A Phylogenetic History of the Tenerife Skink Chalcides viridanus: A multi-locus Coalescent Approach

Abstract

Previous studies on the Tenerife skink Chalcides viridanus found clear links between observed within-island cladogenesis and the geological history of the island. Since these studies there have been many advances in conceptual, numerical and methodological approaches in phylogenetic analyses. This study aims to revisit the phylogeography of the Tenerife skink, using more current *BEAST analysis techniques, and attempt to resolve some unanswered questions about population history of this species. Specifically, using previously unused nuclear gene markers to obtain a more robust phylogeographical history of the species, using new techniques to estimate whether any historic changes in population size can be linked to known geological events and whether there is enough evidence to reclassify any discovered genetic clades as distinct species. A multi-locus approach was undertaken, using more informative mtDNA gene fragments (Cyt-b & ND1 & 2, totalling 1566bp) as well as the sequencing of 5 nuclear loci (PRLR-555bp, Rag-1-761bp, RELN-583bp, EXPH-796bp and SELT-414bp). These sequences were combined with the latest Bayesian methods to estimate divergence times, historical changes in population structure and infer species boundaries. Results from the BAPS and *BEAST analyses identified three main population groups within the island, the geographical distribution of two of which were restricted to the areas representing the ancient precursor islands Teno & Anaga in the North West and North East of Tenerife, respectively. Contrary to previous findings, the divergence estimates reveal the NE lineage was first to diverge ~1.16 Ma, with the NW diverging ~ 0.6 Ma. The third population group is widely geographically distributed across the island consisting of two clades, one previously unidentified. Results show evidence of a substantial population expansion in this population group approximately 0.2 Ma, which ties in with the end of the last major eruptive cycle. Species delimitation analyses seem to favour a one species model for C. viridanus, however the effects of incomplete lineage sorting and low diversity in the nuclear loci make the robustness of this result questionable. Overall this study highlights both how using a multi-locus coalescent approach can reveal a greater insight in to a species’ phylogeographical history, but also how factors such as incomplete lineage sorting can severely limit any intraspecific phylogenetic inferences made on recently diverged population groups.