Diversificación de las especies altoandinas del grupo Pristimantis conspicillatus (Anura: Brachycephalidae): fauna anura primitiva o derivada de la fauna de tierras bajas

Abstract

Abstract One of the fundamental goals of biology is to understand the patterns and processes underlying diversification. These can be studied using phylogenies. Through phylogenies, it is possible to reconstruct the evolutionary history of a group, even in the absence of paleontological information. Given the advances in molecular biology, it is now possible to generate phylogenies for a large number of groups. Since 2005, the amphibian phylogenies involving molecular data increased markedly in number, presenting dramatic change in the systematic and taxonomy of the most groups. The genus Pristimantis Jiménez de la Espada, 1870, was resurrected for the molecular clade including most former South American Eleutherodactylus Duméril and Bibron, 1841 by Heinicke et al., 2007 and Hedges et al., 2008. The genus Pristimantis was subdivided in several species groups and series by Hedges et al (2008). The former E. conspicillatus group is now Pristimantis conspicillatus group, this group contains most of the species recognized in the “E conspicillatus” group by Lynch and Duellman (1997). Hedges et al. (2008) redefined the content of the E. conspicillatus group proposed by Lynch (1986, 1994) and Lynch and Duellman (1997), and allocated 37 species into the group and showed support for the monophyly of this species group. Padial et al. (2014), in their molecular systematics of terraranas, found that P. conspicillatus group could be rendered monophyletic by making a few changes. Members of the Pristimantis conspicillatus group have a large and interesting distribution. These species occur from Costa Rica to Bolivia across the Andean cordillera, the Amazon and Cerrado, the Guianas and in the Atlantic Forest; and have a range from the lowlands up to about 3000 m.a.s.l. Given that species of the P. conspicillatus group are distributed in the highlands and lowlands, we used frogs of the P. conspicillatus group as a model system to address the following questions: 1) Are lowland species of P. conspicillatus group more closely related to other lowland species or to highland species?, 2) What was the most probable ancestral distribution area for P. conspicillatus group?, 3) Does the radiation of P. conspicillatus group follow a south to north pattern of speciation?, 4) Can the P. conspicillatus diversification be attributed to allopatric/vicariant scenario?. To evaluate mode and area relationships of the diversification of P. conspicillatus species group, we reconstructed phylogenetic relationships using the maximum possible number of species that have been assigned to the P. conspicillatus group sensu Lynch 1994, Lynch and Duellman, 1997, Hedges et al. 2008 and Padial et al., 2014. We used an extensive taxon (617 terminals), data set of 17.365 aligned sites of mitochondrial and nuclear DNA, and 32 morphological characters. Phylogenetic analyses were performed using parsimony analysis and maximum likelihood (ML). Pristimantis conspicillatus group was recovered as a monophyletic under both parsimony and ML analysis. We provide a new definition of the P. conspicillatus group based on external morphology and cranial osteology. With a phylogenetic hypothesis for a monophyletic group containing a mixture of lowland and montane species, coupled with detailed geographic and elevational range data, it is possible to understand diversification of montane fauna The present study illustrate how the different geological and climatic events of the Neotropics shaped, at different levels of the phylogeny. According to the new circumscription of the P. conspicillatus group showed in this study, most of the species are distributed in the Amazon basin and Andean slopes of Peru and Bolivia and a few species are distributed on northern Andes. Our results identify Amazonia as the most probable ancestral range of the most recent common ancestor of P. conspicillatus group. Amazonian origin of the P. conspicillatus group was followed by several events of colonization to other habitats. The highland species of P. conspicillatus group are more related to lowland species, where the montane species are derived, supporting the hypothesis of montane regions largely as ‘species pumps’. Within the P. conspicillatus group there is not a generalized patter of speciation, ecological speciation and vicariance models explain the diversification of the P. conspicillatus group.