Molecular phylogeny and taxonomy of wood mice (genus Apodemus Kaup, 1829) based on complete mtDNA cytochrome b sequences, with emphasis on Chinese species

Phylogenetic relationships among 15 species of wood mice (genus Apodemus) were reconstructed to explore some long-standing taxonomic problems. The results provided support for the monophyly of the genus Apodemus, but could not reject the hypothesis of paraphyly for this genus. Our data divided the 15 species into four major groups: (1) the Sylvaemus group (A. sylvaticus, A. flavicollis, A. alpicola, and A. uralensis), (2) the Apodemus group (A. peninsulae, A. chevreri, A. agrarius, A. speciosus, A. draco, A. ilex, A. semotus, A. latronum, and A. mystacinus), (3) A. argenteus, and (4) A. gurkha. Our results also suggested that orestes should be a valid subspecies of A. draco rather than an independent species; in contrast, A. ilex from Yunnan may be regarded as a separate species rather than a synonym of orestes or draco. The species level status of A. latronum, tscherga as synonyms of A. uralensis, and A. chevrieri as a valid species and the closest sibling species of A. agrarius were further corroborated by our data. Applying a molecular clock with the divergences of Mus and Rattus set at 12 million years ago (Mya) as a calibration point, it was estimated that five old lineages (A. mystacinus and four major groups above) diverged in the late Miocene (7.82-12.74 Mya). Then the Apodemus group (excluding A. mystacinus) split into two subgroups: agrarius and draco, at about 7.17-9.95 Mya. Four species of the Sylvaemus group were estimated to diverge at about 2.92-5.21 Mya. The Hengduan Mountains Region was hypothesized to have played important roles in Apodemus evolutionary histories since the Pleistocene. (C) 2004 Elsevier Inc. All rights reserved.