Abstract

Mito-nuclear discordance is a well-recognized phenomenon; however, the limited use of nuclear data in rodent systematics has revealed few documented cases. As such, the extent to which discordant gene trees mislead taxonomic schemes remains unclear. Here, we document an example of such discordance in a taxonomic assessment of species boundaries within the subgenus Angelomys of the genus Abrothrix, based on mitochondrial, nuclear (ultraconserved elements, UCEs), and morphological data. Our analyses revealed strong discordance between the mitochondrial tree and the species tree emerging from the UCE dataset. The latter recovered three main lineages within Angelomys, each phenotypically distinct. Therefore, we propose that the taxonomic scheme that best represents the alpha taxonomy of the subgenus Angelomys recognizes three species rather than the two recently proposed. The three species we recognize are Abrothrix (Angelomys) andina, distributed from the highlands of Santiago in central Chile northward along the Pacific coast to southern Peru; Ab. (An.) dolichonyx, a highland species ranging from Mendoza in west central Argentina northward through the high Andean regions of northern Argentina, northern Chile, western Bolivia and southern Peru; and Ab. (An.) olivacea, distributed from lowland areas of central Chile and Mendoza (Argentina), extending southward through the temperate forests, Patagonia, Tierra del Fuego, nearby islands, and reaching Cape Horn. Finally, we designate a lectotype for Hesperomys dolichonyx (= Ab. dolichonyx) and restricted the type locality of Mus andinus (= Ab. andina) to a locality within the general area indicated in the species description and from which the species has been recorded.