the length of time frozen varied between
3 months and 24 years. To assess whether freezer storage time affected IGF-1 concentrations,
we used a paired t-test to compare variation between randomly chosen subsets of plasma
samples frozen before and after 1999 (n = 30, each). We found no significant difference

27

between the two subsets (t58 = -0.72, P = 0.47), suggesting that storage time was not a
confounding variable. Additionally, a simple linear regression revealed no influence of freezer
storage time on IGF-1 concentrations (r2 = 0.007, t68 = 0.69, P =0.49).
Statistical analyses
Due to the opportunistic nature of our sampling methods, we could only sample juveniles
once so there were no repeated measures in any of our analyses. We first used multiple linear
regression to explore the independent contributions of maternal dominance rank, age measured
in months, and local prey density as predictors of juvenile concentrations of IGF-1. We
calculated age (in months) using known birthdates, and we calculated prey density as the
average local prey density during the 6 months prior to immobilization. Residuals of this
regression model were normally distributed (Shapiro-Wilk W = 0.98, P = 0.32).
We then used multiple linear regression to assess the influences of juvenile IGF-1,
maternal dominance rank at the time of juvenile immobilization, and local prey density on
relative juvenile size at the time of sampling (Table B.3). We included all three predictor
variables in a single model because there was no evidence of multi-collinearity among them,
and we believed them all to be biologically relevant to predicting juvenile size. We used
residuals from two measures of relative juvenile body size at age of blood sampling: body mass
and shoulder height (shown as measure “c” in Figure B.1). An individual that was ‘large for its
age’ had a positive deviation (residual), and an individual ‘small for its age’ had a negative
deviation (Altmann & Alberts 2005). One individual had neither mass nor shoulder height
recorded so was not included in models with size; however, this individual did have an IGF-1
measurement, so was included in all other relevant models. Two other individuals had no
shoulder height recorded but did have IGF-1 measurements so they were included in relevant
models excluding size. Residuals for all the models compared were normally distributed
(Shapiro-Wilk W ≥ 0.96, P ≥ 0.16).

28

To assess how age at first parturition varied among females, we used multiple linear
regression with juvenile concentrations of IGF-1, body mass residuals, maternal dominance
rank, and local prey density as continuous predictor variables. Maternal dominance rank
equaled the rank of the subject’s mother at the time of its immobilization as a juvenile, and local
prey density equaled the density during the 6 months prior to the date of a female’s first
parturition. We included these two variables in both models. We were interested in comparing
direct and indirect effects of IGF-1 on age at first parturition, so we included IGF-1 and body
mass residuals in separate models. We assessed relative model fit using Akaike’s Information
Criteria corrected for small sample sizes (AICc). Models are comparable if ∆AICc<2.0. For
model comparison, we included only individuals with both IGF-1 and body mass. Model
residuals did not differ significantly from a normal distribution (Shapiro-Wilk W ≥ 0.95, P ≥ 0.18).
To analyze the contributions of IGF-1 to survival to and beyond reproductive maturity,
we performed Cox proportional-hazards regression analyses using the ‘coxph’ function in the
‘survival’ package in R (Therneau 2014). We opted to use proportional hazards to model
survival because we did not expect IGF-1 concentrations to have a linear relationship with
lifespan. A negative hazard rate coefficient (β) indicates a decreased probability of survival with
increasing value of the predictor variable; a positive hazard rate coefficient (β) indicates an
increased probability of survival with increasing value of the predictor variable. A hazard ratio
(exp(β)) equal to 2, for example, indicates that individuals are twice as likely to survive with
increasing values of the predictor variable. We included six individuals still alive in January 2014
as right-censored data points. Because we were specifically interested in both direct and
indirect (mediated via body size) influences of IGF-1 on survival, we limited our comparison to
that between two models: one with IGF-1 and one with body mass. We included maternal
dominance rank as a covariate in both analyses due to previous findings on survival in this
population (Watts et al. 2009). Visual inspection of scaled Schoenfeld residuals showed no
violation of the assumption of proportional hazards, influential observations, or nonlinearity.

29

Although both IGF-1 and body size measures were modeled as continuous variables in all
statistical analyses, we divided these predictors into tertiles for illustrative purposes in some of
our figures.
We used multiple linear regression to investigate two different aspects of reproductive
fitness: annual reproductive success and lifetime reproductive success. To predict average
annual reproductive success, we compared three multiple linear regression models with
maternal dominance rank and either IGF-1, age at first parturition or juvenile body mass as a
covariate. In these models, we only included females with IGF-1 values, juvenile body mass
measurements, and at least one litter (n = 29). To predict lifetime reproductive success, we
compared four multiple linear regression models with maternal dominance rank and either IGF1, juvenile body mass, age at first parturition, or lifespan as a covariate. We assessed relative
model fit using Akaike’s Information Criteria corrected for small sample sizes (AICc). Models are
comparable if ∆AICc<2.0.
For all model analyses, we centered all continuous predictor variables with a mean of 0
and standard deviation of 1. Model residuals did not differ significantly from a normal distribution
(Shapiro-Wilk W ≥ 0.94, P ≥ 0.08), and there was no evidence of multicollinearity (square root of
variance inflation factor < 2). For models of lifetime reproductive success, we performed a
square-root transformation of the dependent variable (total number of cubs weaned) in order to
meet assumptions of normality (value of 0.5 added to counts of 0 cubs weaned). Following the
transformation, model residuals did not differ significantly from a normal distribution (ShapiroWilk W ≥ 0.95, P ≥ 0.18).
All statistical analyses were performed in R v. 3.2.3 (R Development Core Team 2014).

30

Results
Predictors of juvenile IGF-1 concentrations
None of our model variables significantly predicted concentrations of IGF-1 in juvenile females
sampled between 5.5 and 12.1 months of age (maternal dominance rank: t65 = 0.28, P = 0.782;
age: t65 = -0.29, P = 0.770; prey density: t65 = -0.30, P = 0.768; Table B.2).
Juvenile body size
Female hyenas with higher juvenile IGF-1 concentrations were heavier, but not taller, for
their age than females with lower juvenile concentrations (mass: t64 = 3.80, P = 0.0003; Figure
3.1A; shoulder height: t62 = 1.28, P = 0.21; Table B.3). Juveniles born to high-ranking mothers
were both heavier and taller than those born to low-ranking mothers (mass: t64 = 2.86, P =
0.006; shoulder height: t62 = 2.06, P = 0.04; Table B.3). Average local prey density during the 6
months before blood collection did not influence either measure of juvenile size (Table B.3).
Age at first parturition
Among wild female hyenas, we found that higher juvenile IGF-1 concentrations predicted
earlier age at first parturition (t34 = -2.15; P = 0.039; Figure 3.1B) after accounting for maternal
rank and average local prey density (Table B.4). This model fit the data better than a model
containing mass instead of IGF-1 (Table B.4a); mass did not significantly predict age at first
parturition (Table 3.4b).
Survival to reproductive maturity
Females that were heavier as juveniles survived proportionally better than did lighter
females (z = -2.48, P = 0.013; Figure 3.2; Table B.5). Maternal rank did not significantly predict
survival to reproductive maturity. This model fit the data better than a model containing IGF-1
instead of mass (Table B.5).
Survival after reproductive maturity
Female hyenas with higher juvenile IGF-1 concentrations had shorter life expectancies
after reaching reproductive maturity than did females with lower IGF-1 concentrations in early

31

life (z = -1.28, P = 0.023; Figure 3.3; Table B.6). Daughters of high-ranking mothers also tended
to survive better than did daughters of low-ranking mothers (z = -1.95, P = 0.052; Table B.6).
Juvenile mass did not predict survival in adulthood and was not included in the best-supported
model.
Fitness measures
Females with younger ages at first parturition experienced higher annual reproductive
success than did females that started breeding when they were older (t25 = -2.82, P = 0.009;
Tables B.7a & B.7b; Figure B.3). Longer life expectancy after reaching reproductive maturity
and higher maternal dominance rank significantly predicted larger numbers of weaned cubs (t26
= 7.55, P <0.0001; Tables B.8a & B.8b; Figure B.4).

32

Figure 3.1. Female hyenas exhibiting higher concentrations of IGF-1 as juveniles were (A)
heavier as juveniles and (B) gave birth to their first litter at younger ages. Females with
higher juvenile plasma IGF-1 concentrations were (A) relatively heavy for their age (t64 = 3.80, P
= 0.0006) and (B) gave birth to their first litters at younger ages (t34 = -2.15; P = 0.039) than did
those with lower juvenile plasma IGF-1 concentrations. Regression lines are from the linear
regression models.

33

1.0
0.8
0.6
0.4

Survivorship

0.2

Relative juvenile mass

0.0

Large (n = 16)
Medium (n = 37)
Small (n = 8)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Age (yr)

Figure 3.2. Survival curves to reproductive maturity for juvenile females of varying
relative mass. Female hyenas that were relatively heavy as juveniles survived to reproductive
maturity at higher rates than those that were relatively light as juveniles (likelihood ratio test: n=
61, χ2 = 6.17, df = 1, P = 0.01). Right-censored tick marks (+) indicate females that survived
past 3 years of age (n = 39). Relative mass was statistically analyzed as a continuous variable
(Table B.5), but partitioned into tertiles for illustrative purposes only.

34

1.0

Juvenile IGF-1

0.6
0.4
0.0

0.2

Survivorship

0.8

High (n = 16)
Medium (n = 15)
Low (n = 8)

0

5

10

15

20

Age (yr)

Figure 3.3. Survival curves after reproductive maturity for females exhibiting varying
concentrations of IGF-1 as juveniles. Female hyenas exhibiting higher concentrations of IGF1 as juveniles experienced significantly lower rates of survival after reproductive maturity than
did hyenas exhibiting lower juvenile IGF-1 concentrations (likelihood ratio test: n= 39, χ2 = 7.88,
df = 2, P = 0.02). Tick marks (+) indicate females alive at the end of the study period (n = 6).
IGF-1 was statistically analyzed as a continuous variable (Table B.6), but classified into tertiles
for illustrative purposes only.

Discussion
IGF-1 concentrations decrease with age over the lifespan in spotted hyenas
(unpublished data) as they do in other species (Bartke et al. 2013), but age did not influence

35

concentrations among our study subjects between 5.5 and 12.1 months of age (t65 = -0.29, P =
0.77; Table B.2). Social dominance rank has been shown to predict IGF-1 concentrations in
adult baboons (Sapolsky & Spencer 1997), but the juvenile hyenas sampled here were still in
the early process of learning their social ranks. We found no influence of maternal dominance
rank on juvenile IGF-1 concentrations (t65 = 0.28, P = 0.78; Table B.2), although maternal
dominance rank did positively influence juvenile size (Figure 3.1A; Table B.3). This suggests
that social rank acts independently of IGF-1 in mediating size; this has also been found in
juvenile mandrills (Bernstein et al. 2012). Lastly, there appeared to be no effect of prey density
on IGF-1 concentrations (t65 = -0.30, P = 0.77; Table B.2).
Postnatal body size is a crucial life-history trait that figures prominently in hallmark
examples of life-history trade-offs (Metcalfe & Monaghan 2003; Dmitriew 2011). Large body size
appears to be one of the first traits that young animals can exploit to optimize their interactions
with their environment and thus survive to reproductive maturity. If IGF-1 indeed mediates the
life-history trade-offs associated with the pace of life continuum, then it should have important
effects on growth (Miller et al. 2002; Dantzer & Swanson 2012). Hyenas undergo much rapid
growth before one year of age (Swanson et al. 2013), and our results suggest that IGF-1
concentrations may be promoting variable gains in body size during this period (Figure 3.1A;
Table B.3). Local autocrine or paracrine secretion of IGF-1 and other components of the
somatotropic axis, such as growth hormone and various IGF-1 binding proteins, are also likely
to affect growth (Hossner, McCusker & Dodson 1997; Yakar et al. 1999; Kappeler et al. 2009),
but these remain unexamined in spotted hyenas.
In laboratory and domesticated mammals, IGF-1 helps stimulate gonadal hormone
production and thus plays a role in mediating the timing of reproductive maturity (Hiney et al.
1996; Taylor et al. 2008; Sparkman, Byars & Ford 2010; Yuan et al. 2012; Bartke et al. 2013).
Here we found that higher juvenile concentrations of IGF-1 predicted earlier ages at first
parturition among female hyenas (Figure 3.1B, Table B.4b). IGF-1 predicted age at first

36

parturition better than did relative juvenile mass, suggesting that reproduction may be more
tightly regulated by internal metabolic signals than by overall body size. Female spotted hyenas
of all social ranks breed year-round, and do not monopolize mating opportunities, so there is no
competitive advantage to breeding at a younger age. Nevertheless, females that give birth to
their first litter earlier in life appear to benefit from higher annual reproductive success than
females that start breeding later in life (Figure B.3). This may give the former a competitive
advantage, and is also characteristic of a ‘fast’ pace of life (Promislow & Harvey 1990; Stearns
1992; Descamps et al. 2006).
Given the importance of reproductive success to fitness, theory predicts that survival to
reproductive maturity should be under strong selection, with selection for survival weakening as
individuals age after reproductive maturity (Williams 1957; Hamilton 1966). Consistent with
predictions of our hypothesis, female hyenas with high juvenile IGF-1 appeared to experience a
trade-off later in life, and had shorter life expectancies after reproductive maturity than did
females with low juvenile IGF-1 values (Figure 3.3, Table B.6). Juvenile mass did not predict
survival in adulthood, suggesting that IGF-1 might be operating independently of mass in
relation to this life-history trait. Spotted hyenas breed throughout their lives, so reduced survival
after puberty was associated with fewer reproductive opportunities and therefore lower lifetime
reproductive success ( Figure B.1, Table B.8). However, it is possible that those females
expressing higher juvenile IGF-1 concentrations, and having earlier ages at first parturition,
offset their shorter lifespans by increasing annual reproductive success (Figure B.3, Table
B.8a).
The pleiotropic effects of IGF-1 appeared to restrict hyenas’ ability to enjoy high survival
both before and after reproductive maturity, which is consistent with the idea that senescence
begins at the onset of reproductive maturity (Williams 1957; Hamilton 1966). Despite living in a
variable environment with fluctuating food abundance and natural predators, our subjects
experienced survivorship rates similar to those seen in laboratory rodents (Sonntag et al. 1999;

37

Panici et al. 2010; Svensson et al. 2011; Yuan et al. 2012; Junnila et al. 2013; Bartke et al.
2013) and expected by theory (Williams 1957; Stearns 1992). The accumulation of cellular
damage has been proposed as one cause of death in animals (Flatt & Heyland 2011; Dantzer &
Swanson 2012), but lions (Panthera leo) and humans represent the largest sources of mortality
for spotted hyenas (Kruuk 1972; Watts & Holekamp 2009), so most die violent deaths. To
explain the relatively poor survival after reproductive maturity of hyenas with high juvenile IGF-1,
we hypothesize that individuals with higher juvenile IGF-1 concentrations might engage in more
risky behaviors, which are characteristic of a high energy metabolism (Svensson et al. 2005;
Stamps 2007; Biro & Stamps 2010; Baldini et al. 2013). However, this idea remains to be
assessed.
We were unable to document within-subject variation in IGF-1 concentrations among our
subjects because we could not repeatedly immobilize young hyenas. Although it would have
been interesting to determine how stable IGF-1 concentrations are in these animals, we find it
encouraging that clear patterns of significance emerged with only one sample per
individual. Furthermore, in contrast to other hormones, which may be highly labile, relative
circulating IGF-1 concentrations vary consistently over time and show high intra-individual
repeatability in other species (Roberts et al. 1990; Obese et al. 2008; Yuan et al. 2009;
Kappeler et al. 2009). Lastly, within-individual IGF-1 concentrations at 6, 12, and 18 months all
correlated with survival in developing mice (only the correlation at 6 months was significant;
Yuan et al. 2009), so our results are not without precedent.
The initial benefits and delayed costs associated with high juvenile concentrations of
IGF-1 in spotted hyenas are consistent with life-history theory, suggesting that IGF-1 may
mediate life history trade-offs within, as well as between, species in the wild. Our results support
the hypothesis that IGF-1 functions to optimize life history variation, and suggest that circulating
IGF-1 concentrations measured during early postnatal development in wild mammals can be
used to predict important adult traits not expressed until many years later. Our work elucidates

38

the fitness correlates of naturally occurring variation in early IGF-1 concentrations, and thus
sheds considerable new light on mechanisms mediating life-history trade-offs in a fashion that
would be impossible in controlled laboratory settings.

39

APPENDIX

40

Supplementary Material for Chapter 3
Table B.1. A commercially available enzyme immunoassay kit was validated for spotted hyenas
using the ‘spike and recovery’ method following kit protocol.

Baseline
229.44
179.57
267.91

Sample 1
Sample 2
Sample 3

Recovered
301.72
252.24
311.74

Expected
254.44
204.57
292.91

% Recovery
118.58%
123.30
106.43

Results from the spike and recovery method for assay validation. Average percent recovery was
116.1% ((recovered value/expected value) x 100), with 70-130% as the acceptable range per kit
instructions.

Table B.2. Concentrations of juvenile IGF-1 concentrations in sampled female hyenas were not
significantly influenced by maternal dominance rank, age at sampling, or prey density.
Significant effects are highlighted in boldface at a significance level of α = 0.05.

Trait
Juvenile
IGF-1

Fixed effect
Intercept
Mat. dom. rank
Age
Prey density

Parameter ± SE
305.31 ± 12.22
6.49 ± 23.35
-2.33 ± 7.92
-0.06 ± 0.19

95% CI
280.90 − 329.71
-40.15 − 53.13
-18.15 − 13.49
-0.43 − 0.322

t
24.24
0.28
-0.29
-0.30

df
65
65
65
65

P
<0.0001
0.782
0.770
0.768

Results from a multiple linear regression model predicting concentrations of IGF-1 among 69
juvenile female spotted hyenas sampled between 5.5 and 12.1 months of age. All predictor
variables were centered with a mean of 0 and standard deviation of 1. Hyenas were assigned
their mother’s dominance rank (Mat. dom. rank) at the time of immobilization; a relative rank of 1
was the highest-ranking animal in the clan, and the individual with a relative rank of -1 was the
lowest-ranking. Age at sampling was calculated in months based on known birthdates. Prey

41

density was calculated as the average local prey density during the 6 months prior to
immobilization.
Table B.3. Juveniles with high concentrations of IGF-1 were heavier, but not taller, than
juveniles with low concentrations of IGF-1. Juveniles born to high-ranking mothers were both
heavier and taller than those born to low-ranking mothers. Mean mass ± SE = 22.88 ± 0.74 kg;
mean shoulder height ± SE = 60.39 ± 0.60 cm. Significant effects are highlighted in boldface at
a significance level of α = 0.05.

Trait
Juvenile
mass

Juvenile
shoulder
height

Fixed effect
Intercept

Parameter ± SE
-7.54 ± 1.90

95% CI
-11.3 − 3.7

Juvenile IGF-1
Mat. dom. rank
Prey density

0.02 ± 0.00
2.67 ± 0.93
0.01 ± 0.01

0.01 − 0.03
0.82 − 4.53
-0.01 − 0.02

Intercept
Juvenile IGF-1

-1.59 ± 1.71
6.1×10-3 ± 4.5×10-3

Mat. dom. rank
Prey density

1.72 ± 0.84
-5.0×10-3 ± 5.0×10-3

-5.0 − 1.83
-2.8×10-3 −
0.02
0.05 − 3.40
-0.18 − 0.01

t
3.96
3.80
2.89
1.23

df
65

P
<0.001

65
65
65

<0.001
0.005
0.224

-0.09
1.36

63
63

0.356
0.178

2.06
-0.73

63
63

0.043
0.470

Results from two separate multiple linear regression models predicting mass among 69 juvenile
female spotted hyenas and shoulder height among 67 juvenile female spotted hyenas. Juvenile
mass and juvenile shoulder height were expressed as residuals from separate linear
regressions predicting body mass and shoulder height by age, respectively. An individual that
was ‘large for its age’ would have a positive deviation (residual) from this line, and an individual
‘small for its age’ would have a negative deviation. Shoulder height data were unavailable for
two individuals. For both analyses shown in Table 3.3, ‘prey density’ was calculated as the
average local prey density during the 6 months prior to immobilization and blood sampling.

42

Table B.4a. Comparison of two multiple linear regression models predicting age at first
parturition using Akaike’s Information Criteria corrected for small sample sizes (AICc). Models
are comparable if ∆AICc<2.0. Only individuals with both IGF-1 and mass measurements were
included in models for comparison (n = 37).

Model 1
Model 2

df
5
5

logLik
-35.30
-37.04

AICc
82.5
86.0

43

∆AICc
0.0
3.5

Weight
0.85
0.15

Table B.4b. Female hyenas were significantly younger at their first parturition when they
exhibited higher juvenile concentrations of IGF-1, were born to higher-ranking mothers, and
experienced higher local prey density during the 6 months before parturition than other females.
Mean age at first parturition among our samples ± SE = 3.67 ± 0.13 yrs. Models correspond to
those in Table 3.4a except that all available data points were used in Model 1 (1 individual with
IGF-1 but no juvenile mass measurement). Significant effects are highlighted in boldface at a
significance level of α = 0.05.
Model 1: n = 38, adjusted R2 = 0.36, p = 0.0004
Fixed effect
Parameter ± SE
95% CI
Intercept
3.67 ± 0.11
3.45 − 3.90
Juvenile IGF-1
-2.6 × 10-3 ± 1.2 ×10-3 -5.1 ×10-3− -0.1 ×10-3
Mat. dom. rank
-0.57 ± 0.21
-1.00 − -0.15
Prey density
-2.7 × 10-3 ± 1.1 ×10-3 -0.4 ×10-3− 5.0 ×10-3

Model 2: n =37, adjusted R2 = 0.32, p = 0.001
Fixed effect
Parameter ± SE
95% CI
Intercept
3.642 ± 1.1×10-1
3.41 − 3.88
Juvenile mass
Mat. dom. rank
Prey density

-2.9×10-5 ± 2.7 ×10-2
-0.68×10-1 ± 0.25
-2.9 × 10-3 ± 1.2 ×10-3

-0.06 − 0.06
-1.18 − -0.18
-0.5 ×10-3− 5.2×10-3

t
33.11
-2.15
-2.75
2.43

df
34
34
34
34

t
31.77

df
33

-1.0 ×10-3
-2.75
2.46

33
33
33

P
<0.0001
0.039
0.009
0.021

P
<0.000
1
0.999
0.010
0.019

Results from two multiple linear regression models predicting age at first parturition, expressed
in years. All predictor variables were centered with a mean of 0 and standard deviation of 1.
Juvenile mass was expressed as residuals from a linear regressions predicting body mass by
age. An individual that was ‘large for its age’ would have a positive deviation (residual) from this
line, and an individual ‘small for its age’ would have a negative deviation. ‘Prey density’ for this
analysis was calculated as the average local prey density during the 6 months prior to the date
of first parturition.

44

Table B.5. Results from two separate Cox proportional hazards models predicting survival to
reproductive maturity. Heavier juvenile female hyenas experienced significantly greater survival
to reproductive maturity than did smaller females. Model 2 was the best-supported model based
on likelihood ratio tests. One individual had no mass recorded so was not included in Model 2,
in which mass was a predictor variable; however, this individual did have an IGF-1 value so was
included in Model 1. Significant effects are highlighted in boldface at a significance level of α =
0.05.

Model 1: n= 62; χ2 = 3.29; df = 2, P = 0.19, model log likelihood = -88.59
Predictor variable
β
Hazard ratio
z
Juvenile IGF-1
Mat. dom. rank

-0.32
-0.40

0.72
0.67

-1.46
-1.06

Model 2: n= 61; χ2 = 6.20; df = 2, P = 0.45, model log likelihood = -86.67
Predictor variable
β
Hazard ratio
z
Juvenile mass
Mat. dom. rank

-0.11
-0.07

0.90
0.93

-2.27
-0.17

P
0.145
0.288

P
0.023
0.868

All predictor variables were z-transformed to a mean of 0 and standard deviation of 1. A
negative hazard rate coefficient (β) indicates a decreased probability of survival with increasing
value of the predictor variable. A hazard ratio (calculated as exp(β)) of 2, for example, indicates
that individuals are twice as likely to survive with increasing values of the predictor variable.

45

Table B.6. Results from two separate Cox proportional hazards models predicting survival past
reproductive maturity. Individuals with higher juvenile concentrations of IGF-1 were more likely
to survive than those with lower juvenile concentrations. Model 1 was the best-supported model
based on likelihood ratio tests. One individual had no mass recorded so was not included in
Model 2, in which mass was a predictor variable; however, this individual did have an IGF-1
value so was included in Model 1. Significant effects are highlighted in boldface at a significance
level of α = 0.05.

Model 1: n= 39; χ2 = 7.88; df = 2, P = 0.02, model log likelihood = -88.98
Predictor variable
β
Hazard ratio
z
IGF-1
Mat. dom. rank

0.45
-0.67

1.57
0.51

2.28
-1.95

Model 2: n= 38; χ2 = 2.98; df = 2, P = 0.23, model log likelihood = -88.23
Predictor variable
β
Hazard ratio
z
Juvenile mass
Mat. dom. rank

0.02
-0.62

1.02
0.54

0.61
-1.79

P
0.023
0.052

P
0.539
0.074

All predictor variables were z-transformed to a mean of 0 and standard deviation of 1. A
negative hazard rate coefficient (β) indicates a decreased probability of survival with increasing
value of the predictor variable. A positive hazard rate coefficient (β) indicates an increased
probability of survival with increasing value of the predictor variable. A hazard ratio (calculated
as exp(β)) of 2, for example, indicates that individuals are twice as likely to survive with
increasing values of the predictor variable.

46

Table B.7a. Comparison of three multiple linear regression models predicting annual
reproductive success using Akaike’s Information Criteria corrected for small sample sizes
(AICc). Models are comparable if ∆AICc<2.0. Females were only included in models for
comparison if they survived past 3 years of age, had known death dates, had successfully
weaned at least one offspring, and had complete IGF-1, mass, and age at first parturition
measurements (n = 25).

Model 1
Model 2
Model 3

df
4
4
4

logLik
8.17
0.52
-0.24

AICc
-6.3
9.0
10.5

47

∆AICc
0.0
15.3
16.8

Weight
1
<0.001
<0.001

Table B.7b. Results from three linear models to determine a female’s annual reproductive
success. Models correspond to those in Table 3.7a except that all available data points for each
model were used below. Significant effects are highlighted in boldface at a significance level of
α = 0.05.
Model 1: n =26, adjusted R2 = 0.54, p = 0.00005
Fixed effect
Parameter ± SE
95% CI
Intercept
0.65 ± 0.04
0.57 − 0.73
Age at first
-0.20 ± 0.05
-0.30− -0.11
parturition
Mat. dom. rank
0.06 ± 0.08
-0.11 − 0.22

t
17.43
-4.34

df
23
23

0.71

23

Model 2: n =26, adjusted R2 = 0.22, p = 0.022
Fixed effect
Parameter ± SE
95% CI
Intercept
0.65 ± 0.05
0.55 − 0.75
Juvenile IGF-1
6.9 × 10-4 ± 5.3 ×10-4 -4.1 ×10-4− 1.8 ×10-3
Mat. dom.
0.23 ± 0.09
0.05 − 0.42
rank
Model 3: n =25, adjusted R2 = 0.17, p = 0.048
Fixed effect
Parameter ± SE
Intercept
0.65 ± 0.05
Juvenile mass
0.008 ± 0.01
Mat. dom. rank
0.22 ± 0.10

95% CI
0.54 − 0.76
-0.02− 0.04
0.02− 0.43

t
13.38
1.30
2.57

t
12.48
0.58
2.26

P
<0.0001
0.0002
0.486

df
23
23
23

df
23
23
23

P
<0.0001
0.207
0.017

P
<0.0001
0.571
0.034

A female’s annual reproductive success was calculated as the number of her weaned cubs
divided by her reproductive lifespan. Both female and male offspring are included.
Results from three multiple linear regression models predicting age at first parturition, expressed
in years. All predictor variables were centered with a mean of 0 and standard deviation of 1.

48

Table B.8a. Comparison of four multiple linear regression models predicting lifetime
reproductive success using Akaike’s Information Criteria corrected for small sample sizes
(AICc). Models are comparable if ∆AICc<2.0. Females were only included in models for
comparison if they survived past 3 years of age, had known death dates, and had complete
IGF-1, mass, and age at first parturition measurements (n = 29).

Model 1
Model 3
Model 2
Model 4

df
4
4
4
4

logLik
-19.73
-31.75
-33.16
-34.07

AICc
49.1
73.2
76.0
77.8

49

∆AICc
0.0
24.0
26.8
28.7

Weight
1
<0.001
<0.001
<0.001

Table B.8b. Results from three linear models to determine a female’s lifetime reproductive
success. All available data points for each model were used below. Significant effects are
highlighted in boldface at a significance level of α = 0.05.
Model 1: n =35, adjusted R2 = 0.78, p = 1.39×10-11
Fixed effect
Parameter ± SE
95% CI
Intercept
2.96 ± 0.25
2.45 − 3.46
Lifespan
0.65 ± 0.07
0.51− 0.79
Mat. dom. rank
1.12 ± 0.47
0.16 − 2.08

t
11.95
9.63
2.39

Model 2: n =32, adjusted R2 = 0.19, p = 0.030
Fixed effect
Parameter ± SE
Intercept
2.86 ± 0.46
Juvenile mass
0.14 ± 0.12
Mat. dom. rank
1.75 ± 0.89

95% CI
1.93 − 3.79
-0.10 − 0.38
-0.08 − 3.57

Model 3: n =34, adjusted R2 = 0.19, p = 0.01
Fixed effect
Parameter ± SE
Intercept
3.06 ± 0.48
Juvenile IGF-1
-5.5 × 10-3 ± 5.4 ×10-3
Mat. dom. rank
2.68 ± 0.89

95% CI
2.08 − 4.05
-0.16 − 5.4 ×10-3
0.87 − 4.50

Model 4: n =30, adjusted R2 = 0.09, p = 0.11
Fixed effect
Parameter ± SE
Intercept
3.26 ± 0.56
Age at first
-0.19 ± 0.69
parturition
Mat. dom. rank
2.01 ± 1.09

t
6.28
1.17
1.96

df
32
32
32

P
<0.0001
<0.0001
0.023

df
29
29
29

P
<0.0001
0.252
0.060

df
31
31
31

t
6.33
-1.03
3.01

P
<0.0001
0.310
0.005

95% CI
2.12 − 4.40
-1.59 − 1.22

t
5.86
-0.27

df
27
27

P
<0.0001
0.788

-0.22− 4.24

1.85

27

0.075

A female’s annual reproductive success was calculated as the number of her weaned cubs
divided by her reproductive lifespan. Both female and male offspring are included.
Results from two multiple linear regression models predicting age at first parturition, expressed
in years. All predictor variables were centered with a mean of 0 and standard deviation of 1.

50

Figure B.1. Morphological traits measured during routine immobilization of study animals.
Shoulder height used for this study indicated by measure “c” above. Reproduced from:
Swanson, E.M., Dworkin, I. & Holekamp, K.E. (2011) Lifetime selection on a hypoallometric size
trait in the spotted hyena. Proceedings of the Royal Society B: Biological Sciences, 278, 3277–
85.

51

Figure B.2. IGF-1 standards (open circles; bottom x-axis) and serial dilutions of a pooled
sample of hyena plasma (filled circles; inset x-axis) plotted against optical density values (yaxis). Samples for this study were diluted to 1:20.

52

1.2

●
●

1.0

●
●

●

0.8

●
●

●

●

●

●

●

0.6

●
●
●
●
●

0.4

Annual reproductive success

●

●

●

●
●
●
●

0.2

●

●

3

4

5

6

Age at first parturition (yr)

Figure B.3. Female spotted hyenas that were younger at first parturition experienced higher
annual reproductive success than others (N = 26 females that survived past 3 years of age,
successfully weaned at least one cub, and had known death dates; r2 = 0.57; t24 = -5.61; P <
0.0001). A female’s annual reproductive success was calculated as the number of her weaned
cubs divided by her reproductive lifespan. Both female and male offspring are included.

53

12

●
●

8
6

●

●

4

●
●●

●●

0

2

Number of Cubs Weaned

10

●

● ●
●

●●

●

●●

●

●●●●●

●

5

●

●

●

● ●

●
●

●
●

10

15

Lifespan (yr)

Figure B.4. Female spotted hyenas that lived longer weaned more cubs than those with
relatively short lifespans (N = 35 females that survived past 3 years of age and with known
death dates; r2 = 0.76; t33 = 10.33, P <0.0001). Data points include five females that did not give
birth to any offspring and four females that failed to wean any offspring. Both female and male
offspring are included.

54

CHAPTER 4: INSULIN-LIKE GROWTH FACTOR-1 (IGF-1) AS A POTENTIAL MEDIATOR OF
FEMALE-BIASED SEXUAL SIZE DIMORPHISM IN WILD SPOTTED HYENAS
Nora Lewin, Eli M. Swanson, Barry L. Williams, and Kay E. Holekamp

55

Introduction
Sexual size dimorphism (SSD) occurs within species when males and females differ in
their average adult body sizes (Fairbairn, Blanckenhorn & Szekely 2007). Among mammals,
SSD is largely male-biased with sex-specific patterns of sexual selection explaining most of this
variation (Darwin 1871; Ralls 1977; Lindenfors, Gittleman & Jones 2007). Male mammals often
secure resources (e.g. mates, territories, etc.) by fighting, and larger males generally win these
fights. Larger male body size is typically accompanied by sexually selected weapons or
ornaments (Andersson 1994). Although less frequent than male-biased SSD, female-biased
SSD does occur in mammals and, in contrast to this phenomenon in males, it rarely appears to
be the consequence of sexual selection (Ralls 1976). The selective pressures favoring femalebiased SSD are currently being explored (Schulte-Hostedde et al. 2002; Kilanowski &
Koprowski 2016), but the mechanisms mediating its occurrence remain poorly understood.
Theoretically, the mechanisms that facilitate male-biased SSD should operate similarly
in both male- and female-biased species due to high genetic overlap and the shared
mechanisms that regulate growth (Badyaev 2002). Indeed, one of most remarkable aspects of
SSD is that such sex-specific phenotypic diversity has evolved despite strong intersexual
genetic correlations. Thus, understanding the proximate mechanisms mediating SSD has the
potential, not only to reveal common targets of selection, but also to illuminate evolutionary
solutions to problems posed by genetic constraints. Furthermore, mechanistic studies may also
inform and stimulate hypotheses explaining the ultimate causes of both female- and malebiased SSD (Badyaev 2002; Bernstein 2010).
Although testosterone has been hypothesized to be the main driver of SSD, inconsistent
results suggest that this relationship is neither direct nor universal across species (Jansson,
Edén & Isaksson 1985; Sadighi et al. 2001; Taylor & Denardo 2005; John-Alder, Cox & Taylor
2007; Cox, Stenquist & Calsbeek 2009; Starostova et al. 2013). One major problem for this
hypothesis is the failure of testosterone to explain female-biased SSD. For example, in the

56

spotted hyena (Crocuta crocuta), females are larger than males (Ralls 1976; Swanson et al.
2013), yet concentrations of testosterone are higher in developing males than in females
(Glickman et al. 1987; Licht et al. 1992). A more plausible scenario might be that gonadal
hormones act as sex-specific modifiers that orchestrate changes in shared anabolic pathways
(Wade 1976; Holzenberger et al. 2001; Cox et al. 2017); however, it remains unknown whether
the pathways associated with male-biased SSD are also associated with female-biased SSD.
The polypeptide hormone, insulin-like growth factor-1 (IGF-1) is a plausible mediator of
SSD. Its positive direct effects on growth rates have been documented in both sexes among a
variety of invertebrate and vertebrate taxa (Sonntag et al. 1999; Tatar et al. 2003; Kenyon 2010;
Panici et al. 2010; Svensson et al. 2011; Yuan et al. 2012; Junnila et al. 2013; Bartke et al.
2013; Lewin et al. 2016). Although the liver is the primary source of circulating concentrations of
IGF-1, most tissues of the body also produce IGF-1 (Cohick & Clemmons 1993). The dual
endocrine and paracrine actions of IGF-1 allow it to direct coordinated and essential
transformations during ontogenetic development (Cohick & Clemmons 1993; Holzenberger et
al. 2003). Perhaps some of the strongest support for a role of IGF-1 in the mediation of SSD
comes from studies of domesticated animals in which body size has been artificially selected
(Gatford et al. 1998). For example, a single nucleotide polymorphism in the IGF1 gene accounts
for a substantial amount of variation in size between giant and small breeds of dogs (Sutter et
al. 2007). Although work in free-living populations is limited, IGF-1 explained variation in size
and life-history patterns among garter snakes (Sparkman et al. 2009, 2010), and a recent study
of sexually dimorphic anole lizards revealed significant sex-biased expression in IGF-1 (Cox et
al. 2017). These findings make IGF-1 a particularly intriguing candidate mechanism for
mediating SSD in natural populations.
Here, we investigated IGF-1 as a potential mediator of female-biased SSD in the spotted
hyena; our work had three specific goals. The first was to describe intra- and inter-sexual
variation in circulating IGF-1 concentrations across the lifespan in this species. In spotted

57

hyenas the period of dimorphic growth is between 1 and 4 years of age, when females grow
faster than males (Swanson et al. 2013); during the first year of postnatal life the sexes do not
differ in size. If IGF-1 mediates the emergence of SSD in spotted hyenas, then we would expect
to find sex differences in IGF1 concentrations during this dimorphic growth period. Profiles of
IGF-1 concentrations across the lifespan are rare among free-ranging mammals, particularly for
both sexes (but see & Kemnitz, 1992; Suzuki & Ishida, 2001), so our data should represent a
substantial contribution to this literature.
Our second goal was to describe the sensitivity of postnatal IGF-1 concentrations to
socio-ecological variables in both sexes. Postnatal development represents not only a time of
significant size growth but also a potential sensitive window during which physiological
processes may be capable of exerting long-lasting effects on later-life phenotypes and fitness
(Bateson 1979; English et al. 2016). Theoretically, IGF-1 could mediate SSD through sexspecific sensitivities during postnatal development. For example, size disparities can result
when a growth trait in one sex responds differently to an environmental stimulus than does the
same trait in the opposite sex (Badyaev et al. 2001). Predicting which sex should be more or
less sensitive is not straightforward (Stillwell et al. 2010), but the adaptive canalization
hypothesis proposes one approach. Over time, the development of traits that are closely linked
to fitness converges on an optimal “canal” leaving development more robust against fluctuating
extrinsic conditions (Waddington 1942; Stearns & Kawecki 1994; Fairbairn 2005). If we were to
extend this hypothesis to spotted hyenas, then we might expect juvenile IGF-1 concentrations in
females to be more resistant to variation in socio-ecological variables than in males because
female size is positively tied to fitness (Swanson et al. 2011; Lewin et al. 2016). Furthermore,
our previous analysis of females revealed that juvenile IGF-1 concentrations did not vary with
maternal rank, age, or prey density (Lewin et al. 2016), suggesting that IGF-1 might be
canalized in females. However, whether and how IGF-1 concentrations might vary with
environmental fluctuations among juvenile male hyenas remains unknown.

58

We selected socio-ecological variables likely to affect food intake because postnatal diet
is known to influence IGF-1 concentrations (Isley, Underwood & Clemmons 1983; Breese,
Ingram & Sonntag 1991; Clemmons & Underwood 1991; Kappeler et al. 2009). Juvenile
hyenas, called cubs, are dependent on their mother’s milk before weaning at around 1 year of
age (Holekamp et al. 1996), so a mother’s resources should be expected to affect her
offspring’s nutritional status (Hinde & Milligan 2011). Spotted hyenas live in groups structured by
linear dominance hierarchies in which an individual’s social rank determines its priority of
access to food resources (Kruuk 1972; Tilson & Hamilton 1984; Frank 1986). As a result,
maternal social rank may influence the IGF-1 concentrations in her developing offspring.
Spotted hyenas typically bear litters of one or two cubs, and a cub’s intra-litter dominance rank
directly determines its access to milk (Frank, Glickman & Licht 1991; Smale et al. 1995; Wahaj
& Holekamp 2006; Benhaiem et al. 2012). Effects of maternal and intra-litter dominance rank on
juveniles are most evident during periods of low resource availability (Holekamp et al. 1996;
Hofer & East 2003; Wahaj & Holekamp 2006) so we also considered local prey abundance as a
variable that might affect IGF-1 concentrations. Finally, maternal parity can affect postnatal IGF1 concentrations in some species (Hyatt et al. 2007; Bernstein et al. 2012) so we also took a
mother’s reproductive history into account.
Our final goal was to investigate the adaptive significance of sex differences in IGF-1
concentrations. Although male size in adulthood is not directly linked to reproductive fitness in
hyenas (Höner et al. 2010), SSD might emerge through sexually dimorphic selection pressures
operating during ontogeny (Lande & Arnold 1985; Badyaev et al. 2001; Karubian & Swaddle
2001; Schulte-Hostedde et al. 2002). Hyenas experience a period of intense selection (i.e. high
mortality) after weaning but before reaching sexual maturity (Watts et al. 2009), during which
SSD emerges in hyenas (Swanson et al. 2013). If patterns of IGF-1 underlie female-biased SSD
in hyenas, we would expect size to have sexually dimorphic effects on survival to sexual
maturity. Among juvenile females, circulating IGF-1 concentrations during the first year of

59

postnatal life correlate positively with relative juvenile mass, which predicts greater probability of
survival to adulthood (Lewin et al. 2016). However, these relationships remain unstudied among
juvenile males. If selection affects males and females differently early in life, then we expected
to see a weaker relationship between juvenile size and survival among males than among
females.
Materials and Methods
Field study site, subjects, and quantification of independent variables
Our subjects were part of a long-term field study of spotted hyenas within the Masai
Mara National Reserve, Kenya. Behavioral, demographic and body size data were collected
along with blood samples from members of three social group, or clans, called the Talek, Mara
River, and Fig Tree clans. The Talek clan has been monitored continuously since 1988; data
from the Mara River and Fig Tree clans have been collected since 2001 and 2007, respectively.
Ages of hyenas in their natal clans were known to ± 7 days based on methods described earlier
(Holekamp et al. 1996). We defined juvenile hyenas as those less than 1 year old. We
estimated ages of immigrant males (± 6 months) using tooth wear data (Van Horn et al. 2003).
Conception dates and reproductive states were calculated by subtracting the 110 day gestation
period from parturition dates (Holekamp et al., 1996). We classified maternal parity as either
‘primiparous’ if the sampled cub was a member of the mother’s first litter, or ‘multiparous’ if the
cub was a member of a subsequent litter.
Within our study population, males typically disperse between 2 and 6 years of age
(Smale et al. 1997; Boydston et al. 2005), but females are philopatric. Natal males were
considered to have dispersed only if they had undertaken exploratory forays outside the natal
territory, were in good health when last seen, and were last seen in the natal territory after 2
years of age (Van Horn et al. 2003). Otherwise males were considered to have died if they
disappeared before they reached 2 years of age. We observed our study clans daily so we
recorded a male’s date last seen as its date of death.

60

As part of our long-term hyena field study, dominance rank matrices were constructed
annually, based on outcomes of dyadic agonistic interactions (Smale et al. 1993). Here we
calculated an adult hyena’s within-sex relative rank as the proportion of adults of the same sex
over which it was dominant during the year of sampling. We standardized social rank to account
for variation in clan size at time of sampling; the individual with a relative rank of 1 was the
highest-ranking animal in the clan at that time, and the individual with a relative rank of -1 was
the lowest ranking. We assigned pre-reproductive females their mothers’ standardized relative
ranks at the time of their births. We determined intra-litter dominance rank based on outcomes
of agonistic interactions between littermate siblings in twin litters (Smale et al. 1995). Cubs were
then assigned as ‘dominant twin’, ‘subordinate twin’, or ‘singleton’ if they had no littermate.
The Masai Mara National Reserve is characterized by open, rolling grasslands, home to
multiple resident ungulate species. For 3 to 4 months each year, large migratory herds of
wildebeest (Connochaetes taurinus) and zebra (Equus burchelli) move into the Reserve
creating a superabundance of available prey for spotted hyenas (Holekamp et al. 1997).
Spotted hyenas appear to hunt whichever species is most abundant (Kruuk 1972; Cooper 1990;
Hofer & East 1993b; Holekamp et al. 1997). We calculated average local prey density as the
mean number of ungulates counted within 100 meters of multiple 4-km transect lines run
through each clan’s territory at biweekly intervals during the six months prior to collection of
each blood sample.
We analyzed a total of 318 plasma samples collected between 1990 and 2014 from 271
unique individuals aged 0.5-15.0 years at the time of immobilization and blood sample
collection. Hyenas younger than 0.5 years old could not be safely immobilized. As part of the
longitudinal study, juvenile hyenas are routinely anaesthetized with Telazol (Zoetis, South Bend,
IN; 6.5 mg/kg) administered in a plastic dart fired from a CO2-powered rifle (Telinject Inc.,
Saugus, CA). All immobilizations were conducted early in the morning and all were carried out
according to guidelines specified by the American Society of Mammalogists (Sikes 2016) and

61

approved by the Institutional Animal Care and Use Committee at Michigan State University
(MSU; AUF #05/14-087-00). Blood samples were taken from the jugular vein in heparinized
vacutainer tubes, centrifuged, and plasma was drawn off, aliquoted, then promptly placed in
liquid nitrogen until it could be shipped on dry ice to Michigan State University, where samples
were stored at -80ºC. Hyena body mass was recorded in kilograms (kg).
Quantification of plasma IGF-1 concentrations
We asssayed plasma IGF-1 concentrations using methods described previously (Lewin
et al. 2016). Briefly, we quantified plasma IGF-1 levels using a commercially available enzyme
immunoassay kit (EIA; ALPCO Diagnostics, Salem, NH). We validated this assay for spotted
hyenas using the ‘spike and recovery’ method following kit protocol. All samples were run in
duplicate and randomly assigned to plates with mean intra- and inter-assay coefficients of
variation of 3.06% and 10.5%, respectively. Plasma samples were kept frozen at -80°C until
analysis with the length of time frozen varying between 3 months and 24 years. We found no
significant influence of freezer storage time (r2 < 0.001, t316 = -0.83, P =0.41).
Data analysis
Our dataset was composed of IGF-1 concentrations for 318 samples from 271
individuals varying in age and number of samples per individual, as some individuals were
sampled multiple times. To compare differences in IGF-1 concentrations between the sexes, we
used separate simple linear regressions among hyenas of five different age groups: 0-0.99
years, 1-1.99 years, 2-2.99 years, 3-3.99 years, and 4+ years. Where IGF-1 concentrations
could not be normalized using traditional transformations we assessed sex differences using
non-parametric Mann-Whitney-Wilcoxon tests. Residuals from all linear regression models
satisfied assumptions of normality based on Shapiro-Wilk tests.
We used a two-tailed paired t-test on data from 9 females to compare within-individual
IGF-1 concentrations between pregnancy and lactation. Where females had more than one
sample per reproductive condition we took an average of their IGF-1 values in each condition.

62

Similarly, we used a paired t-test on 10 pairs of age-matched males to compare adult immigrant
males with adult males still living in their natal territory; ages within these adult male pairs never
differed by more than 12 months.
To determine the influence of juvenile IGF-1 concentrations on juvenile male mass, we
built a multiple linear regression model using juvenile IGF-1, maternal rank, and prey density as
predictor variables, similar to the model previously used in females (Lewin et al. 2016). Mass
measurements were unavailable for 5 males; however, these individuals were included in all
other analyses. To evaluate survival to reproductive maturity among males, we assessed male
survival to 2 years of age with a Cox proportional hazards model identical to that previously
used for analysis of survival among females in our study population (Lewin et al. 2016). All
independent variables were mean centered with a standard deviation of 1.
To evaluate sensitivity of IGF-1 concentrations to socio-ecological variables during early
postnatal life we built separate multiple linear regression models for females and males aged 00.99 years. Residuals from all linear regression models satisfied assumptions of normality
based on Shapiro-Wilk tests. We used Tukey’s post-hoc tests to identify significant differences
among individuals based on intra-litter rank status. Non-significant interaction terms were
removed from models to avoid overestimation (Bolker et al. 2009). To examine the effect of
intra-litter dominance rank on IGF-1, we did a separate analysis with a two-tailed paired t-test
looking at only twins sampled within two weeks of one another (15 litters: 3 male-male pairs, 4
female-female pairs, 8 female-male pairs; 8 litters where the subordinate twin was sampled first,
7 litters where the dominant twin was sampled first). All independent variables were mean
centered with a standard deviation of 1.
Results
Variation in IGF-1 concentrations with age and sex
Both sexes showed a gradual decline in IGF-1 concentrations over the lifespan with the
highest mean values seen in hyenas younger than 4 years old (Table C.1, Figure 4.1). Females

63

had significantly higher IGF-1 concentrations than did males in every age group (Table C.1) and
across the growth period of 1-4 years of age (W = 4151, p = 0.0007). Among 9 females with
IGF-1 concentrations measured during both pregnancy and lactation, IGF-1 concentrations
were significantly higher during pregnancy than were those during lactation (t8 = 4.52, P =
0.002; Figure C.1). A two-tailed paired t-test on 10 pairs of age-matched males revealed no
significant difference in IGF-1 concentrations between immigrant and adult natal males (t9 = 1.09, P = 0.30).

64

500
100

300

IGF-1 (ng/ml)

500
400

0

3
Age (years)

4

200

300

2

0

100

IGF-1 (ng/ml)

1

0

2

4

6

8

10

12

14

Age (years)

Figure 4.1. Patterns of circulating concentrations of IGF-1 across the lifespan in male (filled
circles) and female (open circles) spotted hyenas. Inset is magnified between 1-4 years of age
to better visualize the dimorphic growth period.

65

Sex-specific sensitivity of IGF-1 concentrations among males and females
Among juvenile males, maternal social rank, prey density, maternal parity, and intra-litter
dominance rank significantly predicted IGF-1 concentrations (Table 4.2). Among 15 pairs of
mixed-sex (n=8) and same-sex (n=7) twin littermates sampled within 2 weeks of one another,
dominant twins had significantly higher concentrations of IGF-1 than did their subordinate
littermates (two-tailed paired t-test: t14 = 4.03, P = 0.001). However, this relationship was not
significant when we only looked at same-sex litters (two-tailed paired t-test: t6 = 1.76, P = 0.13).
Among juvenile females, we found no support for age, maternal social rank, prey density,
maternal parity, or intra-litter dominance rank as significant predictors of IGF-1 concentrations
(Table 4.3).

66

Table 4.1. Results from (a) multiple linear regression models to estimate influence of socioecological factors on IGF-1 concentrations among juvenile males, ages 0.5-1.0 years (n = 48).
(b) Contrasts using least-squares means for intra-litter dominance rank. P-values indicating
statistically significant differences at α = 0.05 are bolded.
(a) Model with singleton cub born to a multiparous mother set as the intercept.
Model fit: n = 48, adjusted R2 = 0.22, p = 0.012
Fixed effect
Parameter ± SE
95% CI
Intercept
5.64 ± 0.10
5.44− 5.84
Age
0.95 ± 0.62
-0.30− 2.20
Maternal rank
0.34 ± 0.13
-0.07 −0.61
Intra-litter rank - dom
-0.10 ± 0.26
-0.62− 0.42
Intra-litter rank - sub
-0.41 ± 0.17
-0.74−-0.07
Maternal parity -0.58 ± 0.18
-0.94−-0.22
primiparous
Prey density
1.1×10-3 ± 1.4×10-3 -1.8×10-3−
4.0×10-3

t
57.05
1.54
2.54
-0.40
-2.45
-3.29

df
41
41
41
41
41
41

P
<0.0001
0.13
0.01
0.69
0.02
0.002

0.77

41

0.45

(b) Same model as above with dominant twin set as intercept (all other estimates are identical).
Contrast
Intercept
Intra-litter rank - sub
Singleton cub

Parameter ± SE
5.56 ± 0.13
-0.18± 0.17
0.11± 0.22

95% CI
5.29− 5.83
-0.52− 0.16
-0.33− 0.54

t
41.37
-1.07
0.50

df
41
41
41

P
<0.0001
0.29
0.62

All continuous predictor variables were z-transformed to a mean of 0 and standard deviation of
1. Model residuals did not violate assumptions of normality (Shapiro-Wilk tests of normality: W ≥
0.95, P ≥ 0.06). Juvenile hyenas were assigned their mother’s dominance rank (Maternal rank)
at the time of immobilization; a relative rank of 1 was the highest-ranking animal in the clan, and
the individual with a relative rank of -1 was the lowest ranking. Age was calculated in months
using known birthdates. Maternal parity was coded as either primiparous or multiparous.

67

Table 4.2. Results from (a) multiple linear regression models to estimate influence of socioecological factors on IGF-1 concentrations among juvenile females, ages 0.5-1.0 years (n = 67).
(b) Contrasts using least-squares means for intra-litter dominance rank. P-values indicating
statistically significant differences at α = 0.05 are bolded.
(a) Model with singleton cub born to a multiparous mother set as the intercept.
Model fit: n = 67, adjusted R2 = 0.002, p = 0.418
Fixed effect
Parameter ± SE
95% CI
Intercept
306.27 ± 25.93
254.39 − 358.15
Age
-33.04 ± 96.93
-226.93 − 160.85
Maternal rank
-10.14 ± 23.52
-57.18 − 36.90
Intra-litter-rank - dom
9.92 ± 30.83
- 51.76 − 71.59
Intra-litter-rank - sub
-60.46 ± 36.92
-134.30 − 13.38
Maternal parity 27.69 ± 36.39
-45.10 − 100.48
primiparous
Prey density
0.17 ± 0.19
-0.21− 0.56

t
11.81
-0.34
-0.43
0.32
-1.64
0.76

df
60
60
60
60
60
60

P
<0.0001
0.73
0.67
0.75
0.11
0.45

0.91

60

0.37

(b) Same model as above with dominant twin set as intercept (all other estimates are identical).
Contrast
Parameter ± SE
95% CI
df
t
P
Intercept
316.19 ± 16.98
282.23 − 350.15 18.65 60
<0.0001
Intra-litter rank - sub
-9.92 ± 30.83
-60.52 − 62.57
-0.32 60
0.75
0.02
Singleton cub
-70.38 ± 30.30
-127.16 − -4.43
-2.32 60

All continuous predictor variables were z-transformed to a mean of 0 and standard deviation of
1. Model residuals did not violate assumptions of normality (Shapiro-Wilk tests of normality: W ≥
0.98, P ≥ 0.23). Juvenile hyenas were assigned their mother’s dominance rank (Maternal rank)
at the time of immobilization; a relative rank of 1 was the highest-ranking animal in the clan, and
the individual with a relative rank of -1 was the lowest ranking. Age was calculated in months
using known birthdates. Maternal parity was coded as either primiparous or multiparous.

68

Male survival to reproductive maturity
Juvenile IGF-1 concentrations significantly predicted relative juvenile mass among males
(t39 = 5.15, P < 0.001; Table 4.4), but maternal rank and prey density did not (t39 = -1.76, P =
0.09; t39 = 1.44, P = 0.16, respectively; Table C.2). Neither relative juvenile mass (z = 0.05, P =
0.96) nor maternal rank (z = 0.83, P = 0.43) significantly predicted survival to reproductive
maturity among males (Table C.3; Figure 4.2a), as it did for females (Figure 4.2b). Similarly, a
model with juvenile IGF-1 as a covariate instead of juvenile mass also failed to predict male
survival (z = -1.58, P = 0.12; Table C.3).

69

0.6
0.4

Relative juvenile mass

0.2

Male survivorship

0.8

1.0

A

0.0

Large (n = 8)
Medium (n = 22)
Small (n = 13)

0.0

0.5

1.0

B

1.5

2.0

0.6
0.4
0.2

Female survivorship

0.8

1.0

Age (years)

Relative juvenile mass

0.0

Large (n = 16)
Medium (n = 37)
Small (n = 8)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Age (yr)

Figure 4.2. Juvenile mass did not predict survival to reproductive maturity in (A) males, but did
so in (B) females. Relative mass was statistically analyzed as a continuous variable (Tables 4.4
& 4.5), but partitioned into tertiles for illustrative purposes only.

70

Discussion
In non-primate mammals, sexually dimorphic patterns of IGF-1 are typically male-biased
where males also exhibit larger mean body mass than females (Gatford et al. 1998). In the
present study, we found higher IGF-1 concentrations in the larger sex of a species displaying
female-biased SSD (Table C.1, Figure 4.1), suggesting that similar physiological mechanisms
may be responsible for dimorphic growth in species with either male- or female-biased SSD.
Specifically, female spotted hyenas exhibited consistently higher IGF-1 concentrations than did
males throughout the lifespan, starting as early as the first postnatal year of life (Table C.1;
Figure 4.1). Interestingly, dimorphic growth does not occur until the second year of postnatal life
(Swanson et al. 2013). Previous work in our study population confirmed that females grow faster
than males between 1-4 years of age, rather than longer (Swanson et al. 2013), so our finding
of higher IGF-1 concentrations in females just before and during the growth period implicates a
role for IGF-1 in mediating sex-specific growth rates. Developing tissues may have sex-specific
sensitivities to IGF-1 (Gatford et al. 1998; Holzenberger et al. 2001), but IGF-1 concentrations
had similar slopes in the regression models predicting juvenile mass in males (Table C.2) and
females (Table C.3). These results, along with the similar IGF-1 patterns observed across the
lifespan in both sexes (Figure 4.1), suggest that sex differences in growth rate may be driven
more by absolute IGF-1 concentrations than sex-specific sensitivities in target tissues.
Reproductive status appears to influence circulating IGF-1 concentrations among
female, but not male, spotted hyenas. We observed significantly higher IGF-1 concentrations in
pregnant female hyenas than in the same individuals when they were lactating (Figure C.2). The
increase in IGF-1 levels before, and the subsequent decline after, parturition may be partially
explained by the relationship between maternal IGF-1 and fetal growth. In humans, maternal
IGF-1 levels positively correlate with placental size and surface area, thus potentially facilitating
the exchange of nutrients between the mother and fetus (Liu et al. 1994; Chellakooty et al.
2004).

71

We found no significant difference in IGF-1 concentrations between age-matched natal
and immigrant adult males suggesting that dispersal status alone does not appear to affect IGF1 concentrations in male spotted hyenas. Among adult male spotted hyenas, post-immigration
dispersal status is associated with higher concentrations of plasma testosterone than those
found in adult males still living with their natal clans (Holekamp & Smale 1998; Holekamp & Sisk
2003; Dloniak et al. 2004). Experimental elevation of testosterone raises IGF-1 levels (Hobbs et
al. 1993), so immigrant males would be expected to have higher IGF-1 concentrations than
natal males if those concentrations were being mediated by T.
Maternal and familial variables played a significant role in predicting IGF-1
concentrations, but only among juvenile male hyenas. Specifically, juvenile males born to highranking, multiparous mothers exhibited the highest IGF-1 concentrations (Table 4.3). If food
intake influences IGF-1 concentrations then the greater access to resources afforded to mothers
of higher social rank should be expected to affect offspring concentrations. The differences in
IGF-1 levels between primiparous and multiparous mothers suggests that experienced mothers
may be better at acquiring resources and/or provisioning offspring. Interestingly, we did not find
an effect of prey density on IGF-1 levels, which suggests that mothers may vary in how they
mediate the effects of environmental conditions on their developing offspring. The distribution of
maternal resources via litter size also influenced IGF-1 concentrations. Our within-sex analysis
(Table 4.3) showed that subordinate male cubs displayed significantly lower IGF-1
concentrations than did singleton male cubs but not dominant male cubs. A failure to see
significant differences between siblings of 7 same-sex litters also confirmed that litter size might
be a more important factor in predicting IGF-1 concentrations than intra-litter rank.
Sex-biased sensitivity of growth traits to extrinsic factors has been reported in a range of
wild vertebrate species exhibiting varying degrees of size differences between the sexes
(Leberg & Smith 1993; Sheldon et al. 1998; Post et al. 1999; Toïgo et al. 1999; Badyaev et al.
2001; Kalmbach, Furness & Griffiths 2005; Råberg, Stjernman & Nilsson 2005; Verhulst,

72

Holveck & Riebel 2006; Wilkin & Sheldon 2009; Hinde 2009; Bernstein et al. 2012). In many of
these studies, the precise divergence and duration of male and female growth patterns during
postnatal development are believed to contribute largely to the realized difference, or degree, of
SSD observed in adulthood. For example, in red deer (Cervus elaphus), female and male
growth rates responded differently to shared environmental conditions leading to greater
degrees of SSD during periods of warmer climate and smaller degrees in periods of varying
food abundance (Post et al. 1999). The adaptive canalization hypothesis seeks to explain sexbiased trait sensitivity further by predicting greater trait robustness in the sex for which the
studied trait is positively linked to fitness (Stearns & Kawecki 1994; Fairbairn 2005). Because
juvenile IGF-1 concentrations correlated positively with larger juvenile mass and higher juvenile
survival among female hyenas, we predicted that females would be the more robust sex.
Consistent with this prediction, plasma IGF-1 concentrations of juvenile females did not vary
with environmental variables (Table C.3), whereas those of juvenile males did so (Table C.2).
Sex-biased sensitivity and canalization may be one possible outcome when the selective
optimum for a trait differs between the sexes (Ralls 1977; Lande 1980; Price 1984), such that
any explanation of male- or female-biased SSD should accommodate not only why one sex is
larger but also why the other sex is smaller (Hedrick & Temeles 1989). Here, we found sexspecific relationships between juvenile mass and survival to reproductive maturity. Among
males, we saw a significant relationship between IGF-1 and relative mass (Table C.2), but
neither juvenile IGF-1 nor juvenile mass predicted survival to reproductive maturity (Table C.4;
Figure 4.2A). In contrast, among female spotted hyenas, maternal rank and higher
concentrations of IGF-1 predicted larger relative juvenile mass (Table C.3), which in turn
predicted greater probability of survival to reproductive maturity (Table C.5; Figure 4.3B).
Previous research on our study population found evidence of positive selection on size
traits in male hyenas as well as in females, although it was considerably stronger on females
than males (Swanson 2013). The author used annual reproductive success as a measure of

73

fitness whereas we focused on survival to reproductive maturity so our results are not in conflict.
Rather, taken together our results present a more complex picture and suggest that size may
confer different advantages before and after reproductive maturity. Currently, we can only
speculate why mass is more important to survival among juvenile females than juvenile males,
but sex-biased survivorship is common in the wild (Schulte-Hostedde et al. 2002; Kalmbach et
al. 2005; Råberg et al. 2005; Bonduriansky et al. 2008; Maklakov et al. 2008). Low probability of
survival among small males might be due to small males losing to larger siblings in food
contests, whereas low probability of survival among large males might be due to the costs of
maintaining a large body size, such as higher nutritional demands (Clutton-Brock, Albon &
Guinness 1985; Arendt 1997; Blanckenhorn 2000; Dmitriew 2011).
Due the opportunistic and sensitive nature of immobilizing hyenas in our study
population, we cannot definitively rule out the contributions of IGF-1 outside the ages of our
analyses. Specifically, the period during which development is more sensitive to environmental
conditions may be sexually dimorphic (Clutton-Brock et al. 1985; Gatford et al. 1998; Lindstrom
1999; Badyaev 2005) so it is possible that the IGF-1 system in female hyenas indeed varies
with environmental conditions, but outside our analysis period, which focuses on the first
postnatal year of life. Similarly, IGF-1 levels and mass among juvenile male hyenas may be
sensitive to environmental variables throughout ontogeny and not limited to the age range
studied here. For example, in contrast to previous findings in a Tanzanian population of spotted
hyenas (Höner et al. 2010), maternal rank did not significantly predict relative male mass during
the first postnatal year of life in our study area (Table C.2). However, the Tanzanian study
measured mass until the age of 6 months whereas the youngest male cub in our study was 7
months old, so the contribution of maternal rank to body mass in earlier periods of development
is certainly possible. In either case, our results provide support for postnatal concentrations of
IGF-1 as physiological mediators of SSD in spotted hyenas, and suggest that neutral selection

74

for male juvenile mass and positive selection for female juvenile mass might be one potential
route by which SSD is maintained in this species.

75

APPENDIX

76

Supplementary Material for Chapter 4
Table C.1. Distribution of samples, excluding 12 samples from 9 pregnant females (mean IGF-1
± s.e. (ng/ml) = 176.89 ± 15.98). With the sexes pooled over the growth period of 1-4 years of
age, females displayed significantly higher IGF-1 concentrations than males (W = 4151, p =
0.0007).

Age
range
0-1
1-2
2-3
3-4
4+

Male mean
IGF-1 ± s.e.
(ng/ml)
247.63 ± 14.58
242.94 ± 16.28
182.10 ± 10.61
122.24 ± 6.78
118.09 ± 4.80

No.
samples,
individuals
48, 48
27, 27
26, 26
15, 14
21, 21

Female mean
IGF-1 ± s.e.
(ng/ml)
302.29 ± 11.97
324.31 ± 20.70
273.99 ± 26.06
198.21 ± 9.57
157.47 ± 10.45

No.
samples,
individuals
68, 68
26, 26
12, 12
16, 16
47, 40

Result of model IGF1 ~ Sex
W = 2092, p = 0.01*
t51 = -3.10, p = 0.003
t36 = -3.75, p < 0.001
t29 = -6.40, p < 0.001
W = 683, p = 0.01*

*IGF-1 concentrations could not be normalized using traditional transformations so we assessed
sex differences using a non-parametric Mann-Whitney-Wilcoxon Test.

Table C.2. Results from a multiple linear regression model estimating influence of socioecological factors on relative juvenile mass among 43 juvenile males (ages 0.5-1.0 years).
Mean mass ± SE = 18.90 ± 0.77 kg. Significant effects are highlighted in boldface at a
significance level of α = 0.05.
Model fit: n = 43, adjusted R2 = 0.39, p < 0.0001
Fixed effect
Parameter ± SE
95% CI
-1.85×10-16 ± 0.4
Intercept
-0.88 − 0.88
Juvenile IGF-1
2.25 ± 0.4
1.36 − 3.17
Maternal rank
-0.78 ± 0.4
-1.70 − 0.12
Prey density
0.62 ± 0.4
-0.30 − 1.48

t
0.00
5.15
-1.76
1.44

df
39
39
39
39

P
1.00
<0.0001
0.09
0.16

All predictor variables were z-transformed to a mean of 0 and standard deviation of 1. Model
residuals did not violate assumptions of normality (Shapiro-Wilk test of normality: W = 0.99, P =
0.86).

77

Table C.3. Results from a multiple linear regression model estimating influence of socioecological factors on relative juvenile mass among 69 juvenile females (ages 0.5-1.0 years).
Mean mass ± SE = 22.88 ± 0.74 kg. Significant effects are highlighted in boldface at a
significance level of α = 0.05.

Fixed effect
Intercept
Juvenile IGF-1
Mat. dom. rank
Prey density

Parameter ±
SE
-7.54 ± 1.90
0.02 ± 0.00
2.67 ± 0.93
0.01 ± 0.01

95% CI
-11.3 − 3.7
0.01 − 0.03
0.82 − 4.53
-0.01 − 0.02

t

df

P

-3.96
3.80
2.89
1.23

65
65
65
65

<0.001
<0.001
0.005
0.224

All predictor variables were z-transformed to a mean of 0 and standard deviation of 1. Model
residuals did not violate assumptions of normality (Shapiro-Wilk test of normality: W = 0.99, P =
0.75).

78

Table C.4. Results from a Cox proportional hazards model predicting male survival to
reproductive maturity. Males of larger relative juvenile mass between 0.5 and 1.0 years did not
survive proportionally better or worse than smaller males. Model 2 was the best-supported
model based on log likelihood with a higher value indicating a better fitting model.
Model 1: n= 43; χ2 =0.63; df = 2, P = 0.73, model log likelihood = -65.85
Predictor variable
β
Hazard ratio
z
Juvenile mass
Maternal rank

0.01
-0.19

1.01
0.83

0.05
-0.79

Model 2: n= 43; χ2 = 3.05; df = 2, P = 0.22, model log likelihood = -64.60
Predictor variable
β
Hazard ratio
z
Juvenile IGF-1
Maternal rank

-0.38
-0.15

0.68
0.86

-1.58
-0.57

P
0.96
0.43

P
0.12
0.57

All predictor variables were z-transformed to a mean of 0 and standard deviation of 1. A
negative hazard rate coefficient (β) indicates a decreased probability of survival with increasing
value of the predictor variable. A positive hazard rate coefficient (β) indicates an increased
probability of survival with increasing value of the predictor variable. A hazard ratio (calculated
as exp(β)) of 2, for example, indicates that individuals are twice as likely to survive with
increasing values of the predictor variable.

79

Table C.5. Results from two separate Cox proportional hazards models predicting female
survival to reproductive maturity. Heavier juvenile female hyenas experienced significantly
greater survival to reproductive maturity than did smaller females. Model 2 was the bestsupported model based on likelihood ratio tests. One individual had no mass recorded so was
not included in Model 2, in which mass was a predictor variable; however, this individual did
have an IGF-1 value so was included in Model 1. Significant effects are highlighted in boldface
at a significance level of α = 0.05.
Model 1: n= 62; χ2 = 3.29; df = 2, P = 0.19, model log likelihood = -88.59
Predictor variable
β
Hazard ratio
z
Juvenile IGF-1
Mat. dom. rank

-0.32
-0.40

0.72
0.67

-1.46
-1.06

Model 2: n= 61; χ2 = 6.20; df = 2, P = 0.45, model log likelihood = -86.67
Predictor variable
β
Hazard ratio
z
Juvenile mass
Mat. dom. rank

-0.11
-0.07

0.90
0.93

-2.27
-0.17

P
0.145
0.288

P
0.023
0.868

All predictor variables were z-transformed to a mean of 0 and standard deviation of 1. A
negative hazard rate coefficient (β) indicates a decreased probability of survival with increasing
value of the predictor variable. A hazard ratio (calculated as exp(β)) of 2, for example, indicates
that individuals are twice as likely to survive with increasing values of the predictor variable.

80

300

IGF-1 (ng/ml)

200

100

0
Pregnant

Lactating

Reproductive status

Figure C.1. Plasma IGF-1 concentrations were significantly higher in females while pregnant
(solid circles) than while lactating (open circles). Two-tailed paired t-test: t8 = 4.53, P = 0.002.
Lines represent individual female subjects (n = 9 females).

81

CHAPTER 5: AGE-RELATED REPRODUCTIVE PERFORMANCE VARIES WITH
DOMINANCE RANK AMONG FEMALE SPOTTED HYENAS

82

Introduction
As detailed data accumulate from long-term, individual-based field studies on
iteroparous species, increasingly more studies are documenting, not only how reproductive
performance improves over the lifespan, but also how it declines among the oldest animals in
natural populations (Bronikowski & Promislow 2005; Robbins et al. 2006; McCleery et al. 2008;
Nussey et al. 2008a, 2013; Reed et al. 2008; Dugdale et al. 2011; Roach & Carey 2014;
Hammers et al. 2015; Lee et al. 2016). Evolutionary theories of aging were originally based
upon data from experimental laboratory studies (Stearns 1992; Bronikowski & Promislow 2005),
so the influx of empirical data offers the exciting opportunity to complement this groundbreaking
work with a deeper understanding of how age-related reproductive performance has evolved in
response to, or in spite of, environmental variation (Williams et al. 2006).
Currently, two non-mutually exclusive hypotheses have been offered to explain agerelated improvement and decline in reproductive performance. Early in life, age-related
improvement in reproductive performance may best be explained by the constraint hypothesis
(Forslund & Pärt 1995). Under this hypothesis, reproductive fitness should increase as
developmental constraints (e.g. physiological immaturity, lack of experience rearing young) are
removed, allowing for improvement in traits important for offspring survival. For example, many
female mammals begin reproducing before reaching full adult body size, which creates an
energetic conflict between the needs for maternal growth and offspring survival (Trivers 1972;
Stearns 1989). However, as a female matures and the energy required for her growth
decreases, she is free to allocate more resources toward the survival of her offspring, thus
increasing her realized reproductive success. The constraint hypothesis may also apply to preand postnatal experience rearing young. Inexperienced mothers may not know how to properly
provision or attend to offspring, particularly in species lacking communal rearing of offspring
(Wang & Novak 1994; Cameron et al. 2000).

83

Late-life declines in reproductive performance are often explained by the senescence
hypothesis, which predicts age-related declines in reproductive performance (Medawar 1952;
Hamilton 1966; Charlesworth 1993). Under the senescence hypothesis, declines in reproductive
performance are the realized physiological consequences of investing energy resources into
reproduction rather than into self-maintenance earlier in life. To date, two non-mutually
exclusive hypotheses have been proposed to explain declines in reproductive performance. The
disposable soma theory explains the declines in reproductive performance as the erosion of
essential physiological functions by the lifelong accumulation of cellular damage (Kirkwood &
Holliday 1979), whereas the theory of antagonistic pleiotropy stipulates that the same genes
associated with fitness benefits early in life are linked to negative consequences (i.e.
senescence) later in life (Williams 1957). Regardless of which mechanism is operating, both
generate similar predictions for the optimal allocation of limited energy, with trade-offs being a
central feature (Kirkwood & Rose 1991).
Competition among conspecifics may influence the reproductive patterns exhibited
within a population by restricting individual access to food resources and mating opportunities
(Clutton-Brock 1988; Holekamp et al. 1996; Altmann & Alberts 2003). In many animal societies
structured by rigid linear hierarchies, individuals occupying the top dominance ranks enjoy
priority of access to both, and, as a consequence, also enjoy superior reproductive performance
(Clutton-Brock, Albon & Guinness 1986; Holekamp et al. 1996; Creel et al. 1997; Pusey,
Williams & Goodall 1997; Hofer & East 2003; Sharp & Clutton-Brock 2011). Despite the ubiquity
of this relationship, relatively few studies have explored the degree to which rank-related
performance is attributable to earlier improvements and/or delayed senescence (Holekamp et
al. 1996; Wasser et al. 1998; Côté & Festa-Bianchet 2001; Von Holst et al. 2002; Watts et al.
2009; Sharp & Clutton-Brock 2010; Hoffman et al. 2010).
Here, we use 27 years of individual-based, longitudinal data from a population of spotted
hyenas (Crocuta crocuta) to test for constraint and senescence in several measures of

84

reproductive performance. Unlike other social mammals in which dominant individuals
monopolize breeding (Clutton-Brock & Huchard 2013), all female hyenas are free to breed yearround with equal access to mates (Kruuk 1972), making them a unique system in which to study
age- and social rank-related variation in reproduction. Previous research in this population
shows that high-ranking females start reproducing at younger ages (Holekamp et al. 1996) and
so should be expected to show earlier improvements in reproductive performance than would
low-ranking females (Watts et al. 2009). However, it remains unknown whether earlier
reproductive improvements are accompanied by earlier reproductive declines, and, if so, how
this might vary with social rank across the lifespan.
Materials and Methods
Our subjects were drawn from a long-term field study of a single large social group of
spotted hyenas, the Talek clan, which has been monitored continuously since May 1988 within
the Masai Mara National Reserve, Kenya. Clan size has varied over time (Holekamp et al.
2012) ranging from around 42 to 126 individuals (Green 2015). The home territory occupied by
this clan is primarily open, rolling grassland with permanent water and year-round prey
availability. All hyenas in the study clan were identified by their unique spot patterns, and sex
was identified based on the morphology of the glans of the erect phallus (Frank et al. 1990).
Demographic and behavioral data were collected as part of year-round, routine observations of
the Michigan State University (MSU) Hyena Project started in 1988 by Dr. Kay E. Holekamp and
Dr. Laura Smale.
The gestation period in the spotted hyena is 110 days (Kruuk 1972). Females typically
give birth to small litters containing only one or two cubs, which they nurse for up to 24 months
(Holekamp et al. 1996). A female typically bears her litter in a solitary natal den before
transferring the cubs to a communal den used by other clan members when her cubs are 2-5
weeks old. Birthdates of cubs were estimated to ± 7 days based on their pelage, size, and other
aspects of their behavior and appearance when first seen above ground (Holekamp et al. 1996).

85

We determined age at first parturition based on the initial tearing of the female’s pseudopenis,
indicated by the presentation of pink scar tissue on its posterior surface (Frank & Glickman
1994). This tearing reliably indicates that first parturition has occurred regardless of whether or
not the first litter born is ever seen above ground by observers.
Dominance rank matrices were constructed based on outcomes of dyadic agonistic
interactions (Tilson & Hamilton 1984; Frank 1986). We standardized rank to account for
variation in clan size; an individual with a relative rank of 1 was the highest-ranking animal in the
clan, and the individual with a relative rank of -1 was the lowest ranking. High-ranking females
were considered those with relative ranks of 1 to 0.33, mid-ranking females were those with
relative ranks of 0.33 - 0.33, and low-ranking females were those with relative ranks of -0.33 to 1. Each female was assigned a rank in the adult female hierarchy upon the birth of her first litter
or her third birthday, whichever came first.
Female age at last parturition and lifespan
A female’s age at last parturition was calculated as her age when she gave birth to her
most recent litter if she was still alive, or last recorded litter if she was deceased. Because
female spotted hyenas are philopatric (Smale et al 1997; Boydston et al 2005), and because we
were observing our study clans daily, we recorded each female’s date last seen as her date of
death.
Lifetime reproductive success (LRS)
We calculated lifetime reproductive success as the total number of cubs a female
successfully weaned during her lifespan. We only calculated LRS of females that gave birth to
at least one litter and had full reproductive histories with known death dates.
Age of cub at weaning
We estimated the age at weaning to within 10 days using observations of weaning
conflicts between a mother and her cubs (Holekamp et al. 1996). A singleton litter was defined

86

as a litter from which only one cub successfully weaned, regardless of whether neonatal litter
size was one or two; both cubs were successfully weaned in twin litters.
Inter-birth interval (IBI)
We calculated the inter-birth interval as the number of months between consecutive
parturitions when at least one cub survived to weaning (singleton) or both cubs survived to
weaning (twin litter). If one member of a twin litter died before 6 months of age, but the
remaining cub survived to weaning, then we counted the litter as a singleton litter.
Cub survival
In our study population, males typically disperse between 2 and 6 years of age (Smale et
al. 1997; Boydston et al. 2005), but females are philopatric. Natal males were considered to
have dispersed only if they had undertaken exploratory forays outside the natal territory, were in
good health when last seen, and were last seen in the natal territory after 2 years of age (Van
Horn et al. 2003). Males were considered to have died rather than dispersed if they disappeared
before they reached 2 years of age. We observed our study clans daily so we recorded a male’s
date last seen as its date of death unless it was known to have dispersed.
Statistical analyses
We performed all analyses with R statistical software (Version 3.3.2; R Core Team
2016). For linear mixed effects models (LMM), we used the ‘lme4’ (Bates et al. 2014) package
with p-values estimated by Satterthwaite’s approximation for denominator degrees of freedom
using the ‘lmerTest’ package (Kuznetsova, Brockhoff & Christensen 2015). For survival
analyses, we used the mixed effects Cox model in the ‘coxme’ package (Therneau 2015).
LMMs offer a powerful statistical framework to separate between- and within-individual
variation in traits (van de Pol & Verhulst 2006; Nussey et al. 2008a, 2013). Here, we included
female identity (ID) and litter birth year as random effects to control for individual variation in
reproductive performance and year-to-year environmental variation, respectively (Nussey et al.
2008a). All models (LMM, multiple linear regression, and mixed effects Cox) were initially fitted

87

with a full model of four fixed effects: female rank, female age, an interaction term (female rank
× age) to accommodate potential age variation in rank-related reproduction, and (female age)2
to accommodate a potential nonlinear relationship between reproduction and age (i.e. both
improvement and decline in performance). When the interaction term or quadratic term was
insignificant, we removed it as a fixed effect in a reverse stepwise model simplification method
leaving only the linear terms of female rank and age. We then re-ran the model using restricted
maximum-likelihood methods (Bolker et al. 2009). Residuals for all LMM and multiple linear
regressions were normally distributed based on the Shapiro-Wilk test of normality (W ≥ 0.99, p ≥
0.35 ; Burnham & Anderson 2002).
To analyze LRS, we performed a multiple linear regression. To analyze age at weaning
and inter-birth interval, we performed LMMs with Gaussian distributions.
Results
Generally, female hyenas of all ranks reproduced close to the end of their lifespans and
stopped producing young at around 17 years of age (Figure 5.1). At least eight of 93 deceased
female hyenas in our study population have lived past 16 years of age with one individual living
until 23 years of age (Figure 5.1), but most females live much shorter lifespans in the wild. Mean
± s.e. lifespan among deceased females was 8.25 ± 0.72 years for high-ranking females; 8.72 ±
0.78 years for mid-ranking females; and 8.81 ± 0.89 years for low-ranking females.
During our 27 years of data collection, high-ranking females produced the greatest
number of litters with at least one cub surviving to weaning (Figure 5.2, Table 5.1). High-ranking
females appeared to benefit from high early performance, with over 50% of 113 weaned litters
born to mothers aged 4 years or younger, and peak performance at 3 years of age (Figure 5.2,
Table 5.1). Mid-and low-ranking females also experienced peaks of performance, but at 4 and 5
years of age, respectively, likely owing to older ages at first parturition (Holekamp et al. 1996).

88

Number of litters with at least one cub surviving to weaning

22
20
High rank (113 litters, 35 females)
18

Mid rank (61 litters, 24 females)
Low rank (79 litters, 24 females)

16
14
12
10
8
6
4
2
0
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

Female age at litter birth (years)

Figure 5.1. The number of litters with at least one cub surviving to weaning varied with binned
maternal age (rounded down to the nearest integer value) and maternal rank (n =253 litters born
to 83 unique female hyenas across 27 years). The total number of cubs that survived born to
high-ranking females was 136, with 74 born to mid-ranking females, and 71 born to low-ranking
females. Female rank was analyzed as a continuous variable, but is shown here as a
categorical variable for illustrative purposes only. For example, the sole female in the 1-year
category gave birth when she was 1.88 years.

89

Lifetime reproductive success (LRS)
We analyzed LRS for 48 deceased females, all of which were adults but 12 of which
never successfully weaned any offspring (1 high-ranking, 5 mid-ranking, and 6 low-ranking
females). Mean ± s.e. LRS among all females was 3.24 ± 0.52 cubs. Mean ± s.e. LRS was 4.00
± 0.85 cubs for high-ranking females, 2.53 ± 0.85 cubs for mid-ranking females, and 2.93 ± 1.12
cubs for low-ranking females. Higher female rank and longer lifespans significantly predicted
greater female LRS (rank: t46 = 3.38, P ≤ 0.001; lifespan: t46 = -10.04, P ≤ 0.001; Figure 5.3,
Table 5.3).

90

13
12
11
10

LRS

9
8
7
6
5
4
3
2
1
0
2

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Female lifespan (years)

Figure 5.2. Lifetime reproductive success (LRS) of female spotted hyenas that had given birth
to at least one litter and that had full reproductive histories with known death dates (high-rank: n
= 17, mid-rank: n = 15, low-rank: n = 14). LRS was measured as the total number of cubs
weaned; both male and female offspring are included. Shading denotes binned female rank
(black = high, gray = mid, white = low), which was statistically analyzed as a continuous variable
but shown here as a categorical variable for illustrative purposes only.

91

Age of cub at weaning - singleton
We analyzed cub age at weaning for 125 singleton cubs born to 60 females across 27
years. Mean ± s.e. cub age at weaning among high-ranking females was 10.51 ± 0.43 months,
12.82 ± 0.52 months for mid-ranking females, and 11.86 ± 0.47 months for low-ranking females.
Female ID, but not cub birth year, explained a significant portion of the residual variation (Χ2 =
13.05, P <0.001; Table 5.2), suggesting individual differences in mothering styles among
breeding females. Of the fixed effects, neither (female rank × age) nor (female age)2 was
significant so both variables were eliminated from the final model. Neither female rank (t54.0 = 1.47 P = 0.15) nor female age (t120.6 = 1.63 P = 0.11) significantly predicted cub age at weaning
(Table 5.3).
Age of cub at weaning - twin litter
We analyzed cub age at weaning for 76 twin litters cubs born to 40 females across 24
years. Mean ± s.e. cub age at weaning was 11.45 ± 0.38 months for high-ranking females,
11.99 ± 0.74 months for mid-ranking females, and 13.46 ± 0.73 months for low-ranking females.
As was true for weaning age among singletons, female ID, but not cub birth year, explained a
significant portion of the residual variation (Χ2 = 7.59, P = 0.006; Table 5.2). (Female age)2 was
not significant so was eliminated from the final model. Younger ages at weaning were
significantly predicted by higher female rank (t73.6 = -2.30, P = 0.02) and (female rank × age)
(t73.6 = 2.16, P = 0.03; Table 5.3), but not by age per se.
Inter-birth Interval (IBI) after singleton litter
We analyzed the IBI following the successful weaning of 73 singleton litters born to 40
females across 25 years. Mean ± s.e. IBI was 14.98 ± 0.69 months for high-ranking females,
18.19 ± 0.62 months for mid-ranking females, and 17.79 ± 0.84 months for low-ranking females.
Neither random effect explained a significant portion of the residual variation (Χ2 ≤ 0.64, P ≥
0.11; Table 5.2). Of the fixed effects, neither (female rank × age) nor (female age)2 was
significant, so both variables were eliminated from the final model. IBI was significantly

92

predicted by female rank (t35.2 = -3.44, P = 0.002), but not by female age (t69.3 = 0.32, P = 0.75;
Figure 5.4A, Table 5.3) with high-ranking females exhibiting the shortest IBIs.
Inter-birth Interval (IBI) after twin litter
We analyzed the IBI following the successful weaning of 53 twin litters born to 28
females across 22 years. Mean ± s.e. inter-birth interval was 13.87 ± 0.72 months for highranking females, 18.64 ± 0.90 months for mid-ranking females; and 17.26 ± 1.31 months for
low-ranking females. Neither random effect explained a significant portion of the residual
variation (Χ2 ≤ 0.70, P ≥ 0.40; Table 5.2). Of the fixed effects, neither (female rank × age) nor
(female age)2 was significant, so both variables were eliminated from the final model. Older
females showed significantly longer IBIs (t46.6 = 3.59, P = 0.0008; Figure 5.4B, Table 5.3), but
we saw no significant effect of female rank (t28.7= -0.35, P = 0.73).

93

A

B
30

Interbith interval following weaning of twin litter (months)

Interbith interval following weaning of singleton (months)

30

25

20

15

25

20

15

10

10
2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Female age (years)

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Female age (years)

Figure 5.3. (A) Inter-birth intervals following successful weaning of singleton litters (n=72 litters
born to 39 females; mean inter-birth interval ± s.e. for high-ranking females: 14.88 ± 0.69
months, for mid-ranking females: 18.59 ± 0.62 months, and for low-ranking females: 17.79 ±
0.84 months). (B) Inter-birth intervals following successful weaning of twin litters (n=53 litters
born to 28 females). Shading denotes female rank (black = high, gray = mid, white = low).
Female rank was statistically analyzed as a continuous variable but shown here as a categorical
variable for illustrative purposes only.

94

Cub survival
We analyzed cub survival to 2 years of age among 670 cubs born to 112 females across
27 years. Mean ± s.e. number of cubs that survived was 3.32 ± 0.45 cubs born to high-ranking
females, 2.39 ± 0.34 cubs born to mid-ranking females, and 2.73 ± 0.30 cubs born to lowranking females. Total number of cubs that survived born to high-ranking females: 136 cubs, to
mid-ranking females: 74 cubs, and to low-ranking females: 71 cubs. The random effect of cub
birth year explained a significant portion of the residual variation (Χ2 = 26.76, P < 0.001; Table
5.2) whereas female ID did not (Χ2 = 2.88, P = 0.09; Table 2). Greater cub survival was
significantly predicted by higher female rank (z = -3.14, P = 0.002) and older female age (z = 2.07, P = 0.04), and by (female rank × age) (z = 2.19, P = 0.03), and (female age)2 (z = 2.25, P
= 0.03; Table 5.4), suggesting both age-related improvement and decline in performance.
Discussion
In rank-structured societies, an individual’s rank may have an enormous effect on its
ability to acquire resources and, consequently, on its fitness. Superior priority of access to food
and mating opportunities generally affords high-ranking individuals high reproductive
performance, but whether this competitive edge persists over the entire lifespan is less well
understood. Here, we report both significant rank- and age-related improvement and decline in
reproductive performance among female spotted hyenas occupying all ranks positions within
the social hierarchy, and ranging in age from 3 to 23 years. Our findings suggest that
reproductive senescence does occur in this species and that, in general, high-ranking females
senesce earlier than do lower-ranking females. Interestingly, however, this does not prevent
high-ranking females from out-performing lower-ranking females, as their LRS was nevertheless
25% greater than that of mid- or low-ranked females.
The constraint hypothesis (Forslund & Pärt 1995) predicts age-related improvement in
reproductive performance at the beginning of an individual’s reproductive lifespan. Across the
study period, we saw that females of all dominance ranks experienced age-related increases in

95

the frequency and weaning success rate of litters containing at least one cub that survived to
weaning (Figure 5.2, Table 5.1). Higher-ranking females start breeding at earlier ages
(Holekamp et al. 1996), and, accordingly, we saw that they enjoyed high weaning success
earlier in life with the largest number of weaned litters born to 3-year old females (Figure 5.2;
Table 5.1). Mid- and lower-ranking females weaned far fewer litters early in life with the largest
numbers of weaned litters born to 4- and 5-year old females, respectively (Figure 5.2; Table
5.1). Low-ranking females showed the greatest improvement with age; success rates doubled
from 30% at age 3 to 60% at age 4 and peaked around 79% at age 7 (Table 5.1). However, the
selective disappearance of low-performing individuals may give the false impression of
improvement (Forslund & Pärt 1995; van de Pol & Verhulst 2006), and the reduction in the
number of reproducing low-ranking females from ages 3 to 4 and from ages 6 to 7 (Table 5.1)
certainly makes this plausible. Nevertheless, our results show that at least some low-ranking
females are capable of high reproductive success. Additionally, the reproductive peaks for each
rank group suggest different time points at which females may break free from developmental
constraints, such as negative energy balance (Zehr, Van Meter & Wallen 2005) or immature
levels of ovarian hormones (Van Meter et al. 2008). Future research coupling energetic and
endocrine measures with behavioral data could illuminate not only the mechanisms of
developmental constraint but also potentially those of reproductive senescence (i.e.
menopause).
We found evidence of both reproductive improvement and senescence with advancing
female age in only one measure: cub survival (Table 5.4). Our results mirror patterns found in
some other mammals and birds (Sharp & Clutton-Brock 2010; Stahler et al. 2013; Hammers et
al. 2015), suggesting that female hyenas follow similar age-related trends in offspring survival to
those found in other taxa. Although it remains unclear how this is directly mediated, our finding
of significantly longer IBI with advancing female age following successful weaning of twin litters
may offer some indirect clues. Specifically, the linear trend of increasing IBI with female age

96

(Figure 5.4B; Table 5.3) coupled with the quadratic relationship of cub survival with female age
(Table 5.4) suggests that cub survival may benefit from prolonged IBI, but only to a certain
point. At younger ages, longer IBIs may indicate greater maternal investment (Paul, Kuester &
Podzuweit 1993; Hoffman et al. 2010), but, at older ages, a longer IBI may indicate senescence
in reproductive resilience, energetic condition, or fertility (Altmann et al. 2010; Nenko &
Jasienska 2013). We only calculated IBI for females whose previous litters were successfully
weaned so our results could be biased towards older, successful females that were able to both
wean twin litters and live longer by virtue of elongating their IBI. However, female ID did not
explain a significant portion of the residual variation here (Table 5.2), indicating that all females
experienced this age-related increase in IBI following successful weaning of twin litters.
Although we saw a significant effect of maternal age following the successful weaning of
twin litters, we did not see the same relationship when we looked at weaning of singletons.
Here, instead of maternal age, maternal rank significantly predicted IBI following the successful
weaning of singletons (Figure 5.4A; Table 5.3), which is consistent with earlier findings in this
population (Holekamp et al. 1996). Our observation of age-related increases in IBI, but only
after weaning twin litters, suggests that there may be larger trade-off between self-maintenance
and reproduction as females move closer to senescence. Spotted hyenas are estimated to have
the highest energy output during lactation among terrestrial carnivores (Oftedal & Gittleman
1989) so older females may be undergoing a more severe trade-off between self-maintenance
and reproduction where the latter is prioritized. Given the enormous energetic toll of lactation, it
may be no surprise that we observed longer recovery times in older females following
successful weaning of twin litters but not singletons.
Cubs born to high-ranking females were weaned at earlier ages than were cubs born to
low-ranking females (Table 5.3), but the significant interaction of (female rank × age) reported in
this study suggests that this relationship varies with maternal age (Table 5.3). Specifically, the
age at weaning for cubs of high-ranking females increased with maternal age thereby reducing

97

the disparity among ranks. Despite having an under-developed feeding apparatus, cubs of highranking females are able to wean at younger ages largely due to their mothers’ aggressive
displacement of conspecifics from food (Engh et al. 2000; Watts et al. 2009). Therefore, a
prolonged lactation period among cubs of older high-ranking females may indicate senescence
in the mother’s ability to competitively secure resources for her cub. Because maternal support
exerts an enormous effect on cub survival (Watts et al. 2009) it may be no coincidence that we
also saw a significant interaction effect of between female rank and age on cub survival (Table
5.4).
Although the onset of reproductive senescence varied with dominance rank, all females
ceased reproduction after 17 years (Figure 5.1, Table 5.1), suggesting a terminal point of
reproduction in this species. However, we did observe a small percentage of female hyenas
outliving their ability to reproduce by up to 7 years (Figure 5.1). These females varied in
dominance status and age at last parturition, suggesting that the mechanisms contributing to
post-reproductive lifespans may not be dictated by dominance rank. The occurrence of females
with post-reproductive lifespans is rare among other mammalian species (Cohen 2004) and
may be a random add-on at the end of the life history (Reznick, Bryant & Holmes 2006).
Nevertheless, in killer whales (Foster et al. 2012) and elephants (Lee et al. 2016), the presence
of post-reproductive females in kin groups directly contributes to offspring survival, potentially
through leadership, caring for descendant kin, and the dissemination of knowledge regarding
food resources (Brent et al. 2015) and threats from predators (McComb et al. 2011),
respectively. It remains to be explored in spotted hyenas whether these post-reproductive
females also contribute to their descendants’ survival, or whether their long lives merely allow
them to persist in the population without making contributions to the welfare of their
descendants.
Although some high-ranking females live long lives, most need not have long lifespans
to achieve LRS superior to that of subordinate females. For example, in order to achieve the

98

same LRS as that of high-ranking females, it may take some mid- and low-ranking females
twice as many years (Figure 5.3). Despite the low weaning rate among mid- and low-ranking
females, some females do achieve LRS that is similar to or greater than that of high-ranking
females by virtue of living longer (Figure 5.3), thus highlighting the relative importance of longer
lifespans to the fitness of lower-ranking females. Long lifespans have the potential to drive
population dynamics in wild bats (Schorcht, Bontadina & Schaub 2009) and Soay sheep
(Coulson et al. 2006), so exactly how they may operate in rank-structured societies with multiple
matrilines is worthy of future research.
Our mixed effects modeling approach revealed significant individual variation among
females with respect to the age at which their cubs weaned, but not for any other reproductive
traits (Table 5.2). Although weaning is typically around 13 months of age in our study population
of spotted hyenas (Holekamp et al 1996), it can range from 7.5 to 24 months and appears to be
largely determined by maternal behavior and food availability (Holekamp et al. 1996) rather than
achievement of a fully mature feeding apparatus (Watts et al. 2009). The degree of individual
variation may depend on the life history of the study species (van de Pol & Verhulst 2006).
Here, we attribute the lack of significant inter-individual variation among the other traits to the
large amount of variation explained by female rank and age in this species. In other social
species where dominance rank is more subject to change, we might expect more individual
heterogeneity in reproductive performance particularly in traits that may facilitate both
reproductive success and dominance acquisition (e.g. aggressive behavior, individual physical
quality).
Life-history theory predicts that individuals that start reproducing early and exhibit high
rates of reproduction should also experience early senescence (Williams 1957; Kirkwood &
Rose 1991). Like alpha individuals in other hierarchical societies (Clutton-Brock et al. 1986;
Creel et al. 1997; Pusey et al. 1997; Young, Carlson & Monfort 2006), alpha female hyenas
show uniquely high reproductive performance within a clan (Holekamp et al. 1996). However,

99

contrary to life-history theory, alpha females appear to resemble “Darwinian demons” (Law
1979), which manage to escape fitness trade-offs; that is, alpha female hyenas are able to enjoy
both early, high reproductive success and long lifespans. The average female hyena weans 3
cubs over an 8-year lifespan, but we recorded one alpha female weaning 13 cubs over a 15year (Figure 5.3) and another weaning at least 13 cubs over a 16-year lifespans (data not
shown; female was born before 1988 so we only have a partial reproductive history). It is likely
that the high resource availability (Kruuk 1972; Tilson & Hamilton 1984), low markers of
physiological stress (Lewin et al. 2015), and high immunity (Flies et al. 2016) afforded to alpha
females allow them to escape the early senescence predicted by theory. Alternatively, there
may be costs incurred by high-ranking females that we have yet to identify.

100

APPENDIX

101

Supplementary Material for Chapter 5

Table D.1. Summary table of 426 litters born to 113 unique female hyenas using over 27 years
of longitudinal data. The % of total litters born with at least one cub surviving to weaning is
rounded to the nearest integer value. This % was only calculated for age bins where ≥ 5 litters
were recorded. Maternal rank was statistically analyzed as a continuous variable but shown
here as a categorical variable for illustrative purposes only.

High
Femal
e age
at litter
birth
(years)

#
of
litt
ers
bor
n

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Total:

1
25
31
26
22
18
11
12
10
5
4
4
3
4
1
1
0
17
8
10.
47

Mean:

# of
litters
with >1
cub
surviving
to
weaning
(% of
total)
1
18 (72%)
22 (71%)
17 (64%)
12 (55%)
12 (67%)
5 (45%)
9 (75%)
5 (56%)
3 (60%)
2
3
2
1
1
0
0
113
(63%)
6.64

# of
reproducing
female
s

1
25
26
23
20
17
10
11
9
5
4
4
3
1
1
1
0

9.47

Female social rank
Mid
# of
# of
# of
litters litters
reproborn with >1
ducing
cub
female
surviving s
to
weaning
(% of
total)
0
0
0
6
4 (67%)
6
20
5 (25%)
18
19
11 (58%)
18
10
7 (70%)
10
10
6 (60%)
9
12
8 (67%)
12
6
4 (67%)
6
7
4 (57%)
6
5
2 (40%)
5
6
4 (67%)
6
2
1
2
5
2
4
1
0
1
3
3
3
0
0
0
1
0
1
113 61 (54%)
6.65

4.12

102

6.29

0
3
23
15
18
17
14
8
10
8
5
5
2
4
1
2
0
135

Low
# of
litters
with >1
cub
surviving
to
weaning
(% of
total)
0
1
7 (30%)
9 (60%)
13 (72%)
10 (59%)
11 (79%)
5 (63%)
7 (70%)
6 (75%)
1 (20%)
3 (60%)
1
3
1
1
0
79 (59%)

7.94

4.64

# of
litters
born

# of
reproducing
female
s

0
3
22
15
17
16
13
8
10
8
5
5
2
4
1
2
0

7.71

Table D.2. Effects of the random effects of female identity (ID) and offspring birth year in five
separate linear mixed-effects models measuring reproductive performance. Random effects that
explained a significant portion of the residual variance in each model are highlighted in boldface
at a significance level of α = 0.05.

Trait

Random effect

Variance

Χ2

P

Singleton cub age at weaning
125 cubs born to 60 females
across 27 years

Female ID
Cub birth year

3.70
0.87

13.05
3.55

<0.001
0.06

Twin cub age at weaning
76 cubs born to 40 females
across 24 years

Female ID
Cub birth year

2.65
1.55

7.59
2.15

0.006
0.14

Inter-birth interval after
singleton
70 litters born to 38 females
across 25 years

Female ID
Cub birth year

2.60
1.00

1.27
0.32

0.26
0.57

Inter-birth interval after twin
53 litters born to 28 females
across 22 years

Female ID
Cub birth year

1.05
1.77

0.17
0.70

0.68
0.40

Cub survival to 2 years
690 cubs born to 126 females
across 33 years

Female ID
Cub birth year

0.06
0.21

2.88
26.76

0.09
<0.001

103

Table D.3. Results from models that best predicted reproductive performance in 5 traits.
Significant effects are highlighted in boldface at a significance level of α = 0.05.
Trait
Lifetime
reproductive
success

Variable
Intercept
Female rank
Female lifespan

Estimate
-2.39
1.70
0.69

SE
0.62
0.50
0.70

df
46
46
46

t
-3.85
3.38
10.04

P
<0.001
<0.001
<0.001

Age of cub at
weaning singleton

Intercept
Female rank
Female age

10.87
-0.80
0.14

0.66
0.54
0.08

116.88
54.04
120.61

16.50
-1.47
1.63

<0.0001
0.15
0.11

Age of cub at
weaning - twin
litter

Intercept
Female rank
Female age
Female age × rank

13.41
-3.25
-0.09
0.36

1.12
1.41
0.13
0.17

69.71
73.63
74.41
71.26

11.96
-2.30
-0.66
2.16

<0.0001
0.02
0.51
0.03

Inter-birth interval
after weaning
singleton

Intercept
Female rank
Female age

16.75
-2.81
0.05

1.30
0.82
0.15

67.66
35.21
69.25

12.86
-3.44
0.32

<0.0001
0.002
0.746

Inter-birth interval
after weaning twin
litter

Intercept
Female rank
Female age

10.38
-0.36
0.85

1.91
1.03
0.24

41.98
28.70
46.61

5.45
-0.35
3.59

<0.0001
0.73
0.0008

Table D.4. Results from the best model predicting cub survival to 2 years. Significant effects are
highlighted in boldface at a significance level of α = 0.05.

Trait

Variable

Cub survival to 2
years

Female rank
Female age
Female age × rank
Female age2

β

Hazard ratio

z

-0.66
-0.17
0.06
0.01

0.51
0.85
1.06
1.01

-3.14
-2.07
2.19
2.25

P
0.002
0.04
0.03
0.03

A negative hazard rate coefficient (β) indicates a decreased probability of survival with
increasing value of the predictor variable. A hazard ratio (calculated as exp(β)) of 2, for
example, indicates that individuals are twice as likely to survive with increasing values of the
predictor variable.

104

A

B
23

23

22

22

21

21

20

20

19

19

18

18

17

17

Female lifespan (years)

Female age (years)

16
15
14
13
12
11
10
9
8

16
15
14
13
12
11
10
9
8

7

7

6

6

5

5

4

4

3

3

2

2
2

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

2

Female age at most recent parturition (years)

3

4

5

6

7

8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Female age at last parturition (years)

Figure D.1. Female age at the birth date of her last known litter regardless of whether or not it
was successfully weaned. (A) Females still alive by the end of the study period (high-rank: n =
14, mid-rank: n = 8, low-rank: n = 10), with age at the end of the study period on the y-axis and
age at most recent parturition on the x-axis. (B) Deceased females with known death dates that
had given birth to at least one litter (high-rank: n = 34, mid-rank: n = 29, low-rank: n = 30), with
lifespan on the y-axis and age at last recorded parturition on the x-axis. Shading denotes binned
female rank (black = high, gray = mid, white = low). A point lying on the solid reference line
(intercept = 0, slope = 1) indicates that the female died the same year as she last gave birth.

105

LITERATURE CITED

106

LITERATURE CITED

Altmann, J. & Alberts, S.C. (2003) Variability in reproductive success viewed from a life-history
perspective in baboons. American Journal of Human Biology, 15, 401–409.
Altmann, J. & Alberts, S.C. (2005) Growth rates in a wild primate population: ecological
influences and maternal effects. Behavioral Ecology and Sociobiology, 57, 490–501.
Altmann, J., Gesquiere, L., Galbany, J., Onyango, P.O. & Alberts, S.C. (2010) Life history
context of reproductive aging in a wild primate model. Annals of the New York Academy of
Sciences, 1204, 127–138.
Andersson, M. (1994) Sexual Selection. Princeton University Press, Princeton, NJ.
Angelier, F., Vleck, C.M., Holberton, R.L. & Marra, P.P. (2013) Telomere length, non-breeding
habitat and return rate in male American redstarts (ed J Blount). Functional Ecology, 27,
342–350.
Arendt, J.D. (1997) Adaptive intrinsic growth rates: an integration across taxa. Quarterly Review
of Biology, 72, 149–177.
Badyaev, A. V. (2002) Growing apart: an ontogenetic perspective on the evolution of sexual size
dimorphism. Trends in Ecology & Evolution, 17, 369–378.
Badyaev, A. V. (2005) Maternal Inheritance and Rapid Evolution of Sexual Size Dimorphism :
Passive Effects or Active Strategies? American Naturalist, 166, S17-30.
Badyaev, A. V., Whittingham, L.A. & Hill, G.E. (2001) The evolution of sexual size dimorphism in
the house finch. III. Developmental basis. Evolution, 55, 176.
Baldini, S., Restani, L., Baroncelli, L., Coltelli, M., Franco, R., Cenni, M.C., Maffei, L. & Berardi,
N. (2013) Enriched early life experiences reduce adult anxiety-like behavior in rats: a role
for insulin-like growth factor 1. Journal of Neuroscience, 33, 11715–11723.
Bartke, A., Sun, L.Y. & Longo, V. (2013) Somatotropic signaling: trade-offs between growth,
reproductive development, and longevity. Physiological Reviews, 93, 571–98.
Bates, D., Maechler, M., Bolker, B. & Walker, S. (2014) lme4: Linear mixed-effects models using
Eigen and S4. R package version 1.1-12.
Bateson, P. (1979) How do sensitive periods arise and what are they for? Animal Behaviour, 27,
470–486.
Benhaiem, S., Hofer, H., Kramer-Schadt, S., Brunner, E. & East, M.L. (2012) Sibling rivalry:
training effects, emergence of dominance and incomplete control. Proceedings of the
Royal Society B: Biological Sciences, 279, 3727–3735.
Bernstein, R.M. (2010) The big and small of it: How body size evolves. American Journal of
Physical Anthropology, 143, 46–62.

107

Bernstein, R.M., Setchell, J.M., Verrier, D. & Knapp, L.A. (2012) Maternal effects and the
endocrine regulation of mandrill growth. American Journal of Primatology, 74, 890–900.
Bielby, J., Mace, G.M., Cardillo, M., Gittleman, J.L., Jones, K.E., Orme, C.D.L. & Purvis, A.
(2007) The fast slow continuum in mammalian life history: an empirical reevaluation. The
American Naturalist, 169, 748–757.
Biro, P.A. & Stamps, J.A. (2010) Do consistent individual differences in metabolic rate promote
consistent individual differences in behavior? Trends in Ecology & Evolution, 25, 653–659.
Blanckenhorn, W.U. (2000) The evolution of body size: what keeps organisms small? The
Quarterly Review of Biology, 75, 385–407.
Bolker, B.M., Brooks, M.E., Clark, C.J., Geange, S.W., Poulsen, J.R., Stevens, M.H.H. & White,
J.-S.S. (2009) Generalized linear mixed models: a practical guide for ecology and
evolution. Trends in Ecology & Evolution, 24, 127–35.
Bonduriansky, R., Maklakov, a., Zajitschek, F. & Brooks, R. (2008) Sexual selection, sexual
conflict and the evolution of ageing and life span. Functional Ecology, 22, 443–453.
Boydston, E.E., Kapheim, K.M., Van Horn, R.C., Smale, L. & Holekamp, K.E. (2005) Sexually
dimorphic patterns of space use throughout ontogeny in the spotted hyena ( Crocuta
crocuta ). Journal of Zoology, 267, 271.
Breese, C.R., Ingram, R.L. & Sonntag, W.E. (1991) Influence of age and long-term dietary
restriction on plasma insulin-like growth factor-1 (IGF-1), IGF-1 gene expression, and IGF1 binding proteins. Journal of Gerontology: Biological Sciences, 46, B180-187.
Brent, L.J.N., Franks, D.W., Foster, E.A., Balcomb, K.C., Cant, M.A. & Croft, D.P. (2015)
Ecological knowledge, leadership, and the evolution of menopause in killer whales. Current
Biology, 25, 746–750.
Bronikowski, A.M. & Promislow, D.E.L. (2005) Testing evolutionary theories of aging in wild
populations. Trends in Ecology & Evolution, 20, 271–3.
Burnham, K.P. & Anderson, D.R. (2002) Model Selection and Multimodel Inference. Springer,
New York, NY.
Cameron, E.J., Linklater, W.L., Stafford, K.J. & Minot, E.O. (2000) Aging and improving
reproductive success in horses- declining residual reproductive value or just older and
wiser? Behavioral Ecology and Sociobiology, 47, 243–249.
Cavigelli, S.A. & Chaudhry, H.S. (2012) Social status, glucocorticoids, immune function, and
health: Can animal studies help us understand human socioeconomic-status-related health
disparities? Hormones and Behavior, 62, 295–313.
Charlesworth, B. (1993) Evolutionary mechanisms of senescence. Genetica, 91, 11–19.

108

Chellakooty, M., Vangsgaard, K., Larsen, T., Scheike, T., Falck-Larsen, J., Legarth, J.,
Andersson, a M., Main, K.M., Skakkebaek, N.E. & Juul, A. (2004) A longitudinal study of
intrauterine growth and the placental growth hormone (GH)-insulin-like growth factor I axis
in maternal circulation: association between placental GH and fetal growth. The Journal of
Clinical Endocrinology and Metabolism, 89, 384–91.
Clemmons, D.R. & Underwood, L.E. (1991) Nutritional regulation of IGF-I and IGF binding
proteins. Annual Review of Nutrition, 11, 393–412.
Clutton-Brock, T.H. (ed). (1988) Reproductive Success: Studies of Individual Variation in
Contrasting Breeding System. The University of Chicago Press, Chicago.
Clutton-Brock, T.H., Albon, S.D. & Guinness, F.E. (1985) Parental investment and sex
differences in juvenile mortality in birds and mammals. Nature, 313, 131–133.
Clutton-Brock, T.H., Albon, S.D. & Guinness, F.E. (1986) Great expectations: dominance,
breeding success and offspring sex ratios in red deer. Animal Behaviour, 34, 460–71.
Clutton-Brock, T. & Huchard, E. (2013) Social competition and its consequences in female
mammals. Journal of Zoology, 289, 151–171.
Cohen, A.A. (2004) Female post-reproductive lifespan: a general mammalian trait. Biological
Reviews of the Cambridge Philosophical Society, 79, 733–50.
Cohick, W. & Clemmons, D. (1993) The insulin-like growth factors. Annual Review of
Physiology, 55, 131–153.
Cooper, S.M.M. (1990) The hunting behavior of spotted hyaenas (Crocuta crocuta) in a region
containing both sedentary and migratory populations of herbivores. African Journal of
Ecology, 28, 131–41.
Côté, S.D. & Festa-Bianchet, M. (2001) Reproductive success in female mountain goats: the
influence of age and social rank. Animal Behaviour, 62, 173–181.
Coulson, T., Benton, T.G., Lundberg, P., Kendall, B.E., Dall, S.R.X. & Gaillard, J.-M. (2006)
Estimating individual contributions to population growth: evolutionary fitness in ecological
time. Proceedings of the Royal Society B. Biological sciences, 273, 547–55.
Cox, R.M., Cox, C.L., McGlothlin, J.W., Card, D.C., Andrew, A.L. & Castoe, T.A. (2017)
Hormonally mediated increases in sex-biased gene expression accompany the breakdown
of between-sex genetic correlations in a sexually dimorphic lizard. The American Naturalist,
189, 000–000.
Cox, R.M., Stenquist, D.S. & Calsbeek, R. (2009) Testosterone, growth and the evolution of
sexual size dimorphism. Journal of Evolutionary Biology, 22, 1586–1598.
Creel, S., Creel, N.M., Mills, M.G.L. & Monfort, S.L. (1997) Rank and reproduction in
cooperatively breeding African wild dogs: Behavioral and endorcrine correlates. Behavioral
Ecology, 8, 298–306.

109

Dantzer, B. & Swanson, E.M. (2012) Mediation of vertebrate life histories via insulin-like growth
factor-1. Biological Reviews, 87, 414–429.
Darwin, C. (1871) The Descent of Man, and Selection in Relation to Sex, 1st ed. John Murray,
London, UK.
Davidian, E., Courtiol, A., Wachter, B., Hofer, H. & Höner, O.P. (2016) Why do some males
choose to breed at home when most other males disperse ? Science Advances, 2.
Descamps, S., Boutin, S., Berteaux, D. & Gaillard, J.-M. (2006) Best squirrels trade a long life
for an early reproduction. Proceedings of the Royal Society of London. Series B: Biological
Sciences, 273, 2369–2374.
Dloniak, S.M., French, J.A., Place, N.J., Weldele, M.L., Glickman, S.E. & Holekamp, K.E. (2004)
Non-invasive monitoring of fecal androgens in spotted hyenas (Crocuta crocuta). General
and Comparative Endocrinology, 135, 51–61.
Dmitriew, C.M. (2011) The evolution of growth trajectories: what limits growth rate? Biological
Reviews, 86, 97–116.
Dufty, A.M., Clobert, J. & Møller, A.P. (2002) Hormones, developmental plasticity and
adaptation. Trends in Ecology and Evolution, 17, 190–196.
Dugdale, H.L., Pope, L.C., Newman, C., MacDonald, D.W. & Burke, T. (2011) Age-specific
breeding success in a wild mammalian population: selection, constraint, restraint and
senescence. Molecular Ecology, 20, 3261–3274.
East, M.L. & Hofer, H. (2001) Male spotted hyenas (Crocuta crocuta) queue for status in social
groups dominated by females. Behavioral Ecology, 12, 558–568.
Engh, A.L., Esch, K., Smale, L. & Holekamp, K.E. (2000) Mechanisms of maternal rank
“inheritance” in the spotted hyaena, Crocuta crocuta. Animal Behaviour, 60, 323–332.
English, S., Fawcett, T.W., Higginson, A.D., Trimmer, P.C. & Uller, T. (2016) Adaptive use of
information during growth can explain long-term effects of early life experiences. The
American Naturalist, 187, 620–632.
Fairbairn, D.J. (2005) Allometry for sexual size dimorphism: testing two hypotheses for
Rensch’s rule in the water strider Aquarius remigis. The American Naturalist, 166 Suppl,
S69–S84.
Fairbairn, D.J., Blanckenhorn, W.U. & Szekely, T. (eds). (2007) Sex, Size & Gender Roles.
Oxford University Press, Oxford, United Kingdom.
Flatt, T. & Heyland, A. (2011) Mechanisms of Life History Evolution: The Genetics and
Physiology of Life History Traits and Trade-Offs. Oxford University Press, USA, Oxford,
England.
Flies, A.S., Mansfield, L.S., Flies, E.J., Grant, C.K. & Holekamp, K.E. (2016) Socioecological
predictors of immune defenses in wild spotted hyenas. Functional Ecology, 30, 1449-1557.

110

Forslund, P. & Pärt, T. (1995) Age and reproduction in birds — hypotheses and tests. Trends in
Ecology and Evolution, 10, 374–378.
Foster, E.A., Franks, D.W., Mazzi, S., Darden, S.K., Balcomb, K.C., Ford, J.K.B. & Croft, D.P.
(2012) Adaptive prolonged postreproductive life span in killer whales. Science, 337, 1313–
1313.
Frank, L.G. (1986) Social organization of the spotted hyaena Crocuta crocuta. II. Dominance
and reproduction. Animal Behaviour, 34, 1510–1527.
Frank, L.G. & Glickman, S.E. (1994) Giving birth through a penile clitoris: Parturition and
dystocia in the spotted hyaena. Journal of Zoology, 234, 659–665.
Frank, L.G., Glickman, S.E. & Licht, P. (1991) Fatal sibling aggression, precocial development,
and androgens in neonatal spotted hyenas. Science, 252, 702–704.
Gatford, K.L., Egan, A.R., Clarke, I.J. & Owens, P.C. (1998) Sexual dimorphism of the
somatotrophic axis. The Journal of Endocrinology, 157, 373–89.
Glickman, S.E., Frank, L.G., Davidson, J.M., Smith, E.R. & Siiteri, P.K. (1987) Androstenedione
may organize or activate sex-reversed traits in female spotted hyenas. Proceedings of the
National Academy of Sciences of the United States of America, 84, 3444–3447.
Goymann, W., East, M.L., Wachter, B., Honer, O.P., Mostl, E., Van’t Holf, T.J. & Hofer, H.
(2001) Social, state-dependent and environmental modulation of faecal corticosteroid
levels in free-ranging female spotted hyenas. Proceedings of the Royal Society B:
Biological Sciences, 268, 2453–2459.
Green, D.S. (2015) Anthropogenic Disturbance, Ecological Change, and Wildlife Conservation
at the Edge of the Mara-Serengeti Ecosystem. Michigan State Univeristy.
Greider, C.W. & Blackburn, E.H. (1985) Identification of a specific telomere terminal transferase
activity in Tetrahymena extracts. Cell, 43, 405–413.
Hall, M.E., Nasir, L., Daunt, F., Gault, E.A., Croxall, J.P., Wanless, S. & Monaghan, P. (2004)
Telomere loss in relation to age and early environment in long-lived birds. Proceedings of
the Royal Society B: Biological Sciences, 271, 1571–1576.
Hamilton, W.D. (1966) The moulding of senescence by natural selection. Journal of Theoretical
Biology, 12, 12–45.
Hammers, M., Kingma, S.A., Bebbington, K., van de Crommenacker, J., Spurgin, L.G.,
Richardson, D.S., Burke, T., Dugdale, H.L. & Komdeur, J. (2015) Senescence in the wild:
Insights from a long-term study on Seychelles warblers. Experimental Gerontology, 71, 69–
79.
Harley, C., Futcher, A. & Greider, C. (1990) Telomeres shorten during ageing of human
fibroblasts.

111

Hasmann, M.F., Longenecker, A.S., Marchetto, N.M., Juliano, S.A. & Bowden, R.M. (2012)
Embryonic exposure to corticosterone modifies the juvenile stress response, oxidative
stress and telomere length. Proceedings of the Royal Society B: Biological Sciences, 279,
1447–1456.
Haussmann, M.F. & Marchetto, N.M. (2010) Telomeres: linking stress and survival, ecology and
evolution. Current Zoology.
Hedrick, A. V. & Temeles, E.J. (1989) The evolution of sexual dimorphism in animals:
Hypotheses and tests. Trends in Ecology and Evolution, 4, 136–138.
Herborn, K. a., Heidinger, B.J., Boner, W., Noguera, J.C., Adam, a., Daunt, F. & Monaghan, P.
(2014) Stress exposure in early post-natal life reduces telomere length: an experimental
demonstration in a long-lived seabird. Proceedings of the Royal Society B: Biological
Sciences, 281, 20133151–20133151.
Hinde, K. (2009) Richer milk for sons but more milk for daughters: Sex-biased investment during
lactation varies with maternal life history in rhesus macaques (eds R Sear and MA Gibson).
American Journal of Human Biology, 21, 512–519.
Hinde, K. & Milligan, L. a. (2011) Primate milk: proximate mechanisms and ultimate
perspectives. Evolutionary Anthropology, 20, 9–23.
Hiney, J.K., Srivastava, V., Nyberg, C.L., Ojeda, S.R. & Dees, W.L. (1996) Insulin-like growth
factor I of peripheral origin acts centrally to accelerate the initiation of female puberty.
Endocrinology, 137, 3717–3728.
Hobbs, C.J., Plymate, S.R., Rosen, C.J. & Adler, R.A. (1993) Testosterone administration
increases insulin-like growth factor-I levels in normal men. The Journal of Clinical
Endocrinology and Metabolism, 77, 776–779.
Hofer, H. & East, M.L. (1993a) The commuting system of Serengeti spotted hyaenas: how a
predator copes with migratory prey. II. Intrusion pressure and commuters’ space use.
Animal Behaviour, 46, 559–574.
Hofer, H. & East, M.L. (1993b) The commuting system of Serengeti spotted hyaenas: how a
predator copes with migratory prey. I. Social organization. Animal Behaviour, 46, 547–557.
Hofer, H. & East, M.L. (2003) Behavioral processes and costs of co-existence in female spotted
hyenas: a life history perspective. Evolutionary Ecology, 17, 315–331.
Hoffman, C.L., Higham, J.P., Mas-Rivera, A., Ayala, J.E. & Maestripieri, D. (2010) Terminal
investment and senescence in rhesus macaques (Macaca mulatta) on Cayo Santiago.
Behavioral Ecology, 21, 972–978.
Holekamp, K. & Sisk, C.L. (2003) Effects of dispersal status on pituitary and gonadal function in
the male spotted hyena. Hormones and Behavior, 44, 385–394.
Holekamp, K.E. & Smale, L. (1993) Ontogeny of dominance in free-living spotted hyaenas:
juvenile rank relations with other immature individuals. Animal Behaviour, 46, 451–466.

112

Holekamp, K.E. & Smale, L. (1998) Dispersal status influences hormones and behavior in the
male spotted hyena. Hormones and Behavior, 33, 205–216.
Holekamp, K.E., Smale, L., Berg, R. & Cooper, S.M. (1997) Hunting rates and hunting success
in the spotted hyena (Crocuta crocuta). Journal of Zoology, 242, 1–15.
Holekamp, K.E., Smale, L. & Szykman, M. (1996) Rank and reproduction in the female spotted
hyaena. Journal of Reproduction and Fertility, 108, 229–237.
Holekamp, K.E., Smith, J.E., Strelioff, C.C., Van Horn, R.C. & Watts, H.E. (2012) Society,
demography and genetic structure in the spotted hyena. Molecular Ecology, 21, 613–632.
Von Holst, D., Hutzelmeyer, H., Kaetzke, P., Khaschei, M., Rödel, H.G. & Schrutka, H. (2002)
Social rank, fecundity and lifetime reproductive success in wild European rabbits
(Oryctolagus cuniculus). Behavioral Ecology and Sociobiology, 51, 245–254.
Holzenberger, M., Dupont, J., Ducos, B., Leneuve, P., Géloën, A., Even, P.C., Cervera, P. & Le
Bouc, Y. (2003) IGF-1 receptor regulates lifespan and resistance to oxidative stress in
mice. Nature, 421, 182–187.
Holzenberger, M., Hamard, G., Zaoui, R., Leneuve, P., Ducos, B., Beccavin, C., Périn, L. &
Bouc, Y.L. (2001) Experimental IGF-I receptor deficiency generates a sexually dimorphic
pattern of organ-specific growth deficits in mice, affecting fat tissue in particular.
Endocrinology, 142, 4469–4478.
Höner, O.P., Wachter, B., Hofer, H., Wilhelm, K., Thierer, D., Trillmich, F., Burke, T. & East,
M.L. (2010) The fitness of dispersing spotted hyaena sons is influenced by maternal social
status. Nature Communications, 1, 60.
Van Horn, R.C., Engh, A.L., Scribner, K.T., Funk, S.M. & Holekamp, K.E. (2004) Behavioural
structuring of relatedness in the spotted hyena (Crocuta crocuta) suggests direct fitness
benefits of clan-level cooperation. Molecular Ecology, 13, 449–458.
Van Horn, R.C., McElhinny, T.L. & Holekamp, K.E. (2003) Age estimation and dispersal in the
spotted hyena (Crocuta crocuta). Journal of mammalogy, 84, 1019–1030.
Hossner, K.L., McCusker, R.H. & Dodson, M. V. (1997) Insulin-like growth factors and their
binding proteins in domestic animals. Animal Science, 64, 1–15.
Huchard, E., English, S., Bell, M.B. V, Thavarajah, N.K. & Clutton‐Brock, T.H. (2016)
Competitive growth in a cooperative mammal. Nature, 533, 532–534.
Hyatt, M.A., Budge, H., Walker, D., Stephenson, T. & Symonds, M.E. (2007) Effects of maternal
parity and late gestational nutrition on mRNA abundance for growth factors in the liver of
postnatal sheep. Am J Physiol Regul Integr Comp Physiol., 292, R1934-42.
Isley, W.L., Underwood, L.E. & Clemmons, D.R. (1983) Dietary components that regulate serum
somatomedin-C concentrations in humans. Journal of Clinical Investigation, 71, 175–182.
Jansson, J.O., Edén, S. & Isaksson, O. (1985) Sexual dimorphism in the control of growth
hormone secretion. Endocrine reviews, 6, 128–50.

113

John-Alder, H.B., Cox, R.M. & Taylor, E.N. (2007) Proximate developmental mediators of sexual
dimorphism in size: Case studies from squamate reptiles. Integrative and Comparative
Biology, 47, 258–271.
Junnila, R.K., List, E.O., Berryman, D.E., Murrey, J.W. & Kpochick, J.J. (2013) The GH/IGF-1
axis in ageing and longevity. Nature Reviews Endocrinology, 9, 366–376.
Kalmbach, E., Furness, R.W. & Griffiths, R. (2005) Sex-biased environmental sensitivity: natural
and experimental evidence from a bird species with larger females. Behavioral Ecology,
16, 442–449.
Kappeler, L., De Magalhaes Filho, C., Leneuve, P., Xu, J., Brunel, N., Chatziantoniou, C., Le
Bouc, Y. & Holzenberger, M. (2009) Early postnatal nutrition determines somatotropic
function in mice. Endocrinology, 150, 314–23.
Karubian, J. & Swaddle, J.P. (2001) Selection on females can create “larger males”. Proc. R.
Soc. Lond. B, 268, 725–728.
Kenyon, C. (2005) The plasticity of aging: insights from long-lived mutants. Cell, 120, 449–460.
Kenyon, C. (2010) A pathway that links reproductive status to lifespan in Caenorhabditis
elegans. Annals of the New York Academy of Sciences, 1204, 156–162.
Ketterson, E.D. & Nolan Jr, V. (1992) Hormones and life histories: an integrative approach. The
American Naturalist, 140, S33–S62.
Kilanowski, A.L. & Koprowski, J.L. (2016) Female-biased sexual size dimorphism: ontogeny,
seasonality, and fecundity of the cliff chipmunk (Tamias dorsalis). Journal of Mammalogy,
98, 204–210.
Kimura, M., Stone, R.C., Hunt, S.C., Skurnick, J., Lu, X., Cao, X., Harley, C.B. & Aviv, A. (2010)
Measurement of telomere length by the Southern blot analysis of terminal restriction
fragment lengths. Nature protocols, 5, 1596–607.
Kirkwood, T.B.L. & Holliday, R. (1979) The evolution of ageing and longevity. Proc. R. Soc.
Lond. B, 205, 531–546.
Kirkwood, T.B.L. & Rose, M.R. (1991) Evolution of senescence: late survival sacrificed for
reproduction. Philosophical Transactions of the Royal Society B: Biological Sciences, 332,
15–24.
Kruuk, H. (1972) The Spotted Hyena: A Study of Predation and Social Behavior. University of
Chicago Press, Chicago, Ill.
Kuznetsova, A., Brockhoff, P.B. & Christensen, R.H.B. (2015) lmerTest: Tests in linear mixed
effects models. R package version 2.0-20.
Lande, R. (1980) Sexual dimorphism, sexual selection, and adaptation in polygenic characters.
Evolution, 34, 292–305.

114

Lande, R. & Arnold, S.J. (1985) Evolution of mating preferences and sexual dimorphism.
Journal of Theoretical Biology, 117, 651–664.
Law, R. (1979) Optimal life histories under age-specific predation. The American Naturalist, 114,
399–417.
Leberg, P. & Smith, M. (1993) Influence of density on growth of white-tailed deer. Journal of
Mammalogy, 74, 723–731.
Lee, P.C., Fishlock, V., Webber, C.E. & Moss, C.J. (2016) The reproductive advantages of a
long life: longevity and senescence in wild female African elephants. Behavioral Ecology
and Sociobiology, 337–345.
Lemaître, J.-F., Berger, V., Bonenfant, C., Douhard, M., Gamelon, M., Plard, F. & Gaillard, J.-M.
(2015) Early-late life trade-offs and the evolution of ageing in the wild. Proceedings of the
Royal Society B: Biological Sciences, 282, 20150209.
Lewin, N., Swanson, E.M., Williams, B.L. & Holekamp, K.E. (2016) Juvenile IGF-1
concentrations predict life history trade-offs in a wild mammal. Functional Ecology, 31,
894–902.
Lewin, N., Treidel, L.A., Holekamp, K.E., Place, N.J. & Haussmann, M.F. (2015) Socioecological
variables predict telomere length in wild spotted hyenas. Biology Letters, 11, 20140991.
Licht, P., Frank, L.G., Pavgi, S., Yalcinkaya, T.M., Siiteri, P.K. & Glickman, S.E. (1992)
Hormonal correlates of “masculinization” in female spotted hyaenas (Crocuta crocuta). 2.
Maternal and fetal steroids. Journal of Reproduction and Fertility, 95, 463–474.
Lindenfors, P., Gittleman, J. & Jones, K. (2007) Sexual size dimorphism in mammals. Sex, size
and gender roles: evolutionary studies of sexual size dimorphism, 19–26.
Lindstrom, J. (1999) Early development and fitness in birds and mammals. Trends in Ecology &
Evolution, 14, 343–348.
Liu, L., Harding, J.E., Evans, P.C. & Gluckman, P.D. (1994) Maternal insulin-like growth factor-I
infusion alters feto-placental carbohydrate and protein metabolism in pregnant sheep.
Endocrinology, 135, 895–900.
Maklakov, A.A., Simpson, S.J., Zajitschek, F., Hall, M.D., Dessmann, J., Clissold, F.,
Raubenheimer, D., Bonduriansky, R. & Brooks, R.C. (2008) Sex-specific fitness effects of
nutrient intake on reproduction and lifespan. Current Biology, 18, 1062–1066.
McCleery, R.H., Perrins, C.M., Sheldon, B.C. & Charmantier, a. (2008) Age-specific
reproduction in a long-lived species: the combined effects of senescence and individual
quality. Proceedings. Biological sciences / The Royal Society, 275, 963–970.
McComb, K., Shannon, G., Durant, S.M., Sayialel, K., Slotow, R., Poole, J. & Moss, C. (2011)
Leadership in elephants: the adaptive value of age. Proceedings. Biological sciences / The
Royal Society, 278, 3270–6.
Medawar, P.B. (1952) An Unsolved Problem of Biology. H.K. Lewis and Co., London.

115

Metcalfe, N.B. & Monaghan, P. (2003) Growth versus lifespan: perspectives from evolutionary
ecology. Experimental Gerontology, 38, 935–940.
Van Meter, P.E., French, J.A., Bidali, K. & Weldele, M.L. (2008) Non-invasive measurement of
fecal estrogens in the spotted hyena (Crocuta crocuta). General and Comparative
Endocrinology, 115, 464–471.
Miller, R.A., Harper, J.M., Galecki, A. & Burke, D.T. (2002) Big mice die young: early life body
weight predicts longevity in genetically heterogeneous mice. Aging Cell, 1, 22–29.
Mizutani, Y., Tomita, N., Niizuma, Y. & Yoda, K. (2013) Environmental perturbations influence
telomere dynamics in long-lived birds in their natural habitat. Biology Letters, 9, 20130511.
Monaghan, P. & Haussmann, M.F. (2006) Do telomere dynamics link lifestyle and lifespan?
Trends in Ecology & Evolution, 21, 47–53.
Nenko, I. & Jasienska, G. (2013) First birth interval, an indicator of energetic status, is a
predictor of lifetime reproductive strategy. American Journal of Human Biology, 25, 78–82.
Nussey, D.H., Coulson, T., Festa-Bianchet, M. & Gaillard, J.M. (2008a) Measuring senescence
in wild animal populations: Towards a longitudinal approach. Functional Ecology, 22, 393–
406.
Nussey, D.H., Froy, H., Lemaitre, J.F., Gaillard, J.M. & Austad, S.N. (2013) Senescence in
natural populations of animals: Widespread evidence and its implications for biogerontology. Ageing Research Reviews, 12, 214–225.
Nussey, D.H., Wilson, A.J., Morris, A., Pemberton, J., Clutton-Brock, T. & Kruuk, L.E.. (2008b)
Testing for genetic trade-offs between early- and late-life reproduction in a wild red deer
population. Proceedings of the Royal Society B: Biological Sciences, 275, 745–750.
Obese, F.Y., Rabiee, A.R., Macmillan, K.L., Egan, A.R., Humphrys, S. & Anderson, G.A. (2008)
Variation in plasma concentrations of insulin-like growth factor-I in pasture-fed Holstein
cows. Journal of Dairy Science, 91, 1814–1821.
Oftedal, O.T. & Gittleman, J.L. (1989) Carnivore Behavior, Ecology and Evolution (ed JL
Gittleman). Chapman and Hall, London, UK.
Panici, J.A., Harper, J.M., Miller, R.A., Bartke, A., Spong, A. & Masternak, M.M. (2010) Early life
growth hormone treatment shortens longevity and decreases cellular stress resistance in
long-lived mutant mice. The FASEB Journal, 24, 5073–5079.
Paul, A., Kuester, J. & Podzuweit, D. (1993) Reproductive senescence and terminal investment
in female barbary macaques (Macaca sylvanus) at Salem. International Journal of
Primatology, 14, 105–124.
van de Pol, M. & Verhulst, S. (2006) Age-dependent traits: a new statistical model to separate
within- and between-individual effects. The American Naturalist, 167, 766–73.

116

Post, E., Langvatn, R., Forchhammer, M.C. & Stenseth, N.C. (1999) Environmental variation
shapes sexual dimorphism in red deer. Proceedings of the National Academy of Sciences
of the United States of America, 96, 4467–4471.
Price, T.D. (1984) The evolution of sexual size dimorphism in Darwin’s finches. The American
Naturalist, 123, 500–518.
Promislow, D.E.L. & Harvey, P.H. (1990) Living fast and dying young: A comparative analysis of
life-history variation among mammals. Journal of Zoology, 220, 417–437.
Pusey, A., Williams, J. & Goodall, J. (1997) The influence of dominance rank on the
reproductive success of female chimpanzees. Science, 277, 828–831.
Råberg, L., Stjernman, M. & Nilsson, J.-Å. (2005) Sex and environmental sensitivity in blue tit
nestlings. Oecologia, 145, 496–503.
R Development Core Team. (2014) R: A Language and Environment for Statistical Computing
(R Found Stat Comput). Available at https://www.r-project.org. Accessed November 1,
2016.
Ralls, K. (1976) Mammals in which females are larger than males. The Quarterly Review of
Biology, 51, 245–276.
Ralls, K. (1977) Sexual dimorphism in mammals: avian models and unanswered questions. The
American Naturalist, 111, 917.
Reed, T.E., Kruuk, L.E.B., Wanless, S., Frederiksen, M., Cunningham, E.J. a & Harris, M.P.
(2008) Reproductive senescence in a long-lived seabird: rates of decline in late-life
performance are associated with varying costs of early reproduction. The American
Naturalist, 171, E89–E101.
Reznick, D., Bryant, M. & Holmes, D. (2006) The evolution of senescence and post-reproductive
lifespan in guppies (Poecilia reticulata). PLoS Biology, 4, 0136–0143.
Reznick, D.N., Bryant, M.J., Roff, D., Ghalambor, C.K. & Ghalambor, D.E. (2004) Effect of
extrinsic mortality on the evolution of senescence in guppies. Nature, 431, 1095–1099.
Ricklefs, R.E. & Wikelski, M. (2002) The physiology/life-history nexus. Trends in Ecology &
Evolution, 17, 462–468.
Roach, D.A. & Carey, J.R. (2014) Population Biology of Aging in the Wild. Annual Review of
Ecology, Evolution, and Systematics, 45, 421–443.
Robbins, A.M., Robbins, M.M., Gerald-Steklis, N. & Steklis, H.D. (2006) Age-related patterns of
reproductive success among female mountain gorillas. American Journal of Physical
Anthropology, 131, 5110521.
Roberts, C.A., McCutcheon, S.N., Blair, H.T., Gluckman, P.D. & Breier, B.H. (1990)
Developmental patterns of plasma insulin-like growth factor-1 concentrations in sheep.
Domestic Animal Endocrinology, 7, 457–463.

117

Roff, D.A. & Fairbairn, D.J. (2007) The evolution of trade-offs: where are we? Journal of
Evolutionary Biology, 20, 433–47.
Rose, M.R. (1985) Life history evolution with antagonistic pleiotropy and overlapping
generations. Theoretical Population Biology, 28, 342–358.
Sadighi, M., Li, C., Littlejohn, R.P. & Suttie, J.M. (2001) Effects of testosterone either alone or
with IGF-I on growth of cells derived from the proliferation zone of regenerating antlers in
vitro. Growth Hormone & IGF Research, 11, 240–246.
Sapolsky, R.M. (2005) The influence of social hierarchy on primate health. Science, 308, 648–
652.
Sapolsky, R.M. & Spencer, E.M. (1997) Insulin-like growth factor I is suppressed in socially
subordinate male baboons. The American Journal of Physiology, 273, R1346-51.
Schorcht, W., Bontadina, F. & Schaub, M. (2009) Variation of adult survival drives population
dynamics in a migrating forest bat. Journal of Animal Ecology, 78, 1182–1190.
Schuett, W., Dall, S.R.X., Kloesener, M.H., Baeumer, J., Beinlich, F. & Eggers, T. (2015) Lifehistory trade-offs mediate “personality” variation in two colour morphs of the pea aphid,
Acyrthosiphon pisum. Journal of Animal Ecology, 84, 90–101.
Schulte-Hostedde, A.I., Millar, J.S., Gibbs, H.L. & Shulte-Hostedde, A.I. (2002) Female-biased
sexual size dimorphism in the yellow-pine chipmunk (Tamias amoenus): sex-specific
patterns of annual reproductive success and survival. Evolution, 56, 2519–29.
Schwartz, S.M. & Kemnitz, J.W. (1992) Age- and gender-related changes in body size,
adiposity, and endocrine and metabolic parameters in free-ranging rhesus macaques.
American Journal of Physical Anthropology, 89, 109–121.
Sharp, S.P. & Clutton-Brock, T.H. (2010) Reproductive senescence in a cooperatively breeding
mammal. Journal of Animal Ecology, 79, 176–183.
Sharp, S.P. & Clutton-Brock, T.H. (2011) Competition, breeding success and ageing rates in
female meerkats. Journal of Evolutionary Biology, 24, 1756–1762.
Sheldon, B.C.., Merilä, J., Lindgren, G. & Ellegren, H. (1998) Gender and environmental
sensitivity in nestling collared flycatchers. Ecology, 79, 1939–1948.
Sikes, R.S. (2016) 2016 Guidelines of the American Society of Mammalogists for the use of wild
mammals in research and education. Journal of Mammalogy, 97, 663–688.
Smale, L., Frank, L.G. & Holekamp, K.E. (1993) Ontogeny of dominance in free-living spotted
hyaenas: juvenile rank relations with adult females and immigrant males. Animal
Behaviour, 46, 467–477.
Smale, L., Holekamp, K.E., Weldele, M., Frank, L.G. & Glickman, S.E. (1995) Competition and
cooperation between litter-mates in the spotted hyaena, Crocuta crocuta. Animal
Behaviour, 50, 671–682.

118

Smale, L., Nunes, S. & Holekamp, K.E. (1997) Sexually dimorphic dispersal in mammals:
patterns, causes, and consequeneces. Advances in the Study of Behavior, 26, 181–250.
Sonntag, W.E., Lynch, C.D., Cefalu, W.T., Ingram, R.L., Bennett, S. a, Thornton, P.L. & Khan,
a S. (1999) Pleiotropic effects of growth hormone and insulin-like growth factor (IGF)-1 on
biological aging: inferences from moderate caloric-restricted animals. Journal of
Gerontology: Biological Sciences, 54, B521-38.
Sparkman, A.M., Byars, D. & Ford, N.B. (2010) The role of insulin-like growth factor-1 (IGF-1) in
growth and reproduction in female brown house snakes (Lamprophis fuliginosus). General
and Comparative Endocrinology.
Sparkman, A.M., Vleck, C.M. & Bronikowski, A.M. (2009) Evolutionary ecology of endocrinemediated life-history variation in the garter snake Thamnophis elegans. Ecology, 90, 720–
728.
Stahler, D.R., Macnulty, D.R., Wayne, R.K., vonHoldt, B. & Smith, D.W. (2013) The adaptive
value of morphological, behavioural and life-history traits in reproductive female wolves.
Journal of Animal Ecology, 82, 222–234.
Stamps, J.A. (2007) Growth-mortality tradeoffs and “personality traits” in animals. Ecology
Letters, 10, 355–63.
Starostova, Z., Kubicka, L., Golinski, A. & Kratochvil, L. (2013) Neither male gonadal androgens
nor female reproductive costs drive development of sexual size dimorphism in lizards.
Journal of Experimental Biology, 216, 1872–1880.
Stearns, S.C. (1983) The influence of size and phylogeny on patterns of covariation among lifehistory traits in mammals. Oikos, 41, 173–187.
Stearns, S.C. (1989) Trade-offs in life-history evolution. Functional Ecology, 3, 259–268.
Stearns, S.C. (1992) The Evolution of Life Histories. Oxford University Press, London.
Stearns, S.C. & Kawecki, T.J. (1994) Fitness sensitivity and the canalization of life-history traits.
Evolution, 48, 1438–1450.
Stillwell, R.C., Blanckenhorn, W.U., Teder, T., Davidowitz, G. & Fox, C.W. (2010) Sex
differences in phenotypic plasticity affect variation in sexual size dimorphism in insects:
from physiology to evolution. Annu Rev Entomol, 55, 227–245.
Sutter, N.B., Bustamante, C.D., Chase, K., Gray, M.M., Zhao, K., Zhu, L., Padhukasahasram,
B., Karlins, E., Davis, S., Jones, P.G., Quignon, P., Johnson, G.S., Parker, H.G., Fretwell,
N., Mosher, D.S., Lawler, D.F., Satyaraj, E., Nordborg, M., Lark, K.G., Wayne, R.K. &
Ostrander, E. a. (2007) A single IGF1 allele is a major determinant of small size in dogs.
Science, 316, 112–5.
Suzuki, J. & Ishida, T. (2001) Age changes in plasma IGF-1 concentration in free-ranging
Japanese macaques (Macaca fuscata). Journal of Medical Primatology, 30, 174–178.

119

Svensson, J., Sjögren, K., Fäldt, J., Andersson, N., Isaksson, O., Jansson, J.-O. & Ohlsson, C.
(2011) Liver-derived IGF-I regulates mean life span in mice. PloS one, 6, e22640.
Svensson, J., Söderpalm, B., Sjögren, K., Engel, J. & Ohlsson, C. (2005) Liver-derived IGF-I
regulates exploratory activity in old mice. American Journal of Physiology - Endocrinology
and Metabolism, 289, E466-73.
Swanson, E.M. (2013) Evolution and Development of Body Size in the Spotted Hyena. Michigan
State University.
Swanson, E.M. & Dantzer, B. (2014) Insulin-like growth factor-1 is associated with life-history
variation across Mammalia. Proceedings of the Royal Society B: Biological Sciences, 281,
20132458.
Swanson, E.M., Dworkin, I. & Holekamp, K.E. (2011) Lifetime selection on a hypoallometric size
trait in the spotted hyena. Proceedings of the Royal Society B: Biological Sciences, 278,
3277–85.
Swanson, E., McElhinny, T., Dworkin, I., Weldele, M.L., Glickman, S.E. & Holekamp, K.E.
(2013) Ontogeny of sexual size dimorphism in the spotted hyena (Crocuta crocuta).
Journal of Mammalogy, 94, 1298–1310.
Tatar, M., Bartke, A. & Antebi, A. (2003) The endocrine regulation of aging by insulin-like
signals. Science, 299, 1346–51.
Taylor, E.N. & Denardo, D.F. (2005) Sexual size dimorphism and growth plasticity in snakes: An
experiment on the Western Diamond-backed Rattlesnake (Crotalus atrox). Journal of
Experimental Zoology Part A: Comparative Experimental Biology, 303A, 598–607.
Taylor, J.F., Porter, M.J.R., Bromage, N.R. & Migaud, H. (2008) Relationships between
environmental changes, maturity, growth rate and plasma insulin-like growth factor-I (IGF-I)
in female rainbow trout. General and Comparative Endocrinology, 155, 257–270.
Therneau, T.M. (2014) A package for survival analysis in S. R package version 2.37-7.
Available at http://CRAN.R-project.org/package=survival.
Therneau, T.M. (2015) coxme: Mixed effects cox models. R package 2.2-5
Tilson, R.L. & Hamilton, W. (1984) Social dominance and feeding patterns of spotted hyaenas.
Animal Behaviour, 32, 715–724.
Toïgo, A.C., Gaillard, J. & Michallet, J. (1999) Cohort affects growth of males but not females in
Alpine ibex (Capra ibex ibex). Journal of Mammalogy, 80, 1021–1027.
Trivers, R.L. (1972) Parental investment and sexual selection. Sexual Selection and the
Descent of Man, Campbell B, pp. 136–179. Aldine, Chicago.
Verhulst, S., Holveck, M.-J. & Riebel, K. (2006) Long-term effects of manipulated natal brood
size on metabolic rate in zebra finches. Biology letters, 2, 478–80.

120

Waddington, C. (1942) Canalization of development and the inheritance of acquired characters.
Nature.
Wade, G.N. (1976) Sex Hormones, Regulatory Behaviors, and Body Weight. Advances in the
Study of Behavior, 6, 201–279.
Wahaj, S.A. & Holekamp, K.E. (2006) Functions of sibling aggression in the spotted hyaena,
Crocuta crocuta. Animal Behaviour, 71, 1401–1409.
Wang, Z. & Novak, M. (1994) Parental care and litter development in primiparous and
multiparous prairie voles (Microtus ochrogaster). Journal of Mammalogy, 75, 18–23.
Wasser, S.K., Norton, G.W., Rhine, R.J., Klein, N. & Kleindorfer, S. (1998) Ageing and social
rank effects on the reproductive system of free-ranging yellow baboons (Papio
cynocephalus) at Mikumi National Park, Tanzania. Human Reproduction Update, 4, 430–
438.
Watts, H.E. & Holekamp, K.E. (2009) Ecological determinants of survival and reproduction in
the spotted hyena. Journal of Mammalogy, 90, 461–471.
Watts, H., Tanner, J., Lundrigan, B. & Holekamp, K.E. (2009) Post-weaning maternal effects
and the evolution of female dominance in the spotted hyena. Proceedings of the Royal
Society B: Biological Sciences, 276, 2291–2298.
West, B.T. & Welch, K.B. (2014) Linear Mixed Models: A Practical Guide Using Statistical
Software, Second Ed. Chapman and Hall/CRC.
Western, D. & Ssemakula, J. (1982) Life history patterns in birds and mammals and their
evolutionary interpretation. Oecologia, 54, 281–290.
Wilkin, T.A. & Sheldon, B.C. (2009) Sex differences in the persistence of natal environmental
effects on life histories. Current Biology, 19, 1998–2002.
Williams, G.C. (1957) Pleiotropy, natural selection, and the evolution of senescence. Evolution,
11, 398–411.
Williams, P.D., Day, T., Fletcher, Q. & Rowe, L. (2006) The shaping of senescence in the wild.
Trends in Ecology and Evolution, 21, 458–63.
Yakar, S., Liu, J.L., Stannard, B., Butler, A., Accili, D., Sauer, B. & LeRoith, D. (1999) Normal
growth and development in the absence of hepatic insulin-like growth factor I. Proceedings
of the National Academy of Sciences of the United States of America, 96, 7324–7329.
Young, A.J., Carlson, A.A. & Monfort, S.L. (2006) Stress and the suppression of subordinate
reproduction in cooperatively breeding meerkats. Proceedings of the National Academy of
Sciences of the United States of America, 1093 12005–12010.
Yuan, R., Meng, Q., Nautiyal, J., Flurkey, K., Tsaih, S.-W., Krier, R., Parker, M.G., Harrison,
D.E. & Paigen, B. (2012) Genetic coregulation of age of female sexual maturation and
lifespan through circulating IGF1 among inbred mouse strains. Proceedings of the National
Academy of Sciences of the United States of America, 109, 8224–8229.

121

Yuan, R., Tsaih, S.-W., Petkova, S.B., Marin de Evsikova, C., Xing, S., Marion, M.A., Bogue,
M.A., Mills, K.D., Peters, L.L., Bult, C.J., Rosen, C.J., Sundberg, J.P., Harrison, D.E.,
Churchill, G.A. & Paigen, B. (2009) Aging in inbred strains of mice: study design and
interim report on median lifespans and circulating IGF1 levels. Aging Cell, 8, 277–87.
Zehr, J.L., Van Meter, P.E. & Wallen, K. (2005) Factors regulating the timing of puberty onset in
female rhesus monkeys (Macaca mulatta): role of prenatal androgens, social rank, and
adolescent body weight. Biology of Reproduction, 72, 1087–1094.
Zera, A.J. & Harshman, L.G. (2001) The physiology of life history trade-offs in animals. Annual
Review of Ecology and Systematics, 32, 95–126.

122