Phylogeography of the livebearer Xenophallus umbratilis (Teleostei: Poeciliidae): glacial cycles and sea level change predict diversification of a freshwater tropical fish.

The biogeography of Central America is viewed as a classic case study in understanding the impact of vicariant events on patterns of biotic dispersal. While many biogeographers have focused on community composition and geographical limits of species at broad scales across Central America, much less work has focused on post-colonization diversification patterns at finer scales. The livebearing freshwater fish Xenophallus umbratilis presents an ideal system for determining the impact of recent Earth history events on biodiversity in northern Costa Rica. Here, we test the hypotheses that marine inundation of the San Carlos and northern Limón basins during the Pliocene and Pleistocene has caused genetic fragmentation among X. umbratilis populations, despite contemporary freshwater connections. To test this idea, we collected mitochondrial (cytochrome b) sequence data in 162 individuals taken from 27 localities across northern Costa Rica. We employed a variety of analytical approaches, including: maximum parsimony and maximum likelihood, analysis of molecular variance, and demographic analysis of population size through time. We found four major clades within X. umbratilis, each geographically isolated with no shared haplotypes across drainages. Oddly, clades that occupy adjacent drainages are not always sister taxa in the phylogeny, suggesting that colonization in this species is more complex than a simple model of isolation by distance. All our results are consistent with the hypothesis that changes in sea level associated with glacial eustatic cycles have had an important effect in shaping diversification patterns in this species.