 

 

 

LIBRARY

Mlchlgan State

University

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

MTE DUE

MTE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1/98 cJCIWpGS—p.“

PHYSIOLOGICAL AND ECOLOGICAL MEDIATION OF
NATAL DISPERSAL IN BELDING'S GROUND SQUIRRELS

By

Scott Nunes

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

DOCTOR OF PHILOSOPHY

Department of Zoology

1997

ABSTRACT

PHYSIOLOGICAL AND ECOLOGICAL MEDIATION OF
NATAL DISPERSAL IN BELDING’S GROUND SQUIRRELS

By

Scott Nunes

Natal dispersal, the permanent departure of an individual from its birth site, occurs
in nearly all mammals, and tends to be sexually dimorphic in its occurrence. That is, one
sex usually disperses at higher rates or over greater distances than the other. The
physiological and ecological variables inﬂuencing this behavior are poorly understood.
Here I attempted to elucidate the roles played by these variables in the mediation of natal
dispersal behavior in Belding’s ground squirrels (Spermophilus beldingi).

Males are the predominant dispersers in S. beldingi, as they are in most mammals.
I examined the role of three classes of physiological factors in the mediation of the natal
dispersal behavior of free-living male S. beldingi: energy availability, endogenous timing
mechanisms, and gonadal steroid hormones. I also assessed ecological inﬂuences on natal
dispersal behavior in both juvenile male and female S. beldingi. I manipulated body fat
and energy availability of juvenile S. beldingi by provisioning them with food, and used
unprovisioned juveniles as controls. I compared dispersal between juveniles born early vs.
late during the breeding season. I manipulated early hormone exposure of female S.
beldingi by treating them with testosterone (T) at birth, and used oil-treated females as
controls. I combined hormone and body fat manipulations to evaluate the interaction

between the two. I also collected blood ﬁ'om juveniles to measure circulating

concentrations of T. I regularly evaluated dispersal status, body mass, and body fat of
subjects, while concurrently monitoring several ecological variables.

Juvenile male S. beldingi appear to delay dispersal until they have attained a
threshold body mass. Body fat also importantly inﬂuences the timing of dispersal.
Dispersing males tend to be fatter than same-aged males still residing in the natal area.

The energetic demands of depositing fat for hibernation appear to constrain
dispersal by juvenile male S. beldingi. Juvenile males born late in the year, and which have
little time to prepare for hibernation, exhibit accelerated acquisition of body fat, and
postpone dispersal until their yearling summer. Thus, an endogenous timing mechanism
appears to interact with body fat or fat-related variables in mediating natal dispersal
behavior.

Concentrations of circulating T are undetectable in juvenile S. beldingi, suggesting
that T does not activate natal dispersal during ontogeny. However, treatment of newborn
females with T increases the likelihood that they will emigrate from the natal area.
Moreover, food provisioning advances the timing of natal dispersal among both T-treated
females and males, and these advances in timing appear to be mediated by increases in
body fat.

All male S. beldingi eventually emigrate from their birth sites, regardless of
ecological conditions. By contrast, the probability of dispersal by young female S.
beldingi is inﬂuenced by demand for environmental resources. Young females may

disperse to enhance their access to space in which to establish a territory.

ACKNOWLEDGMENTS

This work was supported by grants-in-aid of research from the American Museum
of Natural History, the American Society of Mammalogists, the Animal Behavior Society,
Sigma Xi, and Michigan State University, and a National Science Foundation Graduate
Research Fellowship to S. Nunes, and by fellowships to K. E. Holekamp from the Searle
Scholars Program/Chicago Community Trust and the David and Lucille Packard
Foundation. Clairol generously donated the hair dye used to mark squirrels.

For their excellent assistance in the ﬁeld, I thank Amy Barber, Lisa Bissett, Tim
Duniec, Anne Engh, Paula Garrett, Co-Diem Ha, David Kent, Eva-Maria Muecke, Ryan
Ojerio, Kurt Reinhart, Sara Schweppe, Amir, Sheikh, Yoshimi Toda, and Peter Zugger. I
thank Lyn Clemens for the use of squirrel cages, Don Beaver for the use of traps, Dan
Hayes and William Taylor for assistance with lipid extractions, and Victor Dom and Gary
Lange for invaluable logistic support.

Erin Boydston, Steve Dobson, Fred Dyer, Anne Engh, Carmen Salsbury, Laura
Smale, Don Straney, Jerry Wolff, and especially Kay Holekamp oﬂ‘ered perceptive
comments and suggestions on the planning, execution, and interpretation of this work. I
thank Dick Hill and Kay Holekamp for their careful readings of an earlier draﬁ of this

dissertation, and for their incisive comments which greatly improved it. For their support

iv

and direction I thank the members of my guidance committee, Lyn Clemens, Fred Dyer,
Dick Hill, Kay Holekamp, Laura Smale, and Don Straney.

For their encouragement I thank my fellow graduate students Erin Boydston, Anne
Engh, Terri McElhinny, Micaela Szykman, Anita Davelos, and Greg Papakostas. I offer
special thanks to Victor Dom and Laura Smale for their support and encouragement
throughout this work, and to my adviser, Kay Holekamp, for being a great role model and
for providing opportunities to explore and understand animal behavior. Finally, I thank

the squirrels for helping to expand our understanding of ecology and behavior.

TABLE OF CONTENTS

LIST OF FIGURES ....................................................................................................... viii

CHAPTER 1:

INTRODUCTION ........................................................................................................... 1
Energy availability and dispersal ............................................................................ 2
Timing mechanisms and dispersal .......................................................................... 3
Gonadal steroid hormones and dispersal ................................................................ 5
Evolutionary causes of natal dispersal .................................................................... 8
Hypotheses ........................................................................................................... 9
Signiﬁcance of the research ................................................................................. 10

CHAPTER 2:

GENERAL METHODS ................................................................................................ 12
Spennophilus beldingi .......................................................................................... 12
Study site and subject animals ............................................................................. 14
Trapping, handling, and recovery of squirrels .......................................................... 15
Evaluation of dispersal status .............................................................................. 16
Observation of behavior ...................................................................................... 17
Estimation of population density ......................................................................... l8
Provisioning ....................................................................................................... 19
Evaluation of body fat ........................................................................................ 20
Hormone treatments ........................................................................................... 22
Blood collection and hormone assays .................................................................... 23
Statistical analyses of data .................................................................................. 23

CHAPTER 3:

ONTOGENY OF MASS AND FAT AQUISITION AND DISPERSAL ....................... 26
Introduction ....................................................................................................... 26
Methods ............................................................................................................. 27
Results ................................................................................................................ 27
Discussion .......................................................................................................... 37

CHAPTER 4:

INTERACTION OF TIME OF YEAR AND ENERGETIC VARIABLES ..................... 44
Introduction ....................................................................................................... 44

vi

Methods ............................................................................................................. 46

Results ............................................................................................................... 46
Discussion .......................................................................................................... 58

CHAPTER 5:

ENDOCRINE INFLUENCES ON DISPERSAL ........................................................... 63
Introduction ....................................................................................................... 63
Methods ............................................................................................................. 64
Results ............................................................................................................... 66
Discussion .......................................................................................................... 73

CHAPTER 6:

INTERACTION OF ENDOCRINE AND ENERGETIC VARIABLES ......................... 77
Introduction ........................................................................................................ 77
Methods ............................................................................................................. 79
Results ............................................................................................................... 79
Discussion .......................................................................................................... 90

CHAPTER 7:

ECOLOGICAL INFLUENCES ON DISPERSAL ........................................................ 96
Introduction ....................................................................................................... 96
Methods ............................................................................................................. 98
Results ............................................................................................................... 98
Discussion ......................................................................................................... l 10

CHAPTER 8:

SUMMARY AND CONCLUSIONS ........................................................................... 116
Mass and fat inﬂuences on dispersal .................................................................. 116
Endogenous timing mechanisms ....................................................................... 117
Endocrine mediation of dispersal ...................................................................... 118
Ecological factors inﬂuencing dispersal ............................................................. 118
Directions for future research ........................................................................... 119

REFERENCES ........................................................................................................... 122

vii

LIST OF FIGURES

Figure 1 - Regression of fat-free mass on conductivity index ............................................ 21
Figure 2 - Body mass of juvenile males, 1993-1994 .......................................................... 29
Figure 3 - Distances of juvenile males from their natal burrows, 1993-1994 ...................... 30
Figure 4 - Dispersal of juvenile males at various ages, 1993-1994 ..................................... 31
Figure 5 - Cumulative dispersal of juvenile males, 1993-1994 ........................................... 32
Figure 6 - Body mass of juvenile males at dispersal, 1993-1994 ........................................ 34
Figure 7 - Body mass of juvenile males relative to dispersal, 1993-1994 ............................. 35
Figure 8 - Body fat of juvenile males, 1994 ...................................................................... 36
Figure 9 - Body fat of juvenile males during the primary dispersal period, 1994 .................... 38
Figure 10 - Body fat of dispersing and natal males, 1994 ................................................ 39
Figure 11 - First emergence of litters from the natal burrow, 1994-1996 .......................... 47
Figure 12 - Body mass of juvenile males, 1994-1996 ........................................................ 49
Figure 13 - Body fat of juvenile males, 1994-1996 ........................................................... 50
Figure 14 - Dispersal of juvenile males, 1994-1996 ........................................................... 52
Figure 15 - Dispersal of juvenile and yearling males in the 1994 and 1995 cohorts ................ 53
Figure 16 - Body fat of juvenile males during the active season, 1994-1996 ...................... 54
Figure 17 - Dispersal of juvenile males during the active season, 1994-1996 ..................... 57
Figure 18 - A model of circannual and energetic regulation of dispersal ............................ 61

viii

Figure 19 - Body mass of yearling males at the end of their juvenile summer .................... 68

Figure 20 - Body fat of juvenile males at the end of their juvenile summer ........................ 69
Figure 21 - Dispersal of yearling males, 1996 ................................................................... 71
Figure 22 - Dispersal of the 1995 cohort ............................................................................. 72
Figure 23 - Body mass of juvenile females, 1996 .............................................................. 81
Figure 24 - Body mass of juvenile males, 1996 ................................................................. 82
Figure 25 - Body fat of juvenile females, 1996 .................................................................. 83
Figure 26 - Body fat of juvenile males, 1996 ..................................................................... 84
Figure 27 - Dispersal of juveniles, 1996 ............................................................................ 86
Figure 28 - Dispersal of T-treated females, 1996 .............................................................. 88
Figure 29 - Dispersal of males, 1996 ................................................................................ 89
Figure 30 - A model of endocrine and energetic regulation of dispersal ............................. 93
Figure 31 - Body mass of juveniles, 1993-1994 ................................................................ 99
Figure 32 - Density of S. beldingr‘, 1993-1995 .................................................................. 101
Figure 33 - Vigilance behavior of adult females, 1995 ...................................................... 103
Figure 34 - Aggressive behavior of adult females, 1995 ................................................... 104
Figure 35 - Dispersal of the 1993 and 1994 cohorts .......................................................... 105
Figure 36 - Body mass of dispersed and natal females, 1993-1994 .................................. 108
Figure 37 - Body mass of juveniles relative to dispersal, 1993-1994 ............................... 109

ix

Chapter 1

INTRODUCTION

Most group-living vertebrates transfer between social units at least once during
their lifetimes. The tendency to leave the natal group, also called dispersal or emigration
behavior, occurs in most mammals and tends to be sexually dimorphic (Greenwood,
1980). Speciﬁcally, male mammals generally disperse with greater frequency or over
greater distances than do female conspeciﬁcs. Although ecological variables promoting
dispersal behavior and adaptive consequences of dispersal have been well studied (e. g.
Pusey, 1987; Stenseth and Lidicker, 1992; Smale et al., 1997), the physiological mediation
of this important and widely-occurring behavior remains poorly understood in mammals
(Dingle, 1996). The goal of this work was to elucidate physiological processes that
underlie emigration from the natal site in male mammals, using Belding’s ground squirrels
(Spermophilus beldingi) as a model species. I examined three classes of physiological
variables as potential regulators of dispersal behavior: (1) availability of energy as
reﬂected in body fat (Nunes and Holekamp, 1996), (2) internal timing mechanisms (Nunes
et al., in press), and (3) gonadal steroid hormones (Smale et al., 1997). In the course of
this work I was also able to observe ecological inﬂuences, such as competition for food

and space, on emigration by those few female S. beldingi that leave the natal area (Nunes

2

et al., 1997). This allowed me to examine causes and consequences of dispersal by female

mammals.

ENERGY A VAILABILI TY AND DISPERSAL

Energy availability often inﬂuences the behavior of animals, which must partition
available energy according to priorities that ensure survival and optimize long-term
reproductive success (Wade and Schneider, 1992). The possibility that the availability of
energy might regulate the timing of dispersal is suggested by the fact that emigration from
the natal area is potentially both risky and energetically costly (Smale et al., 1997).
Emigration is associated with increased locomotor and exploratory behavior (Messier,
1985a; Holekamp, 1986; Rood, 1987). Moreover, dispersers exploring unfamiliar areas
often appear to be more vulnerable to predation or other sources of mortality than are
long-term residents (Metzgar, 1967; Snyder et al., 1976; Messier, 1985b), and dispersing
individuals often encounter aggressive resistance from resident animals when attempting to
immigrate into new areas (Holekamp, 1986; Pusey and Packer, 1987a; Smale et al., 1997).
As a result, individuals may need to reallocate time spent foraging to vigilant and
defensive behaviors during the dispersal movement. Changes in activity associated with
dispersal may therefore increase energy requirements and decrease energy intake. Natural
selection should therefore theoretically favor individuals that delay departure ﬁom the
natal area until their energy reserves are adequate to support the potential demands of the
dispersal movement (Nunes and Holekamp, 1996).

I examined the inﬂuences of energy availability, as reﬂected in body mass and body

fat, on the dispersal behavior of juvenile male S. beldingi. If the timing of dispersal is in

3

fact inﬂuenced by the amount of energy available to an individual, then altering energy
availability should have an impact on when an individual leaves its birth site. I provided
juvenile male S. beldingr' with extra food to increase their reserves of energy stored as fat.
I predicted that these provisioned S. beldingr' should be better able to cope with the
energetic demands of dispersal, and should emigrate ﬁ'om their natal areas at younger ages

than unprovisioned controls.

T MNG MECHANISMS AND DISPERSAL

Many life history events that occur in seasonal or annual cycles are regulated by
endogenous timing mechanisms (Zucker et al., 1991). When dispersal behavior is closely
associated with seasonal events, it might also be inﬂuenced by endogenous timing
mechanisms (Smale et al., 1997). In species exhibiting seasonal reproduction, dispersal
frequently occurs during the breeding period (e. g., Berger, 1986, 1987; Pusey and Packer,
1987a; Sprague, 1992; Matsumura, 1993), and reproduction in seasonal breeders is often
regulated by circannual or photoperiodic timing mechanisms (e. g., Zucker, 1988; Zucker
et al., 1991; Bartness et al., 1993). Thus, it seems reasonable to expect that the timing of
dispersal, puberty and reproduction in some seasonal breeders might be regulated
physiologically by a common timing mechansim. For example, an internal oscillator that
regulates the timing of changes in reproductive status might also mediate temporal
variation in dispersal-related behaviors such as locomotion and exploration. “Dispersal
restlessness” occurs seasonally in birds (Ritchison et al., 1992), and might also occur in

mammals (Smale et al., 1997).

4

Timing mechanisms that regulate cyclic life history events other than reproduction
might also mediate the timing of dispersal behavior. For example, many mammals in
temperate habitats hibernate during predictable periods of limited food availability in
winter, and many hibernators depend entirely on stored fat for survival during dormancy
(Mrosovsky and Fisher, 1970; Davis, 1976; Mrosovsky, 1976; French, 1982).
Hibernation and its associated patterns of fat deposition and usage exhibit annual rhythms
generated by internal oscillators (Zucker et al., 1991). Because hibernation and dispersal
both present the organism with energetic challenges, they might come into conﬂict. That
is, an individual that uses energy to support dispersal behavior may compromise its ability
to build fat reserves adequate to survive the winter (Nunes and Holekamp, 1996). It is
plausible that after a critical date during the annual cycle, when little time remains until
winter, an endogenous timing mechanism initiates fattening for hibernation. If body fat
reserves are low at this critical date, this timing mechanism may inhibit dispersal behavior
so that available energy can be used to build fat stores for overwinter survival. Thus,
energetic and timing mechanisms might interact to mediate dispersal behavior in
hibernators, and in other mammals in which energetic demands vary seasonally.

I tested the hypothesis that endogenous timing mechanisms inﬂuence natal
dispersal in juvenile male S. beldingi. I compared dispersal behavior between three years,
during one of which the active season was substantially delayed. If endogenous timing
mechanisms regulate dispersal behavior, then a very late onset of the active season should
be associated with a decrease in dispersal by juveniles. In particular, when the active
season begins late, internal physiological processes should have been favored by natural

selection which anticipate the approach of winter, and these should direct available energy

5

toward building fat reserves for hibernation rather than toward fueling dispersal behavior.
Thus, a greater proportion of juvenile S. beldingi should delay dispersal until the yearling
summer in years when the juvenile active season begins very late, compared to years in

which the active season is not delayed.

GONADAL STEROID HORMONES AND DISPERSAL

Hormones have been implicated in the control of long-distance movements in
invertebrates, ﬁsh, amphibians, and birds (Dingle, 1996), and might also mediate
mammalian dispersal (Smale et al., 1997). The speciﬁc possibility that mammalian
dispersal might be inﬂuenced by gonadal steroid hormones is suggested by the sexual
dimorphism in dispersal, and by its common occurrence around the time of puberty.
Gonadal steroids can promote sex differences in behavior in two general ways (Phoenix et
al., 1959). First, hormones can alter the physical and biochemical conﬁguration of the
developing central nervous system during fetal or neonatal life to organize tissue which
mediates sexually dimorphic behavior patterns later in life. Second, hormones can act
directly to induce more immediate behavioral changes in the mature animal. Phoenix et al.
(1959) chose the terms “organizational” and “activational,” respectively, to describe these
two modes of hormone action. Organizational effects on behavior are often permanent
and irreversible, whereas activational effects are generally temporary and require the
concurrent presence of activating hormone. Either organziational, activational, or both
effects might promote sexually dimorphic mammalian dispersal.

Temporal coincidence of dispersal with puberty is consistent with the hypothesis

that rising levels of gonadal steroid hormones might activate dispersal behavior. Puberty

6

is characterized by dramatic increases in circulating levels of gonadal steroid hormones
that stimulate a variety of behavioral changes (Bronson and Rissman, 1986; Wilson,
1992). A close temporal association between dispersal and puberty has been observed in
many species of mammals (e. g., Errington, 1963; Harcourt et al., 1976; Harcourt, 1978;
Pusey, 1980; Berger, 1986, 1987; Pusey and Packer, 1987a; Rood, 1987, 1990;
Matsumura, 1993). However, the hypothesis that gonadal steroid hormones activate
dispersal has not yet been empirically tested in any of these species.

When emigration occurs well before puberty, as it does in many sciurid rodents
and dasyurid marsupials (Holekamp, 1984a; Cockbum et al., 1985), gonadal steroid
hormones are not likely to activate dispersal behavior. However, gonadal steroids might
exert organizational effects during the prenatal or perinatal period to promote dispersal
later in life. Androgens organize many non-sexual, dimorphic behaviors that all males of a
species predictably exhibit during speciﬁc stages of development (Beatty, 1992). For
example, perinatal androgen exposure promotes masculine patterns of play behavior in
many mammals (e. g., Thor and Holloway, 1984, 1986). Theoretically, androgens might
also organize a predisposition to emigrate from the natal area in the many species of
mammals in which all males eventually disperse, such as S. beldingr', spotted hyenas
(Crocuta crocuta), lions (Panthera lea), ponies (Equus caballus), marsupial mice (e. g.,
Antechinus stuartii, A. swainsonii), and various species of cercopithecine primates (Smale
et al., 1997). However, in some species of mammals, such as Columbian ground squirrels
(S. columbianus), the dispersal behavior of males is more ﬂexible, and some males

disperse while others remain in their natal areas. In these species, androgens probably do

7

not organize a deterministic predisposition to leave the natal area, but might organize a
masculine response to social or ecological variables which promote dispersal behavior.

Androgen exposure during early development might also masculinize the ways in
which dispersal behavior is inﬂuenced by energetic variables (Smale et al., 1997). In the
numerous species of mammals in which dispersers are predominantly male, the few
females that disperse should confront many of the same energetic problems faced by
males. Nevertheless, patterns of natal dispersal and key life history events often differ
between males and females, and the speciﬁc manner in which fat-related variables regulate
dispersal behavior might also differ. For example, in many mammals, females begin
reproducing at earlier ages than males (e. g., Harcourt et al., 1976; Sherman and Morton,
1984; Berger, 1987; Bronson, 1989; Packer et al., 1995), and disperse at later ages (Smale
et al., 1997). Consequently, females tend to experience the energetic rigors of
reproduction earlier than males, and the energy requirements of emigration from the natal
area might impose greater potential costs on the future reproductive success of young
females than on that of young males. The energetic regulation of dispersal behavior may
be fundamentally similar for young male and female mammals, but sex differences in some
of its ﬁner details might be mediated by gonadal hormones. Thus, endocrine mechanisms
and energetic variables might work together to mediate dispersal behavior, just as timing
mechanisms and energetic variables might interact.

I examined the possibility that testosterone (T) inﬂuences natal dispersal behavior
in S. beldingi, and that the effects of T on dispersal are inﬂuenced by energy availability. I
predicted that if if T activates natal dispersal in juvenile males, then circulating

concentrations of T should increase near the time of dispersal. However, if T organizes

8

natal dispersal behavior, then treatment of female S. beldingi with T at birth should
masculinize physiological systems that promote dispersal behavior, and T-treated females
should exhibit male-typical dispersal behavior later during their ontogeny. Finally, if
energy availability interacts with organizational effects of T on dispersal, then altering
energy availability by food provisioning should advance the timing of dispersal (Nunes and

Holekamp, 1996) in both males and females treated with T at birth.

E VOL UYYONARY CA USES OF DISPERSAL

Three hypotheses are commonly offered in the literature as evolutionary
explanations for natal dispersal: individuals may disperse (1) to increase their access to
environmental resources such as food, territories or nest sites, (2) to improve their access
to mates, or (3) to avoid deleterious effects on ﬁtness of close inbreeding (Lidicker, 1962,
1975; Packer, 1979; Dobson, 1982; Moore and Ali, 1984; Dobson and Jones, 1985;
Pusey, 1987; Pusey and Packer, 1987a; Stenseth and Lidicker, 1992). As suggested by
the striking sexual dimorphisms in dispersal behavior, the evolutionary causes of dispersal
oﬁen differ between males and females (Dobson, 1979; Holekamp, 1984a; Lambin, 1994;
Waser, 1996; Smale et al., 1997). In species in which ultimate causes of dispersal have
been identiﬁed, males often disperse to avoid inbreeding or to enhance the probability of
encountering receptive mates, whereas females usually disperse to gain better access to
environmental resources such as food or space (Smale et al., 1997). As is the case in
many mammals, female S. beldingi appear to leave the natal area to improve their access

to environmental resources critical for raising young, whereas male S. beldingi appear to

9

leave to increase their opportunities to mate with receptive or unrelated females (Nunes et

al., 1997; Smale et al., 1997).

H YPOIHESES

In this work I examined ﬁve hypotheses suggesting explanations for particular
aspects of dispersal behavior in S. beldingi:

(1) variables, reﬂecting availability of energy, such as body fat and body mass

inﬂuence the timing of dispersal by juvenile male S. beldingr',

(2) (a) internal timing mechanisms regulate dispersal behavior in juvenile males and

(b) in doing so, these interact with energetic variables such as body fat,

(3) gonadal steroid hormones exert (a) organizational and/or (b) activational

effects on natal dispersal behavior in S. beldingi,
(4) endocrine mechanisms interact with energetic variables to mediate the
dispersal behavior typical of juvenile male S. beldingr', and

(5) the probability of dispersal by juvenile female but not juvenile male S. beldingi
is inﬂuenced by demand for environmental resources such as (a) food, (b) nest
sites, or (c) space on which to establish a territory.

I tested these hypotheses using various experimental regimens of food provisioning
and hormone manipulations. I gave juvenile S. beldingi extra food to manipulate body
mass and body fat, and observed the inﬂuences of these variables on dispersal behavior.
Because food provisioning is often associated with increases in local population density
(Dobson, 1995; Boutin, 1990; Nunes et al., 1997), I was able to oberve the effects of local

resource competition on dispersal behavior. To evaluate effects of endogenous timing

10

mechanisms, I compared dispersal behavior of provisioned and unprovisioned juveniles
between years with normal and unusually late active seasons. To examine endocrine
inﬂuences on dispersal, I altered androgen levels of some provisioned and unprovisioned
females at birth, and later observed whether they exhibited masculine dispersal behavior

(Holekamp et al., 1984), and I also measured circulating androgen levels of juveniles.

SIGNIFICANCE OF THE RESEARCH

This research represents an attempt to elucidate the relationship between
physiological substrates, ontogenetic development, and adaptive signiﬁcance of an
ecologically signiﬁcant mammalian behavior. Dispersal profoundly inﬂuences the
demography and genetic composition of vertebrate populations (Keith and Tamarin, 1981;
Charlesworth and Charlesworth, 1987; Chepko-Sade and Halpin, 1987; Waser and Elliott,
1991; Nelson, 1993), so an understanding of its proximal and evolutionary causes should
prove directly useful to ecologists and conservation biologists alike, as habitats of
vertebrates become increasingly fragmented and isolated (e. g., Halpin, 1987; Bennett,
1990; Diffendorfer et al., 1995). Moreover, this work helps to ﬁll a virtual information
void in the physiology of mammalian dispersal behavior. It also sheds light on the extent
to which gonadal steroid hormones can mediate nonsexual behavior patterns in mammals,
and further elucidates the inﬂuence that these hormones exert on sex differences in
mammalian behavior. Finally, and most importantly, this work is a novel endeavor to
evaluate in free-living mammals the interaction between endocrine control of behavior and
the regulation of energy balance. To date most studies of this interaction have been

accomplished in the laboratory using species such as hamsters (Mesocricetus auratus;

ll

Wade and Schnieder, 1992; Wade et al., 1996) and mice (Mus musculus; Campﬁeld et al.,
1995; Halaas et al., 1995; Pellymounter et al., 1995) whose ecology is very poorly
understood. As these laboratory studies become progressively more molecular, it is
important not to lose sight of the importance of ecological and evolutionary phenomena

inﬂuencing interactions between physiology and behavior.

Chapter 2

GENERAL METHODS

SPERMOPHIL US BELDINGI

S. beldingi is a diurnal, semi-fossorial sciurid rodent inhabiting alpine and sub-
alpine meadows of the western United States (Jenkins and Eshelman, 1984). Although
populations of S. beldingi do not contain distinct social subdivisions as do those of some
of the more social sciurids (King, 1955; Barash, 1973; Michener, 1983; Hoogland, 1995),
S. beldingi is a group-living animal that engages in daily social interactions with
conspeciﬁcs (Sherman, 1977). Populations of S. beldingi typically contain several clusters
of related females (Michener, 1983). Within such clusters, each breeding female utilizes a
separate burrow system, but interacts more frequently and more amicably with close
female kin than with more distantly related females (Sherman, 197 7, 1981a).

The life history of S. beldingi is characterized by four-month active seasons
interspersed with eight-month hibernation periods (Sherman and Morton, 1984). The
annual cycle of activity is rigidly constrained by extreme seasonal changes in climate
(Morton and Sherman, 197 8), and ontogenetic development after birth is accelerated
compared to that of non-hibernating rodents of similar body size. Between weaning and
entry into hibernation seven to twelve weeks later, the mean body mass of juveniles

increases four to six fold. This rapid development presumably permits juveniles to attain

12

13

body mass and fat reserves necessary for overwinter survival (Morton and Tung, 1971;
Morton et al., 1974; Maxwell and Morton, 1975).

S. beldingi ’s breeding season ordinarily begins one to two weeks after emergence
ﬁom hibernation in the spring, and lasts three to four weeks (Sherman and Morton, 1984).
Adult males are reproductively competent when they emerge from hibernation, whereas
yearling males generally do not exhibit descended testes and do not mate (Morton and
Gallup, 1975). Variance in mating success is high, and only a small proportion of males
successfully copulate during the breeding season (Sherman, 1976). Body mass of male S.
beldingr' is positively correlated with success in winning ﬁghts with other males during the
breeding season, and also with mating success (Sherman, 1976). Adult and yearling
female S. beldingi emerge from hibernation slightly later than do males, and most females
(>60%) begin mating as yearlings. Females exhibit estrous behavior on only one day per
year, and appear to exhibit mate choice behavior, accepting the mating attempts of some
males while rejecting those of others (Sherman, 1976). Litters are multiply sired, typically
by two to three males (Hanken and Sherman, 1981). Gestation lasts 24-25 days, and the
lactation interval is 25-28 days (Holekamp, et al., 1984; Nunes, unpublished data). Young
remain underground in their natal burrows until weaning. Females maintain maternal
territories during gestation and lactation, and often act cooperatively with female kin in
defending these territories (Sherman, 1977, 1981a). Female S. beldingr’ are able to
discriminate between litterrnate sisters, non-littermate sisters, and other female squirrels,
and this capacity to recognize kin appears to facilitate and mediate nepotistic interactions

among females (Holmes and Sherman, 1982; Holmes, 1986a, 1986b).

14

Natal dispersal in S. beldingr’ occurs predominantly among males (Holekamp,
1984a). Most males (40-90%) emigrate from their natal areas as juveniles, and 100% of
surviving males disperse by the end of the yearling year (Holekamp, 1984a; Nunes and
Holekamp, 1996). By contrast, most females remain in their natal areas throughout their
lives, and only about 8% of females ordinarily emigrate (Holekamp, 1984a; Nunes et al.,
1997)

The probability of natal dispersal by male S. beldingi is not proximally inﬂuenced
by social or ecological factors (Holekamp, 1986). All males eventually disperse regardless
of population density, resource availability, ectoparasite load, conspeciﬁc aggression and
the ﬁ'equency of other social interactions with conspeciﬁcs. A suite of behaviors
associated with dispersal, including increased locomotion, exploration and boldness,
appear to be triggered at some point during ontogeny. These behaviors evidently lead to
or otherwise facilitate emigration ﬁom the natal area. The trigger for these behaviors
appears to be attainment of a particular body mass or body composition (Holekamp, 1986;

Nunes and Holekamp, 1996).

STUDY SITE AND SUBJECT POPULA IION

I studied a population of S. beldingr' in a 20 hectare meadow at Tioga Pass (3 8° N,
119° W, elevation 3040 m) in Mono County, California, from 15 May to 26 Sept 1993, 15
May to 3 Sept 1994, 17 May to 5 Oct 1995, and 13 May to 2 Sept 1996. This site has
been described in detail by Morton (1975) and Sherman (1976). I deﬁned a juvenile as an
individual born during the current year, a yearling as one born during the previous year,

and an adult as one born at least two years earlier. I deﬁned a reproductive female as one

15

who was pregnant or lactating. From 1993-1996 I permanently marked 1098 squirrels
(including 365 male and 380 female juveniles from 163 complete litters) with monel metal
ear tags (National Band and Tag Co., Newport, KY).

The growing season at Tioga Pass is very short, and places considerable
constraints on the population of S. beldingr’ that lives there (Morton and Tung, 1971;
Morton et al., 1974; Maxwell and Morton, 1975; Morton, 1975; Sherman, 1976; Morton
and Sherman, 1978; Sherman and Morton, 1979). When squirrels at Tioga Pass ﬁrst
emerge from hibernation in May, much of their habitat is still covered with snow, and they
must rely on fat reserves for energy. During June and July, green vegetation is abundant
and comprises the main part of the squirrels’ diet. In August, the squirrels’ diet shifts to
seeds produced by grasses and other vegetation at that time of year. By September, much
of the vegetation at Tioga Pass has dried out or senesced, and food availability is severely
limited. The extreme brevity of the growing season at Tioga Pass has important
consequences for S. beldr'ngi. The active season of the squirrels is very short (about 3-4
months), and squirrels enter hibernation throughout August and September. Events such
as reproduction, pre-hibemation fattening, dispersal, and physical and behavioral

development must be exactly timed to ﬁt into this very short period of activity.

TRAPPING, HANDLING, AND RECOVERY OF SQUIRRELS
I regularly live-trapped squirrels in the study area using Tomahawk live-traps
(Tomahawk Live-Trap Co., Tomahawk, WI) baited with peanut butter. At each capture, I

weighed squirrels to the nearest gram with spring balance scales (Avinet, Dryden, NY),

l6

and recorded reproductive status (Morton and Gallup, 197 5; Holekamp, 1983) and
capture location (distance and direction from nearest landmarks). I marked individual
squirrels with unique symbols using hair dye to facilitate visual identiﬁcation (Sherman,
1976; Holekamp, 1983).

I observed maternal burrows of lactating females daily to determine the dates on
which their young ﬁrst emerged above ground. I trapped juveniles within four days of
their ﬁrst appearance of the juveniles, during which time they remained close to their natal
burrows and could be unambiguously assigned to litters (Sherman, 1976). I assigned birth
dates to juveniles by subtracting 27 days, the mean duration of lactation, from the date of
ﬁrst emergence from the natal burrow (Holekamp et al., 1984). I observed squirrels from
atop rocks and ladders, with the aid of 7 x 50 binoculars.

I trapped and ear-tagged 48 juvenile males and 47 juvenile females in 1993; 95
juvenile males and 107 juvenile females in 1994, 126 juvenile males and 130 juvenile
females in 1995, and 107 juvenile males and 97 juvenile females in 1996. I considered
juveniles to be recovered each summer if their dispersal status (see below) was known at
10 weeks of age in 1993, 1994, and 1995, or 9 weeks of age in 1996. I recovered 50.0%
of juvenile males and 51.1% of juvenile females in 1993, 54.7% of juvenile males and
70.1% of juvenile females in 1994, 62.7% of juvenile males and 73.1% of juvenile females

in 1995, and 58.9% of juvenile males and 63.9% of juvenile females in 1996.

E VAL UA IION 0F DISPERSAL STA T US
I deﬁned the natal area of each individual as a circle with a radius of 80 meters

centered at the natal burrow, based on maximum size of maternal territories when young

17

were weaned (Holekamp, 1984a, 1986). To determine ages at which juveniles dispersed,
I observed each marked litter in the study area at least twice weekly. In the morning I
noted whether marked squirrels emerged from burrows in their natal areas, and in the
evening I noted whether they returned to their natal areas for the night. I considered
juveniles to have disappeared during the week-long age interval in which they were no
longer observed in their natal areas. I classiﬁed juveniles that had disappeared as
dispersers only if I later recovered them outside their natal areas. To relocate juveniles
that had disappeared, I systematically trapped and visually searched the study site and its
surrounding areas, and examined all road kills on the nearby highway. During morning
and evening observations, and during daily visual searches of the study area, I recorded

the distances at which juveniles were observed from their natal burrows.

OBSER VA IION 0F BEHA VIOR

To evaluate the intensity of resource competition, I observed the behavior of
reproductive females at the study site for a total of 143 hours during the summer of 1995
using Altmann’s (1974) focal animal survey technique and Holekamp’s (1983) ethogram
for S. beldingr’. Speciﬁcally, I deﬁned vigilant behavior as posting or alarm calling, and
aggressive behavior as threatening, chasing, attacking, or ﬁghting a conspeciﬁc. I
attempted to observe each marked reproductive female in the study site for 20 minutes at
least twice weekly, during the morning and late aftemoon when squirrels were most active
(Docekal, 1978). I calculated rates of vigilant and aggressive behavior for individuals as
the number of minutes during which each behavior was observed per hour of observation.

I determined these rates for females during two week intervals comprising the ﬁrst and

18

second halves of both gestation and lactation, and included in calculations only those
females for which I had a total of at least 20 minutes of behavioral observations during the

relevant time interval.

ESTIMATION OF POPULA TION DENSITY

I calculated the densities of females that weaned litters in the study area in 1993
and 1994. I noted the initial appearance above ground of all litters weaned in these areas
during regular observations. I used natural landmarks such as stands of trees and bushes,
ponds, creeks, lakeshore, rock outcrops and other areas not inhabited by S. beldr‘ngi to
delineate the boundaries of local sites within the study area. I estimated geometric areas
of these sites using aerial photographs obtained from the United States Geological Survey
(U SGS), and direct measurements obtained using a 30 meter measuring tape. I calculated
densities as the total number of females that weaned litters per hectare.

I also monitored overall density of S. beldingr‘ in the study area in 1993 and 1994.
I conducted scan samples in the study area hourly during regular observations. During
each scan, I counted all S. beldingr‘ present in a transect of approximately 800 m2 entirely
within the areal was watching. I calculated geometric areas of transects using USGS
aerial photographs and measurements obtained with a 30 meter measuring tape. 1
calculated average density per hectare prior to and after the ﬁrst appearance of juveniles

above ground each year.

19

PRO VISIONING

To manipulate body mass and body fat of juveniles, I provided them with extra
food. I gave additional food to juveniles from 15 litters in 1993, 18 litters in 1994, 21
litters in 1995, and 20 litters in 1996. I provided these juveniles daily with sunﬂower
seeds (700-1000 grams/litter), peanuts (3 50-500 grams/litter), or peanut butter (75-150
grams/litter) ad lib. Sunﬂower seeds were the primary food supplement in 1993 and
1994, and peanuts were the primary supplement in 1995 and 1996. I gave juveniles
peanut butter occasionally when I had no sunﬂower seeds or peanuts. I provided
supplemental food in plywood feeding boxes (25 by 20 by 10 cm) open at one end, and
situated 2 to 3 meters from the entrance of the natal burrow. I covered the feeding boxes
to prevent access by birds.

Reproductive female S. beldingi who receive supplemental food channel excess
energy to their offspring, which are heavier at weaning than offspring of unprovisioned
mothers (Trombulak, 1991). Therefore, I provided each adult female in provisioned areas
daily with extra food (700-1000 grams of sunﬂower seeds, 350-500 grams of peanuts, or
75-150 grams of peanut butter) on her maternal territory from mid-gestation through
lactation. In addition, following their ﬁrst appearance above ground, members of all litters
weaned by provisioned females were further provisioned in the natal area until they were
10 weeks of age. Provisioned areas were separated by at least 100 meters from areas
inhabited by unprovisioned mothers and their young to ensure that unprovisioned
individuals did not have free access to supplemental food. Provisioned and unprovisioned
areas of the study site were alternated between years to control for local environmental

effects.

20

EVALUATION OF BODY FA T

I measured body fat of juvenile S. beldingr' during the summers of 1994-1996 with
a device that analyzes body composition of small animals (model SA-2, EM-SCAN,
Springﬁeld, IL). This device uses electrical conductivity of the body to non-destructively
estimate lean (fat-free) mass of live animals (W alsberg, 1988). I calculated mass of body
fat for live subjects as the difference between total body mass and lean mass, and
expressed fat as a percentage of total body mass.

1 formulated a species-speciﬁc equation for S. beldingi to predict lean mass ﬁom
readings of electrical conductivity of the body (Scott et al., 1991). This equation was
generated from data acquired from 11 squirrels captured near my study site in May and
June of 1994, and from four squirrels captured in August of 1995. I obtained readings of
electrical conductivity from these squirrels, sacriﬁced them via anesthetic overdose, and
stored them at -20° C for approximately 5 months. I dried whole carcasses to constant
mass in a convection oven at 60° C, and then homogenized each carcass. I determined
mass of body fat of carcasses by ether lipid extraction in a Soxtec apparatus (System HT
1043 Extraction Unit, Tecator AB, Hogana, Sweden). Lipid content of carcasses was
similar to that reported by Morton (1975) for S. beldingi captured during May and June,
1970-1972 (approximately 5-10 grams). I obtained lean mass by subtracting fat mass
from mass of total body. My prediction equation was calculated ﬁ'om a linear regression
of lean mass on readings of electrical conductivity (Scott et al., 1991) for S. beldingi
(Figure 1). My determinations of the percentage of body mass composed of fat for

juveniles were similar to those Morton et al. (1974) obtained by solvent lipid extraction

21

  
 
     

 

250 -
y = 0.60 X + 22
ﬁ=os4
F=210.5 . 0'
zoo — df=1 .
’0? P<0.001
E
‘3
5?} 150 —
(D
(D
co
E
m 100 —
cu
t:
+1.
or
u.
50 a
0 — I I I I I I I I

 

0 50 100 150 200 250 300 350 400
Conductivity index

Figure 1 - Regression of fat-free mass on conductivity index in S beldingi.
Values of the conductivity index were generated by a device which
estimates lean body mass by evaluating electrical conductivity of the body.

22

(approximately 2% lipid before pre-hibemation fattening and 20-25% lipid at peak body mass).
For analysis of body composition of S. beldingi, I live-trapped subject animals and

carried them 200-300 meters to a location on the periphery of the study site. I

anesthetized squirrels with methoxyﬂurane (Metofane, Pitman-Moore, Mundelein, IL) in a

two liter anesthesia chamber before measuring body composition. After recovering from

anesthesia (30-90 minutes), squirrels were released at their capture sites. The mean

number times over the summer that body composition was measured in each juvenile was

3.9 1: 1.8 (SD) in 1994, 3.3 i 1.7 (SD) in 1995, and 2.9 (SD) .t 1.5 in 1996.

HORMONE IREA IMEN T S

During the summers of 1993, 1995 and 1996 I treated newborn females with
testosterone (T) to evaluate the effects of perinatal androgen exposure on dispersal
behavior (Holekamp et al., 1984). I trapped pregnant females (1993: n = 16; 1995: n =
69; 1996: n = 51) on their maternal territories and transported them to a lab tent
approximately 16 kilometers from the study site, where they were housed until they gave
birth (1993: 6.1 i 4.1 (SD) days; 1995: 5.2 i 3.2 (SD) days; 1996: 7.9 i 4.4 (SD)
days). Pregnant females were kept individually in cages (60 by 40 by 15 cm), and were
provided with nest boxes (25 by 20 by 10 cm, open at one end) and cotton nest material. I
also gave pregnant females water, peanut butter, mouse chow, and fresh fruit ad lib daily.
I checked nest boxes for the presence of young each day early in the morning, at midday,
and in the evening. The date that young were ﬁrst observed in a nest box was considered

to be their date of birth. At birth, all females of a litter were injected subcutaneously either

23

with T (500 ug in 0.05 ml sesame oil) or with sesame oil alone (0.05 ml). Males of a litter
were not treated.

I returned each mother to her maternal territory with her litter within 24 hours of
treatment, on the day after birth. I transported each mother along with her young and
cotton nest material back to the study area in a small cooler, which was placed on the
maternal territory at 0600 hours and allowed to settle for one hour. The cooler was then
opened, after which the mother foraged and explored the surrounding area, usually for 30-
60 min, before carrying her young and nest material to her maternal burrow. I stayed
within three meters of the cooler until the mother came back for the last of her young. I

warmed the young and defended them from predators and conspeciﬁcs when necessary.

BLOOD COLLECTION AND HORMONE ASSA YS

To evaluate circulating T levels of juveniles in relation to dispersal, I collected 89
blood samples (0.5-1.0 ml) from 71 juvenile males and 62 samples from 52 juvenile
females in 1996. I anesthetized squirrels with methoxyﬂurane near their capture locations,
and collected blood via cardiac puncture. I released squirrels at their capture sites after
they recovered ﬁom anesthesia (30-90 minutes). I stored blood in liquid nitrogen (-1960
C) for one to two months until I returned to Michigan, at which time I stored it at -20° C
for an additional month. I then measured plasma T with I125 radioimmunoassay, using
Coat-A-Count Total Testosterone assay kits (Diagnostic Products, Los Angeles, CA).

Minimum assay sensitivity was 0.2 ng/ml.

24

STATISTICAL ANALYSES OF DA TA

1 monitored various dependent measures, including dispersal status, body mass,
percentage of body mass composed of fat, age at dispersal, hourly rates of behavior,
minimum distance ﬁom the natal burrow, litter size, and the number of females per litter.

I analyzed frequency data (e. g., dispersal status) with chi-square tests or two-tailed
Fisher’s exact tests, depending on sample sizes (Sokal and Rohlf, 1981). I evaluated
normally distributed body mass data with Analysis of Variance (ANOVA). Values of
percent body fat were not distributed normally, so I transformed them to values that
followed the normal distribution using the angular transformation, and analyzed these data
with ANOVA. I analyzed discrete data (e. g., age at dispersal, litter size, number of
females per litter) and data that were not normally distributed (e. g., rates of behavior) with
nonparametric tests. I deﬁned differences between groups to be signiﬁcant when P 5
0.05. I adjusted signiﬁcance levels of multiple pairwise comparisons with Holrn’s
procedure (Neter et al., 1996). Mean values of body mass, body fat, rates of behavior,
and distance from the natal burrow are presented 1 one standard error.

Most parametric and non-parametric statistical tests assume independence of
observations. However, studies of ﬁ'ee-living animals over time routinely yield multiple
observations from the same individuals in the same and different time intervals (Hoogland,
1995). In my statistical analyses, I assumed dependence of data obtained from the same
individual in the same time interval, and so averaged all data for each individual within
each time interval. I assumed independence of data from the same individual in different

time intervals (see Hoogland, 1995). Although data from the same individual are not truly

25

independent across time intervals, in ﬁeld studies of free-living animals they generally act
as though they were (Dobson, personal communication).

In many of my statistical analyses, I used ANOVA to evaluate data over a series of
age or time intervals. Although my data act as though they were independent across these
intervals, they are not truly independent and so violate one of the basic assumptions of
ANOVA. Thus, the results of my analyses with AN OVA should be interpreted with some

caution.

Chapter 3

ONTOGENY OF MASS AND FAT ACQUISITION AND DISPERSAL

INTRODUCTION

Body mass, body size, and body condition may all exert important effects on life
history traits of mammals (e. g., Iverson and Turner, 1974; Armitage et al., 1976; Barnes,
1984; Sauer and Slade, 1987; Trombulak, 1989, 1991; Kenagy et al., 1990a, 1990b; Dark
et al., 1992; Bachman, 1993). It has been proposed that these variables inﬂuence temporal
patterns of natal dispersal in various sciurid rodents. For example, late dispersal by
smaller or lighter than average individuals, or by individuals in poor condition has been
observed in Uinta ground squirrels (S. armatus; Balph and Stokes, 1963), round-tailed
ground squirrels (S. teretr'caudus; Dunford, 1977), Belding’s ground squirrels (S. beldingi;
Holekamp, 1984a), yellow-bellied marrnots (Marmotaﬂaviventris; Downhower and
Armitage, 1981), and Gunnison’s prairie dogs (Cynomys gunnisoni; Rayor, 1985).
Holekamp (1986) noted that of eleven variables examined in juvenile S. beldingi, body
mass was the most reliable predictor of an individual’s dispersal status at any time. On the
basis of this relationship between body mass and dispersal status, she proposed an
hypothesis suggesting that dispersal by male S. beldingi is triggered by attainment during
ontogeny of a particular body mass or body composition, or some combination of these

two variables.

26

27

Here I evaluated the relationship between the ontogenetic development of body
mass, body fat, and dispersal behavior in juvenile male S. beldingi. I manipulated body
mass and body fat of juveniles by provisioning them with food, and used unprovisioned
juveniles as controls. I monitored dispersal status of provisioned and unprovisioned
juveniles throughout the active season. My results suggest that natal dispersal in S.
beldingi is constrained by the energetic demands of depositing fat for hibernation. My
results ﬁrrther indicate that juvenile male S. beldingr' postpone dispersal until they have
attained a threshold body mass, and that the percentage of body mass composed of fat

may importantly inﬂuence the timing of dispersal.

METHODS

I trapped and ear-tagged 135 juvenile males from 65 complete litters during the
summers of 1993 and 1994. I gave extra food to 35 males from 15 litters in 1993, and 40
males from 18 litters in 1994. Iused 15 males ﬁ'om 10 litters in 1993 and 45 males ﬁ'om
22 litters in 1994 as unprovisioned controls. I measured body mass of these males weekly
from 4-11 weeks of age in 1993 and 1994, and also monitored dispersal status weekly
from 6-10 weeks of age and recorded minimum distances from the natal burrows at which
I observed juveniles from 6-10 weeks of age. In 1994, I also measured body fat weekly

from 6-11 weeks of age.

RESULTS
Body mass did not differ signiﬁcantly between 1993 and 1994 for either

provisioned males (t-test for independent samples, P = 0.649) or unprovisioned males (t-

28

test for independent samples, P = 0.126). Therefore, body mass data for the two years
were combined for subsequent statistical analyses. Provisioned males were signiﬁcantly
heavier than unprovisioned males throughout the active season (Figure 2; ANOVA, F =
98.7, df= 1, P < 0.001).

Mean minimum distances from the natal burrow at which I observed juveniles
increased steadily ﬁ'om six to ten weeks of age for both provisioned males (Figure 3;
Kruskal-Wallis test, test statistic = 52.455, P < 0.001) and unprovisioned males (Figure 3;
Kruskal-Wallis test, test statistic = 71.358, P < 0.001). Mean minimum distances from the
natal burrow were signiﬁcantly greater overall for provisioned than unprovisioned males at
6-10 weeks of age (Figure 3; Kruskal-Wallis test, U = 757.5, P = 0.012). In particular,
provisioned juveniles were observed signiﬁcantly farther ﬁ'om the natal burrow at 7 weeks
of age (Figure 3; Kruskal-Wallis test, U = 757.5, P = 0.002).

Ages at which dispersal occurred also differed signiﬁcantly between provisioned
and unprovisioned males (Figure 4; f = 13.437, df= 4, P = 0.009). Provisioned males
began dispersing at a younger age than did control males (Figure 4; 6 weeks vs. 7 weeks),
and had a younger modal age of dispersal (Figure 4; 7 weeks vs. 8 weeks). A larger
cumulative percentage of provisioned than unprovisioned males had dispersed during each
age interval ﬁ'om six to ten weeks of age, and signiﬁcant differences were observed from
six to eight weeks of age (Figure 5; 6 weeks: Fisher’s exact test, P = 0.038; 7 weeks:
Fisher’s exact test, P < 0.001; 8 weeks: 1’ = 5.291, df= 1, P = 0.021). Mean body mass
at dispersal did not differ signiﬁcantly between provisioned and unprovisioned males
(192.0 1 10.1 grams, n = 10, vs. 183.0 1 7.4 grams, n = 15, respectively; t-test for

independent samples, P = 0.470). Body mass at dispersal was constant at about 185

29

300 1

+ Provisioned
—D— Unprovisioned 17 15
250 —-
A
(D
E
E zoo —
3’
(I)
R
E 150 —
>4
.0
o
m
100 -
so -—

 

 

 

I I I I I I I I
4 5 6 7 8 9 10 11

Age (weeks)

Figure 2 - Body mass of juvenile males, 1993-1994. Sample sizes are given
on the graph.

30

  
   

 

 

A 200 . .
E 7 + Provrsroned 34 31
E 175 _ —D—~ Unprovisioned
o
g 34
.o 150 —
E
g 125 —
8
S 100 —
“<73 _ 37
:6 75
E
3 so 5
E
C 55
.— 25 a
E
E o D///-g
T 45
g 53
l l l l l
6 7 8 9 10
Age (weeks)

Figure 3 - Distances of juvenile males from their natal burrows, 1993-1994.
The asterisk indicates a signiﬁcant difference between provisioned and
unprovisioned males. Sample sizes are given on the graph.

31

10 _ —O— Provisioned

 

 

 

—D— Unprovisioned
8 ~ I
c»
.E
1’.’
a) 6 ~
0.
.12
'0 I
L...
m
.o 4 —
E
Z I
2 - I
a
0 Hr I I l l
6 7 8 9 10
Age (weeks)

Figure 4 - Dispersal of juvenile males at various ages, 1993-1994.

32

60—

50-

40—

30—

  

20 —
—o— Provisioned
—1:1— Unprovisioned

Cumulative % dispersed

 

 

I I I
6 7 8 9 10

Age (weeks)

__1
——(

Figure 5 - Cumulative dispersal of juvenile males, 1993-1994. Asterisks
indicate a signiﬁcant difference between provisioned and unprovisioned
males. Samples included 34 provisioned and 41 unprovisioned males.

33

grams, regardless of differences among individuals in age at dispersal (Figure 6; ANOVA,
P = 0.600). This indicates that there may be a mass threshold for dispersal. Provisioned
and unprovisioned juveniles did not differ in body mass relative to time of dispersal
(Figure 7; ANOVA, P = 0.850). This ﬁrrther indicates that body mass importantly
inﬂuences the timing of dispersal.

Seven males in the 1993 cohort and 17 males in the 1994 cohort were known to be
alive at 55 weeks of age. All of these males, in both provisioned and unprovisioned
groups, emigrated from their natal areas before reaching 55 weeks of age. Thus, food
provisioning did not alter the probability that males would eventually leave their birth sites.

Fat, as a percentage of body mass, was greater in provisioned than unprovisioned
juvenile males overall from 6-11 weeks of age (Figure 8; ANOVA, F = 19.554, df= 1, P
< 0.001). In particular, provisioned males were fatter than their unprovisioned
counterparts at 6 and 7 weeks of age (Figure 8; t-tests for independent samples; 6 weeks:
t= -4.214, df= 31, P < 0.001; 7 weeks: t= -4.053, df= 33, P <0.001). There was a
marginally signiﬁcant difference in body mass between provisioned and unprovisioned
males at 8 weeks of age (Figure 8; t-test for independent samples; t= -1.905, df= 29, P =
0.067). At ages 9-11 weeks, however, body fat of provisioned males was similar to that
of unprovisioned males (Figure 8). Thus, food availability or food quality appears to be a
strong determinant of how much body fat a male has early, but not late, during his juvenile
summer. Moreover, there appears to be a close association between body fat and
dispersal. A greater proportion of provisioned than unprovisioned males dispersed early
during ontogeny, by 6-8 weeks of age, when provisioned males tended to be fatter than

unprovisioned males. However, after 8 weeks of age, there was no signiﬁcant difference

34

250 -

200 *

.5

01

o
1

Body mass (grams)

50+

 

 

7 8 9 10
Age at dispersal (weeks)

Figure 6 - Body mass of juvenile males at dispersal, 1993-1994. Sample
sizes are given on the graph.

35

300 —

l
—o— Provisioned : 5
—D—- Unprovisioned I
I
250 — i
A I
w I
E
53 200 -
(I)
(D
(11
E 150 _,
>4
.0
o
m
100 -
so

 

 

 

Time relative to dispersal (weeks)

Figure 7 - Body mass of juvenile males relative to dispersal, 1993-1994.
The vertical dashed line indicates the week of dispersal. Samples sizes
are given on the graph.

36

20-

 

 

 

+ Provisioned 9
—D— Unprovisioned

’61?

g 15 —

E

>4

'0

o

.o

“5 1o -

o\°

4.0

.59

>4

8 5‘

In

0 _ I I I I I I
6 7 6 9 1o 11
Age (weeks)

Figure 8 - Body fat of juvenile males, 1994. Asterisks indicate a signiﬁcant
difference between provisioned and unprovisioned males. Sample sizes are
given on the graph.

37

between provisioned and unprovisioned males either in the proportion dispersed or in
body fat.

During 1994, I recovered 25 dispersed juvenile males, Twenty-one of these (84%)
emigrated between 27 July and 25 August. Fat, as a percentage of body mass, increased
steadily in these juveniles, and differed signiﬁcantly among 10-day intervals within this
period (Figure 9; repeated measures ANOVA; provisioned: F = 3.757, df= 2, P = 0.043;
unprovisioned: F = 13.119, df= 2, P < 0.001). Thus most dispersal behavior occurred
during a period of rapid fat increase.

I compared body fat of males dispersing during the 7 to 10 week age intervals with
that of males still residing in their natal areas. Fat, as a percentage of body mass, was
greater overall for dispersing males than for males that had not yet dispersed Gigure 10;
ANOVA, F = 7.851, df= 1, P = 0.006), and it was signiﬁcantly greater in dispersing
males than in undispersed males at 8 weeks of age (Figure 10; t-test for independent

samples, t = 3.490, df= 23, P = 0.002).

DISCUSSION

Patterns of natal dispersal by juvenile male S. beldingr‘ in 1993-1994 were similar
to those observed by Holekamp (1984a). In both studies, natal dispersal began at 6-7
weeks of age, mean minimum distances at which juvenile males were observed from their
natal burrows increased steadily throughout the active season, and 100% of surviving
males dispersed before 55 weeks of age. This suggests that food provisioning left intact

the normal dispersal behavior of males. This conﬁrms prior studies, in which provisioning

38

N
N
l

+ Provisioned
—D—~ Unprovisioned

-l N
on O
l l

.3
O)
l

.L
N
1

Body fat (% of body mass)

 

 

(D
l

 

 

 

6 I I r
27 July-5Aug 6-15Aug 16-26Aug

Figure 9 - Body fat of juvenile males during the primary dispersal period, 1994.
Samples included repeated measures for 7 provisioned and 14 unprovisioned
males.

    

 

  

6

1 1:] Males in natal area

_ W Dispersing males

 

 

00000000000

2 1 1 1 1 1

Age (weeks)

ed

ng and natal males, 1994. Mean body fat
erval for males dispersing during that
siding in their natal areas. The asterisk
ce between dispersing and natal males.

graph. Samples include both provision

umm
6.2mm
wgne
..maﬁ.ﬁ.ms.
finnvsmne
te mvm
a .m.
fgmcqod
y.nrw‘mem
wmmmcmm
de.lss
.wmumw
0m S.$m
e

s

40

did not affect the probability of dispersal by male S. beldingr' (Trombulak, 1991) or by
male California ground squirrels (S. beecheyi; Dobson, 1979).

Holekamp (1984a, 1986) suggested that there was a relationship between body
mass and dispersal in juvenile male S. beldingr’, but noted that body mass alone could not
explain the timing of natal dispersal, because some heavier than average S. beldingi
delayed dispersal. She proposed that body composition also inﬂuences the timing of natal
dispersal, and that juvenile male S. beldingi disperse only after attainment of some
threshold of body mass or body fat. In this study, food provisioning caused an advance in
the timing of dispersal behavior. I suggest that this effect was mediated, at least in part,
by body fat or some variable correlated with body fat.

My results support the hypothesis that there is a mass threshold for dispersal in S.
beldingi. Provisioned males were heavier than same-aged unprovisioned males, and
dispersed at younger ages. However, relative to time of dispersal, body mass did not
differ between provisioned and unprovisioned males. In addition, body mass at dispersal
was fairly constant among juvenile males, regardless of the ages at which they dispersed.
Thus juvenile males may need to attain a minimum degree of physical grth and
development, as reﬂected in body mass, prior to dispersing.

Body fat appears to exert a strong inﬂuence on the timing of dispersal. A greater
proportion of provisioned than unprovisioned males dispersed in the 6-8 week age
intervals, when provisioned males were fatter than their unprovisioned counterparts.
However, ﬁom 9 to 10 weeks of age, neither body fat nor the cumulative percentage of

males dispersed differed signiﬁcantly between provisioned and unprovisioned males.

41

Moreover, body fat was greater in dispersing males than in same-aged males still residing
in their natal areas.

Juvenile S. beldingr' enter hibernation when body mass is about 200-260 grams,
and when body fat is about 20-25% of total body mass (Morton et al., 1974). Attainment
of a speciﬁc minimum body mass and percentage of body mass composed of fat by the end
of the summer for juveniles may be critical to both the survival and future reproductive
success of male S. beldingr’. Overwinter survival of juveniles is positively correlated with
pre-hibemation body mass in S. beldingr' (Trombulak, 1991), and other hibernators (e. g.,
Sullivan and Sullivan, 1982; Sullivan et al., 1983; Murie and Boag, 1984; Dobson and
Kjelgaard, 1985). Furthermore, a positive correlation between body mass and mating
success has been observed in S. beldingi (Sherman, 1976). Energetic costs of dispersal
behavior may slow the storage of body fat required to sustain the young animal
throughout hibernation. Juvenile males may therefore delay dispersal until they achieve
sufﬁcient body mass and body fat to prevent these costs from compromising mass and fat
acquisition necessary during the juvenile summer.

Early dispersal ﬁ'om the natal site and immigration into a new area may confer a
variety of advantages on juvenile mammals, as it has been proposed to do in some species
of birds. Juvenile crested tits (Parus cristatus) that disperse early are more likely to settle
in favorable habitats, whereas late dispersers are more likely to immigrate into marginal
habitats (Lens and Dhont, 1994). Juvenile male marsh tits (Parus palustris) are less apt to
be targets of aggression and are more successful at establishing themselves in new areas

when they disperse at early ages (Nilsson and Smith, 1988; Nilsson, 1989a, 1989b). Thus,

42

plasticity of dispersal behavior in response to body mass and body fat may enable juveniles
to coordinate physical and physiological readiness with optimal timing of dispersal.

Body mass, body fat, and availability of metabolic ﬁrels have been shown to
inﬂuence the timing of energetically expensive life history events such as reproduction
(e.g. French, 1982; Barnes, 1984; Bushberg and Holmes, 1985; Wade and Schneider,
1992). Little is known about the actual energetic costs of dispersal behavior. However,
the apparent sensitivity of dispersal behavior in S. beldingi to inﬂuences of body mass and
body fat suggests that energetic requirements may constrain dispersal in S. beldingi and
other mammals, as they do reproduction. Natal dispersal is associated with increased
locomotor and exploratory activity (Holekamp, 1986; Ritchison et al., 1992). In addition,
dispersers often encounter elevated levels of conspeciﬁc aggression when attempting to
immigrate into new areas (Holekamp, 1986; Pusey and Packer, 1987a). Thus, dispersing
males may need to reallocate time spent foraging to vigilant and defensive behavior.
Moreover, dispersing individuals may need to be more vigilant as they explore unfamiliar
areas, where they may be more vulnerable to predation (Metzgar, 1967; Snyder et al.,
1976). Bachman (1993) observed a behavioral tradeoff in S. beldr'ngi in which the amount
of time spent alert was inversely proportional to the amount of time spent feeding. Thus,
increases in activity associated with dispersal may not only increase energy requirements,
but may also decrease energy intake.

Kenagy (1980) observed a speciﬁc temporal relationship between circannual cycles
of hibernation and reproduction in male golden-mantled ground squirrels (S. lateralis).
Testicular development in S. lateralis begins during the ﬁnal three months of hibernation,

and is complete when hibernation ends. Energy to support hibernation and testicular

43

development is derived from stored fat. Males also rely on fat reserves during mating,
which occurs immediately after emergence from hibernation. Testicular development is
inhibited in underweight S. lateralis, which apparently do not have sufficient fat reserves
to support both hibernation and reproduction (Barnes, 1984). Thus, the energy demands
of hibernation and reproduction conﬂict in S. lateralis.

My results here suggest that a similar conﬂict occurs in S. beldingi between the
energy demands of dispersal and those of hibernation. Most dispersal by juvenile males in
1993 and 1994 occurred when the percentage of body mass composed of fat was rapidly
increasing, which suggests that dispersal competes with pre-hibemation fattening for
available energy. Dispersal might be suppressed when, for a juvenile with a given
percentage of body mass composed of fat, insufﬁcient time remains in the active season to
take in enough energy both to ﬁrel dispersal behavior and to build fat reserves adequate

for overwinter survival.

Chapter 4

INTERACTION OF TIME OF YEAR AND ENERGETIC VARIABLES

INTRODUCTION

Little is known about the mechanisms that regulate the timing of dispersal
behavior. However, Smale et al. (1997) suggested that in species such as S. beldingi,
which exhibit seasonal cycles of reproduction and behavior, the timing of dispersal might
be inﬂuenced by circannual or other seasonal time-keeping mechanisms. Moreover, the
amount of energy stored as fat should theoretically have an important inﬂuence on the
timing of dispersal. Animals should delay dispersal until their fat reserves are adequate to
meet the potential energy requirements of the dispersal movement (Nunes and Holekamp,
1996). Dispersing juvenile male S. beldingr' do in fact tend to be fatter than same-aged
males still residing in the natal area (Nunes and Holekamp, 1996). Here I explored the
possibility that fat-related energetic variables interact with circannual or other seasonal
time-keeping mechanisms to mediate dispersal in juvenile male S. beldingr'.

Many behaviors, physiological traits, and life history events in ground-dwelling
squirrels exhibit circannual rhythmicity (e. g., Davis, 1976; French, 1982; Barnes and York,
1991; Zucker et al., 1991). Similarly, body mass in ground squirrels undergoes
predictable, endogenously driven annual rhythms, which persist in the absence of
environmental cues (e. g. Pengelley and Fisher, 1963; Davis, 1976; Mrosovsky, 1976;
Mrosovsky and Powley, 1977; Zucker and Boshes, 1982; Smale et al., 1986; Zucker et al.,

44

45

1991). Changes in the size of fat stores are responsible for these changes in body mass
(Dark et al., 1984, 1985; Forger et al., 1986). In juvenile S. beldingr', mass and fat appear
to be regulated at seasonally appropriate levels that steadily increase prior to hibernation,
and gradually decrease during hibernation (French, 1982; Bachman, 1993, 1994; Nunes
and Holekamp, 1996).

Dispersing and depositing fat in preparation for hibernation both present energetic
challenges to S. beldingr'. Increases in activity associated with dispersal may compromise
an individual’s capacity to gain fat needed for survival during winter. Thus, at the phase
of the annual cycle during which squirrels prepare for hibernation, dispersal and fattening
might conﬂict with each other, and emigration might be delayed when it would
compromise an individual’s ability to acquire fat stores necessary for overwinter survival
(Nunes and Holekamp, 1996). Circannual or other time-keeping mechanisms might be
involved in the physiological processes determining whether an individual disperses or
uses available energy to build fat reserves.

Heavy snowfall during the spring and summer of 1995 substantially delayed the
active season of S. beldingi compared to its active seasons in 1994 and 1996. This
fortuitous situation allowed me to examine an hypothesis suggesting that energy stores
and time of year interact in juvenile male squirrels to mediate dispersal behavior. In
particular, I tested the prediction of the hypothesis that juveniles having little time to fatten
for hibernation should use their time and energy to build fat reserves necessary for
overwinter survival rather than for dispersal. Thus, these animals should delay emigration
from the natal area until the yearling summer. By contrast, individuals having time to

store sufficient fat to support both hibernation and dispersal should emigrate as juveniles.

46

METHODS

In this phase of my research I examined body mass, body fat and dispersal status of
juvenile male S. beldingi that were unprovisioned. During the summers of 1994-1996, I
trapped and ear-tagged 182 juvenile males from 79 complete litters. I marked 45 males
ﬁ'om 22 litters in 1994, 84 males from 33 litters in 1995, and 53 males from 24 litters in
1996. I measured body mass weekly ﬁom 4-11 weeks of age, and body fat weekly ﬁom
7-10 weeks of age. I also monitored dispersal status of juveniles each week from 6-10

weeks of age in 1994 and 1995, and from 6-9 weeks of age in 1996.

RESULTS

Compared to 1994 and 1996, initial emergence of litters from natal burrows was
substantially delayed by heavy spring snowfall in 1995 (Figure 11). Litters emerged from
2-22 July in 1994, from 10-24 Aug in 1995, and from 4-26 July in 1996. The median date
of litter emergence was 7 July in 1994 and 9 July in 1996, but was considerably later, 17
Aug, in 1995.

The number of young that females weaned in 1995 was not affected by delays
resulting from heavy spring snowfall. Mean litter size did not differ signiﬁcantly between
years ﬁom 1994-1996 (ANOVA; P = 0.567). Mean litter size was 4.182 : 0.340 (SE) in
1994, 4.667 : 0.353 (SE) in 1995, and 4.667 : 0.339 (SE) in 1996.

Body mass of juveniles overall ﬁom 4-11 weeks of age differed signiﬁcantly
between years from 1994 through 1996 (Figure 12; ANOVA, F = 45.580, df= 2, P <
0.001). I also observed a signiﬁcant interaction of age with year in the comparison of

body mass between years from 1994-1996 (Figure 12; ANOVA, F = 3.136, df= 14, P <

47

- 1994

100

U)
.5 so
9
a)
E
(D 60
E
a:
SF."
"5 4o
40—:
C
a)
3",
0. 2o

 

1-15 16-31 1-15 16-31
July July Aug Aug

Figure 11 - First emergence of litters from the natal burrow, 1994-1996.
Samples included 22 litters in 1994, 33 litters in 1995, and 24 litters in 1996.

48

0.001). In particular, patterns of mass acquisition in 1995 differed ﬁom those in 1994 and
1996. Beginning at about 7-8 weeks of age, male body mass increased more rapidly in
1995 (Figure 12). Juvenile males (aged 8-11 weeks) in 1995 were signiﬁcantly heavier
than were juvenile males in either 1994 or 1996 (Figure 12; t-tests for independent
samples; 1995 vs. 1994: 8 weeks: t= -3.381, df= 63, P = 0.001; 9weeks: t= -3.270, df
= 68, P = 0.002; 10 weeks: t = -3. 167, df= 59, P = 0.002; 11 weeks: I= -3.996, df= 26,
P < 0.001; 1995 vs. 1996: 8 weeks: t= 4.967, df= 56, P < 0.001; 9 weeks: t = 3.743, df
= 55, P < 0.001; 10 weeks: t= 3.872, df= 47, P < 0.001; 11 weeks: t= 3.006, df=10,
P = 0.013). Juvenile females in 1995 also gained body mass more rapidly, beginning at
about 7-8 weeks of age, than did juvenile females in either 1994 or 1996 (Nunes,
unpublished data).

Body fat at ages 7-11 weeks differed signiﬁcantly between years from 1994-1996
(Figure 13; ANOVA, F= 14.862, df= 2, P < 0.001). There was also a signiﬁcant
interaction of age with year in the comparison of body fat between years (Figure 13;
ANOVA, F = 3.417, df= 8, P = 0.001). Starting at about 9 weeks of age, fat deposition
occurred more rapidly in 1995 than in either 1994 or 1996. Body fat at ages 9-11 weeks
was signiﬁcantly greater in 1995 than it was in either 1994 or 1996 (Figure 13; t-tests for
independent samples; 1995 vs. 1994: 9weeks: t= -3.972, df= 58, P < 0.001; 10 weeks:
t= -4.061, df= 57, P < 0.001; 11 weeks: t= -3.508, df= 23, P = 0.002; 1995 vs. 1996:
9 weeks: 1= -4.549, df= 47, P < 0.001; 10 weeks: t= 5.186, df= 45, P < 0.001; 11
weeks: t = 3.031, df= 9, P = 0.014). Juvenile females also fattened more rapidly in

1995, beginning at about 9 weeks of age, than did juvenile females in either 1994 or 1996.

49

275 1

 

 

 

—o— 1994
250 _ + 1995 *
—A— 1996 1
225 —r * ‘
7:?
E zoo -
E
P) 175 -
(D
3 150
E O
'5‘ 125 5 * /A
o
m A-
100 - /'
75 — /2
5o - 2
l l l l T l l l
4 5 6 7 8 9 1o 11
Age (weeks)

Figure 12 - Body mass of juvenile males, 1994-1996. Asterisks indicate
that body mass in 1995 differed signiﬁcantly from than in both 1994 and
1996. Sample sizes are as follows: 1994: 42, 25, 29, 30, 27, 27, 25, 22;
1995: 83, 48, 37, 45, 38, 43, 36, 8; 1996: 45, 32, 23, 20, 20, 14, 13, 4.

50

x.

 

 

 

25 — —o— 1994

—I— 1995
—A— 1996 *

7,; 20 -

(I)

(U

E *

>5

'0 15 —

o

.0

q—

O

3% 1o —

.52

6‘

o 5 —

m

0 _
I l l l l
7 a 9 1o 11
Age (weeks)

Figure 13 - Body fat of juvenile males, 1994-1996. Asterisks indicate
that body fat in 1995 differed signiﬁcantly from that in both 1994 and

1996. Sample sizes are as follows: 1994: 15, 15, 23, 24, 20; 1995:
22, 34, 37, 35, 7; 1996: 18, 16, 12, 12, 4.

51

I evaluated dispersal behavior of juveniles in the 1994-1996 cohorts whose
dispersal status was known at 9 weeks of age. Compared to 1995, juvenile males were
more likely to disperse at 7-9 weeks of age in 1994 (Figure 14; Mantel-Haenszel test, 2’ =
16.058, P < 0.001), and in 1996 (Figure 14; Mantel-Haenszel test, , X = 28.932, P <
0.001). However, dispersal behavior at 7-9 weeks of age did not differ signiﬁcantly
between 1994 and 1996 (Figure 14; Mantel-Haenszel test, ,1? = 1.908, P = 0.167).

I also evaluated dispersal of males in the 1994 and 1995 cohorts whose dispersal
status was known at 10 and 55 weeks of age. A greater proportion of juvenile males
dispersed by 10 weeks of age in 1994 than in 1995 (Figure 15; ,1? = 12.286, df= 1, P <
0.001). All males of the 1994 and 1995 cohorts that were still residing in their natal areas
at 10 weeks of age, and that were known to have survived their ﬁrst winter, were still
present in the natal area at the start of the yearling summer. Moreover, all surviving males
of these cohorts dispersed by 55 weeks of age (Figure 15). Thus, most males born in
1995 postponed dispersal until their yearling year. Rates of survival through the ﬁrst
winter did not diﬂ‘er signiﬁcantly between males in the 1994 and 1995 cohorts (10/30 vs.
13/51, respectively; ,1? = 0.571, df= l, P = 0.450).

I examined the inﬂuence of calendar date on body fat of juvenile male S. beldingi,
beginning when juvenile males ﬁrst reached 6 weeks of age. I monitored body fat during
10-11 day intervals ﬁ'om 10 July to 31 August in 1994 and 1996, and from 21 August to 5
October in 1995 (Figure 16). I ordered these intervals chronologically from ﬁrst to ﬁfth,
and compared body fat between years in corresponding intervals. Overall, body fat
differed signiﬁcantly between years from 1994-1996 (Figure 16; ANOVA, F = 6.903, df=

2, P = 0.001). I also observed a signiﬁcant interaction of time interval with year in my

52

 

 

 

6° 7 —o— 1994
+ 1995
50 _ —A— 1996 ‘
'0
m
2
3 4o —
.92
.0
°\° 30 —~ ’
a)
.2
E .
E 20 —« .
:3
0
1o —
6
0 A/I l
5 7 8 9
Age (weeks)

Figure 14 - Dispersal of juvenile males, 1994-1996. Samples included
32 males in 1994, 51 males in 1995, and 23 males in 1996.

53

1994 cohort
E: 1995 cohort 9 13

 

100 4

80-

60-

40-

Cumulative % dispersed

20-

 

 

 

10 weeks 55 weeks
Age

Figure 15 - Dispersal of males in the 1994 and 1995 cohorts. The
cumulative percentage of individuals dispersed is presented for juvenile
males at 10 weeks of age and for yearling males at 55 weeks of age. The
asterisk indicates a signiﬁcant difference between the 1994 and 1995
cohorts. Sample sizes are given on the graph.

54

 

 

2° " —o— 1994 39 13
—l— 1995
—Ar— 1996
7,; 21
g 15 — 25
E e
5‘ ' ‘
Q 43
“5 10 -* 18
5 5
"“3 11 i
F: 5 ._ ' 53
m ,3 16
18 30 17
32
O
o — 7
l l l T l l l l l
-15- -25- -5- -15- -25- -5- -15- -25- -5-
JUL AUG SEPT OCT

Figure 16 - Body fat of juvenile males during the active season, 1994-1996.
Changes in body fat over time are presented. Sample sizes are given on
the graph.

55
comparison of body fat (Figure 16; ANOVA, F = 4.348, df= 8, P < 0.001). I compared
body fat in corresponding time intervals between years using t-tests for independent
samples, and adjusted signiﬁcance levels using Holm’s procedure (Neter et al., 1996).
Body fat of juvenile males in 1994 and 1996 differed only during the fourth time interval
(Figure 16; t = 2.993, df= 39, P = 0.005). However, juvenile males were heavier in 1995
than in 1994 during the third through ﬁﬁh time intervals (Figure 16; third interval: t= -
3.727, df= 71, P < 0.001; fourth interval: 1= -2. 151, df= 62, P = 0.035, ﬁfth interval: t
= -2.408, df= 25, P = 0.024). This difference in body mass between 1994 and 1995 was
statistically signiﬁcant during the third interval, but was only marginally signiﬁcant during
the fourth and ﬁfth intervals, after signiﬁcance levels were adjusted. Juvenile males were
heavier in 1995 than in 1996 during the fourth and ﬁfth time intervals (Figure 16; fourth
interval: t = 5.783, df= 53, P < 0.001; ﬁfth interval, t= 2.442, df= 17, P = 0.026). This
mass difference between 1995 and 1996 was statistically signiﬁcant during the fourth
interval and marginally signiﬁcant during the ﬁlth interval, after signiﬁcance levels were
adjusted. Thus, patterns of fat deposition differed between years. In particular, juvenile
males acquired fat more rapidly during the last part of the active season in 1995 than did
males in either 1994 or 1996. Males appear to exhibit accelerated pre-hibemation
fattening when they are born late in the year.

I also evaluated the inﬂuence of calendar date on dispersal behavior in juvenile
male S. beldingi. I monitored the cumulative percentage of juvenile males dispersed
during 10-11 day intervals from 10 July to 31 August in 1994 and 1996, and from 21
August to 5 October in 1995 (Figure 17). I ordered these intervals chronologically from

ﬁrst to ﬁfth, and compared dispersal behavior between years. I observed no difference in

56

dispersal between 1994 and 1996 (Figure 17; Mantel-Haenszel test, 13 = 0.227, P =
0.634). However, dispersal in 1995 occurred at signiﬁcantly lower rates than it did in
1994 (Figure 17; Mantel-Haenszel test, ,1) = 47.171, P < 0.001) and 1996 (Figure 17;
Mantel-Haenszel test, 3’ = 5 1 .43 8, P < 0.001). It appears that dispersal seldom occurs
aﬁer early September in the population of S. beldingi at Tioga Pass, regardless of how
much fat juveniles have managed to store by then (Figure 17).

I used discriminant analysis to evaluate variables that might be reliable predictors
of whether an individual dispersed by the end of the juvenile summer. In particular, I
examined date of birth, body mass, and body fat during the last 10 days of the primary
dispersal period each year (20-30 August 1994, 20-30 September 1995, and 15-25 August
1996). I observed no difference between 1994 and 1996 in date of birth (Kruskal-Wallis
test, P = 0.113), body mass (ANOVA, P = 0.250), or body fat (ANOVA, P = 0.927). I
also observed no interaction of year with age in my comparison of mass and fat between
1994 and 1996 (mass: ANOVA, P = 0.900; fat: ANOVA, P = 0.484). Thus, I combined
data from these years for this analysis. In 1994 and 1996, date of birth, and mass and fat
at the end of the dispersal period were dependable predictors of whether or not an
individual dispersed as a juvenile (date of birth: F = 6.486, df= 1, P = 0.015; mass: F =
8.374, df= 1, P = 0.006, fat: F= 7.436, df= I, P = 0.010). In 1995, however, when all
litters were born very late during the year, date of birth, mass, and fat were not reliable
predictors of dispersal status at the end of the juvenile summer (date of birth: F = 2.445,

df= l, P= 0.125; mass: F= 0.219, df= 1, P: 0.642; fat: F= 0.852, df= 1, P= 0.361).

57

60H

50-

4o—

30-*

20—

Cumulative % dispersed

10-

 

 

 

0 ‘ ‘ T I l l I l i
-15- -25- -5- -15- -25- -5- -15- -25- -5-

JUL AUG SEPT OCT
Date

Figure 17 - Dispersal of juvenile males during the active season, 1994-1996.
Changes over time in the percentage of individuals dispersed are presented
for 32 males in 1994, 51 males in 1995, and 23 males in 1996.

58

DISCUSSION

The active season of juvenile S. beldingi began substantially later in 1995 than in
1994 and 1996. As a result little time was available to males in the 1995 cohort for pre-
hibemation fattening, and more than 90% of these males postponed dispersal until the
yearling summer. Moreover, juvenile males gained body mass and deposited fat more
rapidly in 1995 than in 1994 or 1996. In my prior work examining the inﬂuences of body
mass and body fat on the timing of dispersal by young male S. beldingi, most emigration
by juvenile males occurred during the period of pre-hibemation fattening, when body mass
and body fat were rapidly increasing (Chapter 3; Nunes and Holekamp, 1996). Moreover,
dispersing males had greater percentages of their bodies composed of fat than did same-
aged males remaining in their natal areas (Chapter 3; Nunes and Holekamp, 1996).
Together with these earlier results, my data from 1994-1996 are consistent with the
hypothesis that fat reserves interact with time of year to determine whether a male
disperses during his juvenile summer or one year later. Speciﬁcally, when seasonal time
constraints prevent an individual from acquiring enough energy to fuel both dispersal and
pre-hibernation fat deposition, dispersal appears to be inhibited. Presumably, one
consequence is that more energy can be stored as fat to sustain the animal through its ﬁrst
winter.

Delaying dispersal until the yearling summer may potentially be quite costly for
male S. beldingi. Late during the active season when most juvenile dispersal occurs, adult
S. beldingi are hibernating or fattening for the winter, and do not interact extensively with
juveniles (Sherman, 1976; Sherman and Morton, 1979; Holekamp, 1984a; Nunes and

Holekamp, 1996). By contrast, yearling males disperse during the early and middle parts

59

of the active season, when adults are mating or when pregnant and lactating females are
defending maternal territories, and levels of adult aggression are consequently high
(Sherman, 1976, 1977; Holekamp, 1984a; Nunes and Holekamp, 1996; Nunes et al.,
1997). Thus, as they attempt to settle in their new home areas, juvenile dispersers should
encounter less conspeciﬁc aggression than do yearling dispersers. I suggest that this
reduced resistance from conspeciﬁcs might also confer an energetic advantage on juvenile
over yearling immigrants. If indeed juvenile dispersers expend less energy in agonistic
interactions with adults, then they should be able to devote more time to foraging.
Patterns of body mass and body fat gain among juvenile male S. beldingi in 1995
differed signiﬁcantly ﬁ'om those observed in 1994 and 1996. Juvenile males gained mass
and fattened more rapidly in 1995, when they were born late in the season. Thus,
ontogenetic patterns of mass and fat gain in juvenile males are not determined by age
alone, but appear to also be inﬂuenced by time of year. Seasonal cycles in ground
squirrels are generated by an endogenous circannual timing mechanism that is entrained to
changes in daylength (Zucker et al., 1991). My data suggest that, in juvenile male S.
beldingi, this timing mechanism regulates body mass and body fat at seasonally
appropriate levels. Although the growing season of S. beldingi ’s habitat varies between
years and it is possible that the differences I observed in mass and fat gain between years
were due to interannual variation in food availability, I believe this unlikely. The active
season of S. beldingi always coincides with the growing season, so food is abundant while
S. beldingi are active, and squirrels hibernate during periods of limited food availability
(Morton, 1975; Sherman, 1976). Sciurid rodents in general are not limited by food

availability during the active season (e. g., Armitage and Downhower, 1974), and in fact

60

may consume as little as 2% of the total energy available to them (e. g., Kilgore, 1972;
Svendsen, 1973).

My results suggest that a circannual timing mechanism in juvenile male S. beldingi
interacts with a physiological mechanism monitoring fat reserves to regulate the timing of
dispersal behavior. It appears that perhaps the central nervous system assesses body fat in
relation to the amount of time remaining until hibernation, and uses this information in
determining whether dispersal occurs or is delayed. I propose a model (Figure 18) in
which a circannual timing mechanism dictates minimum fat levels appropriate for
particular calendar dates (Mrosovsky, 1976; Dark et al., 1984, 1985; Forger et al., 1986),
and a physiological signal informs the central nervous system about whether these levels of
fat have been achieved (Schneider, 1992; Wade and Schneider, 1992; Campﬁeld et al.,
1995; Wade et al., 1996). During the juvenile summer, dispersal behavior is triggered only
when actual body fat rises above a seasonally appropriate threshold.

Various chemical signals have been identiﬁed that might inform the central nervous
system about energy availability, or about the size of body fat stores. For example,
metabolic fuels such as glucose regulate estrous cycles in golden hamsters (Mesocricetus
auratus) and appear to be detected by the area prostrema of the hind brain (Morin, 1986;
Schneider and Wade, 1989; Schneider, 1992; Wade and Schneider, 1992; Schneider and
Zhu, 1994). Metabolic the] levels might also be involved in the regulation of dispersal
behavior in juvenile male S. beldingw‘. Circulating levels of metabolic fuels, such as
glucose and fatty acids, might vary with adipose tissue mass, and convey to the central
nervous system information about overall body fat mass. Another possibility is that a

chemical signal released by adipose tissue in proportion to overall mass of body fat is

61

20-

15-

Dispersal /
may occur /

 

 

 

 

 

 

 

Body fat (% of body mass)

a

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e A C

Calendar date

Figure 18 - A model of circannual and energetic regulation of dispersal.
According to this model, dispersal is inhibited when body fat is below a
seasonally appropriate threshold, represented on the graph by the solid
curve. When the juvenile active season begins at date "A," the fat
threshold is initially low, and available energy can be used to support
dispersal behavior. When the active season begins at date "B," the fat
threshold increases rapidly, available energy is stored as fat needed for
survival during hibernation, and dispersal behavior is inhibited. Aﬁer
date "C," dispersal is inhibited regardless of the size of fat reserves.

62

detected by the central nervous system (Campﬁeld et al., 1995). Interestingly, a peptide
(leptin) that regulates energy balance via effects on feeding and activity levels has been
isolated in laboratory mice and humans (Halaas et al., 1995; Maffei et al., 1995;
Pellymounter et al., 1995; Cohen et al., 1996). It is tempting to speculate that leptin, or a
molecule with similar functions, might play a role in regulating the timing of dispersal
behavior in S. beldingi.

My results suggest that juvenile dispersal in S. beldingi is restricted to a particular
phase of the annual cycle, and that it is inhibited very late in the active season regardless of
the size of body fat reserves. Seasonal inhibition of dispersal in S. beldingi appears
similar, for example, to seasonal suppression of reproduction in female golden-mantled
ground squirrels (S. Iateralis), in which Iuteinizing hormone (LI-I) is secreted in a
circannual rhythm, limiting estrus to a speciﬁc phase of the circannual cycle (Zucker and
Licht, 1983a). In male S. lateralis, sensitivity of LH secretion to negative feedback from
gonadal steroids also follows a circannual pattern (Zucker and Licht, 1983b), and the
capacity for reproductive maturation appears to be limited to a particular period within the
circannual cycle (Barnes and York, 1990). It appears that dispersal behavior by juvenile
male S. beldingi similarly occurs only within a speciﬁc phase of the circannual cycle, and is

inhibited throughout the remainder of the year.

Chapter 5

ENDOCRINE INFLUENCES ON DISPERSAL

INIRODUC T ION

The male bias typical of mammalian dispersal suggests that it might be mediated by
gonadal androgens (Smale et al., 1997). Early work in behavioral endocrinology revealed
that testosterone (T) can affect sexually dimorphic behavior in mammals by “organizing”
physiological substrates of behavior during early development, or by “activating”
physiological processes that mediate behavior later in life (Phoenix et al., 1959). Because
concentrations of circulating T are higher in males than in females during speciﬁc stages of
ontogeny, T plays an important role in the organization and activation of many male-
typical behaviors (Toran-Allerand, 1978; Arnold, 1980; McEwen, 1981; Arnold and
Breedlove, 1985; Moore, 1991; Beatty, 1992; Breedlove, 1992).

Prior work by Holekamp et al. (1984) indicated that T probably does not exert
activational effects on natal dispersal behavior in juvenile male S. beldingi. Male S.
beldingi that were castrated at weaning, and thus deprived of their primary source of T,
still dispersed from the natal area. This prior work by Holekamp et al. (1984) also
provided preliminary evidence that T might exert organizational effects on the natal
dispersal behavior of S. beldingi. The proportion of females treated with T at birth that
emigrated from the natal area by 60 days of age did not differ signiﬁcantly from the
proportion of males that dispersed by 60 days of age. However, Holekamp et al. (1984)

63

64

released mothers of T-treated females in areas other than where they had been captured.
Holekamp et a1 (1984) used females treated with oil at birth as controls, but all of these

that were weaned disappeared from their natal areas, and none were later found. Thus,

Holekamp et al. (1984) were unable to deﬁnitively rule out the possibility that T-treated
females dispersed in response to release of their mothers at a new site, or some variable
other than perinatal T exposure.

Here, I replicated the ﬁndings of Holekamp et al. (1984), and collected new data
to test hypotheses suggesting that natal dispersal in S. beldingi is mediated by activational
or organizational effects of T. I predicted that, if T exerts activational effects on dispersal
behavior, then concentrations of circulating T should increase around the time of dispersal.
I also exploited variation in reproductive maturation among yearling males to test an
additional prediction of the activational hypothesis. Speciﬁcally, if T activates natal
dispersal behavior, then yearlings with large, descended testes, and thus presumably with
elevated concentrations of circulating T, should be likelier to emigrate from the natal area
than should yearling males showing no testicular development. Finally, I examined the
prediction of the organizational hypothesis that female S. beldingi treated with T at birth

should disperse, whereas oil-treated control females should remain near their birth sites.

METHODS
During the summer of 1995, I ear-tagged 126 juvenile males and 130 juvenile
females from 29 complete litters. During the summer of 1996, I ear-tagged 108 juvenile

males and 95 juvenile females from 46 complete litters.

65

Tests of the Activational Hmothesis

To evaluate circulating T concentrations of juveniles in relation to dispersal, I
collected 89 blood samples (0.5-1 .0 ml) from 71 juvenile males and 62 samples from 52
juvenile females in 1996. I regularly collected blood samples beginning about 1 week after
the ﬁrst young of the year were weaned, and continuing until near the end of the primary
dispersal period (13 July to 20 August). These samples were assayed for T concentrations
at Michigan State University during October of 1996.

I determined whether body mass and body fat of free-living juvenile males prior to
hibernation inﬂuence the likelihood that they will become reproductively mature as
yearlings. I also evaluated whether reproductive status of yearling males inﬂuences the
timing of their emigration from the natal area. During the summer of 1995, I manipulated
body mass of juvenile S. beldingi by providing extra food to 21 litters, and used 33 litters
as unprovisioned controls. I gave supplemental food to mothers of provisioned litters
during lactation, and to their young for 6 weeks after they were weaned. I measured body
mass of juveniles from these litters weekly from ages 4-11 weeks, and body fat weekly
ﬁom ages 6-11 weeks. I then evaluated reproductive condition (Morton and Gallup,
1975) of these squirrels during their yearling summer in 1996. When I observed squirrels
mating during regular observations, I recorded the identity of the squirrels, as well as the

date, time, location, and duration of copulation.

Test of the Organizational Hypothesis
To test the organizational hypothesis, 1 manipulated early androgen exposure of S.

beldingi. During the summer of 1995, I treated newborn females from 11 provisioned

66

litters and 18 unprovisioned litters with testosterone (T), and used oil-treated females from
10 provisioned and 15 unprovisioned litters as controls. I monitored dispersal status of
these females and their brothers weekly from 6-10 weeks of age. I also regularly
monitored dispersal status of these squirrels during their yearling summer in 1996, until 55
weeks of age.

Hormone and provisioning manipulations were combined to elucidate the
interaction between endocrine and energetic variables in the mediation of natal dispersal
behavior. However, because of heavy snowfall during the winter and spring, juvenile
dispersal occurred at very low rates in 1995 (Nunes et al., in press), and I was unable to
evaluate this interaction. I repeated my experimental manipulation of hormones and food

availability in 1996, and discuss the results in Chapter 4.

RES UL T S
Provisioning Effects on Reproductive Development of Males

During May of 1996, I trapped 28 males from the 1995 cohort, 15 of which had
been provisioned as juveniles, and 13 of which had been unprovisioned. None of the
males from the unprovisioned group exhibited signs of reproductive competence (e. g.,
large, scrotal testes and pigmented scrotum; Morton and Gallup, 1975). By contrast, nine
males ﬁ'om the provisioned group were in reproductive condition. Thus, males were more
likely to become reproductively mature as yearlings if they were provisioned as juveniles
(Fisher’s exact test, P = 0.001). Reproductive yearling males in 1996 were signiﬁcantly
heavier than non-reproductive yearlings at the end of their juvenile summer during the

previous year (Figure 19; t-test for independent samples, t = 6.940, df= 22, P < 0.001).

67

Similarly, body fat in 1996 was greater in reproductive than non-reproductive yearlings at
the end of the juvenile summer during the previous year (Figure 20; t-test for independent
samples, t = 2.472, df= 21, P = 0.012). Thus, body mass and body fat at the end of the
juvenile summer appear to inﬂuence whether or not free-living male S .beldingi become

reproductively competent as yearlings.

The Activational ijothesis

Concentrations of circulating T did not differ between juvenile males and juvenile
females of the 1996 cohort, and in all cases fell below assay detectability levels (0.2
ng/rnl). Thus, I found no measurable increases in circulating T associated with natal
dispersal behavior, as predicted by the activational hypothesis.

Of the nine reproductively mature yearlings I trapped in 1996, one dispersed as a
juvenile during the summer of 1995, seven emigrated from the natal area in 1996, and one
disappeared in 1996 and was never recovered. None of the reproductive yearling males
that dispersed or disappeared in 1996 did so before the end of the mating period, which I
estimated to be 12-29 May, based on observations of mating and dates of litter emergence.
Of the 19 non-reproductive yearling males that I trapped in 1996, ﬁve dispersed as
juveniles in 1995, 12 dispersed as yearlings in 1996, and two disappeared in 1996 and
were never recovered. Eight of the 12 non-reproductive yearlings that dispersed in 1996
did so between 12-29 May, during the breeding period. Thus, yearling males with small,
non-scrotal testes were signiﬁcantly more likely to disperse during the breeding season
than were yearling males with large, descended testes (Figure 21; Fisher’s exact test, P =

0.013). In fact, all yearlings with large, scrotal testes delayed dispersal until aﬁer the

68

400 T
350 -
300 -1
250 ~
200 -

150 -

Body mass (grams)

100 -

50-

 

 

Reproductively Not reproductively
mature mature

Figure 19 - Body mass of yearling males at the end of their juvenile summer.
The asterisk indicates a signiﬁcant difference between males that did and did
not achieve reproductive maturity as yearlings. Sample sizes are given on
the graph.

69

Body fat (% of body mass)

 

 

Reproductively Not reproductively
mature mature

Figure 20 - Body fat of yearling males at the end of their juvenile summer.
The asterisk indicates a signiﬁcant difference between males that did and did
not achieve reproductive maturity as yearlings. Sample sizes are given on
the graph.

70

breeding period, when size of testes begins to diminish (Morton and Gallup, 1975), as
concentrations of circulating T presumably do also (e. g., Holekamp and Talamantes,

1992)

The Organizational Hypothesis

Among groups (T-treated females, oil-treated females, and males) in the 1995

cohort, the proportion of juveniles that dispersed at ages 7-10 weeks did not differ
signiﬁcantly between squirrels that were provisioned and those that were unprovisioned
(T-treated females: no T-treated females dispersed; oil-treated females: Mantel-Haenszel
test, [7 = 2.151, P = 0.142; males: Mantel-Haenszel test, 1’ = 0.032, P = 0.859).
Similarly, among these groups, the proportion of individuals that dispersed by 55 weeks of
age did not differ signiﬁcantly between provisioned and unprovisioned squirrels (T-treated
females: Fisher’s exact test, P = 0.249; oil-treated females: Fisher’s exact test, P = 0.538;
males: all surviving males dispersed). Thus, dispersal data were combined for provisioned
and unprovisioned individuals within these groups for statistical analysis.

Heavy winter and spring snowfall had a dramatic effect on the timing of natal
dispersal behavior in 1995 (Nunes et al., in press), and less than 10% of surviving
members of the 1995 cohort dispersed by the end of their juvenile summer (Figure 22).
The proportion of T-treated juvenile females that dispersed ﬁom 7-10 weeks of age in
1995 did not differ signiﬁcantly from that of oil-treated females (Figure 22; Mantel-
Haenszel test, 1’ = 1.634, P = 0.201) or males (Figure 22; Mantel-Haenszel test, ,1? =
0.933, P = 0.334). However, by 55 weeks of age, 80% of surviving T-treated females and

100% of surviving males in the 1995 cohort had dispersed, whereas only 13% of oil-

71

 

 

 

 

 

 

 

 

 

10° 7 —G- Not reproductively j}
mature /
+ Reproductvely //

80 — mature
'o
a)
9
3 so -
.92
"o
°\° 4o - /
g /
'53 /
E //
:3 /
E 20 — /
3 /
0 /

/
o — a
MATING
l l l 1
1-15 16-31 1-15 16-30
May May June June

Figure 21 - Dispersal of yearling males, 1996. The asterisk indicates a
signiﬁcant difference between males that did and did not achieve reproductive
maturity as yearlings. Samples include 7 males that achieved reproductive
maturity and ‘12 that did not.

72

 

 

 

1oo — —o— T-treated females 'P
—E}-' Males /
—A— Oil-treated females /
so — l
'0
m /
£2
8.
.9 60 ~
'0
o\°
m
,5 4o —
£5 /
3 /
E /
8 2° - /
1:1— — — /A
// ———:74—*——— ’
0 _ M — — w
l l l l l i i l
6 7 8 9 1o 55
Age (weeks)

Figure 22 - Dispersal of the 1995 cohort. The cumulative percentages of
juveniles that dispersed from 6 through 10 weeks of age, and of yearling
that dispersed by 55 weeks of age, are presented. Sample of juveniles
included 49 T-treated females, 79 males, and 46 oil-treated females.
Samples of yearlings included 20 T-treated females, 24 males, and 23
oil-treated females.

73

treated females had emigrated (Figure 22). The proportion of T-treated females in the
1995 cohort that dispersed by 55 weeks of age differed signiﬁcantly ﬁ'om that of both oil-
treated females (Figure 22; f = 19.447, P < 0.001) and males (Figure 22; Fisher’s exact
test, P = 0.03 6), but was more similar to that of males (Figure 22). Thus the probability
of natal dispersal by T-treated females more closely resembled that of male than that of

female peers.

DISCUSSION

Together with the results obtained by Holekamp et al. (1984), my data offer
strong evidence that T does not activate natal dispersal behavior in S. beldingi. The
activational hypothesis predicts that early castration should inhibit dispersal
behavior, but this does not occur (Holekamp et al., 1984). Furthermore this
hypothesis predicts that concentrations of circulating T should be relatively high
around the time that dispersal occurs, but my work shows that this is not the casein
this species. Holekamp et al. (1984) measured circulating T throughout the juvenile
summer in S. beldingi that were born in the laboratory to mothers caught in the
wild. Plasma T was undetectable throughout the summer in lab-reared juveniles.
Here I found that circulating T was also undetectable in free-living juveniles
throughout the juvenile summer, when most males usually disperse. The possibility
remains that dispersal is activated by changes in T concentrations below the 0.2
ng/ ml detectability level of the assay I used, but the fact that castrated males

disperse (Holekamp et al., 1984) makes this possibility unlikely. Here I also found

74

that yearling males were less likely to disperse when they had large, scrotal testes
than when their testes were small and undescended. Size of testes in male ground
squirrels is positively correlated with blood concentrations of T (Barnes, 1984;
Holekamp and Talamantes, 1992), and male S. beldingi with large testes
presumably have higher levels of circulating T than do males with small, abdominal
testes. Thus, if T has concurrent effects on natal dispersal behavior in this species,
they appear to be inhibitory rather than stimulatory.

Most male S. beldingi emigrate from the natal area during their juvenile
summer, two years before achieving reproductive maturity (Morton et al. , 1974).
When dispersal occurs well before males reach puberty, the behavior probably is
not likely to be activated by gonadal androgens (Smale et al., 1997). However, in
many species there is a close temporal association between dispersal and puberty
(e.g., Errington, 1963; Pusey and Packer, 1987a, 1987b; Rood, 1987, 1990;
Matsumura, 1993; Smale et al., 1997). The onset of reproductive maturity is
characterized by dramatic increases in circulating concentrations of gonadal steroid
hormones, which stimulate a variety of behavioral changes (Bronson and Rissman,
1986; Wilson, 1992). Thus, elevations in concentrations of circulating androgens
might be likely to promote dispersal behavior in species unlike S. beldingi in which
males typically emigrate from the natal area shortly after attaining puberty (Smale
et al. , 1997).

My results indicate that around the time of birth in S. beldingi, T organizes

masculine patterns of natal dispersal behavior. The proportion of female S.

75

beldingi treated neonatally with T that emigrated from their birth sites by 55 weeks
of age was intermediate between that of females treated with oil at birth, and that of
males, but was closer to that of males. Holekamp et al. (1984) observed that
female S. beldingi treated with T at birth were as likely as males to emigrate from
the natal area by 60 days of age, and also exhibited increases in exploratory and
locomotor behavior and boldness that normally accompany dispersal by young male
S. beldingi. Thus, my work with S. beldingi and the work by Holekamp et al.
(1984) represent the ﬁrst demonstration of organizational T effects on any behavior
pattern in free-living animals. Laboratory studies with captive animals have shown
that perinatal androgen exposure inﬂuences a variety of noncopulatory behaviors
such as play, locomotion, aggression, and learning (e.g., Williams etal., 1990;
Beatty, 1992). My results provide ﬁeld veriﬁcation that androgens can organize
noncopulatory behavior. My results also strongly suggest that natal dispersal in S.
beldingi is among those behaviors, such as play and spatial learning, which
apparently require no subsequent activational supplementation of perinatal
organizational hormone effects (Goy and McEwen, 1980; Holekamp et al., 1984;
Williams et al., 1990).

In many species of mammals, all males eventually leave their natal areas
(e.g., Dobson, 1979; Hanby and Bygott, 1987; Pusey and Packer, 1987a, 1987b;
Matsumura, 1993; Rutberg and Kuiper, 1993; Smale, 1997). Gonadal androgens
might often organize a male-typical propensity to emigrate from the natal area when

all males of a species ultimately disperse (Smale et al., 1997). However, in many

76

mammalian species, not all males leave the natal area (e. g. , Harcourt et al., 1976;
Harcourt, 1978; Wiggett and Boag, 1992, 1993). In species for which the costs of
dispersal sometimes favor philopatric behavior by males, androgenic organization
of a deterministic predisposition to leave the natal area might well be maladaptive
(Holekamp et al., 1984; Smale et al., 1997). However, in these species, androgens
might organize a masculine response to social or ecological variables which
promote dispersal behavior.

The action of gonadal steroid hormones facilitates many features of
reproductive behavior, including courtship, establishment and defense of breeding
territories, sexual partner preference, and copulation (McEwen, 1981; Baum et al. ,
1990; Breedlove, 1992; Mansukhani et al., 1996; Rudd et al., 1996). The role of
T in organizing male-typical dispersal behavior in S. beldingi offers physiological
evidence of a link between dispersal and reproduction (see also O’riain et a1. ,
1996). Prior studies suggest two ways in which natal dispersal might represent an
important male reproductive strategy. First, emigration from the natal area can
improve the quality of a male’s offspring by reducing the probability of inbreeding
(e.g., Pusey, 1980, 1987; Caley, 1987; Charlesworth and Charlesworth, 1987;
Keane, 1990; Wolff, 1992). Second, natal dispersal can enhance a male’s access to
unfamiliar females that are more willing to mate with him than are females in his
natal area (e.g., Peirera and Weiss, 1991; Alberts and Altmann, 1995). One or
both of these might link dispersal behavior with male reproductive success in S.

beldingi.

Chapter 6

INTERACTION OF ENDOCRINE AND ENERGETIC VARIABLES

IN TRODUC TTON

Organizational effects of the gonadal steroid testosterone (T) strongly inﬂuence
the probability that individual S. beldingi will someday emigrate ﬁom the natal area
(Holekamp et al., 1984; Chapter 5). Perinatal treatment of female S. beldingi with T
induces male-like dispersal behavior during the juvenile or yearling summer. The
determinants of when during its early life an individual will disperse are currently
unknown. However, energy availability appears to play an important role in timing
dispersal behavior in S. beldingi (Nunes and Holekamp, 1996; Nunes et al., in press).
Here, I attempted to ascertain whether energetic variables inﬂuence the timing of natal
dispersal behavior in this species, and whether endocrine and energetic factors interact to
mediate the occurrence of this behavior pattern.

Hormones inﬂuence the development of many behaviors, and often interact with
energetic variables in the regulation of reproduction and behavior (e. g., Zucker and
Boshes, 1982; Kenagy, 1980; Barnes, 1984; Smale et al., 1986; Beatty, 1992; Wade and
Schneider,l992; Wade et al., 1996). Because of the potential risks and energetic costs
associated with dispersal, an individual’s decision regarding whether or not to emigrate at
any given time should be inﬂuenced by energy stores available to support this movement in
addition to the normal maintenance and growth functions necessary for survival (Nunes

77

78

and Holekamp, 1996; Smale et al., 1997; Nunes et al., in press). The amount of energy
stored as fat does in fact appear to importantly inﬂuence the timing of natal dispersal by
juvenile male S. beldingi. Dispersing males tend to be fatter than same-aged males still
residing in the natal area (Nunes and Holekamp, 1996). Moreover, juvenile male S.
beldingi born very late in the year delay dispersal until their yearling summer, presumably
because they do not have sufﬁcient time during the juvenile summer to take in enough
energy both to ﬁJCI dispersal behavior and to build fat reserves necessary to survive
overwinter hibernation (Nunes et al., in press).

Energy availability may also inﬂuence the timing of dispersal by those few juvenile
female S. beldingi that leave their natal areas (Smale et al., 1997). Female S. beldingi that
disperse during the juvenile summer tend to be heavier than same-aged females still
residing in their natal areas (Holekamp, 1984; Nunes et al., 1997). However, the timing of
dispersal differs between juvenile male and female S. beldingi. Furthermore, when
juvenile females disperse, they generally do so at signiﬁcantly older ages than do juvenile
males (Nunes et al., 1997).

Here I examined an hypothesis suggesting that early hormone exposure interacts
with current energy availability to mediate the natal dispersal behavior typical of male S.
beldingi. I predicted that the timing of natal dispersal by juvenile female S. beldingi
treated with T at birth should respond to experimental manipulations of body fat in the
same manner as does the timing of dispersal by juvenile males. Juvenile male S. beldingi
fattened by food provisioning disperse from the natal area at younger ages than do
unprovisioned juvenile males (Nunes and Holekamp, 1996; Nunes et al., 1997a). If

perinatal androgen exposure and energy availability interact to mediate natal dispersal,

79

then T-treated females receiving supplemental food should disperse at younger ages than
should unprovisioned T-treated females. Furthermore, T-treated females receiving extra
food should emigrate at younger ages than should any provisioned oil-treated females that

disperse.

METHODS

During the summer of 1996, I trapped and ear-tagged 203 juvenile S. beldingi,
including 108 males and 95 females from 46 complete litters.

I concurrently manipulated body mass, body fat and perinatal androgen exposure
of these free-living juveniles. I treated newborn females with testosterone (T) and used
oil-treated and untreated females as controls. Within each treatment group, I manipulated
body fat of females and their brothers by provisioning them with food, and used
unprovisioned females and their brothers as controls. I treated newborn females from 14
litters with T, treated females from 19 litters with oil, and used females ﬁom l3 litters as
untreated controls. I gave extra food to seven T-treated litters, nine oil-treated litters, and
six untreated litters, and I used seven T-treated litters, ten oil-treated litters, and seven
untreated litters as unprovisioned controls. I monitored body mass of juveniles weekly
from 4-11 weeks of age, and measured body fat weekly from 6-11 weeks of age. I

observed the dispersal status of juveniles weekly from 6-9 weeks of age.

RES UL T S
Body mass did not differ signiﬁcantly between T-treated, oil-treated, and untreated

females (provisioned: ANOVA, P = 0.193; unprovisioned: ANOVA, P = 0.174) and so

80

body mass data of these females were combined within provisioned and unprovisioned
groups for statistical analysis. Overall, provisioned juveniles were signiﬁcantly heavier
than unprovisioned juveniles from 4-9 weeks of age among both females (Figure 23;
ANOVA, F= 129.068, df= 1, P < 0.001) and males (Figure 24; ANOVA, F= 56.087, df
= 1; P < 0.001). Provisioned females were signiﬁcantly heavier than unprovisioned
females in each week-long age interval from 4-9 weeks of age (Figure 23; t—tests for
independent samples; 4 weeks: t = -6.014, df= 74, P < 0.001; 5 weeks: t = 4.868, df=
58, P < 0.001; 6 weeks: I: -4.696, df= 51, P < 0.001; 7 weeks: t= -6.303, df= 41, P <
0.001; 8 weeks: t = -4.190, df= 45, P < 0.001; 9 weeks: t= -2.565, df= 38, P = 0.014).
Similarly, provisioned males were signiﬁcantly heavier than unprovisioned males ﬁom 4-8
weeks of age (Figure 24; t-tests for independent samples; 4 weeks: t = -4.464, df= 97, P
< 0.001; 5 weeks: t = -3.272, df= 75, P = 0.002; 6 weeks: t = -3.046 df= 59, P = 0.003;
7 weeks: t= -3.693, df= 51, P = 0.001; 8 weeks: t= -2.890, df= 44, P =0.006).
Provisioning also had notable effects on body fat, particularly among younger
juveniles. Body fat did not differ between T-treated, oil-treated, and untreated females
(provisioned: ANOVA, P = 0.227; unprovisioned: ANOVA, P = 0.149), and so body fat
data of these females were combined within provisioned and unprovisioned groups for
statistical analysis. Overall, provisioned juveniles had signiﬁcantly greater body fat from
6-9 weeks of age than did their unprovisioned counterparts among both females (Figure
25; ANOVA, F= 12.575, df= 1, P < 0.001) and males (Figure 26; ANOVA, F= 4.784,
df= 1, P = 0.030). However, body fat differed signiﬁcantly only at 6 and 7 weeks of age
among both females (Figure 25; t-tests for independent samples, 6 weeks: t = -2.489, df=

44, P = 0.017; 7 weeks: t = -5.363, df= 38, P < 0.001) and males (Figure 26; t-tests for

81

220 q + Provisioned females

 

 

 

—D—~ Unprovisioned females *
zoo — *
180 -
A
U)
E 150 4
g
v 140 —
U)
3
1 _
E 20
-c>:‘
o 100 -—
m
so -
Cl/
38
40 T F I I I I
4 5 5 7 8 9

Age (weeks)

Figure 23 - Body mass of juvenile females, 1996. Asterisks indicate a
signiﬁcant difference between provisioned and unprovisioned females.
Sample sizes are given on the graph.

 

82

 

 

zoo -— + Provisioned males 20
—D—- Unprovisioned males
180 — 26
A 150 -—
(I)
E
g 140 _.
9
3 120 5
cu
E
> 100 4
'0
8
so -—
so — c//
45
40 W l l l l l
4 5 5 7 a 9
Age (weeks)

Figure 24 - Body mass of juvenile males, 1996. Asterisks indicate a
signiﬁcant difference between provisioned and unprovisioned males.
Sample sizes are given on the graph.

83

 

 

 

14 — —e— Provisioned females
13 _ —-I:I—- Unprovisoned females
12 -
a :3 19
w 11 -
g 10 —
>. 9 —
8
.o 8 ‘
“5 7 —
5‘: 6~
.19 5 ‘
> 4 —
8
m 3 ‘
2 .4
1 _
0 I n I I
5 7 8 9
Age (weeks)

Figure 25 - Body fat of juvenile females, 1996. Asterisks indicate a
signiﬁcant difference between provisioned and unprovisioned females.
Sample sizes are given on the graph.

84

11-

 

 

 

—e— Provisioned males

10 — —D—- Unprovisioned males
A 9 -1 *
(I)
m 30
(U 8 ._
E
> _
.0 7
o
.0 5 _
q—
o
°\° 5 ‘
.... 4
‘9 —1
a a - /
o ______
m 2 __ T—

1 _ 17 13

o , 1 r I

5 7 8 9
Age (weeks)

Figure 26 - Body fat of juvenile males, 1996. Asterisks indicate a
signiﬁcant difference between provisioned and unprovisioned males.
Sample sizes are given on the graph.

85
independent samples, 6 weeks: t = -2.348, df= 43, P = 0.024; 7 weeks: t= -3.407, df=
46, P = 0.001).

Hormone treatments inﬂuenced the probability of dispersal by female S. beldingi.
The cumulative percentage of oil-treated females that dispersed from 7-9 weeks of age did
not differ signiﬁcantly ﬁ'om that of untreated females (provisioned: n = 10 oil-treated, n =
10 untreated, Mantel-Haenszel test, 2’ = 0.134, P = 0.715; unprovisioned: n = 6 oil-
treated, 71 = 13 untreated, no females dispersed), and so oil-treated and untreated females
were combined within provisioned and unprovisioned groups for statistical analyses, and
are referred to here as “control” females. The percentage of T-treated females that
dispersed between 7-9 weeks of age differed signiﬁcantly from that of control females in
both provisioned (Figure 27; Mantel-Haenszel test, 1’ = 13.513, P < 0.001) and
unprovisioned groups (Figure 27; Mantel-Haenszel test, 1" = 14.982, P < 0.001).
However, the percentage of T-treated females that dispersed between 7-9 weeks of age
did not differ signiﬁcantly from that of males in either provisioned (Figure 27 ; Mantel-
Haenszel test, 2’ = 0.253, P = 0.615) or unprovisioned groups (Figure 27; Mantel-
Haenszel test, 2’ = 0.103, P = 0.748). Thus, dispersal behavior of T-treated females more
closely resembled that of males than that of control females.

Provisioning advanced the timing of dispersal among both males and T-treated
females. The median age of dispersal was younger in provisioned than unprovisioned T-
treated females (7 weeks vs. 9 weeks, Kruskal-Wallis test, U = 2.0, P = 0.007). The
median age of dispersal was also younger in provisioned than unprovisioned males (7.5

weeks vs. 9 weeks, Kruskal-Wallis test, U = 80.5, P = 0.023).

Cumulative % dispersed

80*

70-

60—

50-

40-

30—

20—

 

86

+ T-females, provisioned

—O— T-females, unprovisioned

—l—- Males, provisioned

—{:I—- Males, unprovisioned

--A---- Control females, provisioned

—-A—-- Control females, unprovisioned }

 

 

 

Age (weeks)

Figure 27 - Dispersal of juveniles, 1996. Samples included 13 provisioned
T-treated females, 10 unprovisioned T-treated females, 40 provisioned
males, 23 unprovisioned males, 20 provisioned control females, and 19
unprovisioned control females.

87

Provisioning also inﬂuenced the probability of dispersal by control females.
Provisioned control females were more likely to disperse than were unprovisioned control
females (Mantel-Haenszel test, f = 9.556, P = 0.002), but only half as likely to disperse
as were either group of T-treated females (Figure 27). The timing of dispersal by
provisioned control females differed from that by provisioned T-treated females and
provisioned males. Provisioned control females that dispersed did so at a signiﬁcantly
older median age than did provisioned T-treated females (9 weeks vs. 7 weeks,
respectively; Kruskal-Wallis test, U = 39.0, P = 0.006) and at a somewhat older median
age than did provisioned males (9 weeks vs. 7.5 weeks, Kruskal-Wallis test, U = 105.0, P
= 0.068). Thus, perinatal T-treatment appears to have caused the timing of dispersal by
young females to respond to food provisioning in a male-typical way.

I also observed an association between body fat and dispersal. By 7 weeks of age,
when body fat was signiﬁcantly greater in provisioned than unprovisioned juveniles, a
greater proportion of provisioned than unprovisioned juveniles had dispersed among both
T-treated females (Figure 28; Fisher’s exact test, P = 0.046) and males (Figure 29;
Fisher’s exact test, P = 0.022). However, by 9 weeks of age, when body fat no longer
differed between provisioned and unprovisioned juveniles, the proportion of juveniles that
had dispersed was similar in provisioned and unprovisioned groups among both T-treated
females (Figure 28; Fisher’s exact test, P = 1.000) and males (Fig. 29; 1’ = 0.365, P =
0.546). This suggests that the advances in the timing of dispersal associated with food

provisioning are mediated by body fat or fat-related variables.

88

70

 

T-females, provisioned
[:I T-females, unprovisione

Cumulative % dispersed

 

 

 

 

7 weeks 9 weeks
Age

Figure 28 - Dispersal of T-treated females, 1996. The asterisk indicates
a signiﬁcant difference between provisioned and unprovisioned females.
Samples included 13 provisioned and 10 unprovisioned T-treaed females.

89

 

Males, provisioned
[:21 Males, unprovisioned

Cumulative % dispersed

 

 

 

 

 

 

7 weeks 9 weeks
Age

Figure 29 - Dispersal of males, 1996. The asterisk indicates a signiﬁcant
difference between provisioned and unprovisioned males. Samples
included 40 provisioned and 23 unprovisioned males.

90

DISCUSSION

Perinatal T-treatment increased the probability that female S. beldingi would
emigrate from the natal area, supporting the hypothesis that T organizes masculine
dispersal patterns in this species (Holekamp et al., 1984). Furthermore, food provisioning
advanced the timing of dispersal among both T-treated females and males. These results
indicate that T organizes masculine timing of dispersal as well as a male-typical
predisposition to emigrate from the natal area. Provisioned juveniles were heavier than
unprovisioned juveniles throughout the summer. However, provisioned juveniles were
fatter than their unprovisioned counterparts only until 7 weeks of age. Moreover, a
greater proportion of provisioned than unprovisioned males and T-treated females had
dispersed only until 7 weeks of age. Nunes and Holekamp (1996) observed that juvenile
male S. beldingi disperse aﬁer attaining a threshold body mass. My results here indicate
that once a mass threshold for dispersal has been reached, the timing of masculine
dispersal behavior may be more importantly inﬂuenced by body fat than by body mass;
body mass might affect dispersal only insofar as it reﬂects body fat (Smale et al., 1997).
My results further suggest that physiological signals related to body fat convey
information about environmental conditions such as food availability to endogenous
mechanisms mediating dispersal behavior.

Provisioning increased the proportion of control female S. beldingi that dispersed.
This replicates prior work by Nunes et al. (1997) in which food provisioning caused
juvenile female S. beldingi to emigrate from the natal area. Nunes et al. (1997) observed
that population density and density of mothers weaning litters was greater in provisioned

than unprovisioned areas of their study site, and that mothers residing in provisioned areas

91

exhibited higher rates of aggression and vigilance than did mothers in unprovisioned areas.
They suggested that provisioned females emigrated from their natal areas in response to
increased competition for non-food resources such as space in which to establish a
territory. In this prior work, provisioned females that dispersed did so at signiﬁcantly
older ages than did provisioned males. Furthermore, only about 40% of provisioned
females ultimately dispersed, whereas 100% of males eventually left their natal areas.
Here we similarly observed that provisioned control females that dispersed leﬁ home at
older ages and in smaller proportions than did either provisioned T-treated females or
provisioned males. This further suggests that the effects of energy availability on the
timing of dispersal are inﬂuenced by T exposure near the time of birth.

Endocrine factors interact with energetic variables in the mediation of a wide array
of reproductive events among mammals. For example, hormonal and metabolic variables
have been shown to interact in the control of reproductive maturation in both male and
female mammals (Barnes, 1984; Bushberg and Holmes, 1985; Bronson and Rissman,
1986; Bronson, 1987, 1989; Wilson, 1992), and in the control of estrous cycling and
fertility among females (Wade and Schneider, 1992; Wade et al., 1996). Here I show for
the ﬁrst time that endocrine and energetic variables may also interact in the mediation of
dispersal behavior. Moreover, our results reveal in free-living mammals an interaction
between endocrine control of behavior and the regulation of energy balance. To date,
most studies of this interaction have been accomplished in the laboratory, using species
such as hamsters (Mesocricetus auratus) and mice (Mus musculus) about whose ecology
very little is known (e. g. Campﬁeld et al., 1995; Halaas et al., 1995; Pellymounter et al.,

1995; Wade and Schneider, 1992; Wade et al., 1996).

92

It was recently reported that a unique dispersive phenotype occurs occasionally in
colonies of naked mole rats (Heterochephalus glaber; O’riain et al., 1996).
Approximately 95% of these dispersive H. glaber are males. H. glaber that disperse are
morphologically, physiologically, and behaviorally distinct from other colony members.
Dispersers tend to be extraordinarily fat, have elevated blood concentrations of luteinizing
hormone, and are attracted to rather than aggressive toward unfamiliar conspeciﬁcs.
Dispersive phenotypes in H. glaber, S. beldingi, and perhaps many other mammals might
share a suite of deﬁning characteristics that differ between species only in degree. The
situation in H. glaber may represent an extreme development of the same basic control
system operating in S. beldingi, in which endocrine and energetic variables interact to
mediate dispersal. Interestingly, data from both H. glaber and S. beldingi suggest the
existence of key links between mediation of dispersal behavior and reproductive
physiology (O’riain et al., 1996; Chapter 5).

I propose here a model (Figure 30) which describes possible interactions between
endocrine and energetic variables in the mediation of natal dispersal behavior in male
mammals. According to this model, perinatal androgen exposure organizes neural systems
promoting male-typical natal dispersal behavior, as well as the responsiveness of these
systems to energetic variables such as body fat. When an individual male has sufﬁcient fat
stored to meet the potential energetic demands of the dispersal movement, this
information is conveyed by physiological systems that monitor fat stores or fuel availability
to systems that mediate dispersal, which then initiate masculine patterns of dispersal

behavior.

93

 

 

Body Fat Perinatal
Stores Andro gens

/

Systems Systems

Monitoring —y-. Promoting
F uels/F at Dispersal

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DISPERSAL

Figure 30 - A model of endocrine and energetic regulation of dispersal.
According to this model, early androgen exposure masculines
physiological systems that promote dispersal behavior, as well as the
way in which these systems respond to energetic variables such as body
fat. When an individual has sufficient fat stored to meet the potential
energetic demands of dispersing, this information is conveyed by
systems that monitor fuel availability to systems that promote dispersal,
which then initiate male-typical dispersal behavior.

94

This model appears to describe the relationship between hormones and energy
balance in the population of S. beldingi that I studied at Tioga Pass. However, this
population resides at high altitude (3040 m), in a habitat with a very short growing season
(about 3 months). The brevity of the growing season places major energetic constraints
on resident squirrels at Tioga Pass (Morton et al., 1974; Maxwell and Morton, 1975;
Morton, 1975), and energetic demands of dispersal behavior may be especially severe in
these squirrels. Thus, it is not clear whether my model applies to male mammals in
general. However, potential risks and energetic demands are associated with dispersal in
many mammals (e. g., Waser, 1996; Smale et al., 1997), and I suggest that the model may
describe the physiological underpinnings of dispersal in many mammalian species.

My results also suggest interesting parallels between dispersal and migratory
behavior. Hormones directly or indirectly stimulate migratory behavior, and help mobilize
energy to support it, in insects, ﬁsh, amphibians, and birds (Hoar, 1988; Wingﬁeld et al.,
1990; Young et al., 1989; Rankin, 1991; Dingle, 1996). My results indicate that perhaps
T has similarly been “captured” during the evolutionary history of S. beldingi to function
in the mediation of dispersal movements as well as in reproduction. The actual
mechanisms by which perinatal T exposure interacts with lipid stores to inﬂuence male
dispersal behavior in S. beldingi are unknown. One possibility is that concentrations of
circulating metabolic fuels such as glucose or fatty acids, or metabolic hormones such as
insulin or glucagon, are inﬂuenced by the size of fat stores, and are detected by
endogenous mechanisms that mediate dispersal behavior. Circulating glucose in fact
regulates estrous cycles of female hamsters (Mesocricetus auratus) via effects on the area

prostrema of the hind brain (Schneider and Wade, 1989; Wade and Schneider, 1992;

95

Schneider and Zhu, 1994; Wade et al., 1996). A second possibility is that leptin, a
compound released by adipose tissue (Campﬁeld et al., 1995; Halaas et al., 1995;
Pellymounter et al., 1995), plays a role in mediating emigration from the natal area.
Leptin regulates feeding and activity in mice via endocrine effects on the central nervous
system (Campﬁeld et al., 1995). In S. beldingi, the amount of leptin released might be
proportional to the overall amount of body fat, and might convey information about the

overall size of fat stores to endogenous mechanisms mediating natal dispersal behavior.

Chapter 7

ECOLOGICAL INFLUENCES ON DISPERSAL

INTRODUCTION

Because of the striking sexual dimorphisms observed in dispersal behavior, Dobson
(1979) and Holekamp (1984a) proposed that evolutionary explanations of natal dispersal
might oﬁen differ for males and females. Recent evidence suggests that the ultimate
causes of natal dispersal may differ between the sexes among Townsend’s voles Wicrotus
townsendii; Lambin, 1994) and various species of mammalian carnivores (e. g. Waser,
1996). However, because dispersal by both sexes is typically rare among mammals,
collecting data on sex differences in behavior is often difﬁcult and requires long-term study
(e. g., Pusey and Packer, 1987b), fortuitous observations of rare dispersal events (e. g.,
Mills, 1990; Holekamp et al., 1993), or experimental induction of dispersal (e.g., Dobson,
1979). In a food provisioning experiment examining the timing of dispersal behavior by
juvenile male S. beldingi (Nunes and Holekamp, 1996), I observed an unexpected change
in the ﬁ'equency of emigration by juvenile female S. beldingi from areas receiving extra
food. This allowed me to compare the dispersal behavior of male and female S. beldingi
receiving supplemental food with that of unprovisioned animals to determine whether the
ultimate causes of natal dispersal differ between males and females of this species.

Natal dispersal by male S. beldingi appears not to be driven by demand for
environmental resources (Holekamp, 1986). It is more likely that juvenile male S. beldingi

96

97

disperse to avoid inbreeding (Sherman, 1976; Holekamp, 1984b), or to increase their
probability of encountering receptive mates (Holekamp, 1986). On the other hand,
juvenile female ground squirrels may disperse to improve their access to environmental
resources (Dobson, 1979). Iffemale but not male S. beldingi emigrate ﬁ'om the natal area
in response to demand for environmental resources, then altering the availability of
resources such as food, should affect the probability of dispersal by juvenile females but
not by juvenile males.

Female S. beldingi require three critical environmental resources for successﬁil
reproduction: nest sites, food, and space in which to establish maternal territories. If
females emigrate to improve access to nest sites, then they should move from areas
containing few burrows to areas containing abundant burrows. Holekamp (1986) was
able to rule out the nest demand hypothesis as an explanation for dispersal in both male
and female S. beldingi, because numbers of immigrants and emigrants in different study
areas were unrelated to numbers of available burrows. Here I examined the remaining
possibilities that female dispersal improves access to food or space in which to establish
territories.

If female S. beldingi disperse to improve their access to food, then they should
emigrate ﬁom areas in which competition for food is intense, and immigrate into areas
containing abundant food. Thus, provisioning an area with supplemental food should
reduce emigration from that area and increase immigration into it (Arcese, 1989).
However, if female S. beldingi disperse to improve their access to space for territory
establishment, then they should emigrate ﬁ'om areas of high to low population density,

regardless of food abundance. Thus, experimentally increasing population density by food

98

provisioning (Boutin, 1990; Dobson, 1995) should increase the probability of dispersal by

juvenile female S. beldingi from provisioned areas.

METHODS

During the summers of 1993 and 1994, I trapped and tagged 135 juvenile males
and 155 juvenile females ﬁ'om 65 complete litters. I provided extra food to adult and
yearling females in their maternal territories during pregnancy and lactation, and to
offspring of these females in their natal areas for six weeks after weaning. I used
unprovisioned young of unprovisioned mothers as controls. I gave extra food to juveniles
from 15 litters in 1993 and 18 litters in 1994. Iused juveniles from ten litters in 1993 and
22 litters in 1994 as unprovisioned controls. I measured body mass of these juveniles
weekly ﬁ'om 4-11 weeks of age, and monitored dispersal status weekly from 6-10 weeks
of age. Also in 1993 and 1994, I monitored the density of females that weaned litters and
overall density of S. beldingi.

During the summer of 1995, I gave extra food to 20 reproductive females, and
used 34 reproductive females in unprovisioned areas as controls. To estimate the intensity
of resource competition in provisioned areas, I observed the aggressive and vigilant

behavior of these females.

RESULTS
I compared body mass between provisioned and unprovisioned juvenile males and
females from 4-11 weeks of age. Provisioned juveniles were signiﬁcantly heavier than

unprovisioned juveniles during this interval (Figure 31; ANOVA, F = 432.984, df= 1, P <

99

300
7+ Provisioned females
—o—- Provisioned males

+ Unprovisioned females
—EI—‘ Unprovisioned males &

 
 
  

250 1

M

O

O
1

Body mass (grams)

100 e

 

 

50 l l l l T l l l
4 5 5 7 8 9 1o 11

Age (weeks)

Figure 31 - Body mass of juveniles, 1993-1994. Sample sizes are as follows:
provisioned females: 79, 47, 52, 42, 28, 30, 30, 23; provisioned males: 75,
30, 36, 35, 19, 18, 17, 16; unprovisioned females: 60, 30, 37, 38, 35, 34, 35,
29; unprovisioned males: 51, 30, 36, 42, 36, 30, 32, 25.

100

0.001); however, I observed no sex difference in body mass (Fig. 31; ANOVA, F = 0.703,
df= 1, P = 0.402). Thus, provisioning appears to have increased body mass similarly in
both males and females.

The overall density of females weaning litters in provisioned areas was
approximately twice that in unprovisioned areas (73 provisioned females/hectare vs. 36
unprovisioned females/hectare, respectively). This density difference can be explained in
part by the fact that reproductive females in provisioned areas appeared to shift their home
ranges closer to feeding boxes.

The median size of litters at weaning was signiﬁcantly greater for litters of
provisioned than unprovisioned females (5.0, n = 33 vs. 4.0, n = 32, respectively; Kruskal-
Wallis test, U = 303.5, P = 0.002). The median number of juvenile females in each litter at
weaning was also greater for litters of provisioned than unprovisioned females (3.0, n = 33
vs. 2.0, n = 32, respectively; Kruskal-Wallis test, U = 333.5, P = 0.008).

Prior to the ﬁrst emergence of juveniles from the natal burrow, during the period in
which females were pregnant or nursing young, the mean density of adult S. beldingi
above ground was signiﬁcantly higher in provisioned than unprovisioned areas (Figure 32;
84.0/hectare, n = 135 observations vs. 44.8/hectare, n = 187 observations, respectively,
Kmskal-Wallis test, U = 6134.5, P < 0.001). After the ﬁrst emergence of juveniles from
the natal burrow, mean overall density of S. beldingi was also signiﬁcantly higher in
provisioned than unprovisioned areas (Figure 32; 170.5/hectare, n = 47 observations vs.
87 . l/hectare, n = 62 observations, respectively; Kruskal-Wallis test, U = 659.5, P <

0.001). Thus, adult and yearling females and their young experienced signiﬁcantly higher

101

 

200 ,, .. . .
7 :1 PrOVISloned areas *
[:1 Unprovisioned areas

150 —
A
9
to
‘0—0
0
2
i 100 — *
v
.4?
(I)
C
an
O

50 —

o

 

 

 

 

 

 

Pre-weaning Post-weaning

Figure 32 - Density of S. beldingi, 1993-1995. Mean densities both prior
to and after the ﬁrst weaning of young are presented. Asterisks indicate a
signiﬁcant difference between provisioned and unprovisioned areas.

102

densities of both maternal females and all conspeciﬁcs in provisioned than unprovisioned
areas of the study site.

1 evaluated aggressive and vigilant behavior of females that weaned litters during
the summer of 1995. Pregnant and lactating females in provisioned areas exhibited overall
higher rates of vigilance (Figure 33; Kruskal-Wallis test, U = 3630.5, P < 0.001), and
aggressive behavior (Figure 34; Kruskal-Wallis test, U = 3342.0, P < 0.001) than did
maternal females in unprovisioned areas. Provisioned females had signiﬁcantly higher
rates of vigilance during all four observation periods (Figure 33; Kruskal-Wallis tests,
early gestation: U = 90.0, P < 0.001; late gestation: U = 420.5, P < 0.001; early
lactation: U = 307.0, P = 0.001; late lactation: U = 237.0, P = 0.006). Provisioned
females also had somewhat higher rates of aggression during gestation (Figure 34;
Kruskal-Wallis tests, early gestation: U = 217.5, P = 0.033; late gestation: U = 257.5, P =
0.055), and signiﬁcantly higher rates of aggression during the ﬁrst half of lactation (Figure
34; Kmskal-Wallis test, U = 377.5, P = 0.004) than did unprovisioned females.

Provisioning inﬂuenced the timing of dispersal in both sexes, but altered the
probability of dispersal only by females. A signiﬁcantly larger cumulative percentage of
provisioned than unprovisioned females had dispersed by 7-10 weeks and 55 weeks of age
(Figure 35; 7 weeks: Fisher’s exact test, P = 0.006; 8 weeks: 1’ = 11.647, df= l, P =
0.001; 9 weeks: ,1; = 8.746, df= 1, P = 0.004; 10 weeks: ,1? = 8.142, df= 1, P = 0.004; 55
weeks: Fisher’s exact test, P = 0.013). Cumulative dispersal percentages differed
signiﬁcantly between provisioned females and unprovisioned males only during the 55
week age interval (Figure 35; 55 weeks: Fisher’s exact test, P = 0.006). However, a

signiﬁcantly larger cumulative percentage of provisioned males than provisioned females

103

O)
O
A

—o— Provisioned mothers
——I:I—- Unprovisioned mothers

N 00 h 0'1
0 O O O
l l l l
_L
O

Vigilance (# min/hr/individual)

 

O
J

GESTATION LACTATION
I I I I
40-52 27-39 14 -26 1-13

Days prior to weaning of young

 

 

Figure 33 - Vigilance behavior of adult females, 1995. Mean rates of
vigilance for females that weaned litters in provisioned and unprovisioned
areas are presented. Asterisks indicate a signiﬁcant difference between
females in provisioned and unprovisioned areas. Sample sizes are given
on the graph.

104

.5
.b
J

+ Provisioned mothers
—D—' Unprovnsroned mothers

*
24

.3
N
l

.1.
C
1

co
1

20 1o

4— T\\ ///T
18 \\ /-/”i:“
2— E} 21

Aggression (# min/hr/individual)

 

 

9
1 9
0 — GESTATION LACTATION
I I I I
40-52 27-39 14-26 1-1 3

Days prior to weaning of young

Figure 34 - Aggressive behavior of adult females, 1995. Mean rates of
aggression for females that weaned litters in provisioned and unprovisioned
areas are presented. Asterisks indicate a signiﬁcant difference between
females in provisioned and unprovisioned areas. Sample sizes are given

on the graph.

105

 

 

 

1oo — I:
+ Provisioned females //
—O— Unprovisioned females //

'0 80 _ —I—- Provisioned males /

a) —D—- Unprovisioned males .

‘13 /

a)

0.

.2 50 —

'0

o\°

m

'2 40

E

3

E

3

0 20 -

0

Age (weeks)

 

Figure 35 - Dispersal of the 1993 and 1994 cohorts. The cumulative
percentages of provisioned and unprovisioned males and females that
dispersed as juveniles from 6-10 weeks of age, and as yearling by 55
weeks of age, are presented. Samples of juveniles included 51 provisioned
females, 48 unprovisioned females, 35 provisioned males, and 41
unprovisioned males. Samples of yearlings included 13 provisioned
females, 24 unprovisioned females, 12 provisioned males, and 11

unprovisioned males.

106

dispersed by all ages for which we monitored dispersal behavior (Figure 35; 6 weeks:
Fisher’s exact test, P = 0.009; 7 weeks: i = 6.445, df= l, P = 0.011; f = 8.280, df= 1,
P = 0.004; 9 weeks: i = 7.366, df= 1, P = 0.007; 10 weeks: 1’ = 5.664, df= l, P =
0.017; 55 weeks: Fisher’s exact test, P = 0.005). Thus, cumulative dispersal frequencies
of provisioned females during the juvenile summer more closely resembled those of
unprovisioned males than those of unprovisioned females. However, 100% of surviving
males dispersed before the end of the yearling summer, whereas more than half of
provisioned females remained in their natal areas as yearlings.

Of the 21 dispersed juvenile females that I recovered during 1993 and 1994, eight
had sisters that also dispersed as juveniles. I observed six of these after dispersal within
ﬁve meters of a sister that had also dispersed. By contrast, of 38 juvenile males that
dispersed during 1993 and 1994, 17 had brothers that also dispersed as juveniles, but I
never observed brothers to have overlapping home areas after dispersing. Thus, among
same-sex siblings that dispersed, females were signiﬁcantly more likely to settle near a
sister than males were likely to settle near a brother (Fisher’s exact test, P = 0.004). Of
the 38 dispersed juvenile males that I recovered, eight settled in provisioned areas,
whereas all 21 dispersed juvenile females settled in areas not receiving extra food. Thus,
among dispersed juveniles, females were signiﬁcantly less likely to settle in provisioned
areas than were males (Fisher’s exact test, P = 0.041). Juvenile females that dispersed
apparently avoided settling in provisioned areas.

Two juvenile females from provisioned areas established home areas and sleeping
burrows outside their natal areas for > 1 week, and then returned to their natal areas.

These two females spent the remainder of the juvenile summer in their natal areas. One

107

disappeared over the winter, and the other later dispersed as a yearling. By contrast, I
never observed juvenile males to re—establish themselves in their natal areas after
emigrating.

I compared body mass between dispersed and undispersed juvenile females.
Overall, females that had dispersed were signiﬁcantly heavier than were same-aged
females still residing in their natal areas overall from 7-11 weeks of age (Figure 36;
ANOVA, F = 4.289, df= 1, P = 0.039). The difference in body mass between dispersed
and natal females was marginally signiﬁcant at 10 and 11 weeks of age, after signiﬁcance
levels were adjusted (Figure 36; t-tests for independent samples, 10 weeks: t = 2.245, df
= 20, P = 0.036; 11 weeks: t = 2.415, df= 35, P = 0.021). Thus, females that dispersed
apparently did not have limited access to food in either their natal areas, or in their new
home areas.

I evaluated juveniles’ body mass relative to time of dispersal. Body mass of
provisioned females did not differ signiﬁcantly from that of unprovisioned females from
two weeks prior to dispersal to two weeks after dispersal (t-test for independent samples; t
= 1.315, df= 37, P = 0.197), and so I combined body mass data from these two groups
for statistical analysis. Similarly, provisioned and unprovisioned males did not differ in
body mass from two weeks prior to dispersal to two weeks subsequent to dispersal (t—test
for independent samples, t= -l.644, df= 104, P = 0.103), so I also combined body mass
data from these two groups for statistical analysis. Overall, juvenile male S. beldingi were
signiﬁcantly heavier than juvenile female S. beldingi from two weeks prior to dispersal to
two weeks after dispersal (Figure 37; ANOVA, F= 14.102, df= l, P < 0.001). After

signiﬁcance levels of pairwise comparisons were adjusted, differences in body mass were

108

300 —

 

 

 

—e— Dispersed females
—D—- Females in natal area
‘é’ 250 4
‘3
P)
(D
(I)
to
E
>. zoo -
'0
o
m
150 — 78
l l l I l
7 8 9 1o 11
Age (weeks)

Figure 36 - Body mass of dispersed and natal females, 1993-1994. Sample
sizes age given on the graph.

109

 

 

 

300 —
-O— Females :
—I:I—- Males : 5
l
I
250 — I
75 l
E
(U
L.
3’
(D 200 -1
(I)
E
a 2:
o )2 I
150 .4 I
m / 30 g
/ l
l
E/ I
|
100 — 23 .
l l i l l
-2 -1 0 1 2

Time relative to dispersal (weeks)

Figure 37 - Body mass of juveniles relative to dispersal, 1993-1994. The
vertical dashed line indicates the week of dispersal. Sample sizes are given
on the graph.

110

marginally signiﬁcant at l and 2 weeks prior to dispersal and at l and 2 weeks subsequent
to dispersal (Fig. 37; t-tests for independent samples; —2 weeks: t = -2.104, df= 29, P =
0.044; -1 week: t = -1.989, df= 38, P = 0.054; +1 week: t = -2.010, df= 21, P = 0.057;
+2 weeks: t = -2.293, df= 12, P = 0.041).

I also observed sex differences in the ages at which juvenile S. beldingi dispersed.
The median age of dispersal was signiﬁcantly younger for provisioned males than for
provisioned females (7.0 weeks, n = 20 vs. 7 .5 weeks, n = 17, respectively; Kruskal-Wallis
test, U = 219.5, P = 0.044), and was also signiﬁcantly younger for unprovisioned males
than for unprovisioned females (8.0 weeks, n = 18 vs. 9.5 weeks, n = 4, respectively;
Kruskal-Wallis test, U = 62.0, P < 0.001) and between provisioned and unprovisioned
females (7.5 weeks, n = 16 vs. 9.5 weeks, n = 4, respectively; Kruskal-Wallis test, U =
11.0, P = 0.038). Thus, males dispersed at signiﬁcantly younger ages than did females in
both the provisioned and unprovisioned groups, and provisioned animals dispersed at

younger ages in both sexes.

DISCUSSION

I observed a sex difference in the dispersal behavior of S. beldingi in response to
food provisioning. All surviving male S. beldingi in both provisioned and unprovisioned
groups dispersed by the end of their yearling year. Thus, my results conﬁrm the ﬁnding of
Holekamp (1984a, 1986) that all male S. beldingi eventually emigrate from the natal area,
and support an hypothesis that male S. beldingi disperse for reasons other than to enhance
access to environmental resources. Young male S. beldingi disperse probably either to

avoid breeding with close kin (Sherman, 1976; Holekamp, 1984b), or to enhance access to

111

receptive mates (Holekamp, 1986). In contrast to young male S. beldingi, young females
weaned in areas receiving extra food were signiﬁcantly more likely to emigrate ﬁom the
natal area than were unprovisioned females. Holekamp (1984b, 1986) observed that
about 8% of females ordinarily disperse from the natal area by 55 weeks of age. Similarly,
about 8% of unprovisioned females in the 1993 and 1994 cohorts emigrated from the
natal area, whereas over 40% of surviving provisioned females eventually dispersed.

Three lines of evidence suggest that competition for food was not the impetus for
juvenile female dispersal in this study. First, juvenile females were more likely to emigrate
from provisioned areas than unprovisioned areas. Second, no young female S. beldingi
that dispersed settled in a provisioned area. Finally, dispersed juvenile females were
heavier throughout the summer than same-aged females remaining in their natal areas, and
thus, females that dispersed were apparently not those least successful at competing for
food.

The overall density of S. beldingi and the density of female S. beldingi weaning
litters were higher in provisioned than unprovisioned areas of my study site. Rates of
aggression and vigilance were signiﬁcantly higher among provisioned than unprovisioned
mothers. This suggests that competition for non-food resources was greater in
provisioned areas, and that young female S. beldingi may have dispersed to ﬁnd areas in
which this competition was less intense. Factors other than food availability are proximal
stimuli for natal dispersal many female mammals (Wolff, 1993a, 1994). For example,
female white-footed mice (Peromyscus leucopus) disperse in response to the presence of

the opposite-sex parent and the availability of vacant habitat nearby (Wolﬁ’, 1992).

112

I suggest that competition for space may have increased the frequency of dispersal
by provisioned juvenile females in 1993 and 1994. Female S. beldingi defend maternal
territories during gestation and lactation, as do many small mammals (Dobson, 1983;
Wolff, 1993b). Infanticide is also common among small mammals, and reproductive
females typically defend territories to protect their young ﬁom infanticidal conspeciﬁcs
(e.g., Sherman 1981b, 1982; McClean, 1983; Hoogland, 1985, 1995; Dobson, 1990;
Hare, 1991; Vestal, 1991; Wolff, 1993b). Under conditions of high population density,
costs to females of defending territories may be especially high, or the likelihood of
successfully defending young may be relatively low. Young females might disperse from
areas of high density to improve their future chances of obtaining a territory they can
successfully defend. Increases in the relative frequency of dispersal in response to
increases in conspeciﬁc density and aggression are in fact common among small mammals
in unsaturated habitats (Christian, 1970; Lidicker, 1975; Gaines and McClenaghan, 1980).

I did not monitor dispersal behavior of adult and yearling female S. beldingi from
provisioned and unprovisioned areas subsequent to the weaning of their litters. However,
one might expect the relative frequency of emigration by these females to be especially
high in areas with intense competition for space. For example, female Columbian ground
squirrels (S. columbianus) and female red squirrels (T amiasciurus hudsonicus) sometimes
disperse so that their offspring can inherit their territories (Harris and Murie, 1984; Price
and Boutin, 1993). Alternatively, females that have weaned litters might be more
behaviorally experienced than juveniles and thus better able to successfully compete for
resources or defend young in the future, and therefore be less likely to disperse. For

example, primiparous rhesus monkeys (Macaca mulatta) are more likely to be targets of

113

aggression than are multiparous mothers, and primiparous female prairie voles (M.
ochragaster) exhibit less proﬁcient parental care than do multiparous mothers (Hooley
and Simpson, 1981; Wang and Novak, 1994).

In food provisioning experiments with S. beecheyi, Dobson (1979) observed an
increase in immigration of young females into an area receiving extra food. I suggest that
the discrepancy between my results and those of Dobson (1979) may be due to differences
in provisioning protocols between the two studies. I provided extra food throughout
gestation and lactation, and subsequent to weaning of young. Dobson (1979) began
provisioning near the time young were weaned, and put food out openly where many
individuals could access it simultaneously. Immigration in the study of Dobson (1979)
was therefore associated with increased access to food without the cost of facing
territorial aggression from maternal females.

Trombulak (1991) provided extra food to reproductive female S. beldingi during
gestation and lactation, as I did. However, he provisioned individual females randomly
throughout his study site and did not give extra food to all females in a provisioned area.
He observed that juvenile male but not juvenile female S. beldingi began disappearing
from their natal areas 2-3 weeks after weaning. Trombulak (1991) was not speciﬁcally
examining dispersal, so he reported disappearance but not dispersal of young S. beldingi.
Thus, his results are not directly comparable to mine. However, food provisioning studies
of house mice (Mus musculus) support the idea that spatial distribution of resources can
inﬂuence dispersal behavior. M. musculus are more likely to have a rigid social structure
(Stueck and Barrett, 1978; Noyes et al., 1982) and relatively high dispersal rates (Maly et

al., 1985) when extra food is provided at centralized than scattered feeding stations.

114

Dispersed females in my study were heavier than same-aged females remaining in
their natal areas, which suggests that energetic or other costs might be potentially
associated with female dispersal (Nunes and Holekamp, 1996; see also Gaines and
McClenaghan, 1980; Wauters et al., 1994; Isbell and Van Vuren, 1996). Moreover, a
tradeoﬁ’ might exist between the potential costs and beneﬁts of dispersal (Isbell and Van
Vuren, 1996) for young females. Small or average-sized females that reside in areas
characterized by high levels of intraspeciﬁc competition might do better to remain in these
areas until competition declines or until they grow and become more competitive, rather
than risk the potential costs of dispersal. Density and aggression might be proximate
indicators of competition in future years. Immediate and future beneﬁts of dispersal, such
as reduced conspeciﬁc density and competition, might outweigh its costs only for larger or
fatter than average females able to contend effectively with rigors of the dispersal
movement itself.

I observed striking sex differences in the patterns of natal dispersal by S. beldingi.
Females began dispersing at later ages than did males in both provisioned and
unprovisioned groups. Home areas of dispersed sisters, but not of dispersed brothers,
sometimes overlapped after emigration from the natal area. Sisters may settle near each
other to gain beneﬁts to ﬁtness from nepotistic cooperation in defense of maternal
territories (Sherman, 198 la), whereas brothers may avoid settling in the same area to
avoid competing with close kin for mates (Sherman, 1976). These trends are similarly
apparent in the dispersal patterns of many gregarious animals in which females emigrate en
masse via group ﬁssion but males disperse alone or with their male siblings (e. g., Chepko-

Sade and Sade, 1979; Hanby and Bygott, 1987; Pusey and Packer, 1987b; Dittus, 1988;

115

Holekamp et al., 1993). In the present study, I also observed that juvenile female, but not
juvenile male, S. beldingi sometimes abandoned attempts to disperse, and re-established
home sites in the natal area, a pattern also observed among females of other gregarious
mammals (Smale et al., 1997). Young female S. beldingi might return to the natal site
when dispersal fails to improve their immediate or potential ﬁiture ability to compete for
resources. Thus, dispersal by juvenile female S. beldingi appears to be more labile, and

more responsive to social and ecological variables, than does dispersal by juvenile males.

Chapter 8

SUMNIARY AND CONCLUSIONS

INFLUENCES OF MASS AND FA T ON DISPERSAL

I examined the inﬂuences of body mass and body fat on the timing of natal
dispersal behavior in juvenile male Belding’s ground squirrels (Spermophilus beldingi). I
manipulated body mass and body fat of juveniles by provisioning them with food, and I
used unprovisioned juveniles as controls. I evaluated the percentage of body mass
composed of fat using the non-destructive method of measuring electrical conductivity of
the body. Provisioned males were signiﬁcantly heavier than unprovisioned males
throughout the juvenile summer, and dispersed at younger ages. Body mass of juvenile
males at dispersal was constant regardless of the age at which individuals dispersed.
Although body mass differed between provisioned and unprovisioned males when
evaluated with respect to age, it did not differ when considered relative to time of
dispersal. These results indicate that juvenile male S. beldingi postpone dispersal until
they have attained a threshold body mass. A greater proportion of provisioned than
unprovisioned males dispersed in the 6-8 week age intervals, when provisioned males were
fatter than their unprovisioned counterparts. However, at ages 9-10 weeks of age there
was no signiﬁcant difference between provisioned and unprovisioned males in either body

fat or the proportion dispersed. Moreover, dispersing juvenile males were fatter than were

116

117

same-aged juvenile males still residing in their natal areas. This close association between
body fat and dispersal suggests that emigrating from the natal area has potential energetic
costs, and that young males may delay dispersal until their energy reserves, as reﬂected in

body fat, are adequate to meet these energetic demands.

ENDOGENO US TIMING MECHANISMS

I assessed the relationship between dispersal behavior, body mass and body fat of
juvenile male S. beldingi over three consecutive years, one of which had a substantially
delayed active season due to heavy winter and spring snowfall. Body fat was evaluated
using the non-destructive method of measuring electrical conductivity of the body. When
the active season began late, acquisition of body mass and body fat were accelerated, and
dispersal behavior was inhibited. This suggests that ontogenetic trends in patterns of mass
and fat gain in natal dispersal behavior by juvenile male S. beldingi are inﬂuenced by a
seasonal time-keeping mechanism. Energy allocation was heirarchically organized in S.
beldingi, with pre-hibemation fattening taking precedence over dispersal.

My results suggest that physiological signals reﬂecting body fat interact with an
endogenous timing mechansim in S. beldingi to regulate the dispersal behavior of juvenile
males. For example metabolic fuels such as glucose or fatty acids, or metabolic hormones
such as glucagon, insulin, or leptin, might be inﬂuenced by the size of fat reserves, and
might convey information about body fat to time-keeping and other mechanisms in the

central nervous system which inﬂuence natal dispersal behavior.

118

ENDOCRINE MEDIA TION OF DISPERSAL

I tested competing hypotheses suggesting that (l) testosterone (T) activates the
natal dispersal behavior of juvenile male S. beldingi during ontogeny, and (2) that during
neonatal life T organizes tissues which later mediate the natal dispersal behavior typical of
juvenile male S. beldingi, as well as the responsiveness of masculine natal dispersal
behavior to energetic variables such as body fat. I collected blood samples from free-
living juvenile S. beldingi during the primary dispersal period to measure concentrations of
circualting T. I also concurrently manipulated perinatal T exposure and body fat of ﬁ'ee-
living S. beldingi. I treated newborn females with T, and used oil-treated females as
controls. Within T- and oil-treated groups, I manipulated body fat of females and their
brothers by food provisioning, and used unprovisioned females and their brothers as
controls. Plasma T concentrations of juveniles were undetectable, suggesting that T does
not activate natal dispersal behavior during ontogeny. However, treatment of neonatal
females with T induced masculine dispersal behavior, the timing of which varied with the
amount of energy stored as fat. Thus, whereas energy reserves inﬂuenced when dispersal

would occur, the probability of dispersal was determined by early T exposure.

ECOLOGICAL FACTORS INFLUENCING DISPERSAL

I examined the effects of food provisioning on the natal dispersal behavior of
juvenile male and female S. beldingi. I provided extra food to adult and yearling females
in their maternal territories during pregnancy and lactation, and to offspring of these
females in their natal areas for 6 weeks after weaning. I used unprovisioned young of

unprovisioned mothers as controls. Provisioning inﬂuenced the probability of emigration

119

from the natal area by female but not male S. beldingi. All surviving male S. beldingi
dispersed by 55 weeks of age, regardless of whether they and their mothers received extra
food. By contrast, I observed a signiﬁcant trend, beginning 3 weeks after weaning and
continuing through the yearling year, for a greater proportion of provisioned than
unprovisioned female S. beldingi to emigrate from the natal area. Competition for food
did not appear to inﬂuence natal dispersal of females. However, overall population
density, density of females weaning litters, and rates of aggression and vigilance among
these females, were higher in provisioned than unprovisioned areas, suggesting that
competition for non-food resources was unusually intense in provisioned areas. It appears
that juvenile female, but not juvenile male, S. beldingi may emigrate from the natal site to
increase access to areas with low densities of conspeciﬁcs. Together with ﬁndings of
other researchers, these results suggest that spatial and temporal distributions of
environmental resources are important inﬂuences on the dispersal behavior of female

ground squirrels.

DIRECTIONS FOR FUTURE RESEARCH

My results have generated many interesting questions, which could feasibly be
addressed in ﬁxture research using S. beldingi or other species as a model system.

First, my results revealed a close association between body mass and the timing of
natal dispersal in S. beldingi. Juvenile males tend to remain in their natal areas until they
reach a threshold body mass. This indicates that in S. beldingi, and possibly in other
species, a requisite amount of physical and behavioral development may need to occur

before an individual disperses. For example, before dispersing an individual may need to

120

grow to a large enough size to deal with social resistance from conspeciﬁcs in its new
home area, or develop adequate predator evasion skills. However, the exact
developmental prerequisites for dispersal remain to be determined.

1 also observed a close association between the timing of natal dispersal and energy
stored as fat in S. beldingi. In particular, after achieving a threshold body mass, juvenile
males appear to delay dispersal until they have acquired a threshold amount of body fat
appropriate for the particular time of the season. This association between dispersal and
body fat suggests that energetic costs are potentially associated with dispersal in S.
beldingi and possibly other species, and that animals have evolved physiological
mechanisms to deal with the energetic rigors of dispersing. However, the actual energetic
demands of dispersal have not yet been be documented for any species.

Emigration from the natal area occurs at very speciﬁc stages of the ontogenetic
development of animals. Most male S. beldingi disperse as juveniles, and under ordinary
circumstances, all disperse before becoming reproductively mature at two years of age
(Holekamp 1984, 1986). The energetic costs of dispersal seem especially great for
juvenile S. beldingi, which typically have limited time in which to both disperse and fatten
for hibernation. A question that arises, then, is what survival or reproductive beneﬁts are
accrued by male S. beldingi that disperse as juveniles, versus those that disperse as
yearlings? More generally, what are the ultimate factors that determine when during
ontogeny an individual emigrates from its natal area?

The speciﬁc mechanisms by which body fat inﬂuences dispersal behavior are
unknown. My work suggests that fat-related physiological signals inform systems that

promote dispersal behavior about the size of body fat reserves. Metabolic ﬁrels such as

121

glucose, or metabolic hormones such as leptin, are possible candidates for such signals.
However, whether either of these plays a role in mediating dispersal remains to be

determined.

LIST OF REFERENCES

Alberts, S. C. and J. Altmann. 1995. Balancing costs and opportunities: dispersal in
male baboons. American Naturalist, 145:279-306.

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

Arcese, P. 1989. Intrasexual competition, mating systems and natal dispersal in song
sparrows. Animal Behaviour, 38:958-979.

Armitage, K. B. and J. F. Downhower. 1974. Demography of yellow-bellied marmot
populations. Ecology, 55: 1233-1244.

Armitage, K. B., J. F. Downhower, and G. E. Svendsen. 1976. Seasonal changes in
weights of marmots. American Midland Naturalist, 96:36-51.

Arnold, A. P. 1980. Sexual diﬂ‘erences in the brain. American Scientist, 68: 165-173.
Arnold, A. P. and S. M. Breedlove. 1985. Organizational and activational effects of
sex steroids on brain and behavior: a reanalysis. Hormones and Behavior,

19:469-498.

Bachman, G. C. 1993. The effect of body condition on the trade-oﬁ’ between vigilance
and foraging in Belding's ground squirrels. Animal Behaviour, 46:233-244.

Bachman, G. C. 1994. Food restriction effects on the body composition of free-living
ground squirrels, Spermophilus beldingi. Physiological Zoology, 67:756-770.

Balph, D. F. and A. W. Stokes. 1963. On the ethology of a population of Uinta ground
squirrels. American Midland Naturalist, 69: 106-126.

Barash, D. P. 1973. The social biology of the Olympic marmot. Animal Behavior
Monographs, 6: 173-245.

122

123

Barnes, B. M. 1984. Inﬂuence of energy stores on activation of reproductive function in
male golden-mantled ground squirrels. Journal of Comparative Physiology,
B1541421-425.

Barnes, B. M. and A. D. York. 1990. Eﬂ'ect of winter high temperatures on reproduction

and circannual rhythms in hibernating ground squirrels. Journal of Biological
Rhythms, 5:119-130.

Bartness, T. J., J. B. Powers, M. H. Hastings, E. L. Bittman, and B. D. Goldman. 1993.
The timed inﬁrsion paradigm for melatonin delivery: what has it taught us about
the melatonin signal, its reception, and the photoperiodic control of seasonal
responses. Journal of Pineal Research, 15:161-190.

Baum, M. J ., R. S. Carroll, J. A. Cherry, and S. A. Tobet. 1990. Steroidal control of
behavioral, neuroendocrine and brain sexual differentiation: studies in a carnivore,
the ferret. Journal of Neuroendocrinology, 4:401-418.

Beatty, W. W. 1992. Gonadal hormones and sex diﬁ’erences in nonreproductive
behaviors. Pp. 85-128 in Handbook of behavioral neurobiology, volume 11:
sexual differentiation (A. A. Gerall, H. Moltz, and I. L. Ward, eds.). Plenum
Press, NewYork.

Bennett, A. F. 1990. Habitat corridors and the conservation of small mammals in a
fragmented forest environment. Landscape Ecology, 4:109-122.

Berger, J. 1986. Wild horses of the Great Basin. University of Chicago Press, Chicago.

Berger, J. 1987. Reproductive fates of dispersers in a harem-dwelling ungulate: The
wild horse. Pp. 41-54 in Mammalian dispersal patterns: the effects of social
structure on population genetics (D. B. Chepko-Sade and Z. T. Halpin, eds.).
University of Chicago Press, Chicago.

Boutin, S. 1990. Food supplementation experiments with terrestrial vertebrates:
patterns, problems, and the future. Canadian Journal of Zoology, 68:
203-220.

Breedlove, S. M. 1992. Sexual dimorphism in the vertebrate nervous system. Journal of
Neuroscience, 12:4133-4142.

Bronson, F. H. 1987. Puberty in female rats: relative effect of exercise and
restriction. American Journal of Physiology, 252:R140-R144.

Bronson, F. H. 1989. Mammalian reproductive biology. University of Chicago Press,
Chicago.

124

Bronson, F. H. and E. F. Rissman. 1986. The biology of puberty. Biological Reviews,
61:157-195.

Bushberg, D. M. and W. G. Holmes. 1985. Sexual maturation in male Belding's ground
squirrels: Inﬂuences of body weight. Biology of Reproduction, 33 2302-308.

Caley, M. J. 1987. Dispersal and inbreeding avoidance in muskrats. Animal Behaviour,
35:1225-1233.

Campﬁeld, L. A., F. J. Smith, Y. Guisez, R. Devos, and P. Burn. 1995. Recombinant
mouse OB protein: evidence for a peripheral signal linking adiposity and central
neural networks. Science, 269:546-549.

Charlesworth, D. and B. Charlesworth. 1987. Inbreeding depression and its evolutionary
consequences. Annual Review of Ecology and Systematics, 18:23 7-268.

Chepko-Sade, B. D. and D. S. Sade. 1979. Patterns of group splitting within matrilineal
kinship groups. Behavioral Ecology and Sociobiology, 5:67-86.

Chepko-Sade, B. D. and Z. Tang Halpin (Eds). 1987 . Mammalian dispersal patterns:
the effects of social structure on population genetics. The University of Chicago
Press, Chicago.

Christian, J. J. 1970. Social subordination, population density, and mammalian evolution.
Science, 168284-90.

Cockbum, A., M. P. Scott, and D. J. Scotts. 1985. Inbreeding avoidance and male-
biased dispersal in Antechinus spp. (Marsupiala: Dasyuridae). Animal Behaviour,
33 :908-915.

Cohen, B., D. Novick, and M. Rubinstein. 1996. Modulation of insulin activity by leptin.
Science, 274: 1 185-1 188.

Dark, J ., N. G. Forger, and I. Zucker. 1984. Rapid recovery of lipid mass after surgical
removal of adipose tissue in ground squirrels. Proceedings. National Academy of
Sciences 81:2270-2272.

Dark, J ., N. G. Forger, J. S. Stern, and I. Zucker. 1985. Recovery of lipid mass after

removal of adipose tissue in ground squirrels. American Journal of Zoology, 249:
R73 -R78.

Dark, J ., N. F. Ruby, G. N. Wade, P. Licht, and I. Zucker. 1992. Accelerated
reproductive development in juvenile male ground squirrels fed a high fat diet.
American Journal of Physiology, 262:R644-R650.

125

Davis, D. E. 1976. Hibernation and circannual rhythms of food consumption in marmots
and ground squirrels. Quarterly Review of Biology, 51:477-514.

Diffendorfer, J. E., M. S. Gaines, and R. D. Holt. 1995. Habitat ﬁ'agmentation and
movements of three small mammals (Sigmondon, Microtus, and Peromyscus).
Ecology, 76:827-839.

Dingle, H. 1996. Migration: the biology of life on the move. Oxford University Press,
New York.

Dittus, W. P. J. 1988. Groups ﬁssion among wild toque macaques as a consequence of
female resource competition and environmental stress. Animal Behaviour, 36:
1626-1645.

Dobson, F. S. 1979. An experimental study of dispersal in the California ground squirrel.
Ecology, 60: 1 103-1 109.

Dobson, F. S. 1982. Competition for mates and predominant juvenile male dispersal in
mammals. Animal Behaviour, 30: 1 183-1 192.

Dobson, F. S. 1983. Agonsim and territoriality in the California ground squirrel. Journal
of Mammalogy, 64:218-225.

Dobson, F. S. 1990. Environmental inﬂuences on infanticide in Columbian ground
squirrels. Ethology, 84:3-14.

Dobson, F. S. 1995. Regulation of population size: evidence from Columbian ground
squirrels. Oecologia, 102:44-51.

Dobson, F. S. and W. T. Jones. 1985. Multiple causes of dispersal. American Naturalist,
126:855-858.

Dobson, F. S. and J. D. Kjelgaard. 1985. The inﬂuence of food resources on population
dynamics in Columbian ground squirrels. Canadian Journal of Zoology, 63:2095-
2104.

Docekal, E. M. 1978. Home range, activities, and diet of the Belding's ground squirrel,
Spermophilus beldingi beldingi, at Sagehen Creek, California. M.A. thesis.
San Francisco State University, San Francisco, California.

Downhower, J. F., and K. B. Armitage. 1981. Dispersal of yearling yellow-bellied
marmots (Marmotaﬂaviventris). Animal Behaviour, 29: 1064-1069.

Dunford, C. 1977. Behavioral limitation of round-tailed ground squirrel density.
Ecology, 58:1254-1268.

126

Errington, P. L. 1963. Muskrat populations. Iowa State University Press, Ames.

Forger, N. G., J. Dark, and I. Zucker. 1986. Recovery of white adipose tissue after
lipectomy in female ground squirrels. Canadian Journal of Zoology, 64: 128-13 1.

French, AR. 1982. Intraspeciﬁc differences in the pattern of hibernation in the ground
squirrel Spermophilus beldingi. Journal of Comparative Physiology, 148:83-91.

Gaines, M. S. and L. R. McClenaghan, Jr. 1980. Dispersal in small mammals. Annual
Review of Ecology and Systematics, 11:163-196.

Goy, R. W. and B. S. McEwen. 1980. Sexual differentiation of the brain. M. I. T.
Press, Cambridge.

Greenwood, P. J. 1980. Mating systems, philopatry and dispersal in birds and mammals.
Animal Behaviour, 28: 1 140-1 162.

Halaas, J. L., K. S. Gajiwala, M. Maffei, S. L. Cohen, B. T. Chait, D. Rabinowits, R. L.
Lallone, S. K. Burley, and J. M. Friedman. 1995. Weight-reducing effects of
the plasma protein encoded by the obese gene. Science, 269:543-546.

Halpin, Z. T. 1987. Natal dispersal and the formation of new social groups in a newly
established town of black-tailed prairie dogs (Cynomys ludovicianus). Pp.
104-118 in Mammalian dispersal patterns: the effects of social structure on
population genetics (B. D. Chepko-Sade and Z. T. Halpin, eds). The University
of Chicago Press, Chicago.

Hanby, J. P. and J. D. Bygott. 1987. Emigration of subadult lions. Animal Behaviour,
35:161-169.

Hanken, J. and P. W. Sherman. 1981. Multiple paternity in Belding's ground squirrel
litters. Science, 212:3 51-3 53.

Harcourt, A. H. 1978. Strategies of emigration and transfer by primates, with particular
reference to gorillas. Zeitschrift ﬁier Tierpsychologie, 48:401-420.

Harcourt, A. H., K. S. Stewart, and D. Fossey. 1976. Male emigration and female
transfer in wild mountain gorillas. Nature, 263:226-227.

Hare, J. F. 1991. Intraspeciﬁc killing of preweaned juvenile young in the Columbian
ground squirrel, Spermophilus columbianus. Canadian Journal of Zoology,
69:797-800.

Harris, M. A. and J. O. Murie. 1984. Inheritance of nest sites in female ground squirrels.
Behavioral Ecology and Sociobiology, 15297-102.

127

Hoar, W. S. 1988. The physiology of smolting salmonids. Fish Physiology XIB:
273-343.

Holekamp, K. E. 1983. Proximal mechanisms of natal dispersal in Belding's ground
squirrels (Spermophilus beldingi beldingi). Ph. D. dissertation. University of
California, Berkeley, California.

Holekamp, K. E. 19848. Natal dispersal in Belding's ground squirrels (Spermophilus
beldingi). Behavioral Ecology and Sociobiology, 16:21-30.

Holekamp, K. E. 1984b. Dispersal in ground-dwelling sciurids. Pp. 297-320 in The
biology of ground-dwelling sciurids (J. O. Murie and G. R. Michener, eds.)
University of Nebraska Press, Lincoln, Nebraska.

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

Holekamp, K. D. and F. Talamantes. 1992. Seasonal ﬂuctuations in hormones and
behavior of free-living male California ground squirrels (Spermophilus beecheyi).
Hormones and Behavior, 2627-23.

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

Holekamp, K. E., L. Smale, H. Blair Simpson, and N. A. Holekamp. 1984. Hormonal
inﬂuences on natal dispersal in free-living Belding's ground squirrels
(Spermophilus beldingi). Hormones and Behavior, 18:465-483.

Holmes, W. G. 1986a. Identiﬁcation of paternal half-siblings by captive Belding's ground
squirrels. Animal Behaviour, 34:321-327.

Holmes, W. G. 1986b. Kin recognition by phenotype matching in female Belding's
squirrels. Animal Behaviour, 34:38-47.

Holmes, W. G. and P. W. Sherman. 1982. The ontogeny of kin recognition in two
species of ground squirrels. American Zoologist, 22:491-517.

Hoogland, J. L. 1985. Infanticide in prairie dogs: lactating females kill offspring of close
kin. Science, 230: 1037-1040.

Hoogland, J. L. 1995. The black-tailed prairie dog: social life of a burrowing mammal.
University of Chicago Press, Chicago.

128

Hooley, J. M. and M. J. Simpson. 1981. A comparison of primiparous and multiparous
mother-infant dyads in Macaca mulatta. Primates, 22:379-392.

Isbell, L. A. and D. Van Vuren. 1996. Differential costs of locational and social dispersal
and their consequences for group-living primates. Behaviour, 133:1-36.

Iverson, S. L. and B. N. Turner. 1974. Winter weight dynamics in Microtus
pennsylvanicus. Ecology, 55: 1030-1041.

Jenkins, S. H. and B. D. Eshelman. 1984. Spermophilus beldingi. Mammalian Species,
22121-8.

Keane, B. 1990. Dispersal and inbreeding avoidance in the white-footed mouse,
Peromyscus leucopus. Animal Behaviour, 40: 143-152.

Keith, T. P. and R. H. Tamarin. 1981. Genetic and demographic differences between
dispersers and residents in cycling and noncycling vole populations. Journal
of Mammalogy, 62:713-725.

Kenagy, G. J. 1980. Interrelation of endogenous annual rhythms of reproduction and
hibernation in the golden-mantled ground squirrel. Journal of Comparative
Physiology, 135: 333-339.

Kenagy, G. J., D. Masman, S. M. Sharbaugh, and K. A. Nagy. 1990a. Energy
expenditure during lactation in relation to litter size in ﬁ'ee-living golden-
mantled ground squirrels. Journal of Animal Ecology, 59:73-88.

Kenagy, G. J ., S. M. Sharbaugh, and K. A. Nagy. 1990b. Annual cycle of energy and
time expenditure in a golden-mantled ground squirrel population. Oecologia,
78:269-282.

Kilgore, D. L. 1972. Energy dynamics of the yellow-bellied marmot (Marmota
ﬂaviventris). Ph.D. Dissertation. University of Kansas, Lawrence, Kansas.

King, J. A. 1955. Social behavior, social organization, and population dynamics of
a black-tailed prairie dog town in the Black Hills of South Dakota. Contributions:
Laboratory of Vertebrate Biology University of Michigan, 67: 1-123.

Lambin, X. 1994. Natal philopatry, competition for resources, and inbreeding avoidance
in Townsend's voles (Microtus townsendii). Ecology, 75:224-235.

Lens, L. and A. A. Dhondt. 1994. Effects of habitat fragmentation on the timing of
Crested Tit, Parus cristatus natal dispersal. Ibis, 136: 147-152.

129

Lidicker, W. 2., Jr. 1962. Emigration as a possible mechanism permitting the regulation
of population density below carrying capacity. American Naturalist, 96:29-33.

Lidicker, W. 2., Jr. 1975. The role of dispersal in the demography of small mammals.
Pp. 103-108, in Small mammals: their productivity and population dynamics (F.
B. Golley, K. Petrusewicz, and L. Ryszkowski, eds). Cambridge University Press,
Cambridge, United Kingdom.

Maffei, M., J. Halaas, E. Ravussin, R. E. Pratley, G. H. Lee, Y. Zhang, H. Fei., S. Kim,
R. Lallone, S. Ranganathan, P. A. Kern, and J. M. Friedman. 1995. Leptin
levels in human and rodent: measurement of plasma leptin and ob RNA in obese
and weight-reduced subjects. Nature Medecine, 1:1155-1161.

Maly, M. S., B. A. Knuth, and G. W. Barrett. 1985. Effects of resource partitioning on
dispersal behavior of feral house mice. Journal of Mammalogy, 66: 148-153.

Mansukhani, V., E, Adkins-Regan, and S. Yang. 1996. Sexual partner preference in
female zebra ﬁnches: the role or early hormones and social environment.
Hormones and Behavior, 30:506-513.

Matsumura, S. 1993. Intergroup afﬁliative interactions and intergroup transfer of young
male Japanese macaques (Macacafuscata). Primates, 34: 1-10.

Maxwell, C. S. and M. L. Morton. 1975. Comparative thermoregulatory capabilities of
neonatal ground squirrels. Journal of Mammalogy, 56:821-828.

McClean, I. G. 1983. Paternal behavior and killing of young in Arctic ground squirrels.
Animal Behaviour, 31 :32-44.

McEwen, B. S. 1981. Neural gonadal steroid actions. Science, 211:1301-1311.

Messier, F. 1985a. Solitary living and extraterritorial movements of wolves in relation
to social status and prey abundance. Canadian Journal of Zoology, 63:239-245.

Messier, F. 1985b. Social organization, spatial distribution, and population density of
wolves in relation to moose density. Canadian Journal of Zoology, 63: 1068-1077.

Metzgar, L. M. 1967. An experimental comparison of owl predation on resident and

transient white-footed mice (Peromyscus leucopus). Journal of Mammalogy,
48:387-391.

Michener, G. R. 1983. Kin identiﬁcation, matriarchies, and the evolution of sociality
in ground-dwelling sciurids. Pp. 528-572 in Advances in the study of mammalian
behavior (J. F. Eisenberg and D. G. Kleiman, eds). American Society of
Mammalogists Special Publication No. 7.

130

Mills, M. G. L. 1990. Kalahari hyaenas. Unwin Hyman, London.

Moore, J. and R. Ali. 1984. Are dispersal and inbreeding avoidance related? Animal
Behaviour, 32294-112.

Moore, M. C. 1991. Application of organizational -- activational theory to alternative
male reproductive strategies: a review. Hormones and Behavior, 25: 154-179.

Morin, L. P. 1986. Environment and hamster reproduction: responses to phase-
speciﬁc starvtion during estrous cycle. American Journal of Physiology,
251 :R663-R669.

Morton, M. L. 1975. Seasonal cycles of body weights and lipids in Belding ground
squirrels. Bulletin Southern California Academy of Sciences, 74:128-143.

Morton, M. L. and J. S. Gallup. 1975. Reproductive cycle of the Belding ground squirrel
(Spermophilus beldingi beldingi): seasonal and age differences. Great Basin
Naturalist, 35:427-433.

Morton, M. L. and P. W. Sherman. 1978. Effects of a spring snowstorm on behavior,
reproduction, and survival of Belding's ground squirrels. Canadian Journal of
Zoology, 56:2578-2590.

Morton, M. L. and H. L. Tung. 1971. Grth and development in the Belding ground
squirrel (Spermophilus beldingi beldingi). Journal of Mammalogy, 52:611-616.

Morton, M. L., C. S. Maxwell, and C. E. Wade. 1974. Body size, body composition, and
behavior of juvenile Belding ground squirrels. Great Basin Naturalist, 34:121-134.

Mrosovsky, N. 1976. Lipid programmes and life strategies in hibernators. American
Zoologist, 16:685-697.

Mrosovsky, N. and K. C. Fisher. 1970. Sliding set points for body weight in ground
squirrels during the hibernation season. Canadian Journal of Zoology, 48:241-247.

Mrosovsky, N. and T. L. Powley. 1977. Set points for body weight and fat. Behavioral
Biology, 20:205-223.

Murie, J. O. and D. A. Boag. 1984. The relationship of body weight to overwinter
survival in Columbian ground squirrels. Journal of Mammalogy, 65:688-690.

Nelson, M. E. 1993. Natal dispersal and gene ﬂow in white-tailed deer in northeaster
Minnesota. Journal of Mammalogy, 74:316-322.

131

Neter, J ., M. H. Kutner, C. J. Nachtsheim, and W. Wasserrnan. 1996. Applied linear
statistical models, fourth edition. Irwin, Chicago.

Nilsson, J. -A. 1989a. Causes and consequences of natal dispersal in the marsh tit (Paras
palustris). Journal of Animal Ecology, 58:619-636.

Nilsson, J. -A. 1989b. Establishment of juvenile marsh tits in winter ﬂocks: an
experimental study. Animal Behaviour, 38:586-595.

Nilsson, J. -A. and H. G. Smith. 1988. Effects of dispersal date on winter ﬂock
establishment and social dominance in marsh tits Parus Palustris. Journal of
Animal Ecology, 57:917-923.

Noyes, R. F., G. W. Barrett, and D. H. Taylor. 1982. Social structure of feral house
mouse (Mus musculus) populations: effects of resource partitioning. Behavioral
Ecology and Sociobiology, 10:157-163.

Nunes, S. and K. E. Holekamp. 1996. Mass and fat inﬂuence the timing of natal dispersal
in Belding's ground squirrels. Journal of Mammalogy, 77:807-817.

Nunes, S., P. A. Zugger, A. L. Engh, K. O. Reinhart, and K. E. Holekamp. 1997. Why
do female Belding's ground squirrels disperse away from food resources?
Behavioral Ecology and Sociobiology, 40:199-207.

Nunes, S., C.-D. T. Ha, P. J. Garrett, E.-M. Muecke, L. Smale, and K. E. Holekamp.
Body fat and time of year interact to mediate dispersal behaviour in ground
squirrels. Animal Behaviour, in press.

O'riain, M. J ., J. U. M. Jarvis, and C. G. Faulkes. 1996. A dispersive morph in the naked
mole-rat. Nature, 380:619-621.

Packer, C. 1979. Inter-troop transfer and inbreeding avoidance in Papio anubis. Animal
Behaviour, 27: 1-36.

Packer, C., D. A. Collins, A. Sindimwo, and J. Goodall. 1995. Reproductive constraints
on aggressive competition in female baboons: social dominance carries costs as
well as beneﬁts. Nature, 373:60-63.

Pellymounter, M. A., M. J. Cullen, M. B. Baker, R. Hecht, D. Winters, T. Boone, and F.
Collins. 1995. Effects of the obese gene product on body weight regulation in
ob/ob mice. Science, 269:540-543.

Pengelley, E. T. and K. C. Fisher. 1963. The effect of temperature and photoperiod on
the yearly hibernating behavior of captive golden-mantled ground squirrels
(Citellus lateralis tescorum). Canadian Journal of Zoology, 41:1103-1120.

132

Pereira, M. E. and M. L. Weiss. 1991. Female mate choice, male migration, and the
threat of infanticide in ringtailed lemurs. Behavioral Ecology and Sociobiology,
28:141-152.

Phoenix, C. H., R. W. Goy, A. A. Gerall, and W. C. Young. 1959. Organizing action of
prenatally administered testosterone proprionate on the tissues mediating mating
behavior in the female guinea pig. Endocrinology, 65:369-3 92.

Price, K. and S. Boutin. 1993. Territorial bequeathal by red squirrel mothers. Behavioral
Ecology, 4: 144-150.

Pusey, A. E. 1980. Inbreeding avoidance in chimpanzees. Animal Behaviour,
28:543-582.

Pusey, A. E. 1987 . Sex-biased dispersal and inbreeding avoidance in mammals. Trends
in Ecology and Evolution, 2:295-299.

Pusey, A. E. and C. Packer. 1987a. Dispersal and philopatry. Pp. 250-266 in Primate
societies (B. B. Smuts, D. L. Cheney, R. M. Seyfarth, R. W. Wrangham, T. T.
Struhsaker, eds). The University of Chicago Press. Chicago.

Pusey, A. E. and C. Packer. 1987b. The evolution of sex-biased dispersal in lions.
Behaviour, 101:278-310.

Rankin, M. A. 1991. Endocrine effects on migration. American Zoologists, 32:217-230.

Rayor, L. S. 1985. Effects of habitat quality on growth, age of ﬁrst reproduction, and
dispersal in Gunnison’s prairie dogs (Cynomys gunnisoni). Canadian Journal of
Zoology, 63:2835-2840.

Ritchison, G. J. R. Belthoff, and E. J. Sparks. 1992. Dispersal restlessness: evidence for
innate dispersal by juvenile eastern screech owls? Animal Behaviour, 43:57-65.

Rood, J. P. 1987. Dispersal and intergroup transfer in the dwarf mongoose. Pp. 85-103
in Mammalian dispersal patterns: the effects of social structure on population
genetics (B. D. Chepko-Sade and Z. T. Halpin, eds). The University of Chicago
Press. Chicago.

Rood, J. P. 1990. Group size, survival, reproduction, and routes to breeding in dwarf
mongooses. Animal Behaviour, 39:566-572.

Rudd, C. D., R. V. Short, G. Shaw, and M. B. Renfree. 1996. Testosterone control of
male-type sexual behavior in the tammar wallaby (Macropus eugenii). Hormones
and Behavior, 30:446-454.

133

Rutberg, A. T. and R. R. Keiper. 1993. Proximate causes of natal dispersal in feral
ponies: some sex differences. Animal Behaviour, 46:969-975.

Sauer, J. R. and N. A. Slade. 1987. Uinta ground squirrels demography: is body mass
a better categorical variable than age? Ecology, 68:642-650.

Schneider, J. E. 1992. Metabolic sense. Hormones and Behavior, 26: 1-6.

Schneider, J. E. and G. N. Wade. 1989. Availability of metabolic ﬁrels controls estrous
cyclicity of Syrian hamsters. Science, 244: 1326-1328.

Schneider, J. E., G. N. Wade, and Y. Zhu. 1994. Caudal brain stem plays a role in
metabolic control of estrous cycles in Syrian hamsters. Brain Research,
661:70-74.

Scott, I., M. Grant, and P. R. Evans. 1991. Estimation of fat-free mass of live birds: .use
of total body electrical conductivity (TOBEC) measurements in studies of single
species in the ﬁeld. Functional Ecology, 5:314-320.

Sherman, P. W. 1976. Natural selection among some group-living organisms. Ph. D.
dissertation. University of Michigan, Ann Arbor, Michigan.

Sherman, P. W. 1977. Nepotism and the evolution of alarm calls: Science, 197: 1246-
1253.

Sherman, P. W. 19818. Reproductive competition and infanticide in Belding’s ground
squirrels and other animals. Pp. 311-331 in Natural selection and social behavior

(R. D. Alexander and D. W. Tinkle, eds). Chiron Press, New York.

Sherman, P. W. 1981b. Kinship, demography, and Belding’s ground squirrel nepotism.
Behavioral Ecology and Sociobiology, 8:251-259.

Sherman, P. W. 1982. Infanticide in ground squirrels. Animal Behaviour, 30:938-939.

Sherman, P. W. and M. L. Morton. 1979. Four months of the ground squirrel. Natural
History, 88:50-57.

Sherman, P. W. and M. L. Morton. 1984. Demography of Belding’s ground squirrels.
Ecology, 65: 1617-1628.

Shields, W. M. 1982. Philopatry, inbreeding, and the evolution of sex. University of
New York Press, Albany, New York.

134

Smale, L., S. Nunes, and K. E. Holekamp. 1997. Sexually dimorphic dispersal behavior:
patterns, causes and consequences. Pp. 181-250 in Advances in the study of
behavior, volume 26 (P. J. B. Slater, J. S. Rosenblatt, C. T. Snowden, and M.
Milinski, eds). Academic Press, San Diego.

Smale, L., K. Pelz, I. Zucker, and P. Licht. 1986. Neonatal androgenization in ground
squirrels: inﬂuences on sex diﬂ’erences in body mass and luteinizing hormone
levels. Biology of Reproduction, 34:507-511.

Snyder, R. L., W. Jenson, and C. D. Cheney. 1976. Environmental familiarity and
activity: aspects of prey selection for a ferrunginous hawk. Condor, 78: 138-139.

Sokal, R. R. and F. J. Rohlf. 1981. Biometry: the principles and practice of statistics in
biological research. Second ed. W. H. Freeman and Company, San Francisco.

Sprague, D. S. 1992. Life history and male intertroop mobility among Japanese
macaques (Macacafuscata). International Journal of Primatology, 13:437-454.

Stenseth, N. C. and W. Z. Lidicker, Jr. (eds). 1992. Animal dispersal: small mammals
as a model. Chapman and Hall, London.

Stueck, K. L. and G. W. Barrett. 1978. Effects of resource partitioning on the population
dynamics and energy utilization strategies of feral house mice (Mus musculus)
populations under experimental ﬁeld conditions. Ecology, 59:539-551.

Sullivan, T. P. and D. S. Sullivan. 1982. Population dynamics and regulation of the
Douglas squirrel (T amiasciurus douglassii) with supplemental food. Oecologia,
53:264-270.

Sullivan, T. P. , D. S. Sullivan, and C. J. Krebs. 1983. Demographic responses of a
chipmunk (Eutamias townsendii) population with supplemental food. Journal of
Animal Ecology, 52:743-755.

Svendsen, G. E. 1973. Behavioral and environmental factors in the spatial distribution
and population dynamics a yellow-bellied marmot population. Ph. D. dissertation.
University of Kansas, Lawrence, Kansas.

Thor, D. H. and W. R. Holloway, Jr. 1984. Sex and social play in juvenile rats (Rattus
norvegicus). Journal of Comparative Psychology, 98:276-284.

Thor, D. H. and W. R. Holloway, Jr. 1986. Social play soliciting by male and female
juvenile rats: effects of neonatal androgenization and sex of cagemates. Behavioral
Neuroscience, 1002275-279.

135

Toran-Allerand, C. D. 1978. Gonadal hormones and brain development: cellular aspects
of sexual differentiation. American Zoologist, 18:553-565.

Trombulak, S. C. 1989. Running speed and body mass in Belding’s ground squirrels.
Journal of Mammalogy, 70: 194-197.

Trombulak, S. C. 1991. Maternal inﬂuence on juvenile growth rates in Belding’s ground
squirrels (Spermophilus beldingi). Canadian Journal of Zoology, 69:2140-2145.

Vestal, B. M. 1991. Infanticide and cannibalism by male thirteen-lined ground squirrels.
Animal Behaviour, 41:1 103-1104.

Wade, G. N. and J. E. Schneider. 1992. Metabolic ﬁrels and reproduction in female
mammals. Neuroscience and Biobehavioral Reviews, 16:235-272.

Wade, G. N., J. E. Schneider, and H.-Y. Li. 1996. Control of fertility by metabolic cues.
American Journal of Physiology, 270:E1-E19.

Walsberg, G. E. 1988. Evaluation of a nondestructive method for determining fat stores
in small birds and mammals. Physiological Zoology, 61: 153-159.

Wang, Z. and M. A. Novak. 1994. Parental care and litter development in primiparous
and multiparous prairie voles (Microtus ochrogaster). Journal of mammalogy,
75:18-23.

Waser, P. M. 1996. Patterns and consequences of dispersal in gregarious carnivores.
Pp. 267-295 in Carnivore behavior, ecology, and evolution, volume 2. Cornell
University Press, Ithaca, New York.

Waser, P. M. and L. F. Elliott. 1991. Dispersal and genetic structure in kangaroo rats.
Evolution, 45:93 5-943.

Wauters, L., E. Matthyessen, and A. A. Dhondt. 1994. Survival and lifetime reproductive
success in dispersing and resident red squirrels. Behavioral Ecology and
Sociobiology, 34: 197-201.

Wiggett, D. R. and D. A. Boag. 1992. Natal dispersal in Columbian ground squirrels:
is body mass the proximate stimulus? Canadian Journal of Zoology, 70:649-653.

Wiggett, D. R. and D. A. Boag. 1993. The proximate causes of male-biased natal

emigration in Columbian ground squirrels. Canadian Journal of Zoology, 71:
204-218.

136

Williams, C. L., A. M. Burned, and W. H. Meeks. 1990. Organizational effects of
early gonadal secretions on sexual differentiation in spatial memory. Behavioral
Neuroscience, 104:84-97.

Wilson, M. E. 1992. Factors determining the onset of puberty. Pp. 275-313 in

Handbook of behavioral biology, volume 11: sexual differentiation (A. A. Gerall,
H. Moltz, and I. L. Ward, eds). Plenum Press, New York.

Wingﬁeld, J. C., H. Schwabl, and P. W. Mattocks, Jr. 1990. Endocrine mechanisms of
migration. Pp. 232-254 in Bird migration: physiology and ecophysiology (E.
Gwinner, ed.). Springer-Verlag, Berlin.

Wolff, J. O. 1992. Parents suppress reproduction and stimulate dispersal in opposite-sex
juvenile white-footed mice. Nature, 235:409-410.

Wolff, J. O. 1993a. What is the role of adults in mammalian juvenile dispersal? Oikos,
68:173-176.

Wolff, J. O. 1993b. Why are female small mammals territorial? Oikos, 68:364-370.
Wolff, J. O. 1994. More on juvenile dispersal in mammals. Oikos, 71:349-352.

Young, G., B. Bjomsson, P. Prunet, R. J. Lin, and H. A. Bern. 1989. Smoltiﬁcation
and seawater adaptation in coho salmon (Onchorhynchus kisutch): plasma
production, grth hormone, thyroid hormones, and cortisol. General and
Comparative Endocrinology, 74:335-345.

Zucker, I. 1988. Neuroendocrine substrates of circannual rhythms. Pp. 219-251 in
Biological rhythms and mental disorders (D. J. Kupfer, T. H. Monk, and J. D.
Barchas, eds). Guilford Press, New York.

Zucker, I. and M. Boshes. 1982. Circannual body weight rhythms of ground squirrels;
role of gonadal hormones. American Journal of Physiology, 243:R546-R551.

Zucker, I. and P. Licht. 1983a. Circannual and seasonal variation in plasma luteinizing
hormone levels of ovariectomized female ground squirrels (Spermophilus
lateralis). Biology of Reproduction, 28:178-185.

Zucker, I. and P. Licht. 1983b. Seasonal variations in plasma luteinizing hormone levels
of gonadectomized male ground squirrels (Spermophilus lateralis). Biology of
Reproduction, 29:278-285.

137

Zucker, I., T. M. Lee, and J. Dark. 1991. The suprachiasmatic nucleus and annual
rhythms of mammals. Pp. 246-259 in Suprachiasmatic nucleus: the mind’s clock
(D. C. Klein, R. Y. Moore, and S. M. Reppert, eds). Oxford University Press,
New York.

 

nrcuran smr: UNIV. LIBRARIES
lllllllllllllllllllllllllIllllllllllllllllllllll
31293016987343