Abstract
Quaternary environmental changes fundamentally influenced the genetic diversity of temperate-zone terrestrial animals, including those in the Japanese Archipelago. The genetic diversity of present-day populations is taxon- and region-specific, but its determinants are poorly understood. Here, we analyzed cytochrome b gene (Cytb) sequences (1140 bp) of mitochondrial DNA (mtDNA) to elucidate the factors determining the genetic variation in three species of large moles: Mogera imaizumii and Mogera wogura, which occur in central and southern mainland Japan (Honshu, Shikoku, and Kyushu), and Mogera robusta, which occurs on the nearby Asian continent. Network construction with the Cytb sequences revealed 10 star-shaped clusters with apparent geographic affinity. Mismatch distribution analysis showed that modes of pairwise nucleotide differences (τ values) were grouped into five classes in terms of the level, implying the occurrence of five stages for rapid expansion. It is conceivable that severe cold periods and subsequent warm periods during the late Quaternary were responsible for the population expansion events. The first and third oldest events included island-derived haplotypes, indicative of the involvement of land bridge formation between remote islands, hence suggesting an association of the ends of the penultimate (PGM, ca. 130,000 years ago) and last (LGM, ca. 15,000 years ago) glacial maxima, respectively. Since the third event was followed by the fourth, it is plausible that the termination of the Younger Dryas and subsequent abrupt warming ca. 11,500 years ago facilitated the fourth expansion event. The second event most likely corresponded to early marine isotope stage (MIS) 3 (ca. 53,000 years ago) when the glaciation and subsequent warming period were predicted to have influenced biodiversity. Utilization of the critical times of 130,000, 53,000, 15,000, and 11,500 years ago as calibration points yielded evolutionary rates of 0.03, 0.045, 0.10 and 0.10 substitutions/site/million years, respectively, showing a time-dependent manner whose pattern was similar to that seen in small rodents reported in our previous studies. The age of the fifth expansion event was calculated to be 5800 years ago with a rate of 0.10 substitutions/site/million years ago during the mid-Holocene, suggestive of the influence of humans or other unspecified reasons, such as the Jomon marine transgression.