|Distribution:||Mountains of SE China in S Anhui, Zhejiang, Fujian (AMNH, MCZ, USNM), NE Jiangxi, N Guangodong, W Hubei (MCZ), Ghizhou (FMNH, USNM), Sichuan (AMNH, FMNH, MCZ, USNM), S Gansu, and W Yunnan (AMNH, FMNH, USNM) (as per specimens examined, G. M. Allen, 1940, and Zhang et al., 1997). Also in Taiwan (M.-J. Yu, 1996); NE India (Mishmi Hills in Arunachal Pradesh; Agrawal, 2000); N Burma (Ellerman, 1961, identified as cachinus; BMNH, FMNH); NE Burma and Chin Hills, EC Burma (Anthony, 1941; AMNH); N Thailand (summit Doi Inthanon, Chiengmai Prov.; Marshall, 1977a; MCZ, USNM); and extreme NW Vietnam west of the Red River (Osgood, 1932; Dang et al., 1994; FMNH, MCZ). Known altitudinal range 700-3000 m (Kaneko, 2002).|
E. melanogaster species group. Corbet (1978c) suspected that more than one species is represented in what he identified as E. melanogaster, and Corbet and Hill (1992:401) posed their treatment of the species as "very tentative." After checking large museum series, extracting E. cachinus and E. miletus along with their synonyms, we believe most synonyms listed here actually belong with E. melanogaster, allocations in most part concordant with Kaneko’s (2002) revision. Specimens of E. melanogaster are generally small-bodied with small skulls, have dark brown to blackish upperparts, slate gray underparts (washed with buff or brown in some specimens), and possess a short to medium tail (21-42 mm) relative to body length. Cranial size varies geographically, with specimens from Sichuan and Yunnan averaging smaller than those from E China, Burma, N Thailand, and NW Vietnam. Three lingual salient angles on each M3 were considered to be typical for E. melanogaster (G. M. Allen, 1940; Corbet and Hill, 1992), but number of lingual angles varies within the species, as Anthony (1941) noted for his large sample from NE Burma and confirmed by survey of M3s in AMNH and FMNH material: NE Burma, 65 have four angles, 52 have three; WC Burma, one with four angles, seven with three; Sichuan, nine with four angles, 37 with three; Ghizhou, two with four angles; Yunnan, 25 with four angles; Fujian, four with three salient angles; all specimens from N Thailand and NW Vietnam uniformly have four lingual angles. Furthermore, range in expression of the fourth lingual angle varies from a prominent projection, through progressive size reductions, to undetectable. AMNH and FMNH specimens from Yunnan, identified as E. eleusis by G. M. Allen because they have long tails and four lingual angles, are examples of E. melanogaster. Taxonomic significance of the variation in M3 lingual angles, tail length, and cranial size, as outlined here and discussed by Kaneko (2002), should be assessed by careful morphometric and molecular analyses.
The taxa bonzo, colurnus, kanoi, libonatus, and mucronatus have been traditionally included in E. melanogaster (G. M. Allen, 1940; Corbet, 1978c; Hinton, 1923). Ye et al. (2002) listed chenduensis as a form of E. melanogaster and yingjiangensis as a subspecies of E. eleusis; Corbet and Hill (1992) also excluded eleusis from E. melanogaster. Kaneko (2002) used mucronatus as the oldest name for populations we identify as E. miletus, also including chenduensis; neither he nor we have have examined the holotype of chenduensis and its referral here is provisional. Allocation of the problematic taxa aurora, eleusis, and mucronatus requires comment.
Described as a species from W Hubei (G. M. Allen, 1912; see Kaneko, 2002:62, for type locality details), G. M. Allen (1940) later arranged aurora as a subspecies of E. miletus (curiously, Allen was unaware that aurora had priority, an oversight noted by Ellerman and Morrison-Scott, 1951:668). Included in E. melanogaster in most subsequent checklists and faunal studies, excepting Ye et al. (2002), who listed it as a subspecies of E. eleusis. The type series of aurora (MCZ 7788, holotype) consists of individuals that are larger than most E. melanogaster, but they resemble in size and pelage coloration colurnus from E China and samples from farther south in Burma, N Thailand, and NW Vietnam, which all have greater average cranial dimensions than do most samples from Sichuan and Yunnan (study of MCZ specimens). The W Hubei series consists of adults and younger animals, all with four lingual angles on M3; the adults had been originally identified on skin tags as aurora, the younger individuals as melanogaster. The oldest adults of aurora are only slightly smaller than mucronatus from W Sichuan.
Like aurora, G. M. Allen (1912) had originally described mucronatus as a species but subsequently (1940:809) considered it a full synonym of E. melanogaster melanogaster. Although larger than average for E. melanogaster, all specimens in the type series of mucronatus have three lingual angles, as is usual for E. melanogaster. In addition to the holotype, another adult from the type locality had been identified as mucronatus but a younger specimen was referred to E. melanogaster. While recognizing that the holotype of mucronatus (MCZ 7789) is slightly larger than typical E. melanogaster from Sichuan, it does not possess the even larger and stockier skull with high braincase that characterizes E. miletus; contrary to Kaneko’s (2002) revision, we refer mucronatus to E. melanogaster, following G. M. Allen (1940). Kaneko also identified samples from C Burma, N Thailand, and N Vietnam as E. mucronatus, treating libonotus as its synonym. To us, material from mountains outside of China represents isolated populations of E. melanogaster that average slightly larger in size.
The taxon eleusis is based upon a sample from N Yunnan, 21 miles E Chao-tung-fu (= Zhaotong on modern maps, 27°20'N, 103°50' E). Described as a subspecies of E. melanogaster by Thomas (1911e), as also ranked by Hinton (1923, 1926a), and conventionally included therein (Corbet, 1978c; Ellerman, 1941; Ellerman and Morrison-Scott, 1951; Musser and Carleton, 1993; Pavlinov et al., 1995a). G. M. Allen (1940) regarded eleusis as a valid species because of its longer tail and M3 with four lingual salient angles, a status recently followed by others (Corbet and Hill, 1992; Wang, 2003; Ye et al., 2002; Zhang et al., 1997). Fur coloration, body length (less than 100 mm), and skull dimensions (occipitonasal length less than 25 mm) of the type series of eleusis match E. melanogaster (see G. M. Allen, 1940; Thomas, 1911e), but the tails are significantly longer (eleusis—mean = 48.9 mm, range = 46-55, n = 10; melanogaster—36.1 mm, 31-42 mm, n = 34) and every M3 has four lingual salient angles. Although tail length approaches E. cachinus, no other features are common to that species, which appears to be restricted to mountains west of the Salween River Valley, more than four degrees longitude west of Chao-tung-fu. Most E. melanogaster have only three lingual salient angles on the M3, but the number of angles varies within and among geographic samples (Kaneko, 2002; our observations). As Kaneko (2002) demonstrated, the type series of eleusis cannot be separated from other E. melanogaster except for their long tails; either these were inaccurately measured or the original sample represents the extreme in tail length for E. melanogaster.
Liu (1994) used bacular dimensions as useful for dividing Sichuan specimens into age groups. Conventional karyotype of Taiwan sample (2n = 56, FN = 58) like that characterizing species of Myodes (Harada et al., 1991), which prompted Ye et al. (2002) to exclude it from Eothenomys but leave its generic assignment as indeterminate. Ecology, altitudinal distribution and genetic substructure of Taiwanese populations extensively documented by H.-T. Yu (1993, 1994, 1995).