Evolution and ecology of Southeast Asian fruit bats in the genus, Cynopterus (Chiroptera: Pteropodidae)

Old World fruit bats in the genus Cynopterus are distributed throughout Southeast Asia and India and are ubiquitous in a wide range of habitats. However, both phylogenetic and ecological relationships within the genus are poorly defined. Likewise, while two species are known to modify foliage as roosts in anthropogenic habitats, the prevalence of this unique behavior is unknown and the roosting ecology of forest-associated species is undescribed.; To best address the range of evolutionary and ecological questions to which this system is suited, I used an integrative approach, incorporating molecular, morphological and ecological data, and sequentially narrowing the spatiotemporal focus of the research. First, I examined phylogenetic relationships among the three species that account for the full geographic range of the genus. This analysis revealed striking disparity between taxonomy and inferred evolutionary relationships, including unrecognized diversity in a single widespread species, C. brachyotis. Second, I asked how the four species that co-occur on the Malay peninsula responded to major changes in the ratio of forested to open habitats during the Pleistocene: was response similar, or did species-specific habitat associations mediate response to environmental change? The unique population histories identified for each species indicated that response to recent geological change was species-specific and, in two cases, readily interpretable in the context of ecology. Third, I used an ecomorphological approach to investigate separation in resource use among the four species. Results suggest several mechanisms of coexistence among species pairs, whereby convergence along one resource axis is complimented by divergence on another. Fourth, I compared the roosting ecology of the four species and tested the hypothesis that the use of modified roosts should be observed throughout the genus. While occupation of modified roosts was confirmed for all species, it was observed at low frequencies and was not a determinant of social organization. Fifth, I evaluated the relationships between the spatial distribution of resources important to females, female group size and male reproductive strategy. Roost distribution was a strong predictor of male roost fidelity, indicating that males adopt sedentary strategies when females and resources are aggregated, and may track female movements when females and resources are dispersed.