Search

The "out of Tibet" hypothesis that suggests an important role of Tibetan highlands in the evolution of Central Asian cold-adapted fauna is a matter of debate. The long-tailed hamster Cricetulus longicaudatus is an important member of both steppe/semidesert Mongolian and high altitude Tibetan faunas. In this study, we analyze the mitochondrial variation in C. longicaudatus throughout a major part of its distribution range, perform molecular dating of the main divergence events, and model its historical range. The data reveal six genetic lineages, which are estimated to have diverged 100–200 kya. Four of the lineages are distributed in the northeastern part of the Qinghai-Tibetan Plateau. The northern part of the distribution range (Mongolia and southern Siberia) is almost entirely populated by another single lineage; its sister haplogroup is restricted to a small isolate in the Xingan Mountains. Species distribution modeling shows that the southern (Chinese) part of the geographic range of C. longicaudatus was relatively stable over the last 200 kya, while the northern (Mongolian) part demonstrated a high level of temporal variation, contracting considerably during cold glacials. The results suggest that the main center of origin of the long-tailed hamster was located in Tibet and that the northern part of the recent range was colonized through a single dispersal event at the Middle–Late Pleistocene boundary. At the same time, the comparative analysis of the fauna of Tibet and Mongolia indicates that the "out of Tibet" model is not the predominant pattern of geographic range evolution for small mammals.

A range-wide phylogeographic study of the common shrew Sorex araneus was performed using the concatenation of cytb and ND2 mitochondrial genes. The data support two major lineages, one of which (European haplogroup) is distributed continuously from West Europe to West Siberia; the other (Siberian haplogroup) is found in two non-adjacent areas: in central Siberia and in southern Finland and Karelia. Several smaller subclades (East European, Balkan, and West Siberian) were identified within the European lineage. Ancestral area reconstructions were performed under the maximum parsimony criterion using the algorithm that was designed to analyze landmark data. The results suggest that the last common ancestor of the common shrew inhabited the Balkans or the Danubian-Carpathian area. The results of demographic analyses are consistent with the hypothesis of recent expansion throughout the entire distribution range. The genetic diversity indices and the expansion time estimates were higher in the western part of the range. Cumulatively, these findings support the hypothesis that during the Last Glacial Maximum, the range of the common shrew was restricted to a single refugium in Southeastern Europe.

Using a large body of observational data on the occurrence ofSorex shrews in boreal forests, we test two models that predict the structure of small mammal communities along a gradient of increasing habitat productivity. Tilman’s (1982) model predicts a humped curve of species richness along productivity gradients. In contrast, we found a linear increase in species richness with increasing logarithm of the pooled density of shrews, which we use as a measure of habitat productivity for shrews. The model of Hanski and Kaikusalo (1989) assumes a trade-off between exploitative and interference competitive abilities, and it predicts that the size structure of small mammal communities should shift from the dominance of small species (superior in exploitative competition) in unproductive habitats to the dominance of large species (superior in interference competition) in productive habitats. Shrew assemblages show such a shift. Though it is not possible to draw definite conclusions about the role of interspecific competition from our observational data, the changing size structure of local shrew assemblages with increasing habitat productivity is a predictable feature of their community structure.