Phylogenomics at the tips: inferring lineages and their demographic history in a tropical lizard, Carlia amax

High-throughput sequencing approaches offer opportunities to better understand the evolutionary processes driving diversification, particularly in nonmodel organisms. In particular, the 100-1000's of loci that can now be sequenced are providing unprecedented power in population, speciation and phylogenetic studies. Here, we apply an exon capture approach to generate >99% complete sequence and SNP data across >2000 loci from a tropical skink, Carlia amax, and exploit these data to identify divergent lineages and infer their relationships and demographic histories. This is especially relevant to low-dispersal tropical taxa that often have cryptic diversity and spatially dynamic histories. For C. amax, clustering of nuclear SNPs and coalescent-based species delimitation analyses identify four divergent lineages, one fewer than predicted based on geographically coherent mtDNA clades (>9.4% sequence divergence). Three of these lineages are widespread and parapatric on the mainland, whereas the most divergent is restricted to islands off the northeast Northern Territory. Tests for population expansion reject an equilibrium isolation-by-distance model for two of the three widespread lineages and infer refugial expansion sources in the relatively mesic northeast Top End and northwest Kimberley. The latter is already recognized as a hotspot of endemism, but our results also suggest that a stronger focus on the northeast Top End, and adjacent islands is warranted. More generally, our results show how genome-reduction methods such as exon capture can yield insights into the pattern and dynamics of biodiversity across complex landscapes with as yet poorly understood biogeographic history and how exon data can link between population and phylogenetic questions.This research wassupported by an ARC Laureate Fellowship awarded to CM(ARC FL110100104). This work used the Vincent J. CoatesGenomics Sequencing Laboratory at UC Berkeley, supportedby NIH S10 Instrumentation Grants S10RR029668 andS10RR027303, as well as the ACRF Biomolecular ResourceFacility at ANU