UC Berkeley

Islands and archipelagos represent incredible natural experiments with which to study evolution and diversification. The island of Sulawesi is especially unique in that it is a large island that only in the last ~1 Ma fused together from seven or more paleoislands, many of which existed for ~20 Ma. This geographic history led Sulawesi to develop a diverse, endemic fauna comprising taxa that initially dispersed overwater then diversified across the paleoisland archipelago. Species boundaries across vertebrate lineages are found at faults and areas where paleoislands fused indicating their importance of in shaping two-dimensional biogeographic patterns; however, little is known about the impacts on species-groups that have colonized high elevations. A series of rugged mountains often over 3000m in elevation on Sulawesi provide a unique replicated system to study biogeographic drivers of diversification. I investigated and compared the evolutionary histories of two largely unstudied genera of skink lizards (Sphenomorphus, and Tytthoscincus) on Sulawesi to understand the dominant biogeographic forces that shaped their diversification. These groups span elevational gradients and represent distinct size classes, dispersal capabilities, and behavioral patterns that provide an ideal comparative system to investigate island and mountain diversification.

In my first chapter I developed a novel long-amplicon mitochondrial genome (mtGenome) sequencing method to cost-effectively screen more than 600 complete mtGenomes in parallel. These data revealed two astounding radiations of lizards on Sulawesi with both Sphenomorphus and Tytthoscincus each comprising 30–40 candidate species lineages. On Sulawesi, there are presently only eight lineages of Sphenomorphus and Tytthoscincus with names available, suggesting there are 60 or more undescribed species in these two genera. Future work to delimit and describe these species will be critical to conservation efforts as it will triple the number of lizard species known from Sulawesi. Phylogenetic estimates using mtGenomes were significantly better supported than from single genes highlighting the effectiveness of this new method. Still, important components of the tree were unsupported and needed to be resolved using genome-wide sequencing to test biogeographic hypotheses.

In my second chapter, I integrated thermal tolerance and morphometric data to investigate elevational adaptation across these radiations in a phylogenetic framework. Morphometric analyses display the variety of unique morphospaces skinks evolved into as they diversified across Sulawesi. Closely related species tend to share similar areas of morphospace indicating that these skinks do not represent an adaptive radiation, but rather waves of species groups diverging in allopatry then forced into sympatry as paleoislands fused. Highland species tend to have shared morphological characteristics such as longer hindlimbs indicating convergent responses to inhabiting high elevation habitats. The thermal tolerance analysis found that species in sympatry within an elevational band have divergent thermal tolerance patterns. This supports the existence of multiple adaptive routes to inhabiting higher elevation and the importance of interspecific interactions in shaping thermal physiology.

In my final chapter, I developed a powerful hybrid capture probeset to target samples within each clade from the mtGenome tree to produce robust phylogenomic estimates and test alternative biogeographic hypotheses on the evolutionary processes that generated this incredible diversity of species. The new probeset consisted of several sets of phyologenomic loci, including 284 rapidly evolving long exon loci I developed that optimize gene tree and species tree estimates. These loci produced a robust species tree estimate of these radiations and recovered divergence times for dispersal events that were exceptionally concordant with predicted paleo-island formation and movements. I recovered extreme elevational stratification of lineages, with 2–7 species in each genus inhabiting a single mountain. The biogeographic histories of both Sphenomorphus and Tytthoscincus are remarkably similar, highlighting the importance of the processes that generated diversity in both radiations. I found that early divergences set the foundation of lineages that persisted across different paleo-islands regions to this day. This foundation was built upon when the Central Core of Sulawesi uplifted and further subdivided many lineages through vicariance across the newly formed mountains. This led to a rapid burst of diversification containing a huge assemblage of species primarily in the Central Core. As Sulawesi further fused together these new lineages spread asymmetrically into the different regions of Sulawesi, possibly linked to diversity-dependent priority effects. Together these results highlight several of the most important processes leading to biodiversity.