Phylogeography and evolution of the Red Salamander (Pseudotriton ruber)

Abstract

Phylogeographic studies frequently result in the elevation of subspecific taxa to species given monophyly, or the synonymy of subspecies that are not monophyletic. However, given limited or incongruent datasets, retention of subspecies can be useful to describe hypothesized incipient species or to illustrate interesting biological phenomena driving morphological diversity. Four subspecific taxa have been used to describe largely allopatric geographic variation within the species Pseudotriton ruber, a plethodontid salamander occupying stream and spring habitats across eastern North America: P. r. vioscai occurs in lowland Coastal Plain habitats, while P. r. ruber, P. r. nitidus, and P. r. schencki occupy upland regions in and around the Appalachian Mountains. Pseudotriton ruber co-occurs through its distribution with the aposematic newt Notophthalmus viridescens, and both species are hypothesized to be part of a Müllerian mimicry complex. In this study, we sequenced regions of two mitochondrial (cytochrome b, NADH dehydrogenase subunit 2) and one single copy nuclear protein-coding gene (pro-opiomelanocortin) from individuals sampled across much of the distribution of P. ruber and then used maximum-likelihood and Bayesian phylogenetic inference to test the monophyly of subspecies, reconstruct biogeographic history, and make inferences about morphological evolution. Phylogeographic hypotheses from mitochondrial and nuclear datasets described structure among populations of P. ruber which separated Coastal Plain and upland Appalachian populations, but subspecies were not monophyletic. Biogeographic reconstruction estimated the ancestor of all populations to have occupied and initially diverged in the Coastal Plain during the Pliocene (∼3.6mya), before one lineage subsequently invaded upland areas of Appalachia. Bold bright coloration of high elevation subspecies P. r. nitidus and P. r. schencki appears to have evolved twice. We hypothesize that the Müllerian mimicry complex with N. viridescens and P. ruber may provide a selective mechanism driving the co-evolution of striking bright and dull morphological variation among populations of both species. While P. ruber subspecies were not consistent with our criteria for diagnosing species (monophyly) and therefore could not be elevated to species, we advocate for the retention of subspecies because they describe hypotheses about an incipient species (P. r. vioscai) and how Müllerian mimicry may shape morphological diversity of species.