Leptomys paulus, Musser & Helgen & Lunde, 2008

Musser, Guy G., Helgen, Kristofer M. & Lunde, Darrin P., 2008, Systematic Review of New Guinea Leptomys (Muridae, Murinae) with Descriptions of Two New Species, American Museum Novitates 3624 (1), pp. 1-60 : 40-49

Leptomys paulus , new species

HOLOTYPE AND TYPE LOCALITY: The holotype is AMNH 158202 , an adult female collected by Hobart M. Van Deusen (original number 12583) on July 8, 1953. The specimen consists of a stuffed museum study skin, cranium, and mandible. There is a sewn tear in the basal third of the tail; otherwise the skin is in good condition as is the skull (figs. 5, 7, 8); the teeth are intact, and although the molars are worn, the cusp patterns persist. External measurements are listed in table 1, cranial and dental measurements are included in the univariate summaries listed in tables 5 and 11.

The type locality is ‘‘Number 2 Camp’’ (fide Brass 1956: 130), labeled ‘‘Middle Camp’’ on skin tags, 1370 m, north slopes of Mount Dayman, 09 ° 499S, 149 ° 169E, in the Maneau Range (also spelled ‘‘Maneao’’), at eastern end of the Owen Stanley Mountains, Milne Bay Province, eastern Papua New Guinea (locality 23 in gazetteer and fig. 3). The actual Number 2 Camp was at 1550 m ‘‘on the crest of the leading spur ridge we had followed up from lower levels’’ ( Brass, 1956: 130). The trapping locality at 1370 m was in ‘‘the deep Atairo ravine, …a beautiful place, very mossy, and crowded with ferns,’’ about an hour’s walk from the camp at 1550 m ( Brass, 1956: 131). The Maneau Range is part of the Owen Stanley Range and, as Brass (1956: 111–112) noted, has four primary summit peaks (Maneau, Gadmarau, Mana-man, and Dayman), ‘‘grouped around the head of the Gwariu Valley’’ that are all parts of one large mountain. The entire Mount Dayman with its peaks, and the section of the Owen Stanley Range dominated by the mountain, was called the Maneau Range by early explorers, after Maneau Peak, which is only one and not the highest of the four peaks. Another perspective of the region is provided by Engilis and Cole (1997: 1) who worked there in 1985 and wrote that Mount Dayman is one of three isolated mountain peaks that dominate the extremity of the eastern peninsula of New Guinea. Viewed from the air, the summit is actually the apex of a complex series of dissected ridges. All but the summit alpine grasslands are covered in dense forest. Human settlements are in valleys below 1600 m; most are isolated by steep ridges.

REFERRED SPECIMENS: Along with the holotype, three other specimens of L. paulus were obtained by Van Deusen at Number 2 Camp, 1540 m: AMNH 158199 , the stuffed skin, skull, and partial postcranial skeleton of a young adult male collected on June 27, 1953 ; AMNH 158200 , the stuffed skin and skull of a juvenile male collected on June 29, 1953 ; and AMNH 158201 , the stuffed skin, skull, and partial postcranial skeleton of an adult female collected on July 5, 1953 .

An additional seven specimens were collect- ed during August of 1992 by Pavel German at 1300 m in the Agaun region of the Maneau Range at two places: Mount Waruman and Kamaya Village (see gazetteer). These are also discussed by Flannery (1995) in his account of ‘‘ Leptomys ernstmayeri ’’. Cole et al. (1997) reported other specimens collected in March, 1985, from the vicinity of Agaun, which are deposited at BBM-NG (see gazetteer) as ‘‘ L. ernstmayri ’’; these too represent L. paulus . The Australian National Wildlife Collection holds a single specimen of L. paulus , represented by a skull and postcranial skeleton ( CSIRO 29433, adult female), obtained at Dumae Creek, Agaun (see gazetteer) in January 1986.

The northernmost record of L. paulus derives from Kagi, 1500 m, in the Astrolabe Range (locality 22 in gazetteer and fig. 3), where two individuals were collected by G. H. H. Tate in 1937 ( AMNH 108451 and 108452).

ETYMOLOGY: The Latin, paulus means ‘‘little’’, and is used here to highlight the small body size of the species occurring in the Maneau Range and elsewhere in the eastern Owen Stanley highlands relative to all the other species of Leptomys , which are physically larger.

DIAGNOSIS: Leptomys paulus is smaller in body size than L. elegans , L. signatus , L. arfakensis , and L. ernstmayri (tables 1–5). It lacks the short buffy gray dorsal fur and white blaze on the head that is diagnostic for L. signatus and in pelage coloration more closely resembles L. elegans , L. arfakensis , n. sp., and L. ernstmayri (figs. 4–6). Of these, L. paulus is ecologically and morphologically most like L. ernstmayri , but differs by its smaller body size, relatively longer tail (compared with samples from the Adelbert Range and Central Cordillera), rostral coloration at base of mystacial vibrissae (unpigmented to pale in L. paulus , dark brown in L. ernstmayri ), longer dorsal coat that is soft and silky to the touch (shorter and velvety in L. ernstmayri ), the part in the fur between ears and shoulders (inconspicuous because of the longer fur in L. paulus , a prominent feature in all the other species), missing or tiny hypothenar pad on the plantar surface (present and usually large in most examples of L. ernstmayri ), cusp t7 and anterolabial cingulum absent from each second upper molar (table 6, fig. 11), and cranial and dental proportions (figs. 16, 20, 21).

GEOGRAPHIC DISTRIBUTION: Leptomys paulus is documented in museum collections primarily by specimens from the Maneau Range, including the type series from the northern slopes of Mount Dayman between 1370 and 1540 m, and from several localities in the vicinity of the village of Agaun between 1240 and 1525 m (see fig. 3 and gazetteer) .

Two specimens from Kagi, at 1500 m in the Astrolabe Range, along the Kokoda Trail (locality 22 in gazetteer and fig. 3) demonstrate that the geographic range of L. paulus extends northeast of the Maneau Range into the main body of the Owen Stanley Range. These specimens are unambiguously referred to L. paulus rather than to Central Cordilleran L. ernstmayri (which is also recorded farther to the west in the Owen Stanley Range) on the combined evidence of their small body size, lack of cusp t7 and an anterolabial cingulum in the second upper molar (table 6), proportionally long tail (115%–121% of head and body length), relative length of white tail tip (19%–29% of tail length; table 3), and their morphometric associations in principal components analysis (fig. 16).

The mean altitude of nine vouchered collecting sites for L. paulus (see gazetteer) is 1408 m (median 5 1400 m, SD 5 112 m).

DESCRIPTION AND COMPARISONS: Leptomys paulus is a small-bodied, long-tailed montane representative of Leptomys (fig. 19). It is generally similar in external proportions as well as pelage texture and coloration to all the other species of Leptomys except L. signatus . Adult dorsal pelage is thick (up to 12 mm long) and very soft, silky to the touch (dense but slightly shorter—not more than 10 mm long—which renders coat a velvety rather than silky texture, in the other species; L. signatus has an even shorter coat, 5–7 mm). As a result of its longer fur, L. paulus is the only species of the genus in which the balding pattern on the back of the head, neck, or shoulders is not evident in study skins (however, a balding pattern similar to the other species can be found by parting the fur over the anterior dorsum manually to reveal the underfur). The upperparts are dark tawny brown over the back and rump and buffy along sides of body, thighs, and upper arms; overhairs are mostly gray with brown to dark buffy tips (dorsal coloration is similar to all other species except L. signatus , which has pale buffy gray upperparts without the rich buff and dark brown tones of the other species). The head is the same color as the back (all other species are similar except L. signatus , which has a conspicuous and large white blaze on the head; figs. 4–6). The large, dark brown pinnae appear naked but are scantily covered with minute dark hairs. The mystacial and superciliary vibrissae, 55– 60 mm at longest, extend well beyond the ears when laid against the head. A dark brown mask surrounds the eyes but does not extend onto the muzzle. The muzzle at the base of vibrissae appears unpigmented because only short scattered pale hairs cover it (the muzzle is brown or grayish brown in all the other species due to the dense covering of pigmented hairs), and the cheeks are white. Coloration of the ventral coat ranges from that seen in the holotype of L. paulus (as well as in AMNH 158200 and BBM-NG 109478), which is grayish white over most of the venter, except for a pure white area from chin to chest and a small white patch in the inguinal region (fig. 5), to whitish gray with white chest and throat in most specimens (in the pigmented areas, the hairs are gray basally with either short unpigmented tips—producing the grayish white affect—or longer white tips, responsible for the paler gray wash), to pure white throughout, except for grayish white fur on the undersides of the hind limbs (BBM-NG 109514 and 109522, AMNH 108451 and 108452). The range in ventral coloration resembles that characteristic for many specimens of L. ernstmayri with the exception that some individuals of that species exhibit a buffy wash over the grayish white underparts. In the sample of L. ernstmayri from Mount Rawlinson, for example, the underparts range from buffy gray through grayish white to whitish gray with the same white pattern on chin, chest, and inguinal area as exhibited by L. paulus . Specimens of L. ernstmayri from the Okapa region and the Kratke Range demonstrate the range from pure white venters to whitish gray underparts that are broken by the same white pattern seen in L. paulus .

The tail is longer than the head and body in L. paulus (table 3), and covered in annuli of very small flat and inconspicuous scales (18– 20 per cm). The scale hairs are fine and laid flat against the tail so that it appears naked. About one-fourth (19%–29%, averaging 26%; table 3) of the distal portion of tail is all white, the dorsal proximal region is brownish gray, and the ventral surface below this brownish gray segment ranges from white to mottled. General coloration of the tail and its patterning are similar to examples of other Leptomys .

Dorsal surfaces of the front and hind feet, including the digits and claws, are all white and densely covered with short silvery hairs, which form short and sparse tufts at the base of each claw. The area proximal to the ankle and wrist is brown. The second digit of the front foot is half the length of the third, which is the longest; the fourth is slightly shorter, and the fifth is the shortest, reaching to only base of adjacent digit. The hind foot is very long and narrow in L. paulus and all the other species except in L. signatus where the hind foot is longer and wider relative to body size (this contrast between that species and the others is striking). In L. paulus and the other species, the first digit is very short, its claw barely or not quite reaching the base of the adjacent digit; the three middle digits are longest and subequal in length; the fifth digit is short but longer than the first, its claw reaching to proximal third of the second phalanx. The configurations of manal and pedal digits are similar in all species. A naked plantar surface, which is partially brown and adorned with conspicuous interdigital pads is also common to all species of Leptomys , but presence or absence and size of the hypothenar is variable among samples. In all species, the thenar is elongate and conspicuous. In 14 specimens of L. paulus (at AM, AMNH, and BBM), the hypothenar was absent from eleven examples, present but small in two, and of normal size in one. In specimens of L. signatus examined for this trait (at AMNH only), three lacked a hypothenar, and one had a tiny remnant. Six of nine L. elegans from the Papuan Peninsula (at AMNH and MSCN) lacked the hypothenar, but a small series of fluid-preserved specimens from Mt. Sisa (at AM) all had a hypothenar. Both specimens of L. arfakensis appear to have a hypothenar. For specimens of L. ernstmayri that we scored, 11 of 13 from the Huon Peninsula (at AMNH) had a hypothenar, both Adelbert specimens in alcohol that we studied (at BBM) had a hypothenar, but all seven specimens from the Central Cordillera (at AM, AMNH, and BBM) lacked a hypothenar.

Qualitative shape of the skull and mandible of L. paulus , along with most dental traits, resemble those of the other species of Leptomys (figs. 7–10); characteristics of skulls and dentitions were described in the section detailing generic characteristics. The primary qualitative difference we observed was in presence or absence of cusp t7, and the anterolabial cingulum, on the second upper molar. All mature L. paulus lack a cusp t7; it is present in incipient form on the upper left molar in one juvenile (BBM-NG 109478). In contrast, that cusp is present in all specimens of L. ernstmayri from the Huon Peninsula, in most of those constituting the samples from the Central Cordillera, and in two of the 10 examples from the Adelbert Range (table 6). An anterolabial cingulum is not present in any example of L. paulus (and is also typically absent in L. elegans , L. signatus , and L. arfakensis , n. sp.). Specimens of L. ernstmayri from the Central Cordillera usually lack an anterolabial cingulum (14 of 18), but all those from the Huon Peninsula exhibit a conspicuous cingulum, as do seven of the 10 specimens from the Adelbert Range (table 6).

The disparity in body size between L. paulus and specimens of L. ernstmayri from the Huon Peninsula, Adelbert Range, and Central Cordillera is reflected by the much shorter head and body length in L. paulus , its shorter hind foot, and lesser mass (table 3). Length of tail is an exception in this comparative context. While the tail is, on average, absolutely shorter than in the sample of L. ernstmayri from the Huon Peninsula, it is slightly longer than the series from the Adelbert Range and only slightly shorter than the Cordilleran sample. The proportional result is a very long tail relative to body size, which, on average, is markedly greater than the proportion in samples from the Adelbert Range and Central Cordillera but closely similar to body and tail proportions in examples of the Huon sample (table 3).

The dissimilarity in cranial and mandibular dimensions between the two species is illustrated by the larger skull of an L. ernstmayri from the Huon Peninsula contrasted with the appreciably smaller holotype of L. paulus from Mount Dayman that are aligned in figures 7 and 8. Measurements of cranial and dental variables provide quantifiable univariate and multivariate (principal and discriminant-function analyses) portrayals of the size distinction. In absolute size, L. paulus matches L. ernstmayri in a few variables (lengths of nasals and incisive foramina, height of braincase) but is significantly smaller in most other metrics (table 5).

Differences in cranial and dental dimensions among all population samples of L. paulus and L. ernstmayri are summarized in the scatter plot of specimen scores projected onto the first and second principal components, where all specimens referred to L. paulus cluster apart from geographic samples referred to L. ernstmayri (table 8, fig. 16). Covariation in size of most cranial and dental variables, as indicated by the high values is responsible for the spread of scores along the first axis. This pattern, scores for L. paulus at one end of the plot and those for the geographic samples of L. ernstmayri clustered to the left, reflects the contrast in size that is apparent when skulls of the two species are compared side by side and in the table of measurements— L. paulus is a smaller animal than L. ernstmayri in nearly all variables we examined.

Patterns of covariation along the second canonical axis extracted from discriminantfunction analysis clearly isolate specimen scores representing L. paulus from the Huon, Adelbert, and Cordilleran samples of L. ernstmayri (fig. 20). Size primarily affects the spread of scores, as indicated by the positive and strong correlations of most variables with size (table 10); the analysis again quantitatively emphasizes the magnitude of the difference in most variables between the smaller L. paulus and larger L. ernstmayri .

Proportional differences between the two species can be extracted from the discriminant-function analysis (the relatively longer nasals in L. paulus , for example), but are better visualized in the ratio diagram that contrasts the series of L. ernstmayri from the Huon Peninsula (the provenance of the holotype) with the AMNH sample of L. paulus , which includes the holotype (fig. 21; see also table 11). Compared with the Huon sample, L. paulus has significantly longer nasals relative to occipitonasal length as well as any of the other measured dimensions (length of nasals is a reasonable estimate of rostral length, which we did not measure directly), a wider braincase relative to its height, longer incisive foramina relative to both their width and length of diastema, and a longer bulla relative to occipitonasal length. Leptomys paulus also has a significantly lesser zygomatic breadth relative to occipitonasal length or breadth of braincase; narrower rostrum relative to length of nasals or occipitonasal length; a narrower zygomatic plate relative to most of the other dimensions; a shorter diastema, bony palate, and postpalatal region relative to occipitonasal length; and a longer molar row relative to length of bony palate. These are the striking contrasts; other proportional differences and similarities are also evident in the ratio diagram.

HABITAT AND BIOLOGY: Leptomys paulus inhabits forests; published descriptions of collection sites correspond to definitions of tropical lower montane rainforest ( Whitmore, 1984).

Two of the AMNH examples of L. paulus were collected at ‘‘ Number 2 Camp’ ’ where members of the Fourth Archbold Expedition to New Guinea (1953) lived and worked from June 20 to July 12, 1953. Most of the mammals obtained at this elevation were trapped or shot within easy range of camp because of the unfavorable climatic conditions, so aptly described by Brass (1956: 130– 132) :

But for three of the five days following the establishment of our collecting base here and three rainless days in July, the weather can be summarized as generally dull, misty, rainy, and downright disagreeable. The climate evidently was much wetter at this altitude than at our upper camp on the mountain [at 2230 m], at least at this time of year. Many tall trees were felled to let light and wind into the camp, but for days together only rare brief bursts of sunshine would break through the clinging murk of cloud and mist and enter the clearing. On better days, by 10 A.M. often and by noon nearly always, the usual uphill air movement brought mist that turned into drizzle or was accompanied by showers. Substantial amounts of rain fell on nine days and during five nights, but as a rule the sky cleared soon after nightfall. Some of the rain was from general disturbances, for heavy falls were expe- rienced on the lowlands at times during this period.…It will be noted that owing no doubt to effects on solar radiation of the all too constant diurnal blanket of cloud and mist, maximum temperatures were slightly lower here than at the Top Camp, 680 meters higher on the mountain.…Traffic on the trails soon wore through a surface layer of 5 to 10 cm. of peaty brown humus and exposed slippery yellow clay. In camp, we corduroyed the most trampled ground with sticks, between which the mire squelched and spurted as we walked in the mist. It became necessary to burn lamps in the tents all day and much of the night to dry specimens and clothing, and to have extra kerosene carried up the mountain for this purpose. Until we did this, mammal specimens pinned to drying boards remained limp, and mosses flourished on drying trays hung under the peak of my tent to take advantage of the almost non-existent heat of the sun.

Leeches, standing on end on the ground humus and waving back and

TABLE 10 Results of Discriminant-Function Analysis Derived from 39 Adult Leptomys with Smaller Body Size, Leptomys ernstmayri and Leptomys paulus (Correlations, eigenvalues, and proportions of variance are explained for two canonical roots; see table 5, fig. 20.)

forth in readiness to attach themselves to a blood supply, were active, particularly in hollows and gullies.

Mean maximum and minimum temperatures were 20.7 ° C (69 ° F) and 14.1 ° C (59 ° F), respectively (‘‘thermometer-house temperatures in the camp clearing,’’ Brass, 1956: 130–131).

Brass (1956: 131) provided this description of the topography and vegetation near camp:

Except to the west where the slopes were much cut by deep ravines, the terrain about camp held some steep little gullies but for the most part consisted of fairly easy ridges. Many large trees over 30 meters tall occurred in a generally well-spaced forest dominated by a Castanopsis and two or three species of oaks. Engelhardtia , a frequent associate of these trees in New Guinea, was present, with representatives of Gordonia, Sloanea , and other genera. A big, red-flowered Metrosideros made occasional blazes of color in ravines. Many dead trees, standing, or fallen and covered with mosses and ferns, gave parts of the forest an over-age appearance. Especially on ridge crests, the bases of the trees were mounded with a network of surface roots and covered deep with bryophytes. An abundance of leaf and twig litter lay on the ground over a carpet of mosses and hepatics. Tree trunks and branches, and a predominantly woody undergrowth were lightly to heavily mossed. Especially abundant on ridge crests and climbing to 20 meters or more, the scrambling

Table 11 Descriptive Statistics for Cranial and Dental Measurements (mm) used to derive the Ratio Diagram in Figure 21 for Adults in AMNH Samples of Leptomys ernstmayri and Leptomys paulus a (Mean plus or minus one SD and observed range in parentheses are listed.)

bamboo of this altitude had larger leaves than that of the Top Camp.…

These mid-mountain forests of the ridges differed almost entirely in floristics from the mossy forests and the beech forests of our Top Camp [2230 m], only a few ferns and other minor elements occurring in common. But in gullies and ravines, where mixed rain forest replaced the mid-mountain forest, a lush undergrowth contained rather numerous ferns, including Leptopteris, Vandenboschia, Nephrolepis , and large Asplenium species , and fleshy-leaved shrubs and large herbs, such as Medinilla, Saurauia , and Begonia , already familiar in similar habitat niches in the beech and mossy forests. The altitude was still too high or the climate too misty for palms to figure prominently, even in moist gullies in which one species ascended as high as Top Camp, and only a Calyptrocalyx , a small Linospadix , and one other were present in addition to one or two climbing Calamus species. Pandanus , likewise, was not prominent.

Brass (1956: 131) lacked direct observations about the floristics in the deep Atairo ravine, about a one-hour walk from camp where the holotype of L. paulus was captured, because an injured shoulder prevented him from visiting the site and his description comes from his colleague’s observations:

The deep Atairo ravine.…was a beautiful place, very mossy, and crowded with ferns. My boys collected many rain-forest elements there which we did not find elsewhere on the mountain, one conspicuous plant being an Impatiens with large white flowers, another the fig-like Dammaropsis kingiana , the great leaves of which served well as bath mats in our open air ‘shower room.’ Of special interest as high mountain herbs descending low in this streamway were a buttercup ( Ranunculus ), Nertera granadensis , and Gunnera macrophylla . Ranunculus , an alpine genus in New Guinea, was not met with elsewhere on Mt. Dayman.

Members of the Fourth Archbold Expedition did not find L. paulus at lower camps on Mount Dayman that were situated in tropical lowland evergreen rain forest. At their 700 m ‘‘Number 3 Camp’’ at the upper limit of evergreen rain forest, the larger-bodied L. elegans was encountered (see that species account).

According to their associated field notes, specimens of L. paulus from Mount Waruman at Kamaya Village and near Munimun Village (deposited in AM) were collected in ‘‘primary forest’’ on the forest floor in snap-traps baited with beetle larvae, or dug out from a nest in the ground near a stream (apparently five individuals had been using the same nest; see Flannery, 1995).

Of the seven examples of Leptomys paulus obtained on the Garatin Ridge reported by Cole et al. (1997:12), one was live-trapped at the base of a tree in primary forest. The others were purchased with their nests from Daga hunters. All of the nests were reportedly dug up from burrow systems that contained a central nest chamber about 1 m beneath ground surface. One nest, from primary forest habitat, contained three individuals (adult female and two juvenile males): the remaining three nests, from secondary forest areas, contained one adult each. Our hunters told us that some animals escaped capture during nest excavation.

Two of the female L. paulus collected by Cole et al. (1997) in March 1997 were pregnant, one (BBM-NG 109273) with two embryos, one in each horn of the uterus, another with one embryo in the right horn (BBM-NG 109514). Engilis and Cole (1997) provided a brief description of the forest on the Garatin Ridge where the examples of Leptomys reported by Cole et al. (1997) were taken, and some climatological data.

Smales (2006) described a new helminth, Labiobulura (Archeobulura) leptomyidis , taken from the caeca of AM and BBM-NG specimens of Leptomys paulus from Munimun Village and Iana River, respectively (see gazetteer).

No other kinds of biological information are yet available for L. paulus .

Figures

Fig. 21. Ratio diagram illustrating some proportional relationships in cranial and dental dimensions between samples of adult Leptomys. The standard is the AMNH sample of L. ernstmayri from the Huon Peninsula (N 5 11) contrasted with the AMNH sample of L. paulus from Mount Dayman (N 53). How the diagram was constructed and how to read it is explained in Materials and Procedures.