Leptomys elegans Thomas, 1897

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 : 24-31

Leptomys elegans Thomas, 1897

HOLOTYPE AND TYPE LOCALITY: The holotype of L. elegans is MSCN 32988, an intact skull and study skin of an adult female, collected by Dr. Lamberto Loria between 1890 and 1893 in the Astrolabe Range behind Port Moresby (Central Province, Papua New Guinea); measurements are listed in table 1. The tag of the holotype bears no definite locality other than ‘‘British N.G. ’’, and no exact provenance was published in the original description of L. elegans , although Thomas (1897: 607) wrote that Loria’s localities ‘‘are mostly between the Owen Stanley Range and the sea, in or near the watershed of the Kemp Welch river’’, an area that would include the Astrolabe Range and adjacent Sogeri Plateau. Tate (1951: 223) observed that ‘‘Loria collect- ed in the Astrolabe Range behind Port Moresby’’, and we regard specimens from the Astrolabe Range (available at AMNH and ZMB) as topotypical.

DIAGNOSIS AND CONTRASTS: Of the known species in the genus, L. elegans is the largest in body size (as indexed by length of head and body along with mass), but has the shortest tail relative to body length (tables 2, 3). On average, the tail is shorter than head and body in L. elegans , about coequal in L. signatus and L. arfakensis , n. sp., and appreciably longer in L. ernstmayri and L. paulus , n. sp. The coat covering upperparts of head and body is moderately long, soft, dense, velvety in texture, overall tawny brown, and usually without decoration (some specimens in AM, AMNH, and PNGNM series have a discernable but small, pale marking on the head between the eyes). This coloration and texture is shared with all the other species of Leptomys except L. signatus , which has shorter, buffy gray dorsal pelage marked by an expansive white blaze on the head (fig. 4), and L. paulus , n. sp., which has longer and softer pelage that is silky to the touch. Pure white underparts are characteristic of L. elegans (also L. signatus and L. arfakensis , n. sp., figs. 4, 6); the fur covering the venter in most examples of the smaller-bodied L. ernstmayri and L. paulus , n. sp., is white suffused with varying intensities of gray or buff (fig. 5).

Most cranial and dental dimensions characterizing L. elegans are, on average, appreciably greater than those associated with L. ernstmayri and L. paulus , n. sp., the species of smaller body size (see skulls compared in figs. 7, 8), and more equitable with those of L. signatus and L. arfakensis , n. sp. (tables 4, 5). But significant proportional contrasts shown by L. signatus (relatively short nasals, narrow zygomatic plate, and larger molars), along with an absolutely narrower rostrum, highlight its morphometric distinction from L. elegans , and the much smaller molars in L. arfakensis , n. sp., help mark its clear separation from L. elegans (tables 4, 7; figs. 13, 14).

Most specimens in our sample of L. elegans have a small cusp t7 on the second upper molar but lack an anterolabial cingulum (table 6). That latter structure is absent in all but one of the 17 specimens we surveyed, and in that individual (AM M32103) it is very small. Occurrence of cusp t 7 in L. ernstmayri and L. signatus is comparable to its frequency in L. elegans , but the cusp is not present in all specimens examined representing the two new species, L. paulus and L. arfakensis . The low frequency of occurrence of the anterolabial cingulum in L. elegans is comparable to that seen in samples of all the other species except L. ernstmayri (found in only one of 38 examples of L. elegans , L. signatus , L. arfakensis , n. sp., and L. paulus , n. sp.; occurs in about half of the total samples for L. ernstmayri ).

GEOGRAPHIC DISTRIBUTION: Leptomys elegans is recorded from localities scattered throughout the Papuan Peninsula of southeastern New Guinea (along the slopes of the Astrolabe, Wharton, and Maneau ranges, as well as the outlying peak of Mount Victory) and along the southern slopes of the Central Cordillera in the Kikori River Basin (see gazetteer and map in fig. 2). The altitudinal range of the species, as documented by voucher specimens, extends from lowlands near sea level to uplands at 1500 m. The average altitude of 16 trapping sites, from which specimens of L. elegans were preserved, is 840 m (median 5 775 m, SD 5 426 m).

GEOGRAPHIC VARIATION: Leptomys elegans is recorded by specimens from two broad regions in New Guinea. One is the Papuan Peninsula, basically southeastern New Guinea. The other is the Kikori River Basin on the southern slopes of the Central Cordillera in east-central New Guinea (fig. 2). Some consistent morphological differences are apparent between these two geographically disjunct populations. In the discriminantfunction analysis depicted in figures 13 and 14, for example, the dispersion of specimen scores along the first canonical variate indicates narrow separation between samples from the Papuan Peninsula and Kikori River Basin that is influenced primarily by the Papuan specimens’possession of smaller molars but

TABLE 2 Descriptive Statistics for External Measurements (mm), Mass (grams), and Proportions of Adult Leptomys with Larger Body Size: Leptomys elegans , Leptomys signatus , and Leptomys arfakensis (Mean plus or minus one SD, observed range in parentheses, and size of sample are provided.)

broader incisive foramina relative to specimens from the Kikori River Basin (see table 7), a difference also reflected in means of measurements listed in table 4. Scatter of specimen scores along the second canonical axis forms two barely overlapping groups, affected mostly by the relatively longer nasals in the Kikori sample, its narrower zygomatic plate, and longer molar rows but narrower first upper molars. If the morphological traits expressed in our samples reflect variation present in the populations from which they were derived then the population occurring on the Papuan Peninsula is distinguished from that in the Kikori River Basin by its relatively smaller molars, broader incisive foramina and zygomatic plate, and shorter nasals (indicating a shorter rostrum). We currently regard these distinctions to reflect geographic variation between conspecific populations. Should they become available, series of L. elegans from intermediate localities (such as from Gulf Province) might demonstrate these contrasts to be clinal.

The single specimen collected at 800 m on Mount Victory (on the Papuan Peninsula; locality 9 in fig. 2) differs from all other examples of L. elegans in having longer and narrower incisive foramina and a broader braincase. Mount Victory is an isolated peak entirely separated from the montane forests of the Central Cordillera by lowland tropical rainforest, the biota of which remains poorly surveyed ( Diamond, 1985). Series from this outlying mountain block will be needed to assess patterns of morphological covariation in Leptomys from Mount Victory compared to samples of L. elegans collected along the margin of the Central Cordillera. However, because L. elegans is known from several lowland localities (see gazetteer), it may range throughout the lowland rainforest corridor between the two lower montane regions.

HABITAT AND BIOLOGY: Information gleaned from collection sites indicates that L. elegans inhabits tropical lowland and lower montane evergreen rain forests, including both mature and secondary forested habitats. All collection localities for this species lie on mountain slopes, and all of the specimens were collected between 20 m and 1500 m, which would seem to be the core altitudinal distribution of L. elegans .

The lowest record is BBM-NG 60171 , collected at ‘‘ Karema , Brown River Forest Station’ ’ (locality 8 in gazetteer and fig. 2). The tag of this specimen bears the annotation ‘‘lowland evergreen rainforest’’. Although the altitude of the trapping locality is unrecorded, Zweifel (1972) gave the altitude in the vicinity of Karema as ca. 20 m. Additional records in the vicinity of the Brown River include

TABLE 3 Descriptive Statistics for External Measurements (mm), Mass (grams), and Proportions of Adults in Samples of Leptomys with Smaller Body Size: Leptomys ernstmayri and Leptomys paulus . (Mean plus or minus one SD, observed range in parentheses, and size of sample are provided.)

specimens taken on the Sogeri Plateau (410– 450 m) and in the Astrolabe Range (520 m), localities associated with lowland evergreen rainforest (K. P. Aplin, personal commun.). Other specimens collected nearby came from the Wharton Range at Mafulu (1253 m), situated in lower montane forest ( Archbold and Rand, 1935).

The highest records, 1400–1500 m, come from Mount Sisa, situated along the southern slopes of the Kikori River Basin (locality 5 in fig. 2 and gazetteer). Dwyer (1984) recorded the distribution of New Guinea rodents according to successional phases in the transition from gardens to forest over 20 years on the southern slope of Mount Sisa. Examples of L. elegans were taken in second-growth formations estimated at 10 and 12.5 years regrowth from garden environments, 25-yearold secondary forest, and primary forest, but not in any of the environments representing the transition from gardens and grassy stages through to early regrowth and very low secondary forest, an estimated period of five to seven years. Flannery (1990: 177; 1995: 241) portrayed a living individual that was trapped on Mount Sisa.

Leary and Seri (1997) reported additional specimens of L. elegans from the Kikori River Basin. Within that expansive area, the species has now been recorded at several sites, including those on Mount Sisa, situated between 450 and 1500 m (see gazetteer). Habitats at these various places correspond to lowland evergreen rainforest, lower montane forest, and the transition between the two (K. P. Aplin, personal commun.). Leptomys signatus is also represented by specimens from the Kikori River Basin, on the north-northwestern slopes of Mount Bosavi and on the Darai Plateau (localities 4 and 5, respectively, on the map in fig. 2); whether the distributions of these two species closely abut or overlap within the region remains unknown.

The single specimen trapped in the Maneau Range in far southeastern New Guinea (AMNH 158204) was caught at 700 m (‘‘Number 3 Camp’’ or ‘‘Bottom Camp’’) on the north slopes of Mount Dayman by H. M. Van Deusen, a member of the Fourth Archbold Expedition to New Guinea. This camp, as described by Brass (1956: 133),

was situated close below the junction of the mixed rain forests of the lower slopes and the oak and Castanopsis forest of the middle slopes. As evidenced by secondgrowth rain forest beginning immediately below the camp, we were at the exact upper limit of former cultivation on this part of the mountain by the Daga people who now live in the Biniguni group of

TABLE 4 Descriptive Statistics for Cranial and Dental Measurements (mm) of Adults in Samples of Leptomys with larger body size: Leptomys elegans , Leptomys signatus , and Leptomys arfakensis (Mean plus or minus one SD, observed range in parentheses, and size of sample are listed.)

villages on the lowlands. …With great regularity in late morning or early afternoon a pall of heavy dark cloud enveloped the slopes above us, its lower edge often resting in the treetops of the oak- Castanopsis forest only 50 meters higher on the slopes, but we were below the zone of regular mists and seldom had any in camp. A brown haze obscured the lowlands and the coast nearly always, and at no time were the mountains of the Cape

Nelson Peninsula visible, though distant less than 40 miles to the north.

Between July 19 and July 26, mean maximum temperature was 25.1 ° C (77 ° F), mean minimum was 17.2 ° C (63 ° F). Brass provided additional description of the nearby forest, which indicates that Number 3 Camp was situated at the upper limits of what Whitmore (1984) designates as tropical lowland evergreen rain forest and just below tropical lower

TABLE 5 Descriptive Statistics for Cranial and Dental Measurements (mm) of Adults in Samples of Leptomys with Smaller Body Size: Leptomys ernstmayri and Leptomys paulus . (Mean plus or minus one SD, observed range in parentheses, and size of sample are listed.)

montane rain forest (Brass’s ‘‘mid-montane’’ forest ‘‘at once distinguished from the more mixed rain forest by its generally more open character under the canopy and the distinctive appearance of its dominant oaks and Castanopsis ’’; Brass, 1956: 134). Some interdigitation of the two formations characterized the forest near the camp ( Brass, 1956: 134):

Rain-forest canopy trees entered the mid-mountain forest and extended sporadically far up the mountainside. Scattered large trees of one of the oaks, with small acorns, grew in the upper parts of the rain forest at camp level. Many minor floristic elements such as woody and herbaceous undergrowth species, epiphytic orchids and ferns, and some large lianas occurred in both types of forest, but others did not. The lower limits of the zone of regular cloud formation on the mountainside marked, here as apparently everywhere in New Guinea where such cloud conditions prevail, the contact line of two readily distinguishable forest types .

TABLE 6 Frequency of Cusp t7 and a Discrete Recognizable Anterolabial Cingulum on the Second Upper Molar in Samples of Leptomys (See fig. 13. Specimens are stored in AM, AMNH, BBM, BMNH, MCZ, MSCN, and USNM.)

The smaller-bodied Leptomys paulus , n. sp., was encountered at Number 2 Camp, higher on Mount Dayman in tropical lower montane rain forest (see account of L. paulus , below).

The specimen of L. elegans from Mount Dayman was the only Leptomys that Van Deusen, its collector, had ever encountered (partial manuscript in AMNH Mammalogy archives). Of his experience, he wrote:

While jacklighting at 8:00 on the evening of July 17, I observed and then shot a large male (158204) as he was hopping on the forest floor. This individual may well have been frightened to such action by my presence. I cannot say with certainty that hopping is the normal rapid gait of Leptomys .

Because the tail slightly exceeds length of head and body, Van Deusen suspected ‘‘hopping may often be the preferred gait, especially to escape predators’’.

According to its associated field data, the single available specimen from Mount Victory , an outlying peak north of the Central Cordillera in far southeastern New Guinea ( Diamond, 1985), was taken in an Elliott trap set ‘‘on mossy ground in secondary forest’’ at 800 m. This elevation probably marks the approximate zone of transition between lowland and lower montane evergreen rain forest on this small isolated mountain block .

Apart from habitat associations, only meager biological data is available for L. elegans . The only reproductive information is from two sources. Dwyer (1984) reported a mean litter size of 1.75, based on four pregnant females taken on Mount Sisa. Leary and Seri (1997: 89) noted a single young born to another pregnant female caught during March, 1995, in the Kikori River Basin.

During Dwyer’s study of rodents and plant succession on the southern slopes of Mount Sisa in 1979 and 1980 (see Dwyer, 1984, and references cited there), he described behavior and diet of a juvenile male (extracted from his field notes on file at CSIRO). The animal slept in a ‘‘sitting’’ position, resting upright on the full extent of the hind feet with its back curled and the head inserted down between the front legs; his sketch suggested that the tail was extended rather than curled in (this sleeping position of the head and body may possibly explain the balding pattern of the fur between ears and shoulders). However, he also noted that when at rest, the tail was curled forward along the right side of the body all the way around the head to ear level on the left side. The animal walked about ‘‘elegantly’’, on its digits with most of the plantar portion off the ground, and the tail arched with the tip pointing downward.

Available evidence gleaned from published and unpublished sources indicates Leptomys elegans to be carnivorous. Leary and Seri (1997) captured two specimens from the Kikori River Basin in traps baited with ‘‘sago grubs’’ and a specimen from Mount Dayman (AMNH 158203) was taken in a live trap baited with meat. Dwyer’s observations recorded in his field notes are definitive. During one afternoon and night his captive male ‘‘had eaten 20 or so small grasshoppers’’ and one preying mantis (3 inches long), pouncing on each insect and attacking at the ‘‘head end first’’; it also gnawed cooked flesh off a cuscus (phalanger) vertebrae. Over the following day and night the rat ate an entire carcass of a juvenile male Rattus niobe (weighing 26 g; the Leptomys weighed 39 g when caught and 2 g less when released a few days later), ‘‘leaving skin, feet and tail but chewing up smaller bones’’. Dwyer’s captive also devoured the raw flesh off skulls. It ignored pawpaw (a fruit). Another of Dwyer’s specimens, a female, had ‘‘insect stuff and other [unidentified matter]’’ in her stomach. Although Leptomys has been caught in traps baited with coconut— two specimens from the Kikori River Basin ( Leary and Seri, 1997: 89) and an example from Mount Victory—we suspect the rats in these instances were attracted to invertebrates that had gathered on and around the coconut bait.

Smales (2006) lists helminth parasites discovered in the digestive tract of ‘‘ L. elegans ’’ (taken from samples of Leptomys at AM and BBM), but Smales’s results are impossible to interpret as presented, because her host sample was composite, including two specimens of L. paulus , n. sp., from the Maneau Range, two individuals of L. ernstmayri from the Adelbert Range, and two examples of L. elegans , one from Mount Victory and the other from the Kikori River Basin.