Effects of experience on the development of social behaviour of house-dwelling bats

This study aims at clarifying the mechanisms underlying the formation and maintenance of stable social units in bats (Chiroptera), a characteristic shared by most species of this group of mammals, by investigating the effects of the early social environment, i.e. pursuing the hypothesis that imprinting-like influences on the development of social behaviour exist and may have a profound impact on the social lives of bats. First I present an overview on bats' social behaviour and systems: these mammals comprise a high number of species and thus constitute an excellent group for testing general hypotheses about evolution and development of social behaviour. Bat social systems in fact range from solitary species to others aggregating in conspicuous groups of up to millions individuals. Such complexity leads to a variety of social behaviours rarely found in other taxonomic groups: there is increasing evidence that bats are able of cooperative social behaviours such as allogrooming, communal nursing, group hunting and social learning, all interactions that require high-level cognitive skills. Investigating such a complex system needs a multi-disciplinary approach, fundamental for disentangling the mechanisms through which bat sociality develops. I performed a series of experiments and used classical ethological and statistical methods (ethogram composition, general linear models) together with social network analysis (SNA), developing the analyses of social interactions on an individual-based approach. Experiment 1 deals with the ability of bats to modulate their behaviour (e.g. aggressiveness) according to intrinsic (e.g. age, sex) and extrinsic (familiarity) factors during a social interaction. I performed dyadic arena-encounters where two bats per test were allowed to freely interact. I recorded aggressive and affiliative behaviours and measured the degree of affiliation towards familiar and unfamiliar individuals of captive European free-tailed bats 9 (Tadarida teniotis). By testing individuals from different captive colonies and of different age, I measured the effects of familiarity, sex and age on the aggressive behaviour of this species, using aggressiveness as a proxy for xenophobia. I found significant effects of all the selected factors upon the degree of affiliation among individuals: familiar bats were more prone to perform affiliative behaviours, behaving xenophobically towards non-group members. This xenophobic attitude was lower in females and almost null towards juveniles. These results indicate that adult bats' behaviour is influenced by previous social experiences, also suggesting that social bonds formed inside colonies are long lasting regardless of genetic relatedness existing between individuals. With experiment 2 I investigate the mechanisms leading to the formation of social subunits in groups bats, using Pipistrellus kuhlii as a model species. By manipulating the early social environment of young bats and describing their pattern of association inside artificial roost as well as measuring their rates of interaction, I demonstrate that spatial proximity inside roosts promotes social cohesion. These associations are maintained by bats throughout adulthood by means of cooperative behaviours such as allogrooming and social thermoregulation. Both classical approach and social network analysis of interacting bats indicate that physical contacts and cooperative behaviours among bats inside a colony are non-random and are more frequently performed between individuals that already had contacts at a very young age. Following the same approach and techniques of Experiment 2, with Experiment 3 I test the hypothesis that the same mechanisms that produce group cohesion inside roosts can lead to the formation of multi-specific associations of bats. Such multi-specific groups are widespread among mammals, and in bats they are assumed to form due to eco-physiological reasons (i.e. species sharing micro-climatic requirements). I manipulated the early social environment of two species that naturally occur inside the same roosts; very young Kuhl's (P. kuhlii) and Savi's (Hypsugo savii) bats were exposed to artificial multi-specific social contacts in captivity. I demonstrate that early social experience does influence social bonding also beyond the species' boundaries. Independent young bats in fact selected previous groupmembers for social thermoregulation and reciprocal grooming, regardless of species membership. Results from all experiment clearly indicate a strong effect of early social environment on the interaction and association patterns in bats, both at short (Experiments 2-3) and long (Experiment 1) time scales, suggesting the existence of imprinting-like mechanisms. Such mechanisms lead to the formation of cryptic social subunits within bat colonies and probably enhance the cohesion of the entire social structure, with obvious and strong consequences on behavioural and ecological (e.g. demographic and epidemiological) scales.