Dendrelaphis papuensis Boulenger 1895: 409

Dendrelaphis papuensis Boulenger 1895: 409 View in CoL

Figs. 1 View FIGURE 1 , 2A, B View FIGURE 2 , 3A View FIGURE 3

Lectotype. BMNH 1946.1 .6.58, obtained by A.S. Meek in Trobriand Islands, Milne Bay Province, Papua New Guinea, designated herein.

Paralectotypes. BMNH 1946.1 .6.57, 1946.1.6.59–61, same data as lectotype .

Additional Material. Papua New Guinea: Central Province: Edevu, 9.2130° S, 147.3097° E ( NHM 2013.275 ); GoogleMaps Milne Bay Province: Alotau , 10.3067° S, 150.4381° E, 5 m a.s.l. ( BPBM 20817 About BPBM ); GoogleMaps Opea Island , 10.6034° S, 150.0113° E, 0–1 m a.s.l. ( UMMZ 245411 View Materials ). GoogleMaps Photographs only: Papua New Guinea: Milne Bay Province: Menapi, Cape Vogel , 9.767° S, 149.917° E ( AMNH 73940 About AMNH ) GoogleMaps ; Iamalele #1, 9.489° S, 150.543° E, Fergusson Island ( AMNH 76642 About AMNH ) GoogleMaps ; Morobe Province: Garaina , 7.88° S, 147.14° E ( AMNH 95613 About AMNH , 107175 About AMNH ) GoogleMaps .

Diagnosis. A moderately sized species of Dendrelaphis (adult SVL up to 856 mm, tail up to 386 mm; TL/SVL = 0.30–0.36); eye approximately equal in diameter to eye-naris distance (EY/EN = 0.92–1.11, mean = 1.01); ventrals 189 in sole male, 182–190 in seven females; subcaudals 120 in sole male, 125–133 in five females; hemipenis without a terminal awn, ornamented proximally with a few whorls of large spines; black bars absent from neck; dark postocular stripe present; venter uniformly tan or yellow, unspotted; dorsum tan; anterior vertebral scales paler than remainder of dorsum, imparting the appearance of a pale anterior vertebral stripe; head tan, lacking dark speckles; supralabials cream with brown or black markings dorsally; chin yellow, often with a few gray specks.

Comparisons with other species. Dendrelaphis papuensis differs from D. lineolatus and D. striolatus in lacking black bars on the neck and in having fewer subcaudals (120–133 vs. 144–151 in D. lineolatus and 133–147 in D. striolatus ); from D. lorentzii in lacking dark speckling on the head, having more ventrals (182–190 vs. 156–181 in D. lorentzii ), and in having an eye subequal in size to the EN distance (vs. eye much smaller than EN distance in D. lorentzii ); from D. keiensis , D. macrops , and D. punctulatus in having a dark postocular stripe; from D. gastrostictus in lacking dense black spotting on the venter; and from D. calligastra in having fewer subcaudals (120–133 vs. 134–156 in D. calligastra ), no terminal awn on hemipenis (vs. present in D. calligastra ), and hemipenis with a few whorls of large spines proximally (vs. with numerous tiny spines proximally in D. calligastra ). The pale anterior vertebral stripe and the large spines on the base of the hemipenes distinguish D. papuensis from all other Melanesian members of the genus.

Redescription of the lectotype. Adult female. Dorsal scale rows 15-13-11, reduction to 13 rows occurs at the level of Ventral 13 and to 11 rows at Ventral 115; all scales smooth. Vertebral scales hexagonal; paravertebral scales elongated, semi-hexagonal; remainder narrowly rhomboidal and oblique to body axis. Rostral slightly wider (2.9 mm) than high (2.4 mm); internasals, frontal, supraoculars, and parietals longer than wide; prefrontals wider than long; lateral extension of parietal contacts upper postocular; nasals divided by large nares, with short dorsal and ventral sutures extending from top and bottom of naris to internasal and first supralabial, respectively; one loreal on each side, longer than high; preocular single, higher than long; postoculars two, upper more than twice as large as lower; anterior temporals two, upper smaller than lower ( Fig. 1 View FIGURE 1 ). Supralabials 8, 4th and 5th below eye; infralabials ten, five contact anterior genials. Posterior genials in contact with infralabials 5 (point contact only) and 6, separated along their posterior half by pair of intergenials; single lateral gular separates posterior genials from infralabial 7. Many small tubercles present on all head shields except frontal.

Vertebral scales hexagonal; remaining dorsal scales oblique. Dorsal scales on tail in ten rows at level of cloaca, four rows at midtail, and two rows near tip. Ventrals 187; cloacal scale divided; subcaudals 123+, last two or three subcaudals missing; ventrals and subcaudals each with a ventrolateral ridge.

Total length 1029 mm; snout-vent length 724 mm; tail length 305 mm, small portion of tip missing.

Dorsal ground color in preservative pale brown, heavily dusted with black on lateral and dorsolateral scale rows beginning around ventral 5 and gradually decreasing posteriorly into sparse flecks; this black field extending approximately 25% of body. Vertebral scale row lighter than surrounding scales anteriorly, margined in black, forming pale stripe to approximately ventral 43; vertebral scales with elliptical white blotch on each anterolateral margin, these decreasing in size posteriorly and disappearing at approximately midbody. Similarly, anterior of each lateral scale with an elliptical white blotch on ventral margin, these too decreasing in size posteriorly and disappearing at midbody. Black postocular stripe extends from eye to end at black lateral field that begins at ventral 6. Labials, chin, and throat pale yellow. All supralabials except the first margined above with black. Infralabials and genials with few gray flecks. Yellow throat changes posteriorly to tan shortly beyond neck, which continues through tail. Each ventral with narrow brown streak on ventrolateral ridge of each scale; these marks absent on subcaudals, the ridges of which are white. Iris black.

Variation. The paralectotypes show little meristic variation ( Table 1 View TABLE 1 ) and are virtually identical to the lectotype in coloration. The sole neonate ( BMNH 1946.1.6.59) has a slightly shorter tail (tail/ SVL = 0.28) than do the adults (tail/ SVL = 0.30–0.36). Scales missing the corneous epidermis are pale gray instead of tan. The venter is paler yellow in the smallest specimen ( BMNH 1946.1.6.59), quickly changing posteriorly to cream and then tan. For the specimen from Opea Island ( Fig. 2A View FIGURE 2 ), just off the southernmost tip of New Guinea ( UMMZ 245411 View Materials ) the black anterolateral color field and the black postocular stripe are not as densely developed as in the type series from the Trobriand Islands . The iris has a narrow margin of silver around the pupil. The venter is pale gray irregularly marked with small flecks or streaks of darker gray or black ( Fig. 2B View FIGURE 2 ). Linear streaks on the ventrolateral ridges are black instead of brown. Overall , this specimen is grayer than the type series, which may reflect either geographic variation in color or the longer period of time in preservative of the type series. The neonate has narrow black preocular, postocular, and neck stripes; whereas adults maintain a wide black postocular stripe, which is beginning to be broken up with brown in the largest specimen ( BMNH 1946.1.6.57) and in the specimen from Opea Island.

Color in life. Judging from a photo in life, this species looks much like it does in preservative, being tan dorsally with a paler tan vertebral stripe and wider dorsolateral stripe that is black anteriorly and soon changes posteriorly to become brown with each scale having a central white area ( Fig. 3A View FIGURE 3 ). It has a narrow black postocular stripe, and the labials and lower sides are white anteriorly.

Etymology. The species name denotes that this species comes from Papua.

Range. This species was conservatively interpreted by van Rooijen et al. (2015) to be restricted to the Trobriand Islands, but I have found it to also occur in southeastern mainland New Guinea ( Fig. 4 View FIGURE 4 ), as reported by McDowell (1984), who stated its distribution to be throughout the Papuan Peninsula as well as in Western Province. That this species is not endemic to the Trobriand Islands is unsurprising considering that those islands were entirely continuous with New Guinea during the last glaciation. Based on current evidence—including photographs of specimens seen by me—this species seems to occur across most or all of the Papuan Peninsula of New Guinea with certainty, but McDowell’s report of this species from farther west—apparently based only on reports from Fred Parker but not direct examination of specimens—requires verification.

Ecology. Like many Papuan Dendrelaphis , this species seems to adapt well to human-modified areas so long as trees and shrubs remain. The Trobriand Islands are largely converted to gardens, my specimen from Alotau came from a well-landscaped lodge, and the specimen from Opea Island came from an offshore islet with native strand vegetation.

Remarks. Boulenger (1895) made clear that he based his description of the species on “several specimens”, these several apparently being the five BMNH specimens referred to as syntypes by van Rooijen et al. (2015) and The Reptile Database ( Uetz et al. 2024). However, although he gave ranges of variation for numbers of ventrals and subcaudals, Boulenger (1895) provided length information for only one specimen, giving values of total length of 1050 mm and tail length of 310 mm, leaving the SVL to be 740 mm by subtraction. These values do not correspond exactly to my measurements of those specimens, but they are closest to that I have chosen to be the lectotype (BMNH 1946.1.6.58), for which I get values of 1029, 305, and 724 mm, respectively. These values are consistent with some shrinkage/distortion of the specimen in the intervening 130 years; other specimens are either too large or much too small to have been the specimen for which Boulenger provided measurements ( Table 1 View TABLE 1 ). Boulenger (1895) did not provide information on the sex of his specimens, so that could not be used in verifying my choice of lectotype.

Identification of this species in the literature has been problematic. The pale anterior vertebral stripe that I find diagnostic for the species was noted in the original description by Boulenger (1895), but he made no special note of its importance. Presence of this feature was repeated by de Rooij (1917) and O’Shea (1996), though neither noted its diagnostic importance. Furthermore, this character was not mentioned by van Rooijen et al. (2015), whose diagnoses among many of the Melanesian Dendrelaphis otherwise relied on several color-pattern features. McDowell (1984) also did not emphasize this attribute because his study was focused on the snakes of the Huon Peninsula, on which D. papuensis is lacking. Despite the fact that his explication of the geographic distribution of this species seems largely to have relied on examination of hemipenes (with the exception of snakes from Western Province, as noted above), information on the presence or absence of the pale vertebral stripe on specimens from mainland New Guinea is lacking, although I have verified its presence on several specimens from throughout the Papuan Peninsula ( Fig. 4 View FIGURE 4 ). McDowell (1984) mentioned that this color-pattern feature was seen in many, but not all, specimens of this species, but inasmuch as he included in his concept of that species three of the species described below—two of which are unstriped—does not allow one to infer whether other specimens he examined from mainland New Guinea also lacked the vertebral stripe. Consequently, as noted above, additional examination of specimens from New Guinea must be had to clarify whether this feature truly varies in D. papuensis , what the exact geographic distribution of that species is on New Guinea, and whether additional species may be involved in samples from that island. Thus, the distribution for this species provided by Tallowin et al. (2016) for the IUCN Red List must be viewed as provisional and requiring verification. What is untenable—and clearly refuted by the data presented here—is that D. papuensis is endemic to the low-elevation Trobriand Islands, which were broadly connected to the New Guinea mainland during the last glaciation.