Kinship, association, and social complexity in bats

Abstract: Among mammals, bats exhibit extreme variation in sociality, with some species living largely solitary lives while others form colonies of more than a million individuals. Some tropical species form groups during the day that persist throughout the year while many temperate species only gather into groups during hibernation or parturition. How groups form and then persist has now been described for a number of species, but the degree to which kinship explains patterns of association has never been quantified across species. Here, we use social network analysis and genetic data to determine the extent to which relatedness contributes to associations among individuals estimated from free-ranging animals across nine species from four families of bats. Network analysis reveals that all species show evidence of emergent social structure. Variation in the strength of the relationship between genetic relatedness and social association appears to be related to the degree of roost switching, i.e., species in which individuals change roosts frequently tend to exhibit higher levels of association among relatives. Sex-biased dispersal determines whether associations were between male or female relatives. The strength of associations among kin does not predict known occurrence of complex behaviors, such as dominance or various types of cooperation, indicating that kinship is not a prerequisite for social complexity in bats. Significance statement: The number of differentiated relationships has been proposed as a way to measure social complexity. Among primates, relationships can be differentiated on the basis of rank, age, kinship, or association. Application of this approach to other groups of mammals that vary in sociality could help reveal ecological, behavioral, or cognitive similarities and differences between species. As a first step toward this approach, we used social network analysis on long-term individual records and estimated relatedness using genetic markers for nine species of bats. We confirmed nonrandom emergent social structure in all species. Kinship was predictive of social association among individuals of the same sex in a few species, but largely independent of the occurrence of complex behaviors, such as dominance among males or cooperation among females. Complex social behavior in bats appears to require frequent interactions among a small number of individuals that roost together for multiple years.