干旱区生态环境知识资源中心

Aim: To develop a robust phylogeny for the iconic Australian water skinks (Eulamprus) and to explore the influence of landscape evolution of eastern Australia on phylogeographic patterns.

Location: Eastern and south-eastern Australia.

Methods: We used Sanger methods to sequence a mitochondrial DNA (mtDNA) locus for 386 individuals across the five Eulamprus species to elucidate phylogeographic structure. We also sequenced a second mtDNA locus and four nuclear DNA (nDNA) loci for a subset of individuals to help inform our sampling strategy for next-generation sequencing. Finally, we generated an anchored hybrid enrichment (AHE) approach to sequence 378 loci for 25 individuals representing the major lineages identified in our Sanger dataset. These data were used to resolve the phylogenetic relationships among the species using coalescent-based species tree inference in *BEAST and ASTRAL.

Results: The relationships between Eulamprus species were resolved with a high level of confidence using our AHE dataset. In addition, our extensive mtDNA sampling revealed substantial phylogeographic structure in all species, with the exception of the geographically highly restricted E.leuraensis. Ratios of patristic distances (mtDNA/nDNA) indicate on average a 30-fold greater distance as estimated using the mtDNA locus ND4.

Main conclusions: The major divergences between lineages strongly support previously identified biogeographic barriers in eastern Australia based on studies of other taxa. These breaks appear to correlate with regions where the Great Escarpment is absent or obscure, suggesting topographic lowlands and the accompanying dry woodlands are a major barrier to dispersal for water skinks. While some river corridors, such as the Hunter Valley, were likely historically dry enough to inhibit the movement of Eulamprus populations, our data indicate that others, such as the Murray and Darling Rivers, are able to facilitate extensive gene flow through the vast arid and semi-arid lowlands of New South Wales and South Australia. Comparing the patristic distances between the mitochondrial and AHE datasets highlights the continued value in analysing both types of data.