TESTING BIOGEOGRAPHIC PREDICTIONS DERIVED FROM QUATERNARY ISLAND CONNECTIVITY USING GENOMIC DATA FROM MAMMALS IN THE SOUTHERN PHILIPPINES

Abstract

Island archipelagos have long piqued the attention of evolutionary biologists interested in biogeographical patterns implying processes of diversification in terrestrial vertebrates. One model island archipelago, the Philippines, has emerged as an ideal geographical theater for testing hypotheses related to a 30 year biogeographical paradigm, derived from climatic oscillations of the Pleistocene and species pump expectations. To test predictions from the Pleistocene Aggregate Island Complex (PAIC) Paradigm, we generated the first ever Rad-seq datasets of thousands of genomewide loci from forest mammals restricted to the southern portions of the archipelago. We used genomic population-level statistics and demographic model selection techniques to test assumptions of the PAIC Paradigm and predicted differences among focal taxa based on life history and ecological preferences. Predictably, our analyses revealed pronounced differences in historical population demographics of forest rodents and understory bats in terms of geographical structure of populations and gene flow among populations, suggesting taxon-specific response to climatic and environmental variation across recent evolutionary timescales. In contrast to previous inferences derived from single-locus phylogeography of the same populations, our new genomic data warn against uncritical assumptions of highly divergent, geographically structured island populations. They also suggest that individual, species-specific, idiosyncratic response of small forest mammal lineages to the geographical and environmental template may become more prevalent in structured island systems, in strong contrast to the popularly hypothesized shared mechanisms of diversification that have been so prevalent over the last century.