Molecular phylogeny of long-tailed shrews (genus Sorex) from México and Guatemala

Abstract

We present a molecular phylogeny of North American species of long-tailed shrews of the genus Sorex. Our focus is on Mexican and Guatemalan species to begin understanding their evolutionary relationships and to test the validity of nominal species. Seventy-seven sequences of the mitochondrial cytochrome b gene were analyzed, including 19 specimens representing nine Mexican and one Guatemalan species. Phylogenetic analyses using parsimony, maximum likelihood and Bayesian approaches revealed two major clades of North American species, all within the subgenus Otisorex. The first major clade includes S. trowbridgii and southern species (S. macrodon from Oaxaca; S. veraecrucis from Nuevo León, Michoacán, Chiapas, S. saussurei from Jalisco and Guatemala; S. veraepacis from Guerrero and Guatemala). Relatively deep branches among taxa characterize this clade and suggest that their early divergence from other North American shrews was soon after arrival of the ancestral stock from the Beringian region. The other major clade includes all other North American species of Sorex we examined, with two Mexican species, S. milleri and S. emarginatus, grouped in a subclade with the S. cinereus complex. Sorex veraecrucis is not, however, a monophyletic taxon because specimens of this nominal species were included in both the major clades. The Isthmus of Tehuantepec has likely played a role as a biogeographic barrier in the evolutionary history of Mexican shrews. This study of mitochondrial variation in southern North American shrews of the genus Sorex indicates there is substantial, previously undetected diversity that necessitates a revision of the taxonomy of S. veraecrucis and S. veraepacis.