Significance of pre-Quaternary climate change for montane species diversity: insights from Asian salamanders (Salamandridae: Pachytriton).

Despite extensive focus on the genetic legacy of Pleistocene glaciation, impacts of earlier climatic change on biodiversity are poorly understood. Because amphibians are highly sensitive to variations in precipitation and temperature, we use a genus of Chinese montane salamanders (Salamandridae: Pachytriton) to study paleoclimatic change in East Asia, which experienced intensification of its monsoon circulation in the late Miocene associated with subsequent Pliocene warming. Using both nuclear and mitochondrial DNA sequences, we reconstruct the species tree under a coalescent model and demonstrate that all major lineages originated before the Quaternary. Initial speciation within the genus occurred after the summer monsoon entered a stage of substantial intensification. Heavy summer precipitation established temporary water connectivity through overflows between adjacent stream systems, which may facilitate geographic range expansion by aquatic species such as Pachytriton. Species were formed in allopatry likely through vicariant isolation during or after range expansion. To evaluate the influence of Pliocene warming on these cold-adapted salamanders, we construct a novel temperature buffer-zone model, which suggests widespread physiological stress or even extinction during the warming period. A significant deceleration of species accumulation rate is consistent with Pliocene range contraction, which affected P. granulosus and P. archospotus the most because they lack large temperature buffer zones. In contrast, demographic growth occurred in species for which refugia persist. The buffer-zone model reveals the Huangshan Mountain as a potential climatic refugium, which is similar to that found for other East Asian organisms. Our approach can incorporate future climatic data to evaluate the potential impact of ongoing global warming on montane species (particularly amphibians) and to predict possible population declines.