UC Berkeley

Species evolve in landscapes and environments that change through time. This spatial backdrop has profound effects on the diversification and merging of lineages. Populations become isolated by climate change, moving continents, and rising oceans, all of these factors can impact the evolutionary dynamics of lineages. While the spatial dimension of evolution has been a subject of interest since the inception of biogeography, the approaches to addressing how spatial heterogeneities affect lineages have advanced considerably in recent decades as genetic tools, computational tools, and new methodologies have allowed for explicit spatial hypotheses to be generated using GIS methods that can then be tested with genetic data.

The focus of this dissertation examines amphibian evolution at deep temporal scales in relation to a changing global landscape, and sets the stage for examining the spatial dimension across the Asian landscape at shallow temporal scales. Chapter one investigates amphibian evolution and diversification within an explicit spatiotemporal framework in order to understand how spatial variation drives evolutionary patterns of amphibians at global scales. At macroevolutionary scales, global amphibian diversity is strongly correlated with the area of forest rather than the longevity of forests, and signatures of the extensive forest area of the Eocene underlies the dramatic amphibian latitudinal diversity gradient.

Chapter two models the changing Asian landscape throughout the Pleistocene in order to understand how glacial and interglacial cycles impact the distribution of habitat types throughout the region. Satellite data on current habitats form the basis for this work. Palynological data are used to validate the models, providing a measure of confidence that the palaeo-predictive models are doing a good job in hind-casting habitats through changing climatic conditions. Asia differs from many other systems in not having been covered in extensive ice sheets during glacial periods, and provides a very different set of habitat dynamics, setting a unique stage for evolutionary dynamics. Habitats across Asia responded in dramatic fashion to changing palaeo-climates, with some habitats undergoing massive expansions and others contracting. The past 140 thousand years have witnessed a highly dynamic landscape.

Chapter three delves deeper into palaeo-ecological modeling, and spatially locates stable habitat refugia across Asia for multiple tropical forest, temperate forest, and non-forested habitat types. These refugia are examined with respect to climatic stability and latitude to evaluate if stable climates give rise to stable habitats. Refugia are further examined with respect to terrestrial vertebrate species richness to understand if stable refugia have helped structure contemporary diversity patterns. Strong relationships between habitat stability and species richness were found, indicating that the spatial backdrop has played a pivotal role in contemporary diversity patterns. This indicates that these refugia may be important buffers against climate change and are probably important for the conservation of diversity in Asia.

Chapter four focuses in on glacial and interglacial habitat dynamics on the Sundashelf, where dramatic changes in sea-level affect the connectivity of landmasses providing potential migration corridors between mainland SE Asia and insular Asia. I examine how these changing sea levels impact the distribution of habitats across the region and how in turn these habitat changes structure genetic data across the area. I found that high levels of genetic diversity correspond with stable habitats.

Chapter five turns its attention to phylogenetics of the family Dicroglossidae, a widespread Old World frog family with high diversity across Asia. This chapter examines the generic assignments and familial monophyly and evaluates how nucleotide substitution models and partitioning affect the phylogenetic informativeness of markers and how this impacts phylogenetic inference. Both partitioning scheme and model selection were found to have profound effects on the phylogenetic informativenss of the genetic data. It was found that there are systemic biases among model selection criteria with the BIC commonly selecting highly parameterized models. It was also found that models of nucleotide substitution are often inadequate in capturing the complexity of empirical data. The monophyly of the Dicroglossidae was supported in these analyses. Many of the genera within the family were found to be monophyletic with the exceptions of the frogs of the Painii group and the Fejervarya.

Together, these chapters provide the spatial foundation which sets the stage for research evaluating diversification processes in Dicroglossid frogs across the Old World and within Asia at biogeographic, phylogeographic and population genetics time scales. These analyses provide the first deeper time model-based evaluation of landscape processes for Asia.