Novitates PUBLISHED by the AMERICAN MUSEUM of NATURAL HISTORY CENTRAL PARK WEST at 79TH STREET, NEW YORK, N.Y

Home , Bunomys, New Guinean rat, Palm rat, Papagomys, Tateomys

AMERICAN MUSEUM Novitates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2850, pp. 1-10, figs. 1-5, tables 1, 2 June 30, 1986

Paulamys, a Replacement Name for Floresomys Musser, 1981 (Muridae), and New Matenral of that Taxon from Flores, Indonesia

GUY G. MUSSER,' A. VAN DE WEERD,2 AND ELIZABETH STRASSER3

ABSTRACT Paulamys, a new name, is proposed to replace gara, Indonesia. New specimens of the species are Floresomys Musser (1981), the generic name for listed and describcd and their significance to un- a species of long-nosed murid known by subfossil derstanding the native Floresian murids discussed. fragments from the island of Flores in Nusateng- INTRODUCTION Twelve species ofmurids are recorded from- subfossil and recent specimens and still lives Flores, one ofthe larger islands in Nusateng- on Flores. Papagomys theodorverhoeveni, gara (Lesser Sunda Islands) east of Bali and Floresomys naso, Komodomys rintjanus, and south ofSulawesi. Six ofthem (Rattus rattus, Spelaeomys florensis are represented by Rattus argentiventer, Rattus exulans, Rattus subfossil fragments, and Hooijeromys nusa- norvegicus, Mus caroli, and Mus musculus tenggara by Pleistocene samples. Except for castaneus) were introduced and now live on Komodomys, which now occurs on Rintja and Flores in ecological associations tied to hu- Padar, the small islands between Flores and mans-towns, villages, gardens, agricultural Komodo, all the native species have been fields, scrub, and disturbed forest and savan- found on Flores and nowhere else (Musser, na (Musser, 1981). The other six species are 1981). native. Papagomys armandvillei is known by

' Archbold Curator, Department of Mammalogy, the American Museum of Natural History. 2 UNOCAL, Locked Bay Service No. 3, Orchard Point Postoffice, Singapore, 9123. 3 Participant, Undergraduate-Graduate Research Program, American Museum of Natural History.

The subfossil and Pleistocene fragments D. Carleton, and James L. Patton gave us were collected by Dr. Theodorus Verhoeven, intelligent and helpful criticisms ofthe manu- described by Hooijer (1957, 1967), and then script; we appreciate their contributions. Pe- redescribed by Musser (1981). That material ter Goldberg photographed the specimens and came from Liang Toge (Liang is Indonesian provided his usual excellent prints. for cave). Dr. Verhoeven also collected spec- Ms. Elizabeth Strasser thanks the Green- imens of rats from other caves in Flores and wall Foundation for her award (administered these samples were sent to the Instituut voor by the American Museum of Natural His- Aardwetenschappen at the Rijksuniversiteit tory) from the Undergraduate-Graduate Re- Utrecht, where the specimens were cleaned, search Program. catalogued, and recently loaned to Musser. The present report is number 116 in Re- Examples of the six native species discussed sults of the Archbold Expeditions. by Musser (1981) are in the new samples, and some ofthe species are represented by abun- THE REPLACEMENT NAME dant material. There are also samples ofthree (BY GUY G. MUSSER) new species. Two of these are related to Pa- pagomys armandvillei and the third is a shrew During the last three years, Dr. V. Fahl- rat in which the morphology of the dentary busch and Dr. R. Hutterer, among others, resembles the large-bodied shrew rats of the kindly pointed out to me that Floresomys, Philippines (Chrotomys) and the smaller the generic name I had applied to a murid shrew mice native to New Guinea (Neohy- from Flores, was preoccupied by Floresomys dromys). These new specimens reinforce the Fries, Hibbard, and Dunkle (1955), which view that the native Floresian murids com- had been proposed for a species of sciuravid. prise two primary groups. One is clustered I read that report long ago but forgot about around Papagomys, which is phylogenetical- the name, for which I apologize. Floresomys ly tied to faunas found to the north and west Musser (1981) should be replaced by the fol- on Sulawesi and the Sunda Shelf; the other lowing. is represented by Spelaeomys and the shrew rat, which are linked with murids endemic Paulamys Musser, 1986 to the New Guinea and Australian region, The diagnosis, genotype, and included and possibly to the Philippines. species for Floresomys as proposed by Mus- Among the new specimens are seven den- ser (1981) apply to the replacement name tary fragments of Floresomys naso, all less Paulamys. than 4000 years old, collected from three In 1980, Dr. Verhoeven and his wife, Paula caves. Although some of the specimens are Hamerlinck, visited me at the American Mu- no more complete than the four examples seum ofNatural History. I had recently sub- described by Musser in 1981, one dentary is mitted a manuscript to the publisher that nearly complete and enough remains of two contained descriptions ofspecimens Dr. Ver- others to provide new information about cer- hoeven had collected from Liang Toge in tain structural aspects of the species. western Flores during the 1950s. We talked In this report, we place on record the new about those fragments and his excavations. I specimens ofFloresomys naso and clear up a showed him and his wife some of the new nomenclatorial problem. Several persons had material that he had collected, which hadjust mentioned to us that Floresomys is preoc- been sent to me from Utrecht. I related how cupied. Musser provides a replacement name important this undescribed material was in that is needed at this time to satisfy the re- understanding the nature of the murids na- quests of those, including ourselves, who are tive to Flores, and that some of the samples preparing classifications and faunal lists in represented new species which provided spe- which the name will be included. cial insight into the past evolutionary history We are indebted to Dr. Verhoeven who of the rats. Dr. Verhoeven asked questions allowed us to study the new material and about the fauna and spoke ofhis experiences related details about the caves and their ex- working in the different caves. The time cavations. Drs. Karl F. Koopman, Michael passed quickly and before my guests left, Dr. 1986 MUSSER, VAN DE WEERD, AND STRASSER: PAULAMYS 3

TABLE 1 Measurements (in Millimeters) and Ratios (in Percent) of Lower Molars and Incisors from Adult Paulamys nasoa Ratio Cave and Length and breadth BM2 BM3 BM3 specimen CLMI-3 ALMI-3 BI BM1 BM2 BM3 BM, BM, BM2 Liang Toge Spec. lb 7.6 7.9 1.0 2.0 2.1 1.9 105 95 90 Spec. 2 7.5 7.9 1.2 2.1 2.2 1.8 105 86 82 Spec. 3 - - 1.1 2.0 2.2 - 110 - - Spec.4 7.5 7.7 1.2 2.0 - 1.8 - 90 -

LT 47 - - 1.2 ------Liang Soki LS 9 7.3 7.5 - 1.9 2.1 1.8 111 95 86 Liang Bua LB 90 7.6 8.0 1.2 1.8 2.0 1.8 111 100 90 LB 15 - - - 1.9 2.2 - 116 - - LB 20 - 8.1 1.2 - 2.2 2.0 - - 91 LB 13 - 7.7 - 1.8 - - - - -

LB 85 - - - - - 1.8 - - - a Abbreviations: CLM, crown length ofmolar row; ALM, alveolar length ofmolar row; BI, breadth ofincisor taken where the tooth emerges from the ramus; BM, breadth of molar. b Holotype. Specimens 1-4 are described and illustrated by Musser (1981).

Verhoeven asked if it might be possible to LLANG TOGE: This is a rock shelter located name one of the rats after Paula. I would be in the northwestern part of the Ngada Dis- happy to do so, I told him. By combining the trict, 2 km south of the abandoned village Greek mys, for mouse (or rat), with the name Lepa, near Teong. The shelter was discovered Paula, I honor the companion of Dr. Ver- in 1952 by Dr. Verhoeven and excavated hoeven, the man who has contributed so during 1954 and 1960. The rock ceiling of much to our knowledge ofthe recent past on the cave has an overhang of about 4 m, and Flores. the shelter itself is about 11 m broad. Its deposits are divisible into three units. The top unit is 1-1.6 m thick and is the only one THE CAVES that concerns us here. A human skeleton, described by Jacob (1967), was found about 1 The caves yielding the specimens we report m deep in this unit and it is from this level here are located in the Lepidocyclina-lime- that a radiocarbon date of 3550 ± 525 years stone (Ehrat, 1928), a formation of Miocene old was obtained. Specimens 1-4 of Paula- age occurring in the northern part of the mys naso described by Musser (1981) were Manggarai and Ngada districts of western collected in Unit 1 but the level at which they Flores. Ehrat (1928) reported that the lime- were found is unknown. The additional spec- stone formation has been intensively eroded imen (table 1) we report here was obtained into a karst topography containing numerous from either the surface or top level ofUnit 1. caves. Sediments in some of the caves had LIANG BUA: This cave is north ofthe village been explored for remains ofartifacts (mainly ofTeras, 10-12 km northwest ofRuteng, the Mesolithic) and other human remains (Ver- capital of the Manggarai District. About 30 hoeven, 1968). Fragments of postcranial m broad and 34 m deep, with a ceiling 10 m skeletal elements, crania, mandibles, and iso- high, the cave was used as a school building lated molars and incisors are common in some by the Teras residents and was visited by Dr. of the cave deposits described below. Verhoeven in 1950 and excavated in 1965. 4 AMERICAN MUSEUM NOVITATES NO. 2850

I !

LB90

V43Wli? N.

_.t LB20

LB13

- o a" , 'V ftMA..- .

lw LS9 N Okk-

-Alo-, .1 .1 000 LB85 1.

'a.^ low.-AdSLK 4MML~ ~ S~~~~~ ~ LT47

.1 LB15 Fig. 1. Dentary fragments ofPaulamys naso from Liang Bua, Liang Soki, and Liang Toge. See table 1. Approximately x 2.

Several Neolithic graves were discovered at Paleolithic, grave was found at approximate- a depth of 90-115 cm. An older, probably ly 2 m. Five specimens of Paulamys naso 1986 MUSSER, VAN DE WEERD, AND STRASSER: PAULAMYS 5

CUp

ar %V, I

cp ;r* -me ;ap men

cdp

man

',..

'. ", I f .:il .i:.Wjv---- il .... rwipw-- Fig. 2. Labial (top) and lingual (bottom) views of the dentary ofPapagomys armandvillei (Museum Zoologicum Bogoriense 2395). Approximately x 2. Compare this dentary with those of Paulamys naso on the opposite page. Abbreviations: ap, angular process; ar, ascending ramus; br, body of ramus; cdp, condylar process; cp, capsular process (encloses base of the incisor); crp, coronoid process; Ir, lingual ridge; man, mandibular foramen; men, mental foramen; mr, dorsal and ventral masseteric ridges. from this cave (table 1) were discovered in mus is absent in three and present but badly the sediments above the graves; we do not eroded in one. The bony covering of the in- know the exact level at which they were col- cisor anterior to the molar row is present in lected. two specimens but incomplete in the other LIANG SOKI: We know little about this cave two. The fragments were originally deter- except that it is located about 15 km north mined to come from species ofRattus (Hooi- ofRuteng in the Manggarai District. A single jer, 1967; Musser, 1972) but the "shapes of specimen ofPaulamys naso (table 1) was dis- the first two laminae on the first molar, the covered in a shallow test excavation. absence ofan anterolabial cusp from the third molar, the proportion of small molar size to THE SPECIMENS large dentary, the elongated segment of the dentary in front ofthe toothrow, and the slim The four fragments of Paulamys naso de- incisors is a combination offeatures not found scribed by Musser in 1981 (Specimens 1-4 in Rattus" (Musser, 1981, p. 1 8). from Liang Toge) represent adults. Each con- Although fragmentary, enough characters sists of a partial mandibular ramus with a are preserved in Specimens 1-4 to indicate a molar row that is either complete or missing species characterized by an elongate mandi- the second or third molar. The ascending ra- ble, which implies a long rostrum. 6 AMERICAN MUSEUM NOVITATES NO. 2850

A B Fig. 3. Mandibular molar rows ofPaulamys naso. A, LS 9; B, LB 90; C, LB 20; and D, LB 15. Arrow points to small anterolabial cusp. Approximately x 8. Measurements are listed in table 2.

The seven new pieces represent seven in- cess is associated with LB 90 and it is small, dividuals that range in age from young to old thin, and delicate. The angular process is bro- adults (table 1). Four of the pieces (LS 9, LT ken on all specimens but there is enough re- 47, LB 15, and LB 85) are small fragments maining on LB 90 and LB 13 to suggest that (fig. 1) and contain no additional information it was shaped like the process in Papagomys beyond that derived from the original Spec- armandvillei (fig. 2). The coronoid process is imens 1-4. Three ofthese four (LS 9, LT 47, missing from all the pieces. The back edge of and LB 15) have enough of the dentary re- the dentary is probably shallowly concave, maining anterior to the molar row to indicate judging from the gently dished intact segment the dentary was elongate and closely similar that forms the top half of LB 90. to Specimens 1-4. Parts of the lingual surface of the dentary The other new fragments (LB 90, LB 20, are preserved in the specimens. A prominent and LB 13) are relatively complete (table 1; feature is a lingual shelflike ridge that extends fig. 1) and together provide details about from behind the molar row, slants gradually mandibular configuration. The bony sheath upward, and fades into the base of the con- containing the incisor anterior to the molar dylar process. Dorsal to this ridge and near row is long and slim. A large mental foramen the base of the condylar process is the large is present just anterior to the junction of the oval mandibular foramen. At the front ofthe strong dorsal and ventral masseteric ridges. dentary, the elongate conformation of the Projecting from the labial side ofthe dentary bony incisor sheath is conspicuous in lingual is a large, globular capsule containing the base view. ofthe incisor. A remnant ofthe condylar pro- Incisors in the new samples resemble those 1986 MUSSER, VAN DE WEERD, AND STRASSER: PA ULAMYS 7 already described by Musser in 1981. The TABLE 2 three most complete incisors are long, slen- The New Specimens of Paulamys naso from der, and 1.2 mm thick (measuredjust beyond Flores (see fig. 1) the end of the bony sheath). In LB 20, the Cave and tooth ends in a long, gently curved wear facet specimen Age Description and sharp tip (fig. 1). The incisor enamel is pale or dark orange in the new material as it Liang Toge is in Specimens 1-4 from Liang Toge. LT 47 adult anterior portion of right a of Molars (fig. 3) of the new specimens are dentary with segment similar in size, occlusal patterns, and number incisor; no molars ofroots to those in the holotype and the three Liang Soki other specimens described previously (Mus- LS 9 adult part of right ramus with in- ser, 1981). One ofthe new pieces has a shorter tact molar row; no inci- molar row than any ofthe others and the first sor; inferior masseteric mental fo- molars offour fragments are slightly narrow- ridge evident; ramen intact; occlusal er, increasing the observed range in length of surfaces of molars moder- row molar and molar breadth (table 2). Cusps ately worn (fig. 3) and cusplets in the new sample resemble those of Specimens 1-4 with one exception. The Liang Bua third molar ofLB 90 has a small anterolabial LB 90 adult a nearly complete right den- is cusp (fig. 3), which is absent in third molars tary; part of incisor ofthe other specimens, including the original missing as are the coronoid and angular process- material from Liang Toge. es; molar row is intact, its surface moderately worn SIGNIFICANCE (fig. 3) LB 20 old adult anterior three-fourths of with com- We have been unable to identify any cra- right dentary plete incisor and second nial or postcranial fragments that might be- and third molars; masse- naso. long to Paulamys Pieces of crania and teric ridges evident, men- body skeletons, as well as upper molars and tal and mandibular fo- incisors of the other Floresian species, are ramina present; molar either absent or uncommon; dentary frag- surfaces moderately worn ments, lower molars, and lower incisors con- (fig. 3) stitute the bulk of the more than 300 pieces. LB 13 old adult ramus of right dentary with This is unfortunate because elements of the first molar only; masseter- cranium and postcranial skeleton would pro- ic ridges and capsular vide more information about the overall projection of incisor al- morphology of Paulamys naso and its phy- veolus are intact; mental logenetic position within the cluster ofnative foramen evident; molar surface very worn Floresian murids. The new material we re- of ramus cord here nonetheless expands our informa- LB 85 adult fragment right with intact third molar, tion to a different set of limits. We better roots of first molar, and a understand the shape ofthe dentary and know short segment of incisor; more about variation in molar size and the molar surface moderately occurrence of certain cusplets. Information worn derived from the new material reaffirms the LB 15 young adult anterior part of left dentary original hypothesis about the possible ecol- with intact first and sec- ogy of the species (Musser, 1981). ond molars; no incisor; The entire dentary ofPaulamys naso is low mental foramen present; and long, compared with that of Papagomys parts of masseteric ridges armandvillei, for example (fig. 2). This con- intact; occlusal surfaces of formation is especially evident in the elon- molars slightly worn gate and gracefully curved bony sheath that (fig. 3) 8 AMERICAN MUSEUM NOVITATES NO. 2850

Fig. 4. Bunomys chrysocomus. An adult caught in primary tropical evergreen rainforest at 1000 m in Lindu Valley, Central Sulawesi. Photographed by Margareta Becker. encloses the incisor anterior to the molar row, ofprimary forests and feeds mainly on insects the relatively low ramus beneath the first mo- (beetle larvae, moths, cicadids, crickets, ka- lar, and the lingual ridge behind the molars, tydids, cockroaches, and wasps), earth- which extends posteriorly and then gradually worms, snails, small frogs and lizards, and dorsad. In Papagomys, the incisor capsule is occasionally fruit. The mandible is low and chunky, bends dorsad at a sharper angle, and elongate, the lower incisor long and sharp, is shorter relative to the body of the ramus; the rostrum long (fig. 5), and the molars sim- the body of the ramus is relatively higher; ple in occlusal patterns (see the figures in and the shelf behind the molar row extends Musser and Newcomb, 1983, pp. 398-399). back and dorsad at a sharper angle. The com- The rat is a good invertebrate predator, but bination of elongate dentary; long, slender, does not have the extreme specializations and sharp lower incisor; and small simple characteristic ofthe native Sulawesian shrew teeth set in a long dentary as seen in P. naso rats such as Tateomys rhinogradoides and T. recalls characters associated with murids that macrocercus, which apparently feed only on have a long rostrum and attendant adapta- earthworms (Musser, 1982). tions pointing to terrestrial life in wet forest The new material provides little additional and a diet of invertebrates (Musser, 1982). data useful in identifying the closest relative A possible living analog ofPaulamys naso of Paulamys naso. The samples only rein- is the terrestrial Bunomys chrysocomus, which force Musser's (1981) earlier conclusion that is native to Sulawesi. About the size of a there are no specialized characters ofthe den- houserat (see the measurements listed by tary, incisor, or molars which associate P. Musser and Newcomb, 1983, p. 394), B. naso closely with any murid known from the chrysocomus is soft-furred and dark; it has a Indo-Australian region outside ofFlores, and tail shorter than length of head and body, that among Floresian murids, P. naso mor- small eyes, and a long snout (fig. 4). It is phologically clusters with the described common under dense cover in the wetter parts species of Papagomys, Komodomys, and 1986 MUSSER, VAN DE WEERD, AND STRASSER: PA ULAMYS 9

, AF t. : i;. m %}9

S51 k

i L 1ll ''' : E:,, }e # " t f..^\

S4 t *

Lt lll.m . dt 9 ..! t' A, .4, k .11., ..'. "'T 01,11,

" N.". . .4v . Fig. 5. Cranium and left dentary of Bunomys chrysocomus (American Museum of Natural History 224698) from Central Sulawesi. Approximately x 2.

Hooijeromys rather than with Spelaeomys. naso provide limited phylogenetic informa- We hypothesize that P. naso may be related tion, they support a view that P. naso is a to species of Papagomys. Among the new unique member of an assemblage of murids samples ofPapagomys which we will discuss native to Flores and possibly other islands in in a later paper, there are two undescribed Nusatenggara. The long-nosed morphology species with molar characteristics resembling of P. naso, implied from dentary conforma- those seen in P. naso. We may be able to say tion, reflects a particular ecological position more about the phylogenetic relationship of within a murid fauna that includes a gerbil- P. naso in the context of our final report, like savanna species (Komodomys rintjanus), which will identify all the taxa in the new at least one and possibly more terrestrial for- samples. est herbivores (some species of Papagomys), Although the new specimens of Paulamys arboreal herbivores (Spelaeomysflorensis and 10 AMERICAN MUSEUM NOVITATES NO. 2850 possibly one species of Papagomys), and a problems pertaining to the racial history terrestrial shrew rat (forest or grassland) that ofthe Indonesian Region. Doctoral dis- is specialized (more so than Paulamys naso) sertation, Rijksuniversiteit, Utrecht. for a diet of earthworms. Other than Papa- Jacob, T. gomys armandvillei on Flores and Komod- 1967. Some problems pertaining to the racial history of the Indonesian Region. Doc- omys rintjanus on the islands of Padar and toral dissertation, Rijksuniversiteit, Rintja, we do not know ifPaulamys naso or Utrecht, pp. 1-156, pls. 1-49. any ofthe other rats that are represented only Musser, Guy G. by subfossil fragments still survive on Flores 1972. Identities oftaxa associated with Rattus or other islands in Nusatenggara. Flores re- rattus (Rodentia, Muridae) ofSumba Is- quires better biological survey than it has re- land, Indonesia. Jour. Mammal., vol. ceived to determine just what murids are still 53, pp. 861-865. living there. 1981. The giant rat of Flores and its relatives east of Borneo and Bali. Bull. Amer. LITERATURE CITED Mus. Nat. Hist., vol. 169, pp. 67-176, figs. 1-40. Ehrat, H. 1982. Results of the Archbold Expeditions. 1928. Geologisch-mijbouwkundige onder- No. 110. Crunomys and the small-bod- zoekingen op Flores. Jaarb. Mijnw., ied shrew rats native to the Philippine 1925, Verh. 2, pp. 221-315. Islands and Sulawesi (Celebes). Ibid., Fries, Carl, Jr., Claude W. Hibbard, and David H. vol. 174, pp. 1-95, figs. 1-60. Dunkle Musser, Guy G., and Cameron Newcomb 1955. Early Cenozoic vertebrates in the Red 1983. Malaysian murids and the giant rat of Conglomerate at Guanajuato, Mexico. Sumatra. Bull. Amer. Mus. Nat. Hist., Smith. Misc. Coll., vol. 123, no. 7, pp. vol. 174, pp. 327-598, figs. 1-117. 1-25, figs. 1-6, 1 pl. Verhoeven, Theodor Hooijer, D. A. 1968. Vorgeschichtliche Forschungen aufFlo- 1957. Three new giant prehistoric rats from res, Timor und Sumba. In Anthropica, Flores Lesser Sunda Islands. Zool. Med., Gedenkschrift zum 100. Geburtstag von vol. 35, pp. 229-314, pls. 1-2. P. W. Schmidt (Studia Instituti Anthro- 1967. Mammalian remains from Liang Toge, pos, vol. 21), St. Augustin, pp. 393-403, Flores. Appendix II. In T. Jacob, Some 1 fig.

Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates, Bulletin, and Anthropological Papers published during the last five years are available free of charge. Address orders to: American Museum of Natural History Library, Department D, Central Park West at 79th St., New York, New York 10024.