Testing diversification models of endemic Philippine forest mice (Apomys) with nuclear phylogenies across elevational gradients reveals repeated colonization of isolated mountain ranges

Abstract

AbstractAimOur aims were to document the extent of diversification in an endemic clade of small mammals (Apomys, subgenus Megapomys) on a large oceanic island and to test whether speciation occurred primarily (1) along isolated elevational gradients or (2) among montane areas, as would be expected if diversification were driven by glacial cycles.LocationThe Philippine archipelago, particularly Luzon Island and two smaller adjacent islands (Mindoro and Lubang).MethodsWe analysed newly generated nuclear DNA sequences from five unlinked genes and mitochondrial cytochrome b using concatenation (likelihood and Bayesian) and coalescent‐based methods to estimate the species tree for all 13 species. We tested a priori models of speciation using both topological constraints and reconstructed elevational ranges of ancestors.ResultsAll individual gene trees recovered at least four of the morphologically defined Megapomys species as monophyletic, while the concatenated approaches recovered all 13 species as monophyletic. Each species was confined to a single mountain range or off‐shore island. Most mountain ranges had two species, but these species pairs usually were not sisters. Megapomys originated at medium to high elevation followed by three transitions into lower elevations and possibly one to high elevation. Both a priori models of speciation (elevational gradient and glacial cycle) were rejected by topology tests. The speciation rate was approximately constant through time.Main conclusionsNuclear sequence data strongly corroborated the species status of recently described taxa. A well‐supported phylogeny showed that Megapomys diversified by splitting into a predominantly high‐elevation clade and an entirely low‐elevation clade. Neither adaptation along elevational gradients on single mountain ranges nor vicariance of high‐elevation species following glacial cycle‐induced dispersals fitted the data. Rather, the most likely process explaining species distributions is repeated colonization of isolated mountain ranges by distantly related species.