Female philopatry may influence antipredatory behavior in a solitary mammal

Introduction

Relatedness between neighboring conspecifics has a number of ecological and evolutionary consequences (Lawson Handley & Perrin, 2007; Clutton-Brock & Lukas, 2012; Nichols, Zecherle & Arbuckle, 2016), affecting fitness (van Noordwijk et al., 2012; Jungwirth et al., 2023; Walmsley et al., 2023) and behavior (Thorington et al., 2010; Williams et al., 2013), including spatial and social organization (Støen et al., 2005; de Oliveira et al., 2022; Ausband, 2024). High relatedness between neighbors often results from philopatry, in which adults settle close to their natal home range, and dispersal is often sex-biased (Lawson Handley & Perrin, 2007). Mammals typically exhibit male-biased dispersal (Lawson Handley & Perrin, 2007; Mabry et al., 2013), and daughters can benefit from remaining close to their mothers through behaviors such as inheriting territory (Støen et al., 2005; Goodrich et al., 2010) or breeding position (Ausband, 2024). Antipredatory behavior can also be affected by sociospatial patterns resulting from philopatry and may represent an additional benefit of philopatry. Alarm vocalizations may benefit neighboring conspecifics by functioning as a warning that a predator is present, directly deterring predators, or both (Sherman, 1977; Schel, Candiotti & Zuberbühler, 2010; Fuong, Maldonado-Chaparro & Blumstein, 2015; Isbell & Bidner, 2016). Alarm calling behavior can therefore be influenced by kin selection where species live within close proximity of relatives (Dunford, 1977; Sherman, 1977; Wheeler, 2008).

Ground squirrels of the family Sciuridae emit alarm vocalizations in response to predators, and many ground squirrels live in close proximity to their relatives. Kin selection is therefore thought to be a strong evolutionary driver of calling behavior (Dunford, 1977; Sherman, 1977). Male ground squirrels often disperse away from natal burrows whereas females typically exhibit philopatry (e.g., Holekamp, 1984; Shriner & Stacey, 1991; Neuhaus, 2006; Nguyen & Van Vuren, 2024). As a result, adult males may be unrelated to surrounding conspecifics, so alarm calling would not benefit adult males via indirect fitness. Thus, individuals emitting alarm vocalizations may be more likely to be female in species with male-biased dispersal (Dunford, 1977; Sherman, 1977). The rate of male-biased dispersal can vary substantially across ground squirrel species, however, with more social species exhibiting higher rates of male dispersal (Devillard et al., 2004). Further, not all squirrel species exhibit close relatedness between neighboring squirrels (Munroe & Koprowski, 2014; Glover, 2018). For example, the round-tailed ground squirrel (Xerospermophilus tereticaudus) lives gregariously in groups of unrelated individuals (Munroe & Koprowski, 2014). Round-tailed ground squirrels also emit alarm signals, but because they live in groups of unrelated individuals, kin selection is not likely to be a primary driver of alarm calling behavior. Other evolutionary drivers of alarm vocalizations could include direct fitness (Sherman, 1977; Blumstein et al., 1997) or reciprocity (Trivers, 1971, but see discussion in Blumstein, 2007). Alarm calls may also be directed at predators (Wheeler, 2008; Digweed & Rendall, 2009; Burnett & Koprowski, 2020), and predator-prey interactions could be the primary driver of alarm calling behavior in some squirrel species.

The Harris’s antelope squirrel (Ammospermophilus harrisii) is a solitary species that emits alarm calls throughout the year, regardless of juvenile presence, indicating that alarm calls may be directed toward predators as a deterrent (Burnett & Koprowski, 2020). Opportunistic observations of alarm vocalizations showed a higher proportion of female callers (Burnett & Koprowski, 2020), suggesting that calling behavior could be influenced by female philopatry and kin selection. Although adult A. harrisii live solitarily in large home ranges, neighboring home ranges overlap (Burnett & Koprowski, 2024). Further, A. harrisii alarm vocalizations can be high amplitude (A. Burnett, 2017, personal observation) and are structured to carry over long distances (Bolles, 1988), potentially benefiting neighbors if individuals use alarm vocalizations as a warning or if predators are deterred from the area. However, whether neighboring A. harrisii are related is unknown. To understand whether philopatry could influence calling propensity in A. harrisii, we sampled the genomes of neighboring individuals to analyze their relatedness and estimate any sex-dependency. Given our observation of more female antelope squirrels alarm calling (Burnett & Koprowski, 2020), we expect patterns in relatedness of A. harrisii to be indicative of female philopatry, in which female-female relatedness is higher than male-male relatedness and the relationship between genetic and geographic distance is sex-dependent, showing relatedness between neighboring females but not males. Conversely, if neighboring individuals are not related or dispersal is not male-biased, higher call propensity in females may be maintained primarily through direct fitness or alternative selection pressures (e.g., Blumstein et al., 1997).

Materials and Methods

Sample collection and DNA extraction

We sampled A. harrisii in the Santa Rita Experimental Range as part of a larger research study conducted in 2017 and 2018, using the same focal areas as those in Burnett & Koprowski (2020) in which squirrels were observed to exhibit sex-biased calling. We chose to sample in the Santa Rita Experimental Range because we were specifically interested in whether the sex-biased calling observed in this population could be influenced by relatedness between individuals, and whether other populations also exhibit sex-biased calling has not yet been investigated. We focused our trapping efforts around four focal areas, where we experienced much higher trapping success compared to other locations in the study area (Burnett & Koprowski, 2020, 2024). Harris’s antelope squirrels were baited and captured with Tomahawk live traps (model No. 201; Tomahawk Live Trap, Hazelhurst, WI, USA) in the Santa Rita Experimental Range (SRER), located in the Sonoran Desert approximately 65 km south of Tucson, Arizona (Fig. 1). Traps were checked frequently (once/hour) and shaded with vegetation and/or shade cloth to prevent heat stress and sun exposure. Once captured, we used a cloth handling cone (Koprowski, 2002) to minimize stress during handling. We recorded weight, sex, life stage, and reproductive status for each animal captured and collected ear tissue samples with an ear punch (Fisherbrand Animal Ear Punch, 1 mm, Thermo Fisher Scientific, Waltham, MA, USA). Squirrels were tagged with sterile passive integrated transponder (PIT) tags (HPT9, 8.4 × 1.4 mm, 0.02–0.04% body weight, Biomark, Inc., Boise, ID, USA) before being released. Some adult individuals (>110 g) were additionally fitted with a radio collar for VHF tracking (Wildlife Materials; <5% body weight; see Burnett & Koprowski, 2024). We took no more than one tissue sample per ear from squirrels captured more than once (n = 47 individuals, 51 tissue samples prior to removal of duplicate individuals). We received approval from University of Arizona Institutional Animal Care and Use Committee (16–169) and complied with the Animal Welfare Act for all procedures. We additionally followed ethical guidelines for trapping and handling small mammals published by the American Society of Mammalogists (Sikes & Animal Care and Use Committee, 2016). We obtained a scientific collecting permit from Arizona Game and Fish Department (SP501610 & SP611944). We did not give the animals anesthesia or analgesia because the effects of these agents are not well-studied in Harris’s antelope squirrels. Animals were not chemically immobilized due to the nature of the procedures (i.e., momentary pain). Lack of chemical immobilization also limits time spent handling, risk of thermoregulatory distress, and additional stress caused by immobilization procedures (Sikes & Animal Care and Use Committee, 2016). The DNA of the collected tissue samples was extracted at the University of Arizona Conservation Genetic Laboratory using a Qiagen DNeasy blood and tissue extraction kit (Qiagen Inc., Hilden, Germany). We used a Qubit fluorometer (Invitrogen™, Thermo Fisher Scientific Inc., Waltham, MA, USA) to quantify DNA products before sequencing.

Results

Isolation by distance

Genetic distance between squirrels showed a significant relationship with geographic location (Mantel test; R = 0.18, p = 0.001, n = 47), such that squirrels at closer distances were more related (Fig. S3). When squirrels were separated by sex, correlation between genetic and geographic distance strengthened and significant patterns remained for females (Mantel test; R = 0.29, p = 1e−4, n = 29; Fig. 2), but not males (Mantel test; R = 0.03, p = 0.39, n = 18). When geographic distances were grouped into distance classes, female-female genetic distance was positively correlated (Mantel correlogram analysis; R = 0.46, p = 0.001, n = 86) with geographic distance at close distances (0–100 m), but negatively correlated (Mantel correlogram analysis; R = −0.26, p < 0.01, n = 238) at further distances (3–6 km; Table S1). Male-male genetic distance exhibited very low or no correlation with geographic distance (Mantel correlogram analysis; R < 0.1, p > 0.05) regardless of distance class (Fig. S4, Table S2).

Discussion

Harris’s antelope squirrels in the SRER showed genetic diversity reflective of small, isolated and fragmented populations, similar to that of northern and southern Idaho ground squirrels (Garner, Rachlow & Waits, 2005; Barbosa et al., 2021). Our population may exhibit similar genetic diversity due to major physical barriers in all cardinal directions, with cities and major highways positioned to the north and west of the SRER and the Santa Rita Mountains extending from the southern boundary of the SRER to the northeastern boundary.

The sex-dependent relationship between geographic distance and genetic distance, as well as the higher relatedness observed between female-female pairs compared to female-male or male-male pairs provides evidence that A. harrisii exhibits male-biased dispersal, or female philopatry. The results of the Mantel correlogram analysis provide further supportive evidence of male-biased dispersal and spatial clustering of female relatives in our population of A. harrisii, with female squirrels that are further distances apart being largely unrelated. Female philopatry is a widespread mammalian trait, particularly in species with polygynous mating systems (Lawson Handley & Perrin, 2007; Mabry et al., 2013) like those found in ground squirrels. Male-biased dispersal is exhibited in a number of other ground squirrel species (Devillard et al., 2004), including Golden-mantled ground squirrels (Callospermophilus lateralis; Nguyen & Van Vuren, 2024), Columbian ground squirrels (Waterman, 1992; Neuhaus, 2006), rock squirrels (Otospermophilus variegatus; Shriner & Stacey, 1991), and Belding’s ground squirrels (Urocitellus beldingi; Holekamp, 1984). Male-biased dispersal in ground squirrels may help prevent inbreeding (Holekamp, 1984) or reduce exposure to female aggression (Neuhaus, 2006). Relationships between geographic and genetic distance vary across ground squirrel species, however. Speckled ground squirrels (Spermophilus suslicus) show significant positive correlations between genetic and geographic distances (Matrosova et al., 2016), whereas northern and southern Idaho ground squirrels (Urocitellus brunneus and U. endemicus, respectively; Garner, Rachlow & Waits, 2005), round-tailed ground squirrels (Xerospermophilus tereticaudus; Munroe & Koprowski, 2014), and California ground squirrels (Otospermophilus beecheyi; Glover, 2018) do not exhibit positive relationships between genetic and geographic distances. Variation in the relationship between geographic distance and relatedness among squirrel species could be due to differences in how sex-biased dispersal is expressed, which can vary substantially among species (Devillard et al., 2004; Lawson Handley & Perrin, 2007). Although female philopatry is widespread, the magnitude of bias (e.g., whether one sex exhibits complete philopatry or not) and the ratio of dispersal distances between males and females varies among species (see Lawson Handley & Perrin, 2007 for review). Variation in the degree of sex-bias may explain the relatively weak correlation (R = 0.29) between female relatedness and geographic distance. Adult females were captured within close proximity (<300 m) of other adult females of both high (e.g., r = 0.25) and low (e.g., r = 0) relatedness, indicating that clustering among related individuals is relatively relaxed and some female offspring may also disperse depending on resource availability, mate competition, and aggression from conspecifics (Lawson Handley & Perrin, 2007).

Our results support kin selection theory, in which solitary females that exhibit overlapping territories are expected to be related, such that tolerance of neighbors benefits females via indirect fitness (Hamilton, 1964). Our study area exhibited a uniform distribution of resources at a low spatial density, and antelope squirrels in our study area maintain large, overlapping home ranges ranging from ~0.5 to 7 ha (Burnett & Koprowski, 2024). Other solitary species similarly benefit from female philopatry in a number of ways, including allo-parental care (Gilchrist, 2007; Schradin, Vuarin & Rimbach, 2018), territory acquisition (Lutermann et al., 2006; Goodrich et al., 2010; le Roex et al., 2022; Payne et al., 2024) and thermoregulation (Williams et al., 2013). Females in a number of solitary species, including bobcats (Lynx rufus; Janečka et al., 2007; Payne et al., 2024), Amur tigers (Panthera tigris altaica; Goodrich et al., 2010), and brown bears (Ursus arctos; Støen et al., 2005), share home ranges with their daughters, resulting in kin-related spatial structure that could have important repercussions for indirect fitness and social relationships (Støen et al., 2005; de Oliveira et al., 2022). Although female philopatry and kinship theory is useful for understanding the social relationships in some solitary mammals, other solitary species exhibit adaptive social strategies that are maintained by familiarity with neighbors (Siracusa et al., 2019) or reciprocity (Elbroch et al., 2017). For example, North American red squirrels (Tamiasciurus hudsonicus) are highly territorial but exhibit behavioral plasticity, reducing effort spent defending their territory (i.e., emitting territorial vocalizations) and increasing time spent in the nest as familiarity with their neighbors increases over time (Siracusa et al., 2019). Pumas (Puma concolor) cofeed at kill sites with unrelated individuals and maintain social networks via reciprocity (Elbroch et al., 2017). Some species, like North American red squirrels, can exhibit social strategies governed by both kinship and familiarity simultaneously (Walmsley et al., 2023). Thus, a number of ecological factors can select for social structure to evolve across the spectrum of sociality (Makuya & Schradin, 2024).

Our findings indicate that kin selection resulting from high genetic relatedness between neighboring females may be partially responsible for sex differences in alarm calling behavior in A. harrisii (Burnett & Koprowski, 2020), highlighting the nuanced role that ecological patterns like female philopatry can play in species behavior. Kin selection may have a strong evolutionary influence on antipredatory behavior in many ground squirrels, especially those for which alarm vocalizations likely serve as a warning to surrounding relatives and provide predator details (Ackers & Slobodchikoff, 1999). However, A. harrisii is largely solitary and emits alarm vocalizations at a wide range of amplitudes, including very low amplitudes that do not travel across the landscape and would be difficult for neighboring squirrels to detect (personal observation). Additionally, sex bias in calling propensity is nonsignificant under high-risk contexts (e.g., when confined; Burnett & Koprowski, 2020), indicating that while females may have a lower threshold for calling than males, both sexes vocalize in response to immediate threats. Alarm vocalizations in this species are thought to be directed toward the predator (Burnett & Koprowski, 2020), thus predation pressure may still be primarily responsible for the maintenance of alarm vocalizations in this species. Although these results show alarm calling behavior in our population of antelope squirrels is likely subject to kin selection, whether alarm vocalizations serve multiple functions is still unclear. Alarm calls were frequently given by multiple individuals (Burnett & Koprowski, 2020), but whether alarm calls elicit further alarm calling by neighboring squirrels or whether squirrels were simply calling in response to the same threat is unknown. During our field observations over the course of 2 years, we did not observe a clear behavioral response to alarm calls from neighboring antelope squirrels, however, relatives may benefit from alarm vocalizations if predators leave the immediate area to hunt elsewhere (Blumstein et al., 1997; Zuberbühler, Jenny & Bshary, 1999; Isbell & Bidner, 2016). Although kin selection is an important component underlying the evolution of alarm signals, other evolutionary pressures such as those imposed by the environment and the intended receiver (e.g., conspecifics, predators, or both) are also important to take into consideration (Guilford & Dawkins, 1991; Patricelli & Hebets, 2016).

Conclusions

Our study aimed to determine whether female philopatry influences alarm calling behavior in Harris’s antelope squirrels. We hypothesized that relatedness between squirrels would be correlated with geographic distance and that females would be more closely related to neighboring squirrels than males, based on previous findings of greater calling propensity in female antelope squirrels. We found that dispersal in our population of Harris’s antelope squirrel is male-biased, and genetic distance and geographic distance were positively correlated for female squirrels but not males, indicating that kin selection resulting from female philopatry may be responsible for sex differences in calling behavior. Relatedness between neighboring females supports kin selection theory predicting that solitary females with overlapping home ranges are likely to be related. Our results complement other studies showing that female philopatry plays an important role in the formation of sociospatial patterns in solitary mammals. We additionally found low genetic diversity in our population, suggesting that our population may be somewhat isolated from other populations due to topographical barriers. Our findings add to the body of literature underlining how sociospatial organization may shape the behavior of solitary mammals and highlight the subtle ways that kinship and genetic structure influence the social landscape experienced by solitary mammals.

Supplemental Information