Kin and Social Network Structure in Two Populations of Octodon Degus

KIN AND SOCIAL NETWORK STRUCTURE IN TWO POPULATIONS OF OCTODON DEGUS

Garrett Davis

Loren Hayes Hope Klug Associate Professor Assistant Professor Biological and Environmental Sciences Biological and Environmental Sciences (Chair) (Committee Member)

Margaret Kovach Professor Biological and Environmental Sciences (Committee Member)

KIN AND SOCIAL NETWORK STRUCTURE IN TWO POPULATIONS OF OCTODON DEGUS

Garrett Davis

A Thesis Submitted to the Faculty of the University of Tennessee at Chattanooga in Partial Fulfillment of the Requirements of the Degree of the Master of Science: Environmental Science

The University of Tennessee at Chattanooga Chattanooga, Tennessee

December 2014

ABSTRACT

A growing body of evidence showing that individuals of some species live in non-kin groups

suggests kin selection is not required in all species for sociality to evolve. Here I investigate two

populations of Octodon degus , a South American rodent which has been shown to form kin and non-kin

groups. I quantified genetic relatedness within social groups in two populations as well as social network

parameters (association, strength, and clustering coefficient) in order to determine if these aspects of

sociality were driven by kinship. I analyzed social network parameters relative to ecological conditions at

burrow systems used by individuals to determine if ecological characteristics could explain variation in sociality. In both populations, genetic relatedness among individuals within social groups was not significantly higher than randomly selected individuals from the background population, suggesting non- kin structure is common in degus. In both populations, I found significant relationships between habitat characteristics of burrow systems and social network characteristics of individuals.

iii

ACKNOWLEDGEMENTS

I would first like to thank my advisor (Dr. Loren Hayes) and my other committee members, Dr.

Margaret Kovach and Dr. Hope Klug, for their knowledge and guidance. Additionally, I would like to

thank Dr. Luis Ebensperger, Dr. Rodrigo Vasquez, Dr. Elie Poulin, and the members of the Molecular

Ecology Lab at Universidad de Chile for their assistance with my research while in Chile. Lastly, I thank

my friends and family for their support and encouragement.

TABLE OF CONTENTS

ABSTRACT………………………………………………………………………………………….………………………….iii

ACKNOWLEDGEMENTS………………………………………………………….………………………………………iv

LIST OF TABLES………………………………………………………….………………………………………………….vii

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

LIST OF TERMS……………………………………………………………………………………………………………….ix

CHAPTER

I. INTRODUCTION………………………………………………………………………………………………………1

Sociality models………………………………………………………………………………………………….…1 Emlen’s theory of the family………………………………………………………………….………………2 Ecological constraints…………………………………………………………………………………….….3 Kin selection and inclusive fitness……………………………………………………………………..3 Reproductive skew……………………………………………………………………………………….…..4 Conflicting evidence……………………………………………………………………………………………...5 Trade-offs of sociality……………………………………………………………………………………….……6 Thesis objectives……………………………………………………………………………………………….....8 Study sites……………………………………………………………………………………………………………..9 Significance………………………………………………………………………………………………………….10

II. KIN AND SOCIAL NETWORK STRUCTURE IN TWO POPULATIONS OF OCTODON DEGUS ………………………………………………………………………………………..……….12

Introduction…………………………………………………………………………………………………………12 Intraspecific variation and sociality………………………………………………………………..14 Social networks……………………………………………………………………………………………...14 Objectives and study organism…………………………………………………………………….…15 Methods……………………………………………………………………………………………………..………17 Study populations……………………………………………………………………………………….….17 Ecological sampling…………………………………………………………………………………….….17 Social group determination………………………………………………………………….………..18 Social network analysis………………………………………………………………………….……...19 Genetic analysis…………………………………………………………………………………….……….20

Statistical analysis………………………………………………………………………………….……...21 Results………………………………………………………………………………………………………………22 Microsatellite variation…………………………………………………………………………………..22 Descriptive data………………………………………………………………………………………………23 Relatedness and social structure…………………………………………………………………….23 Ecology and network structure……………………………………………………………………….25 Discussion………………………………………………………………………………………………………….28 Social structure and kinship…………………………………………………………………………….28 Social networks and habitat conditions…………………………………………………………..32 Concluding remarks………………………………………………………………………………………..34

III. CONSERVATION IMPLICATIONS…………………………………………………………………………..35

REFERENCES………………………………………………………………………………………………………………...38

APPENDIX

A. SOCIAL GROUPS…………………………………………………………………………………………………48

B. GENETIC PROCEDURES…………………………………………………………………………………….…50

C. GENETIC DATA……………………………………………………………………………………………….…..53

D. ECOLOGICAL DATA…………………………………………………………………………………………….57

E. NETWORK DATA…………………………………………………………………………………………………61

F. STATISTICAL ANALYSIS…………………..…………………………………………………………………..66

VITA……………………………………………………………………………………………………………………………..69

LIST OF TABLES

2.1. Group size and within-group relatedness of social groups at Rinconada and Los Molles 2007-2008………………………………………………………………………………………………………..24

2.2 Multiple regression statistics for individuals’ network parameters vs. weighted habitat characteristics at Rinconada and Los Molles 2007-2008……………………………………………….28

1 2 3 4

LIST OF FIGURES

1.1 Conceptual framework for the drivers and outcomes of kin structure…………………………..………2

1.2 Geographic location of Los Molles (top) and Rinconada (bottom) in central Chile…………….…10

2.1 Scatterplot showing the statistically significant relationship between (a) soil hardness and strength and (b) food biomass and clustering coefficient for individuals at Rinconada in 2007-2008………………………………....………………26

2.2 Scatterplot showing the statistically significant relationship between (a) burrow density and strength and (b) food biomass and clustering coefficient for individuals at Los Molles in 2007-2008…………………………………………………...…………………..27

2.3 Social network structure of (a) a theoretical social group exhibiting kin structure and (b) an actual social group from the Los Molles 2007 population …………...……...... 29

3.1 Example of the Allee effect, demonstrating a reduced population growth rate at low density (from Kuussaari et al. 1998) …..………………………………………………………….…………….37

viii

LIST OF TERMS

Association – social network parameter which measures the amount of spatial and temporal overlap between two individuals, calculated from trapping and telemetry overlap

Coefficient of relatedness (r) – the proportion of genetic material (e.g. alleles) shared between two individuals

Clustering coefficient – social network parameter which measures the level of association between an individual’s associates, i.e. how closely one’s associates interact with each other

Direct fitness – an individual’s contribution to the gene pool of a subsequent generation based only on the reproductive success of the individual itself

Fitness – an individual’s contribution to the gene pool of a subsequent generation

Group-size effects – the costs and benefits experienced as a result of living in close proximity with conspecifics

Inclusive fitness – the sum of an individual’s direct and indirect fitness

Indirect fitness – benefit to the reproductive success of an individual’s relative due to help provided by the individual

Kin selection – concept that natural selection may favor sociality when an individual maximizes its inclusive fitness by improving the reproductive success of its relatives

Natal philopatry – offspring remain at the birthplace beyond sexual maturity

Plural breeding – breeding system in which multiple individuals breed

Singular breeding – breeding system in which one or few individuals dominate breeding

Social network analysis – the mathematical measuring of relationships between individuals within a given area during a given time

1 ix 2 3

LIST OF TERMS

Sociality – two or more conspecifics living in close proximity

Strength – social network parameter which measures the sum of an individual’s social interactions, based on the total number and intensity of interactions

CHAPTER I

INTRODUCTION

Sociality models

Animal sociality involves the cooperative interactions between conspecifics that result in social relationships of varying degree and duration within populations of social species. Animal social systems derive from three main components: (i) composition of social groups, (ii) social structure (interactions between individuals), and (iii) mating system (Schradin, 2013). A major aim of behavioral ecology is to determine the factors that cause inter- and intraspecific variation in social systems and the reproductive consequences of this variation.

Each of these components, and in turn the social system as a whole, may be influenced by many factors both intrinsic and extrinsic to individuals within the population. Ecological conditions, life history traits, and the evolutionary history of the species may all influence individuals’ behavior and the type and extent of sociality seen in a population (Figure 1.1). For example, whether offspring decide to remain in their natal territory or disperse to a new territory may be dependent on the quality of resources (e.g. food, mates) available in each area (Emlen, 1982). Certain life history traits, such as brain size, may influence the capacity of individuals to maintain complex social relationships, dictating the extent to which sociality is possible (Dunbar and Shultz, 2007). Moreover, other life history traits, such as longevity, contribute to aspects of sociality (Arnold and Owens, 1998). For example, in long-lived species, opportunities to establish a breeding territory may be reduced by low adult turnover rates, which may influence rates of philopatry and dispersal. Finally, a species propensity for sociality may also be driven

by the characteristics of its evolutionary ancestors (Shultz et al., 2011). A comprehensive model to

explain how these factors may converge and result in animal sociality is Emlen’s (1995) ‘integrated

theory of family social dynamics.’ A major goal of my study is to test components of Emlen’s model,

specifically the ecological and social factors underlying kin structure. Herein, I describe the major

framework underlying the model and how each component contributes to the evolution of sociality.

Ecological conditions Life history Evolutionary history

Behavioral mechanisms of group formation

Kin structure

Individual differences in inclusive fitness

Figure 1.1 Conceptual framework for the drivers and outcomes of kin structure

Emlen’s theory of the family

For decades, the study of animal sociality has largely worked under the paradigm outlined in

Emlen’s (1995) model, which posits that social groups form when offspring remain philopatric to the

natal nest (natal philopatry), resulting in the formation of extended family groups. Emlen’s model integrates three major theories: ecological constraints, inclusive fitness theory, and reproductive skew.

Ecological constraints

The first component of Emlen’s model is ‘ecological constraints theory,’ which suggests that natal philopatry originally arises in response to ecological conditions that constrain offspring dispersal from the natal nest (Emlen, 1982). Under these conditions, it is more beneficial for offspring to remain in the natal nest rather than disperse and attempt to breed in a new territory . For example, if an individual is born into a natal group that maintains a high quality territory with good access to food resources and protection from predation, and the surrounding territories are of lesser quality, the individual stands to gain the greatest fitness by remaining philopatric, leading to the formation of social groups composed of close relatives (Emlen, 1982). Thus, when conditions do not favor offspring dispersal and independent breeding, we expect more and larger social groups to form. Evidence supporting ecological constraints theory comes from research on a variety of taxa, including insects

(Field et al., 1998), fish (Bergmuller et al., 2005), birds (Komdeur et al., 1995; Arnold and Owens, 1999), and mammals (Chapman et al., 1995; Faulkes et al., 1997; Randall et al., 2005; Lucia et al., 2008).

Kin selection and inclusive fitnes s

Emlen’s model also incorporates kin selection theory to explain that philopatric individuals may further increase fitness indirectly by providing care to offspring produced by closely-related kin

(Hamilton, 1964; Maynard-Smith, 1964). Hamilton’s rule states that care should be provided by an individual when the benefit of the care to the recipient (B), multiplied by the relatedness of the individual to the recipient (coefficient of relatedness, r), is greater than the cost (C) to the individual for providing the care (rB-C>0) (Hamilton, 1964). In other words, an individual may gain indirect fitness by

enhancing a close relative’s direct fitness through providing care to the relative’s offspring, who share

some of the same genetic material as the individual providing care. Thus, selection is predicted to favor

strategies to maximize inclusive fitness, the sum of indirect and direct fitness (Hamilton, 1964).

Although controversial (West et al., 2002; Wilson, 2005; Nowak et al., 2010), kin selection and inclusive fitness are foundations of animal social theory (Emlen, 1995; Abbot et al., 2011). Decades of research have supported the role of kin selection as a driver of sociality in numerous species, including invertebrates (Trivers and Hare, 1976), birds (Brown, 1987), and mammals (Solomon, 2003; Kappeler,

2008). Based on this research, natal philopatry and kin structure are generally accepted as defining characteristics of social groups in most social species (Lacey and Sherman, 2007).

Reproductive skew

The final component of Emlen’s model, reproductive skew theory, builds upon aspects of both

ecological constraints and kin selection. Reproductive skew describes how much direct reproduction is

shared among individuals in group. High skew is when one or a few individuals breed, whereas low

skew is when reproduction is shared more evenly among several group members. The central idea

behind reproductive skew theory is that, following the formation of social groups due to ecological

constraints and natal philopatry, dominant breeders in the group may share reproduction with

subordinates under certain conditions. In particular, dominant individuals may allow subordinates to

breed if it induces subordinates to remain in the group, and if the presence of subordinates increases

the dominant’s inclusive fitness above what it would be if they were to disperse (Emlen, 1995). More

specific aspects of the theory, such as decreasing ecological constraints leading to increasing shared

reproduction, have been supported in multiple species (Curry, 1988; Emlen and Wrege, 1991).

Regardless, group stability may vary considerably in relation to competition for resources or mates.

Such competition will influence the amount of reproductive skew and the potential costs and benefits to group members, which in turn influences group stability and whether groups disband or remain together (Emlen, 1995).

Conflicting evidence

Emlen’s model is not universally accepted, and a growing body of evidence suggests that natal

philopatry is not the only mechanism underlying the formation of social groups. Some invertebrates

(Queller et al., 2000; Seppa et al., 2008), fish (Avise and Shapiro, 1986), birds (Griesser et al., 2008), and

mammals (Faulkes and Bennett, 2001; Kappeler, 2008; Ebensperger et al., 2009) form groups when

adults move into existing social groups or establish new groups with unrelated conspecifics. For

example, adult nutria (Myocastor coypus) migrate between social groups (Guichon et al., 2003), with

subsequent research showing groups that lack kin structure (Tunez et al., 2009). In such cases,

individuals cannot benefit from indirect fitness gains and thus, their inclusive fitness derives solely

from direct fitness.

Further, recent research has also called into question the extent to which inclusive fitness

adequately explains group-living even amongst related individuals (Nowak et al., 2010). Although the

benefits of helping kin are intrinsically obvious, empirically determining the extent and result of such help

is difficult (West and Griffin, 2002). For example, helping affects fitness in numerous ways, including

indirectly (Hamilton, 1964) or from group augmentation (see group size effects below), which in turn may

influence survival and reproductive success. Helping may also influence the future direct fitness of group

members if it contributes to improved breeding position within the group. Further, it is likely that indirect

benefits of helping kin have largely been overestimated. For example, estimates of inclusive fitness have

often included the effects of helping on direct descendants as well as those of kin, resulting in artificially

high estimates of the indirect benefits (Clutton-Brock, 2002). Conversely, the direct fitness benefits of group living have likely been underestimated, with some research demonstrating that helpers may increase their direct fitness overall by improving their likelihood for survival and successful reproduction

or increasing their likelihood of successful dispersal in the future (Grinnell et al., 1995; Komdeur, 1996;

Clutton-Brock, 2002). The inherent trade-offs that arise from group living have rarely been compared to the indirect benefits gained by group members, a comparison that is necessary to fully understand the costs and benefits of such a system and ultimately explain its evolutionary significance (West et al., 2001) .

Trade-offs of sociality

Regardless of kinship, sociality may result in increased direct fitness of individuals through a

number of mechanisms. These fitness benefits are largely tied to group-size effects i.e., costs and

benefits derived simply from living in close proximity to other individuals (Krause and Ruxton, 2002). For

example, individuals may benefit from a reduced need for vigilance and reduced predation risk simply

due to the collective vigilance of the group as a whole (‘many eyes hypothesis’). This improved vigilance

as a product of group living has been demonstrated in both vertebrate and invertebrate species

(Kenward, 1978; Treherne and Foster, 1980). Further, reduced predation risk may stem simply from

mathematically reducing an individual’s likelihood of being depredated relative to other individuals

around it, known as the dilution effect (Williams, 1966). For example, a solitary individual that

encounters a predator has a certain likelihood of being depredated (N), whereas in a group of several

others that individual’s likelihood of being depredated is reduced to 1/N simply due to probability. For

example, Calvert et al. (1979) demonstrated that, in winter aggregations of monarch butterflies,

predation rates per individual decreased with increasing colony size, even though larger colonies

attracted more predators.

Sociality may also benefit individuals by providing better access to resources through several

mechanisms. If individuals in a group share information amongst themselves, the group as a whole may

benefit from an increased ability to find high quality resources, such as food. In such cases, groups serve

as a sort of information center and group members benefit from the information gathered by other

group members (Ward and Zahavi, 1973). Further, in predatory species, sociality may result in improved

rates of prey capture. For example, in groups of black-headed gulls ( Chroicocephalus ridibundus ), an individual’s hunting success on prey fish increases with the number of predators in the group (Gotmark et al., 1986). Similar effects can be seen in pack-hunting animals, such as spotted hyena ( Crocuta

crocuta ) groups assembled to hunt larger prey (Kruuk, 1972).

At the same time, sociality may be costly to individuals under some environmental conditions.

Most obviously, individuals surrounded by other conspecifics are still in competition with each other for

resources, including food and mates. However, there are also less obvious potential detriments to living

socially. Being in close proximity to several individuals may increase the likelihood of disease

transmission and parasitism within a group (Côté and Poulin, 1995). For example, Brown and Brown

(1986) showed that the number of parasites on nestling cliff swallows ( Hirundo pyrrhonota ) increased

with increased colony size and that subsequently nestling body mass decreased as a result of increased

parasitism. Further, the behavioral traits necessary to maintain sociality may in fact result in a trade-off

to an individual’s success in other aspects of life. For example, Magurran and Seghers (1991)

demonstrated that guppies ( Poecilia reticulata ) prone to shoaling were also less aggressive when in

competition for food. Additionally, group living may result in an increased risk of cannibalism (Thiel

2011) and infanticide (Van Schaik and Kappeler, 1997). Lastly, living with close relatives incurs the

inherent risk of inbreeding, though this may be reduced or avoided through sex-biased dispersal and/or

reproductive suppression (Wolff, 1992).

While many of these potential trade-offs to sociality may seem logical, from an evolutionary

perspective they become particularly interesting in cases in which sociality results in reduced direct

fitness of group members and indirect fitness is unlikely. In cases where social groups are not composed of close relatives (i.e. non-kin groups), individuals will not benefit from indirect fitness gains, and thus

may actually experience net fitness costs under conditions in which sociality reduces their direct fitness.

While not common, the presence of non-kin groups has been demonstrated in several taxa (Davies and

Lundberg, 1984; Túnez et al., 2009). In particular, analysis across bird taxa has demonstrated that

approximately 15% of cooperatively-breeding bird species primarily nest with non-related individuals

(Riehl, 2013). This seems to present an evolutionary paradox, as an individual’s inclusive fitness may be

reduced by group-living and selection should not favor the trait. Thus, it is crucial to fully investigate such situations in order to understand the interactions between a given social system and the costs and benefits that arise from it in order to ultimately understand how they may evolve.

Thesis objectives

The goal of this research is to describe the social structure in two populations of degus

(Octodon degus ), a social rodent endemic to central Chile. Degus are plural breeders that live socially in underground burrows, with female group members providing care to both related and unrelated offspring. Previous research on one degu population indicates that social groups consist of unrelated individuals (Quirici et al., 2011) and that some individuals in large groups experience direct fitness costs (Hayes et al., 2009). This is particularly interesting given that degus are short-lived (lifespan generally less than 1 year in the wild) and typically only reproduce once during a lifetime; individuals who lose opportunities to maximize fitness are unlikely to benefit during a future reproductive cycle.

However, degus live in ecologically distinct populations throughout their geographic range, with little

known about the genetic structure of social associations in other populations experiencing different

ecological conditions. Further, no research has been conducted investigating the relationship between

within-site and between-site ecological conditions and sociality at the scale of individual associations

(as opposed to social groups as a whole). Thus, the specific aims of this study were two-fold: 1) determine the genetic composition of social groups in a second population in a different habitat, and 2) use social network analysis (Sih et al., 2009) to investigate individual and group-level relationships between genetic relatedness, ecological conditions, and social structure in two populations. In doing so, I can test the working hypotheses that kin structure and local habitat conditions influence social interactions among individuals, and therefore the social network structure of the population.

Study sites

This study was conducted on two degu populations in geographically distinct sites (400 km

apart) in north-central Chile, Rinconada de Maipú (33°23′S, 70°31′W, altitude 495 m) and Bocatoma

Los Molles (30°45′S, 70°15′W, altitude 2,600 m; see Figure 1.2). The sites are characterized by differences in several ecological traits, with Rinconada having harder soil, greater food abundance, greater distance from burrows to overhead cover, and lower density of burrow openings than Los

Molles. Previous research has also demonstrated differences in sociality between the two populations, with slightly larger social groups at Rinconada than Los Molles (Ebensperger et al., 2012a). Prior to this study, there had been no genetic analyses to determine kin structure within the Los Molles population.

Further, there had been no examination of the relationships between local ecological conditions, kin structure, and social structure at the individual level for either population.

Figure 1.2 Geographic location of Los Molles (top) and Rinconada (bottom) in central Chile

Significance

Although rare, the existence of non-kin groups challenges the established paradigm of kin selection as the driving force of sociality in these species. While there are potential direct fitness benefits to sociality regardless of kinship (e.g., decreased predation risk, increased for aging efficiency), indirect benefits are minimal or non-existent in groups of distantly related or unrelated individuals.

Moreover, the relationship between sociality and fitness varies considerably across species , and group 10

size indices are often unlikely to capture all variation within a social system (Krause and Ruxton, 2002;

Ebensperger et al., 2012b). For example, there may be direct fitness costs to individuals living in large groups (Rasa, 1989; Lacey, 2004). By investigating social relationships on the scale of individuals, a

major goal of this study, it may be possible to deduce links between local ecological conditions, genetic

relatedness, and social structure that were not apparent when using other indices of sociality (e.g.

group size). This may reveal relationships that were previously unseen, and thus lead to a more

comprehensive understanding of the evolution of sociality.

CHAPTER II

KINSHIP AND SOCIAL NETWORK STRUCTURE IN TWO POPULATIONS OF OCTODON DEGUS

Introduction

Social structure summarizes the nature and extent to which animals interact with others within a population (Whitehead, 2008; Schradin, 2013). In social species, the cooperative interactions among individuals in a population are the background upon which foraging, mating, and reproductive interactions take place (Wolf et al., 2007). Thus, determining the factors that influence these interactions is crucial to developing a comprehensive understanding of the evolution of sociality. One well-established model explaining potential conditions leading to and favoring sociality is Emlen’s (1995)

“integrated theory of family social dynamics”. This model incorporates two important theories - ecological constraints (Emlen, 1982) and kin selection (Hamilton, 1964) - to explain the evolution of animal sociality. The model posits that extended family groups (kin groups) form when juveniles remain philopatric to the natal group under conditions that limit direct reproduction (ecological constraints;

Emlen, 1982). Under these conditions, kin selection theory predicts that breeders benefit when philopatric individuals assist with offspring care (alloparental care) and philopatric individuals benefit indirectly by providing care to non-descendent offspring produced by closely related kin. Thus, parental care directed towards closely related kin is predicted to increase an individual’s inclusive fitness

(Hamilton, 1964; Maynard-Smith, 1964).

The two main thrusts of Emlen’s model – ecological constraints and kin selection – have been the subject of considerable theoretical and empirical work for decades. Thus, the impacts of ecological

constraints on animal sociality have been demonstrated in invertebrates (Emlen, 1982; Rehan and

Schwarz, 2011) and vertebrates (Komdeur, 1992; Travis et al., 1995; Lucia et al., 2008; Schoepf and

Schradin, 2012). Consequently, ecological constraints are often viewed as a primary driver for social group formation. Regarding the influence of kin selection, decades of research have validated that groups typically consist of extended families as has been observed in invertebrates (Trivers and Hare,

1976), birds (Brown, 1987; Stacey and Ligon, 1991), and mammals (Solomon, 2003; Kappeler, 2008).

Taken together, these observations suggest that natal philopatry and inclusive fitness benefits are the defining characteristics of groups in most social species (Emlen, 1995; Lacey and Sherman, 2007; but see: Griffin and West, 2002; Nowak et al., 2010).

However, a growing body of evidence suggests that Emlen’s model is not universal for social animals and that natal philopatry is not the only mechanism underlying the formation of social groups.

In some invertebrates (Queller et al., 2000; Seppa et al., 2008), fish (Avise and Shapiro, 1986), birds

(Griesser et al., 2008), and mammals (Faulkes et al., 2001; Guichon et al., 2003; Ebensperger et al.,

2009), adults move into existing social groups or establish new groups with unrelated conspecifics

(Ebensperger and Hayes, 2008). Consequently, some of these species may live in non-kin groups.

Mammals with low mean levels of relatedness within social groups include rodents (Tunez et al., 2009;

Quirici et al., 2011), white-tailed deer ( Odocoileus virginianus), and river otters ( Lontra Canadensis)

(Smith, 2014).

In non-kin groups, the potential costs of group living (e.g. competition for resources) and cooperation (e.g. cheating) are not outweighed by indirect fitness benefits of cooperating with kin.

However, cooperation may exist irrespective of the genetic relatedness of individuals (West and Griffin,

2002). For example, it has been shown that the amount of parental care provided by subordinates in meerkat ( Suricata suricatta ) social groups is not driven by the relatedness of the subordinate to the

offspring that is receiving care (Clutton-Brock et al., 2001). Thus, it may be necessary to investigate

other avenues of fitness gains outside of kinship in cases where indirect fitness is unlikely to explain the

evolution of sociality.

Intraspecific variation and sociality

Historically, research on animal social systems has focused on single populations, assuming that

the social system is fixed by natural selection and consistent across species. However, intraspecific

variation in social systems has been observed in reptiles (Shine and Fitzgerald, 1995), birds (Komdeur,

1992) and mammals (Travis and Slobodchikoff, 1993; Brashares and Arcese, 2002; Ophir et al., 2007).

Such variation is expected if the associated costs and benefits of sociality depend on local ecological conditions and result in differential selection on the behaviors influencing group formation (Emlen and

Oring, 1977; Lott, 1991). At the proximate level, intraspecific variation in social structure may arise due to genetic variation and/or varying levels of phenotypic plasticity between populations (Schradin, 2013).

Since non-kin groups provide a challenge to the fundamental model of group formation, it is critical to investigate the site-specific social systems and environmental conditions characteristics of multiple populations in species that exhibit non-kin social structure. This approach will allow researchers to determine if non-kin groups are common or the rare product of local conditions acting upon individual populations.

Social networks

The most common metric of sociality – group size – provides only one dimension of an animal’s social system. A challenge to developing a comprehensive model for sociality is to determine social relationships at varying scales and dimensions that collectively make up the social structure of a

population. A quantifiable method of analysis for such questions is social network analysis (Whitehead,

2008; Wey et al., 2008; Sih et al., 2009). Social networks model the ways in which individuals interact with other individuals in the population, allowing researchers to quantify the strength and extent of relationships in ways that traditional methods of social determination cannot (Sih et al., 2009). By analyzing the network dynamics at both the individual and social group level (calculated as means from the values of each group member), it is possible to answer questions about an individual’s social connections as well as questions about how the individual associations interact at varying levels to form the social structure of the population as a whole (Wey et al., 2008). For example, social network analysis has provided insights into complex patterns of sociality, including quantifying distinct structural layers within a population’s social system (Wolf & Trillmich, 2008) and determining how social associations predict patterns of cooperation (Croft et al., 2006). Based on theory, we expect stronger social interactions among kin than non-kin, a prediction that can be tested by comparing within-group relatedness (calculated as a mean from the pairwise relatedness values of group members) with group- level social network parameters (association, strength) which quantify the proportion of time individuals spend in close proximity with each other. Further, ecological characteristics may influence social network parameters by dictating how individuals move through their environment, both spatially and temporally, ultimately affecting the extent to which they interact with other individuals in a given area.

Thus, habitat use may influence a given social network parameter (Croft et al., 2004).

Objectives and study organism

The degu ( Octodon degus ) is a group-living caviomorph rodent endemic to central Chile (Hayes

et al., 2011). Degus are widespread, occurring in ecologically distinct habitats throughout their

geographic range (Meserve et al., 1984; Ebensperger et al., 2012a), making them a good model

organism for examining how local conditions influence the formation and composition of groups and

social associations at the population level. In one population (Rinconada de Maipú, Chile; 33°23′S,

70°31′W), the immigration and emigration of adults into and out of groups is a more important driver of

group formation than natal philopatry by offspring (Ebensperger et al., 2009). Consequently, genetic

relatedness (R) of individuals within groups is similar to that of the background population, indicating an

absence of kin structure (Quirici et al., 2011). A recent study comparing social groups in Rinconada and a

second population (Bocatoma Los Molles) revealed that groups differ in size between these populations

(Ebensperger et al., 2012a). This observation suggests some degree of intraspecific variation in degu

social organization. To date, no one has investigated if these differences in social organization and local

ecological conditions are linked to differences in kin and social network structure. The objectives of this study were to determine if kin structure differed between two degu populations and to use social network analysis to investigate possible links between ecological conditions, kinship, and social associations at both individual and group-level scales within each population.

While non-kin groups are prevalent at Rinconada, the kin structure of social groups at Los Molles was previously unknown. Ebensperger et al. (2012a) also showed that the sites differ in predation risk and distribution of food resources. Based on Emlen’s model, we then expect differences in sociality between the two populations. In particular, since resources are more patchily distributed at Los Molles than Rinconada, I expect greater natal philopatry at Los Molles and predict that a greater percentage of groups are kin-based in this population. Additionally, if kinship is driving social interactions among group

members, I predict a positive relationship between within-group relatedness and group-level strength

and association in both populations. Lastly, individuals inhabiting high quality habitats may experience

stronger and/or more social interactions. If social network structure at the individual level is driven by

within-site ecological conditions (e.g. food availability near burrows), I predict a positive relationship

between the habitat quality of burrows used by an individual and that individual’s social network parameters (strength and clustering coefficient).

Methods

Study populations

I determined the social network and kin structure of social groups from two degu populations in central-north Chile, Estación Experimental Rinconada de Maipú (33˚23′S, 70˚31′W, altitude 495 m)

(hereafter Rinconada) and Bocatoma Los Molles (30˚45′S, 70˚15′W, altitude 2,600 m) (hereafter Los

Molles). The environmental conditions at these sites differ in several ways, with Rinconada having harder soil, greater food abundance, greater distance from burrows to overhead cover, and lower density of burrow openings than Los Molles (Ebensperger et al., 2012a). Predator sightings are more frequent at Rinconada than Los Molles (Ebensperger et al., 2012a). The fieldwork was conducted in 2007 and 2008, during the time when females were in late pregnancy or lactating (i.e. September–October at

Rinconada and November–December at Los Molles).

Ecological sampling

I used data on ecological conditions at burrow systems (food availability, burrow density, and soil hardness; Ebensperger et al., 2012a) to determine if local ecological variation predicted social structure within each site. To quantify food availability, a 250 x 250-mm quadrat was placed at 3 meters and 9 meters from the center of each burrow system in one of the cardinal directions

(randomly selected for each distance at each burrow system), and all above-ground green herbs were removed, dried, and weighed for biomass. Burrow density (openings per square meter) was quantified by counting the number of burrow openings within a 9-m radius from the center of each

burrow system. Soil hardness was sampled similarly to food availability, with a soil penetrability

measurement taken at 3 and 9-m from the center of each burrow system in one of the cardinal

directions. Distance to overhead cover was measured from the center of each burrow system using

a 100-m measuring tape (Ebensperger et al., 2012a).

Social group determination

Degus are diurnal and remain in underground burrows with conspecifics overnight (Ebensperger

et al., 2004). Thus, the main criterion used to assign degus to social groups was the sharing of burrow systems during the nighttime (Ebensperger et al., 2004). To determine social group membership, I used a combination of night-time telemetry and early morning burrow trapping. During burrow trapping, a burrow system was defined as a group of burrows surrounding a central location where individuals were repeatedly found during telemetry (Hayes et al., 2007). Burrow systems were trapped an average of

31.4′±′1.2 (mean ± standard error) days in 2007 and 45.3′±′1.6 days in 2008 at Rinconada, and for 30 days in 2007 and 21 days in 2008 at Los Molles. Tomahawk live-traps (model 201, Tomahawk Live Trap

Company, Tomahawk, Wisconsin, USA) were set prior to the emergence of adults during the early morning hours (0700-0730) and were checked and closed after 1.5 hours. The identity, sex, body mass, and reproductive condition of all individuals were determined at first capture. Additionally, a small tissue sample was taken from each individual’s ear the first time it was captured and stored in 99%

ethanol at 0 °C. Adults weighing more than 170 g were fitted with 8 g (BR radio-collars, AVM Instrument

Co., Colfax, California, USA) or 7–9 g radio-transmitters (RI-2D, Holohil Systems Limited, Carp, Ontario,

Canada) with unique frequencies.

During night-time telemetry, females were radio-tracked to burrow systems. Previous studies at

Rinconada have demonstrated that telemetry locations represent sites where degus remain

underground throughout the night (Ebensperger et al., 2004). Locations were determined once per night approximately 1 hour after sunset using an LA 12-Q receiver (for radio collars tuned to 150.000–

151.999 MHz frequency; AVM Instrument Co., USA) and a hand held, 3-element Yagi antenna (AVM instrument Co.).

To determine social group membership, I created a similarity matrix of pairwise associations of the burrow locations of all adult degus during trapping and telemetry (Whitehead, 2009). Associations were determined using the “simple ratio” association index (Ginsberg and Young, 1992), i.e. the number of times that two individuals are captured or tracked via telemetry at the same burrow system on the same day divided by the total number of times each is captured/tracked on the same day regardless of burrow system. Only associations with a value greater than 0.1 (i.e. 10% overlap of trapping/telemetry locations) were included. Since social network parameters are calculated based on trapping data, only degus that were trapped on at least five days were included in analyses to exclude poorly sampled individuals (Wey et al., 2013). Thus, some individuals and social groups previously used in Ebensperger et al. 2012a were excluded from our study to avoid biasing the network data. For example, Ebensperger et al. 2012a reported four social groups at Los Molles in 2007; however, one of these groups consisted of a solitary individual, and several individuals within two other groups had poor trapping data. Thus, I use only one social group from this year in my study. Social groups were determined using a hierarchical cluster analysis in SOCPROG 2.0 software (Whitehead, 2009).

Social network analysis

Social network analysis was used to look for patterns of sociality at the individual level, including pairwise relationships between group members and non-group members. For each individual, I calculated the strength – the sum of associations (Whitehead, 2008) - calculated from the pairwise

association networks. High strength indicates a high total amount of spatial and temporal overlap with other individuals, resulting from strong associations, many associations, or a combination of both. For

each individual, I also calculated the clustering coefficient, a measure of how connected an individual’s

associates are to each other (e.g. an individual with a high clustering coefficient has close associations

with individuals who also associate closely with each other, forming a “cluster”). For each social group, I

calculated the mean association (from each pair of group members, based on the “simple ratio” explained above) and the mean strength, based on the individual values for each group member.

Network parameters were calculated from pairwise similarity matrices in SOCPROG 2.0.

Genetic analysis

Genetic analyses to determine relatedness (R) were conducted in the Molecular Ecology lab at the Universidad de Chile in Santiago, Chile. Analyses were conducted on tissue samples collected from n=14 and n=26 individuals at Los Molles and n=21 and n=29 individuals at Rinconada in 2007 and 2008, respectively. Genomic DNA was extracted from tissue using a standard salt extraction protocol.

Amplification of DNA was achieved using polymerase chain reaction of 100 ng of DNA from each

individual using the conditions recommended by Quan et al (2009). Amplification was confirmed with

agarose gel electrophoresis. Individuals were genotyped using 5 degu microsatellite loci (OCDE3, OCDE6,

OCDE11, OCDE12, OCDE13; Quan et al. 2009). These loci were used because they were polymorphic and

showed no linkage disequilibrium during previous studies (Quan et al., 2009; Quirici et al., 2011). Allele

quantification and testing for linkage disequilibrium were performed in GENEPOP 4.2 (Raymond and

Russet, 1995). Microsatellite sequencing was performed by Macrogen, Inc. (Seoul, South Korea). Allele

sizes were determined and genotypes assigned using PeakScanner 2.4 software. Deviations from Hardy-

Weinberg equilibrium and the pairwise coefficient of relatedness ( R) among individuals were calculated using the ML-Relate software (Kalinowski et al., 2006).

Statistical analysis

To determine if social groups consisted of closely-related kin, mean pairwise relatedness across

group members was compared to the relatedness of the background population consisting of all

individuals for which there was genetic data. To determine the relatedness of the background

population, bootstrapping analysis (n=1000 permutations, with replacement) was performed on

randomly selected pairs of individuals irrespective of social group, with sample sizes dependent on the

number of individuals in each social group (e.g. 3 randomly selected pairs for group size = 3) using R

version 3.1.1 statistical software. Groups with mean pairwise relatedness that fell outside of the 95%

confidence interval for the randomly-selected background population were considered statistically

different from the background population.

To examine how population, group size, and genetic relatedness influence group-level social

network parameters (mean association and mean strength), I first used Akaike Information Criterion

(Akaike, 1974) to determine the best fit model for both mean association and mean strength. Each

possible combination of factors and interactions was tested and the model with the lowest AIC value

was selected. The best fit model for each network parameter was then used to test for significant main

effects and interactions. For mean strength, the best fit model included population, group size, and

relatedness as factors (R 2 = 0.90, AIC = 2.59). Thus, an ANCOVA was performed with each of those three

fixed factors as well as the population*relatedness interaction. For mean association, the best fit model

included only population as a fixed factor (R 2 = 0.85, AIC = -61.60), and thus group size and relatedness

were removed as factors influencing mean association. With only one fixed factor, an independent

samples t-test was run to test for differences in mean association between the two populations.

Additionally, simple linear regressions were run on each population to test for relationships between group size and relatedness. All group-level analyses were conducted using SAS 9.3 (SAS Institute, Inc.,

Cary, NC).

To evaluate the relationship between an individual’s social network parameters (strength and clustering coefficient) and the ecology (food biomass, burrow density, soil hardness) of its burrow systems within both populations, I conducted multiple regressions with all weighted (based on proportion of captures at burrow systems) ecological characteristics as independent variables and each network parameter as the dependent variable. To test the assumption that the model was linear, I visually inspected a plot of standardized residuals. To test for autocorrelation, the Durbin–Watson statistic (d) had to range between 1.5 and 2.5. To test for homoscedasticity, I visually inspected the data point spread showing the regression standardized residual vs. the regression standardized predicted value. Variables were considered collinear if the variability inﬂation factor (VIF) was greater than 4.0.

Multiple regressions were conducted with SPSS Statistics 22.2 (IBM, Inc., Chicago, IL). For all analyses, I set the alpha level to P=0.05. Throughout, I report means with standard errors (SE).

Results

Microsatellite variation

There was no evidence of linkage disequilibrium across all 5 loci screened (P > 0.05 for each loci). The number of alleles per locus ranged from 5-12. The observed heterozygosity of loci ranged from

0.36-0.79 for Rinconada and 0.48-0.91 for Los Molles. In both populations, deviations from Hardy-

Weinberg equilibrium were detected for two loci (OCDE6 and OCDE12, P < 0.01). Therefore, estimations of pairwise relatedness were adjusted to account for potential null alleles using the ML-Relate software.

Descriptive data

The pairwise relatedness between individuals ranged from 0.00-0.52 and 0.00-0.58 for

Rinconada and Los Molles, respectively. The mean group-level relatedness ranged from 0.07 to 0.21 at

Rinconada and from 0.09 to 0.25 at Los Molles.

Individual network strength ranged from 0.07-4.13 (mean±SE=1.92±0.81) and 1.01-8.00

(4.55±1.44) for Rinconada and Los Molles, respectively. The clustering coefficient for individuals ranged

from 0.01-0.94 (0.34±0.17) for Rinconanda and 0.41-1.00 (0.81±0.37) for Los Molles. At the group level,

mean strength ranged from 0.09-3.48 (2.23±0.67) at Rinconada and 1.00-6.79 (3.36±1.40) at Los Molles.

Mean association ranged from 0.17-0.96 (0.44±0.51) at Rinconada and 0.45-1.00 (0.88±0.46) at Los

Molles.

Relatedness and social structure

Mean group-level relatedness (R) ranged from 0.07-0.25 across all social groups examined.

Bootstrapping analysis indicated that social group members were not significantly more related to each

other compared to randomly selected individuals from the background population, with mean pairwise

relatedness of all groups falling within the 95% confidence intervals of the background population (Table

2.1). Additionally, there was not a statistically significant relationship between group size and

relatedness at either Los Molles (β = -0.79, R2 = 0.63, P = 0.06) or Rinconada (β = -0.27, R2 = 0.07, P =

0.30).

Table 2.1 Group size and within-group relatedness of social groups at Rinconada and Los Molles 2007- 2008

(i) Rinconada Year Social Group Group Size Genetic Relatedness 95% CI 2007 1 3 0.19 0-0.31 2007 2 7 0.08 0-0.21 2007 3 6 0.15 0-0.22 2007 4 4 0.17 0-0.25 2007 5 3 0.14 0-0.31 2007 6 3 0.11 0-0.31 2007 7 6 0.09 0-0.22 2008 8 5 0.11 0-0.24 2008 9 2 0.07 0-0.41 2008 10 2 0.21 0-0.41 2008 11 3 0.16 0-0.31 2008 12 3 0.09 0-0.31 2008 13 3 0.11 0-0.31 2008 14 2 0.12 0-0.41 2008 15 2 0.14 0-0.41 2008 16 3 0.09 0-0.31 2008 17 4 0.07 0-0.25 (ii) Los Molles Year Social Group Group Size Genetic Relatedness 95% CI 2007 1 4 0.09 0-0.23 2008 2 6 0.09 0-0.23 2008 3 2 0.21 0-0.46 2008 4 3 0.25 0-0.33 2008 5 6 0.09 0-0.23 2008 6 7 0.10 0-0.23

The ANCOVA revealed that group size (F 1, 18 = 4.65, P = 0.05) and the population*relatedness interaction (F 1, 18 = 7.44, P = 0.01) were statistically significant predictors of mean strength. Post-hoc simple linear regressions of relatedness and group size on mean strength for each population separately showed a statistically significant positive relationship between group size and strength at Rinconada (β =

0.56, R2 = 0.31, P = 0.02) but not at Los Molles (β = 0.74, R2 = 0.55, P = 0.09). Relatedness alone was not a statistically significant predictor of mean strength at Rinconada (β = -0.08, R2 = 0.01, P = 0.76) but

showed a statistically significant negative relationship at Los Molles (β = -0.85, R2 = 0.72, P = 0.03). The

independent samples t-test for mean association revealed a statistically significant difference (P < 0.01)

between the two populations, with Los Molles (mean = 0.88±0.09) social groups having greater mean

association than groups at Rinconada (mean = 0.44±0.07).

Ecology and network structure

Multiple regression analyses reasonably met the regression model assumptions. For the

2 strength analysis, model-level significance was detected at both Los Molles (F3,29 = 47.47, R = 0.85, P <

2 0.01) and Rinconada (F 3,79 = 6.45, R = 0.20, P < 0.01). Similarly, the model for clustering coefficient was

2 2 significant at both Los Molles (F 3,27 = 11.64, R = 0.59, P <0.01) and Rinconanda (F 3,75 = 15.10, R = 0.39, P

< 0.01). At both sites, analyses revealed statistically significant relationships between the ecological characteristics of burrow systems used by individuals and the individuals’ network parameters. At

Rinconada, there was a statistically significant negative relationship between strength and soil hardness and a statistically significant positive relationship between clustering coefficient and food biomass

(Table 2.2; Figure 2.1a-b). In other words, at Rinconada, as soil hardness increased, individuals’ network strength decreased, whereas when food availability increased, individuals’ clustering coefficient also increased. At Los Molles, there was a statistically significant positive relationship between network strength and soil hardness, food biomass, and burrow density, and a statistically significant negative relationship between clustering coefficient and both food biomass and burrow density (Table 2.2; Figure

2.2a-b). In other words, individuals’ network strength increased with increasing food availability,

increasing soil hardness, and increasing burrow density. However, individuals’ clustering coefficient decreased with increasing food availability and increasing burr ow density.

(a)

(b)

Figure 2.1 Scatterplot showing the statistically significant relationship between (a) soil hardness and strength and (b) food biomass and clustering coefficient for indivi duals at Rinconada in 2007 - 2008

(a)

(b)

Figure 2.2 Scatterplot showing the statistically significant relationship between (a) burrow density and strength and (b) food biomass and clustering coefficient for individ uals at Los Molles in 2007 - 2008

Table 2.2 Multiple regression statistics for individuals’ network parameters vs. weighted habitat characteristics at Rinconada and Los Molles 2007-2008

Rinconada Los Molles Predictor Variable Partial r Beta t-value p-value Partial r Beta t-value p-value Soil hardness Strength -0.41 -0.41 -3.97 < 0.01 0.77 0.64 6.11 < 0.01 Clustering coefficient -0.16 -0.13 -1.38 0.17 -0.33 -0.36 -1.69 0.10

Burrow density Strength 0.07 0.07 0.65 0.52 0.90 1.27 10.69 < 0.01 Clustering coefficient -0.19 -0.16 -1.6 0.11 -0.44 -0.60 -2.39 0.03

Food biomass Strength 0.18 0.17 1.6 0.11 0.43 0.24 2.44 0.02 Clustering coefficient 0.54 0.53 5.49 < 0.01 -0.74 -1.04 -5.40 < 0.01

Discussio n

This study confirms previous evidence for non-kin structure in degu social groups, challenging

the importance of kin selection on this species (see Figure 2.3). My study did not reveal the underlying

social network and ecological drivers of variation in kin structure in degus. However, this study yielded

important insights about the social structure of degus and the impact of habitat quality on degu social

networks. Thus, my study does provide insight into the evolution of non-kin sociality.

Social structure and kinship

In my study, mean pairwise relatedness within social groups was not significantly greater than

would be expected from random pairwise comparisons of individuals selected from the background

population in both Rinconada and Los Molles. These observations, and those made previously in

Rinconada (Quirici et al., 2011), suggest that non-kin group structure is typical of degu sociality and not just a characteristic of one population. Additionally, my observations that group size is not a significant

predictor of group relatedness for either population are consistent with previous findings regarding the mechanisms of group formation in degus at Rinconada. Although natal philopatry plays a role in the formation of degu groups, non-sex bias ed dispersal and the movement of adults between groups are important drivers ( Ebensperger et al. , 2009; Quirici et al., 2011 ). Under these conditions, a negative relationship between group size and relatedness is not expected, as the composition of groups varies based on the relative influence of each mechanism on group formation. Further analysis is need ed to determine the extent to which each mechanism influences group composition, particularly at Los

Molles.

Figure 2.3. Social network structure of (a) a theoretical social group exhibiting kin structure and (b) an actual social group from the Los Molles 2007 population

While my study revealed significant population -level differences in the relationship between vario us social and genetic factors an d group-level network characteristics, these results should be

interpreted with some caution due to the influence of inherent population differences, as well as the

characteristics of the network parameters themselves. For example, while mean association was

significantly higher at Los Molles than Rinconada, this may have been influenced by differences in degu

abundance between populations. Abundance was considerably lower at Los Molles than at Rinconada

across both years. It is likely that higher association values at Los Molles were the result of individuals

having fewer potential associates and thus, being captured in the same burrow systems with individuals

proportionally more often than would occur in populations with more individuals. Similarly, while there

was a significant relationship between group size and strength at Rinconada but not Los Molles, the R 2

of the regression was considerably higher at Los Molles than Rinconada (0.55 vs. 0.31), suggesting that

the lack of significance may in part be linked to a small sample size (n = 6 groups at Los Molles, n = 17

groups at Rinconada).

Theory predicts that kin groups form as a result of natal philopatry, usually by one sex (Emlen,

1995). However, a growing body of evidence calls into question the validity of kin selection as the ultimate driver of sociality across taxa (West and Griffin, 2002; Wilson, 2005). Although natal philopatry

(and the resultant kin groups) remains a common mechanism of group formation in many species, other mechanisms of group formation, including the immigration and emigration of adults, influence group structure in some invertebrates (Trivers and Hare, 1976), birds (Brown, 1987), and mammals (Solomon,

2003; Kappeler, 2008, Ebensperger and Hayes, 2008). At Rinconada, the dispersal of degu offspring is not sex-biased, with both sexes dispersing at roughly the same rate (Ebensperger et al., 2009; Quirici et al.,

2011). Further, the primary determinant of group formation and composition in Rinconada is the disappearance of adults and the movement of adults between social groups. As a result, annual turnover of adults comprising social groups is typically high (Ebensperger et al., 2009), likely explaining low kin structure in this population. Although I did not monitor these behaviors at Los Molles, I expect similar

mechanisms to have evolved to maintain non-kin structure in this population. A test of this hypothesis would require a multi-year study to track individuals and their social affiliations between and within seasons (Ebensperger et al., 2009).

Theory also predicts that relatedness between individuals will facilitate the evolution of cooperation (Emlen, 1982). Closely related individuals are thought to benefit from direct benefits and indirect benefits associated with cooperation. However, in species in which groups form social groups in the absence of kin selection, including degus, unrelated individuals cooperate with little or no chance for indirect fitness benefits. Several studies on cooperative breeders in birds and mammals have shown that helpers may care for unrelated young and that unrelated helpers often invest as heavily in offspring care as close relatives (Dunn et al., 1995; Clutton-Brock, 2000). One possible explanation for this involves potential future direct fitness gains, where an unrelated helper may forfeit direct fitness during a given breeding cycle in order to improve potential future direct fitness (Kokko et al., 2002). For example, Reyer (1984) found that in pied kingfisher ( Ceryle rudis ) social groups, unrelated ‘secondary’ helpers increased their future direct fitness when opportunities to mate became available in subsequent breeding seasons.

Life history may explain the evolution of non-kin groups in some cases where kin selection does not provide an adequate explanation for group-living. For example, kin structure is expected in long-lived species in which social groups experience low turnover rates. Evidence for this hypothesis comes from studies showing kin structure in long-lived species such as African elephants ( Loxodonta Africana ) (Archie et al., 2006), coypus ( Myocastor coypus ) (Tunez et al., 2009), and multiple primate species (Silk, 2002). In contrast, due to high turnover rates, social structure in species with short lifespans often lacks kin structure, as has been seen in woodrats ( Neotoma macrotis ) (Matocq and Lacey, 2004) and black grouse

(Tetrao tetrix ) (Legibre et al., 2008). Similarly, the breeding strategies of animals can have significant effects on the genetic structure of social groups. Singular breeders, species with high levels of

reproductive skew (Brown, 1987; Hayes, 2000; Silk, 2007) are expected to show high levels of group

relatedness since most or all offspring come from one dominant breeder. In contrast, plural breeders, species with low skew in groups, are expected to have lower levels of relatedness between group members (Ross, 2001).

Evidence suggests that a combination of these factors could explain non-kin structure in degus. In

terms of life history, degus have low survival (Ebensperger et al., 2009; Ebensperger et al., 2013) and high

turnover rates from year to year (Ebensperger et al., 2009). In terms of breeding system, degus are plural

breeders with communal care, with most females within social groups showing physical signs of pregnancy and lactation (Hayes et al., 2009). Interestingly, a recent long-term study suggests that degu sociality may have evolved as a strategy to deal with harsh mean environmental conditions (Ebensperger et al., 2014), possibly explaining why group-living persists despite low kin structure.

Social networks and habitat conditions

Contrary to previous work in which local ecological conditions had little predictive power for group sizes (Hayes et al., 2009; Ebensperger et al., 2012a), I observed that social network structure was influenced by local ecological conditions in both populations (see Figure 2.1 and 2.2, Table 2.2). At

Rinconada, the negative relationship between strength and soil hardness suggests that individuals inhabiting burrow systems with softer soil experience stronger and/or more social associations. Previous studies found that the energetic cost of digging in hard soil is greater than digging in soft soils

(Ebensperger and Bozinovic, 2000a) and that degus digging in groups remove more soil per capita than solitary individuals (Ebensperger and Bozinovic, 2000b). Thus, softer soil may provide better habitat and result in a greater degree of sociality. Similarly, the positive relationship between food biomass and clustering coefficient suggests that individuals may be clustering together around burrows where food

resources are abundant. My observation is in agreement with previous studies on invertebrates (Tanner et al., 2011) and vertebrates (Foster et al., 2012) showing that food availability influences a population’s

social network structure.

At Los Molles, the positive relationships between network strength and both food biomass and

burrow density suggest a similar trend that individuals inhabiting high quality habitats having stronger

and/or more social associations. In contrast to the observed trend at Rinconada, the relationship between network strength and soil hardness at Los Molles was positive. This difference may be

explained by site-level differences in ecological conditions (Ebensperger et al., 2012a). Overall, the soil

at Los Molles is softer than Rinconada. Since some level of soil hardness is necessary to maintain the

structure of burrows, it is possible that harder soil provides better habitat quality at Los Molles, as softer

soil may not maintain the burrow structures. Other relationships between ecological conditions and

social network structure observed at Los Molles, but not at Rinconada, are more difficult to interpret.

The negative relationships between clustering coefficient and both food biomass and burrow density

suggest that individuals are not clustering more strongly in areas with abundant food and burrows. It is

possible that differences in predation risk (Ebensperger et al., 2012a) influence the distribution of degus

and thus, social network structure at Los Molles. Examinations of the relationship between spatial and

temporal variation in predator abundance and social network structure are needed to test this

hypothesis. Alternatively, it is possible that these unexpected trends were driven by the effect of small

sample sizes (N=29 individuals) on multiple regression analysis.

Regardless, my results suggest that local ecological conditions influence social interactions and

help shape social structure in degu populations. Intraspecific social variation in response to local

ecological conditions has also been demonstrated in numerous taxa (Lott, 1991; Schradin, 2013),

including reptiles (Shine and Fitzgerald, 1995), birds (Davies and Lundberg, 1984) and mammals

(MacDonald, 1979; Roberts et al., 1998). Regarding social networks, Henzi et al. (2009) found that

female associations of chacma baboons ( Papio hamydra ursinus ) varied cyclically in relation to temporal

variation in food abundance. In this sense, degu social structure seems to fit within a common theme,

that local ecological conditions are a significant driver of social variation across species. Future work

should aim to determine if the processes (e.g. phenotypic plasticity) underlying intraspecfic variation in social structure (Schradin, 2013) differ between sites. Such work could have important implications for fully explaining the drivers of social variation.

Concluding remarks

The major take-home point of this study is that degu kin structure is consistently non-kin based and largely insensitive to intraspecific variation in ecological and social network structure. Thus, the results of this and previous studies on degus (Ebensperger et al., 2009; Quirici et al., 2011) suggest a social system that does not conform to principles of long-standing paradigms for animal sociality (Hamilton, 1964;

Emlen, 1982). However, my findings also demonstrate that degu social network structure is influenced by local ecological conditions, and that these influences may result in population-specific social structure at the individual level. To fully understand these relationships, future work should investigate how degu social networks vary in relation to temporal changes in ecological conditions across populations. At the broader scale, researchers need to further examine the complex relationships between life history, ecological conditions, and social/kin structure. To accomplish this, future research should make use of large comparative databases (e.g. PanTHERIA, Jones et al., 2009; Lukas and Clutton-Brock, 2012) to determine if the relationships between these factors are consistent across taxa.

CHAPTER III

CONSERVATION IMPLICATIONS

Understanding the variables that shape a population’s social structure has wide-ranging implications for the conservation and management of social species. Population viability is closely tied to effective population sizes, which in part may be determined by aspects of sociality, such as reproductive skew and genetic variation (Anthony and Blumstein, 2000). Thus, social species present different challenges regarding conservation than solitary species, as in many cases populations must be considered not solely by the total number of individuals but by the number of social units (e.g. social groups).

Conservation biologists need to consider the extent to which animals are social, including whether a species is ‘socially obligate’ or ‘socially flexible.’ In socially obligate species (i.e. species in which individuals cannot survive under solitary conditions), the effective population size is significantly reduced, as each individual can only survive within a social group with a minimum number of other individuals. This presents a significant constraint on populations of such species, as below a certain population density threshold it may be impossible to maintain sufficient group sizes, and the population will experience the consequences of an Allee effect (i.e. negative growth rate at low densities, Allee et al., 1949; Courchamp et al., 1999; Figure 3.1). In response to Allee effects, greater aggregations of individuals may also change fundamental aspects of the population’s ecology, such as its mating system and susceptibility to disease outbreaks (Stephens and Sutherland, 1999).

Such constraints imposed by sociality have been documented across diverse taxa. White-winged

choughs ( Corcorax melanorhamphos ) are incapable of successfully breeding in groups of less than four

(Heinsohn, 1992). Cant (1998) found that banded mongoose ( Mungos mungo ) groups with less than six individuals failed to successfully raise any pups despite several breeding attempts over multiple years.

Further, it has been suggested that high rates of group extinction within such species are a direct result of this minimum threshold related to sociality (Clutton-Brock et al., 1999). Thus, adequately understanding the social dynamics of species may be critical to determining the population densities necessary to prevent local extinctions.

In addition to determining the conditions under which a species may survive, social structure can also affect the potential impacts of anthropogenic changes on population viability. Behavioral traits in social species may be affected by human-induced habitat changes. For example, habitat alteration or fragmentation may lead to changes in a population’s mating system and rates of migration and dispersal, which in turn may affect the actual population size, the levels of reproductive skew, or the population growth rate (Anthony and Blumstein, 2000), all of which may impact the likelihood of the population to grow or diminish.

Finally, wildlife managers may be able to use the social structure of a species or population to potentially improve conservation efforts. In particular, since this study and others have demonstrated how local ecological conditions may affect social structure, managers may use this knowledge to create conditions in which survival and reproduction are maximized, potentially helping to conserve imperiled species. For example, if it is known that individuals are attracted to areas with other conspecifics, it may be possible to induce migrations to higher quality habitat by translocating individuals and/or using decoys to attract larger sects of the population (Stephens and Sutherland, 1999). Understanding the underlying causes of reproductive skew and suppression may also help managers create conditions

under which reproductive success is maximized. For example, if it is known that a social structure is typically affected by dominant individuals suppressing the reproduction of subordinates in large groups, altering resource distributions in a manner that creates several smaller groups rather than a few large groups may result in fewer individuals being reproductively suppressed (Anthony and Blumstein, 2000).

Thus, this and similar work provides the opportunity to both further investigate the evolutionary significance of sociality and also gain practical knowledge toward improving conservation efforts in social species.

Figure 3.1 Example of the Allee effect, demonstrating a reduced population growth rate at low density (from Kuussaari et al. 1998)

REFERENCES

Abbot, P., Abe, J., Alcock, J., Alizon, S., Alpedrinha, J. A., Andersson, M., Andre, J.P., Van Balalen, M., Balloux, F., et al. 2011. Inclusive fitness theory and eusociality. Nature . 472: E1-E4.

Akaike, H. 1974. A new look at the statistical model identification. Automatic Control, IEEE Transactions. 19: 716-723.

Allee, W. C., Park, O., Emerson, A. E., Park, T., and Schmidt, K. P. 1949. Principles of Animal Ecology . WB Saundere Co. Ltd.

Anthony, L. L., and Blumstein, D. T. 2000. Integrating behaviour into wildlife conservation: the multiple

ways that behaviour can reduce N e. Biological Conservation . 95: 303-315.

Archie, E. A., Moss, C. J., and Alberts, S. C. 2006. The ties that bind: genetic relatedness predicts the fission and fusion of social groups in wild African elephants. Proceedings of the Royal Society B: Biological Sciences. 273: 513-522.

Arnold, K. E., and Owens, I. P. 1998. Cooperative breeding in birds: a comparative test of the life history hypothesis. Proceedings of the Royal Society of London. Series B: Biological Sciences . 265: 739- 745.

Arnold, K. E., and Owens, I. P. 1999. Cooperative breeding in birds: the role of ecology. Behavioral Ecology . 10: 465-471.

Avise, J. C., and Shapiro, D. Y. 1986. Evaluating kinship of newly settled juveniles within social groups of the coral reef fish, Anthias squamipinnis . Evolution . 40: 1051-1059.

Bergmüller, R., Heg, D., and Taborsky, M. 2005. Helpers in a cooperatively breeding cichlid stay and pay or disperse and breed, depending on ecological constraints. Proceedings of the Royal Society B: Biological Sciences . 272: 325-331.

Brashares, J. S., and Arcese, P. 2002. Role of forage, habitat and predation in the behavioural plasticity of a small African antelope. Journal of Animal Ecology . 71: 626-638.

Brown, C. R., and Brown, M. B. 1986. Ectoparasitism as a cost of coloniality in cliff swallows ( Hirundo pyrrhonota ). Ecology . 67: 1206-1218.

Brown, J. L. 1987. Helping and communal breeding in birds. Princeton University Press.

Calvert, W. H., Hedrick, L. E., and Brower, L. P. 1979. Mortality of the monarch butterfly ( Danaus plexippus ): avian predation at five overwintering sites in Mexico. Science . 204: 847-851.

Cant, M. A. 1998. A model for the evolution of reproductive skew without reproductive suppression. Animal Behaviour . 55: 163-169.

Chapman, C. A., Chapman, L. J., and Wrangham, R. W. 1995. Ecological constraints on group size: an analysis of spider monkey and chimpanzee subgroups. Behavioral Ecology and Sociobiology . 36: 59-70.

Chaverri, G. 2010. Comparative social network analysis in a leaf-roosting bat. Behavioral Ecology and Sociobiology . 64: 1619-1630.

Clutton-Brock, T.H. 2002. Breeding together: kin selection and mutualism in cooperative vertebrates. Science . 296: 69-72.

Clutton-Brock, T. H., Brotherton, P. N. M., O'riain, M. J., Griffin, A. S., Gaynor, D., Kansky, R., Sharpe, L. and McIlrath, G. M. 2001. Contributions to cooperative rearing in meerkats. Animal Behaviour . 61: 705-710.

Clutton-Brock, T. H., Brotherton, P. N. M., O'Riain, M. J., Griffin, A. S., Gaynor, D., Sharpe, L., Kansky, R., Manser, M.B., and McIlrath, G. M. 2000. Individual contributions to babysitting in a cooperative mongoose, Suricata suricatta . Proceedings of the Royal Society of London. Series B: Biological Sciences . 267: 301-305.

Clutton-Brock, T. H., Gaynor, D., McIlrath, G. M., Maccoll, A. D. C., Kansky, R., Chadwick, P., Manser, M., Skinner, J.D., and Brotherton, P.N.M. 1999. Predation, group size and mortality in a cooperative mongoose, Suricata suricatta . Journal of Animal Ecology . 68: 672-683.

Côté, I. M., and Poulin, R. 1995. Parasitism and group size in social animals: a meta-analysis. Behavioral Ecology . 6: 159-165.

Courchamp, F., Clutton-Brock, T.H., and Grenfell, B. 1999. Inverse density dependence and the Allee effect. Trends in Ecology & Evolution . 14: 405-410.

Croft, D. P., and Krause, M. S. B. J. 2004. Is sexual segregation in the guppy, Poecilia reticulata , consistent with the predation risk hypothesis? Environmental Biology of Fishes . 71: 127-133.

Croft, D. P., James, R., Thomas, P. O. R., Hathaway, C., Mawdsley, D., Laland, K. N., and Krause, J. 2006. Social structure and co-operative interactions in a wild population of guppies ( Poecilia reticulata ). Behavioral Ecology and Sociobiology . 59: 644-650.

Curry, R. L. 1988. Group structure, within-group conflict and reproductive tactics in cooperatively breeding Galápagos mockingbirds, Nesomimus parvulus . Animal Behaviour . 36: 1708-1728.

Davies, N. B., and Lundberg, A. 1984. Food distribution and a variable mating system in the dunnock, Prunella modularis . The Journal of Animal Ecology . 53: 895-912.

Dunbar, R. I., and Shultz, S. 2007. Evolution in the social brain. Science. 317: 1344-1347. Dunn, P. O., Cockburn, A., and Mulder, R. A. 1995. Fairy-wren helpers often care for young to which they are unrelated. Proceedings of the Royal Society of London. Series B: Biological Sciences . 259: 339-343.

Ebensperger, L. A., and Bozinovic, F. 2000a. Communal burrowing in the hystricognath rodent, Octodon degus : a benefit of sociality? Behavioral Ecology and Sociobiology . 47: 365-369.

Ebensperger, L. A., and Bozinovic, F. 2000b. Energetics and burrowing behaviour in the semifossorial degu Octodon degus (Rodentia: Octodontidae). Journal of Zoology . 252: 179-186.

Ebensperger, L. A., and Hayes, L. D. 2008. On the dynamics of rodent social groups. Behavioural Processes . 79: 85-92.

Ebensperger, L. A., Chesh, A. S., Castro, R. A., Tolhuysen, L. O., Quirici, V., Burger, J. R., and Hayes, L. D. 2009. Instability rules social groups in the communal breeder rodent Octodon degus . Ethology . 115: 540-554.

Ebensperger, L. A., Chesh, A. S., Castro, R. A., Tolhuysen, L. O., Quirici, V., Burger, J. R., Sobrero, R., and Hayes, L. D. 2011. Burrow limitations and group living in the communally rearing rodent, Octodon degus . Journal of Mammalogy . 92: 21-30.

Ebensperger, L. A., Hurtado, M. J., Soto-Gamboa, M., Lacey, E. A., and Chang, A. T. 2004. Communal nesting and kinship in degus ( Octodon degus ). Naturwissenschaften . 91: 391-395.

Ebensperger, L. A., Tapia, D., Ramírez-Estrada, J., León, C., Soto-Gamboa, M., and Hayes, L. D. 2013. Fecal cortisol levels predict breeding but not survival of females in the short-lived rodent, Octodon degus. General and Comparative Endocrinology . 186: 164-171.

Ebensperger, L. A., Villegas, Á., Abades, S., and Hayes, L. D. 2014. Mean ecological conditions modulate the effects of group living and communal rearing on offspring production and survival. Behavioral Ecology . 25: 862-870.

Ebensperger, L. A., Sobrero, R., Quirici, V., Castro, R. A., Tolhuysen, L. O., Vargas, F., Burger, J.R., Quispe, R., Villavicencio, C.P., Vasquez, R.A., and Hayes, L.D. 2012a. Ecological drivers of group living in two populations of the communally rearing rodent, Octodon degus . Behavioral Ecology and Sociobiology . 66: 261-274.

Ebensperger, L. A., Rivera, D. S., and Hayes, L. D. 2012b. Direct fitness of group living mammals varies with breeding strategy, climate and fitness estimates. Journal of Animal Ecology . 81: 1013-1023.

Edenbrow, M., Darden, S. K., Ramnarine, I. W., Evans, J. P., James, R., and Croft, D. P. 2011. Environmental effects on social interaction networks and male reproductive behaviour in guppies, Poecilia reticulata . Animal Behaviour . 81: 551-558.

Emlen, S. T. 1982. The evolution of helping. I. An ecological constraints model. American Naturalist . 119: 29-39.

Emlen, S. T. 1995. An evolutionary theory of the family. Proceedings of the National Academy of Sciences . 92: 8092-8099.

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

Emlen, S. T., and Wrege, P. H. 1991. Breeding biology of white-fronted bee-eaters at Nakuru: the influence of helpers on breeder fitness. The Journal of Animal Ecology . 60: 309-326.

Faulkes, C. G., and Bennett, N. C. 2001. Family values: group dynamics and social control of reproduction in African mole-rats. Trends in Ecology & Evolution . 16: 184-190.

Faulkes, C. G., Bennett, N. C., Bruford, M. W., O'brien, H. P., Aguilar, G. H., and Jarvis, J. U. M. 1997. Ecological constraints drive social evolution in the African mole–rats. Proceedings of the Royal Society of London. Series B: Biological Sciences . 264: 1619-1627.

Field, J., Foster, W., Shreeves, G., and Sumner, S. 1998. Ecological constraints on independent nesting in facultatively eusocial hover wasps. Proceedings of the Royal Society of London. Series B: Biological Sciences . 265: 973-977. Foster, E. A., Franks, D. W., Morrell, L. J., Balcomb, K. C., Parsons, K. M., van Ginneken, A., and Croft, D. P. 2012. Social network correlates of food availability in an endangered population of killer whales, Orcinus orca . Animal Behaviour . 83: 731-736.

Ginsberg, J. R., and Young, T. P. 1992. Measuring association between individuals or groups in behavioural studies. Animal Behaviour . 44: 377-379.

Götmark, F., Winkler, D. W., and Andersson, M. 1986. Flock-feeding on fish schools increases individual success in gulls. Nature . 319: 589-591.

Griesser, M., Nystrand, M., Eggers, S., and Ekman, J. 2008. Social constraints limit dispersal and settlement decisions in a group-living bird species. Behavioral Ecology . 19: 317-324.

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

Guichón, M. L., Borgnia, M., Righi, C. F., Cassini, G. H., and Cassini, M. H. 2003. Social behavior and group formation in the coypu ( Myocastor coypus ) in the Argentinean Pampas. Journal of Mammalogy . 84: 254-262.

Hamilton, W. D. 1964. The genetical evolution of social behaviour I. Journal of Theoretical Biology . 7: 1- 16.

Hayes, L. D. 2000. To nest communally or not to nest communally: a review of rodent communal nesting and nursing. Animal Behaviour . 59: 677-688.

Hayes, L. D., Burger, J. R., Soto-Gamboa, M., Sobrero, R., and Ebensperger, L. A. 2011. Towards an integrative model of sociality in caviomorph rodents. Journal of Mammalogy . 92: 65-77.

Hayes, L. D., Chesh, A. S., and Ebensperger, L. A. 2007. Ecological predictors of range areas and use of burrow systems in the diurnal rodent, Octodon degus. Ethology . 113: 155-165.

Hayes, L. D., Chesh, A. S., Castro, R. A., Tolhuysen, L. O., Burger, J. R., Bhattacharjee, J., and Ebensperger, L. A. 2009. Fitness consequences of group living in the degu, Octodon degus , a plural breeder rodent with communal care. Animal Behaviour . 78: 131-139.

Heinsohn, R. G. 1992. Cooperative enhancement of reproductive success in white-winged choughs. Evolutionary Ecology . 6: 97-114.

Henzi, S. P., Lusseau, D., Weingrill, T., Van Schaik, C. P., and Barrett, L. 2009. Cyclicity in the structure of female baboon social networks. Behavioral Ecology and Sociobiology . 63: 1015-1021.

Jones, K. E., Bielby, J., Cardillo, M., Fritz, S. A., O'Dell, J., Orme, C. D. L., Safi, K., Sechrest, W., Boakes, E.H. et al. 2009. PanTHERIA: a species-level database of life history, ecology, and geography of extant and recently extinct mammals. Ecology . 90: 2648-2648.

Kalinowski, S. T., Wagner, A. P., and Taper, M. L. 2006. ML-RELATE: a computer program for maximum likelihood estimation of relatedness and relationship. Molecular Ecology Notes . 6: 576-579.

Kappeler, P.M. 2008. Genetic and ecological determinants of primate social systems. In Ecology and Social Evolution . 225-243. Springer Berlin Heidelberg.

Kenward, R. E. 1978. Hawks and doves: factors affecting success and selection in goshawk attacks on woodpigeons. The Journal of Animal Ecology . 47: 449-460.

Kokko, H., Johnstone, R. A., and Wright, J. 2002. The evolution of parental and alloparental effort in cooperatively breeding groups: when should helpers pay to stay? Behavioral Ecology . 13: 291- 300.

Komdeur, J. 1992. Importance of habitat saturation and territory quality for evolution of cooperative breeding in the Seychelles warbler. Nature . 358: 493-495.

Komdeur, J. 1996. Influence of age on reproductive performance in the Seychelles warbler. Behavioral Ecology . 7: 417-425.

Komdeur, J., Huffstadt, A., Prast, W., Castle, G., Mileto, R., and Wattel, J. 1995. Transfer experiments of Seychelles warblers to new islands: changes in dispersal and helping behaviour. Animal Behaviour . 49: 695-708.

Krause, J., and Ruxton, G. D. 2002. Living in groups. Oxford University Press. 42

Kruuk, H. 1972. The spotted hyena: a study of predation and social behavior. University of Chicago Press.

Kuussaari, M., Saccheri, I., Camara, M., and Hanski, I. 1998. Allee effect and population dynamics in the Glanville fritillary butterfly. Oikos . 82: 384-392.

Lacey, E. A. 2004. Sociality reduces individual direct fitness in a communally breeding rodent, the colonial tuco-tuco ( Ctenomys sociabilis ). Behavioral Ecology and Sociobiology . 56: 449-457.

Lacey, E.A., and Sherman, P.W. 2007. The ecology of sociality in rodents. In Rodent societies: an ecological and evolutionary perspective . 243-254. University of Chicago Press.

Lebigre, C., Alatalo, R. V., Forss, H. E., and Siitari, H. 2008. Low levels of relatedness on black grouse leks despite male philopatry. Molecular Ecology . 17: 4512-4521.

Lott, D. F. 1991. Intraspecific variation in the social systems of wild vertebrates. Cambridge University Press.

Lucia, K. E., Keane, B., Hayes, L. D., Lin, Y. K., Schaefer, R. L., and Solomon, N. G. 2008. Philopatry in prairie voles: an evaluation of the habitat saturation hypothesis. Behavioral Ecology . 19: 774- 783.

Lukas, D., and Clutton-Brock, T.H. 2012. Life histories and the evolution of cooperative breeding in mammals. Proceedings of the Royal Society B: Biological Sciences . 279: 4065-4070.

Macdonald, D. W. 1979. The flexible social system of the golden jackal, Canis aureus . Behavioral Ecology and Sociobiology . 5: 17-38.

Magurran, A. E., and Seghers, B. H. 1991. Variation in schooling and aggression amongst guppy ( Poecilia reticulata ) populations in Trinidad. Behaviour . 118: 214-234.

Matocq, M. D., and Lacey, E. A. 2004. Philopatry, kin clusters, and genetic relatedness in a population of woodrats ( Neotoma macrotis ). Behavioral Ecology . 15: 647-653.

Maynard-Smith, J. 1964. Group selection and kin selection. Nature . 201: 1145-1147.

Meserve, P. L., Martin, R. E., and Rodriguez, J. A. I. M. E. 1984. Comparative ecology of the caviomorph rodent Octodon degus in two Chilean mediterranean-type communities. Revista Chilena de Historia Natural . 57: 79-89.

Nowak, M. A., Tarnita, C. E., and Wilson, E. O. 2010. The evolution of eusociality. Nature . 466: 1057- 1062.

Ophir, A. G., Phelps, S. M., Sorin, A. B., and Wolff, J. O. 2007. Morphological, genetic, and behavioral comparisons of two prairie vole populations in the field and laboratory. Journal of Mammalogy . 88: 989-999.

Quan, Y. F., MacManes, M. D., Ebensperger, L. A., Lacey, E. A., and Hayes, L. D. 2009. Isolation and characterization of polymorphic microsatellite loci from Octodon degus . Molecular Ecology Resources . 9: 999-1001.

Queller, D. C. 2000. Relatedness and the fraternal major transitions. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences . 355: 1647-1655.

Quirici, V., Faugeron, S., Hayes, L. D., and Ebensperger, L. A. 2011. Absence of kin structure in a population of the group-living rodent Octodon degus . Behavioral Ecology . 22: 248-254.

Quirici, V., Faugeron, S., Hayes, L. D., and Ebensperger, L. A. 2011. The influence of group size on natal dispersal in the communally rearing and semifossorial rodent, Octodon degus . Behavioral Ecology and Sociobiology . 65: 787-798.

Quispe, R., Villavicencio, C.P., Vasquez, R.A., and Hayes, L. D. 2012. Ecological drivers of group living in two populations of the communally rearing rodent, Octodon degus . Behavioral Ecology and Sociobiology . 66: 261-274.

Randall, J. A., Rogovin, K., Parker, P. G., and Eimes, J. A. 2005. Flexible social structure of a desert rodent, Rhombomys opimus : philopatry, kinship, and ecological constraints. Behavioral Ecology . 16: 961- 973.

Rasa, O. A. E. 1989. The costs and effectiveness of vigilance behaviour in the dwarf mongoose: implications for fitness and optimal group size. Ethology Ecology & Evolution . 1: 265-282.

Raymond, M., and Rousset, F. 1995. GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity . 86: 248-249.

Rehan, S. M., Schwarz, M. P., and Richards, M. H. 2011. Fitness consequences of ecological constraints and implications for the evolution of sociality in an incipiently social bee. Biological Journal of the Linnean Society . 103: 57-67.

Reyer, H. U. 1984. Investment and relatedness: A cost/benefit analysis of breeding and helping in the pied kingfisher ( Ceryle rudis ). Animal Behaviour . 32: 1163-1178.

Riehl, C. 2013. Evolutionary routes to non-kin cooperative breeding in birds. Proceedings of the Royal Society B: Biological Sciences . 280: 1772.

Roberts, R. L., Williams, J. R., Wang, A. K., and Carter, C. 1998. Cooperative breeding and monogamy in prairie voles: influence of the sire and geographical variation. Animal Behaviour . 55: 1131-1140.

Ross, K. G. 2001. Molecular ecology of social behaviour: analyses of breeding systems and genetic structure. Molecular Ecology . 10: 265-284.

Schoepf, I., and Schradin, C. 2012. Better off alone! Reproductive competition and ecological constraints determine sociality in the African striped mouse ( Rhabdomys pumilio ). Journal of Animal Ecology . 81: 649-656. 44

Schradin, C. 2013. Intraspecific variation in social organization by genetic variation, developmental plasticity, social flexibility or entirely extrinsic factors. Philosophical Transactions of the Royal Society B: Biological Sciences . 368: 1618-1629.

Seppä, P., Fernández-Escudero, I., Gyllenstrand, N., and Pamilo, P. 2008. Colony fission affects kinship in a social insect. Behavioral Ecology and Sociobiology . 62: 589-597.

Shine, R., and Fitzgerald, M. 1995. Variation in mating systems and sexual size dimorphism between populations of the Australian python Morelia spilota (Serpentes: Pythonidae). Oecologia . 103: 490-498.

Shultz, S., Opie, C., and Atkinson, Q. D. 2011. Stepwise evolution of stable sociality in primates. Nature . 479: 219-222. Sih, A., Hanser, S. F., and McHugh, K. A. 2009. Social network theory: new insights and issues for behavioral ecologists. Behavioral Ecology and Sociobiology . 63: 975-988.

Silk, J. B. 2002. Kin selection in primate groups. International Journal of Primatology . 23: 849-875.

Silk, J. B. 2007. The adaptive value of sociality in mammalian groups. Philosophical Transactions of the Royal Society B: Biological Sciences . 362: 539-559.

Smith, J. E. 2014. Hamilton's legacy: kinship, cooperation and social tolerance in mammalian groups. Animal Behaviour . 92: 291-304.

Solomon, N. G. 2003. A reexamination of factors influencing philopatry in rodents. Journal of Mammalogy . 84: 1182-1197.

Stacey, P. B., and Ligon, J. D. 1991. The benefits-of-philopatry hypothesis for the evolution of cooperative breeding: variation in territory quality and group size effects. American Naturalist . 137: 831-846.

Stephens, P. A., and Sutherland, W. J. 1999. Consequences of the Allee effect for behaviour, ecology and conservation. Trends in Ecology & Evolution . 14: 401-405.

Tanner, C. J., Salali, G. D., and Jackson, A. L. 2011. Feeding and non-feeding aggression can be induced in invasive shore crabs by altering food distribution. Behavioral Ecology and Sociobiology . 65: 249- 256.

Thiel, M. 2011. The evolution of sociality: peracarid crustaceans as model organisms. New Frontiers in Crustacean Biology . 285-297.

Travis, S. E., and Slobodchikoff, C. N. 1993. Effects of food resource distribution on the social system of Gunnison's prairie dog ( Cynomys gunnisoni ). Canadian Journal of Zoology . 71: 1186-1192.

Travis, S. E., Slobodchikoff, C. N., and Keim, P. 1995. Ecological and demographic effects on intraspecific variation in the social system of prairie dogs. Ecology . 76: 1794-1803.

Treherne, J. E., and Foster, W. A. 1980. The effects of group size on predator avoidance in a marine insect. Animal Behaviour . 28: 1119-1122.

Trivers, R. L., and Hare, H. 1976. Haploidploidy and the evolution of the social insect. Science , 191: 249- 263.

Túnez, J. I., Guichón, M. L., Centrón, D., Henderson, A. P., Callahan, C., and Cassini, M. H. 2009. Relatedness and social organization of coypus in the Argentinean pampas. Molecular Ecology . 18: 147-155.

Van Schaik, C. P., and Kappeler, P. M. 1997. Infanticide risk and the evolution of male–female association in primates. Proceedings of the Royal Society of London. Series B: Biological Sciences . 264: 1687- 1694.

Ward, P., and Zahavi, A. 1973. The importance of certain assemblages of birds as “information-centres” for food-finding. Ibis . 115: 517-534.

West, S. A., Murray, M. G., Machado, C. A., Griffin, A. S., and Herre, E. A. 2001. Testing Hamilton's rule with competition between relatives. Nature . 409: 510-513.

West, S. A., Pen, I., and Griffin, A. S. 2002. Cooperation and competition between relatives. Science . 296: 72-75.

Wey, T. W., Burger, J. R., Ebensperger, L. A., and Hayes, L. D. 2013. Reproductive correlates of social network variation in plurally breeding degus ( Octodon degus ). Animal Behavior . 85: 1407-1414.

Wey, T., Blumstein, D. T., Shen, W., and Jordán, F. 2008. Social network analysis of animal behaviour: a promising tool for the study of sociality. Animal Behaviour . 75: 333-344.

Whitehead, H. 2008. Analyzing animal societies: quantitative methods for vertebrate social analysis. University of Chicago Press.

Whitehead, H. 2009. SOCPROG programs: analysing animal social structures. Behavioral Ecology and Sociobiology . 63: 765-778.

Williams, G.C. 1966. Adaptation and natural selection. Princeton University Press.

Wilson, E. O. 2005. Kin selection as the key to altruism: its rise and fall. Social Research: An International Quarterly . 72: 1-8.

Wolf, J. B., and Trillmich, F. 2008. Kin in space: social viscosity in a spatially and genetically substructured network. Proceedings of the Royal Society B: Biological Sciences . 275: 2063-2069.

Wolf, J. B., Mawdsley, D., Trillmich, F., and James, R. 2007. Social structure in a colonial mammal: unravelling hidden structural layers and their foundations by network analysis. Animal Behaviour . 74: 1293-1302.

Wolff, J. O. 1992. Parents suppress reproduction and stimulate dispersal in opposite-sex juvenile white- footed mice. Nature . 359: 409-410.

APPENDIX A

SOCIAL GROUPS

ID = individual’s ear tag number, individuals in the same column represent social groups Rinconada 2007: ID ID ID ID ID ID ID 1012 4052 2434 2223 1155 4002 2202 2454 4012 143 4441 1255 1111 134 2352 4010 3323 1312 3325 2453 2341 155 1253 1314 2053 4443 1041 1445 154 1354 3011 141 Rinconada 2008:

ID ID ID ID ID ID ID ID ID ID 315 4325 1021 4243 4144 1015 3355 412 4422 3221 1023 3400 410 1041 1311 243 124 252 4423 255 240 3252 4322 54 4455 3005 251 214 100 Los Molles 2007:

ID 1 2200 1200 54 Los Molles 2008:

ID ID ID ID ID 3062 3066 3068 2489 2418 3037 3049 2200 3046 3100 3061 3036 3041 3063 3050 3042 3045 3069 2488 3070 810 3040 3071 400

APPENDIX B

GENETIC PROCEDURES

DNA Extraction Protocol:

1) Cut tissue sample into small pieces 2) Heat sample at 50 degrees C for 30 minutes 3) Add 440ml extraction buffer 4) Add 44ul SDS 20% 5) Add 12ul Proteinase K 6) Vortex for ~30 seconds 7) Heat at 60 degrees C for 60 minutes 8) Add 300ul 6M saline solution 9) Vortex for at least 60 seconds 10) Centrifuge for 30 minutes at 10,000 RPM and 4 degrees C 11) Pipette 750ul into new tubes 12) Add 750 ul isopropanol 13) Invert tubes to mix 14) Incubate at -20 degrees C for 30 minutes 15) Centrifuge for 20 minutes at 13,000 RPM and 4 degrees C 16) Pour out excess solution, careful to keep DNA precipitate in tube 17) Dry inside of tube with paper towel 18) Add 200ul 70% ethanol 19) Centrifuge for 10 minutes at 13,000 RPM and 4 degrees C 20) Remove excess liquid again with paper towel 21) Dry sample for 1-2 hours at 60 degrees C 22 Dissolve DNA in 100-300ul of water

PCR Primer Mixture Volumes and Annealing Temperatures:

Locus Mg dTNP Primer Taq DNA Temp OCDE3 2.0 0.8 0.8 0.75 30 62 OCDE6 1.75 0.4 0.53 0.5 30 58 OCDE11 1.75 0.4 0.53 0.5 30 58.5 OCDE12 1.75 0.4 0.53 0.5 50 63

OCDE13 1.5 0.2 0.53 0.5 20 58

Gel Electrophoresis:

1) To make 1.5% agarose gel, mix 1.5g agarose and 100ml TAE 1x in flask and heat until completely dissolved 2) Add 2.5ul ethidium bromide to flask 3) Pour mixture into gel box, allow to solidify for ~45 minutes 4) Fill electrophoresis chamber with TAE 1x, add gel 5) Put drops of 2ul buffer on parafilm (1 for each sample) 6) Add 4ul PCR product for each sample to drops, then load in gel (leaving first slot for molecular ladder) 51

7) Add 2 ul ladder to 2ul buffer and load in first slot 8) Run electrophoresis for 45 minutes at 110 volts

APPENDIX C

GENETIC DATA

Mean Relatedness of Social Groups:

2007 Rinconada:

Grp Relatedness 1 0.186 2 0.082 3 0.148 4 0.166 5 0.139 6 0.114 7 0.088 8 0.112 9 0.069

2008 Rinconada:

Grp Relatedness 1 0.114 2 0.071 3 0.212 4 0.155 5 0.088 6 0.11 7 0.122 8 0.136 9 0.094 10 0.073

2007 Los Molles:

Grp Relatedness 1 0.094

2008 Los Molles:

1 0.091 2 0.213 3 0.246 4 0.89 5 0.102

Background Relatedness Confidence Intervals:

GS = group size, 95 CI = 95% confidence interval

Rinconanda:

GS 95 CI 3 0-0.31 7 0-0.21 6 0-0.22 4 0-0.25 3 0-0.31 3 0-0.31 6 0-0.22 5 0-0.24 2 0-0.41 2 0-0.41 3 0-0.31 3 0-0.31 3 0-0.31 2 0-0.41 2 0-0.41 3 0-0.31 4 0-0.25

Los Molles:

GS 95 CI 4 0-0.23 6 0-0.23 2 0-0.46 3 0-0.33 6 0-0.23 55

7 0-0.23

Primer data: Ho and He are the observed and expected levels of heterozygosity. Rinconada:

# Primer alleles Ho He P value OCDE 3 8 0.71 0.84 0.38 OCDE 6 5 0.45 0.74 <0.01 OCDE 11 8 0.59 0.63 0.44 OCDE 12 7 0.47 0.72 <0.01 OCDE 13 12 0.79 0.84 0.29

Los Molles:

# Primer alleles Ho He P value OCDE 3 7 0.54 0.66 0.32 OCDE 6 6 0.48 0.79 <0.01 OCDE 11 9 0.67 0.79 0.28 OCDE 12 6 0.51 0.68 <0.01 OCDE 13 10 0.91 0.81 0.52

APPENDIX D

ECOLOGICAL DATA

Rinconada

Biomass Burrows Soil 2 2 (g/m ) (#/m ) (kPa) 79.27 0.16 2971.90 81.51 0.15 3052.01 106.83 0.07 2976.81 108.42 0.14 3059.76 93.10 0.12 3067.14 87.89 0.17 2988.10 111.73 0.06 2977.22 107.73 0.06 3001.27 100.73 0.07 3044.40 124.68 0.18 3086.28 146.16 0.08 2932.89 133.36 0.19 3055.96 121.38 0.16 3094.33 131.52 0.18 3079.69 111.23 0.13 3112.47 99.87 0.14 3095.54 131.16 0.16 3067.14 130.49 0.16 3068.93 87.44 0.18 2956.95 83.14 0.17 2963.87 110.96 0.12 3057.11 91.84 0.15 3057.11 84.55 0.14 3062.32 79.05 0.17 3051.46 79.31 0.11 3130.77 89.17 0.16 3047.84 99.07 0.06 3070.28 99.03 0.15 3049.44 62.84 0.17 3064.29 114.53 0.15 3116.28 107.64 0.08 3056.33

82.49 0.16 3037.87 133.45 0.19 3055.74 126.09 0.19 3053.63 58

82.56 0.17 3015.70 91.81 0.14 3096.29 65.94 0.17 3088.38 118.78 0.17 3062.85 99.67 0.16 3087.36 104.71 0.15 3052.50 116.08 0.14 3042.05 126.78 0.17 3084.64 132.93 0.17 3046.01 58.33 0.18 3104.00 105.26 0.07 3008.44 74.60 0.14 3121.56 92.80 0.07 3089.10 148.48 0.09 3038.29 101.73 0.13 3133.80 69.26 0.13 3116.19 281.60 0.09 3060.87 242.24 0.10 3050.98 117.60 0.10 2919.72 92.80 0.07 3089.10 117.60 0.10 2919.72 133.68 0.09 3052.05 281.60 0.09 3060.87 136.34 0.10 2935.85 54.46 0.17 3111.54 196.10 0.13 3106.15 130.83 0.08 2950.81 92.80 0.07 3089.10 61.94 0.15 3126.74

117.60 0.10 2919.72 57.29 0.09 3126.33 139.09 0.14 3139.20 164.20 0.13 3112.62

281.60 0.09 3060.87 281.60 0.09 3060.87 84.75 0.09 3137.26 87.81 0.13 2940.05 111.20 0.09 3131.97 59

126.52 0.14 3137.54 63.60 0.10 3123.97 154.49 0.16 3145.56 91.89 0.10 3139.73 96.36 0.10 3130.59 82.61 0.14 3130.21 85.34 0.11 3130.13 241.45 0.09 3016.62

Los Molles

Biomass Burrows Soil (g/m 2) (#/m 2) (kPa) 0.00 0.15 2428.65 0.00 0.15 2428.65 0.00 0.15 2428.65 0.00 0.15 2428.65 509.71 0.16 1585.38 39.20 0.20 2180.09 39.20 0.20 2180.09 0.00 0.27 1186.40 296.08 0.16 2180.47 601.60 0.16 1329.43 39.20 0.20 2180.09 8.80 0.20 1111.12 376.48 0.16 1956.51 568.18 0.16 1341.56 340.30 0.16 2057.29 0.00 0.27 1186.40 0.00 0.27 1186.40 280.00 0.16 2225.26 134.40 0.11 1743.47 39.20 0.20 2180.09 8.80 0.20 1111.12 8.80 0.20 1111.12 0.00 0.27 1186.40 134.40 0.11 1743.47 39.20 0.20 2180.09 39.20 0.20 2180.09 0.00 0.27 1186.40 60

APPENDIX E

NETWORK DATA

Rinconada:

Clustering ID (sex) Strength Eig Centrality Reach Coeff Affinity 0124 (F) 1.25 0 2.85 0.26 2.28 0134 (F) 2.84 0.01 6.85 0.22 2.41 0141 (F) 2.34 0 4.96 0.23 2.12 0143 (F) 1.43 0 2.23 0.18 1.56 0151 (F) 1.62 0 3.32 0.21 2.05 0153 (F) 1.76 0 3.76 0.16 2.14 0154 (F) 1.38 0 3.24 0.34 2.35 0155 (F) 2.57 0 5.3 0.19 2.06 0432 (M) 2.41 0 5.36 0.26 2.22 1012 (M) 2.62 0 5.51 0.18 2.1 1014 (F) 1.83 0.1 4.91 0.14 2.68 1033 (F) 1.61 0.01 2.64 0.2 1.64 1041 (F) 1.53 0.21 4.12 0.31 2.69 1042 (F) 3.57 0.51 9.49 0.36 2.66 1103 (F) 2.9 0.02 6.11 0.38 2.11 1111 (F) 1.79 0.03 4.27 0.41 2.39 1115 (M) 2.78 0.02 5.82 0.37 2.09 1155 (F) 2.37 0 5.54 0.24 2.34 1251 (F) 0.87 0 1.21 0.07 1.4 1252 (M) 0.41 0.02 0.66 0 1.62 1253 (F) 2.78 0.38 7.17 0.39 2.58 1255 (F) 2.35 0.03 4.91 0.24 2.09 1311 (F) 1.89 0.04 4.12 0.35 2.18 1312 (F) 2.79 0 5.72 0.21 2.05 1314 (F) 2.26 0 4.65 0.29 2.06 1354 (F) 2.98 0.44 8.22 0.43 2.76 1412 (M) 2.99 0.02 8.34 0.19 2.79 1445 (M) 1.92 0.01 4.11 0.15 2.14 2053 (F) 2.73 0.02 7.22 0.22 2.65 2054 (F) 2.03 0 5.23 0.2 2.57 2202 (F) 3 0.02 6.94 0.21 2.31 2223 (F) 2.36 0 6.15 0.2 2.6 2341 (F) 2.1 0.01 5.29 0.25 2.52 2352 (F) 2.4 0 5.2 0.23 2.17

2434 (F) 1.61 0 2.85 0.16 1.77 2452 (F) 1.85 0 4.14 0.2 2.23 2453 (F) 3.05 0.06 5.87 0.19 1.92 2454 (F) 2.02 0 4.48 0.22 2.22 3011 (F) 2.04 0.01 4.77 0.24 2.34 0.91 0.04 2.54 0 2.79 62

3(34)15 3102 (F) 1.75 0.01 4.9 0.81 2.8 3114 (F) 1.98 0 4.13 0.28 2.09 3210 (F) 1.89 0.02 5.18 0.59 2.74 3251 (F) 1.33 0 2.85 0.09 2.15 3323 (M) 1.98 0 4.4 0.16 2.22 3325 (M) 2.27 0 5.8 0.35 2.55 4002 (M) 1.59 0.02 3.33 0.21 2.09 4010 (F) 2.24 0.32 6.31 0.41 2.82 4012 (F) 2.59 0.31 6.27 0.27 2.42 4043 (F) 1 0 1.52 0.5 1.52 4052 (M) 2.54 0.37 7.09 0.45 2.79 4114 (M) 1.23 0.01 2.07 0.31 1.69 4201 (M) 1.68 0.01 2.47 0.17 1.47 4311 (M) 2.3 0 5.55 0.31 2.41 4441 (F) 3.36 0.01 7.77 0.21 2.31 4443 (F) 1.53 0 3.58 0.2 2.34

Eig Clustering ID (sex) Strength Centrality Reach Coeff Affinity 0001(F) 1.13 0 1.26 0.27 1.12 0043(M) 0 0 0 NaN NaN 0054(M) 2 0 4.42 0.92 2.21 0100(M) 1.3 0.2 4.44 0.87 3.42 0124(F) 0.1 0 0.11 NaN 1.08 0130(F) 0 0 0 NaN NaN 0203(F) 1.07 0 1.19 0.24 1.11 0214(F) 2.88 0.1 8.45 0.88 2.93 0215(F) 1.39 0 1.76 0.1 1.27 0220(F) 1 0.17 4.13 NaN 4.13 0240(F) 3.31 0.47 10.73 0.61 3.24

0243(F) 1.99 0 4.38 0.87 2.2 0251(F) 3.32 0.46 10.31 0.57 3.11 0252(F) 2.07 0.1 4.02 0.33 1.94 0255(F) 3 0.09 8.57 0.88 2.86 0315(F) 4.13 0.49 10.35 0.33 2.5 0410(F) 1.35 0 1.93 0.24 1.43 0411(F) 0.89 0 0.59 0.01 0.67 0412(F) 1.92 0.13 4.61 0.4 2.4 948(M) 0.5 0 1.21 NaN 2.42 1015(F) 2.42 0 4.09 0.49 1.69 63

1021(M) 1.48 0 1.98 0.19 1.34 1023(F) 3.14 0.45 10.02 0.63 3.2 1041(F) 0.56 0 0.64 0.33 1.14 1311(F) 0.41 0 0.31 0.17 0.77 1445(M) 0.7 0 0.62 NaN 0.89 2040(F) 1.56 0.06 3.13 0.71 2.02 3005(F) 2.83 0.09 8.26 0.94 2.92 3221(F) 2.85 0.09 8.11 0.81 2.84 3252(M) 1.48 0 1.32 0.03 0.89 3323(M) 2 0.01 3.9 0.73 1.95 3355(F) 0.07 0 0.13 NaN 1.95 3400(F) 0.67 0 0.87 NaN 1.31 4144(F) 0.66 0 0.49 0.07 0.74 4243(F) 0.53 0 0.61 0.36 1.14 4322(F) 0.51 0 0.33 0.1 0.65 4325(F) 1.31 0 1.38 0.11 1.05 4422(F) 1.22 0 1.27 0.23 1.04 4423(F) 1.08 0 1.11 0.2 1.03 4455(F) 0.64 0 0.81 0.15 1.27 022(01)(F) 1.95 0.02 3.86 0.66 1.98 13(15)1(F) 1.95 0.03 3.97 0.65 2.04

Los Molles

Clustering ID (sex) Strength Eig Centrality Reach Coeff Affinity 0001(F) 4.5 0.4 21.08 0.95 4.69 0011(F) 3 0 11 1 3.67 0024(M) 4 0.36 19.11 0.99 4.78 0054(F) 4.5 0.4 21.08 0.95 4.69 0055(F) 1.33 0 1.83 0.4 1.38 0100(F) 1.17 0 1.81 0.33 1.55 0300(F) 4 0 12 0.67 3 1000(F) 4 0 12 0.67 3 1200(F) 5 0.43 22.67 0.88 4.53 1300(M) 3 0 11 1 3.67 2200(F) 4.67 0.41 21.67 0.93 4.64 2400(F) 1.83 0 2.36 0.22 1.29 3100(F) 2 0 8 1 4 3300(F) 1.33 0 2.28 0.33 1.71 4100(F) 5 0.43 22.67 0.88 4.53

Clustering ID (sex) Strength Eig Centrality Reach Coeff Affinity 64

0400(F) 3.54 0 13.69 0.49 3.87 2200(F) 6 0 34 0.93 5.67 2418(F) 7.75 0.36 51.6 0.77 6.66 2488(F) 2.88 0 10.25 0.69 3.56 2489(F) 4.14 0 16.61 0.69 4.01 3036(F) 5 0 30 1 6 3037(F) 2 0 4 1 2 3040(M) 4.32 0 15.52 0.41 3.6 3041(F) 4.15 0 16.79 0.7 4.05 3042(F) 3.46 0 12.38 0.56 3.58 3043(F) 2 0 4 1 2 3044(F) 6.92 0.3 43.92 0.66 6.35 3045(M) 8 0.37 53.83 0.76 6.73 3046(F) 3.08 0 10.82 0.66 3.51

3048(F) 2 0 4 1 2 3049(F) 1 0 1 NaN 1 3050(M) 6 0 34 0.93 5.67 3061(M) 2 0 4 1 2 3062(F) 2 0 4 1 2 3063(F) 7 0.34 49.13 0.87 7.02 3066(F) 1 0 1 NaN 1 3068(F) 6 0 34 0.93 5.67 3069(F) 6 0 34 0.93 5.67 3070(F) 7.67 0.36 51.11 0.77 6.67 3071(M) 7 0.35 50.33 0.9 7.19 3072(M) 7 0.35 50.33 0.9 7.19 3074(F) 2 0 4 1 2 3099(F) 4.33 0 17.79 0.55 4.11 3100(F) 3 0.12 19.13 0.76 6.38 4400(F) 5 0 30 1 6 801(F) 4.5 0.19 29.12 0.69 6.47 806(F) 6 0.31 44.42 1 7.4 810(F) 6 0 34 0.93 5.67 3100d(F) 4.03 0 16.43 0.74 4.07

APPENDIX F

STATISTICAL ANALYSIS

Bootstrapping code for calculating confidence intervals of background relatedness, using R v. 3.1:

> bstrap <- c() > for (i in 1:1000){ + # take the sample + bsample <- sample(x,7,replace=T) (x previously defined as “relatedness” from column in spreadsheet) + # calculate the bootstrap (here n=7 for GS=7) + bestimate <- mean(bsample) + bstrap <- c(bstrap,bestimate)} > #lower bound for 95%: > quantile(bstrap,.025) > #upper bound for 95% > quantile(bstrap,.975)

AIC Code for SAS v. 9.3: Data (GS= group size): DATA groupr; INPUT Population GS Relatedness Association Strength; DATALINES ; 1 3 0.19 0.57 1.51 1 7 0.08 0.30 2.12 1 6 0.15 0.28 2.44 1 4 0.17 0.23 2.69 1 3 0.14 0.06 2.33 1 3 0.11 0.39 2.14 1 6 0.09 0.27 2.44 1 5 0.11 0.91 3.48 1 2 0.07 0.50 0.99 1 2 0.21 0.25 1.42 1 3 0.16 0.31 0.86 1 3 0.09 0.32 0.53 1 3 0.11 0.96 2.14 1 2 0.12 0.83 2.00 1 2 0.14 0.35 1.10 1 3 0.09 0.20 0.98 1 4 0.07 0.82 2.89 2 6 0.09 1.00 6.79 2 2 0.21 1.00 1.00 2 3 0.25 1.00 2.00 2 6 0.09 0.92 5.83 2 7 0.10 0.44 3.65 2 4 0.09 0.94 4.67 ;

Strength Analysis: 67

proc reg data=groupr outest=est; model Strength = population GS relatedness/ selection=adjrsq sse aic ; output out=out p=p r=r; run; quit; proc reg data=groupr outest=est0; model Strength = population GS relatedness/ noint selection=adjrsq sse aic ; output out=out0 p=p r=r; run; quit; data estout; set est est0; run; proc sort data=estout; by _aic_; proc print data=estout(obs=23); run;

Association Analysis: proc reg data=groupr outest=est; model Association = population GS relatedness/ selection=adjrsq sse aic ; output out=out p=p r=r; run; quit; proc reg data=groupr outest=est0; model Association = population GS relatedness/ noint selection=adjrsq sse aic ; output out=out0 p=p r=r; run; quit; data estout; set est est0; run; proc sort data=estout; by _aic_; proc print data=estout(obs=23); run;

VITA

Garrett Davis was born in Springville, New York to the parents of Margaret Ronan and Dennis

Davis. He attended Pioneer High School in Yorkshire, New York and, after graduation, obtained a

Bachelor of Science degree in Wildlife Science from the SUNY College of Environmental Science &

Forestry in Syracuse, New York. Garrett conducted fieldwork in Alabama, California, and Costa Rica before entering the Environmental Sciences graduate program at the University of Tennessee at

Chattanooga in January 2012. He will graduate with his Master of Science degree in Environmental

Science in December 2014.