Prosciurillus murinus

Musser, Guy G., Durden, Lance A., Holden, Mary Ellen & Light, Jessica E., 2010, Systematic Review of Endemic Sulawesi Squirrels (Rodentia, Sciuridae), with Descriptions of New Species of Associated Sucking Lice (Insecta, Anoplura), and Phylogenetic and Zoogeographic Assessments of Sciurid Lice, Bulletin of the American Museum of Natural History 2010 (339), pp. 1-260 : 1-260

The Prosciurillus murinus Group

Two species comprise this cluster, both the smallest in body size among all the tree squirrels and ground squirrels endemic to Sulawesi. One is Prosciurillus murinus , named and described by Müller and Schlegel in

TABLE 34 Allocation of Scientific Names Associated with Species in the Prosciurillus murinus Group as Presented in Primary Published Checklists of Mammals from 1940 to the Present Generic and specific names as originally published are listed in their published chronological order. See the species accounts for names of authors and dates of publication.

1844, which has been collected from most of mainland Sulawesi where it occurs through an altitudinal gradient from coastal lowlands to mountains and within the range of forest formations from tropical lowland evergreen rain forest to upper montane rain forest; specimens were also taken from two small islands off the northeastern coast of the northern peninsula. The other is Prosciurillus abstrusus , named and described by Moore in 1958; it has only been collected at 1500 m and 2000 m in montane forest on Pegunungan Mekongga , the highest range on the southeastern peninsula of Sulawesi .

Although originally described as a species of Sciurus , the genus used for most new species of squirrels described during the 1800s and early 1900s, murinus was eventually transferred to Sciurillus ( Ellerman, 1940: 318) then subsequently designated the typespecies of Prosciurillus ( Ellerman, 1947: 259) where it has remained (see table 34 and introduction to Prosciurillus ).

Three scientific names have been associat- ed with murinus , one described as a separate species, the other two as subspecies of murinus ; after 1947 all were either recognized as subspecies or synonyms (table 34). From our study of skins and skulls, we can delineate a single species, P. murinus , and similar kinds of data sources indicate P. abstrusus to be a separate but related species. Concordant with the morphological data identifying two distinctive species are their sucking lice parasites: each species of squirrel supports a distinctive species of Hoplopleura (see the descriptions of sucking lice).

Definitions of the species limits of P. murinus and P. abstrusus , based on qualitative evaluation of variation in coat color combined with morphometric analyses of cranial and dental dimensions, along with summaries of their geographic and altitudinal distributions derived from collection localities of voucher specimens, and contrasts between the two, are provided in the following accounts.

GAZETTEER AND SPECIMENS EXAMINED: Collection localities for the 197 specimens of P. murinus and the 30 examples of P. abstrusus studied are listed below. The number preceding each locality keys to a symgbol on the map in figure 30.

Prosciurillus murinus

1. Pulau Talise, 01 ° 509N, 125 ° 039E, sea level (highest point on island is 264 m): BMNH 87.7.15.1.

2. Teteamoet (several miles southeast of Likupang; see Raven’s brief description of the place in Miller, 1917: 29), 01 ° 409N, 125 ° 059E (estimated from Raven’s map), on coastal plain near sea level: USNM 216793–95 .

3. Manado (also spelled ‘‘Menado’’), 01 ° 309N, 124 ° 509E, coastal plain near sea level: ANZP 14156; BMNH 88.7.15.1, 97.1.2. 18; MZB 5975, 5978; RMNH 13216–18 (specimens ‘‘ d–f ’’ in Jentink’s [1888: 22] catalog; specimen ‘‘ e ’’ is actually from ‘‘Ménado-Langowan).

4. Gunung Klabat (‘‘Goenoeng Kalabat’’ on specimen labels), 01 ° 289N, 125 ° 029E: AMNH 196512 (‘‘Koemersot,’’ 250 m on skin label); USNM 217809 (5600 ft [1707 m]). This specimen was collected by H.C. Raven, who provided this note (quoted in Riley, 1924: 1): ‘‘My camp was at about 1,700 meters (5,600 feet), where the forest is only semitropical in appearance. Practically all the trees are heavily coated with moss and are not tall. Kalabat is the highest mountain in Minahassa, having an altitude of 2,020 (6,617) feet.’’

5. Pulau Lembeh, 01 ° 269N, 125 ° 139E (elevation of collection site is not recorded but highest point of the island is 447 m): BMNH 40.631, 40.632; SNSD 3081; USNM 217810–13, 217814 (holotype of Sciurus evidens ), 217815.

6. Rurukan, 01 ° 219N, 124 ° 529E: AMNH 196510 (900 m), 196511 (1000 m); BMNH 97.1.2.17, no elevation is recorded for the specimen, but Rurukan is at 1000 m; Sheet NA 51-12).

7. Tonsealama (also known as ‘‘Tonsea’’), 01 ° 199N, 124 ° 559E, 600–700 m (estimated from Sheet NA 51-12): MZB 5976, 5977.

8. Tomohon, 01 ° 199N, 124 ° 499E, 700–800 m (estimated from Sheet NA 51-12): BMNH 99.10.1.6. Meyer (1898: 21) reported specimens from Tomohon.

9. Temboan (on Kuala Kalait, ‘‘is a new clearing of eight houses and lies from Mt. Sapoetan south, 55 ° west and about six miles from Loboe,’’ wrote Raven in his field journal, 1916: 3 (in mammal division library, USNM), 01 ° 039N, 124 ° 339E (estimated from Raven’s map), 500 m (estimated from Sheet NA 51-12): USNM 217816–21.

10. Amurang, 01 ° 119N, 124 ° 359E, on coastal plain near sea level: MZB 1502–1505.

11. Pinogu, Bone Valley (also spelled Pinogoe), 00 ° 329N, 123 ° 259E, 200 m (estimated from Sheet NA 51-14): BMNH 99.10.1.23.

12. Gorontalo, 00 ° 319N, 123 ° 039E, coastal plain near sea level: RMNH 13215 (specimen ‘‘ c ’’ in Jentink’s [1888: 22] catalog).

13. Bumbulan, 00 ° 299N, 122 ° 049E, coastal plain near sea level: AMNH 152921, 152932, 152938–43, 153281, 153282; MZB 5973; RMNH 9827 (formerly MZB 5974; holotype of Sciurillus murinus griseus ).

14. Sungai Paleleh (‘‘I went inland about four or five miles over the mountains and made camp at the edge of the Paleleh River, which is a small brook … with steep mountains or hills on all sides’’ [Raven, in Miller, 1915: 23]), 00 ° 599N, 121 ° 499E (estimated from Raven’s map), 700 m (estimated from Sheet NA 51-14): USNM 200275–78.

15. Matinan (also spelled Matinang), 01 ° 059N, 121 ° 429E (for village near coast; specimens came from highlands south of village in foothills of Pegunungan Peleleh at 600 m): AMNH 196513–19. Meyer (1898: 21) report- ed a specimen from ‘‘Matinangkette’’ (Matinan range, part of the larger Pegunungan Peleleh) at 1000 m.

16. Gunung Ile-Ile, 00 ° 589N, 121 ° 489E (part of the larger Pegunungan Peleleh forming the mountainous backbone of the northwestern portion of the northern peninsula): AMNH 196524–28 (500 m), 196520–23 (1700 m).

17. Valley of Sungai Miu, Sungai Oha Kecil (small tributary on left side of Sungai Miu), 01 ° 229S, 119 ° 579E (near confluence with Sungai Miu; estimated from Sheet SA 50- 8), 290 m: AMNH 224589–91.

18. Sungai Oha Kecil, 1300 ft (396 m): AMNH 224592.

19. Sungai Oha Kecil, 1400 ft (427 m): AMNH 224593.

20. Sungai Oha Kecil, 1600 ft (488 m): AMNH 224594, 224595.

21. Valley of Sungai Miu, Sungai Miu (right side), 01 ° 239S, 119 ° 589E (estimated from Sheet SA 50-8), 350 m: AMNH 224046–49.

22. Valley of Sungai Miu, Sungai Sadaunta (also spelled ‘‘Sidaonta’’ or ‘‘Sidaunta’’; tributary on right side of Sungai Miu), 01 ° 239S, 119 ° 589E (estimated from Sheet SA 50-8), 675 m: AMNH 224043–45.

23. Sungai Sadaunta, 2500 ft (762 m): AMNH 224601–603; 2550 ft (777 m), AMNH 224604.

24. Sungai Sadaunta, 2700 ft (823 m): AMNH 224596, 224597, 224605–607.

25. Sungai Sadaunta: 2800 ft (854 m), AMNH 224598, 224608, 224609; 2850 ft (869 m), AMNH 224610.

26. Sungai Sadaunta: 2900 ft (884 m), AMNH 224599, 224611.

27. Sungai Sadaunta: 3000 ft (915 m), AMNH 224600, 224612, 224613; 3050 ft (930 m), AMNH 226836.

28. Sungai Sadaunta, 3200 ft (976 m), AMNH 224614, 226837; 3250 ft (991 m), AMNH 224615.

29. Sungai Sadaunta, 3300 ft (1006 m): AMNH 224616.

30. Sungai Sadaunta, 3400 ft (1037 m): AMNH 224617.

31. Valley of Danau Lindu, Tomado (a village on western shore of Danau Lindu), 01 ° 199S, 120 ° 039E (estimated from Sheet SA 50-8), 1000 m: AMNH 223021, 223022, 223475, 224050, 224051.

32. Gunung Kanino, 01 ° 179S, 120 ° 089E (estimated from Sheet SA 50-8), 4600 ft (1402 m): AMNH 225492; 4650 ft (1418 m), AMNH 225493–96.

33. Gunung Kanino, 4750 ft (1448 m): AMNH 225499; 4800 ft (1463 m), AMNH 223536, 223538, 223539.

34. Gunung Kanino, 4900 ft (1494 m), AMNH 225500.

35. Gunung Kanino, 5200 ft (1585 m): AMNH 223537.

36. Gunung Nokilalaki, 01 ° 139S, 120 ° 089E, 7000 ft (2134 m): AMNH 225497; 7100 ft (2187 m), AMNH 223498.

37. Malakosa, Kuala Navusu, 00 ° 589S, 120 ° 279E (estimated from Sheet SA 51-1): 100 ft (30 m), AMNH 226063–66; 125 ft (38 m), AMNH 226067, 226068; 150 ft (46 m), AMNH 226069; 180 ft (55 m), AMNH 226070; 200 ft (61 m), AMNH 225909; 350 ft (107 m), AMNH 226071; 450 ft (139 m), AMNH 226072, 226073; 600 ft (182 m), AMNH 226074; 750 ft (229 m), AMNH 226075.

38. Tolai, Sungai Tolewonu, 01 ° 049S, 120 ° 279E (estimated from Sheet SA 50-8): 500 ft (152 m), AMNH 226500–502; 540 ft (166 m), ASE 4174; 600 ft (183 m), AMNH 226503, 226504; 650 ft (198 m), AMNH 226505; 700 ft (213 m), AMNH 226506; 750 ft (229 m), AMNH 226507; 800 ft (244 m), AMNH 226508; 1000 ft (305 m), AMNH 226509.

39. Pinedapa, 01 ° 259S, 120 ° 359E (estimated from Raven’s map), 100 ft (31 m): USNM 219504– 507, 219509.

40. Rano Rano 01 ° 309S, 120 ° 289E, 6000 ft (1829 m): USNM 219508, 219510.

41. Kulawi, 01 ° 279S, 119 ° 599E, 500 m: USNM 218713.

42. Gunung Lehio, 01 ° 339S, 119 ° 539E, above 6000 ft (1829 m): USNM 218712 (holotype of Sciurus murinus necopinus ).

43. Wawo, on the plain between the coast and western foothills of Pegunungan Mekongga (also spelled ‘‘Mengkoka’’), 03 ° 419S, 121 ° 029E, 50 m: AMNH 101390–93.

44. Pegunungan Mekongga, Masembo, which is southeast of Wawo and the highest place in the southern portion of Pegunungan Mekongga (see maps and discussion in Heinrich [1932] and Stresemann [1940]) 03 ° 359S, 121 ° 159E (for the Pegunungan), 550 m: AMNH 101379, 101394.

45. Pegunungan Takolekaju (‘‘Molengraff Range,’’ the old the European name for the mountains, is notated on the specimen tags; no more precise collection locality is available; P. Jenkins, in litt., 2008), an extensive mountainous region in the western portion of the central core with the north-south boundaries between approximately 01 ° and 02 ° 159S, and extending slightly diagonally from about 119 ° 459E in the north to 120 ° near its southern ramparts: BMNH 40.636 (1500 ft, 457 m), 40.365 (3200 ft, 976 m), and 40.367 (3500 ft, 1067 m).

46. Pegunungan Latimojong, 03 ° 309S, 121 ° 239E, 2200 m: AMNH 196529–33.

47. Lombasang (also spelled ‘‘Lambasang’’), northwest of Gunung Lompobatang, 05 ° 169S, 119 ° 509E, 1100 m: AMNH 101380–84, 101387–89.

48. Gunung Lompobatang (also spelled Lompobattang; see Fraser and Henson, 1996, for example; spelled ‘‘Pik v. Bonthain’’ or ‘‘Pik v. Bantaeng’’ on old maps), Wawokaraeng, 05 ° 209S, 119 ° 559E: 2000 m, AMNH 101385, 101386. Tasoso (spelled ‘‘Dasoso’’ on the skin tag; a village on the lower slopes and one of Alfred Everett’s collection localities [see Hartert, 1896]); 05 ° 179S, 119 ° 589E, 4000 ft (1320 m): BMNH 97.1.3.9.

The following specimens were studied but the places where they were caught have not been mapped.

A. ‘‘Nord de Célèbes’’: RMNH 13213 (specimen ‘‘ a ’’ in Jentink’s [1887: 190, 1888: 22] catalog), lectotype of Sciurus murinus (see account of P. murinus ). Sody (1949: 77) restricted the type locality to ‘‘the extreme N.E. part of the island.’’ ‘‘NE Celebes’’: RMNH 13214 (specimen ‘‘ b ’’ in Jentink’s [1888: 22] catalog), paralectotype of Sciurus murinus ; RMNH 13219 (skull ‘‘ c ’’ in Jentink’s [1887: 190] osteological catalog), paralectotype of Sciurus murinus . ‘‘ Celebes ’’: NMB 1172.

B. Northeast peninsula, Minahasa District (Kabupaten Minahasa, between 1 ° and 2 ° north latitude): BMNH 40.633, 40.634 ; RMNH 24435, 24436; ZMB 92611–13.

C. Central core, Tamalanti: BMNH 40.639, 40.640 (3300 ft, 1006 m); 40.641 (3800 ft, 1159 m). Laurie and Hill (1954: 156) noted that Tamalanti is a ‘‘Plantation between Rantekaroa [02 ° 509S, 119 ° 509E] and Koelawi [01 ° 279S, 119 ° 599E],’’ which was also the only information Musser and Paula Jenkins found when they searched through field-notes and other documents at BMNH.

Prosciurillus abstrusus 1. Pegunungan Mekongga (also spelled ‘‘Mengkoka’’), Tanke Salokko (the highest spot in

Pegunungan Mekongga; see the maps and discussions in Heinrich [1932] and Stresemann [1940]), 03 ° 359S, 121 ° 159E (for the

Pegunungan): 1500 m, AMNH 101350,

101351; 2000 m, AMNH 101349, 101352–77,

101378 (holotype of Prosciurillus abstrusus ).

Prosciurillus murinus

(Müller and Schlegel, 1844)

Sciurus murinus Müller and Schlegel, 1844: 87 . Sciurus murinus necopinus Miller and Hollister,

1921: 98. Sciurus evidens Miller and Hollister, 1921: 99 . Sciurillus murinus griseus Sody, 1949: 77 .

LECTOTYPE AND TYPE LOCALITY: The lectotype of Prosciurillus murinus is an adult female (RMNH 13213, specimen ‘‘ a ’’ in Jentink’s [1888: 22] catalog) obtained by E.A. Forsten sometime during 1840–1842. It consists of a skin mounted in a live pose and a damaged cranium, the latter also listed as specimen ‘‘ a ’’ in Jentink’s (1887: 190) osteological catalog. The right zygomatic region is missing, the basicranial area is fragmented, the right bullar capsule is gone, and the cranium is cracked in several other places; both dentaries are missing. Values from dental and some cranial measurements are listed in table 35.

Specimen ‘‘ a ’’ was one of three specimens cataloged as types (5 syntypes) by Jentink (1887: 190, 1888: 22). At Chris Smeenk’s suggestion (in litt., 2008), we select specimen ‘‘ a ’’ as the lectotype following the rules promulgated in Article 74.1 of the Code ( ICZN, 1999: 82); G.H.H. Tate had scribbled ‘‘lectotype’’ on the specimen tag during his visit to Leiden in 1951 (C. Smeenk, in litt., 2008). The other two specimens, collected by E.A. Forsten between 1840 and 1842, become paralectotypes: specimen ‘‘ b ’’ ( Jentink, 1888: 22), RMNH 13214, is an adult male skin mounted in live pose with skull still inside the

TABLE 35

Age, Sex, and External, Cranial, and Dental Measurements (mm) for Types of Prosciurillus murinus (includes necopinus and evidens ), and Prosciurillus abstrusus mount; the osteological specimen ‘‘ c ’’ ( Jentink, 1887: 190), RMNH 13219, is a skull only. See the account of P. leucomus for a description of Forsten’s travels through northeastern Sulawesi.

The type locality is ‘‘Nord de Célèbes,’’ as listed in Jentink’s (1887: 190, 1888: 22) catalogs. Forsten obtained most of his material in the region near the tip of the northeastern peninsula, which is in the administrative district of ‘‘Minahasa’’ (Kabupaten Minahasa), between 1 ° and 2 ° N. Sody (1949: 77) restricted the type locality to ‘‘the extreme N.E. part of the island,’’ which effectively places the type locality in Kabupaten Minahasa, Propinsi Sulawesi Utara, Indonesia. However, the type locality should rather be defined as ‘‘NE Celebes’’ because Forsten also worked in the environs of Gorontalo and as far west as Paguat (September–November, 1841; see Van Steenis-Kruseman [1950: 179] and Forsten’s unpublished diary in the archives of the Leiden Museum); hence, we cannot exclude that one of the specimens is from that region (C. Smeenk, in litt., 2008).

EMENDED DIAGNOSIS: The smallest-bodied of the species of tree squirrels endemic to Sulawesi (see table 3), and further characterized by: (1) uniform dark brown upperparts flecked with buff, ochraceous, and black; (2) lack of ear tufts, nape patches, middorsal black stripe, or any other patterning formed by different colored stripes and patches; (3) dark grayish buff to ochraceous gray underparts; (4) tail that is shorter than length of head and body; and (5) small, gracile skull (figs. 12–14).

GEOGRAPHIC AND ALTITUDINAL DISTRI- BUTIONS: Except for the east-central peninsular region, collection localities for Prosciurillus murinus are scattered throughout mainland Sulawesi and are also from Talise and Lembeh islands off the coast of the northeastern peninsula (see gazetteer and map in fig. 30). Likely the species can be encountered anywhere on Sulawesi where forest persists, even the east-central peninsula, which has been poorly surveyed for small mammals, especially squirrels.

Prosciurillus murinus is found in forest formations ranging from tropical lowland evergreen rain forest throughout the coasts and middle altitudes to tropical upper montane rain forests in high mountains. On the northern peninsula, voucher specimens come from the coastal plain near sea level to 1700 m on Gunung Klabat and Gunung IleIle, although most examples were collected at places that are scattered between the coastal plain and 1000 m. Kuala Navusu at 30 m and Gunung Latimojong at 2200 bracket the range of elevations at which samples of P. murinus have been obtained in the central core of Sulawesi. On the southwestern peninsula, specimens come from 1100 m (Lombasang) and 2000 m (Gunung Lompobatang). Before most of the peninsula was converted from forests to agriculture (see account of Prosciurillus sp. ), P. murinus likely occurred in forests below 1100 m, down to and along the coastal plain.

Only two places on the southeastern peninsula have yielded voucher specimens: Wawo at 50 m near the coast and Masembo at 500 m inland, and this region forms an exception to the altitudinal range of P. murinus , extending from coastal plain to mountains, documented in the central core of the island and on the northern peninsula. Montane forests on Pegunungan Mekongga, the mountain range looming above Wawo and Masembo, support populations of P. abstrusus , a close relative of P. murinus ; samples of the former are from 1500 and 2000 m. The lower altitudinal limit of P. abstrusus in this highland region and the upper limit of P. murinus are unknown.

Pulau Talisei, 8 km northeast of the northwestern tip of the Sulawesi mainland, and Pulau Lembeh, directly off the western coast of the northeastern mainland tip (see map in fig. 28), are the collection sites of the only two insular samples of Prosciurillus murinus . Another squirrel, P. leucomus , and several other native Sulawesi mammals are recorded from Pulau Lembeh (see the account of P. leucomus ), but samples of only P. murinus and the macaque, Macaca nigra , come from Pulau Talisei. That island is separated from the mainland by two smaller islands and water depths of less than 50 m, mostly only 10 m (Sheet NA 51-12). As with Pulau Lembeh, Talisei was most likely part of mainland Sulawesi during Pleistocene glacial intervals when sea level was depressed by at least 120 m, and separated during warm inter-glacial periods (see the account of P. leucomus ). Neither squirrel nor macaque exhibits any significant morphological contrasts with mainland samples, suggesting relatively recent isolation on Pulau Talisei.

DESCRIPTION: Müller and Schlegel (1844: 87) described murinus with this sentence (translated from the original Dutch by Chris Smeenk, in litt., 2008; see appendix 2 for a rendition of the text in Dutch):

11.) Sciurus murinus, Forsten , n. sp. Appearance and shape as in the preceding species [ Sciurus modestus , currently a synonym of Sundasciurus tenuis , native to peninsular Thailand, Malaya, Sumatra, Borneo, and some smaller islands on the Sunda Shelf; see Corbet and Hill, 1992: 296], but slightly smaller; the tail slightly shorter; the ears less hairy, and the colour more uniform; above mouse-grey and the hairs with whitish yellow tips; below ash-gray. Celebes.

Prosciurillus murinus is the smallest in body size of the species of tree squirrels endemic to Sulawesi (length of head and body, 102–150 mm; length of hind foot, 32– 38 mm; weight, 42–110 g; extremes from table 36), and one of the smallest of the tree squirrels and ground squirrels that are native to the Indomalayan region. A few species of Sundaic Callosciurus ( C. adamsi and C. orestes ) and Sundasciurus ( S. lowii , S. tenis , S. jentinki , and S. brookei ) are similar to P. murinus in body size; only the species of pygmy tree squirrels in Nannosciurus and Exilisciurus are smaller ( Medway, 1969; Heaney, 1985; Payne et al., 1985).

TABLE 36 Descriptive Statistics for Measurements (mm) of Lengths of Head and Body, Tail, Hind Foot, and Ear, and for Weight (g), Derived from Samples of Prosciurillus murinus and Prosciurillus abtrusus a Mean ± 1 SD, observed range (in parentheses), and size of sample are listed. Mean values were used to compute LT/LHB. Specimens measured are listed in footnotes.b

Uniformly dark upperparts, unbroken by contrasting ear tufts, nape patches, midventral black strip, or any other contrasting adornment, and grayish underparts characterize the small and dark P. murinus . The dense and short (8–10 mm thick in most specimens) fur covering the upperparts from nose to base of the tail, including the ears and all appendages, is dark brown flecked with buff or ochraceous tints and peppered with black, the middorsal region slightly darker than sides of the head and body. The dark hue, conspicuously darker than seen in species of the P. leucomus group or Rubrisciurus rubriventer , results from the overhairs that are dark gray for most of their lengths, have a subterminal buff or ochraceous band, and are tipped with black. No tufts or postauricular patches distinguish the small ears from the rest of the dorsum. The eyes are circled by buffy rings.

The dense and short (up to 5 mm thick) fur covering the underparts, from chin to base of the tail, ranges from dark gray washed with pale buff to dark ochraceous gray in most samples (hairs are dark gray for most of their length and tipped with buff or ochraceous tints); a few specimens have dark grayish white venters (tips of the hairs are unpigmented). Most specimens fall in the range between dark grayish buff and ochraceous gray. The demarcation between upperparts and underparts in the latter is not sharp; for example, USNM 219508 from Rano Rano has a solid dark ochraceous gray venter that inconspicuously merges with coloration of the dorsal coat.

The tail is shorter than the length of the head and body (LT/LHB 5 68%–86%, see table 36) with the same color pattern as the dorsal fur—long black hairs with alternating distal ochraceous and black bands. Hairs forming tip of the tail are all black.

Females exhibit three pairs of teats, just as in members of the P. leucomus complex: one postaxillary pair and two inguinal pairs. A single embryo was found in each of the few females examined.

Views of the small, gracile skull are illustrated in figures 12–14; summary statistics for cranial and dental measurements are listed in tables 37 and 38. Conformations of the skull and teeth are basically miniatures of those elements in the P. leucomus group.

COMPARISONS: Prosciurillus murinus is markedly smaller in body size than any other species of tree squirrel or ground squirrel endemic to Sulawesi except for the small-bodied P. abstrusus , which resembles P. murinus not only in body size but coat color. It is between these two that comparisons are required, which we document here.

Prosciurillus abstrusus is still known only by the original 30 specimens collected by Gerd Heinrich in 1932 at 1500 and 2000 m on Pegunungan Mekongga, the highest range on the southeastern peninsula of Sulawesi. In diagnosing the species, Moore (1958: 3) wrote that it ‘‘has short white pelage on the backs of the ears which distinguish it from skins of … Prosciurillus murinus ; otherwise only size distinguishes the stuffed skins, P. abstrusus being larger than P. murinus . Moore also noted that P. abstrusus had underparts less intensely washed with buff or ochraceous hues.

Mean values of external measurements are greater when the large sample of P. abstrusus is contrasted with the very small sample of P. murinus from the lowlands of the southwestern peninsula, but such distinctions are not evident between P. abstrusus and the larger samples of P. murinus from the central core of Sulawesi and the northern peninsula (table 36). The pelage differences described by Moore, however, consistently distinguish the two species. Backs of the ears are covered with fine, short white hairs that form a white mat on all specimens of P. abstrusus , but we have not seen this pattern on any example of P. murinus —in that species, the fur covering the ears is the same color as the rest of the upperparts. Both P. abstrusus and P. murinus have similar dark brown dorsal fur in which the buffy or ochraceous hair tips provide a flecking over the dark brown background. The coat averages thicker in P. abstrusus compared with lowland samples of P. murinus (10–12 mm versus 8–10 mm, respectively), but is comparable in thickness to high mountain samples of P. murinus . All examples of P. abstrusus have dark grayish white venters (hairs are dark gray basally and unpigmented at their tips); some specimens show a pale buffy wash across the chest, and a few others have spots or larger patches of pale buff scattered over the venter. Most examples of P. murinus have brighter underparts, dark gray washed with buff or richer ochraceous tints; a few match the average in P. abstrusus —dark grayish white underparts.

While contrasts between the two species in lengths of head and body, tail, and hind feet are not impressive, P. abstrusus has an appreciably larger skull and longer tooth rows than does P. murinus . This difference can be appreciated qualitatively by visually comparing skulls of the two species side-by-side (figs. 12–14) and quantitatively by the contrasting mean values for cranial and dental dimensions summarized in tables 37 and 38. Nearly all cranial measurements average greater in the sample of P. abstrusus compared with P. murinus , and there is no overlap between the samples in the range of values for lengths of bulla and tooth row, and breadth of bony palate. A few dimensions are similar or smaller in P. abstrusus compared with P. murinus . Typically, the Mekongga squirrel has a relatively narrower interorbit, shorter nasals, rostrum, and diastema relative to occipitonasal length, and an absolutely shorter orbit.

The cranial and dental distinction between the two species is also reflected in the results of multivariate analyses of cranial and dental variables expressed by the distribution of specimen scores projected onto the first and second principal components extracted from principal-components analysis (fig. 31). The projection of scores along the first axis results in two discrete clusters, the one on the left representing all samples of P. murinus , that on the right signaling the larger specimens of P. abstrusus . Covariation in most dimensions influences this segregation of scores along the first principal component; lengths of bony palate, bulla, and tooth row along with breadths of rostrum and bony palate are especially strong (table 39). Prosciurillus abstrusus has relatively shorter nasals, rostrum, diastema, and orbit compared with P. murinus as indicated by their large and positive scores along the second axis, and relatively longer tooth rows as reflected by the large negative score for that dimension (table 39).

An osseous qualitative trait can be added to the absolute and proportional cranial dimensional differences between the two species. The smaller-bodied P. murinus has a conspicuously higher jugal component of the zygomatic arch, the dorsal border typically forming a high process. The larger-bodied P. abstrusus has a lower jugal, its dorsal border smooth or with only a low dorsal process (fig. 14).

Prosciurillus abstrusus inhabits montane forests on Pegunungan Mekongga where it replaces its close relative P. murinus . The latter was sampled by G. Heinrich in tropical lowland evergreen rain forest covering the foothills of the Pegunungan at Masembo, 550 m (four specimens), and on the coastal plain at Wawo, 50 m (two individuals). Color and length of the fur of these lowland samples closely resemble those in samples of P. murinus from lowlands elsewhere on Sulawesi; the ears of all, for example, are the same color as the neck and back, and the pelage is 8–10 mm thick. We could not quantify cranial differences between the two species because skulls of the six lowland specimens collected by Heinrich are too damaged to obtain complete sets of cranial measurements. The tooth rows, however, remain intact: mean and extreme values for the six (5.1 ± 0.06, 5.0– 5.2 mm) are nested within the range of variation shown by this variable for P. murinus (5.2 ± 0.17, 4.7– 5.6 mm, N 5 57) and not for P. abstrusus (6.2 ± 0.13, 5.9–6.3 mm, N 5 14).

At present, there is no information revealing how high P. murinus extends onto the mountain range or the lowest altitude to which P. abstrusus descends. We suspect the altitudinal ranges to be mutually exclusive, their boundaries approximately coinciding with the limits of lower montane forest and lowland evergreen rain forest. This altitudinal exclusion of P. murinus from a mountain range and its replacement at those higher altitudes by a different but closely related species is unique to the landscape of the southeastern peninsula. Elsewhere on mainland Sulawesi, populations of P. murinus occupy tropical lowland evergreen rain forests on the coastal plain, in middle-altitude hill forests, and up into the mountains where montane forest formations prevail.

TABLE 37 Descriptive Statistics for Cranial and Dental Measurements (mm) Derived from Population Samples of Prosciurillus murinus Mean ± 1 SD and observed range (in parentheses) are listed.

GEOGRAPHIC VARIATION: In addition to murinus , three other scientific names have been attached to geographic samples of the species, diagnosed primarily by chromatic distinctions of the fur and a cranial measurement, diagnoses inferring the presence of detectable and possibly significant geographic variation within P. murinus . Our visual inspection of coat coloration combined with multivariate analyses of data derived from cranial and dental measurements in samples from throughout Sulawesi, however, do not reveal significant geographic variation in these features. Why we disregard the three scientific names (necopinus, evidens , and griseus) as identifying distinctive populations within P. murinus is detailed in the section covering synonyms. We explain here why we cannot, using the specimens at hand, identify diagnostic geographic subsets of the species.

In stark contrast to species in the Prosciurillus leucomus group, all samples of P. murinus closely resemble one another in body size (comparing adults; see table 36) and coloration of the fur covering upperparts and tail—all samples contain a small-bodied, dark squirrel with a tail shorter than the length of head and body. There is a difference

TABLE 38 Descriptive Statistics for Cranial and Dental Measurements (mm) Derived from Population Samples of Prosciurillus murinus , and from the Sample of Prosciurillus abstrusus Mean ± 1 SD and observed range (in parentheses) are listed.

in thickness of the dorsal coat that is correlated with altitude. The few samples from high mountains, those taken on Gunung Nokilalaki, Pegunungan Latimojong, and Gunung Lompobatang (see gazetteer and map in figure 30), for example, are inseparable in coat color from those collected at lower altitudes, but have a slightly thicker dorsal coat, up to 15 mm in the montane samples; 8–10 mm is the thickness typical of specimens collected in lowlands. But where samples are available from an altitudinal gradient, along Musser’s transect from Sungai Oha Kecil at 290 m to Gunung Nokilalaki at 2187 m (see gazetteer and fig. 30), for example, the thicker coat gradually gives way to a shorter coat at the lower collection sites.

There is some geographic variation in coloration of the fur covering underparts of the head and body. Dark gray washed with buff predominates in most samples; others are slightly paler (dark gray tinged with buff) or exhibit a richer hue (dark gray washed with ochraceous). The latter characterizes some samples from the central core of Sulawesi but also a few samples from the northeastern peninsula; paler venters predominate in samples obtained from terrain between those regions. A gradation from grayish buff to ochraceous gray can even be found within a single population sample. While of potential interest to studies involving coat-color variation, we cannot resolve the chromatic differences we see into geographic patterns that might reflect populations with distinct genetic histories.

Our comparison of data derived from measuring cranial and dental dimensions reveals variation in these quantitative traits that is associated more with age within our adult category than with geographic provenances of the samples. Specimen scores derived from population samples (identified in table 1) and projected onto the first and second principal components extracted from principal-components analysis illustrated in figure 32 form, with an exception, a single constellation without significant substructure pointing to identifiable clusters bounded by particular geographic regions. Scores representing specimens from the northern peninsula, including Pulau Lembeh off the northeastern coast, comingle with those identifying individuals from the central core of Sulawesi. Covariation in most variables spreads the scores along the first principal component (table 40), a measure of size, reflecting both individual variation within an age class, and variation due to age (the range in age from young adults to old adults constitutes the samples). The only striking departure from a single constellation is the score representing the animal from Gunung Lehio, the holotype of necopinus, and the only sample collected from that mountain. Its position far to the left in the ordination reflects its smaller skull because the squirrel is younger relative to all the other specimens measured. Although the squirrel is clothed in adult pelage, the cheek teeth are slightly worn and the basisphenoid-

TABLE 39 Results of Principal-Components Analysis Contrasting Population Samples of Prosciurillus murinus with the Sample of Prosciurillus abtrusus Principal components are extracted from a covariance matrix of log-transformed values for 15 cranial and 1 dental variable; see figure 31.

presphenoid and basisphenoid-basioccipital sutures incompletely fused.

Three samples we studied are not included in the principal-components analysis. Specimens from Wawo and Masembo in lowlands of the southwestern peninsula (see gazetteer) have badly damaged skulls and we measured only the lengths of maxillary tooth rows. Mean and extreme values for six specimens (5.1 ± 0.06, 5.0– 5.2 mm) fall within the range of variation calculated for other geographic samples of P. murinus (see tables 37 and 38). We also measured length of the hind foot on the six dry skins (other body measurements recorded by the collector are unreliable and we could not obtain accurate values from the stuffed skins). The mean (33.5 ± 1.05 mm) and extremes (32–35 mm) are also comparable to the range in variation seen in other population samples of P. murinus (table 36).

Damaged, juvenile, or missing skulls also excluded samples from Pegunungan Latimojong (2200 m) in the central core of Sulawesi, and Gunung Lompobatang (2000 m) on the southwestern peninsula ( Lombasang , 1100 m, Tasoso , 1320 m, and Wawokaraeng , 2000 m; see gazetteer and map in figure 30) from the multivariate analysis using all cranial variables. We did measure lengths of maxillary tooth rows. Two specimens from Pegunungan Latimojong (each with a tooth row of 5.5 mm), and eight specimens from Gunung Lompobatang (5.5 ± 0.15, 5.3–5.7 mm) average slightly greater than in other geographic samples of P. murinus (tables 37, 38). We detected no significant difference in coloration of fur between these highland samples and those from comparable altitudes in the central core and northern peninsula of the island, and saw no significant size or conformational distinctions when we compared, side-by-side, the few less damaged skulls with those in other geographic samples. We also obtained mean and extreme values for length of hind foot, which we measured on the dry skins and used as an estimate of body size. (The samples were collected by G. Heinrich and his values for length of head and body along with length of tail are suspect, and we could not remeasure accurately these dimensions on the skins). We measured four specimens from the Latimojong range (34.8 ± 0.50, 34–35 mm) and 10 from the Lompobatang volcano (35.6 ± 0.52, 35–36 mm), which are comparable to the ranges of variation calculated for other population samples of P. murinus (table 36) .

The pattern of geographic variation in traits associated with coloration of fur and cranial and dental dimensions obtained from samples of P. murinus collected from most of Sulawesi is similar to the pattern derived from samples of Rubrisciurus rubriventer . Individuals of P. murinus , the smallest-bodied of the endemic Sulawesian tree squirrels, and R. rubriventer , the largest Sulawesi tree squirrel in body size, are instantly recognizable wherever they are encountered on the island. Unlike species in the P. leucomus group where particular color patterns, along with average morphometric distinctions, are tied to different regions of mainland Sulawesi — P. leucomus on the northern peninsula, P. topapuensis in the highlands forming the western mountain block in the central core of the island, P. alstoni occurring over the eastern part of the central core and on the southeastern peninsula, and P. weberi at the lower end of the central core—no patterns of pelage color and morphometric variation in cranial and dental variables concordant with geographic regions characterize the available samples of P. murinus and R. rubrisciurus .

ECOLOGY: The diurnal and arboreal Prosciurillus murinus is common in understory habitats of primary forest that bracket the range from lowland evergreen rain forest habitats (figs. 8, 22, 23, 33) to montane landscapes (figs. 27, 34, 41, 42), and warm to cool ambient temperatures (table 2); forests along streams, on hillsides, and on ridgetops were all occupied. Musser saw or heard P. murinus in the crowns of understory trees, low on the large trunks of high canopy and emergent trees, traveling over trunks and limbs on the forest floor, or foraging on the ground. The squirrels use the limbs and branches of understory trees and their attached woody vines as pathways through the understory canopy. On the ground they

TABLE 40 Results of Principal-Components Analysis Contrasting Population Samples of Prosciurillus murinus Principal components are extracted from a covariance matrix of log-transformed values for 15 cranial and 1 dental variable; see figure 32.

dash along tops of trunks and limbs lying in undergrowth on the forest floor, or dart from beneath one rotting tree trunk or pile of limbs to another. Many squirrels were trapped on tree trunks and limbs lying across streams and ravines (see summary of trapping sites in table 41). Musser saw one squirrel about 40 ft from the ground on the trunk of a canopy tree, but never noticed them any higher. The squirrels may forage into the upper canopy level and crowns of emergent trees, but they move very quickly, and would be difficult to locate unless moving slowly while concentrating on foraging.

In contrast to the wary and quiet Rubrisciurus rubriventer , a relative giant inhabiting the same understory habitats as P. murinus , the diminutive squirrels are aggressively vocal and not shy. Musser frequently encountered them when checking his traps. Once the squirrels detected his presence they became agitated, scolding from the ground hidden in the undergrowth, from looping woody vines and crowns of small trees, or on the lower trunks of the large canopy formers and emergents. Usually Musser heard the squirrels but could not see them. Those exposed on large tree trunks would soon tire of the confrontation and scurry down the trunk to the ground where they disappeared from sight in the undergrowth. Musser never saw a squirrel run up the trunk into the upper canopy. Vocalizations within the context of these interactions ranged from high-pitched chucks to almost whistles, to bird-like stacatto trills. Durden sometimes saw P. murinus in Bogani Nani Wartabone (then Dumoga-Bone) National Park in northern Sulawesi. As he sat quietly on a log removing ectoparasites from anaesthetized forest rats, individual P. murinus would occasionally descend to nearby tree trunks and chatter at the human intruder.

Musser occasionally saw P. murinus foraging on the ground but mostly in crowns of understory trees or on surfaces of large trunks. Two instances are illustrative. About 30 minutes before dark in lower montane rain forest on Gunung Kanino at 1500 m, Musser watched a squirrel working about 40 ft from the ground in the canopy of a small tree. The squirrel was moving head-first down a tree limb, its body pressed against the bark. It climbed over the lower branches and trunk in slow, jerky movements with its muzzle against the bark and appeared to be scoring the bark or digging into it. The squirrel would cover 2 ft of bark this way, then scamper to another part of the tree and work another section. Musser could not determine whether the squirrel was after a substance in the bark or listening for the movement of insect larvae beneath it.

Another time at about 4:00 P.M., Musser watched three squirrels foraging in the crowns of understory figs that canopied a section of the Sadaunta stream in lowland evergreen rain forest at 800 m. Each squirrel crawled along a branch with its body appressed against the bark, head down. The tail would alternately be held flat out, then quickly twitched upward towards the head, the motion accompanied by a bird-like, loud chirp. As the squirrels crawled along the branch, they occasionally stopped and dug their incisors into the bark, then moved on. They worked along a large branch to where it divided into smaller limbs, then jumped to other branches and began again to crawl and gnaw. The three worked over the branches for about 30 minutes, continually flagging their tails and chirping as they went, and occasionally emitting series of deeper chucks and chirps similar to the alarm call. None showed any aggression toward the other; often two would approach one another on the same branch, one working along the top of the limb, the other crawling on the limb’s undersurface, passing by and continuing on.

Stomach contents reveal a diet of soft fruits, seeds, and insects (tables 41, 57). Some stomachs contained unidentifiable brown or tan fruit mash from soft fruits; others had remains of figs (the tissuelike endosperm, rubbery rind, and tiny seeds) and large, hard pyramidal seeds from another kind of fruit. Insects recovered from stomachs consisted of macrolepidopteran larvae (caterpillars); geometrid larvae (inch-worm moths); a primitive group of scale insects (Coccoidea, Sternorrhyncha, Margarodidae ); arboreal termites (workers and soldiers in Termitidae ); small adult beetles (represented by elytras, sclerite fragments, wings, legs, and antennal segments); large adult beetles chopped into pieces; larvae of buprestid beetles (small to large instars); and other kinds of beetle larvae, cockroaches, and cranefly larvae ( Tipulidae ). Both groups of caterpillars feed on foliage. All the scale insects found are sessile females. Appressed to the surface of bark and leaf stems, they feed by embedding their mouthparts into the vascular system of the tree. Nests of termites are attached to limbs and smaller branches. The buprestid larvae reside beneath bark, other kinds of beetle larvae could be in decaying sections of limbs or beneath bark, and adult beetles roam over surfaces of leaves and bark. The dipteran tipulid larvae live in damp soil, very decayed leaf litter, or very rotting wood on the ground (D. Grimaldi, personal commun., 2009). Cockroaches occur at all levels in the forest.

Some stomachs contained mostly fruit with only a few insects; others were packed with remains of caterpillars, scale insects, or termites along with a bit of fruit. The small caterpillars and beetle larvae were ingested whole, the larger species were usually chewed into pieces; scale insects were either intact or had been rendered into chewed fragments. Adult beetles were always represented by chewed fragments in the stomachs. The squirrels glean the caterpillars from the leaves, pick off the scale insects from surfaces of branches, scratch the termite nests and grab the insects as they pour out, dig into bark for the beetle larvae, and snatch adult beetles and cockroaches from leaves and tree limbs. On the ground they would scratch out the tipulid larvae, and also pick off any adult beetles or cockroaches encountered.

Musser caught only one squirrel in a live-trap. It ignored all foods offered and was released the next day.

ECTOPARASITES: A sucking louse Hoplopleura murinus , n. sp. (see following section for description), uniquely parasitizes Prosciurillus murinus (table 56). This small-bodied squirrel is also the host for two species of flea ( Siphonaptera , Pygiopsyllidae ): Farhangia quatturordecimdentata is recorded from voucher hosts collected at Tomado and Sungai Oha Kecil in central Sulawesi ( Mardon and Durden, 2003), and F. celebensis was found on hosts obtained from Pulau Lembeh off the northeastern coast of the northern peninsula ( Ewing, 1924; Traub, 1980). Farhangia quatturordecimdentata is also parasitic on the arboreal murid rodent, Margaretamys beccarii , another Sulawesian endemic confined to lowland tropical evergreen rain forest ( Musser, 1982), but Farhangia celebensis is known only from Prosciurillus murinus ( Mardon and Durden, 2003) . Farhangia spp. are ‘‘nest fleas’’ ( Traub, 1980) and would therefore be expected to be more common in nests of P. murinus than on the host. Ectoparasitic laelapid mites were recovered from P. murinus pelt AMNH 224589 collect- ed at Sungai Oha Kecil in central Sulawesi in 1974 at an elevation of 290 m.

SYMPATRY: Over mainland Sulawesi, the geographic range of P. murinus overlaps those of Rubrisciurus rubriventer , both species of Hyosciurus , and all species of Prosciurillus except for P. abstrusus (table 6). Prosciurillus murinus and P. leucomus are also sympatric on Pulau Lembeh, off the coast of the northeastern peninsula (table 5). Along his transect in the northern portion of central Sulawesi, Musser trapped P. murinus in the same traplines, and sometimes in the same

TABLE 41 Summary of Habitat at Trapping Sites, Stomach Contents, and Other Relevant Information for Specimens of Prosciurillus murinus Collected by Musser in Central Sulawesi, 1973–1976 Collection locality, specimen number, elevation, and month and year of collection are included. Descriptions of the trapping sites and contents of stomachs, slightly edited, are from Musser’s field journals (in mammalogy archives at AMNH). Six of the collection localities (Sungai Oha Kecil, Sungai Miu, Sungai Sadaunta, Tomado, Kuala Navusu, and Sungai Tolewonu) are in tropical lowland evergreen rain forest; lower montane rain forest describes the places on Gunung Kanino, and Gunung Nokilalaki is mantled in tropical upper montane rain forest. With exceptions as indicated, all squirrels were caught during the day in Conibear traps (rats taken in the same traps were caught during the night). Unless noted differently, trapping sites were in primary forest formations. Locality, AMNH Elevation and (ASE) numbers (m) Date Trap site and other information

Sungai Oha Kecil On small, dry but rotting trunk lying across stream.

224589 (1757) Aug. 1974

224590 (1867) 290

224591 (1709) 290 Aug. 1974 In runway beneath rotting trunk lying on ground in partially cut forest covering stream terrace.

224593 (1877) 427 Aug. 1974 On living trunk sprouting leafy branches lying across stream in good forest.

224592 (1886) 396 Aug. 1974 On trunk, still sprouting shoots, lying over stream suspended between rocks about 10–12 ft above water surface.

224594 (1898) 488 Aug. 1974 On rotten Pandanus trunk (smooth, damp, no moss, vines, or other vegetation covering it) lying across stream 3 ft above water level. Streamside slopes steep but well covered with good undergrowth and forest of short trees, palms (including Pigafetta ) and Pandanus . The rat, Paruromys dominator , was trapped at same spot.

224595 (1887) 488 Aug. 1974 On sprouting and vine-covered trunk lying across stream in good streamside forest.

Sungai Miu Mar. 1974 On tree growing over stream in streamside forest.

224046 (1483) 350

224047 (1484) 350 Mar. 1974 On top of pile of dead limbs at the bottom of a steep bank next to river. Stomach: empty except for remnants of reddish brown fruit.

224048 (1511) 350 Mar. 1974 On wet, decaying trunk straddling stream, caught in early morning. Stomach: nearly empty, some old fruit mash and a sprinkling of termite ( Termitidae ) remains.

224049 (1536) 350 Mar. 1974 On wet, decaying trunk lying across stream in thick streamside forest. Stomach: nearly empty, masticated pieces of one or two small adult beetles; no fruit.

Sungai Sadaunta 675 Feb. 1974 Five ft above ground on top of huge rotting trunk of giant

224043 (1434) canopy tree lying in dense understory (shrubs, ferns, crisscrossed by woody vines) in primary forest adjacent to stream. A Rubrisciurus rubriventer and the rat, Taeromys celebensis , were caught in same spot. Stomach: full of reddish brown soft fruit remains, some pieces of fig, numerous macrolepidopteran caterpillar larvae and a few termitid termite soldiers.

224044 (1435) 675 Feb. 1974 On rotting trunk lying across stream in primary streamside forest. Stomach empty.

TABLE 41 (Continued)

Locality, AMNH Elevation and (ASE) numbers (m) Date Trap site and other information

224045 (1439) 675 Feb. 1974 On rotting, dead limb, covered with leafy vines, lying across stream in primary streamside forest. The rats, Rattus hoffmanni and Maxomys musschenbroekii were trapped on same spot.

Stomach: partially full of same reddish brown fruit mash as found in other P. murinus stomachs along with remains of a macrolepidoptern caterpillar and legs from small adult insects, probably beetles.

224601 (2089) 762 Oct. 1974 On rotting, smooth and wet trunk lying across a steep ravine in hillside forest below trail and above camp. The rat, Maxomys musschenbroekii was caught at same spot.

224602 (2092) 762 Oct.1974 On wet, smooth and rotting trunk lying across stream near camp in streamside forest. A Prosciurillus topapuensis and a rat, Maxomys musschenbroekii , were taken in same spot.

224603 (2124) 762 Oct. 1974 On trunk lying across stream near camp, caught about 10:00 a.m.

Stomach: fullof fig remains (mostly the rubbery rind with some tiny black seeds still attached to the inner tissue), a few hard and black pyramidal seeds enclosed in tissue, and pieces of at least one small adult beetle (sclerites, elytra, wings, legs, antennae).

224604 (2399) 777 Dec. 1974 On rotting, moss-covered trunk, partially covered by vines and ferns, lying across Sungai Sadaunta upstream from camp. A Rubrisciurus rubriventer and the rat, Maxomys musschenbroeki , were trapped in same spot.

224596 (2062) 823 Sept. 1974 On wet, decomposing wanga ( Pigafetta filaris ) palm trunk lying across Sungai Sadaunta about 200 ft upstream from where Rubrisciurus ASE 2033 was trapped (see table 10). A Prosciurillus topapuensis was trapped at same spot on a different day.

224597 (2303) 823 Nov. 1974 On rotting, wet trunk (free of vegetation) lying across stream in dense understory of streamside forest. Just downstream the rat, Crunomys celebensis , was trapped on the stream terrace.

224605 (2060) 823 Sept. 1974 On huge, rotting, moss-covered trunk lying across Sungai Sadaunta in good streamside forest. Several examples of the rat, Paruromys dominator , were trapped nearby.

224606 (2072) 823 Sept. 1974 On wet, rotting wanga palm trunk lying on top of rocky terrace next to and above Sungai Sadaunta. Just downstream is a long trunk on which Rubrisciurus rubriventer ASE 2033 was caught (see table 10).

224607 (2154) 823 Oct. 1974 On wet, rotting wanga palm trunk straddling deep ravine running into Sungai Sadaunta. The rats, Bunomys chrysocomus and Rattus marmosurus were trapped at same spot; several B.

chrysocomus were trapped on an adjacent terrace.

224598 (2224) 854 Oct. 1974 On decaying, wet trunk lying through thick undergrowth of ferns, shrubs, and gingers over small stream just above trail in steep hillside forest. Stomach: full of tan fruit mash with chewed and cut pieces of a few macrolepidopteran caterpillars.

224608 (2079) 854 Sept. 1974 On long, wet, and decaying wanga palm trunk lying 5 ft above and across Sungai Sadaunta just upstream from area where some rats, Paruromys dominator , were trapped. The rat, Rattus hoffmanni was trapped in same spot as ASE 2079.

224609 (2170) 854 Oct. 1974 On wet and decaying wanga palm trunk lying across stream; trunk is smooth and free of vegetation.

TABLE 41 (Continued)

Locality, AMNH Elevation and (ASE) numbers (m) Date Trap site and other information

224610 (2171) 869 Oct. 1974 On rotting, half-shredded wanga palm trunk lying across Sungai Sadaunta only a few feet upstream from where a rat, Paruromys dominator , was caught on a different trunk bridging the stream. Prosciurillus topapuensis was caught at same spot on a different day.

224599 (2265) 884 Oct. 1974 On same moss-covered rotting limb and trunk lying on terrace adjacent to stream where two of the rats, Paruromys dominator , were also trapped, just upstream from where the rat, Echiothrix centrosa , was taken.

224611 (2125) 884 Oct. 1974 On moss and fern and epiphyte-covered large root from a huge canopy fig growing on edge of the stream terrace; root projects from base of tree across stream to other terrace, 7–10 ft above the water; the straddling root is nearly concealed by a covering mantle of moss, ferns, and epiphytes.

224600 (2266) 915 Oct. 1974 On limb of understory tree growing from one bank to the other

224612 (2311) over the main upper tributary of the Sungai Sadaunta. The trunk springs from one margin of the stream, leaning over it with its upper branches reclining on the opposite terrace. The main connecting limb (3–5 in. diameter) is covered by moss that is trampled down, indicating frequent use by small mammals. Trap was set about 7 ft above stream level. No. ASE 2311 was caught during November. The arboreal rat, Rattus marmosurus , was trapped in same spot.

224613 (2316) 915 Nov. 1974 On thick, smooth, dead limbs lying across upper tributary of Sungai Sadaunta. Limb is free of vegetation or moss, extends from one eroded bank up and over 2 ft above the water and down to the other bank; each bank is at stream level in a wide ravine below high narrow terraces. The rats, Bunomys sp. and Maxomys sp. , were trapped on same spot.

226836 (4341) 930 Mar. 1976 On limb of understory tree growing across ravine containing main upper tributary of the Sungai Sadaunta; base of the trunk is at edge of stream, with the trunk leaning over the water and its upper branches reclining on the opposite terrace. The main connecting limb (3–5 ft in diameter) on which the trap was placed is mossy, and the moss is trampled, indicating frequent use; trap set about 7 ft from stream level. In 1974, examples of three kinds of squirrels (the ground squirrel, Hyosciurus ileile ; the tree squirrels, Prosciurillus topapuensis and Rubrisciurus rubriventer ) and an arboreal rat ( Rattus marmosurus ) were taken in the same spot.

224614 (2361) 976 Nov. 1974 On rotting large trunk covered with moss, shrubs, and ferns lying across headwater tributary of Sungai Sadaunta—steep sides of stream ravine supports dense wanga palms; figs form a dominant component of the understory.

226837 (4347) 976 Mar. 1976 Caught during morning before 7:00 a.m. on top of huge rotting trunk (4 ft diameter) that extends from one high terrace across stream to the opposite terrace. Stomach: nearly empty, a bit of fruit debris mixed with a few remains of macrolepidopteran caterpillars.

TABLE 41 (Continued)

Locality, AMNH Elevation and (ASE) numbers (m) Date Trap site and other information

224615 (2317) 991 Nov. 1974 On wet, moss-covered decaying limb lying across main headwater tributary of Sungai Sadaunta. Canyon here is Vshaped, covered with short forest typical of steep slopes near ridgetops. Understory figs are abundant along stream and form canopy over it; woody vines, the palm Caryota , occasional sugar palms, scattered wanga palms, many ferns, shrubs, and rattan comprise the streamside vegetation. The rat, Bunomys sp. , was trapped at same spot.

224616 (2328c) 1006 Nov. 1974 On decaying trunk lying on stream terrace in thick ferns and rattan upstream about 10 ft from where the rat, Taeromys celebensis , was caught.

224617 (2336) 1037 Nov. 1974 On branch lying through understory and extending across a shallow ravine, 3 ft off ground, in low shrubby forest between stream and trail.

Tomado 1000 Aug. 1973 Two traps set on two trunks straddling narrow stream in mature

223021 (524) forest. Both squirrels were caught during same day.

223022 (525)

223475 (552) 1000 Aug. 1973 On trunk lying in mature forest; trunk is part of large tree-fall that opened canopy in forest.

224050 (1621) 1000 May 1974 On decaying, moss-covered trunk bridging narrow steam in primary forest. Stomach: full of reddish brown soft fruit mash, packed with whole and chewed scale insects (Coccoidea, Sternorrhyncha, Margarodidae ), and a few small geometrid caterpillars (larvae of inch-worm moths) representing two species.

224051 (1622) 1000 May 1974 On decaying, moss-covered trunk lying across narrow steam in primary forest, just upstream from where ASE1621 was caught. Stomach: full of reddish brown remains of fig; pieces of macrolepidopteran caterpillars.

Gunung Kanino 1402 Feb. 1975 On small living branch of understory tree growing low across

225492 (2535) Sungai Salubeka. Stomach: partially filled with tan and gray fruit mash.

225493 (2446) 1418 Jan. 1975 On decaying, wet jumble of trunks and limbs from an old tree-

225496 (2449) fall straddling narrow stream below camp. A Rubrisciurus rubriventer was trapped on same spot on a different day. Stomach of ASE 2449: full of tan fruit mash.

225494 (2478) 1418 Feb. 1975 On wet, clear and rotting trunk lying across Sungai Salubeka near camp.

225499 (2517) 1418 Feb. 1975 On rotting trunk, wet and clear of moss and other vegetation, lying across small stream at bottom of shallow ravine with very wet and muddy slopes.

225492 (2535) 1402 Feb. 1975 On small living branch of understory tree growing low across Sungai Salubeka. Stomach: partially full of tan and gray fruit; no insects.

223536 (849) 1463 Nov. 1973 On trunk 2 ft from forest floor; caught about 8:00 a.m.

223538 (924) 1463 Nov. 1973 On dry but decaying trunk 2 ft from ground, lying over slope

223539 (935) near hillside next to river. The rat, Bunomys penitus , was trapped at same spot.

225500 (3046) 1494 May 1975 On decaying trunk (6 in. diameter, top clear, sides mossy) lying across stream extending from terrace to the opposite terrace and 5 ft above water surface. A Prosciurillus topapuensis was trapped at same spot on a different day.

TABLE 41 (Continued)

Locality, AMNH Elevation and (ASE) numbers (m) Date Trap site and other information

223537 (875) 1585 Nov. 1973 In runway beneath large rotting, moss-covered trunk in rattan understory of dammar ( Agathis ) grove.

Gunung Nokilalaki 2134 Apr. 1975 On ground in front of earth and rock face just above supply stop

225497 (2856) on trail.

225498 (2587) 2187 Mar. 1975 Sherman live-trap on ground in damp runway alongside rotting, moss-covered trunk lying on wet, rocky slope in moss forest. Caught about 10:00 a.m.

Kuala Navusu 31 Sept. 1975 On trunk (diameter 5 in.) of tree that had fallen across

226063 (3091) stream, connecting the stream terrace on one side to the opposite bank and bordering forest, trunk 5 ft above surface of water, just up from camp; caught yesterday in evening. Four Prosciurillus alstoni were trapped at same spot during different days.

226064 (3123) 31 Sept. 1975 On live trunk (6 in. diameter) growing across dry ravine on steep hillside below camp. Good hillside forest, broken in a few places by rotting tree-falls. Stomach: full of mostly brown fruit mash, many tiny larvae (small versions of those in ASE 4174), a few termitid termite workers (heads), several small macrolepidopteran caterpillars.

226065 (3155) 31 Sept. 1975 On a decomposing Livstonia palm trunk lying across stream near end of trapline. High, dense scrub covers streambanks in hillside forest partially thinned of canopy trees. Rubrisciurus rubriventer and rat, Maxomys hellwaldii were trapped at same spot. Stomach: full of reddish brown fruit mash mixed with remains of numerous small macrolepidopteran caterpillars.

226066 (3384) 31 Oct. 1975 On long, smooth trunk of Livistona palm that had fallen across part of stream terrace, connecting one terrace with that on the other side of the stream; trunk about 10 ft above water surface. A Prosciurillus alstoni was caught in same spot on a different day. Stomach: nearly empty, two geometrid caterpillars (larvae of inch-worm moths), tiny bits of debris.

226067 (3317) 38 Sept. 1975 On rotting trunk (10 in. diameter) bridging tributary stream. Caught during late morning drizzle. Stomach: full of mostly fruit mash and remains of a few scale insects (Coccoidea, Sternorrhyncha).

226068 (3347) 38 Oct. 1975 On decaying section of tree limb (10–12 in. diameter) lying across creek; caught in morning between 7 and 10:00. Stomach: full of semi-dry compact mass of purple and pale yellow fruit mash with pieces of skin.

225909 (3354) 61 Oct. 1975 On decaying branch straddling wet ravine. The branch is part of a rotting, tall canopy tree that fell and bridged the ravine connecting a ridge with the opposite hillside. Forest is open on either side of the huge trunk, limbs, and branches, and only scrub has regrown. Gingers and elephant ears cover slopes near the trunk, young Macaranga and hereau ( Pinanga ) palms are common along margins of the opening. Tall understory trees form a closed canopy at either side of the tree-fall opening; slopes are steep and muddy, partial leaf cover. The rat, Maxomys hellwaldiii , was trapped on the ground beneath the trunk.

TABLE 41 (Continued)

Locality, AMNH Elevation and (ASE) numbers (m) Date Trap site and other information

226069 (3583) 46 Nov. 1975 On trunk (6 in. diameter) lying across rocky tributary bounded by steep hillsides. Examples of the rat, Bunomys andrewsi , were caught nearby. Stomach: packed with remains of geometrid caterpillars (larvae of inch-worm moths) mixed with a little fruit mash.

226070 (3178) 55 Sept. 1975 On long, decaying trunk lying on ground down side of slope above tributary stream. The trunk is part of a tall, huge canopy tree that fell down the hillside from upslope. The bole still lays down the hill providing a path on which rats and squirrels can run downslope above the ground; surrounded by tall hillside forest. Examples of the rat, Paruromys dominator , were caught in on an adjacent limb of the same tree. Stomach: partially full of unidentifiable carmel-colored soft fruit mash and numerous chewed pieces of scale insects (Coccoidea, Sternorrhyncha, Margarodidae ).

226071 (3596) 107 Nov. 1975 On mossy, decaying trunk (1 ft diameter) that lay across the rocky stream bed of main stream where it was becoming steep. Good hillside forest on either side. Squirrel caught about 8:00 a.m., just before we got to trap. Stomach:

empty.

226072 (3348) 139 Oct. 1975 On smooth trunk (8 in. diameter) lying across rocky slope

226073 (3368) in open understory above Siuri and near base of large,

clifflike outcrop. Trunk lay about 3 ft above ground for most of its length, about 60 ft long. Caught ASE 3348 day before in late afternoon, ASE 3368 in early morning a few days later.

Stomach of ASE 3348: partially full of mostly mash from small figs.

226074 (3407) 182 Oct. 1975 On rotten section of trunk (2 ft diameter) lying horizontally on steep slope and extending across deeply cut rocky streambed to other slope; part of an old tree-fall that spread its trunk sections down the ravine. Good understory canopy. Stomach: partially full: gray, claylike substance, unidentifiable.

226075 (3296) 229 Sept. 1975 On rotten trunk (6 in. diameter) lying across damp ravine that is well shaded by rattan and understory trees on steep slope above tributary stream. Caught Prosciurillus alstoni at same spot on a different day. The rat, Maxomys hellwaldii , was taken on ground nearby.

Sungai Tolewonu 152 Jan. 1976 On decaying trunk lying across deep tributary, near another

226500 (3915) trunk where a rat, Paruromys dominator , was caught.

Stomach: partially full of pink fruit mash and a chunk of pinkish purple buprestid beetle larva (like those in P. alstoni ASE 3366).

226501 (3929) 152 Jan. 1976 On trunk of living Pometia pinnata (10 in. diameter) that had been uprooted and fallen across ravine; trunk is covered with moss and branches that have sprouted leaves all the way along trunk, and sits about 15 ft above water in bottom of ravine.

Short forest on either side; thick understory; very rocky.

Stomach: full of soft fruit remains.

TABLE 41 (Continued)

Locality, AMNH Elevation and (ASE) numbers (m) Date Trap site and other information

226502 (4258) 152 Feb. 1976 On rotting trunk of canopy tree (1.5–2 ft diameter, 50 ft long) bridging the main river about midway between first and second large tributaries. Trunk rests 4–7 ft above water level and is covered with thick moss on two sides and supports a few small shrubs, ferns, and palm seedlings scattered along its length—not enough that provides any decent cover. Along middle of trunk is a moss-free runway, about 5 in. wide that extends entire length (apparently a path frequently used by rodents to cross the river). On one side of the river is terrace forest, on the other side is steep, moss-covered rocky slope that gives way above to hill forest. Five Rubrisciurus rubriventer and a rat, Paruromys dominator , were taken in same spot. Stomach: full of tan fruit mash and chopped segments of a large adult beetle (mostly pieces of thorax and abdomen, two legs, no head).

(4174) 166 Feb. 1976 On large rotting Pometia trunk (1.5 ft diameter, 45 ft long) bridging main river, extending from flat river terrace on one side to rocky hillside on the other bank, and about 5 ft above water level. Trunk is free of vegetation; a limb from a live tree runs along the side of the trunk and its leafy branches provide partial cover over the top surface of the Pometia . The tree squirrels, Prosciurillus alstoni and Rubrisciurus rubriventer , and the rat, Maxomys hellwaldii , were taken on same trunk. Stomach: full of brown fruit mash and some worker termitid termites.

226503 (3909) 183 Jan. 1976 On large limb (10 in. diameter) of Pometia that had fallen from a

226504 (3952) canopy tree on bank; limb is wet and smooth, no moss. The stream is wide here with a vegetation-covered rocky island in middle; branch rests on rocks, one end in stream, the other ends on sections of old, rotten and moss-covered trunk that lay across from rocks to steep slope. ASE 3952 caught about 6:30 a.m. Stomach of ASE 3952: full of brown fruit mash mixed with one small beetle larva. Stomach of ASE 3909: full of reddish brown fruit mash, one large macrolepidopteran caterpillar, remains of a few scale insects (Coccoidea).

226505 (4138) 198 Feb. 1976 Caught before 8:00 a.m. on trunk lying across left fork of third long tributary. Trunk (8 in. diameter, 30 ft long) is dry and free of moss, extending from low hillside across the stream to a terrace; trunk rests a foot above water surface. Prosciurillus murinus is common in the understory forest bordering the tributary; the squirrels are heard calling but rarely seen. Stomach: partially filled with brown fruit mash and numerous remains of termitid termites.

226506 (4070) 213 Jan. 1976 On rotten section of wet, mossless trunk (40 ft long, 8–10 in. diameter) that lay in understory on slope above tributary stream; one end of trunk rests on side of streambed, the other on slope in dense understory. Several pieces of rotten wood are scattered here, part of an old tree-fall; ground is wet and crumbling; good forest; trunk is about 2 ft off ground for most of its length. Stomach: full of mostly tan and brown fruit mash, looks like the same kind of stuff found in stomachs of the rats Maxomys hellwaldii and Paruromys dominator ; also remains of a few small beetles.

TABLE 41 (Continued)

traps, as the tree squirrels R. rubriventer , P. topapuensis , P. alstoni , and the ground squirrels H. heinrichi and H. ileile (see table 41).

SYNONYMS: Information covering the three synonyms of P. murinus , and reasons behind their allocations are summarized below.

Sciurus murinus necopinus Miller and Hollister, 1921: 98 . HOLOTYPE: USNM 218712 (museum study skin and complete skull; measurements are listed in table 35), a very young adult male collected by H.C. Raven (original number 3107) on January 20, 1917. TYPE LOCALITY: Gunung Lehio, 01 ° 339S, 119 ° 539E, above 6000 ft (1829 m; locality 42 in gazetteer and map in figure 30), in the western portion of the central core of Sulawesi, Propinsi Sulawesi Tengah, Indonesia.

Based on nine specimens from Gunung Lehio, Kulawi, Pinedapa, and Rano Rano in the central core of Sulawesi, Miller and Hollister (1921: 98) diagnosed necopinus as ‘‘Like Sciurus murinus murinus Müller and Schlegel , from Menado, North Celebes, but skull with decidedly smaller auditory bullae.’’ After presenting measurements, they remarked (p. 99) that ‘‘This subspecies reaches its extreme form in the high mountains of the interior of Middle Celebes. Specimens from Pinedapa near the coast of the Gulf of Tomini, near Mapane, show an approach toward the typical race of North Celebes.’’

For Miller and Hollister, the name necopinus identified USNM samples from the central core of Sulawesi. Multivariate analyses of cranial and dental variables, which includes an estimate of bullar size, for the USNM specimens examined by Miller and Hollister and those collected by Musser along his transect from lowlands at the Sungai Oha Kecil to the highlands of Gunung Kanino and Gunung Nokilalaki (see tables 37 and 38), did not separate the central Sulawesian samples from those collected on the northern peninsula, the northeastern tip, or Pulau Lembeh (see map in fig. 30). In the graph of specimen scores projected onto the first and second principal components extracted from principal-components analysis illustrat- ed in figure 32, the only striking departure from a single constellation of scores is the score representing the animal from Gunung Lehio, the holotype of necopinus. Its position far to the left in the ordination (table 40) reflects its smaller skull because the squirrel is younger relative to all the other specimens measured. Although clothed in adult pelage, the cheek teeth are slightly worn and the basisphenoid-presphenoid and basisphenoid-basioccipital sutures are open.

Squirrels in some samples from the central core of Sulawesi tend to have darker underparts (a rich ochraceous gray) than samples from other geographic regions, but there is geographic overlap in this chromatic feature. The series from Teteamoet in the northeastern limb of the northern peninsula, for example, is inseparable from the richly pigmented specimens collected at Pinedapa in the central core of the island.

We have not uncovered unique chromatic or morphometric traits identifying samples from the central core of Sulawesi as representing a population significantly different from that ranging over the northern peninsula and on Pulau Lembeh.

Sciurus evidens Miller and Hollister, 1921: 99 . HOLOTYPE: USNM 217814 (museum study skin and complete skull; see table 35 for measurements), an adult female collected by H.C. Raven (original number 3176) on August 23, 1917. TYPE LOCALITY: Pulau Lembeh, 01 ° 269N, 125 ° 139E (elevation of collection site is not recorded but highest point of the island is 447 m; locality 5 in gazetteer and map in figure 30), of the coast of the northeastern end of the northern peninsula of Sulawesi, Propinsi Sulawesi Utara, Indonesia.

‘‘Like Sciurus murinus Müller and Schlegel , from the mainland of Celebes, but paler, decidedly more yellowish or rusty brown’’ is how Miller and Hollister (1921: 99) diagnosed evidens based on six specimens from Pulau Lembeh. They remarked that ‘‘The paler, more yellowish brown coloration of this form makes the animal conspicuously different from the dark, richly colored Sciurus murinus of the neighboring mainland of Celebes.’’

We examined all of Miller and Hollister’s specimens from Pulau Lembeh and the adjacent mainland and failed to see their diagnostic chromatic distinction between island and mainland samples. Covariation in cranial and dental dimensions also do not result in two separate clusters of specimen scores projected onto first and second principal components extracted from principal-components analysis (fig. 32, table 40). Scores representing specimens in the type series of evidens are scattered throughout the compact constellation of points for squirrels in our samples from the northern peninsula and central core of Sulawesi. We have no evidence to indicate that the squirrels collect- ed by Raven on Pulau Lembeh represent anything other than an insular population of Prosciurillus murinus , an assessment made earlier by Ellerman (1947: 259): ‘‘A specimen of Prosciurillus evidens Miller and Hollister , from Pulau Lembeh off North Celebes has been received at the Museum, and I consider it to be a synonym of P. murinus .’’

Sciurillus murinus griseus Sody, 1949: 77 . HOLOTYPE: RMNH 9827 (formerly MZB 5974; museum study skin, skull missing; skull and mandible were at Leiden in 1951 when G.H.H. Tate photographed them, but both elements cannot now be located in the collection; C. Smeenk, in litt., 2008), an adult female collected by J.J. Menden on October 24, 1939. TYPE LOCALITY: Bumbulan, 00 ° 299N, 122 ° 049E, coastal plain near sea level (locality 13 in gazetteer and map in figure 30), on the southern coast of the northern peninsula of Sulawesi, Propinsi Sulawesi Utara, Indonesia.

In addition to the holotype, Sody studied two other females collected by Menden from Bumbulan. Those three are part of a larger sample from Bumbulan obtained by Menden in 1939, ten of them currently stored in AMNH (see gazetteer). Sody (1949: 77) diagnosed griseus this way:

While in freshly collected murinus the underside is mousecoloured, suffused with tawny buff, in griseus the hairs on the underside, which have also mouse-colored bases, possess rather long grey tips slightly tinged with buff. In general appearance the ventral side is much lighter than in the typical race.

The auditory bullae are a trifle smaller than in typical murinus . Strange enough, for necopinus Miller & Hollister, the authors do not mention difference in colour from murinus of Menado.

In our samples of P. murinus from throughout Sulawesi, the underparts range from dark gray speckled with white or pale buff to dark ochraceous gray, with most specimens falling within the range between dark grayish buff and ochraceous gray. The specimens from Bumbulan, including the holotype of griseus, fall close to the dark grayish buff end of the spectrum and in this respect are inseparable from some samples collected at places in the northern, central, and southeastern regions of the island.

We could not measure the skull from the holotype of griseus, but did obtain values for cranial and dental dimensions from four intact adult skulls in Menden’s sample from Bumbulan housed at AMNH (the other six were so damaged by shotgun pellets that we could not obtain a full set of measurements). In the graph of specimen scores projected onto the first and second principal components extracted from principal-components analysis shown in figure 32, scores for the four specimens from Bumbulan (indicated by delicate arrows) are clumped with scores representing squirrels from the central core of Sulawesi, the northern peninsula, and the northeastern arm.

The sample from Bumbulan does not depart in any significant way from the extent of variation in color of underparts and dimensions of the skull and tooth rows expressed by other geographic samples of P. murinus .