In small mammals, mid-elevation peaks in species richness are commonly seen along elevational gradients. The processes underlying this pattern, however, are often uncertain. The Philippines offer a unique opportunity to study the biogeography and diversification of small mammals, as home to two broad radiations of endemic murine rodents. We focus on one of these radiations, the earthworm mice (Chrotomyini), to investigate the elevational pattern of diversification. Previous detailed, standardized elevational transect studies conducted across the Philippines have provided extensive information on both the elevational ranges and details of the habitats in which this clade is found. The Philippines are highly topographically heterogeneous, and while primary habitat types are consistent (lowland, montane, and mossy forests), transitions between them occur at variable elevations, since mountain height and degree of isolation can impact the specific elevations at which habitat transitions take place. Using both elevational range and habitat association as characters, we conducted maximum likelihood ancestral character state analysis on maximum likelihood and Bayesian inference trees of the earthworm mice clade in order to evaluate their pattern of diversification along elevational gradients as well as how well elevation may function as a proxy for habitat.
Results/Conclusions
Reconstruction of ancestral states using elevational range showed a single elevational zone preference for all but one modern species but yielded equivocal results for earthworm mice overall. Reconstructions using habitat, however, showed an association with mossy forest as the clade’s ancestral state, as well as a general lack of single-habitat specialization in this clade. The disparate outcomes of the ancestral state reconstructions illustrate the danger of using elevation as a proxy for habitat, especially in studies that incorporate data from more than one mountain. In the Philippines, a system with peaks of variable height and isolation, it is clear that absolute elevation does not dictate the forest transition zones on mountains. Massenehrebung effects and local edaphic conditions make the delimitation of elevational ranges for multi-peak studies somewhat arbitrary, as they may fail to reflect actual habitat distributions across these mountain ranges. This is particularly relevant to studies that make use of museum specimens, which often have only locality information (including elevation) and no data regarding actual habitat.