Sylvaemus group. The taxon witherbyi was originally described as a subspecies of Mus sylvaticus, subsequently arranged as a subspecies of A. sylvaticus (Ellerman, 1941), treated as a synonym of A. sylvaticus arianus (Ellerman and Morrison-Scott, 1951) or A. sylvaticus (Corbet, 1978c), and listed as a synonym of A. arianus (Musser and Carleton, 1993; Pavlinov et al., 1995a). Zagorodnyuk (1996a) finally identified the holotype as an example of A. uralensis. This is the species identified as A. arianus by Musser and Carleton (1993), who separated it from A. sylvaticus on the basis of its distinctive pelage and smaller body size. Recently, Zagorodnyuk (1996a), Zagorodnyuk et al. (1997), and Mezhzherin (1997a) elucidated the morphological and distributional boundaries of A. witherbyi (as arianus) and their definition incorporates the names and ranges of fulvipectus, falzfeini, and hermonensis, each of which has been treated as a species by various authors (Chelomina et al., 1998a; Filippucci et al. 1989, 1996; Mezhzherin and Zagorodnyuk, 1989; Musser and Carleton, 1993). Mezhzherin and Zagorodnyuk (1989) described falzfeini as a species, which O. Rossolimo (in litt.) considered to be the same as A. fulvipectus. Subsequent results from genetic studies led Mezhzherin and Zykov (1991) to treat falzfeini as identical with chorassanicus, which was originally described as a subspecies of A. sylvaticus; the identity of chorassanicus with fulvipectus and hermonensis is supported by allozymic data (see Filippucci et al., 2002:407). Vorontsov et al. (1992) recognized fulvipectus as the oldest name for this species; it was one of the distinct electrophoretic forms in the Caucasus revealed by Vorontsov et al. (1989). Apodemus hermonensis was described from the alpine "tragacanthic" belt at about 2000 m on Mt Hermon and was considered to be morphologically and genetically closely related to A. flavicollis, which it displaces on Mt Hermon (and possibly also Lebanon and Antilebanon mountain ranges) at elevations above 1900 m (Filippucci et al., 1989). Based on allozymic and morphological data, hermonensis was subsequently identified from Anatolian Turkey and N Iran and suspected to be synonymous with fulvipectus (Filippucci et al., 1996; Macholán et al., 2001b); Michaux et al. (2002a), however, retained the two as separate species based upon analysis of mtDNA cytochrome b sequences.

Kryštufek (2002a) has convincingly demonstrated that the holotype of arianus is an example of A. flavicollis, that the holotype of tauricus Barrett-Hamilton, 1900, is the same as hermonensis, and is an older name but preoccupied by tauricus Pallas, 1811, which is associated with A. flavicollis. He suggested that iconicus Heptner, 1948 was the oldest available name for the species, but also noted that the holotype of witherbyi Thomas, 1902 might be identical with hermonensis. The holotype is similar to hermonensis in all external, cranial, and dental traits (including the stephanodont pattern on M1), except for its longer bullae (4.8 mm) and shorter molar row (3.4 mm), and Kryštufek was reluctant to declare witherbyi and hermonensis as being the same. Two specimens we studied from the Zagros Mtns of W Iran, USNM 369849 (Baneh, Kordistan Province) and USNM 350595 (Borujerd region, Luristan Province), have all the chromatic and morphological attributes of hermonensis; molar row is 3.5 mm in the former, bullar length is 4.8 mm in the latter. Like these two, the holotype of witherbyi is at one end of the range of variation in lengths of molar row and bullae. We use witherbyi as the oldest name for what has been called arianus (Musser and Carleton, 1993; Zagorodnyuk et al., 1997; Mezhzherin, 1997a), hermonensis, fulvipectus or iconicus (Filippucci et al., 1996; Kryštufek and Vohralík, 2001); no other species of Apodemus occurs in the Fars region, and all of its other traits match those of hermonensis and the other material we assemble here.

Apodemus witherbyi may be the species represented by the sample from Qazvin, N Iraq, that Darviche et al. (1979) separated electrophoretically from A. sylvaticus and A. flavicollis, but still considered closer to the latter. A close alliance between A. flavicollis and A. witherbyi is supported by other studies employing protein electrophoresis (Filippucci et al., 1989, 2002; Macholán et al., 2001b; Mezhzherin, 1997a) and discriminant function analysis of morphometric traits (Frynta et al., 2001), but not mtDNA cytochrome b gene sequences in which A. witherbyi is isolated within subgenus Sylvaemus (Michaux et al., 2002a; reported as A. hermonensis). Included in Mezhzherin’s (1997a) revision of N Eurasian Apodemus. Compared (as A. fulvipectus) with A. ponticus and A. argenteus by restriction analysis of nuclear DNA (Chelomina et al., 1995), and with A. ponticus and A. uralensis by comparison of isozymic, chromosomal and molecular divergence (Chelomina et al., 1998a). Phylogenetic affinity of A. witherbyi requires sharper resolution, particularly concerning the nature of the relationship with A. pallipes from the Pamirs and Himalayas and the Himalayan A. rusiges.

Easternmost record of A. witherbyi is from the Quetta region of WC Pakistan, and the four specimens from there match those in the large samples from Iran in chromatic traits, cranial morphology, and molar occlusal patterns. In Iran, A. witherbyi is the most widespread Apodemus, occurs syntopically with A. flavicollis in the Zagros Mtns, and is altitudinally parapatric with A. hyrcanicus (see those accounts); these are the only three Apodemus currently recorded from Iran. Lay (1967:186) noted that a bright, small Apodemus, and a larger, darker species occur in N Iran, and considered them ecological subspecies; the brighter one is A. witherbyi, the darker A. hycanicus (see that account).

Apodemus witherbyi, A. mystacinus, and A. flavicollis are the only Apodemus that now occur in Israel (see accounts of latter two). Tchernov (1968, 1979, 1986, 1996) has identified fossilized Apodemus recovered from Pleistocene Israeli cave sediments. Three size classes of molars are represented. Tchernov identified the largest as A. mystacinus, the next in size as A. flavicollis, and the smallest as A. sylvaticus (Tchernov, 1986:263), which he regarded as part of the modern Israeli fauna. He (Tchernov, 1986) also synonymized Bate’s (1942b) Pleistocene A. caesareanus with A. sylvaticus. The latter, however, does not occur in Israel or anywhere in the Middle East (Filippucci et al., 1989; Macholán et al., 2001b), and Kryštufek (2002, in litt.) thinks the smallest morphotype in the Pleistocene samples represents A. witherbyi, not A. sylvaticus. We agree. Certainly the illustrations of caesareanus molar rows (Tchernov, 1968:52, 54) exhibit some of the traits characteristic of A. witherbyi and the molars are similar in dimensions (Filippucci et al., 1996; Kryštufek, 2002a).