The Philippine Archipelago represents a natural model system for the study of the processes of diversification. The islands harbor one of the greatest concentrations of biodiversity in the world and have a dynamic, yet well-understood, geological history. During the last 500,000 years, sets of islands in the Philippines have been repeatedly connected and isolated by changing sea levels. The researchers will use molecular data from populations of six species (three lizards and three frogs) co-distributed across one such set of islands and novel analytical techniques to test hypotheses about how recent climate change and associated sea-level change has influenced diversification in this dynamic system. This research will yield insights into the processes responsible for the formation of new species, and investigate the effect of climate and sea-level change on biodiversity. This information is vitally important for understanding the long-term effects that current climate change will have on global biodiversity. One of the analytical techniques developed for this research will represent an important tool for delimiting species in a way that is both biologically and statistically meaningful, which will have important implications for both basic and conservation research.
The goal of this project was to use comparative genetic data to better understand the processes that generated the rich biodiversity of the Philippine Archipelago, and to assess and advance statistical methods of inferring models of diversification. Genetic data was collected from a suite of vertebrate species distributed across the Philippine Islands. If island fragmentation promoted speciation during Pleistocene inter-glacial rises in sea level, the divergences across these species are expected to have been recent and temporally clustered. A statistical model-choice method was used to test this prediction using the comparative genetic data. The empirical results indicated strong statistical support for a highly clustered model of divergence across 22 distantly related vertebrate taxa distributed across the Philippine Islands, seemingly corroborating the hypothesis of Pleistocene climate-driven diversification. However, extensive simulation-based assessment of the accuracy, power, and robustness of the method revealed low power to detect differences in divergence times given the data, and bias toward estimating temporally clustered models of diversification. Given this finding, the investigators developed a nonparametric Bayesian model-choice approach that places a flexible Dirichlet-process prior over all possible models of divergence. Furthermore, genome-wide data was collected from two groups of geckos co-distributed throughout the Philippines. The new method and added utility of genomic data will provide better resolution in testing the climate-driven model of diversification. This award greatly enhanced the doctoral dissertation research of co-Principal Investigator, Jamie Oaks, and provided research opportunities for several University of Kansas undergraduate students. Methods developed for this project will be of general interest to researchers interested in understanding the mechanisms that generate and maintain biodiversity and influence community assembly. All methods are implemented in freely available, open-source software packages, including dpp-msbayes (https://github.com/joaks1/dpp-msbayes), PyMsBayes (https://github.com/joaks1/PyMsBayes), ABACUS (https://github.com/joaks1/abacus), and SeqSift (https://github.com/joaks1/SeqSift).
