Differentiation of the Akodon mollis Species Group (Rodentia: Cricetidae: Sigmodontinae) in the Andes of Peru.

Abstract

The mechanisms responsible for the species composition of montane biotas remain poorly understood. Parapatric and allopatric speciation, in the forms of the Ecological Gradients and Montane Vicariance hypotheses respectively, have been suggested to explain the diversification of Andean fauna. This dissertation reviews the possible role of these mechanisms in the speciation of mammals in Andean habitats. The soft grass mouse, Akodon mollis, was chosen as the test organism because of its wide elevational and latitudinal range. A. mollis inhabits the Andean mountains of North and Central Peru, and the Andes of Ecuador; however, this work focuses exclusively on Peruvian populations. Geometric morphometrics and environmental analyses revealed that despite the wide elevational range of these mice, their skull morphology does not vary consistently across elevation except for a sharp change at the treeline (the edge of the páramo/puna, ~4000m). Partial least squares and multivariate multiple regression analyses showed a significant association between skull shape and environment. The most important variables were associated with temperature, and these almost perfectly correlated with elevation. Nevertheless, a clear association between morphology and elevation was not found. Phylogenetic analyses of the mitochondrial Cytochrome b gene showed spatial structuring of A. mollis populations. Deeper lineages at the southern limit of the species coincide with the oldest part of the Andes, while shallower clades were found in the more recent northern Andes. An unambiguous test of the Ecological Gradient hypothesis was not possible; however, Montane Vicariance was suggested in two instances where mice at similar elevations on different mountains were closely related. Geographic distance did not explain patterns of genetic variation in Andean populations. Finally, AMOVA suggested differentiation between populations at higher southern elevations (>3500) and lower northern elevations (<3500m). More extensive sampling, including more elevational belts and mountain replicates, additional molecular markers, and the inclusion of isolation by resistance methods, might help to shed further light on speciation by Andean mammals.