Demographic processes, refugia and dispersal routes within the sigmodontine rodent assemblage from South American Pampas during the Pleistocene

Abstract

The geographical distribution of a taxon is the result of the interaction among geological, environmental and ecological processes in a space-time dimension. Theoretically, co-distributed species should have congruent phylogeographic patterns as a result of being subject to common environmental and geological changes. However, these patterns may be a_ected by ecological di_erences among the species, such as dispersal abilities and food or habitat preferences, and by the possible instability of the assemblage itself through time. South America presents one of the most extended subtropical-temperate grasslands in the world, the Pampas. In the Pleistocene, this biome would have experienced expansions over forest areas during cold and generally dry (glacial) periods and retractions during warm and generally humid (interglacial) climates. The sigmodontine rodents inhabiting the Pampas are one of the best known rodent assemblages in South America, being a good study model to analyze the impact of the Pleistocene climate changes in the evolutionary history of the region through comparative phylogeography. We studied the phylogeographic patterns of seven rodent species inhabiting the Pampas; our aims were to evaluate the concordance of phylogeographic patterns among them, assess if they experienced congruent demographic changes, estimate the time in which this demographic change occurred, and infer if the dispersal routes are shared among them. Four of the seven species evidenced recent demographic expansion. However, no temporal synchrony was observed since Oligoryzomys nigripes and Oxymycterus nasutus experienced a population increase about 250,000 years ago, whereas for Calomys musculinus and Oligoryzomys avescens it was about 125,000-100,000 years ago. We observed three common centers of origin for the species populations, which would have acted as grassland refugia where species would have developed isolated lineages. The reconstruction of lineage dispersal showed common dispersion routes for all of the species, following the grassland expansion to North, West and South. Our results support the idea that recent glacial cycles had a more moderate impact in South America because of its latitude and continentality, and that open biomes such as grasslands would have been more stable in comparison with tropical forests. In this context, ecological di_erences could have gained a greater prominence, producing a more complex scenario, like the one observed in the present study.