Marsupials and Rodents of the Admiralty Islands, Papua New Guinea H. L. Ronald Pine, Andrew Mack, and Robert M. Timm Abstract We provide the first account of all non-volant, non-marine mammals recorded, whether reliably, questionably, or erroneously, from the Admiralty Islands, Papua New Guinea. Species recorded with certainty, or near certainty, are the bandicoot Echymipera cf. kalubu, the wide¬ spread cuscus Phalanger orientalis, the endemic (?) cuscus Spilocuscus kraemeri, the endemic rat Melomys matambuai, a recently described species of endemic rat Rattus detentus, and the commensal rats Rattus exulans and Rattus rattus. Species erroneously reported from the islands or whose presence has yet to be confirmed are the rats Melomys bougainville, Rattus mordax, Rattus praetor, and Uromys neobrittanicus. Included additional specimens to those previously reported in the literature are of Spilocuscus kraemeri and two new specimens of Melomys mat¬ ambuai, previously known only from the holotype and a paratype, and new specimens of Rattus exulans. The identity of a specimen previously thought to be of Spilocuscus kraemeri and said to have been taken on Bali, an island off the coast of West New Britain, does appear to be of that species, although this taxon is generally thought of as occurring only in the Admiralties and vicinity. Summaries from the literature and new information are provided on the morphology, variation, ecology, and zoogeography of the species treated. Key words: Austro-Papuan region, biogeography, Bismarck Archipelago, Echymipera kalubu, marsupial dentition, morphology, Phalanger orientalis, Phalangeridae, Rattus detentus, taxonomy Introduction The Admiralty Islands, a group of 18 islands of high biological importance, has been highlighted by low elevation and located 275-300 km north of the Johns (1993) and Wikramanayake et al. (2002). island of New Guinea, and politically part of Papua New Guinea, have a poorly known mammal fauna. The Admiralties have a number of vertebrates Manus Island is the largest of the Admiralties and it that are endemic or of limited distribution elsewhere, has been designated by the World Wildlife Fund as an including the cuscus Spilocuscus kraemeri (Schwarz, area of high biological importance (WWF 2014). Some 1910) (Phalangeridae); the bats: Dobsonia anderseni four-fifths of Manus remained forested at least until the Thomas, 1914 (Pteropodidae), Pteropus admiralita- mid-1990s, but many of the smaller adjacent islands tum Thomas, 1894 (Pteropodidae), and Emballonura have been converted into coconut plantations (Rannells serii Flannery, 1994 (Emballonuridae); and the rats: 1995). The original vegetation of the Admiralty Islands Melomys matambuai Flannery, Colgan, and Trimble, is evergreen forest, including tropical and subtropical 1994 (Muridae) and the recently described endemic moist broadleaf forest (Wikramanayake et al. 2002); species Rattus detentus Timm, Weijola, Aplin, Flan¬ however, the only significant forests remain on Manus. nery, and Pine, 2016 (Muridae). Three of the six bird The need to study the Tropical and Subtropical Moist species endemic to the Admiralties have been listed Broadleaf Forests, especially in central Manus, from as Vulnerable on the IUCN Red List—Manus Fantail Mt. Dremsel to the northern coast, in order to protect (Rhipidura semirubra). Superb Pitta {Pitta superba), the Calophyllum and Sararanga forests as an area of and Manus Masked Owl (Tyto manusi). The large, 2 Occasional Papers, Museum of Texas Tech University brightly-colored land snail Papustyla pulcherrima lands, summarize and comment on the widely scattered is listed as endangered owing to over-collecting for and sometimes obscure and/or misleading published the commercial trade. The recent description of the literature, and to report on newly acquired specimens. endemic Admiralties Rat Rattus detentus adds further In particular, we provide: new information on coat weight to calls for continued study and protection of color and dentition; cranial images of certain species; the unique flora and fauna of these islands. the first color photograph of the skin of a female Spi- locuscus kraemeri,; and the first such photograph of the Our goals herein are to document what is known skin of a Melomys matambuai and a comparison of that about the marsupials and rodents of the Admiralty Is¬ species with M. rufescens. Materials and Methods English names.—English names follow Flannery premolars in the maxilla, the first two will be designated (1995b), except for that of Rattus detentus, which fol¬ as PI and P2. The minute, functionless teeth in the lows Timm et al. (2016). Not all English names are mandible, posterior to the large procumbent anterior in keeping with the principles recommended for their incisors, which might be interpreted as constituting construction by Duckworth and Pine (2003). one or more other incisors, possibly a canine, and one or more premolars, will simply be referred to as “uni¬ Taxon sampling.—Specimens were examined cuspids.” The first tooth after the “unicuspids” will be at and/or on loan from the following museums (acro¬ designated as p3 and the teeth posterior to this tooth will nyms for U.S. institutions follow Hafner et al. 1997): be designated as ml, m2, m3, and m4. This clarification American Museum of Natural History (AMNH), New is given here because various hypothetical schemes as York, New York; Bernice P. Bishop Museum (BPBM), to the homologies and “actual identity” of various teeth Honolulu, Hawaii; Field Museum (FMNH), Chicago, in marsupials have caused inconsistency resulting in Illinois; Los Angeles County Museum (LACM), Los various conflicting, confusing, and misleading treat¬ Angeles, California; Natural Science Research Labora¬ ments. Our scheme of terminology is not intended to tory at the Museum of Texas Tech University (TTU), involve hypothetical considerations in regard to homol¬ Lubbock, Texas; Papua New Guinea National Museum ogy and phylogeny but is adopted purely for effective & Art Gallery (PNGMAG), Port Moresby; University communication, convenience, and comprehensibility. of Kansas Natural History Museum (KU), Lawrence, An example of what we are trying to avoid can be Kansas; and University of Wisconsin Zoological Mu¬ found in Nowak (1999:89), in which he presents the seum (UWZM), Madison, Wisconsin. dental formula of Phalangeridae as “i 3/1-2, c 1/0-1, pm 2-3/3, m 5/5,” and gives the total number of teeth Morphologic evaluations.—All linear measure¬ as 40-46, citing Archer (1984a) as his authority. No ments presented herein are in millimeters, and masses more than 40 teeth, however, are normally present in a are given in grams. Skull/dental measurements were phalangerid skull. The discrepancy is caused by Nowak taken with dial calipers to the nearest 0.1 mm. Those believing that both a lower canine and a lower second for specimens of Rattus are as defined by Taylor et al. incisor can exist in the same dentary, although actually (1982); those for Melomys follow those of Thomas only one tooth is involved, and it is uncertain whether (1905) and Hershkovitz (1962) or are self-explanatory. to call it an incisor or a canine. Also, Archer (1984a, b) External measurements are given as recorded on the regarded the deciduous teeth, above and below, which tags by collectors. are displaced by P3 and p3, as representing an Ml and ml. Thus, Archer maintained that there were, in fact, Tooth terminology in Phalangeridae.—For the five teeth above and below which could be regarded purposes of this paper, adult phalangerids will be treated as molars, but all five would not be present after the as having no more than three premolars and four molars putative “Ml” and “ml” were replaced by the P3 and in the maxilla and the last upper premolar above and p3, but Nowak erroneously counted them in the adult below will be designated as the third. If there are three dentition. Although they did not explicitly say so, it ap- Pine et al.—Marsupials and Rodents of Admiralty Islands 3 pears that Lidicker and Ziegler (1968) reported all of the last premolars as fourth premolars, apparently believ¬ teeth that we treat as “unicuspids” as being premolars, ing that the “missing” most anterior premolar should because they treat the deciduous teeth and permanent be numbered premolar number 1. Species Accounts PERAMELEMORPHIA this Perameles cockerelli [see also Thomas (1914)]. Peramelidae The correct spelling is cockerellii (see Ramsay 1877). Echymipera cf. kalubu (Fischer, 1829) Echymipera kalubu cockerelli [sic] is also one of the three subspecies recognized by Groves (2005b), who Fischer (1829:68) named “P\erameles\P Kalubu” did not mention the Admiralties as within the species’ from “insula Waigiou” (= Waigeo) and attributed the range. Williams (1999) reported E. kalubu as dating name to Lesson (“Less, in Diet, class. XIII. P. 200”). back to “12400 +/- 480 BP” (also given as “at approx We follow Husson (1955) in regarding Fischer as the 12,000 BP”) at the Pamwak archeological site on author of the name. Manus. This corresponds in time with archeological evidence that would mean that the inhabitants had Thomas (1914) reported a specimen of'Echymi¬ become more seafaring. Williams wrote (p. 248) that pera cockerelli [sic] from “Admiralty Island,” presum¬ “The implication of the bandicoot presence is that it ably meaning Manus. was brought to the island by people accidentally or as a food source.” Helgen and Flannery (2004:831), how¬ Flannery (1995b:68) wrote “Two specimens are ever, wrote “... Echymipera kalubu is common at all known from Manus: BMNH 14.4.1.24, a male col¬ levels throughout the Pamwak deposit [“a rockshelter lected by Meek and Eichhorn on 23 September 1913 site in southern Manus”] back to about 12,400 years [the specimen reported by Thomas (1914)]; and BZM ago; it could be native to Manus ... and its taxonomic [apparently a lapsus for MZB = Museum Zoologicum status deserves study.” Aplin et al. (2015:56) wrote Bogoriense] 15061, a male collected by H. Schodde that clarification of the identity of this animal “is a [apparently a lapsus for H. Schoede, see Aplin et al. high priority for future research.” However, Dickman (2015), and earlier authors] at Seeadlerhafn. There is (2015) seems to have little doubt as to the specific also a specimen (unregistered) held in the collections identity of the Manus bandicoot, because he assigns it of the Museum fur Naturkunde, Berlin, from the nearby the subspecies E. k. cockerelli [sic] Ramsay, 1877. He island of Lou. [Aplin et al. (2015) misquote Flannery hypothesized that it reached the Admiralties by human in regard to these specimens]. Preliminary investiga¬ agency in the late Pleistocene. tions of archeological sites on Manus suggest that it has been introduced there ... .” Earlier, Fredericksen Timm et al. (2016) reported that snares are set by et al. (1993) had written, in regard to evidence from local trappers on Manus, around Kawaliep Village, in the Pamwark archaeological site, that their findings hopes of catching these bandicoots and rats and that suggested that the bandicoot had first appeared there both are eaten. However, Aplin et al. (2015) indicated at about 11,000 BP, when the inhabitants had shifted that this species appears to be scarce on Mt. Sobomu, from flaked “stone” to obsidian artifacts, and that the stating that landowners in 2014 claimed that they are animal had presumably been introduced by humans, present in the forest, primarily in rocky areas and only from the mainland of New Guinea. Flannery et al. in low numbers, such that they were not worth trying (1994:29) further stated that evidence suggested that to hunt or set snares for. Aplin et al. however stated E. kalubu had “been introduced by humans during the that locals at their Yeri River site in 2014 said that Holocene.” Flannery (1995b) would apparently have bandicoots were plentiful and could be snared along classified the Admiralty Echymipera as E. k. cockerel- runways. Efforts by Aplin et al. at spotlighting, looking lii Ramsay, 1877, which Flannery consistently, and for sign, and camera trapping that year also produced apparently inadvertently, referred to as “Perameles no evidence of the presence of this species. cockerellii.” Laurie and Hill (1954) and others spelled 4 Occasional Papers, Museum of Texas Tech University DIPROTODONTIA thus we thought it useful to summarize, analyze, and Phalangeridae annotate the published pertinent information on the Phalanger orientalis (Pallas, 1766) species. Schwarz (1910) named “Phalanger maculatus Northern Common Cuscus krameri” from the “Admiraltats- und Hermit-Inseln” and on the basis of 35 skins and 7 skulls in the Berlin It is clear that Tate (1945:13) knew of no Pha¬ Museum [given as “32 specimens” by Aplin et al. langer orientalis from the Admiralties, for he wrote: (2015:50)] and collected by “... Thilenius, Schoede und “If orientalis is represented on the Admiralty Islands ... KrAmer.” The new taxon was said to be most closely and St. Matthias it may be expected to conform to this related to “Ph[alanger]. m[aculatus\.papuensis” [now general type [like animals from ‘Long Island and Ruk,’ generally regarded as a full species, Spilocuscuspapu- etc.].” Kisokau (1974), however, recorded both “Pha¬ ensis (Desmarest, 1822)] “from Waigeoe” (= Waigeo). langer maculatcT (presumably = Spilocuscus kraemeri) The male of the new species was described as having a and Phalanger orientalis from “Rambutso Island” [ = white to yellowish ground color with moderately large Rambutyo Island] and “Phalanger maculata” (again, blackish-brown spots with faded borders. The snout, presumably = S. kraemeri) from Lou. Menzies and cheeks, and forehead were said to be mostly a paler Pemetta (1986) seem to be the only authors who have rusty brown; the limbs spotted, (the spots?) somewhat taken note of Kisokau’s report. They wrote that they paler than [on?] the back; the tail with large brown had seen no specimens of P. orientalis from the Ad¬ spots. The chin and throat white, the chest and belly miralties but that they would expect them to be of the yellowish. The female was described as unspotted with “subspecies” P o. intercastellanus Thomas, 1895 (now the snout and forehead reddish brown, the top of the regarded as a species in its own right—see Colgan et al. head, nape of the neck, and the back blackish brown, 1993). One might wonder if some or all of the animals not shiny. The rear part of the back was described as reported by Kisokau, as Phalanger orientalis, might be reddish brown, the tail somewhat darker, the limbs red¬ misidentified Spilocuscus kraemeri—young individuals dish brown, the chin white, the throat and upper part and/or adult females, and/or individuals of the nearly of the chest blackish brown, the ventral portion of the uniformly silver gray color phase mentioned by Cohn chest and belly yellowish, the underside of the base of (1914). However, as a native of Manus (see Flannery the tail dark brown. Skull measurements were given 2011), Kisokau is one to be expected to know how to for the holotype, an old male, variously identified as tell Spilocuscus kraemeri from Phalanger orientalis, “Berl. Mus. A. 12.6.09” and “A. 126 09”, “gesammelt and so we are inclined to accept his having regarded von Schoede;” and for a subadult female, “Berl. Mus. the two as sympatric on Rambutyo. In this regard, A. 19.07.2.” No definite localities, even to island or Helgen and Flannery (2004), who think that Spilocus¬ island group, were given for any specimen. Although cus kraemeri may occur on New Britain, explained its earlier in the description Schwarz had stated that rarity there, if it is found there, as possibly owing to the specimens were from both the Admiralties and competitive interactions with Phalanger orientalis on the Hermit Islands (p. 406), he later (p. 408) wrote that island. Accordingly, Helgen and Flannery also “Hab.: Admiral itats-Inseln.” The skull and dentition hypothesized that the commonness of S. kraemeri on were described and the former compared with those Manus could be explained by the absence of P. orien¬ of specimens of Spilocuscus maculatus (sensu lato?) talis. However, Helgen and Jackson (2015) listed P. that Schwarz had seen from other localities. Although o. breviceps Thomas, 1888, for Manus, and assumed Schwarz did not indicate a type locality, Feiler (1978) that it had been introduced. We know of no evidence, provided a color drawing of an individual identified as however, that it has ever occurred there. They did not Berlin Museum “12609” and from Manus, presumably list Rambutyo as an island occupied by this species. the holotype. Spilocuscus kraemeri (Schwarz, 1910) Thomas (1914) listed four specimens of “Phalan¬ Admiralty Cuscus ger maculatus krameri, Schwarz” from Manus. Two in the BMNH were collected by Meek and Eichhom and The literature on this species is complex, con¬ two were listed as being in the Tring Museum. fusing, widely scattered, and in various languages; Pine et al.—Marsupials and Rodents of Admiralty Islands 5 Cohn (1914), unaware of Schwarz’s (1910) young in July-August, with pouch young observed in description of S. kraemeri, redescribed his animal as September-December (then carried by mothers on their Phalanger maculatus minor, on the basis of at least 80 backs); young achieve independence in June.” specimens from Manus. These specimens presumably are or were housed in Bremen. However, our attempts According to Cohn, the females are more nu¬ to confirm this have not met with success. He also merous and larger than the males. The males were stated that over 60 living individuals from Manus had described as having dark brown legs with little white passed through his hands. No type specimen was des¬ on them, the belly white with very small brown spots, ignated, no catalog numbers or museum(s) of deposi¬ the tail as a rusty-brown. In these respects, Cohn’s tion were mentioned, and no localities, other than the description differed somewhat from that of Schwarz names of islands, were given. In addition to Manus, (1910). Two male specimens out of considerably more Cohn stated that he had seen this species on Ponam, than 20 were a uniform silver gray with small dark dots and on Lou, where he found it to be quite common. He on the back; the base of the tail with a faint brownish attributed its presence on these islands to the persistence wash. Both of these males were half grown and were of primary forest, although Flannery (1995b: 105) found taken to Cohn at the same time and he speculated that this species to be “common in all vegetation types, from they might have been from the same litter. gardens to almost undisturbed forest” on Manus. Cohn stated that on smaller, more heavily populated islands, Cabrera (1919) synonymized Phalanger macu¬ the species had been extirpated, but mentioned only latus krameri Schwarz with Phalanger maculatus Baluan as being one of these islands (see also Helgen minor Cohn and placed [Phalanger] “minor y fused’ and Jackson 2015). It is unclear whether Cohn actu¬ Oken, 1821 in the synonymy of Phalanger orientalis ally had any evidence of the former occurrence of S. (Pallas, 1766). In this, he seems to have been follow¬ kraemeri on any of these islands or merely surmised ing Thomas (1888:201), who had placed “Phalangista that it had once occurred there. The latter seems more minor and fusca (Oken) Schinz, Cuv. Thierr.... (1821)” likely, although Flannery (1995b) accepted the former in “Phalanger orientalis.” presence of S. kraemeri on Baluan, based on Cohn’s statement (see also Helgen and Jackson 2015). Cohn According to Schwarz (1934:90), in the course found S. kraemeri to be very common on Manus and of his idiosyncratic, essentialistic speculations con¬ to be the most important game animal for the locals. cerning the phylogeny of and also supposedly directly Cohn was on Manus from September until December environmentally-induced color patterns in cuscuses, of 1912. During this time, he examined many pouch certain spotted patterns belong “to the segmental young ranging in size down to 95 mm head plus body type, being homologous to the rosettes of a leopard, length and 80 mm tail length. These smallest ones were and representative of the ‘English’ type [undefined] blind and hairless. Cohn concluded that this species did of piebald pattern ... these spots evolve a tendency to not breed all year long. He stated that the heavy rains spread, thus in the end producing a true piebald pattern, began in November (with the dry season starting around such as is found in the males of Ph. maculatus. [new March/April) and it appeared to him that breeding must paragraph] That the very distinct markings found in have ceased a short time before this. He concluded that this species are derived from the ones just described is a second breeding season must exist early in the year, shown by the females of Ph. m. papuensis Desmarest making the species diestrous, because in September he from Waigen and of Ph. m. krameri Schwarz from the found numerous already half-grown animals. Flannery Admiralty Islands, some of which develop spots, but (1995b), in discussing Cohn’s findings, apparently of a less progressive type than is found in the males, thought that all of the small September-December and more resembling that of the male Ph. o. ornatus.” young were around the minimum size given and omit¬ Schwarz (1934:90-91) went on to write “Males which ted the evidence for a second breeding season. Aplin are nearly or perfectly white are ... found in Ph. macu¬ et al. (2015:56) reported a “fully furred pouch young” latus, although in that species perfect albinism is rare, during the period 2-6 October 2014. According to and a stage similar to the ‘Siamese’ or ‘Marten’ type Helgen and Jackson (2015:496), “Limited observations of cats and rabbits is the rule ... the individuals show [unspecified] suggest the Admiralty Spotted Cuscus the normal piebald pattern earlier in life.” breeds seasonally, giving birth generally to a single 6 Occasional Papers, Museum of Texas Tech University Tate and Archbold (1937:381) noted that “Minor of “Young male, Museum Godefroy, Hamburgh, 1869. Cohn from Admiralty Islands would appear to be a Collected by Captain Alfred Tetens. Head, ears, hands, synonym of krameri Schwarz ...” and stated that the feet and spots on tail and on outerside of legs red; spots Archbold collection held topotypes of “krameri.” They on back and nape of neck brownish red... Adult female. figured the skull, minus the mandible, ofAMNH 99900, Museum Godefroy, 1869. Collected by Captain Alfred seen from below. Concerning “krameri” they wrote (p. Tetens. Tail injured. Crown of head, nape of neck, 383) that the pelage “... appears to be thoroughly and back and flanks sooty red; face and legs red; ears and consistently distinct from that of our Arfak and south tail yellowish red; underparts like in other females.” New Guinea specimens ... A striking characteristic of krameri in both sexes is the nearly black head on which Laurie and Hill (1954) used the name Phalanger the cinnamon-colored ears stand out contrastingly, and maculatus kraemeri for this form and, presumably fol¬ the deep chestnut-colored fore limbs.” The Archbold lowing Tate (1945), treated it as not even subspecifi- holdings were listed (p. 384) as “Admiralty Islands cally distinct from Spilocuscus rufoniger and listed it (Manus Isl.): 1 y. ad. S', 3 y. ad. $s (all except one 5 as occurring in the Echiquier (Ninigo) Islands and on skins only); 3 y. Ss, 1 y. ? (skulls only); 1 ad. albino the Huon Peninsula. They placed Balantia minor Oken, P .” Concerning the last, Tate and Archbold wrote (p. 1816 in the synonymy of Phalanger orientalis (Pallas). 384) “The muzzle and the insertions of the superciliary and labial vibrissae are reddish brown ... and the skin Haltenorth (1958:28) wrote, concerning this ani¬ and hairs of the hands and feet are yellowish ... .” The mal, “Huon-F[leckenkuskus].,P. m. krameri Schwarz, supposed “albino” is presumably AMNH 99749, which 1910; Huon-Halbinsel, Kaiser Wilhelms-Land, Admi- was recorded as having a brown iris, apparently the nor¬ ralitats- u. Echiquier-Inseln.” mal color—the collector, W. F. Coultas, also recorded a brown iris for specimens with normally colored pelage. Collins (1973:177) gave the range of “Phalanger Our notes on this specimen state “All white dorsally maculatus” as including “the Ninigo and Admiralty except some very faint bufify areas on rump and base of Islands, New Britain, New Ireland, and some of the tail. Ventrally, base of tail and adjacent area faint bufify. Bismarck Archipelago.” Some buffy toward end of tail. Grayish around pouch area and posterior to that but maybe soiled.” In their Kisokau (1974) stated that he had collected “Pha¬ tables of measurements (pp. 448-449), the following langer maculata” during the period 22 December 1971 specimens are listed: “A.M. 99832 juv. S ... Manus”, to 19 January 1972 on “Rambutso Island” [= Rambutyo “A.M. 99831 juv. c? ... Manus”, “A.M. 99829y. ad. 5 Island] and Lou. No information concerning where ... Admiralty Is.”, “A.M. 99900 y. ad. $ ... Admiralty the specimens were deposited was given but it seems Is.”, “A.M. 99749 ad. $ (albino) ... Admiralty Is.” likely that it was in a collection that was housed in the old Department of Environment holdings at Moitaka, Tate (1945:27) mentioned “Our [AMNH] original much of which was lost to mold, decay, and general small series of skins and skulls from Manus ... includ¬ lack of curation by the 1990s (information provided ing one white female.” Tate, who treated this species by an anonymous reviewer). All subsequent authors as “Phalanger maculatus krameri,” mentioned the except for Menzies and Pernetta (1986) seem to have Echiquier (Ninigo) Islands as included in its range but overlooked Kisokau’s publication. provided no further information. Tate also, somewhat tentatively, synonymyzed Phalanger maculatus rufo¬ Feiler (1978) examined 32 skins and 2 skulls niger Zimaxa., 1937 [= Spilocuscus rufoniger (Zimara)] from Manus. These were some of the 35 skins and 7 of mainland New Guinea, with S. kraemeri. He (p. skulls that had been the basis for Schwarz’s original 19) characterized S. kraemeri (including rufoniger) description of the species. Feiler gave the type local¬ as “Male with dark head, most spots dark, on white ity as “Admiralitats-Inseln.” He mentioned one small ground, female dark throughout, with reddish tail, white individual (“gelblichweiss oben und unten”) hands, and feet. A white aberration occurs.” Concern¬ which he thought was a young animal but said that ing the Ninigo Islands, Jentink (1885:117-118) had the remainder were spotted. Feiler made it seem that reported “Cuscus maculatus” from there on the basis the three females mentioned by Schwarz (1910), and 7 Pine et al.—Marsupials and Rodents of Admiralty Islands showing the unspotted female pattern, were unavail¬ seldom the case in specimens from Manus. The tail able for some reason. Schwarz did not mention a white was said to be reddish brown as in the specimens from specimen. Feiler stated incorrectly that Schwarz had Manus. written that S. kraemerfs range reached to eastern New Guinea (the island). Feiler also mentioned two female, Feiler discussed two skulls from what he called Berlin Museum, skins damaged by the war (“ZMB, 11. the “Hermit-Inseln,” one from “Echiquier,” “MNHL” 9.1899”), collected by Thilenius, presumably from the (apparently a typo for MNHG, meaning from the Museo Admiralties, that he said that Schwarz must have seen. Civico di Storia Naturale “Giacomo Doria,” Genoa) Feiler then divided the males into four groups on the taken by Godeffroy, and a specimen in the Berlin Mu¬ basis of color pattern. For some reason, he listed only seum, “Agoma [?] A 19. 07.” taken by Kramer. 30 of the 32 specimens he had examined. His first group consisted of the single whitish specimen. His second Koopman (1979) referred to animals from the Ad¬ was individuals with the palest [or most pronounced?] miralty andNinigo Islands as P[halanger\. maculatus. spots, with the snout rusty red, rust-colored to reddish- black; the front feet and nape of the neck more or less George (1987) treated S. kraemeri as belonging in paler or darker rusty (from golden yellow to reddish- the species S. maculatus (E. Geoffroy St.-Hilaire) and, brown); with only isolated white spots; pale rusty-red tentatively, as constituting a recognizable subspecies. spots on the back, isolated darker spots between them; about 60% of its surface white, tail almost solidly rusty Flannery and Calaby (1987) discussed one young red with isolated pale spots. female (AM 3693) and two male specimens (AM 3563, 5117). AM 5117 was labeled as from New Britain The third group had the face until above the eyes (which Flannery and Calaby thought was an extremely rusty red, the neck rusty to almost black and without unlikely provenance) and the other two were Taronga spots; the forefeet darker, reaching, in two cases, a Zoo (in Mosman, New South Wales) specimens of blackish reddish brown, in general like in his second unknown origin. These specimens were identified as group. The ground color cream with pale golden red¬ of S. kraemeri because they “conform closely to the dish brown tinge, 60-80% of the surface of the back is morphology of kraemeri as described by Tate (1945) white with large spots; the tail reddish brown more or [see Tate’s brief statement quoted above].” In addition less paler or darker and with white spots. Feiler men¬ to there being these Australian zoo specimens, the Field tioned one specimen from “12.6.1909” (recall that the Museum in Chicago houses a male specimen FMNH catalog number of the holotype was variously given by 74914 (Figs. 1A, 2), represented by a skin and skull, Schwarz as “A. 12.6.09” and “A. 126 09”) and taken from the Brookfield Zoo and labeled as from Australia, by Schoede, identified to locality only by “Deutsch- that clearly represents S. kraemeri. The history of this Neuguinea,” as having the ground coloration more specimen may be tied in with that of the Australian yellow and with the dorsal spots not reddish-black Taronga Zoo specimens examined by Flannery and but a deep and sharply contrasting black with 80% of Calaby (1987). Paradiso and Fisher (1972) reported, its surface white. for example, the importation of eight live “Phalanger maculatus” into the U.S. in 1970. Collins (1973) also The fourth group was made up of animals with discussed S. maculatus (including S. kraemeri?) being dark skins, the forefeet a dark rusty red, large spots, held in the Taronga Zoo and in at least one U.S. zoo. with 30-50% of the skin white, intensely black spots on FMNH 7494 is discussed further in this species account a yellow-white background, sometimes with isolated under “Specimens examined.” Pine has seen a speci¬ reddish spots in between, the tail reddish above, a pale men (apparently of S. maculatus sensu stricto) from yellowish-white underneath, the hind end above and a U.S. pet store and now housed in the Los Angeles below the same, one individual from “D.-Neuguinea” County Museum. with the tail underneath also mostly reddish brown. A skin from “Hermit-Inseln,” “Berlin Museum 5. 7. Flannery and Calaby (1987) also mentioned a 1909,” taken by Kramer resembled the specimens from specimen from Mussau Island in the St. Matthias group Manus, the nape of the neck was very rusty, which is as possibly belonging to S. kraemeri, but Flannery and 8 Occasional Papers, Museum of Texas Tech University Figure 1. Skins of male Spilocuscus kraemeri. A—Stuffed skin of immature male (FMNH 74914), specific locality unknown. B—Incomplete skin of adult male (UWZM M27790) said to be from the island of Bali, off the north coast of New Britain (see text for additional details). Figure 2. Dorsal, ventral, and lateral views of the skull of an immature male Spilocuscus kraemeri, specific locality unknown (FMNH 74914—see Fig. 1A). Condyloincisive length of skull = 75 mm. Pine et al.—Marsupials and Rodents of Admiralty Islands 9 White (1991), Flannery (1995b), Helgen and Flannery the color patterns of the specimens available to them. (2004), and Helgen and Jackson (2015) reported only Their photograph of the juvenile male shows a gray S. maculatus from there. This last publication’s treat¬ animal with a white cheek. Singadan (1996) wrote that ment of the cuscuses on Mussau is odd, considering juvenile males are normally gray with grayish brown that Helgen (2007) treated the cuscus on Mussau as spots and patches. The two more-or-less adult males belonging to an unnamed introduced species. described by Flannery and Calaby are said to be (p. 554) with “limbs and tail ... brownish-red. The crown ... Flannery and Calaby (1987:555) wrote “... we a darker shade of brown ... base coloration ... white consider it extremely unlikely that kraemeri reached ... dark mottling ... the same colour as the dark area Manus with the aid of man, as it is a highly distinctive of the crown over the shoulders ... becomes darker, form, and given the lack of distinctive species to have almost black over ... lower back. Near ... base of... evolved on islands isolated since the end of the Pleis¬ tail... spots lighten again [this variation in darkness of tocene in Australia, it appears to have been isolated on dorsal spots not evident in their published color photo], Manus for a considerably longer period.” Nonetheless, Ventrally ... white, with brown hairs on ... lower sides Flannery and Calaby continued to treat kraemeri as a of... neck.” They described (p. 554) a juvenile female subspecies of S. maculatus. They noted (p. 555) that as “... ventral surface has a white patch on the chest “Two color morphs have been recorded for adult female with a longer white patch over the belly. A darker line Spilocuscus from Manus. Feiler (1978) mentions a ... surrounds these light patches. Overall... dark red¬ single white female from Manus. [In fact, Feiler in¬ dish-brown, with silvery red guard hairs. On ... base dicated that the specimen was of a male—Tate (1945) of... tail... lighter patch of fur ... under fur ... lighter mentioned a white female.] However, most individuals coloured.” Although Singadan (p. 78) wrote that the seem to be very dark-colored, with a large black saddle tail in both male and female S. kraemeri is “white with area on a reddish background.” Although Flannery shades of brown or yellow,” the specimens and figures and Calaby (1987) provided color photographs of a we have seen show warm pale brown or more or less skin of a typically spotted adult male [AM 3563] and tan tails, although some white may be on the tails of a living juvenile male from an unknown locality, and some. Singadan gave body mass as 2.2-3.2 kg. Be¬ Flannery (1995b) published a color photograph of a ing unaware of the later archaeological discoveries to living adult and typically spotted male from Manus, be made on Manus, Singadan hypothesized that the we have found no photographs of an animal identified Admiralties cucuses had been introduced from the as an adult female. Feiler (1978) and Flannery (1994), mainland about 3,000 years ago and had acquired their however, provided colored drawings of adult males distinguishing characteristics since then. and females. Feiler’s figured female from “Admiral- itatsinseln” was Berlin Museum 11.9. 1899, the male, Menzies (1991) provided a map indicating the Berlin Museum 12609 from Manus. Although Flannery range of S. m. kraemeri to consist of Manus and an (1994:218) stated that “The male [of the species] has unidentified dot in the ocean which is probably intended small blackish spots on a white background ...,” the to represent Wuvulu, an island that is questioned as a spots on the male figured do not appear small and the habitat for S. kraemeri by Flannery (1995b), although earlier literature reports size of the spots as ranging it is listed from there by Helgen and Jackson (2015). from small to large. Flannery and Calaby (1987:554) Menzies gave “Admiralty Islands” as the type locality. wrote “It is possible that the white females from Manus He stated (p. 68) that isolated populations of Spilo¬ and Mussau represent a distinct species related to the cuscus maculatus occur on “Wuvulu, Manus Islands mainland S. maculatus and that the highly coloured [sic] but not New Britain.” Writing of S. maculatus in kraemeri represents a separate species.” Singadan general and presumably including S. kraemeri, Menzies (1996), however, assumed that white specimens of stated (p. 69) that “Male spotted cuscuses are bigger both sexes were only color variants in animals of the than females ... .” Helgen and Jackson (2015) stated S. maculatus species group on the mainland of New the opposite, as had Jentink (1885), Cohn (1914), and Guinea, the Admiralties, and “the St. Matthais Islands.” Helgen and Flannery (2004). According to Helgen Flannery and Calaby (1987) provided descriptions of (2007), females average 14% heavier than males. 10 Occasional Papers, Museum of Texas Tech University Flannery (1995b: 104) treated kraemeri as a fairly precise field locality published for it. Flannery full species, stating “These species [kraemeri and (1994) provided measurements from at least two males maculatus] differ in dentition and skull features and and two females from Manus. Natural history informa¬ are probably not closely related.” The animal from tion is provided by Flannery (1994, 1995b). Mussau Island was then apparently treated as S. macu¬ latus maculatus. It was unequivocally treated as S. Singadan (1996) noted that S. kraemeri is a low¬ m. maculatus by Flannery (1994), apparently, also by land animal, usually occupying lowland rainforests, Helgen and Flannery (2004), and definitely by Helgen usually below 1,000 m. However, the maximum eleva¬ and Jackson (2015). tion of Manus is only about 720 m and no other island inhabited by this species is even that high. Singadan Flannery (1994, 1995b) made no mention of also noted that there were at least 52 specimens held in the existence of white female individuals. He stated the collections of the University of Papua New Guinea (1995b: 105) “... archeological evidence suggests that and PNGMAG. Assuming that the 80 or so specimens it [S. kraemeri] has reached Manus recently, perhaps seen by Cohn are still in existence, then the number of in the last one or two thousand years ... .” Flannery available specimens is indeed impressive. (1994) and Flannery et al. (1994) had earlier expressed this same view. Helgen and Flannery (2004:831), Williams (1999) reported S. kraemeri remains however, wrote that at “Pamwak, a rockshelter site in from as far back as ca. 11,000 ybp, at the Pamwak southern Manus ... S. kraemeri is represented errati¬ archeological site on Manus. cally in the record at least as far back as about 11,000 years ago. This ... suggests to us that S. kraemeri might Anthony (2001) discussed the purchase, in a be a natural element in the Manus fauna.” In addition market in Kimbe, New Britain, of a skin of a cuscus to Manus, Flannery (1995b: 105) listed Lou and Luf which had reportedly been speared on the small island (Hermit Islands) as places where S. kraemeri occurs of Bali off the north coast of New Britain. (Bali is an and “Ninigo” and Wuvulu Island as places where it alternative name for the island generally known as possibly occurs. Flannery (1994, 1995b) cited Cohn Unea or Uneapa in West New Britain.) Judging from (consistently misspelled “Kohn” in 1994) (1914) as the photograph provided of this skin and our direct the authority for the species having been extirpated examination of the specimen (UWZM M27790; Fig. on Baluan and other, unspecified, islands. Groves IB), it is that of a male S. kraemeri, although Anthony (2005a) listed it only for Manus and Lou, as had Flan¬ identified it as that of an S. maculatus. nery (1994). Helgen and Flannery (2004:828) wrote, concern¬ As late as 1988, Flannery (1995b: 105) found S. ing Spilocuscus wilsoni, that the “adult male coat is kraemeri to be common on Manus “in all vegetation pure white dorsally and ventrally ... as in some ... S. types, from gardens to almost undisturbed forest.” m. maculatus ... ”, taking no note of white individuals The IUCN Red List of Threatened Species lists it as of both sexes, reported of S. kraemeri, although Flan¬ “Near Threatened” (see Helgen et al. 2008). Helgen nery and Calaby (1987) had earlier mentioned white and Jackson (2015:480) wrote that this species “... Spilocuscus from Manus and which they supposed appears still to be common in all parts of its limited might represent a second species there. [Somewhat distribution ... .” incongruously, Helgen and Jackson (2015) in their discussion of S. kraemeri, mention that there are white Flannery (1994,1995b) gave “Yowat” as the Nali individuals of S. rufoniger (Zimara, 1937), but make no language name (“South Manus”) for this animal. mention of white S. kraemeri.] Helgen and Flannery (2004:830) also stated “Two species of Spilocuscus The colored drawing of the male, in Flannery’s occur in the Bismarck Archipelago of eastern Papua 1994 book, is based on AM M20925, from Polomou, New Guinea: S. maculatus maculatus, which occurs Manus. The female depicted was acquired at the Lo- on the islands of Mussau ... and New Ireland; and rengau Market on Manus. The specimen from Polomou S. kraemeri of the Admiralty group ... ”. They stated is the only one of which we are aware which has had a that the Bismarck S. maculatus apparently had been