Uromys emmae, Groves & Flannery, 1994

Groves, C. P. & Flannery, Tim F., 1994, A revision of the genus Uromys Peters, 1867 (Muridae: Mammalia) with descriptions of two new species, Records of the Australian Museum 46 (2), pp. 145-169 : 159-161

Uromys emmae n.sp.

Figs 10-12 , Table 2

Type material. HOLOTYPE, AM M7200 , adult female skin and skull. Collected by Col. C.B. ~hillips of the USA Typhus Commission on Owi Island (1016'E 136°13'S), Schouten Group , Geelvinck Bay, Irian Jaya, Indonesia. The specimen was registered on 1 July 1946, and was probably collected during 1944-1945.

Etymology. For the junior author's daughter, Emma.

Diagnosis. Uromys emmae can be distinguished from all other species of Uromys except U. caudimaculatus in having a mottled section of the tail, where the light tip contacts the darker proximal part. It differs from U. caudimaculatus in the following ways: i) hindfoot shorter, broader; ii) body fur shorter, coarser; iii) white mottling on tail restricted to distal third, where it is limited in extent, and terminal 1 cm brown (in other species white tail section much more extensive); iv) rostrum relatively and absolutely shorter; v) preorbital foramen hidden behind zygomatic plates (most of it visible in other species); vi) frontals not inflated, supraorbital ridging better developed, incisive foramina shorter (only 18.4% of palate length); vii) molars relatively larger (21% of condylobasal length); viii) posterior ends of nasals extend further posteriorly; preorbital bar thinner; ix) zygomatic arches less swung downwards.

Description. The holotype skin is in good condition except that the right pinna is missing, the interorbital region has been damaged by a rat trap, and part of the venter is soiled ( Figs 10 , 11 ). The guard hairs of the dorsum are tipped with Clove Brown, while the fur is tipped with Straw. The bases of all of the dorsal fur is Dull Grey. The overall effect of the dorsal colouration is somewhat less rufescent than most U. caudimaculatus . A ring of slightly darker hairs surrounds the eyes. The hindquarters are Prout's Brown, with the tipping of the hairs being less conspicuous than anteriorly, and the guard hairs reduced in number. The venter is sparsely furred, the hairs being dirty white from the anus to the chin. The vibrissae are black. The dorsal surface of the feet are clothed in fine, pure white hairs except on the metatarsum proximal to digits 2-3, where some light brown hairs are present. The feet are unusually broad and short with smooth plantar pads ( Fig. 13 ). They are only 50.5 mm long (su) but 13.5 mm wide at the base of the fifth metatarsal. Tate (1951) gives measurements of the hindfoot for the holotypes of U. nero , U. aruensis , U. scaphax , U. prolixus , U. ductor , U. seibersi and U. sherrini . The length-width ratio for this sample is as follows: X = 0.18, range = 0.15-0.19, s.d. = 0.014 (n=8). Measurements of U. caudimaculatus held in the Australian Museum conform closely to Tate's sample. This compares with a ratio of 0.27 for U. emmae . It is highly unlikely that preparation method could have affected these measurements, both because the measurements of U. emmae taken in the field and from the study skins are similar, and because the proportions of the foot are altered little by preparation relative to other body parts. The mammary formula is 0 + 2 4 and, to judge from the size of the nipples, the animal was lactating when caught. Over the proximal 120 mm of the tail the tail scales are raised to a conical point, the tail resembling a file. The scales become progressively more flattened distally, forming irregular tesserae near the tip. Limited white mottling is present over the distal one third of the tail, but white encircles the tail only in two narrow places. The terminal 1 cm is dark. There is a single hair per tail scale, visible only under magnification, which is one half to one third of a scale in length.

The skull is cracked through just anterior to the parietal-frontal suture; damage typically inflicted by the bar of a rat trap. The teeth are moderately worn and several have been glued into their sockets. The rostrum is short and narrow relative to that of other species of Uromys . The premaxillary/maxillary suture is positioned just anterior to the zygomatic plate, the entire masseteric foramen being hidden behind the zygomatic plate when the skull is viewed from the side ( Fig. 12 ). The incisive foramina are remarkably short, narrow and parallel sided. The frontals in the interorbital region are slightly dished, and weak supraorbital ridging developed. The parietal crests are weakly developed and are subparallel. The palate and upper molars are essentially similar in morphology to those of Uromys caudimaculatus . The mesopterygoid fossa is narrow relative to U. caudimaculatus . The pterygoids are damaged. The bullae are small and resemble those of U. caudimaculatus . Except in their smaller size, the dentary and lower dentition do not differ from those of U. caudimaculatus .

Discussion. As in the case of U. boeadii , careful consideration has been given as to whether the taxon U. emmae should be recognised as a new species, or elongated hindfeet and a long skull are typical of as a subspecies of U. caudimaculatus , always a terrestrial taxa (eg, Xenuromys barbatus , Melomys difficult question to resolve when dealing with lorentzi). The combination ofa short, broad hindfoot and insular taxa for which the criterion of reproductive short snout may well indicate that U. emmae is more isolation is immaterial and sample size is small. We arboreal than U. caudimaculatus . (3) It shares decided that U. emmae should be described as a full plesiomorphic features with U hadrourus : nasofrontal species for the following reasons. 1) It differs from suture well posterior to the level of the preorbital Uromys caudimaculatus in a number of features that foramen; thin preorbital bar; uninflated interorbital are not observed to vary among the previously profile; weaker zygomatic arches; and broader feet, recognised subspecies of the latter taxa. These features recalling U. ( Cyromys ). It retains other apparently include the short, very broad hindfoot; short rostrum; plesiomorphic states not seen in either U. hadrourus or recession of the masseteric foramen behind the U. caudimaculatus , including relatively small ears zygomatic plate, and short, sparse coat. 2) Some of (39% of condylobasal length), short incisive foramina, the distinguishing features of U. emmae argue for a and relatively large teeth. In all of these features it quite different ecological niche from that of resembles U. anak and U. neobrittanicus .

U. caudimaculatus ; in particular the short, broad In addition to U. boeadii and U. emmae , the hindfoot and the shortened rostrum denote terrestrial mammalian fauna of the Geelvinck Bay considerable specialisation. Short, broad feet and a islands, including Owi Island ( Palau Awai) and Biak foreshortened snout are commonly seen in Melanesian Island (P. Biak) includes a number of endemic taxa. A murid taxa that are highly arboreal (eg, the species of highly distinctive and diminutive race of spiny Pogonomys , Chiruromys and Melomys rufescens ), while bandicoot (Echymipera kalubu philipi) has previously been named (Troughton, 1945). Troughton (1945) also described Petaurus kohlsi from Owi, but this is a junior synonym of Petaurus breviceps biacensis Ulmer, 1940 , a taxon so distinctive that it may represent a distinct species unique to Owi and Biak Islands. Two Rattus species, R. jobiensis and R. exulans , are also found on Owi. The former is known only from Owi, Biak and nearby Japen Islands, while the latter species is a widespread Human commensal. Thus the degree of endemism among the mammals of the Owi-Biak Island group (which were connected during the late Pleistocene), is very high. The occurrence of R. jobiensis on Yapen as well as Owi-Biak is intriguing, especially as it is Uromys caudimaculatus (as documented by BM 46.642, which is a large individual of cbI = 63.0, bzw = 32) is present on Japen, the two endemic Geelvinck Uromys species not being known from there. It is possible that the occurrence of R. jobiensis on Yapen could be due to accidental human transportation, as it is the only mammal taxon with such a distribution, and other species of Rattus have been widely distributed by this means in other parts of the Pacific.

Figures

Fig. 10. Study skins in dorsal view of A, holotype of U. emmae and B, U. caudimaculatus papuanus Mount Karimui area, AM M13812.

Fig. 11. Study skins in ventral views of A, holotype of U. emmae and B, U. caudimaculatus papuanus Mount Karimui area, AM M13812.

Fig. 12. Holotype cranium of U. emmae in: A - occlusal, B - dorsal and C - lateral views, and dentary in D - occlusal, and E - lateral views. F - pes of U. caudimaculatus AM M138l2, Mount Karimui, and G - holotype of U. emmae.

Fig. 13. Most parsimonious cladogram for the genus Uromys produced by Hennig86.