GENETIC-STRUCTURE OF GYMNURES (GENUS HYLOMYS, ERINACEIDAE) ON CONTINENTAL ISLANDS OF SOUTHEAST-ASIA - HISTORICAL EFFECTS OF FRAGMENTATION

Eustatic sea level changes during Pleistocene climatic fluctuations pr oduced several cycles of connection-isolation among continental island s of the Sunda shelf. To explore the potential effects of these fluctu ations, we reconstructed a model of the vicariant events that separate d these islands, based on bathymetric information. Among many possible scenarios, two opposite phylogenetic patterns of evolution were predi cted for terrestrial organisms living in this region: one is based on the classical allopatric speciation mode of evolution, while the other is the outcome of a sequential dispersal colonization of the archipel ago. We tested the applicability of these predictions with an analysis of sequence variation of the cytochrome b gene from several taxa of H ylomys. They were sampled throughout SE-Asia and the Sunda islands. Hi gh levels of haplotype differentiation characterize the different isla nd taxa. Such levels of differentiation support the existence of sever al allopatric species, as was suggested by previous allozyme and morph ological data. Also in accordance with previous results, the occurrenc e of two sympatric species from Sumatra is suggested by their strongly divergent haplotypes. One species, Hylomys suillus maxi, is found bot h on Sumatra and in Peninsular Malaysia, while the other, H. parvus, i s endemic to Sumatra. Its closest relative is H. suillus dorsalis from Borneo. Phylogenetic reconstructions also demonstrate the existence o f a Sundaic clade composed of all island taxa, as opposed to those fro m the continent. Although there is no statistical support for either p roposed biogeographic model of evolution, we argue that the sequential dispersal scenario is more appropriate to describe the genetic variat ion found among the Hylomys taxa. However, despite strong differentiat ion among island haplotypes, the cladistic relationships between some island taxa could not be resolved. We argue that this is evidence of a rapid radiation, suggesting that the separation of the islands may ha ve been perceived as a simultaneous event rather than as a succession of vicariant events. Furthermore, the estimates of divergence times be tween the haplotypes of these taxa suggest that this radiation may act ually have predated the climatic fluctuations of the Pleistocene. Furt her refinement of the initial palaeogeographic models of evolution are therefore needed to account for these results.