Closing the ring: historical biogeography of the salamander ring species Ensatina eschscholtzii

Aim The salamander Ensatina eschscholtzii Gray is a classic example of a ring species, or a species that has expanded around a central barrier to form a secondary contact characterized by species‐level divergence. In the original formulation of the ring species scenario, an explicit biogeographical model was proposed to account for the occurrence of intraspecific sympatry between two subspecies in southern California (the ‘southern closure’ model). Here we develop an alternative ring species model that is informed by the geomorphological development of the California Coast Ranges, and which situates the point of ring closure in the Monterey Bay region of central coastal California (the ‘Monterey closure’ model). Our study has two aims. The first is to use phylogenetic methods to evaluate the two competing biogeographical models. The second is to describe patterns of phylogeographical diversity throughout the range of the Ensatina complex, and to compare these patterns with previously published molecular systematic data. Location Western North America, with a focus on the state of California, USA. Methods We obtained mitochondrial DNA sequence data from 385 individuals from 224 populations. A phylogeny was inferred using Bayesian techniques, and the geographical distributions of haplotypes and clades were mapped. The two biogeographical ring species models were tested against our Bayesian topology, including the associated Bayesian 95% credible set of trees. Results High levels of phylogeographical diversity were revealed, especially in central coastal and northern California. Our Bayesian topology contradicts the Monterey closure model; however, 0.08% of the trees in our Bayesian 95% credible set are consistent with this model. In contrast, the classic ring species biogeographical model (the southern closure model) is consistent with our Bayesian topology, as were 99.92% of the trees in our 95% credible set. Main conclusions Our Bayesian phylogenetic analysis most strongly supports the classic ring species model, modified to accommodate an improved understanding of the complex geomorphological evolution of the California Coast Ranges. In addition, high levels of phylogeographical diversity in central and northern California were identified, which is consistent with the striking levels of allozymic differentiation reported previously from those regions.