Abstract
Jeju Island is located at a marginal edge of the distributional range of East Asian evergreen broad-leaved forests. The low genetic diversity of such edge populations is predicted to have resulted from genetic drift and reduced gene flow when compared to core populations. To test this hypothesis, we examined the levels of genetic diversity of marginal-edge populations of Quercus gilva, restricted to a few habitats on Jeju Island, and compared them with the southern Kyushu populations. We also evaluated their evolutionary potential and conservation value. The genetic diversity and structure were analyzed using 40 polymorphic microsatellite markers developed in this study. Ecological Niche Modeling (ENM) has been employed to develop our insights, which can be inferred from historical distribution changes. Contrary to our expectations, we detected a similar level of genetic diversity in the Jeju populations, comparable to that of the southern Kyushu populations, which have been regarded as long-term glacial refugia with a high genetic variability of East Asian evergreen trees. We found no signatures of recent bottlenecks in the Jeju populations. The results of STRUCTURE, neighbor-joining phylogeny, and Principal Coordinate Analysis (PCoA) with a significant barrier clearly demonstrated that the Jeju and Kyushu regions are genetically distinct. However, ENM showed that the probability value for the distribution of the trees on Jeju Island during the Last Glacial Maximum (LGM) converge was zero. In consideration of these results, we hypothesize that independent massive postglacial colonization from a separate large genetic source, other than Kyushu, could have led to the current genetic diversity of Jeju Island. Therefore, we suggest that the Jeju populations deserve to be separately managed and designated as a level of management unit (MU). These findings improve our understanding of the paleovegetation of East Asian evergreen forests, and the microevolution of oaks.