Type species of Phaulomys Thomas (1905b), diagnosed as a subgenus of Evotomys. Imaizumi (1949) transferred Phaulomys to Eothenomys, an action largely ignored (e.g., Ellerman and Morrison-Scott, 1951) until adopted by Corbet (1978c) and Aimi (1980). Phaulomys was reinstated as a genus, containing andersoni and smithii, by Kawamura (1988), who thought the two extant species with rootless molars (a key trait of Eothenomys) were derived from the middle Pleistocene japonicus, which has rooted molars. Molar and external traits also led Tanaka (1971) to recognize Phaulomys as a genus distinct from Eothenomys, primarily because smithii combined characteristics of both Myodes and Eothenomys. Chromosomal analyses of smithii suggest that Myodes and Phaulomys are derived from a common ancestor (Ando et al., 1988).
We (1993:532) previously recognized Phaulomys because "The removal of Japanese Phaulomys from Eothenomys, whose species-diversity centers in the mountains of China, and its proposed association with Clethrionomys [= Myodes] is zoogeographically plausible and outlines a precise hypothesis that can be tested by analyses of other data sets." In just such a taxonomically and geographically broad study, using mitochondrial and nuclear ribosomal DNA, Suzuki et al. (1999b) demonstrated: 1) that andersoni, imaizumii, and smithii are closely related to Japanese and Asian M. rufocanus and to Korean M. regulus; 2) that this clade is phyletically remote from M. rutilus and Eothenomys melanogaster; and 3) that no monophyletic group corresponding to Phaulomys as a separate genus is supported. They concluded (1999b:520) that the allocations of Japanese red-backed voles to Myodes, Eothenomys, or Phaulomys should be reevaluated. The incorporation of smithii into Myodes had earlier been urged by Yoshida et al. (1989), based on chromosomal and allozymic data, and was so arranged by Pavlinov et al. (1995a). The molecular relationships reported by Suzuki et al. imply that the evergrowing molars in andersoni and smithii are independently derived from a rooted Myodes ancestor endemic to Japan, as earlier hypothesized by Kawamura (1988).
Species revised, as Eothenomys, and many name combinations traced by Aimi (1980). Morphology, altitudinal and geographic distribution, taxonomic comparisons, and nomenclatural history meticulously investigated by Kaneko (1992b, 1994, 1996a). Synonymy of kageus demonstrated by Aimi (1980), Kaneko (1985) and Ando et al. (1988); later Kaneko (1994, as Eothenomys) recognized kageus as occurring in E Honshu and smithii in W Honshu. The taxon okiensis, described as a subspecies of rufocanus, was placed in synonymy by Aimi (1980). Chromosomal data widely available, in comparisons among arvicoline species and within samples of smithii (Ando et al., 1988, 1991; Iwasa and Tsuchiya, 2000; Iwasa et al., 1999a, b; Tsuchiya, 1981; Vorontsov et al., 1980; Yoshida et al., 1989). See Kawamura (1991, 1994) for discussion of Holocene occurrence in archeological sites on Honshu, Shikoku, and Kyushu. Both M. smithii and M. andersoni are discussed by Dobson (1994, as Phaulomys) in context of elucidating patterns of distribution in Japanese land mammals.