Myodini. Closely related to Alticola, Caryomys, Hyperacrius, and Myodes, all of which are usually placed in Clethrionomyini (Gromov and Polyakov, 1977; Hooper and Hart, 1962; Koenigswald, 1980) or the subtribe Myodina, Prometheomyini (Pavlinov and Rossolimo, 1998; Pavlinov et al., 1995a). Under Eothenomys, Corbet (1978c) included Phaulomys (placed by Ye et al., 2002, in Clethrionomys and by Japanese systematists in Eothenomys), Aschizomys (listed here in Alticola), and Caryomys (treated as a separate genus). Anteliomys was proposed as a subgenus of Microtus (Hinton, 1923), then tranferred to Evotomys (Hinton, 1926a), and finally treated as a subgenus of Eothenomys (Osgood, 1932; G. M. Allen, 1940); recently, Ye et al. (2002) also observed subgeneric usage for Anteliomys, although they suggested its reinstatement as genus. Phallic morphology of E. melanogaster and E. miletus is described by Yang et al. (1992) and contrasted with Caryomys eva.
Molar patterns and the lack of pectoral and postaxillary mammae are among the traits distinguishing Eothenomys from Myodes (see account of M. shanseius), but most systematists emphasize the degree of molar hypsodonty in adults, rootless (evergrowing) in Eothenomys versus rooted in Myodes (Corbet and Hill, 1992; Hinton, 1926a; Kaneko, 1990, 1992c, 1996b). Rootless molars, however, have apparently been acquired independently in some lineages of Myodes (see that account). The endemic species on the Korean Peninsula (revised by Kaneko, 1990, as E. regulus), the N Chinese shanseius (transferred to Eothenomys by Kaneko, 1992c), and all the Japanese endemics formerly considered in Eothenomys or Phaulomys (andersoni, imaizumii, and smithii) have rootless molars, and their allocation to Eothenomys has not been questioned until recently (all have also been regarded as species of Clethrionomys, here = Myodes; see Ellerman and Morrison-Scott, 1951). Phylogenetic analyses of mitochrondrial and nuclear ribosomal DNA convincingly demonstrated that these Korean and Japanese endemics are closely related to species of Myodes and form a monophyletic group that excludes Eothenomys melanogaster (Suzuki et al., 1999b; Wakana et al., 1996).
Does Eothenomys simply represent a southern continental extension of Myodes? The Palearctic distribution of Myodes mostly lies above 45°N latitude, with southeastern projections onto the Korean Peninsula, adjacent mainland China, and the main Japanese isls (see Gromov and Erbajeva, 1995). In contrast, species belonging to Eothenomys, as we view it, are distributed in the mountains of S China, south of about 40°N latitude, with limited extensions into NE India, N and WC Burma, N Thailand, NW Vietnam (west of the Red River), and probably EC Burma (east of the Salween River), and N Laos. In phylogenetic analyses of mitochondrial and ribosomal DNA sequences, melanogaster (type species of Eothenomys) was nested between rutilus (type species of Myodes) and the monophyletic Myodes rufocanus cluster (andersoni, smithii, imaizumii, regulus, rex) (Suzuki et al., 1999b). Systematists should examine other species of Eothenomys in a phylogenetic framework and revisit the significance of rootless versus rooted molars in our generic constructs.
Corbet and Hill (1992:401) wrote of Eothenomys that "The southern forms are badly in need of revision, especially to establish the limits of E. melanogaster in which Ellerman & Morrison-Scott (1951) included a great diversity of forms that seem unlikely to be conspecific." Thanks to Kaneko’s careful studies (1992c, 1996b, 2002), most of those species have been delineated by diagnostic morphological traits and their geographic distributions empirically documented. Kaneko’s revisions and the species we recognize here (based on specimen examination) in some ways reflect the treatment of Chinese Eothenomys by G. M. Allen (1940), which for the time was an insightful portrayal of species diversity. Consult also the review by Ye et al. (2002), which derives from an earlier revision that we have not seen.
Anteliomys is sometimes used as a subgenus but authors have differently allocated the species (compare G. M. Allen, 1940, and Pavlinov et al., 1995a). We arrange the species into two groups, based largely on molar patterns as described by Kaneko (1996b): E. melanogaster group (E. cachinus, E. melanogaster, E. miletus) and E. chinensis group (E. chinensis, E. custos, E. olitor, E. proditor, E. wardi). Kaneko (1996b:106) also recognized two groups of species occurring in Sichuan and Yunnan: one consists of the five species in his E. chinensis group; the other "is the E. melanogaster group, which includes confinii, eleusis, fidelis, miletus and mucronatus." Kaneko (2002) has recently revised this latter complex, but our interpretation of species limits and distributions, based upon specimen study and pertinent literature, differs somewhat from his. These two assemblages generally mirror former subgeneric groups; their apparent monophyly should be tested with analyses of a variety of morphological, chromosomal, and molecular data. Eothenomys dates from the late Pliocene in the Sichuan-Guizhou region, S China, where it is also represented by Pleistocene fossils identified as extinct E. praechinensis and living E. chinensis (both in E. chinensis species group), and by samples from late Pliocene and Pleistocene cave sediments of extant E. melanogaster (Zheng, 1993; Zheng and Li, 1990).