Species limits in the Andean toad genus Osornophryne (Bufonidae)

As Darwin observed, the differentiation among varieties, subspecies, and species seems, often times, arbitrary. Nowadays, however, novel tools provide the possibility of testing hypotheses of species. Using the Andean toad genus Osornophryne, we address the following questions: (1) How many species are within the genus? (2) Are morphological and molecular traits congruent when delimiting species? (3) Which morphological traits are the most divergent among species? We use recently developed methods for testing species boundaries and relationships using a multilocus data set consisting of two mitochondrial genes (12S, 16S; 1647. bp aligned matrix), one exon (RAG-1; 923 aligned matrix), and one intron (RPL3. Int5; 1410. bp aligned matrix). As another line of evidence for species delimitation, we integrated analyses of 12 morphometric variables and 10 discrete traits commonly used in amphibian systematics. The molecular and morphological approaches support the validity of most of the described species in Osornophryne. We find, however, contradictory lines of evidence regarding the status of O. angel. Within O. guacamayo, we found a genetically divergent population that, we argue, represents a new species. We consider that O. bufoniformis represents a species complex that deserves further study. We highlight the importance of incorporating morphological data when delimiting species, especially for lineages that have a recent origin and have not achieved reciprocal monophyly in molecular phylogenies. Finally, the most divergent morphological traits among Osornophryne species are associated with locomotion (finger, toes and limbs) and feeding (head), suggesting an association between morphology and the ecological habits of the species. © 2012 Elsevier Inc.