TESTING PHYLOGEOGRAPHIC AND NICHE MODEL RECONSTRUCTIONS OF GLACIAL REFUGIA USING THE FOSSIL RECORD OF NORTH AMERICAN SMALL MAMMALS

Recent work in phylogeography and environmental niche modeling has tried to interpret the locations of Pleistocene refugia. Phylogeographic efforts have focused on interpreting area cladograms based on the lineage structures of populations, with the assumption that all living populations of a temperate species will have spread from a refugium since the last glacial maximum. Niche modeling attempts, in contrast, have applied environmental constraints derived from the current ranges of species to reconstructed climate maps of the last glacial maximum. These workers argue that, if niches are conserved, the species could only have lived in those locations at that time. We have tested some of these reconstructed glacial refugia using fossil data compiled in the FAUNMAP II database. Our results do not consistently support or reject any of the phylogeographic or niche modeling methods, supporting the idea that species behave individualistically in response to climate change: some follow the assumptions of these methods and some decisively do not. The fossil distribution of Blarina brevicauda largely agrees with the phylogeographic hypothesis except along coastal TX/LA, but it does not agree with the climate envelope hypothesis. The fossil distribution of Glaucomys sabrinus largely agrees with both hypotheses, but the reconstructed distributions are so large that they do not provide a good test. The fossil distribution of Glaucomys volans does not align with either type of hypothesis of refugia. The fossil record of Lepus arcticus aproximately agrees with the climate envelope hypotheses, but does not fit at all with phylogeographic hypotheses. Martes americana lines up well with phylogeographic hypotheses and less well with climate envelope hypotheses. Myodes (Cleithrodontomys) americana does not line up with either hypothesis. The database is limited to the USA and Canada, so we are unable to completely test some of the ranges, which were projected to extend into Mexico.