ABSTRACT

PHILOPATRY IN PRAIRIE VOLES: AN EVALUATION OF THE HABITAT SATURATION HYPOTHESIS

By: Kristen E. Lucia

Philopatry, or delayed dispersal of sexually mature offspring, may be due to ecological constraints on dispersal. In this study I manipulated the population density of prairie voles (Microtus ochrogaster) living in experimental enclosures to test the predictions of the habitat saturation hypothesis that philopatry and subsequent group formation in this cooperatively breeding mammal is affected by the availability of suitable territories. A significantly greater proportion of offspring remained philopatric at high densities, with females being more philopatric than males at all densities. This increase in philopatry led to a significant increase in the proportion of social units that were groups as well as a significant increase in group size. These results provide the strongest evidence for a mammal of a causal affect of density on dispersal and group formation and suggest that habitat saturation is at least a partial explanation for philopatry in prairie voles.

PHILOPATRY IN PRAIRIE VOLES: AN EVALUATION OF THE HABITAT

SATURATION HYPOTHESIS

A Thesis

Submitted to the

Faculty of Miami University

in partial fulfillment of

the requirements for the degree of

Master of Science

Department of Zoology

By

Kristen E. Lucia

Miami University

Oxford, Ohio

2007

Co-Advisor Reader ______Brian Keane

Co-Advisor ______Nancy G. Solomon

Reader ______Robert L. Schaefer TABLE OF CONTENTS

List of Tables iii

List of Figures iv

Dedication v

Acknowledgements vi

Introduction: Cooperative Breeding and Philopatry 1

References 7

Chapter 1: An Evaluation of the Habitat Saturation Hypothesis 12

Introduction 12

Methods 15

Results 22

Discussion 24

References 29

Tables 35

Figures 38

General Conclusions 43

References 45

ii LIST OF TABLES

Table Page

1 Vegetative data. 35

2 Prairie vole social structure. 37

iii LIST OF FIGURES

Figure Page

1 Trapping schedule for study. 38

2 Mean prairie vole group size (± SE) within enclosures as a function of 39

density and year.

3 Proportion of prairie vole social units within an enclosure that were 40

groups as a function of density and year.

4 Proportion of prairie vole offspring within each enclosure that were 41

classified as philopatric during the three week residency period

as a function of density.

5 Proportion of male and female prairie vole offspring within each 42

enclosure that were classified as philopatric during the three week

residency period as a function of density.

iv DEDICATION

To my parents, Maureen and Art Lucia, who always encouraged me to follow my own path.

v ACKNOWLEDGMENTS

I would like to thank my advisors, Brian Keane and Nancy Solomon for their encouragement, advice, and support throughout my time at Miami. I also thank Bob Schaefer for his patient guidance through the statistical gauntlet that is animal behavior research. I would also like to thank Ann Rypstra, Rodney Kolb, and the ERC staff for providing technical support during my two field seasons. I thank Tom Crist and Todd Levine, who helped with density estimations and program MARK. I also thank the animal care staff for their aid in maintaining the prairie vole colony in Boyd. I thank Ashley Richmond, Michelle Edwards, Tony Fries, John Williams, Lisa Aschemeier, and all the vole wranglers who made this research possible. I also thank Craig Streatfeild who let me bounce countless ideas off him and was always willing to help with the residency saga. I am grossly indebted to my good friends Kathy, Jen, Jenn, Sam, Harry, Will, Andy, Janelle, Padre, and Phill, who supported me through the last three years and provided as much aid as they could (and when all else failed joined me in my insanity). I offer many thanks to Jack Cranford, who helped foster my love for mammalogy, and all those at Virginia Tech who helped mold me into the scientist I am today. I also thank my family for their love and support throughout my graduate career. I thank Maddie for her love and for putting up with my long days at school and in the field and all my friends who have if nothing else put up with the tiresome excuse of “I can’t ______, I’m working on my thesis.” The research was financially supported by Miami University Summer Workshop fund, managed by Dave Berg, the National Science Foundation, and the American Society for Mammalogists Grants-in-Aid fund.

vi INTRODUCTION: COOPERATIVE BREEDING AND PHILOPATRY

Cooperative breeding is a social system typically defined by three characteristics: philopatry (delayed dispersal of offspring), reproductive suppression (when mature offspring do not reproduce), and provision of care to offspring that are not one’s own, commonly referred to as alloparental care (Solomon and French 1997). Cooperative breeding is found across the animal kingdom in species of insects and fish as well as birds and mammals (Koenig and Dickinson 2004; Queller 1994; Russell 2004; Taborsky and Limberger 1981). While occurring in only around three percent of bird and mammalian species (Jennions and Macdonald 1994; Koenig and Dickinson 2004; Russell 2004) the distribution of cooperative breeding is not random, but is instead concentrated in certain families, such as Maluridae and Lybiidae in birds and Canidae and Callitrichidae in mammals (Arnold and Owens 1998; Ligon and Burt 2004; Riedman 1982). Why an individual would forgo reproduction and remain at the natal nest to help rear the offspring of others is difficult to explain in light of Darwinian selection. Extensive empirical and theoretical research has been conducted in attempts to determine the factors that led to the evolution of philopatry, and a wide range of hypotheses have been proposed. Almost all of the explanations for philopatry can be categorized into one of three general hypotheses: those focused on life history characteristics, those emphasizing the benefits of philopatry and those based on ecological constraints on dispersal. The life history hypothesis proposes that cooperative breeding evolves in species with similar dispersal patterns and longevity (see review by Hatchwell and Komdeur 2000). The benefits of philopatry hypothesis emphasizes fitness benefits that can be obtained by an individual that remains philopatric (Stacey and Ligon 1991). These benefits can include increased indirect fitness by helping closely related individuals survive and reproduce (kin selection theory – see review by Clutton-Brock 2002). Direct benefits may include the acquisition of an important resource, such as improved survival while on the home territory (Emlen 1995; Kokko and Ekman 2002), access to food stores while on the natal territory (Stacey and Ligon 1987), the development of an important skill such as hunting through practice with group members (Bednarz and Ligon 1988), increased probability of obtaining a territory (Rood 1990), experience raising young or mating access to unrelated opposite sex conspecifics (Jennions

1 and Macdonald 1994). The ecological constraints hypothesis proposes that offspring will remain philopatric when ecological conditions limit the probability of successful dispersal and independent breeding (Emlen 1982). The ecological constraints hypothesis is the only one of the three hypotheses that proposes that dispersal is based on environmental limitations. Emlen (1982) formally developed the ecological constraints hypothesis based on the ideas of Selander (1964), Brown (1969), and Koenig and Pitelka (1981). The ecological constraints hypothesis focuses on four factors that could potentially determine whether an offspring will disperse and attempt to breed or remain philopatric within its natal group: (1) the cost/risk of dispersal, (2) the probability of successfully becoming established on a suitable territory, (3) the probability of obtaining a mate, and (4) the likelihood of successful reproduction once an individual obtains a territory and a mate (Emlen 1982). Many different types of constraints have been investigated in the past. Some researchers have looked at seasonal dynamics like temperature to evaluate changes in dispersal patterns and group size, because individuals may form groups to decrease thermoregulatory costs associated with decreasing temperatures (Getz et al. 1987; Getz et al. 1993). The aridity-food distribution hypothesis states that eusociality in naked mole-rats (Heterocephalus glaber) and Damaraland mole-rats (Cryptomys damarensis) may be due to the constraint on digging imposed by living in an arid environment with patchy food distribution (Jarvis et al. 1994). Tunnel expansion is primarily limited to times immediately following sporadic rainfall events. Group-living individuals work together and make large expansions to the burrow system in the brief window of opportunity a rain provides, thus lowering the cost of foraging and locating patches of food that can be shared among the group members (Jarvis et al. 1994, Lacey and Sherman 1997). A similar hypothesis has been proposed for the gundi (Ctenodactylus gundi), although it remains untested (Nutt 2005). The mate limitation hypothesis states that philopatry occurs due to a limited number of available mates (Marra and Holmes 1997), as seen in the superb fairy-wren, Malurus cyaneus (Pruett- Jones and Lewis 1990). Although poorly studied to date, support for the mate limitation hypothesis is slowly growing as experimental manipulations of available mates are performed (Dickinson and Hatchwell 2004).

2 The best studied ecological constraint is the habitat saturation hypothesis, which states that philopatry occurs due to an absence of suitable breeding locations (i.e., territories, nest sites, or nest cavities), often a result of relatively high population densities that saturate the habitat (Brown 1969; Emlen 1994; Koenig and Pitelka 1981). Much of the support for the habitat saturation hypothesis has come from studies of avian cooperative breeders. The strongest data comes from field studies that manipulate the availability of breeding locations. For example, shortages of hard-to-construct nest cavities in pine trees (multiple species) causes philopatry in red-cockaded woodpeckers (Picoides borealis), but the creation of new nest cavities in areas without pre-constructed cavities led to the dispersal of individuals (Walters et al. 1992). More frequently, support for the habitat saturation hypothesis comes from studies showing a correlation between density and philopatry or group formation. When high quality territories, as defined by characteristics such as food resources and amount of protection from predators, are unavailable due to high adult survival and low territory turnover, the Seychelles warbler (Acrocephalus sechellensis), the splendid fairy-wren (Malurus splendens) and the groove-billed anis (Crotophaga sulcirostris), remain philopatric (Koford et al. 1986; Komdeur 1992; Russell and Rowley 1993). Support for the habitat saturation hypothesis has also been found across a diverse array of mammalian species, where a variety of ecological variables can lead to high density populations and constrain dispersal. Similar to the limitations imposed on the red-cockaded woodpecker, by the cost of construction of nest cavities, mammalian species that utilize underground burrow systems like the woodland vole (Microtus pinetorum) and the European rabbit (Oryctolagus cuniculus) may form groups when no burrows or warrens are available because the cost of digging a new burrow system is so high in some soil types (Cowan 1987; Powell and Fried 1992; Roberts 1987). Variation in food abundance and distribution can also influence philopatry. When food resources are clustered, population densities at the clusters can increase leading to group formation as seen in the great gerbil (Rhombomys opumus), where groups form in higher density populations when food is not limited (Randall et al. 2005). For territorial species, space limitations can also saturate a habitat, as seen with deer mice (Peromyscus maniculatus) and white-footed mice (Peromyscus leucopus), which both exhibit philopatry when densities of resident adults are high (Wolff 1992, 1994). Space limitations are especially constraining in species with high adult survival like the saddle-back

3 tamarin (Saguinus fuscicollis), where breeding vacancies in high density populations are rare and offspring remain philopatric (Goldizen and Terborgh 1989). Although there is extensive support for the habitat saturation hypothesis within the literature, there are also studies that fail to support the habitat saturation hypothesis and/or list it as secondary to other factors leading to philopatry of offspring. For instance, Western American crows (Corvus brachyrhynchos hesperis) exist in nonbreeding flocks or are philopatric and serve as helpers at the nest in areas where known territories (previously used breeding sites) are available (Caffrey 1992). Likewise, the stripe-backed wren (Campylorhynchus nuchalis) and the dwarf mongoose (Helogale parvula) both form groups to protect offspring from predators, and therefore offspring remain philopatric even in the presence of available territories (Rabenold 1984; Rood 1990; Wiley and Rabenold 1984). Deer mice, which Wolff (1994) reported as constrained by habitat saturation, have also been reported to form groups even when population density does not saturate the habitat (Millar and Derrickson 1992). Furthermore, Stacey and Ligon (1991) questioned the utility of the habitat saturation hypothesis as an explanation for the evolution of philopatry and cooperative breeding, citing the aforementioned cooperative breeding species that remain philopatric regardless of ecological constraints like habitat saturation and habitat quality, as well as nomadic and migratory cooperative breeders that should be unaffected by density. Additionally, species like the striped mouse remain philopatric due to benefits of philopatry in addition to ecological constraints like habitat saturation (Schradin and Pillay 2004, 2005). Males and females may respond differently to constraints on dispersal. In mammals, females are traditionally the more philopatric sex, which contrasts the male-biased philopatry typically found in avian species (Greenwood 1980). However, monogamous species generally do not display sex-biased philopatry (Greenwood 1980). Among cooperatively breeding mammals, female-biased philopatry is most common (Clutton-Brock et al. 2002; Russell 2004), as seen in many of the microtine rodents (Lambin 1994). However in some species, like golden and silver-backed jackals (Canis aureus and C. mesomelas), both sexes are equally philopatric (Moehlman and Hofer 1997). Male-biased philopatry is relatively rare, but is common among the cooperatively breeding large canids, like the gray wolf, Canis lupus, and the African wild dog, Lycaon pictus (Moehlman and Hofer 1997).

4 The hypotheses proposed to explain philopatry are not mutually exclusive and the precise role of habitat saturation in determining whether or not an individual will remain philopatric compared to other explanations like benefits of philopatry remains unclear (Solomon 2003). The strongest support for the habitat saturation hypothesis comes from work with avian species, and the evidence for mammals is weaker (Russell 2004). The mammalian literature base addressing the habitat saturation hypothesis to date has primarily relied on correlations with density, like group formation (Cochran and Solomon 2000; Russell 2004), while the avian literature includes studies which found support for the habitat saturation hypothesis through the manipulation of field conditions, like Walters et al. (1992). The differences between birds and mammals, including viviparity and the energy intensive nutritional component to offspring care provided by lactating females, have led researchers to question the ability to extrapolate results from avian cooperative breeding studies to mammalian cooperative breeders (Mumme 1997; Russell 2004). Therefore, the purpose of my study is to experimentally manipulate population density in order to determine the role habitat saturation plays in philopatry in prairie voles (Microtus ochrogaster). This field experiment is the first study to manipulate initial population densities in a cooperatively breeding mammalian species to test predictions of the habitat saturation hypothesis and should provide the strongest test to date of the habitat saturation predictions. Prairie voles are a good model species to test predictions of the habitat saturation hypothesis because their populations naturally vary widely in density, their philopatry rates frequently show a positive correlation with population density (Cochran and Solomon 2000; Getz et al. 1993; Lin and Batzli 2001; McGuire et al. 1993; but see Gaines et al. 1979), and the populations exhibit plasticity in their within-population social organization (Cochran and Solomon 2000; Getz et al. 1979; Getz et al. 1993). Furthermore, recent studies have shown habitat availability to be stronger in dictating prairie vole social interactions than the mate limitation hypothesis, as observed during experimental creation of territory vacancies, with and without potential male mates (Jacquot and Solomon 2004). Females in the study moved onto vacant territories without regard to the presence or absence of males. Additionally, based on the results of their food supplementation experiment Cochran and Solomon (2000) determined that food resources were not determinates of group formation in prairie voles, but noted a positive correlation between density and group formation.

5 Male and female prairie voles may also respond differently to the manipulation of population density. Prairie voles are socially monogamous (Getz et al. 1987) and previous studies of prairie voles have found no sex bias in philopatry in Illinois populations (Lin and Batzli 2004; McGuire et al. 1993). However, Boonstra et al. (1987) demonstrated female- biased philopatry in prairie voles in Kansas populations. Similar results were found in an enclosure study using prairie voles that descended from an Illinois population (Lin et al. 2006). Solomon et al. (2004) demonstrated that while prairie voles are socially monogamous, they are not genetically monogamous. The absence of genetic monogamy adds a level of promiscuity to reproduction in prairie voles and may explain why female-biased philopatry could exist. When trying to understand the ultimate causes of characteristics that several species have in common, researchers should evaluate the proximate factors that lead to the ultimate causes in order to identify similarities between the species and thus better understand the ultimate causation (Brown and Balda 1977). My study will be the first experimental manipulation of initial density, a proximate factor, to test predictions of the habitat saturation hypothesis with a mammalian species. My empirical tests of the habitat saturation hypothesis should help to provide the strongest test to date of one of the factors potentially underlying philopatry in cooperative breeding mammals.

6 REFERENCES

Arnold KE, Owens IPF. 1998. Cooperative breeding in birds: a comparative test of the life history hypothesis. Proc R Soc B. 265:739-745. Bednarz JC, Ligon JD. 1988. A study of the ecological bases of cooperative breeding in the Harris’ Hawk. Ecology. 69:1176-1187. Boonstra R, Krebs CJ, Gaines MS, Johnson ML, Craine ITM. 1987. Natal philopatry and breeding systems in voles (Microtus spp.). J Anim Ecol. 56:655-673. Brown JL. 1969. Territorial behavior and population regulation in birds: a review and re-evaluation. Wilson Bull. 81:293-329. Brown JL, Balda RP. 1977. The relationship of habitat quality to group size in Hall’s babbler (Pomatostomus halli). Condor. 79:312-320. Caffrey C. 1992. Female-biased delayed dispersal and helping in American crows. Auk. 109:609-619. Clutton-Brock T. 2002. Breeding together: kin selection and mutualism in cooperative vertebrates. Science. 296:69-72. Clutton-Brock TH, Russell AF, Sharpe LL, Young AJ, Balmforth Z, McIlrath GM. 2002. Evolution and development of sex differences in cooperative behavior in meerkats. Science. 297:253-256. Cochran GR, Solomon NG. 2000. Effects of food supplementation on the social organization of prairie voles (Microtus ochrogaster). J Mammal. 81:746-757. Cowan D P. 1987. Group living in the European rabbit (Oryctolagus cuniculus): mutual benefit or resource localization? J Anim Ecol. 56:779-795. Dickinson JL, Hatchwell BJ. 2004. Fitness consequences of helping. In: Koenig WD, Dickinson JL. Ecology and Evolution of Cooperative Breeding in Birds. United Kingdom: Cambridge University Press. p. 48-66. Emlen ST. 1982. The evolution of helping. I. an ecological constraints model. Am Nat. 119:29-39. Emlen ST. 1994. Benefits, constraints and the evolution of family. Trends Ecol Evol. 9:282-285. Emlen ST. 1995. An evolutionary theory of the family. Proc Nat Acad Sci. 92:8092-8099. Gaines MS, Vivas AM, Baker CL. 1979. An experimental analysis of dispersal in fluctuating vole populations: demographic parameters. Ecology. 60:814:828. Getz LL, Hofmann JE, Carter CS. 1987. Mating system and population fluctuations of the prairie vole, Microtus ochrogaster. Amer Zool. 27:909-920.

7 Getz LL, McGuire B, Pizzuto T, Hofmann JE, Frase B.1993. Social organization of the prairie vole (Microtus ochrogaster). J Mammal. 74:44-58. Getz LL, Verner L, Cole FR, Hofmann JE, Avalos DE. 1979. Comparisons of population demography of Microtus ochrogaster and M. pennsylvanicus. Acta Theriol. 24:319-349. Goldizen AW, Terborgh J. 1989. Demography and dispersal patterns of a tamarin population: possible causes of delayed breeding. Am Nat. 134:208-224. Greenwood PJ. 1980. Mating systems, philopatry and dispersal in birds and mammals. Anim Behav. 28:1140-1162. Hatchwell BJ, Komdeur J. 2000. Ecological constraints, life history traits and the evolution of cooperative breeding. Anim Behav. 59:1079-1086. Jacquot JJ, Solomon NG. 2004. Experimental manipulation of territory occupancy: effects of immigration of female prairie voles. J Mammal. 85:1009-1014. Jarvis JUM, O’Riain MJ, Bennett NC, Sherman PW. 1994. Mammalian eusociality: a family affair. Trends Ecol Evol. 9:47-51. Jennions MD, Macdonald DW. 1994. Cooperative breeding in mammals. Trends Ecol Evol. 9:89-93. Koenig WD, Dickinson JL. 2004. Ecology and Evolution of Cooperative Breeding in Birds. United Kingdom: Cambridge University Press. Koenig WD, Pitelka FA. 1981. Ecological factors and kin selection in the evolution of cooperative breeding in birds. In: Alexander RD, Tinkle DW. Natural Selection and Social Behavior. New York, NY: Chiron Press Inc. p. 261-280. Koford RR, Bowen BS, Vehrencamp SL. 1986. Habitat saturation in groove-billed anis (Crotophaga sulcirostris). Amer Nat. 127:317-337. Kokko H, Ekman J. 2002. Delayed dispersal as a route to breeding: territorial inheritance, safe havens, and ecological constraints. Am Nat. 160:468-484. Komdeur J. 1992. Importance of habitat saturation and territory quality for evolution of cooperative breeding in the Seychelles warbler. Nature. 358:493-495. Lacey EA, Sherman PW. 1997. Cooperative breeding in naked mole-rats: implications for vertebrate and invertebrate sociality. In: Solomon NG, French JA. Cooperative Breeding in Mammals. New York: Cambridge University Press. p. 267-301. Lambin X. 1994. Natal philopatry, competition for resources, and inbreeding avoidance in Townsend’s voles (Microtus townsendii). Ecology. 75:224-235.

8 Ligon JD, Burt DB. 2004. Evolutionary origins. In: Koenig WD, Dickinson JL. Ecology and Evolution of Cooperative Breeding in Birds. United Kingdom: Cambridge University Press. p. 5-34 Lin YK, Batzli GO. 2001. The influence of habitat quality on dispersal and population densities of voles. Ecol. Monog. 71:245-275. Lin YK, Batzli GO. 2004.. Movement of voles across habitat boundaries: effects of food and cover. J Mammal. 85:216-224. Lin YK, Keane B, Isenhour A, Solomon NG. 2006. Effects of patch quality on dispersal and social organization of prairie voles: an experimental approach. J Mammal. 87:446-453. Marra PP, Holmes RT. 1997. Avian removal experiments: do they test for habitat saturation or female availability? Ecology. 78:947-952. McGuire B, Getz LL, Hofmann JE, Pizzuto T, Frase B. 1993. Natal dispersal and philopatry in prairie voles (Microtus ochrogaster) in relation to population density, season, and natal social environment. Behav Ecol Sociobiol. 32:293-302. Millar JS, Derrickson EM. 1992. Group nesting in Peromyscus maniculatus. J Mammal. 73:403-407. Moehlman PD, Hofer H. 1997. Cooperative breeding, reproductive suppression, and body mass in canids. In: Solomon NG, French JA. Cooperative Breeding in Mammals. New York: Cambridge University Press. p. 76-128. Mumme RL.1997. A bird's-eye view of mammalian cooperative breeding. In: Solomon NG, French JA. Cooperative Breeding in Mammals. New York, NY: Cambridge University Press. p. 364-388. Nutt KJ. 2005. Philopatry of both sexes leads to the formation of multimale, multifemale groups in Ctenodactylus gundi (Rodentia: Ctenodactylidae). J Mammal. 86:961-968. Powell RA, Fried JJ. 1992. Helping by juvenile pine voles (Microtus pinetorum), growth and survival of younger siblings, and the evolution of pine vole sociality. Behav Ecol. 3:325-333. Pruett-Jones SG, Lewis MJ. 1990. Sex ratio and habitat limitation promote delayed dispersal in superb fairy-wrens. Nature. 348:541-542. Queller DC. 1994. Extended parental care and the origin of eusociality. Proc R Soc B. 256:105-111. Rabenold KN. 1984. Cooperative enhancement of reproductive success in tropical wren societies. Ecology. 65:871-885. Randall JA, Rogovin K, Parker PG, Eimes JA. 2005. Flexible social structure of a desert rodent, Rhombomys opimus: philopatry, kinship, and ecological constraints. Behav Ecol. 16:961-973.

9 Riedman ML. 1982. The evolution of alloparental care and adoption in mammals and birds. The Q Rev Biol. 57:405-435. Roberts SC. 1987. Group-living and consortships in two populations of the European rabbit (Oryctolagus cuniculus). J Mammal. 68:28-38. Rood JP. 1990. Group size, survival, reproduction, and routes to breeding in dwarf mongooses. Anim Behav. 39:566-572. Russell AF. 2004. Mammals: comparisons and contrasts. In: Koenig WD, Dickinson JL. Ecology and Evolution of Cooperative Breeding in Birds. United Kingdom: Cambridge University Press. p. 210-227. Russell EM, Rowley I. 1993. Philopatry or dispersal: competition for territory vacancies in the splendid fairy-wren, Malurus splendens. Anim Behav. 45:519-539. Schradin C, Pillay N. 2004. The striped mouse (Rhabdomys pumilo) from the succulent karoo, South Africa: a territorial group-living solitary forager with communal breeding and helpers at the nest. J Comp Psychol. 118:37-47. Schradin C, Pillay N. 2005. Intraspecific variation in the spatial and social organization of the African striped mouse. J Mammal. 86:99-107. Selander RK. 1964. Speciation in wrens of the genus Campylorynchus. Univ Calif Publ Zool. 74:1-224. Solomon NG. 2003. A reexamination of factors influencing philopatry in rodents. J Mammal. 84:1182- 1197. Solomon NG, French JA. 1997. Cooperative Breeding in Mammals. United Kingdom: Cambridge University Press. Solomon NG, Keane B, Knoch LR, Hogan PJ. 2004. Multiple paternity in socially monogamous prairie voles (Microtus ochrogaster). Can J Zool. 82:1667-1671. Stacey PB, Ligon JD. 1987. Territory quality and dispersal options in the acorn woodpecker, and a challenge to the habitat-saturation model of cooperative breeding. Am Nat. 130:654-676. Stacey PB, Ligon JD. 1991. The benefits-of-philopatry hypothesis for the evolution of cooperative breeding: variation in territory quality and group size effects. Am Nat. 137:831-846. Taborsky M, Limberger D. 1981. Helpers in fish. Behav Ecol Sociobiol. 8:143-145. Walters JR, Copeyon CK, Carter JH III. 1992. Test of the ecological basis of cooperative breeding in red-cockaded woodpeckers. Auk. 109:90-97.

10 Wiley RH, Rabenold KN. 1984. The evolution of cooperative breeding by delayed reciprocity and queuing for favorable social positions. Evolution. 38:609-621. Wolff JO. 1992. Parents suppress reproduction and stimulate dispersal in opposite-sex juvenile white- footed mice. Nature. 359:409-410. Wolff JO. 1994. Reproductive success of solitarily and communally nesting white-footed mice and deer mice. Behav Ecol. 5:206-209.

11 CHAPTER 1: AN EVALUATION OF THE HABITAT SATURATION HYPOTHESIS

Philopatry, also referred to as delayed dispersal of offspring, is one of the key tenets of cooperative breeding because it can result in multiple adults living together in a group and contributing to the care of offspring (Koenig et al. 1992; Solomon 2003). In singular cooperative breeders, reproductive suppression typically occurs and only a single male-female pair reproduces within a group (Brown 1978). Cooperative breeding in these species appears to pose a challenge to a Darwinian explanation for the evolution of philopatry since philopatric individuals typically forgo reproduction while remaining in the natal group past sexual maturity (Brown 1987; Emlen 1982). Because cooperative breeding often arises from philopatry of offspring, understanding the basis of philopatry is the first step towards understanding cooperative breeding in many species (Emlen 1982; Gaston 1978; Lacey and Sherman 2007; Mumme 1997). Three categories of hypotheses have been most commonly proposed to explain philopatry. The benefits of philopatry hypotheses emphasize the benefits that accrue to nondispersers (Stacey and Ligon 1987, 1991). The life history hypothesis focuses on life history traits, like longevity, that increase recruitment relative to death rates (Arnold and Owens 1998; Hatchwell and Komdeur 2000). Lastly, the ecological constraints hypotheses stress that if there are constraints on dispersal, such as an absence of vacant breeding territories, the direct fitness cost of remaining philopatric may be small, especially compared to the costs associated with dispersal such as increased predation risk (Emlen 1982, 1995). These categories of hypotheses are not mutually exclusive, and the relative importance of each has been widely debated (Arnold and Owens 1998; Emlen 1994; Jennions and Macdonald 1994; Koenig et al. 1992; Solomon 2003; Stacey and Ligon 1991). Of the aforementioned hypotheses, the ecological constraints hypothesis is the only one that proposes that the decision to remain philopatric or disperse is primarily a function of some environmental limitation. The ecological constraints hypothesis focuses on four factors that could potentially determine whether an offspring will disperse and attempt to breed or remain philopatric within its natal group: (1) the cost/risk of dispersal, (2) the probability of successfully becoming established on a suitable territory, (3) the probability of obtaining a

12 mate, and (4) the likelihood of successful reproduction once an individual obtains a territory and a mate (Emlen 1982). To date, studies have been conducted to evaluate the influence of each of the four factors on philopatry, either individually or in combination (DuPlessis 1992; Jacquot and Solomon 2004; Komdeur 1992; Pruett-Jones and Lewis 1990; Stacey and Ligon 1987; Walters et al. 1992). However, habitat saturation, the unavailability of a resource critical for reproducing (e.g., territory, nest site), which affects the probability of successfully finding a suitable breeding territory, is the best studied of these factors (Emlen 1991). Although support for the habitat saturation hypothesis was recently found in a cooperatively breeding cichlid (Neolamprologus pulcher - Bergmüller et al. 2005), most support comes from studies using avian model systems (e.g., Baglione et al. 2005; Carrete et al. 2006; Doerr and Doerr 2006; Kinnaird and Grant 1982; Moreira 2006; Pruett-Jones and Lewis 1990; Walters et al. 1992; Yáber and Rabenold 2002). The most common evidence from avian studies cited as support for the predictions of the habitat saturation hypothesis comes from detecting a positive correlation between density and philopatry or group size (e.g., Carrete et al. 2006; Koford et al. 1986; Moreira 2006). However, while such a correlation is consistent with the predictions of the habitat saturation hypothesis, these studies cannot determine causation. For example, when observing population demographics of the group-territorial blue korhaan (Eupodotis caerulescens), Moreira (2006) found a positive correlation between population density and group size. While is it possible that group size in blue korhaan is increasing due to increasing philopatry in response to saturation of the habitat, the two factors were also both correlated with the distance from the core of the distribution range, where a high altitude plateau occurs. Therefore, the cause of both increasing density and group size could be some characteristic of the core of their range. Field manipulations of habitat availability in avian studies have provided much stronger tests of the habitat saturation hypothesis. For example, Walters et al. (1992) showed that red-cockaded woodpeckers (Picoides borealis) remain philopatric at high densities when nest cavities become limiting, but disperse when cavities become available by creating nest cavities. Likewise, when Seychelles warblers (Acrocephalus sechellensis) were transferred to an unoccupied island, philopatry was not observed until after all high-quality areas were occupied (Komdeur 1992). While avian studies have been critical in advancing our understanding of the factors that influence an individual’s decision about whether to remain philopatric or disperse, some

13 investigators have questioned the ability of researchers to extrapolate and apply results from avian systems to mammalian systems (Hayes 2000; Jennions and Macdonald 1994; Mumme 1997; Russell 2004). Viviparity and lactation in mammals skew the balance of parental care towards the female and prevent the male social partner and offspring from previous litter(s) from providing an energetic investment in the young similar to that provided by females (Mumme 1997). For this reason, it is important that we examine the factors leading to the development of philopatry in mammalian species more thoroughly. The cooperatively breeding prairie vole (Microtus ochrogaster) is an excellent mammalian model species to test predictions of the habitat saturation hypothesis for several reasons. Populations vary widely in density, with free-living populations ranging in density from 10 voles/ha to as many as 624 voles/ha (Getz and McGuire 1997; Getz et al. 1979; McGuire et al. 1990). Although there is some evidence that dispersal rates in prairie voles may be positively correlated with density (Gaines et al. 1979), most data indicates that dispersal in prairie voles is negatively correlated with population density (Cochran and Solomon 2000; Getz et al. 1993; Lin and Batzli 2001; McGuire et al. 1993). For example, McGuire et al. (1993) showed that dispersal rates were higher at low population density than at high population density, indicating a potential ecological constraint on dispersal. In addition, they reported that males and females dispersed with equal probability (see also Lin and Batzli 2004), contrary to previous findings that showed female-biased philopatry (Boonstra et al. 1987). Prairie voles are territorial and Getz et al. (1993) described three different types of social units for prairie voles in a free-living population in east-central Illinois: male-female pairs with or without offspring present, single females with or without offspring, and groups (at least two adults of the same sex) with or without offspring present. Individuals that are not residents of a single nest, called wanderers, also occurred within this population (Getz et al. 1993). All three types of prairie vole social units and wanderers have been observed within experimental enclosures as well (Cochran and Solomon 2000). Prairie vole populations exhibit plasticity in their intra-population social structure, with relative proportions of the three types of social units varying among studies (Cochran and Solomon 2000; Getz et al. 1993). Plasticity in the social structure is particularly important, since it could result from individuals responding directly to ecological conditions. Finally, prairie voles are easily maintained in outdoor enclosures (Desy and Batzli 1989) providing semi-

14 natural conditions for experimental manipulations, which cannot be performed with most mammalian cooperative breeders such as mongooses, canids and callitrichids. The goal of this study is to examine predictions of the habitat saturation hypothesis for philopatry and the formation of cooperatively breeding groups in prairie voles living in outdoor enclosures at different densities. According to the habitat saturation hypothesis, higher population densities should limit available territories and thus increase the proportion of philopatric offspring. Further, sex differences in philopatry may occur if males and females respond differently to changing ecological conditions. If philopatry is female biased, as found by Boonstra et al. (1987), I predict that males should be more affected than females by an increase in density, and that males should display a greater proportional increase in philopatry from low to high densities compared to females. Conversely, if males and females disperse at equal rates, as McGuire et al. (1993) found, I would expect to see no sex difference in the proportion of offspring remaining philopatric in response to high population densities. Finally, an increase in the proportion of philopatric offspring could increase average group size and/or increase the proportion of social units that are groups.

METHODS

Study area This study was conducted in two sets of small mammal enclosures from June through September 2004 and 2005 at the Miami University Ecology Research Center (ERC) in Oxford, OH (39º 30’N, 84º 44’W). Each set of enclosures was located in a different field (Fields 1 and 2) and the two fields were approximately 225 m apart. During each year, I used 4 of the 8 enclosures in Field 1, and 6 of the 8 enclosures in Field 2. The 0.1 ha enclosures at the ERC were constructed of 20-gauge galvanized steel panels, which extended 45 cm below the ground’s surface and 75 cm above the surface to prevent movement of voles among enclosures (Cochran and Solomon 2000). Vegetation in the enclosures consisted primarily of goldenrod (Solidago spp. L.), bluegrass (Poa pratensis L.), clover (Trifolium spp. L.), fescue (Festuca spp. L.), timothy (Phleum spp. L.), and ryegrass (Elymus spp. L.). Each enclosure had a one meter border, mowed to 5-10 cm in height, adjacent to the walls of each enclosure to help prevent voles from digging near the walls (Cochran and Solomon 2000). There was

15 an electric fence around the outside perimeter of each set of enclosures to prevent some mammalian predators (primarily raccoons, Procyon lotor) from entering enclosures and disturbing traps. Aside from the exclusion of species that would disturb traps, all other predators (e.g., raptors, snakes, shrews) had access to enclosures, helping to maintain semi- natural conditions for this study. I live-trapped in all enclosures for two weeks prior to the start of the study to remove all Microtus spp.

Experimental design

F1 generation prairie voles, which were descended from wild caught animals collected in east- central Illinois, were paired and bred in Miami University’s wild animal facilities each year between January and May. Animals were housed in polycarbonate cages (17 x 28 x 13 cm) containing processed paper bedding (Cell Sorb Plus, A & W Products, Inc., New Philadelphia, OH), dried alfalfa, and a cotton Nestlet (Ancare Corp., North Bellmore, NY) for nesting material. The animal room was maintained at 25 ± 3 ºC with a 14:10 h light/dark period (lights on at 0600). Rodent chow (Rodent Breeder Diet #5013, PMI Nutritional

International, Brentwood, MO) and water were provided ad libitum. All offspring (F2) were separated from their parents at 21 days of age and were toe clipped for identification, removing no more than one toe per foot.

In June of 2004 and 2005, F2 individuals were released into enclosures, which were established at two different initial densities, with the two years serving as temporal replicates. Four of the ten enclosures were established at a low density (4 founding males and 4 founding females per 0.1 ha enclosure), and the other six were established at a high density (12 founding males and 12 founding females per 0.1 ha enclosure). These densities were designated as low or high based on previous studies of natural populations, which looked at peak densities and sizes of home ranges of prairie voles (Desy et al. 1990; Getz et al. 1979). The low and high density designations are in accordance with McGuire and Getz (1998), who defined low density for prairie voles as less than 100 voles/ha, and high density as greater than 100 voles/ha. I considered a density of 80 voles/ha to be low because the average home range size for prairie voles is approximately 100 m2 and 4 pairs of voles should not occupy the entire 0.1 ha enclosure (Hofmann et al. 1984, Jike et al. 1988). Conversely, a starting

16 density of 240 voles/ha should lead to increased competition for space due to saturation of the habitat. Each year, the treatments (low or high density) were randomly assigned to the enclosures. Two enclosures of each density treatment were contained within each set of enclosures. Field 2 contained the two additional high density enclosures in both years. Individuals were randomly assigned to enclosures, although steps were taken to minimize the number of related individuals within each enclosure. When siblings were assigned to the same enclosure, they were never from the same litter and were always of the same sex to prevent inbreeding. None of the founding voles placed in the low density enclosures were half or full siblings but each of the high density enclosures (n = 12 enclosures) contained several (5-10) pairs or trios of full siblings among the founders. The founding voles were between the ages of 35 and 117 days at the time of release. Since the minimum age for breeding under laboratory conditions is 31 days of age (Solomon 1991), all founders should have been old enough to initiate breeding following release. Due to early mortality of founding individuals or escape of animals through previously undetected holes in enclosure walls, an additional release took place in five enclosures in 2004 after the first week of grid trapping to maintain starting densities (35 replacement individuals released, approximately 85% of replacements due to disappearance of individuals). Selection of replacement individuals followed the same protocol as selection of founders to limit relatedness and inbreeding within the enclosures. Populations were monitored by live-trapping using Ugglan multiple-capture traps (Grahnab, Sweden) baited with cracked corn, a low quality food item (Desy and Batzli 1989). Cotton batting was added to the traps when temperatures were predicted to be below 10°C (Cochran and Solomon 2000). For each individual captured, I recorded the location of capture, ID, sex, reproductive status (males – scrotal or non-scrotal; females – pregnant, lactating, formerly lactating, or non-reproductive), age class, any injuries or other unique characteristics, and individuals with which they were captured. Offspring were toe-clipped for identification at their first capture, removing no more than one toe per foot until all possible number combinations were exhausted, at which point some individuals had two toes removed from one of their feet. Additionally, all individuals were weighed to the nearest gram using a spring scale (Pesola micro-line spring scale, Forestry Suppliers Inc, Jackson,

17 MS). During the first year, founding adults were weighed once per week. However, during the second year adult males were weighed weekly, while adult females were weighed every day that they were captured to help track pregnancy and births. Both years, offspring were weighed every time they were captured until they reached adulthood (defined by body mass) and then were weighed as described for founding adult males and females. Age classes used were juveniles (less than 20g), subadults (20g to 29g), and adults (30g and greater) as defined by Getz et al. (1993). Live-trapping during each temporal replicate (2004 and 2005) lasted 14 weeks, beginning 1.5 weeks after the initial release of animals. During those 14 weeks, I used a combination of grid and nest trapping to monitor populations (see Figure 1). Grid trapping was conducted to monitor population density and to trap founding females in order to locate their nests. Nest trapping functioned to identify nest residents and capture offspring when they emerged. During grid trapping, traps were located in a 5x5 grid pattern with 5 m spacing between traps (25 traps per enclosure). Five trap checks occurred during every grid trapping week, with traps open from 2000 on Sunday until the 0700 trap check on Monday. Traps were reset at 1800 on Monday and Tuesday nights with trap checks taking place at 2100 that night and 0700 the following morning (Tuesday and Wednesday mornings). Ultraviolet reflective powder was used in combination with radio tracking to locate the nests of all founding females beginning the first week of trapping. I powdered females with uniquely colored UV reflective powder (Radiant Color, Richmond, California) during the morning trap check, released them at the point of capture immediately following powdering. The powder trails of females were followed after dark using an ultraviolet lamp (ML-49, UVP, Inc., San Gabriel, CA) to locate females’ nests (as described by Lemen and Freeman 1985). Females to be radio tracked were also collected during the morning trap check, anesthetized using Isoflurane (Phoenix Pharmaceutical, Inc., St. Joseph, MO), and radio collared with a small mammal radio collar (model PD-2C, Holohil Systems Ltd., Ontario, Canada). The radiocollars weighed approximately 3 g and were always less than 10% of a collared female’s body mass so as not to influence their daily energy expenditure (Berteaux et al. 1996). Females were given several hours to recover from anesthesia in a temperature controlled animal room at the ERC, and then each vole was re-released at her site of capture. The locations of collared individuals were identified via triangulation of transmitter signals

18 with two receivers (Fieldmaster 16 channel receiver, Johnson’s Telemetry, El Dorado Springs, MO) and 3 element Yagi antennas (Johnson’s Telemetry, El Dorado Spring, MO) between 1100 and 1500. Fixes were taken at least three times per day, at 30-60 minute intervals, for three consecutive days. I assumed that during this time period, females were likely to be in their nests to avoid the heat. I searched areas to which individuals were tracked for nests or nest entrances. Once nests were located by either powder tracking or radio tracking, three Ugglan traps were placed in surface runways less than 1 m from the entrance to nests. After three consecutive weeks of grid trapping, nest trapping and grid trapping were alternated on a weekly basis for eight weeks (Figure 1). Each nest trapping week had 10 trap checks, divided into two periods, each with five trap checks. During the first period, traps were set at 1800 and checked at 2100 Sunday, Monday, and Tuesday as well as at 0700 on Monday and Tuesday. This schedule was then repeated Wednesday night through Friday night. I conducted only nest trapping during the final three weeks of the study (trapping weeks 12-14; residency period) to determine the individuals residing at each nest, the composition of social units, and the proportion of offspring that remained philopatric. At the end of each field season, all surviving voles were live-trapped and removed from enclosures. One high density enclosure from 2004 and two high density enclosures from 2005 were removed from all analyses due to predation by raccoons in 2004 and a long-tailed weasel (Mustela frenata) in 2005. All research procedures involving live animals followed the guidelines of the American Society of Mammalogists for the use of wild animals in research (Gannon et al. 2007) and were approved by the Miami University Institutional Animal Care and Use Committee.

Vegetation sampling Vegetation sampling was conducted in June of each year in all enclosures to determine if there were differences between the two sets of enclosures (Fields 1 and 2), between years, or between treatments. Square quadrats (0.36 m2) were randomly sampled within each 5 m x 5 m section of the grid (n = 36 sections/enclosure). Within each quadrat, the proportion of vegetation from class Liliopsida (monocots) and class Magnoliopsida (dicots) were estimated. The proportion of monocots and dicots provided an index of habitat quality. Dicots were

19 indicative of high quality habitat since they comprise the majority of summer diet for prairie voles living in the prairie, as compared to blue grass or alfalfa fields (Batzli 1985). Because the proportions of monocots and dicots were based only on the area within a quadrat that had vegetative cover, the total percentage of monocots and dicots in each quadrat always equaled 100%. I also measured the maximum height of vegetation within each quadrat to evaluate the amount of cover. Variation in the amount of cover between enclosures could influence the fitness of individuals within those enclosures because prairie vole survival and reproduction is positively correlated with the amount of cover available (Lin and Batzli 2001).

Demography To avoid the potential bias generated by using the minimum number known alive to calculate population density (Pocock et al. 2004; Slade and Blair 2000), I used a robust design model (Huggins closed capture estimator – see Huggins 1989 and Huggins 1991) within the computer program MARK (White and Burnham 1999) to estimate population density based on data from the seven grid trapping weeks. The Huggins closed capture estimator was chosen because it is more stable when sample sizes are small (Kendall in litt.). The derived population estimate for each enclosure from week 11 was used as the density estimate for all analyses of the proportion of philopatric animals and social structure since week 11 is the final grid week before the start of the three week residency period. At the beginning of the study, enclosures could be categorically divided into those with low or high densities. Later in the study, however, the population densities within treatments had increased more quickly in some enclosures than in others, and population density estimates for some low density enclosures overlapped estimates for some of the high density enclosures. Due to this gradient of densities and loss of distinct categorical division, density was used as a continuous variable in all analyses. Trappability (the probability of capturing an individual at least once during a grid trapping week) and survival of all individuals (founders and offspring) between grid trapping weeks were also calculated for each enclosure based on data from the grid trapping weeks using program MARK (Kendall 2001). Further, offspring survival was also calculated as the proportion of offspring that survived from first capture to the three week residency period. Only offspring that could have potentially reached sexually maturity, 30 days of age as

20 estimated from date of first capture and body mass, by the three week residency period were included in this analysis.

Social structure and philopatry For subadults and adults, residency at a nest site was defined as having ≥ 75% of all captures at one nest site for all individuals captured at least once per week (Cochran and Solomon 2000) during the three week residency period (weeks 12-14). Individuals that were captured at least once per week but did not have ≥ 75% of their captures at one nest site were termed wanderers (Cochran and Solomon 2000). Individuals with less than one capture per week were not classified and were excluded from analyses of social structure. Composition of social units was determined based on the number of adults and subadults of each sex that were residents at the same nest site. Social units were defined as single male or single female with or without offspring, male-female pair with or without offspring, or group, i.e., at least two adults or subadults of the same sex, with or without offspring. Group size was equal to the number of adult and subadult residents at a nest. One low density enclosure in 2004 had no groups, so analysis of mean group size within enclosures is based on 8 enclosures for 2004 and 8 enclosures for 2005. The proportion of groups in each enclosure was determined by dividing the number of social units that were groups by the total number of social units within the enclosure. The natal nest of each offspring was the nest at which it was first captured as a juvenile, since prairie voles typically do not leave the natal nest prior to reaching 20 g (McGuire et al. 1993). Any offspring with a body mass greater than or equal to 20 g and classified as a resident of its natal nest during the three week residency period was considered philopatric. Conversely, all offspring with body masses greater than or equal to 20 g and classified as residents of a nest other than their natal nest or that were classified as wanderers during the three week residency period were considered dispersers. Furthermore, only offspring that were at least 30 days old, the age of sexual maturity, during the three week residency period were included in the philopatry analysis (McGuire et al. 1993). Of 17 enclosures included in the data analysis, only 15 contained individuals that met the requirements for inclusion in the analysis of philopatry.

21 Statistical analysis Vegetative data were statistically compared using a three-way ANOVA to determine if there was an effect of initial population density, enclosure location (Fields 1 and 2), and year. Bonferroni post-hoc analyses using least squares means were used to investigate significant interactions. Based on the results of the vegetation analysis, year, enclosure location (Fields 1 and 2) and its interaction were included in all the models to account for any variation in the analyses due to these factors. Analysis of covariance (ANCOVA) was used to determine if population density predicted differences in group size. Group sizes within an enclosure were averaged prior to analysis because groups within enclosures were not independent. Logistic regression was used to determine if the probability that social units were groups was predicted by density. Logistic regression was also used to determine if population density predicted the proportion of surviving offspring that were philopatric. Sex was also used as a factor in the regression model for the analysis of philopatry to see if the proportion of offspring remaining philopatric differed between females and males. Pearson’s correlation analysis was used to determine relationships between group size, the proportion of social units that were groups and the proportion of offspring remaining philopatric. For all regression analyses, only statistically significant differences are presented since non-significant interactions and main effects were removed from the models through backwards elimination for parsimony. All means are presented ± standard error. The ANOVA and ANCOVAs were checked for normality and constant variance and data presented met the assumptions of the analyses used. Statistical significance was set at p < 0.05.

RESULTS

Vegetation sampling

There were significant main effects of year (F1,605 = 12.44, p = 0.0005) and enclosure location

(F1,605 = 704.03, p < 0.0001) on the percentage of dicots. The overall percentage of dicots was significantly less in 2004 than in 2005 (Table 1a). Furthermore, the percentage of dicots in Field 2 was approximately 16 time greater than the percentage of dicots in Field 1 (Table

22 1b). There was a significant year x density interaction with regard to the maximum height of vegetation (F1,605 = 9.57, p = 0.0021). Although vegetation height was not significantly different between low and high density enclosures in 2004, vegetation in the low density enclosures in 2005 was significantly taller than the vegetation in the high density enclosures (Table 1c). In addition, there was a significant year x enclosure location interaction in the prediction of the maximum height of vegetation (F1,605 = 24.77, p < 0.0001). In 2004, the maximum height of the vegetation in Field 2 was significantly lower than the maximum height of the vegetation in Field 1 (Table 1d). The maximum height of the vegetation did not significantly differ between Fields 1 and 2 in 2005 (Table 1d).

Demography Although initial densities were either 8 or 24 voles/0.1 ha, by the final week of grid trapping (week 11), density ranged from 6 to 77 voles/0.1 ha. The trappability of voles was high throughout the experiment, averaging 80% ± 2% for the 17 enclosures. Survival of all individuals from one grid trapping week to the next was also high (95% ± 1%). Offspring survival, based on offspring that could have been at least 30 days old during the residency period (trapping weeks 12-14), was estimated to be 65% ± 6%. Of the 579 subadults and adults in the enclosures during the residency period for the two years of the study, 354 (61%) were classified as residents, 106 (18%) were classified as wanderers, and 119 (21%) could not be classified because they did not have at least one capture per week.

Social structure and philopatry Groups ranged in size from 2 to 19 voles, with an overall group size of 4.2 ± 0.4. Density had significantly different effects on group size between years (density x year, F1,12 = 6.58, p = 0.0248). Although there was a positive relationship between group size and density in both years, group size showed a greater response to increasing density in 2005 than in 2004 (Figure 2). Enclosure location was not a significant predictor of group size. On average, groups contained 1.4 ± 0.1 adult females, 1.3 ± 0.2 adult males, 0.7 ± 0.2 subadult females, and 0.8 ± 0.1 subadult males. The mean proportion of adult and subadult females in groups was 0.51 ± 0.05, indicating that groups tended to have an approximately equal sex ratio, with 84% of the 63 groups having at least one male and one female. When

23 looking at only mature offspring within groups, the mean proportion of females was 0.50 ± 0.08, similar to the overall sex ratio within groups. Most groups (86%) had at least one founding individual within the group, and 40% of the groups contained both a male and female founder. Each enclosure contained 5.6 ± 0.6 social units with a range from 3 to 9 social units (Table 2). Density and year were both significant predictors of the proportion of social units that were groups within an enclosure (χ2 = 10.16, df = 1, p = 0.0014 and χ2 = 7.14, df = 1, p = 0.0075, respectively). The proportion of groups increased as density increased in both years, but at a given density the proportion of groups was greater in 2005 than in 2004 (Figure 3). Enclosure location was not a significant predictor of the proportion of social units that were groups and there were no statistically significant interactions. Of density, year and enclosure location, density was the only predictor that affected the proportion of offspring that remained philopatric (χ2 = 15.77, df = 1, p < 0.0001), with the proportion of philopatric individuals increasing as density increased (Figure 4). There were no statistically significant interactions among factors. Sex was a significant predictor of the proportion of offspring remaining philopatric (χ2 = 4.86, df = 1, p = 0.0275), with females being more philopatric than males at all densities (Figure 5). However, density had significantly different effects on the proportion of offspring remaining philopatric between years when sex was included in the model (density x year, χ2 = 5.07, df = 1, p = 0.0244), with individuals from 2005 being less philopatric at low densities and more philopatric at high densities than individuals from 2004. Enclosure location was not a significant predictor of philopatry when analyzed by sex. Group size was significantly correlated with the proportion of social units that were groups (Pearson’s correlation: r = 0.6312, p = 0.0087, n = 16) and the proportion of offspring remaining philopatric (Pearson’s correlation: r = 0.5432, p = 0.0447, n = 14). The proportion of social units that were groups was significantly correlated with the proportion of offspring remaining philopatric (Pearson’s correlation: r = 0.6650, p = 0.0095, n = 14).

DISCUSSION

24 The results of this experiment provide strong evidence for a causal effect of population density on natal dispersal and are consistent with the predictions of the habitat saturation hypothesis. The predictions from the habitat saturation hypothesis are based on the assumption that as population density increases the amount of available habitat decreases. Although I did not directly measure habitat use, because prairie voles are territorial, I assumed the amount of space available should decrease at higher densities. This assumption is supported by the finding that within the experimental enclosures used in my study, the home range size of prairie voles decreased at high densities, with a mean home range size of 45 ± 16 m2 in the lowest density enclosure and 19 ± 4 m2 in the highest density enclosure when calculated using the mean squared distance from the center of activity (AR Richmond, personal communication). The strongest support for the habitat saturation hypothesis is that increasing the population density of prairie voles in experimental enclosures increased the proportion of offspring remaining philopatric. The retention of offspring at the natal nests led to an increase in the proportion of social units that are classified as groups, as well as an increase in group size. Taken together, these results suggest that as habitat became occupied due to the increase in population densities, mature offspring were more likely to remain at the natal nest as opposed to dispersing and establishing their own territory or wandering within the population (sensu Getz et al. 1993; Solomon and Jacquot 2002). The results from this study are consistent with data from other studies of prairie voles. McGuire et al. (1993) found that philopatry was positively correlated with population density in unenclosed populations in Illinois. Additionally, although they found no difference in social structure between populations of prairie voles with and without supplemental food, Cochran and Solomon (2000) noted a correlation between increasing population density and an increase in the number of groups as well as an increase in group size during their experiment. Getz and colleagues also found a correlation between population density and group formation in prairie voles (Getz et al. 1990; Getz et al. 1993; Getz and McGuire 1997; Getz et al. 2005). They concluded that group formation and the increase in population density were ultimately due to increased survival of juveniles in late autumn, when snakes, a predator of juvenile prairie voles, entered hibernation. However, the three week residency period during which I evaluated social structure, was in September (early autumn), when juvenile mortality should still be high due to snake predation. Further, any effects that may be caused

25 by seasonal trends should be negated by the concurrent study of high and low density enclosures as opposed to Getz and colleagues’ studies where density changed with time and could not be easily isolated from seasonal changes. Therefore, I would argue that my data support a causal relationship between density and changes in social structure, as opposed to differential offspring survival leading to an increase in group formation and higher densities. In my study, philopatry in both males and females was affected by increased population density. Philopatry was female biased independent of density, however, at high population densities, both males and females are more likely to remain philopatric than they were at low densities. I had predicted that if philopatry was female biased, habitat saturation should result in a stronger effect on males than females at high population densities, but this prediction was not supported, as evidenced by the lack of a sex x density interaction. While evolutionary theory predicts that monogamous species will not have sex biased philopatry (Greenwood 1980) and prairie voles are socially monogamous, they are not genetically monogamous (Solomon et al. 2004). Therefore, female-biased philopatry in prairie voles is consistent with evolutionary theory for non-monogamous mammal systems (Greenwood 1980; Handley and Perrin 2007). However, female-biased philopatry in prairie voles diverges from the tendency for no-sex biased philopatry in singular breeding rodents (Lacey and Sherman 2007; Russell 2004). Habitat saturation has been evaluated as a possible reason for philopatry in several cooperatively breeding mammals. While most studies find evidence supporting habitat saturation as a potential cause of philopatry in mammalian species (e.g., Goldizen and Terborg 1989; Jacquot and Solomon 2004; Jannett 1978; Macdonald and Moehlman 1982; Randall et al. 2005; Schradin and Pillay 2005; Wolff 1992, 1994), the data from a few contradict the predictions of the habitat saturation hypothesis (e.g., Jennions and Macdonald 1994; Millar and Derrickson 1992). The strongest data supporting the habitat saturation hypothesis comes from field studies that remove individuals from a territory. For example, Jacquot and Solomon (2004) concurrently tested the mate limitation hypothesis and habitat limitation hypothesis by removing either all adults or all adult females from a group. They compared the response of remaining voles to vacancies with the control treatments, where a group was left intact and no vacancies were created. They found that immigration into vacancies occurred regardless of whether or not a potential mate was present, indicating that

26 availability of a suitable territory was a more important factor limiting dispersal than mate availability. Similarly, Wolff (1992) removed adults from a population of white-footed mice (Peromyscus leucopus) and compared movement patterns of remaining individuals to the patterns of individuals in control populations. He found that more juvenile males dispersed when adults were removed than in control populations, indicating that philopatry at high population densities may be due to habitat saturation. Likewise, when breeding vacancies were created in woodland vole (Microtus pinetorum) populations by the removal of the breeding female from groups, the vacancies were rapidly filled and the new females became reproductive (Solomon et al. 1998). The other evidence supporting the predictions of the habitat saturation hypothesis comes from studies where a correlation between philopatry and density is found, but a cause and effect is not demonstrated. For example, Jannett (1978) found that while female montane voles (Microtus montanus nanus) in low density conditions will abandon their pups when they are around 15 days old, when population density is high, the female remains with or very close to her pups and extended maternal family groups form. My results go beyond the evidence presented by the mammalian correlative and removal studies by demonstrating that high population densities not only increase philopatry but additionally cause an increase in group formation and the size of groups. These results provide the strongest test to date of the habitat saturation hypothesis in a cooperatively breeding mammalian species and show it to be a significant factor influencing philopatry in prairie voles. Although the results of this study support the habitat saturation hypothesis, the results do not eliminate the possibility that other factors contribute to philopatry in prairie voles, as predicted by the delayed-dispersal threshold model which proposes that a combination of benefits and constraints may influence philopatry in cooperatively breeding species (Koenig et al. 1992). Since voles in the lower density enclosures should not have experienced ecological constraints on dispersal, the presence of philopatric individuals within these enclosures suggests that other factors are may also be influencing the decision of prairie voles to delay dispersal. Aside from constraints on dispersal imposed by density, philopatric prairie voles may also accrue benefits by remaining at the natal nest. Benefits of philopatry for prairie voles may include direct fitness benefits such as such as improved survival (Emlen 1995; Kokko and Ekman 2002), increased body mass prior to reproduction which has been

27 shown to increase pup weight gain in prairie voles (Hayes and Solomon 2004; Solomon 1991), thus improving the quality of offspring produced, and the experience of raising young, which in the case of female prairie voles led them to exhibit greater parental investment in their pups and thus produce pups with higher survival rates that develop more rapidly and are larger at weaning (Wang and Novak 1994). Indirect benefits of philopatry also may occur if philopatric individuals help close relatives survive and reproduce, since prairie vole pups reared with juveniles at the nest develop faster and weigh more at weaning than those reared without helpers present (Solomon 1991). Prairie voles would not be the first mammalian species to have philopatry decisions potentially influenced by both habitat saturation and benefits of philopatry. Striped mouse populations in the succulent karoo are at significantly higher densities than populations in the mesic grasslands, suggesting that habitat saturation influences philopatry (Schradin and Pillay 2004, 2005). Philopatric striped mice might also benefit from predator avoidance through increased vigilance of group members and are hypothesized to gain thermoregulatory benefits from nest sharing (Schradin and Pillay 2004). Habitat saturation has been suggested as a factor influencing philopatry in both cooperatively breeding birds and mammals. However, most of the evidence supporting the habitat saturation hypothesis, particularly in mammals, has been relatively weak (Russell 2004), because they do not directly demonstrate that variations in density cause changes in philopatry and social structure. Through the manipulation of population density, the results presented here provide the strongest support of the habitat saturation hypothesis to date by showing a causal effect of density on philopatry, group size and the proportion of groups. Therefore, ecological constraints, at least partially, cause philopatry in prairie voles. Furthermore, these results show that changes in ecological conditions, such as increasing population density, can influence the social structure of a species, leading to group formation and cooperative breeding. Philopatry is likely to be the result of both constraints due to habitat saturation and benefits of philopatry. Future analyses of parentage and relatedness should allow for the simultaneous evaluation of the benefits of philopatry and ecological constraints, thus furthering the understanding of factors leading to philopatry in cooperatively breeding mammals.

28 REFERENCES

Arnold KE, Owens IPF. 1998. Cooperative breeding in birds: a comparative test of the life history hypothesis. Proc R Soc B. 265:739-745. Baglione V, Marcos JM, Canestrari D, Griesser M, Andreotti G, Bardini C, Bogliani G. 2005. Does year-round territoriality rather than habitat saturation explain delayed natal dispersal and cooperative breeding in the carrion crow? J Anim Ecol. 74:842-851. Batzli GO.1985. Nutrition. In: Tamarin RH. Biology of New World Microtus. Shippensburg, PA: Special Publication No. 8, The American Society of Mammalogists. p. 779-811. Bergmüller R, Heg D, Taborsky M. 2005. Helpers in a cooperatively breeding cichlid stay and pay or disperse and breed, depending on ecological constraints. Proc R Soc B. 272:325-331. Berteaux D, Masseboeuf F, Bonzom J, Bergeron J, Thomas DW, Lapierre H. 1996. Effect of carrying a radiocollar on expenditure of energy by meadow voles. J Mammal. 77:359- 363. Boonstra R, Krebs CJ, Gaines MS, Johnson ML, Craine ITM. 1987. Natal philopatry and breeding systems in voles (Microtus spp.). J Anim Ecol. 56:655-673. Brown JL. 1978. Avian communal breeding systems. Ann Rev Ecol Syst. 9:123-55. Brown JL. 1987. Helping and Communal Breeding in Birds. Princeton, NJ: Princeton University Press. Carrete M, Donázar JA, Margalida A, Bertran J. 2006. Linking ecology, behavior and conservation: does habitat saturation change the mating system of bearded vultures? Biol Lett. 2:624-627. Clutton-Brock TH, Russell AF, Sharpe LL, Young AJ, Balmforth Z, McIlrath GM. 2002. Evolution and development of sex differences in cooperative behavior in meerkats. Science. 297:253-256. Cochran GR, Solomon NG. 2000. Effects of food supplementation on the social organization of prairie voles (Microtus ochrogaster). J Mammal. 81:746-757. Desy EA, Batzli GO. 1989. Effects of food availability and predation on prairie vole demography: a field experiment. Ecology. 70:411-421. Desy EA, Batzli GO, Liu J. 1990. Effects of food and predation on behavior of prairie voles: a field experiment. Oikos. 58:159-168. Doerr ED, Doerr VAJ. 2006. Comparative demography of treecreepers: evaluating hypotheses for the evolution and maintenance of cooperative breeding. Anim Behav. 72:147-159.

29 DuPlessis MA. 1992. Obligate cavity-roosting as a constraint on dispersal of green woodhoopoes: consequences for philopatry and the likelihood of inbreeding. Oecologia. 90:205-211. Emlen ST. 1982. The evolution of helping. I. an ecological constraints model. Am Nat. 119:29-39. Emlen ST. 1991. Evolution of cooperative breeding in birds and mammals. In: Krebs JR, Davies NB. Behavioural Ecology: An Evolutionary Approach. 3rd ed. Cambridge, MA: Blackwell Scientific Publications. p. 301-337. Emlen ST. 1994. Benefits, constraints and the evolution of family. Trends Ecol Evol. 9:282-285. Emlen ST. 1995. An evolutionary theory of the family. Proc Nat Acad Sci. 92:8092-8099. Gaines MS, Vivas AM, Baker CL. 1979. An experimental analysis of dispersal in fluctuating vole populations: demographic parameters. Ecology. 60:814:828. Gannon WL, Sikes RS, The Animal Care and Use Committee of the American Society of Mammalogists. 2007. Guidelines of the American Society of Mammalogists for the use of wild mammals in research. J Mammal. 88:809-823. Gaston AJ. 1978. The evolution of group territorial behavior and cooperative breeding. Am Nat. 112:1091-1100. Getz LL, McGuire B. 1997. Communal nesting in prairie voles (Microtus ochrogaster): formation, composition, and persistence of communal groups. Can J Zool. 75:525-534. Getz LL, McGuire B, Carter CS. 2005. Social organization and mating system of free-living prairie voles Microtus ochrogaster: a review. Acta Zool. 51:178-186. Getz LL, McGuire B, Pizzuto T, Hofmann JE, Frase B.1993. Social organization of the prairie vole (Microtus ochrogaster). J Mammal. 74:44-58. Getz LL, Solomon NG, Pizzuto TM. 1990. The effects of predation of snakes on social organization of the prairie vole, Microtus ochrogaster. Am Midl Nat. 123:365-371. Getz LL, Verner L, Cole FR, Hofmann JE, Avalos DE. 1979. Comparisons of population demography of Microtus ochrogaster and M. pennsylvanicus. Acta Theriol. 24:319-349. Goldizen AW, Terborgh J. 1989. Demography and dispersal patterns of a tamarin population: possible causes of delayed breeding. Am Nat. 134:208-224. Greenwood PJ. 1980. Mating systems, philopatry and dispersal in birds and mammals. Anim Behav. 28:1140-1162. Handley LJL, Perrin N. 2007. Advances in our understanding of mammalian sex-biased dispersal. Mol Ecol. 16:1559-1578.

30 Hatchwell BJ, Komdeur J. 2000. Ecological constraints, life history traits and the evolution of cooperative breeding. Anim Behav. 59:1079-1086. Hayes LD. 2000. To nest communally or not to nest communally: a review of rodent communal nesting and nursing. Anim Behav. 59:677-688. Hayes LD, Solomon NG. 2004. Costs and benefits of communal rearing to female prairie voles (Microtus ochrogaster). Behav Ecol Sociobiol. 56:585-593. Hofmann JE, Getz LL, Gavish L. 1984. Home range overlap and nest cohabitation of male and female prairie voles. Am Midl Nat. 112:314-319. Huggins RM. 1989. On the statistical analysis of capture experiments. Biometrika. 76:133.140. Huggins RM. 1991. Some practical aspects of a conditional likelihood approach to capture experiments. Biometrics. 47:725-732. Jacquot JJ, Solomon NG. 2004. Experimental manipulation of territory occupancy: effects of immigration of female prairie voles. J Mammal. 85:1009-1014. Jannett FJ Jr. 1978. The density-dependent formation of extended maternal families of the montane vole, Microtus montanus nanus. Behav Ecol Sociobiol. 3:245-263. Jennions MD, Macdonald DW. 1994. Cooperative breeding in mammals. Trends Ecol Evol. 9:89-93. Jike L, Batzli GO, Getz LL. 1988. Home ranges of prairie voles as determined by radiotracking and by powdertracking. J Mammal. 69:183-186. Kendall WL. 2001. The robust design for capture-recapture studies: analysis using program MARK. In: Warren RJ, Okarma H, Sievert PR. Wildlife, land, and people: priorities for the 21st century. Bethesda, MD: The Wildlife Society. p. 357-360. Kinnaird MF, Grant PR. 1982. Cooperative breeding by the Galápagos mockingbird, Nesomimus parvulus. Behav Ecol Sociobiol. 10:65-73. Koenig WD, Pitelka FA, Carmen WJ, Mumme RL, Stanback MT. 1992. The evolution of delayed dispersal in cooperative breeders. Q Rev Biol. 67:111-150. Koford RR, Bowen BS, Vehrencamp SL. 1986. Habitat saturation in groove-billed anis (Crotophaga sulcirostris). Amer Nat. 127:317-337. Kokko H, Ekman J. 2002. Delayed dispersal as a route to breeding: territorial inheritance, safe havens, and ecological constraints. Am Nat. 160:468-484. Komdeur J. 1992. Importance of habitat saturation and territory quality for evolution of cooperative breeding in the Seychelles warbler. Nature. 358:493-495.

31 Lacey EA, Sherman PW. 2007. The ecology of sociality in rodents. In: Wolff JO, Sherman PW. Rodent societies: an ecological and evolutionary perspective. Chicago, IL: The University of Chicago Press. p. 243-254. Lemen CA, Freeman PW. 1985. Tracking mammals with fluorescent pigments: a new technique. J Mammal. 66:134-136. Lin YK, Batzli GO. 2001. The influence of habitat quality on dispersal and population densities of voles. Ecol. Monog. 71:245-275. Lin YK, Batzli GO. 2004. Movements of voles across habitat boundaries: effects of food and cover. J. Mammal. 85:216-224. Macdonald DW, Moehlman PD. 1982. Cooperation, altruism, and restraint in the reproduction of carnivores. Persp Ethol. 5:433-67. McGuire B, Getz LL. 1998. The nature and frequency of social interactions among free-living prairie voles (Microtus ochrogaster). Behav Ecol Sociobiol. 43:271-279. McGuire B, Getz LL, Hofmann JE, Pizzuto T, Frase B. 1993. Natal dispersal and philopatry in prairie voles (Microtus ochrogaster) in relation to population density, season, and natal social environment. Behav Ecol Sociobiol. 32:293-302. McGuire B, Pizzuto T, Getz LL. 1990. Potential for social interaction in a natural population of prairie voles (Microtus ochrogaster). Can J Zool. 68:391-398. Millar JS, Derrickson EM. 1992. Group nesting in Peromyscus maniculatus. J Mammal. 73:403-407. Moreira F. 2006. Group size and composition are correlated with population density in the group- territorial blue korhaan (Eupodotis caerulescens). Afr J Ecol. 44:444-451. Mumme RL.1997. A bird's-eye view of mammalian cooperative breeding. In: Solomon NG, French JA. Cooperative Breeding in Mammals. New York, NY: Cambridge University Press. p. 364-388. Pocock MJO, Frantz AC, Cowan DP, White PCL, Searle JB. 2004. Tapering bias inherent in minimum number alive (MNA) population indices. J Mammal. 85:959-962. Pruett-Jones SG, Lewis MJ. 1990. Sex ratio and habitat limitation promote delayed dispersal in superb fairy-wrens. Nature. 348:541-542. Randall JA, Rogovin K, Parker PG, Eimes JA. 2005. Flexible social structure of a desert rodent, Rhombomys opimus: philopatry, kinship, and ecological constraints. Behav Ecol. 16:961-973.

32 Russell AF. 2004. Mammals: comparisons and contrasts. In: Koenig WD, Dickinson JL. Ecology and Evolution of Cooperative Breeding in Birds. United Kingdom: Cambridge University Press. p. 210-227. Schradin C, Pillay N. 2004. The striped mouse (Rhabdomys pumilo) from the succulent karoo, South Africa: a territorial group-living solitary forager with communal breeding and helpers at the nest. J Comp Psychol. 118:37-47. Schradin C, Pillay N. 2005. Intraspecific variation in the spatial and social organization of the African striped mouse. J Mammal. 86:99-107. Slade NA, Blair SM. 2000. An empirical test of using counts of individuals captured as indices of population size. J Mammal. 81:1035-1045. Solomon NG. 1991. Current indirect fitness benefits associated with philopatry in juvenile prairie voles. Behav Ecol Sociobiol. 29:277-282. Solomon NG. 2003. A reexamination of factors influencing philopatry in rodents. J Mammal. 84:1182- 1197. Solomon NG, Jacquot JJ. 2002. Characteristics of resident and wandering prairie voles, Microtus ochrogaster. Can J Zool. 80:951-955. Solomon NG, Keane B, Knoch LR, Hogan PJ. 2004. Multiple paternity in socially monogamous prairie voles (Microtus ochrogaster). Can J Zool. 82:1667-1671. Solomon NG, Vandenbergh JG, Sullivan WT. 1998. Social influences on intergroup transfer by pine voles (Microtus pinetorum). Can J Zool. 76:2131-2136. Stacey PB, Ligon JD. 1987. Territory quality and dispersal options in the acorn woodpecker, and a challenge to the habitat-saturation model of cooperative breeding. Am Nat. 130:654-676. Stacey PB, Ligon JD. 1991. The benefits-of-philopatry hypothesis for the evolution of cooperative breeding: variation in territory quality and group size effects. Am Nat. 137:831-846. Walters JR, Copeyon CK, Carter JH III. 1992. Test of the ecological basis of cooperative breeding in red-cockaded woodpeckers. Auk. 109:90-97. Wang Z, Novak MA. 1994. Parental care and litter development in primiparous and multiparous prairie voles (Microtus ochrogaster). J Mammal. 75:18-23. White GC, Burnham KP. 1999. Program MARK: Survival estimation from populations of marked animals. Bird Study. 46 Supplement, 120-138.

33 Wolff JO. 1992. Parents suppress reproduction and stimulate dispersal in opposite-sex juvenile white- footed mice. Nature. 359:409-410. Wolff JO. 1994. Reproductive success of solitarily and communally nesting white-footed mice and deer mice. Behav Ecol. 5:206-209. Yáber MC, Rabenold KN. 2002. Effects of sociality on short-distance, female-biased dispersal in tropical wrens. J Anim Ecol. 71:1042-1055.

34 Table 1 Vegetative data. The percentage of vegetation within an enclosure that was comprised of dicots presented (a) by year and (b) by enclosure location. The maximum height of vegetation in meters by year for (c) all low (8 voles/0.1 ha) and high (24 voles/0.1 ha) enclosures and (d) the location of the enclosures (Fields 1 and 2). In all parts of this table, n = number of enclosures; 36 measurements/enclosure.

(a) Year n % dicot

2004 9 40.4 ± 11.7

2005 8 44.8 ± 10.0

(b) Enclosure Location n % dicot

1 6 4.1 ± 1.4

2 11 63.4 ± 4.2

35 ( c) Year Initial Density n Vegetation height (m)

2004 Low 4 0.84 ± 0.06

High 5 0.83 ± 0.05

2005 Low 4 1.01 ± 0.08

High 4 0.91 ± 0.06

(d) Year Enclosure Location n Vegetation height (m)

2004 1 4 0.93 ± 0.04

2 5 0.76 ± 0.02

2005 1 2 0.97 ± 0.13

2 6 0.96 ± 0.06

36 Table 2 Prairie vole social structure. Population density and the number of each type of social unit within enclosures, designated as either groups (at least two adults of the same sex), male-female pairs (M-F pairs), single females (single F), or single males (single M) and the total number of social units within each enclosure. All social units may or may not have offspring present.

Total Social Year Enclosure Density Groups M-F Pairs Single F Single M Units

2004 OE5 6.0 1 1 1 0 3

2004 NE6 10.1 0 0 2 1 3

2004 OE7 13.0 1 2 0 0 3

2005 OE5 14.0 2 2 0 0 4

2004 NE2 16.0 3 1 0 0 4

2004 OE3 18.0 1 2 5 0 8

2004 NE4 20.0 3 1 1 0 5

2004 OE1 20.1 1 0 1 1 3

2005 NE1 22.3 4 0 0 0 4

2005 OE7 27.3 4 1 0 0 5

2005 NE5 28.8 3 1 1 0 5

2005 NE6 39.9 6 2 0 0 8

2004 NE5 40.0 2 4 0 0 6

2005 NE2 49.7 8 0 0 0 8

2005 NE4 64.2 8 1 0 0 9

2004 NE3 73.6 8 1 0 0 9

2005 NE3 76.3 8 0 0 0 8

3.7 ± 0.7 1.1 ± 0.3 0.7 ± 0.3 0.1 ± 0.1 5.6 ± 0.6

37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 Density Residency Grid Nest estimate period

Figure 1 Trapping schedule for study. Hatched boxes represent 1.5 weeks between release of prairie voles and initiation of 14 consecutive weeks of trapping. Grey boxes represent grid trapping weeks (5 trap checks/week). White boxes represent nest trapping weeks (10 trap checks/week). The density estimates used in the analyses came from week 11, the final grid week. The three week residency period, which consisted of three consecutive nest trapping weeks, occurred during trapping weeks 12-14.

38 12

2004 10 2005

8

6

4 Mean Group Size Group Mean

2

0 0 1020304050607080 Density (voles/0.1ha)

Figure 2 Mean prairie vole group size (± SE) within enclosures as a function of density and year. For data from 2004, n = 8 enclosures since one enclosure had no groups. For data from 2005, n = 8 enclosures. The dotted line is the predicted effect from the 2004 ANCOVA model and the solid line is the predicted effect for the 2005 ANCOVA model.

39 1 0.9 0.8 0.7 0.6 0.5

Groups 0.4 0.3 0.2 2004

Proportion of Social Units that are are that Units Social of Proportion 0.1 2005 0 0 1020304050607080 Density (voles/0.1ha)

Figure 3 Proportion of prairie vole social units within an enclosure that were groups (≥ 2 adults of the same sex) as a function of density and year. The dotted line is the predicted relationship from the 2004 logistic model and the solid line is the predicted relationship for the 2005 logistic model.

40 1.2

1

0.8

0.6

0.4

0.2 Proportion of Offspring Remaining Philopatric Remaining Offspring of Proportion 0 0 102030405060708090 Density (voles/0.1ha)

Figure 4 Proportion of prairie vole offspring within each enclosure that were classified as philopatric during the three week residency period (weeks 12-14) as a function of density (n = 15 enclosures since 2 enclosures had no offspring meeting requirements for inclusion in the analysis). The line is the predicted logistic equation.

41 (a)

1.2 2004 1

0.8

0.6 Philopatric 0.4

Males 0.2 Females Proportion of Offspring Remaining Remaining Offspring of Proportion 0 0 1020304050607080 Density (voles/0.1ha)

(b)

1.2 2005 1

0.8

0.6 Philopatric 0.4

Males 0.2 Females Proportion of Offspring Remaining Remaining Offspring of Proportion 0 0 1020304050607080 Density (voles/0.1ha)

Figure 5 Proportion of male and female prairie vole offspring within each enclosure that were classified as philopatric during the three week residency period (weeks 12-14) as a function of density. (a) n = 8 enclosures for males; n = 5 enclosures for females; based on enclosures with offspring that met the requirements for inclusion in the analysis, (b) n = 7 enclosures for males; n = 7 enclosures for females; based on enclosures with offspring that met the requirements for inclusion in the analysis. The solid line is the predicted logistic relationship between density and the proportion of females remaining philopatric from the logistic model, the dashed line is the predicted logistic relationship between density and the proportion of males remaining philopatric.

42 GENERAL CONCLUSIONS

Cooperative breeding is a social system where individuals assist in rearing offspring that are not their own and has been documented across taxa in species of insects, fish, birds, and mammals (Jennions and Macdonald 1994; Koenig and Dickinson 2004; Queller 1994; Taborsky and Limberger 1981). Cooperative breeding arises from group formation, which often arises from philopatry of offspring (Emlen 1982; Lacey and Sherman 2007; Mumme 1997). The challenge has been how to explain why sexually mature animals would not disperse and forgo reproduction. This study is the first test of the habitat saturation hypotheses for philopatry in a cooperatively breeding mammal using an experimental manipulation of population density. At high population densities, the proportion of offspring remaining philopatric was greater than at low population densities. The proportion of philopatric offspring was positively correlated with the proportion of social units that were groups, as well as with the size of groups. These findings are all in accordance with predictions of the habitat saturation hypothesis and show a causal effect of population density on philopatry and the social structure of prairie voles. Furthermore, these findings help explain why an individual that is sexually mature may remain at its natal nest as opposed to dispersing and breeding independently. When the costs of dispersal are sufficiently high, as seen at high population densities where securing a territory for reproduction is unlikely, philopatry appears to be the best option for some individuals. This conclusion is in accordance with Emlen’s (1994) prediction that groups formed in response to limitations and the benefits of philopatry arose subsequently as a way of gaining fitness from the presence of group members. My findings provide strong support that habitat saturation, an ecological constraint, is a proximate factor contributing to philopatry in prairie voles. Because many cooperatively breeding mammals are large, such as members of the Canidae family, density manipulations are often difficult, and previous evidence for the habitat saturation hypothesis in cooperatively breeding mammals has been primarily correlative. However, this causal link between high population densities and greater proportions of individuals remaining philopatric is indicative that ecological factors may be influencing other cooperatively breeding mammals as well.

43 However, support for habitat saturation does not rule out other factors influencing philopatry. The delayed-dispersal threshold model (Koenig et al. 1992), proposes that multiple factors influence the probability of philopatry. Therefore, it is possible that in addition to habitat saturation, other factors are also a part of the decision to remain philopatric. For example, should an individual help with the care of related offspring, it may accrue indirect fitness benefits (Clutton-Brock 2002). Independent of helping, the individual may gain other benefits of philopatry including increased direct fitness, such as improved survival and access to resources (Emlen 1995; Kokko and Ekman 2002; Stacey and Ligon 1987). Future analyses of data from this study will help examine the relative importance of these benefits of philopatry in relation to the ecological constraints imposed by habitat saturation.

44 REFERENCES

Clutton-Brock T. 2002. Breeding together: kin selection and mutualism in cooperative vertebrates. Science. 296:69-72. Emlen ST. 1982. The evolution of helping. I. an ecological constraints model. Am Nat. 119:29-39. Emlen ST. 1994. Benefits, constraints and the evolution of family. Trends Ecol Evol. 9:282-285. Emlen ST. 1995. An evolutionary theory of the family. Proc Nat Acad Sci. 92:8092-8099. Jennions MD, Macdonald DW. 1994. Cooperative breeding in mammals. Trends Ecol Evol. 9:89-93. Koenig WD, Dickinson JL. 2004. Ecology and Evolution of Cooperative Breeding in Birds. United Kingdom: Cambridge University Press. Koenig WD, Pitelka FA, Carmen WJ, Mumme RL, Stanback MT. 1992. The evolution of delayed dispersal in cooperative breeders. Q Rev Biol. 67:111-150. Kokko H, Ekman J. 2002. Delayed dispersal as a route to breeding: territorial inheritance, safe havens, and ecological constraints. Am Nat. 160:468-484. Lacey EA, Sherman PW. 2007. The ecology of sociality in rodents. In: Wolff JO, Sherman PW. Rodent societies: an ecological and evolutionary perspective. Chicago, IL: The University of Chicago Press. p. 243-254. Mumme RL.1997. A bird's-eye view of mammalian cooperative breeding. In: Solomon NG, French JA. Cooperative Breeding in Mammals. New York, NY: Cambridge University Press. p. 364-388. Queller DC. 1994. Extended parental care and the origin of eusociality. Proc R Soc B. 256:105-111. Stacey PB, Ligon JD. 1987. Territory quality and dispersal options in the acorn woodpecker, and a challenge to the habitat-saturation model of cooperative breeding. Am Nat. 130:654-676. Taborsky M, Limberger D. 1981. Helpers in fish. Behav Ecol Sociobiol. 8:143-145.

45