Do greater spear-nosed bats have societies?

Abstract

Societies have been recognized in many animals but rarely used to describe social systems of bats. Here I discuss how the spatial and temporal relationships among greater spear-nosed bats, Phyllostomus hastatus, reflect societal features shared with other animal species in this special issue. These bats exhibit a social system in which one or more roosting groups of unrelated females or males occur within a cave or cave-like structure. The female members of a roosting group cooperate to defend young while often remaining together for a decade or more and can persist for multiple generations. A single male defends the group and monopolizes mating with them for up to 4 years. Females learn to match calls given during flight that are distinctive for their roosting group and use them to coordinate foraging trips. Females within a roosting group not only utilize shared space in a cave, in some populations, they also share foraging areas. While natal dispersal of both males and females, as occurs in greater spear-nosed bats, is unusual, other bat species likely exhibit some of these same societal features and warrant additional study.

Greater spear-nosed bats, Phyllostomus hastatus, exhibit many features of a society. This large (80—140 g), omnivorous bat occurs throughout much of the wet Neotropics. While these bats occasionally roost during the day in hollow trees, several hundred individuals can be found together in caves or cave-like structures (Eisenberg, 1989). These aggregations commonly consist of multiple roosting groups (Fig. 1). Roosting groups within a cave contain 15—30 adult females with a single reproductively dominant ‘harem’ male, although unstable clusters of 2—30 ‘bachelor’ males also occur (McCracken & Bradbury, 1977, 1981). Each harem male aggressively defends a female roosting group year-round for up to 4 years before another male replaces him, almost certainly after an injurious fight (Wilkinson et al., 2016). Adult females have much higher survival rates than males (Adams et al., 2025) and can live for at least 22 years (Wilkinson & Adams, 2019).

Figure 1.

Photographs of (a) a colony of greater spear-nosed bats roosting in the ceiling of a cave and (b) a closer view of one of the roosting groups consisting of a reproductively dominant male stationed adjacent to a cluster of females occupying a solution cavity in the cave.

Females within a roosting group can include individuals from multiple generations. How and why unrelated females form a roosting group and then either remain together or join another group is key to understanding this intriguing social system.

GROUP FORMATION AND PERSISTENCE

Unlike most mammals (Clutton-Brock & Lukas, 2012; Lukas & Clutton-Brock, 2011), including most bats, both male and female greater spear-nosed bat offspring disperse from their natal group. First-year males typically join loose assemblages of males of different ages (McCracken & Bradbury, 1981; Wilkinson et al., 2016). In contrast, first-year females often form a new roosting group with other like-age females away from the other groups or occasionally join an existing roosting group. Consequently, most females within a roosting group are unrelated to each other (Bohn et al., 2009; McCracken & Bradbury, 1981; Wilkinson et al., 2019). Adult females can remain together in the same roosting group for many years and likely most of their lives. For example, we captured eight individuals in the same cave ceiling cavity in Trinidad 12 years after they were initially captured together. The average community size of a female roosting group is 13.9 (Wilkinson et al., 2019). Even though harem males typically father the majority of young in a roosting group and aggressively defend females throughout the year, average extragroup paternity varies from 0% to 16% across caves and is positively related to age heterogeneity of females in a roosting group (Adams & Wilkinson, 2020). Such heterogeneity in paternity could be due to a few females switching roosting groups after mating, which occurs in December or January, but before parturition in April. What precipitates such movements has yet to be determined.

Despite being unrelated, females within a roosting group exhibit several cooperative behaviours that likely confer direct benefits (Wilkinson et al., 2016). For example, group members have been observed allogrooming (Wilkinson, 1988), pup guarding (Bohn et al., 2009), babysitting (Wilkinson et al., 2016) and group foraging (Boughman, 2006; Wilkinson & Boughman, 1998). Pup guarding and babysitting are facilitated by the dramatic reproductive synchrony exhibited each year by 2-year-old or older females in the same roosting group, in which they all typically give birth to a single pup within a week of each other (Porter & Wilkinson, 2001). The average date of parturition often differs among roosting groups within a cave (McCracken & Bradbury, 1981; Porter & Wilkinson, 2001).

Females do not actively cooperate across roosting groups, and instead compete during the nursing period for offspring survival. We witnessed females attacking and sometimes killing pups from other roosting groups that had inadvertently fallen to the cave floor (Bohn et al., 2009). However, fallen pups are defended by their mother or another roosting group member (Bohn et al., 2009). Given that both sexes leave their natal group but rarely leave their natal cave (e.g. over a 35-year period, 11 of 3482 recaptured females and 2 of 946 recaptured males were found in a cave other than where they were born), we suspect that such infanticide reduces future competition. All young bats give distinctive vocalizations, referred to as isolation calls, that females use for locating, identifying and retrieving young (Wilkinson, 2003). In greater spear-nosed bats, these isolation calls acoustically differ among individual pups (Bohn et al., 2007). Psychoacoustic experiments in the laboratory show that females can reliably discriminate calls from the different pups in a roosting group, but they are not as good at discriminating calls from pups taken from other roosting groups (Bohn et al., 2007). However, calls also change with age, so it is possible that females use age-related acoustic features to identify and attack pups from other groups.

In contrast to the cooperative behaviour of females in their roosting group, bachelor males that are roosting together frequently exhibit aggressive pushing and biting that can result in wounds, scars and broken canines. A male likely joins a roosting group of females by expelling the current resident. On one occasion we removed a resident harem male from a cave. The next day, an unmarked male had taken his place in the roosting group with no change in female membership. Harem males are, on average, in better body condition than bachelor males (Adams & Wilkinson, 2020) and have higher survival and more stable epigenomes than bachelor males (Adams et al., 2025). Among harem males, better body condition is associated with higher paternity (Adams & Wilkinson, 2020).

GROUP RECOGNITION

The extraordinary persistence of individuals that were not raised together suggests an ability either to keep track of each other individually or to use of some type of group recognition signal, such as the calls learned by yellow-naped parrots, Amazona auropalliata, that use the same communal roost (Wright et al., 2005). Social network analyses (Wilkinson et al., 2019), as well as the cooperative behaviours described above, indicate that bats within a roosting group recognize each other as individuals likely using acoustic or olfactory information. In addition, group recognition signals exist. During flight, greater spear-nosed bats give loud, audible ‘screech’ calls, especially while departing from a roost site and while foraging at feeding sites, such as at flowering trees (Wilkinson & Boughman, 1998). Screech calls span a wide frequency range but exhibit distinctive amplitude-frequency spectra that differ between roosting groups (Boughman, 1997) and from different caves (Boughman & Wilkinson, 1998). Habituation-discrimination playback experiments using females from two roosting groups kept in captivity revealed that females do not use screech calls to discriminate among individuals from the same roosting group, but they can use them to discriminate individuals from other roosting groups (Boughman & Wilkinson, 1998). Transfer of weaned bats to a different roosting group further revealed that females learn to match the screech call spectra of their new group within a few months (Boughman, 1998), consistent with adult vocal learning (Vernes & Wilkinson, 2019). We suspect these learned vocalizations enable individuals to recognize roosting group members at foraging sites (Wilkinson & Boughman, 1998). Additional studies are needed to determine whether bats recognize individuals from other caves, consistent with a multilevel society.

In addition to acoustic discrimination, bats within a roosting group also have distinctive group-specific odours. Greater spear-nosed bats possess a gland on the chest that in sexually mature adult males produces an odoriferous secretion. Harem males rub this secretion onto the backs of females in their group. Gas chromatography-mass spectrometry (GC-MS) analysis of those secretions revealed that each male has a distinct chemical profile and that reproductively dominant harem males can be distinguished from bachelor males (Adams et al., 2018). Given that screech calls are only given during flight, we suspect chemical cues are important for maintaining group cohesion while in the roost. Roosting sites are often adjacent (cf. Fig. 1a), so interactions between members of different roosting groups are common. Unpublished infrared-illuminated video recordings of interactions among females and bachelor males during the mating season indicate that females sometimes behave aggressively to females from other roosting groups and drive them away with bites and wing beats, indicating localized defence of their roost site. Nevertheless, a female will sometimes spend several days in another roosting group and occasionally switch groups (Fig. 2), which indicates that females are familiar with individuals from more than one group in a cave. What a female must do to avoid aggression and eventually join another roosting group has not yet been documented.

Figure 2.

Network diagram for co-roosting associations among female P. hastatus in a cave roost in Trinidad, West Indies between March 2001 and June 2004. Colour-coded roosting groups are identified by modularity analysis (reproduced from Wilkinson et al., 2019). Females were identified by numbered bands and unique fur markings. Individuals were assigned as roosting together either from capture data, daily censuses of marked animals or observations using night-vision recording equipment inside the cave.

FORAGING

Greater spear-nosed bats are omnivorous and consume a variety of animal (e.g. large beetles, alate termites and ants) and plant material (e.g. the pollen, nectar or fruit or various trees; Gardner, 1977). While the locations of fruiting or flowering trees or leaf-cutter ant colonies are static, the number of open flowers, ripe fruit or emerging alates available on a given night can vary and potentially create resources that are sufficiently rich to provide food for multiple individuals over several nights.

The perceived extent to which members of a roosting group utilize and defend a joint feeding area has likely been influenced by how data on space use by a flying nocturnal mammal has been collected over the past 50 years. Radiotracking studies conducted in Trinidad in the 1970s indicated that females of the same roosting group foraged in adjacent or overlapping areas 1—7 km from the cave, and members of different roosting groups used separate foraging areas, suggesting potential control of different physical spaces both within and between groups (McCracken & Bradbury, 1981). Two decades later, direct observations of light-tagged bats revealed that pairs or trios of females from the same roosting group sometimes departed from the roost together and were captured at the same flowering tree (Wilkinson & Boughman, 1998). But, females from more than one roosting group were also sometimes captured, indicating that roosting groups can converge at a rich feeding site (Wilkinson & Boughman, 1998). By monitoring the faecal material deposited every day by bats in three different roosting groups over 5 months, we found that the food eaten can differ among groups roosting in the same cave. For example, one group fed primarily on Rollinia multiflora fruit while other groups fed primarily on balsa tree, Ochroma pyramidale, pollen and nectar (Fig. 3). These results are consistent with information about food availability being acquired by individuals in the same roosting group, perhaps because bats often return to their day roost carrying fruit or covered in pollen (cf. Wilkinson & Boughman, 1998), but direct evidence of active defence of foraging sites is lacking.

Figure 3.

Proportion of faecal material collected each day beneath three female roosting groups (G22, G11, G10) in January 1993. At that time of year, pollen and nectar from O. pyramidale trees commonly make up much of the diet of greater spear-nosed bats (Wilkinson & Boughman, 1998), but as this figure shows, the majority of bats in a roosting group can quickly switch and feed on other food sources, consistent with some form of information sharing.

More recent GPS tracking of P. hastatus in Bocas del Toro, Panama revealed that not only did bats from different roosting groups share foraging areas, but bats from two different caves also foraged in the same location, while bats from a third equidistant cave foraged in a different location (Calderón-Capote et al., 2024). In contrast to Trinidad, these bats travel much longer distances (25 km from island caves to the mainland) to feed on balsa tree flowers (O’Mara & Dechmann, 2023). Additional study of greater spear-nosed bat foraging behaviour is warranted given that there are substantial morphological differences between these populations and they likely utilize at least some different food sources (Calderón-Capote et al., 2024; Wilkinson et al., 2024).

DO OTHER BATS FORM SOCIETIES?

To my knowledge, no other bat species form stable groups consisting of unrelated females, so greater spear-nosed bats may be exceptional in that regard. However, many tropical bats roost in aggregations containing multiple groups of females, each defended by a male (McCracken & Wilkinson, 2000). At least one of these species, the greater sac-winged bat, Saccopteryx bilineata, exhibits some of the same characteristics of a society as greater spear-nosed bats. These bats often roost between the buttresses of large trees, and colonies can contain 50 or more individuals and include multiple subgroups of females defended by single males (Bradbury & Emmons, 1974; Bradbury & Vehrencamp, 1977) as well as peripheral males. Males produce elaborate hovering displays in front of females (Voigt et al., 2008) during which they give vocalizations that resemble high-frequency birdsong (Behr et al., 2006; Davidson & Wilkinson, 2004). These songs exhibit differences among males within a colony and between colonies, which suggests the colonies have distinct long-term memberships (Davidson & Wilkinson, 2002; Knörnschild et al., 2012). Males often do not disperse from their natal colonies (Nagy et al., 2007) so colony differences in songs could be the result of genetic similarity or social learning (Knörnschild et al., 2010), or both. Nevertheless, male songs exhibit dialects and colony roosting sites are occupied by successive generations of males.

I suspect that colonies containing multiple social units connected by periodic associations are typical of some other bat species, such as common vampire bats, Desmodus rotundus (Carter et al., 2020, 2024; Hartman et al., 2024; Ripperger & Carter, 2021; Wilkinson, 1985), lesser spear-nosed bats, Phyllostomus discolor (Wilkinson, 1987), grey-headed flying foxes, Pteropus poliocephalus (Welbergen, 2005), and possibly others. In addition to describing patterns of association, how those associations form and are maintained needs to be determined. But, given that the social systems for the vast majority of the more than 1400 bat species (Wilson & Mittermeier, 2019) remain undescribed (Dorrestein et al., 2024; McCracken & Wilkinson, 2000), future work is likely to reveal additional cases of societies in bats.

Data Availability

No data were used for the research described in the article.