Journal of Herpetology, Vol. 42, No. 4, pp. 645–667, 2008 Copyright 2008 Society for the Study of Amphibians and Reptiles

Revision of the Genus Leiopython Hubrecht 1879 (Serpentes: Pythonidae) with the Redescription of Taxa Recently Described by Hoser (2000) and the Description of New Species

WULF D. SCHLEIP

Fichtenweg 11, 53340 Meckenheim, Germany; E-mail: [email protected]

ABSTRACT.—This study is the first comprehensive systematic study on the python genus Leiopython Hubrecht 1879 native to New Guinea. The taxonomic arrangement recently made is critically reviewed, and proper descriptions for taxa herein recognized as valid are provided. Twenty external morphological characters were recorded from 90 preserved specimens from throughout most of the distribution of the genus. Thirteen characters were used with principal coordinate analysis to test the diversity of populations from different distributions. Additional evidence for some species was obtained by maximum parsimony and maximum likelihood analysis of mitochondrial DNA sequences (cytochrome b gene) taken from GenBank. Besides three conventional taxa, two new species from the mainland, and one new island species were recognized in accordance with the evolutionary species concept. Additionally, a new locality record is provided.

The island of New Guinea is known for its thon on the northern Torres Strait Islands of extremely diverse flora and fauna and for its Australia. A detailed description of such spec- high level of species endemism (e.g., McDowell, imens can be found in Barker and Barker (1994), 1984; Heads, 2002; Kraus and Allison, 2002; stating that populations would occur on the Austin, 2006). This megadiversity may result islands of Dauan, Saibai, and ‘‘possibly Boigu.’’ from a variety of climates (McApline et al., 1983; Nevertheless, reports of the occurrence on the Prentice and Hope, 2007) and habitats (Austin, mainland of Australia around Pascoe River, 2006) such as tropical savannahs, mixed low- Cape York Peninsula (Ehmann, 1992) remain land forests, rain forests, montane forests, and unconfirmed (Barker and Barker, 1994), as does even alpine vegetation at higher altitude. The the occurrence on Aru Island, mentioned by herpetofauna of New Guinea currently consists Hoser (2000). Some authors assumed that, in the of 109 snake species (Allison, 2007), including former case, specimens may have been con- several endemic python species (Rawlings et al., fused with Liasis fuscus Peters (Kend, 1997; 2004; Allison, 2007). However, some areas are O’Shea, 2007). However, the occurrence of difficult to access, and it is likely that many Leiopython on Normanby Island (McDowell, species remain undiscovered (e.g., Austin, 1975; McDiarmid et al., 1999) appears dubious 2006). Just recently, O’Shea et al. (2004) reported because McDowell (1975) had erroneously the occurrence of Antaresia maculosa in the assigned Bara Bara to the Normanby Island, Western Province, previously unknown from rather than to the mainland of Papua New New Guinea. Guinea (PNG) at the Milne Bay Province as The genus Leiopython (Serpentes: Pythonidae) stated by Boulenger (1898) and Koopman represents medium-sized, unpatterned, terres- (1982). trial, and nocturnal snakes (Barker and Barker, Although specimens of the genus Leiopython 1994; O’Shea, 1996; Kend, 1997) that are widely Hubrecht 1879 are well represented in natural distributed throughout tropical and subtropical history museum collections around the world, New Guinea and closely associated with vari- this genus remains taxonomically rather under- ous wetland habitats. Several populations occur studied. Even though the genus was considered on offshore islands (McDowell, 1975; Barker monotypic until recently, two distinct popula- and Barker, 1994; O’Shea, 1996), such as tions, separated by the Central Mountain Range, Salawati Island (terra typical of Leiopython were recognized in the international pet trade gracilis Hubrecht, 1879), Biak Island (Bron- for over the last 30 years (McDowell, 1975; gersma, 1956), and the remote Mussau Island Parker, 1982; Barker and Barker, 1994; O’Shea, of the Bismarck Archipelago (McDowell, 1975). 1996), called the ‘‘golden’’ or ‘‘northern race’’ Various authors (e.g., McDowell, 1975; Cogger and the ‘‘black’’ or ‘‘southern race.’’ Neverthe- et al., 1983; Barker and Barker, 1994; O’Shea, less, published locality data of specimens and 1996) have mentioned the occurrence of Leiopy- distributional ranges were always assigned to 646 WULF D. SCHLEIP

Leiopython albertisii in the literature (e.g., Parker, eny of pythonid snakes (Rawlings et al., 2008) 1982; McDowell, 1984; Barker and Barker, 1994; identified Leiopython Hubrecht and Bothrochilus A. Allison, Reptiles and amphibians of the Fitzinger as sister species. Rawlings et al. (2008) trans-Fly region, New Guinea, report sub- suggest their placement into a single genus, mitted to World Wildlife Fund, South Pacific namely Bothrochilus Fitzinger, as the oldest Program, WWF PNG Madang Office, Madang, available synonym. However, this placement Papua New Guinea, available at http://www. will require further study. wwfpacific.org.fj/publications/png/Transfly_ Aim of the Study.—In this paper, I critically reptiles_report.pdf [February 2008], unpubl. evaluate the current taxonomic arrangement data. 2006). and provide detailed descriptions of the taxa Taxonomic History.—Hubrecht (1879), obvi- recognized as valid. I also provide new mor- ously unaware of the prior description of Liasis phological and ecological data on the genus that albertisii by Peters and Doria 1878, described a may serve for future research. The examination new species based on a single specimen from of a larger number of specimens from through- Salawati Island. He introduced it to his new out the distribution of the genus makes this the genus Leiopython, an intermediate genus be- first comprehensive systematic study of the tween Liasis Gray 1842 and Nardoa Gray 1842 genus. It is hypothesized that the genus forms (generic name was preoccupied for a starfish, a species complex comprising several species Nardoa Gray 1840) as Leiopython gracilis, charac- under the evolutionary species concept (ESC; terized by a pitted rostral and the presence of sensu Frost and Kluge, 1994). The hypothesis of pits in the supralabials. However, Boulenger species diversity was tested using principal (1893) did not follow Hubrecht’s proposal and coordinate analysis of morphological characters. synonymized Leiopython with Liasis. Later, Stull Additionally, the analysis of available mito- (1935) placed the taxon albertisii as a subspecies chondrial DNA (cytochrome b) sequences pro- of Liasis fuscus Peters 1873 and, although she vides evidence for the separation of the so- had not provided reasoning for her action, called northern and southern races. Stull’s proposal was widely followed by subse- quent workers (e.g., Loveridge, 1948; De Haas, 1950; Brongersma, 1953, 1956). Nevertheless, MATERIALS AND METHODS after examining the skulls of the two taxa, Museum Abbreviations.—Museum abbrevia- Worrell (1961) reassigned specific level to tions used in this paper are those of Leviton et albertisii. The ambiguous taxonomic status of al. (1985) with the following additions of Liasis the type species of (Gray had not International Commission on Zoological No- designated a type species for Liasis) prompted menclature (ICZN); School of Biological Scienc- Cogger et al. (1983) to synonymize Liasis with es, Bangor University, Wales, UK (UWB); ‘‘Vida Bothrochilus Fitzinger 1843 (for discussion, see Preciosa’’ International, Boerne, Texas, USA Stimson and McDowell, 1986). Later, Under- (VPI); and Southeastern Louisiana University, wood and Stimson (1990) conducted a phyloge- Hammond, Louisiana, USA (SELU). netic study and synonymized Bothrochilus and Liasis with Morelia Gray 1842. Kluge (1993) Specimens Examined.—A total of 90 preserved contradicted this study in parts, examining 121 specimens from natural history museum collec- external and internal morphological and behav- tions was examined (Appendix 1). Additional- ioral characters in a phylogenetic study, finding ly, data was obtained from literature sources evidence for the distinction of the taxon albertisii (Peters and Doria, 1878; Hubrecht, 1879; Bron- from other python species and, therefore, gersma, 1953, 1956; McDowell, 1975; Under- resurrected the oldest available synonym Leio- wood and Stimson, 1990; Kluge, 1993). Further- python Hubrecht. More recently, Hoser (2000) more, curatorial staff members of natural introduced two new subspecies and one species history museums provided scale counts along to the genus, but subsequent workers did not with pictures of some specimens (see Acknowl- follow this taxonomic arrangement because of edgments). The numbers of available specimens inadequate descriptions and the lack of evi- from some localities (Salawati, Biak, Mussau dence for the taxa described (see Wu¨ ster et al., and Emirau Islands, Enga Province, PNG, the 2001; Williams et al., 2006). Besides the weakly upper Fly-River region of the Western Province, defined diagnoses and erroneous authorship PNG, and Merauke, Indonesia) were small, and (i.e., assigning Leiopython albertisii to Gray 1842 specimens from other localities (e.g., Torres instead of Peters and Doria 1878), emendation Strait Islands, Australia and southwestern Pa- of the subspecific names is required (Wu¨ ster et pua) were unavailable for this study. Data and al., 2001) because of erroneous latinization pictures from one specimen (road kill) from the (barkeri and bennetti were both named after sets Milne Bay Province, near Alotau, were provid- of two persons). A recent study on the phylog- ed by anonymous. REVISION OF THE GENUS LEIOPYTHON HUBRECHT 1879 647

TABLE 1. Overview of morphological characters used for the multivariate morphometric analysis.

Character Acronym Character states As characterized by Number of dorsal midbody rows DMB Meristic Underwood and Stimson (1990) Character 37; Kluge (1993) Character 98 Number of ventrals VEN Meristic Underwood and Stimson (1990) Character 34; following Dowling (1951) Number of subcaudals SCA Meristic Underwood and Stimson (1990) Character 35 (in part); Kluge (1993) Character 96 Number of supralabials SPL Meristic Underwood and Stimson (1990) Character 7–9; Kluge (1993) Characters 88–90 (in part) Number of infralabials INL Meristic Number of postoculars POC Meristic Underwood and Stimson (1990) Character 11; Kluge (1993) Character 77 Number of supralabials SLE Meristic Kluge (1993) Character 90 entering the eye Presence of suboculars SOC 0 5 absent; Kluge (1993) Character 79 1 5 present Number of loreals LOR Meristic Kluge (1993) Character 87 Number of prefrontals PFR Meristic Kluge (1993) character 84 and 85 Number of parietal scale pairs PAR Meristic Underwood and Stimson (1990) Character 4; Kluge (1993) Character 76 (in part) Parietal scales that border the PML 0 5 no; Underwood and Stimson (1990) Character 5, frontal in contact at the median 1 5 yes included in Kluge’s mutlistate Character 76 line Whitish spot on the postoculars SPT 0 5 absent; Whitish markings on the postoculars and 1 5 present sometimes on the supraocular as well.

Morphological Characters.—Twenty external Basic Statistics.—The morphological charac- morphological characters were recorded based ters of each population (with N . 7) were tested on characters found taxonomically important by for outliers using Grubbs’s test (with 95% previous workers (e.g., McDowell, 1975; Under- significance). Specimens that showed outliers wood and Stimson, 1990; Kluge, 1993). Species- in one or more characters were marked accord- specific characters such as the whitish markings ingly. Statistically significant differences in on the postoculars were used as well. The meristic characters between populations were definition of head scales follows Underwood tested with Kruskal-Wallis (KW)-test and with and Stimson (1990) and Kluge (1993; Table 1). chi-square (x2)-test for binary coded characters. Loreal scales were defined as the scales between KyPlot 2.0 beta 15 (Yoshioka, 2002) was used for the nasal scales and the preoculars. Interpar- the basic statistics. Means are given 6 1 SD. ietals separate the parietal scales at the median Construction of OTUs.—Thirteen operational line or lay between the anterior and posterior taxonomic units (OTU’s) were constructed a pair of parietal scales. Anterior temporals are priori based on distribution (e.g., geographically scales that directly follow the postoculars and isolated or disjunctive populations); currently are in between the supralabials and the parietal accepted knowledge of the genus (recognition scales. Posterior temporals directly followed the of two color races in common usage as stated anterior ones (also see O’Shea, 1996; Fig. 1). earlier) and abiotic data (temperatures, humid- Ventral scale counts were in accordance with ity and annual rainfall) were obtained from Dowling (1951). The whitish markings on McAlpine et al. (1983). Additionally, the ‘‘UN- the postoculars are either present or absent. KNOWN’’ OTU consists of two specimens that Based on my own observations in juveniles and were included in the analysis without adequate adults, these markings develop within the first locality data (BMNH 1892.3.15.1–2 labeled only year of age, depending on the growth of the ‘‘New Guinea’’) to test their species affiliation. specimens. Another OTU contained a single specimen from Previous workers (e.g., Brongersma, 1956; Bulolo, Morobe Province, PNG (Table 2). Three McDowell, 1975; Shine and Slip, 1990) did not preserved specimens from the pet trade (ZFMK find sexual dimorphism in size in this species. 20331–20332, SFM 72616) although with inade- In fact, Brongersma (1956) reported distinctly quate locality data (only labeled ‘‘Irian Jaya’’) larger anal spurs in males, but this seems to were included in the analysis as ‘‘NORTH1’’ be only an insignificant trend (Shine and Slip, OTU. 1990). Therefore, males and females were Morphometric Analysis.—Principal coordinate analyzed jointly. analysis (PCoA; Gower, 1966) was performed 648 WULF D. SCHLEIP

FIG. 1. Simplified illustration of the head scalation in Leiopython (RMNH R-4796, holotype of Leiopython gracilis Hubrecht) A: dorsal surface B: right lateral view. aP - anterior parietal, aT - anterior temporal, f - frontal, lor - loreal, inl - infralabial, pfr - prefrontal, poc - postocular, pro - preocular, pT - posterior temporal, spl - supralabial, spo - supraocular. with MVSP 3.13p (Kovach, 1999), using the Molecular Data.—Three sequences of mtDNA Gower General Similarity Coefficient for mixed (cytochrome b gene) were obtained from the data sets. Data were Log-e transformed before NCBI GenBank. Additionally, two sequences, analysis. PCoA has an advantage over PCA one from a specimen from Madang, and a because it computes the distances between second from a specimen from the Milne Bay OTU’s rather than the correlation between Province near Alotau, were provided (Appen- characters. Therefore, all characters found sta- dix 2). Two of the GenBank sequences (U69835– tistically significant among populations were U69836), submitted by Campbell (1997) were used. Furthermore, PCoA does not assume provided without proper locality data and, normal distribution of the data and, there- therefore, were backtracked to the LSUMZ (B. fore, is applicable for mixed data sets (Gower, N. Campbell, pers. comm.). D. Dittmann from 1966). the LSUMZ provided locality data from the

TABLE 2. Operational taxonomic units (OTU’s) used for the morphometric analysis. The last column (N) shows the sample size of each OTU. Additionally, two specimens (BMNH 1892.3.15.1–2) were included to the analysis without adequate locality to test their affiliation.

Id Acronym of OTUs Localities N 1 NORTH1 Papua (Salawati Island, Sorong, Manokwari, Fak Fak) 13 2 NORTH2 Sepik and Madang Provinces, Eastern Jayapura and north coast of PNG10 3 ENGA Enga Province, Wabag, Sau River Region 2 4 WAU Morobe Province, Wau, Highlands 5 5 BULOLO Morobe Province, Bulolo, Highlands 1 6 KARIM Simbu Province, Karimui, Bomai Tive Plateau 13 7 SOUTH1 Central Province, (Port Moresby Sogeri), Milne Bay Province, 18 PNG, (Alotau), Oro Province (Popondetta) 8 SOUTH2 Western Province, trans-Fly river region 4 9 HUONP Morobe Province: Huon Peninsula, Lae, Finschhafen, 15 Madang Province: Ramu River (near Dumpu) 10 BIAK Biak Island 2 11 MUSS Mussau Island 2 12 EMIR Emirau Island 3 13 UNKNOWN Two specimens with uncertain locality data: BMNH 1892.3.15.1–2, 2 unknown locality Total 90 REVISION OF THE GENUS LEIOPYTHON HUBRECHT 1879 649 tissue sample database of these two samples (pers. comm.). The locality data of the sequence U69836 was found erroneous (stating that Leiopython would occur at New Britain) and, therefore, was verified and corrected to the locality of Madang by H. G. Cogger (pers. comm.). One sequence (AF241406), used as the outgroup in Harvey et al. (2000), is assumed to be from a specimen from Merauke, Indonesia (D. G. Barker, pers. comm.). Two cytochrome b gene GenBank sequences (AF241399 and AF241404) from Morelia amethistina were used as outgroups (Appendix 2). Phylogenetic Analysis.—Sequences were ini- tially aligned by eye and by the built-in alignment algorithm (ClustalW) in MEGA 3.1 (Kumar et al., 2004). Because of variation in FIG. 2. Simplified general patterns of the parietal region. A: two pairs of parietals with the anterior ones sequence length (642–1,114 base pairs), only the (aP) in contact at the median line and the posterior first 642 base pairs (bp) were used, and sites ones (pP) in either contact at the median line or containing gaps were excluded. Phylogenetic separated by one or more small interparietals (‘‘alber- analysis of the remaining 624 sites was con- tisii type’’). B: one pair of large parietal scales followed ducted using maximum parsimony (MP) and by two small scales separated at the median line by an maximum likelihood (ML) methods as imple- elongate scale that gets wider posteriory (‘‘hoserae mented in PAUP* 4.0b10 for Windows (D. L. type’’). C: one pair of large parietal scales followed by Swofford, Sinauer Associates, Sunderland, MA, small, irregular scales (‘‘huonensis type’’). D: two pairs of parietal scales, the anterior ones large, the posterior unpubl. data, 2002). Trees were rooted using ones slender and elongate often separated at the Morelia amethistina sequences as the outgroup median line by an elongate interparietal about the (Appendix 2). In the MP analysis, data were length of the posterior parietal scales. Small interpar- treated as equally weighted, and phylogenetic ietals are often present in the center of the anterior and inference was performed by exhaustive search. posterior parietals (‘‘fredparkeri type’’). Additionally, 2,000 bootstrap replicates were performed using heuristic search with tree bisection-reconnection (TBR) branch swapping median line followed by a second pair of large and random stepwise addition. ML was per- scales often separated from the median line by formed using these parameters as well along one or more small interparietals (Fig. 2A). with the substitution model (HKY85 + G) Specimens from the southern part show a large estimated by Modeltest 3.7 (Posada and Cran- pair of parietals followed by two smaller scales dall, 1998). This model revealed 47 distinct data separated from the median line by an elongate patterns. A genetic distance matrix was com- scale that gets wider posteriorly (96%, N 5 22). puted using the substitution model estimated This pattern (Fig. 2B) was found in all southern by Modeltest. specimens from Alotau on the east coast of PNG to Merauke in the Papua Province. Specimens from the Huon Peninsula (93%, N 5 15), as well RESULTS as two specimens from the Enga Province General Patterns of Parietal Scales.—The exam- (100%, N 5 2), showed a large pair of parietal ination of specimens from throughout the scales followed by small, irregular scales distribution uncovered specific patterns in the (Fig. 2C). Specimens from Karimui and the Tive scale arrangement of the parietal region (Fig. 2). Plateau both in the Simbu Province at an In specimens from the northern part of New elevation of 1,000–1,500 m have two pairs of Guinea, including the populations from the parietal scales with the posterior ones being Bismarck Archipelago, Biak, and Salawati Is- slender and elongate (83%, N 5 12), often lands, two pairs of parietals are present, separated at the median line by an elongate whereas only one pair was recognized in interparietal scale (Fig. 2D). Further information specimens from the southern parts of New and schemata of the parietal structures can be Guinea and from the Huon Peninsula. This is in found at http://www.leiopython.de. agreement with McDowell’s (1975) findings, Morphological Analysis.—Some characters (num- and O’Shea (1996: fig. 12) identified two pairs ber of anterior and posterior temporals, parietal of parietals in a specimen from Siar, Madang scales in contact with the postoculars) were Province, PNG. In northern species, the anterior excluded from the analysis because of little pair is most often (93%, N 5 30) in contact at the variation within the genus or because of random 650 WULF D. SCHLEIP

FIG. 3. Scatter plot of the PCoA. The three axes comprised 49.63% of the total variance. See text for abbreviations. distribution throughout different populations. OTU’s suggested an affiliation of these speci- Head and body color were excluded from the mens to this taxon. However, the ‘‘BULOLO’’ analysis, because of possible fading of colors OTU was also found within this cluster, but it is during prolonged storage in preservative. Nev- difficult to draw conclusions on the taxonomic ertheless, color pattern was considered in live placement of this population based on only one specimens. The patterns of the parietal region as specimen. The third cluster (C) primarily described above may be helpful to identify contained the ‘‘KARIM’’ OTU. However, the species but were not included in the morpho- analysis was not able to separate two speci- metric analysis. mens, from Kokoda (MCZ R-84333, ‘‘SOUTH1’’) In the scatter plot (Fig. 3) of the PCoA, the and another from Mussau Island (ZMUC R- first axis (PCo 1) summarized 25.7% (eigenvalue 5445, ‘‘MUSS’’) from this OTU in morphometric 5 6.311), the second axis explained 14.3% space. Nevertheless, the Karimui population (eigenvalue 5 3.499), and the third axis (PCo was assumed geographically isolated from 3) explained 9.6% (eigenvalue 5 2.357) of the other populations, and, along with diagnosable total variance. The scatter plot revealed the differences in morphology, it is herein regarded existence of seven clusters. Besides the ‘‘WAU’’ as a new species Leiopython fredparkeri sp. nov. OTU, cluster A, which represented the taxon The two specimens from the Enga Province bennettorum, also included two specimens from (‘‘ENGA’’ OTU) formed a tight cluster (D) in the ‘‘SOUTH’’ OTU’s, one from Port Moresby morphometric space but were morphologically (MCZ R-145941, ‘‘SOUTH1’’) and another one indistinguishable from specimens from the from Oslobip (MCZ R-129393, ‘‘SOUTH2’’) Huon Peninsula (‘‘HUONP’’ OTU), represented because of their aberrant morphology from by the neighboring cluster (E) and, therefore, other specimens of the ‘‘SOUTH’’ OTU’s. Both were not recognized taxonomically. One spec- specimens showed two pairs of loreals and imen (AMNH R-107148) from the ‘‘HUONP’’ prefrontals (see subsection Variation in Leiopy- OTU was detected as an outlier (Grubbs-test, thon hoserae). Cluster B comprised four OTU’s, 95% significance; N 5 15, PG 5 2.372) in several of which the two ‘‘SOUTH’’ OTU’s correspond- morphological characters (see also McDowell, ed to the taxon hoserae. The considerable overlap 1975) and was specially marked in the ordina- of the ‘‘UNKNOWN’’ OTU with the ‘‘SOUTH’’ tion plot. The cluster F comprised four REVISION OF THE GENUS LEIOPYTHON HUBRECHT 1879 651

OTUs (‘‘NORTH1,’’ ‘‘NORTH2,’’ ‘‘MUSS,’’ and

‘‘EMIR’’) with the ‘‘NORTH1’’ and ‘‘NORTH2’’ 1.41 0.35 3.54 1.1 1.41 )

OTUs corresponding to the conventional taxon 6 6 6 6 % albertisii. These four OTUs formed a tight cluster 6 n in Table 1. sp. nov. with considerable overlap, indicating few mor- 0 1 1 without ZMUC present phological differences between the populations 3

represented by the OTUs. Therefore, specimens biakensis from Mussau (‘‘MUSS’’ OTU) and Emirau (‘‘EMIR’’ OTU) islands are herein considered conspecific with albertisii, but further research is 0.26 2 0.450.24 11–12 (11.8) 3 2.993.94 45–47 (46) 0.59 65–70 (67.5) 14–16 (15.3) needed to determine the actual taxonomic status 6.12 270–272 (271) 6 6 6 6 6 6 of these island populations. Nevertheless, spec- 6 ) yes (100 5, 6 5 5 5 imens from Biak Island (‘‘BIAK’’ OTU) were % without MVZ 40848, sp. nov. found distinct from albertisii forming a separate 2 cluster (G) and are herein regarded as a separate species Leiopython biakensis sp. nov. Clinal Variation.—Clinal variation was found huonensis in scalation patterns in the two widespread species Leiopython albertisii and Leiopython hoserae. In Leiopython albertisii, ventral scale counts 0.43 2–3 (2.9) 0.320.45 12–13 (12.7) 3–4 (3.2) 3.350.59 65–78 (70.6) 15–17 (16.1) 3.15 258–282 (268.2) 0.52 0 0.13 1 increase from east to west as reported previously 1.04 43–55 (48.3) without MVZ 74917, 6 6 6 6 6 6 6 6 6 1 by McDowell (1975). Furthermore, the number of ) yes (93 % subcaudal scales and maximum midbody scale sp. nov. rows were highest in the west at Salawati Island and in the latter case were lowest in the east at Emirau Island. In Leiopython hoserae, the number except CAS 103384 and 135211.

of subcaudal scales was higher in specimens of 7 the Western Province than they were in speci- mens found in the east, at the Central and Milne 0.37 2–3 (2.7) 0.370.35 12–14 (13.0) 2–4 (3.1) 1.503.33 47–51 (49.1) 0.66 63–76 (69.8) 16–18 (17.2) 4.30 266–277 (270.5) 0.360.21 0–1 (0.5) 2 1 2 0.35 1–2 (1.1) Bay Provinces. No significant variation was 0.35 1 1 (BMNH 1892.3.15.1–2 were excluded, see text). Character acronyms as give 6 6 6 6 6 6 6 6 6 6 6 observed in other characters. ) yes (85 Polymorphism.—Polymorphism was not detect- % ed in populations of Leiopython albertisii.Howev- er, polymorphic patterns were observed in head Leiopython scalation in populations of Leiopython hoserae from the Central and the Western Provinces. Variation within populations was found in the number of infra- and supralabials, the number of suprala- 0.35 2–4 (3.0) 2.303.39 45–52 (48.1) 0.35 64–77 (68.7) 15–18 (16.6) 4.45 264–278 (272.6) 0.45 1–2 (1.1)

bials entering the eye, and in the number of except BMNH 1922.11.24.29, and 6 6 6 6 6 6 6 postoculars, but little variation was found in ) yes (96 is SD. The following exceptions must be considered: other morphological characters (Table 3). % 01 0–1 (0.2) 1–2 (1.0) Phylogenetic Analysis.—Of the 624 bp, 525 6 were constant; 31 sites were parsimony unin- absent absent absent present formative, and 68 were parsimony informative. The most parsimonious tree (Fig. 4A) was obtained by exhaustive search (tree length [TL] 5 117; consistency index [CI] 5 0.940; homo- 0.18 3 2–3 (2.8) 0.280.25 13 3–4 (3.5) 12–14 (13.1) 2.532.85 49–54 (51.6) 0.38 62–70 (66.0) 16–18 (17.0) plasy index [HI] 5 0.060; retention index [RI] 5 4.83 263–274 (268.4) except AMNH R-107148, 6 6 6 6 6 6 0.921; rescaled consistency [RC] index 5 0.965). 6 5 ) no (100 + 3 Modeltest estimated HKY85 G (base frequen- % 7 5 5 5 5 4 cies: A 0.3108, C 0.3315, G 0.1149, T (47.1) (71.5) (15.9) (274.5) 1 1

5 5 1 0.2428; Ti/Tv ratio 8.7913; shape parameter albertisii bennettorum hoserae fredparkeri 0.0139) as the most adequate substitution model 1, 2 for the ML analysis. Seventy-four distinct data patterns were found under this model. The best ML tree (Fig. 4B) was evaluated (of 945 trees) by except MVZ 40847, 3. Scale comparison and basic statistics of the taxa of the genus exhaustive search (TL 5 119, likelihood-ln L 4 1404.64980, CI 5 0.924, RI 5 0.898, RC 5 0.830, 5 52 ABLE HI 0.076, G-fit 65.750). The genetic dis- T Character SOC 0 IFLSLE 15–17 2–3 (3.0) SPLPOC 12–13 (12.9) 2–4 (3.0) VEN 262–283 Sample sizeDMB 30 43–51 5 22 13 15 2 PMLSPT yes (93 present SCA 65–79 Numbers in parentheses represent mean values and LORPFRPAR 1 1 2 2–3 (2.3) 2 1–2 (1.1) tance was about 0.6% within species, whereas R-5445, 652 WULF D. SCHLEIP

FIG. 4. Phylograms of the maximum parsimony (MP) (A) and the maximum likelihood (ML) tree (B). Numbers above the branches represent bootstrap proportions. Bootstrap proportions less 50% are not shown. Numbers to the right of the internal nodes in the MP tree, and below the branches in the ML tree represent branch lengths. The ML tree differs from the MP tree only by the arrangement of the Merauke and Western Province samples within the branch representing Leiopython hoserae. the distance between the species was between Synonymy.—Liasis alberitisii Peters and Doria, 8.7 and 9.3%. 1878 Leiopython gracilis Hubrecht, 1879 SPECIES ACCOUNTS Liasis albertisii Boulenger, 1893 Leiopython albertisii Peters and Doria 1878 Liasis fuscus albertisii Stull, 1935 REVISION OF THE GENUS LEIOPYTHON HUBRECHT 1879 653

Liasis fuscus albertisi Capocaccia, 1961 infralabials, seventh to 12th pitted and in a skin Liasis fuscus albertisii Stimson, 1969 fold; single loreal and preocular scale; usually Liasis albertisii Switak, 1973 three postoculars; four to five anterior and Liasis albertisii McDowell, 1975 posterior temporals; one pair of elongate pre- Bothrochilus albertisii Cogger, Cameron, and frontals, two to three times longer than the Cogger, 1983 internasals; bell-shaped frontal; two pairs of Lisalia albertisii Wells and Wellington, 1984 parietals (see also Peters and Doria, 1878: pl. III, Morelia albertisii Underwood and Stimson, 1990 fig 2), anterior ones in median contact but most Leiopython albertisii Kluge, 1993 often not in contact with the uppermost Leiopython albertisii McDiarmid et al., 1999 postoculars, posterior parietals sometimes sep- Leiopython albertisii albertisii Hoser, 2000 arated by one or more interparietals at median line. Scales are smooth and arranged in 43–51 Syntypes.—Two specimens from Papua col- longitudinal dorsal rows; 262–283 ventrals; anal lected at Kapoar [5 Fak Fak regency], Onin entire; 65–79 subcaudals mostly paired. peninsula (MSNG 29990) by Mr. L. M. D’Alber- Color in Life.—The color of the dorsum is tis on April 1872 and at Andai (near Dorei) yellowish or reddish-brown to purplish-brown [5Manokwari regency], Birds Head Peninsula with an iridescence gleam, fading to lighter (MSNG 29989) collected during June 1875 by yellowish laterally, and white ventrally. Speci- Beccari. The present state of disposition of the mens from the Sorong peninsula are often syntypes is unknown (see also McDowell, 1975). lighter yellowish in color dorsally (D. G. Barker, Diagnosis.—Leiopython albertisii is distinguish- pers. comm.). Specimens from the Fak Fak and able from Leiopython hoserae, Leiopython bennet- Manokwari regencies, as well as from the torum, and Leiopython huonensis sp. nov. by the Madang Province are more brownish-violet presence of two pairs of parietals. A pair of dorsally fading to brownish-yellow laterally large scales often separated from the median and white ventrally. The dorsal surface of the line by one or more small interparietal follows head is shiny black with iridescence gleam. the anterior pair (Fig. 2A). It further differs from Supra- and infralabials are white with black the former two species and from Leiopython vertical markings on anterior edge of suprala- fredparkeri sp. nov. by the presence of whitish bials, reaching into infralabials. Chin is white. postocular spots, and can easily be distin- Whitish markings on postoculars (sometimes on guished from Leiopython bennettorum in the supraoculars, too) were found in all but one number of loreals and prefrontals (Table 3), specimen (ZMUC R-5445, Mussau Island). This the average number of postoculars (KW-test: species also shows physiological color change 2 x 1 5 14.22, P , 0.001), dorsal midbody rows by turning lighter at night and darker brownish- 2 (KW-test: x 1 5 7.98, P , 0.01), ventrals (KW- violet within a period of 20 min during the day 2 test: x 1 5 5.59, P , 0.05), and subcaudal scales (pers. obs.). Pictures of this species in the wild 2 (KW-test: x 1 5 8.54, P , 0.01). Leiopython alber- can be found in O’Shea (1996, 2007). tisii further differs from Leiopython hoserae and Color in Preservative.—Specimens in preserva- from Leiopython fredparkeri in lighter dorsal color tive show only little fading in body color, but and in having a yellowish flank (Parker, 1982; ventral scales sometimes have turned yellowish. Barker and Barker, 1994; O’Shea, 1996, 2007), The dorsal surface of the head is black and the smaller average body size in hatchlings and chin is white or off-white. adults, and additionally from Leiopython hoserae Variation.—In general, only little variation in by molecular evidence (see Fig. 4). Leiopython morphological characters occurs in Leiopython albertisii can be distinguished from Leiopython albertisii throughout its distribution. Minor biakensis sp. nov. by higher average subcaudal differences in the presence/absence and in the (71.5 6 2.85; range 5 65–79, N 5 30 vs. 67.5 6 number of interparietals were recognized. Ad- 3.54; range 5 65–70, N 5 2), and supralabial scale ditionally, as already stated by McDowell counts (12.9 6 0.28; range 5 12–13, N 5 30 vs. (1975), ventral counts of specimens from the 11.8 6 0.35; range 5 11–12, N 5 2) along with a western part of Papua (Birds Head [including higher number of supralabials entering the eye Salawati Island] and the Onin Peninsulas) (3.0 6 0.18; range 5 2–3, N 5 30 vs. 2, N 5 2). exceeded those found in specimens from east- Description.—Medium-sized unpatterned py- ern Jayapura (near the border to PNG) and from 2 thon; elongate head and snout; rostral with two PNG significantly (KW-test: x 1 5 14.80, P , to three diagonal pits; nasals with diagonal 0.001). Nevertheless, no significant variation slitlike pits; one pair of internasals; 12–13 was found in other characters, but two speci- supralabials, first and second or first to third mens from the Sorong peninsula (pictures only) pitted and fifth to seventh entering the orbit, were morphologically aberrant from Leiopython lateral depression dorsally on fifth and some- albertisii. UTA R-36312 had a second pair of times also on the sixth supralabial; 16–17 small posterior prefrontals and a third pair of 654 WULF D. SCHLEIP

FIG. 5. Distribution map of specimens examined. The line in the center of the island separates Papua New Guinea (east) from Indonesian Papua (west). The question marks indicate populations of unclear taxonomic status. See text for details. parietals bordering the first and second pairs Madang Province (Siar, Madang, and Alexisha- laterally on one side and the loreal and fen), although no specimens were sighted in the preocular on the other side. In UTA R-44692, Kau Wildlife area, a lowland rain-forest area posterior parietals were fused, and enlarged near Madang (Austin, 2006). In the Morobe temporals bordered the anterior and posterior Province, this species is replaced by Leiopython parietals laterally. huonensis sp. nov. at the southern Huon Penin- Ontogeny.—Ontogenetic color change was sula and by Leiopython bennettorum at Wau. observed in Leiopython albertisii, but not in Emirau Island Population.—The occurrence of Leiopython hoserae. Hatchlings are uniformly Leiopython at Emirau Island, St. Matthias group very dark grayish-black in color, including the in the Bismarck Archipelago is first reported ventral and subcaudal scales. The whitish spots herein. This contradicts statements made by are absent in juveniles but develop with age and previous authors that the genus does not occur growth. The dorsal color turns into a brownish- elsewhere in the Bismarck Archipelago than on violet, and ventral and subcaudal scales turn Mussau Island (McDowell, 1975; O’Shea, 1996). white in semiadult specimens (pers. obs.). The Three specimens were collected in 1944 by the yellowish color ventrolaterally develops at ornithologist C. G. Sibley and were stored at the last. MVZ (MVZ 40847–40849). The following brief Distribution.—Leiopython albertisii is a wide- description of the specimens may serve for spread species ranging from Sorong (including future research. Specimens of this population Salawati Island) and the Fak Fak regency on the are similar to Leiopython albertisii in general ap- Onin Peninsula to Madang at the east coast of pearance (Fig. 6A, B). Scale counts are 13/13 (12/ PNG (Fig. 5). However, it remains unclear 12 in MVZ 40847) supralabials, first and second whether the distribution is disjunctive at the pitted, fifth with lateral depression, fifth to southeastern part of the Manokwari regency seventh (fifth to sixth in MVZ 40847) entering and along the north coast of the Yapen- the orbit; 15/15 infralabials (lower jaw absent in Waropen and Nabire regencies. Nevertheless, MVZ 40848), seventh to 11th pitted; single loreal specimens were collected at the Jayapura and preocular; single supraocular; 3/3 postocu- regency (Toem, Doromena, Joka and around lars; 4–5 anterior temporals; 4–5 posterior tem- the Lake Santiani), the Sanduan (former West porals; two pairs of parietals, both in median Sepik) and East Sepik Provinces (on the foot of contact enclosing a small rhombic interparietal in the Torricelli Mountains, Aitape, Mt. Somoro the center; maximum dorsal scale rows 43–44, and Lumi, Wewak, Marienberg) and at the ventrals 273–275, anal entire, subcaudals 68–70, REVISION OF THE GENUS LEIOPYTHON HUBRECHT 1879 655

FIG. 6. Dorsolateral or dorsal (left) and lateral (right) views of the head of specimens. (A, B): Leiopython albertisii from Emirau Island, MVZ 40849, (C, D): holotype of Leiopython hoserae AMNH R-107150, (E, F): holotype of Leiopython bennettorum BPBM 5452, (G, H): holotype of Leiopython fredparkeri CAS 118906, (I, J): holotype of Leiopython huonensis AMNH R-95535, and (K, L): specimen of Leiopython albertisii from Mussau Island ZMUC R- 5444. 656 WULF D. SCHLEIP mostly paired. The color of the dorsum is light scale; enlarged anterior temporals; midbody 50; reddish-brown fading to yellowish laterally and ventrals 264; anal entire; subcaudals 73. off-white ventrally in preservative. The dorsal Description of Paratype.—CAS 118910: supra- surface of the head is black. labials 13/13, first two pitted, fifth to seventh entering the orbit with a lateral depression at Leiopython hoserae Hoser 2000 the fifth to sixth scute; infralabials 16/16, with Figure 6C, D 8–12/9–13 pitted and in a skin fold; loreals 1/1; Holotype.—AMNH R-107150, a large male preoculars 1/1; postoculars 3/3 without a specimen from Wipim, Western Province, whitish spot; supraoculars 1/1; one pair of PNG, 2.41 m in length, collected by F. Parker prefrontals; one pair of parietals separated by in August 1969. two small interparietals at the median line, an Paratype.—CAS 118910, an adult (sex un- elongate scale behind the posterior interparietal known) specimen from the Laloki River/Brown that gets wider posteriorly and is laterally River Road, Central District, PNG, collected by bordered by two small scales posteriorly in F. Parker on 29 August 1967. contact with the parietals; parietals do not meet Diagnosis.—Leiopython hoserae can be distin- the uppermost postoculars, midbody scale rows guished from Leiopython fredparkeri, Leiopython 47; ventrals 277; anal entire; subcaudals 66. albertisii, and Leiopython biakensis by the pres- Color in Life.—The body color of this species is ence of only one pair of parietals followed by a blackish-blue dorsally fading to grayish ventro- characteristic scale pattern of two small scales laterally and white laterally. The dorsal surface separated from the median line by an elongate of the head is shiny black and the chin is white. scale that gets wider posteriorly (see Fig. 2B, The whitish spot behind the eye is absent. 6C). It further differs from the former species in Pictures of specimens in the wild can be found a lower average number of dorsal midbody in Barker and Barker (1994), and in O’Shea 2 (1996, 2007). rows (KW-test: x 1 5 5.68, P , 0.05), and from the latter two species by the absence of the Color in Preservative.—The holotype’s body whitish postocular spot. Furthermore, Leiopy- color in life was, according to Parker (1982), very dark grey to black and iridescent. The thon hoserae exceeds Leiopython albertisii and dorsal surface of the head was shiny black. The Leiopython biakensis in adult and hatchling body preserved specimen today shows a brownish- size and is darker in color (Parker, 1982; Barker violet head and body color, the head was only and Barker, 1994; O’Shea, 1996). Molecular slightly darker than the body, rich brown color evidence also supports the separation of Leiopy- ventrolateral fading to yellowish ventrally. thon hoserae Leiopython albertisii from (genetic Similar color fading was also found in the % distance of up to 9.3 ). It differs from Leiopy- paratype. thon bennettorum in the number of loreals and Variation.—Variation in morphological char- prefrontals as well as in lower midbody scale acters was found in several specimens. Three x2 5 , row counts (KW-test: 1 8.92, P 0.01) and specimens (AMNH R-107150 from Wipim, in the average number of postoculars (KW-test: 2 USNM 213396 and USNM 213397 from Port x 1 5 7.19, P , 0.01). Leiopython hoserae can be Moresby) had only two supralabials entering distinguished from Leiopython huonensis by the the eye, but the latter two specimens with a absence of the whitish postocular spot, the small subocular scale on each side. Another two characteristic scale arrangement in the parietal specimens (USNM 213398 and USNM 213399 region (Leiopython huonensis has one pair of from Laloki River, near Port Moresby) showed parietals followed by small, irregular scales) 2 three supralabials entering the eye on one side, and in higher ventral scale counts (KW-test: x 1 but only two on the other side with an addi- 5 6.62, P , 0.05). Leiopython hoserae also occurs tional small subocular. The number of posto- in drier and hotter climate conditions than other culars was more variable than in Leiopython taxa of the genus (detailed below). albertisii, so were the supralabial counts. Three Redescription of Holotype.—Supralabials 13/13, specimens, one from Port Moresby (MCZ R- first two pitted, fifth to sixth entering the orbit 145941), and the other two from the Western and with a lateral depression; infralabials 18/18; Province, PNG from Abam (USNM 195754) and loreals 1/1; preoculars 1/1; postoculars 3/3, from Oslobip (MCZ R-129393) had a small without the whitish postocular spots; supra- triangular scale bordering the prefrontal, loreal, oculars 1/1; one pair of prefrontals; anterior and preocular scales laterally on each side. parietals in median contact anteriorly, but do These scales might be considered a second small not meet uppermost postoculars, right parietal loreal scale. USNM 195754 also showed four scale larger than the left; two scales behind the postoculars on the right side, whereas MCZ R- parietals, the left smaller than the right, both 145941 and MCZ R-129393 had a small scale separated at the median line by an elongate posteriory bordering the prefrontal scales at the REVISION OF THE GENUS LEIOPYTHON HUBRECHT 1879 657 median line and anteriorly bordering the In PNG, Leiopython hoserae is a common frontal. Furthermore, MCZ R-145941 and MCZ species around the Port Moresby/Sogeri area, R-129393 as well as one specimen from Sogeri where it replaces the more commonly found (CAS 118939) showed aberrant morphology Papuan Carpet Python (Morelia spilota variegata) from Leiopython hoserae and could be confused in the lowland (O’Shea, 1996). It is also found with Leiopython bennettorum in having a second around Alotau (Milne Bay Province) and Po- pair of prefrontals. Specimens from the Western pondetta (Oro Province). In the highlands of the Province (N 5 4) showed significantly higher Owen Stanley Range specimens were found at 2 average subcaudal scale counts (KW-test: x 1 5 Garaina and Kokoda, but these specimens are 6.46, P , 0.05) than found in populations from somewhat aberrant from other specimens of this the Central and Milne Bay Provinces (68.1 6 species (see remarks below). Although Leiopy- 2.76; range 5 64–73, N 5 18 vs. 73.3 6 2.76; thon hoserae was found in the Gulf Province at range 5 70–77, N 5 4). Furthermore, the Malalaua (W. Wu¨ ster, pers. comm.), near the average number of infralabials was higher in border to the Central Province, and at Omati Western Province specimens than in the latter (O’Shea, 1996), near to the border of the Western two populations (16.5 6 0.58; range 5 15–18, N Province, this species seems to be absent from 5 18 vs. 17.1 6 0.85; range 5 16–18, N 5 4). All major parts of the Gulf Province (see also PNG specimens of Leiopython hoserae had only distribution maps in O’Shea, 1996). Heads one pair of parietals showing a characteristic (2001) gave a possible explanation for this arrangement, also seen in a captive bred disjunct distribution pattern assuming that this specimen (parents both from the Gulf region) is because of sinking ocean floors in the Gulf shown in Barker and Barker (1994:73) and in a Province. In the Western Province Leiopython specimen from Merauke (UTA R-195941, pic- hoserae is commonly found in the southern tures only). A small triangular interparietal trans-Fly River region (Oriomo Plateau). Spec- separated the parietal scales along the median imens were reported from Daru (mainland), line in three specimens from around the Brown Abam, Boze, Wipim, and Morehead. North- River (near Port Moresby; MCZ R-150792, MCZ wards it was found at Emeti and at the upper R-145941 and USNM 195609, but not in USNM Fly River region around Lake Murray, Kiunga, 195611 from Boroko, Brown River). In two Ningerum, and Oslobip but is not a common specimens, one from Kokoda (MCZ R-84333) species there (Parker, 1982). Barker and Barker and the other one from Garaina (AMNH R- (1994) gave a detailed description of specimens 101073) two pairs of parietals are present (for from the northern islands of the Torres Strait, the latter see also McDowell, 1975). In the politically belonging to Australia. In the Indo- Kokoda specimen, the anterior parietal scales nesian part of New Guinea, this species occurs were also separated along the median line by around Merauke, but it remains unclear how far three small interparietals. Furthermore, this the distribution of Leiopython hoserae extends specimen showed an unusual head color, westward. However, specimens were collected perhaps caused by preservation. Moreover, this at the coastal area of Timika but showed specimen also had several black markings on morphological differences to Leiopython hoserae the genial scales of the throat, only seen in one from PNG. Nevertheless, because of inadequate other specimen examined from Popondetta sample size, specimens from the south coast of (MCZ R-140775). A single specimen from Papua are considered conspecific with Leiopy- Timika, southwest Papua (UMMZ 227653, pic- thon hoserae, but further research is needed to tures only), as well as a number of live speci- determine the actual status of these populations. mens exported for pet trade from Indonesia Leiopython hoserae is primarily found in drier (examined by the author) as the so-called black areas, receiving lower annual rainfall than race White-Lipped Python showed two pairs of Leiopython albertisii (1,800–2,000 mm vs. . 2200 parietals, not seen in specimens of Leiopython mm) and having a pronounced dry season with hoserae from PNG and Merauke. The whitish mean monthly rainfall of 100 mm or less (vs. spot found in Leiopython albertisii was absent in . 100 mm for Leiopython albertisii, except the all specimens of Leiopython hoserae examined. Wewak region) (McAlpine et al., 1983). Both the Distribution and Ecology.—Leiopython hoserae Port Moresby/Sogeri and the trans-Fly area are differs not only morphologically from Leiopy- called the ‘‘dry strips’’ in PNG (Nix, 1982:56; thon albertisii but is also geographically separat- Crisp et al., 2001: fig. 10), and according to Heads ed from the latter taxon by the Central (2002), these areas show the highest diversity in Mountain Range, a known zoogeographic bar- snakes (for the trans-Fly region, also see A. rier (Austin, 2000; Whittier et al., 2000; Rawlings Allison, Reptiles and amphibians of the trans-Fly and Donnellan, 2003) that limits the distribu- region, New Guinea, report submitted to World tional range of Leiopython hoserae to the south of Wildlife Fund, South Pacific Program, WWF the mountain range. PNG Madang Office, Madang, Papua New 658 WULF D. SCHLEIP

Guinea, available at http://www.wwfpacific. R-142857) in 1997 (C. Kishinami, pers. comm.). org.fj/publications/png/Transfly_reptiles_report. H. Clissold collected this specimen on 24 pdf [February 2008], unpubl. data, 2006). Leiopy- August 1963 at an elevation of about 1,066 m thon hoserae shares its distribution with many in a forest habitat. The snake is skinned out other snake species (e.g., Morelia spilota variegata, except for most of the skull. Demansia vestigiata, Pseudechis papuanus, and Diagnosis.—Leiopython bennettorum is easily others) that do not occur north of the Central distinguishable from all other members of the Mountain Range and are absent from much or all genus Leiopython by higher loreal scale count of Gulf Province (see also distribution maps in and by a second pair of small lateral prefrontals O’Shea, 1996). (Fig. 6E, F). Furthermore, higher average mid- The vegetation found throughout the drier body scale row and postocular scale counts se- parts of the distribution consists of dry ever- parate this species from Leiopython hoserae (KW- 2 2 green forests, woodlands, and mixed or Euca- test: x 1 5 8.92, P , 0.01 and x 1 5 7.19, P , 2 lypt savannahs in the dry season that are 0.05), Leiopython huonensis (KW-test: x 15 4.48, P 2 frequently burning. During the wet season, , 0.05 and x 1 5 3.95, P , 0.05), Leiopython 2 2 these areas are only infrequently flooded for fredparkeri (KW-test: x 1 5 5.30, P , 0.05 and x 1 short periods (see McAlpine et al., 1983), at 5 4.28, P , 0.05), and Leiopython albertisii (KW- 2 2 which the vegetation turns more into rain test: x 1 5 7.98, P , 0.01 and x 1 5 14.22, P , forests (Paijmans, 1976). 0.001). It can also be distinguished from the Remarks.—Hoser’s (2000) introduction of Leio- latter two species in having only one pair of python hoserae was heavily criticized by subse- parietals and, additionally, differs from Leiopy- quent workers because of a lack of evidence thon albertisii and Leiopython biakensis by the (e.g., Wu¨ ster et al., 2001; Williams et al., 2006). absence of the whitish postocular spot (contra The author had distinguished this taxon from Hoser, 2000). Leiopython albertisii by body color and by larger Redescription of Holotype.—Total length 635 average size only. These characters alone are mm, SVL 540 mm head length 21 mm; 13/13 somewhat problematic, at least in preserved supralabials, first and second pitted, fifth to specimens, because color fade may have taken seventh entering the orbit; 17/18 infralabials, 8– place in specimens preserved for a long time, as 14/8–15 pitted in a skin fold; 3/3 loreals (a shown above. Nonetheless, dorsal color was second smaller one dorsolaterally bordering the considered as a significant character in live large loreal on one side, laterally bordering the specimens. Body size is subject to environmen- prefrontal on the other side, the third loreal tal factors and life-history traits (e.g., Forsman, posteriorly bordering the first, dorsolaterally 1991; Madsen and Shine, 2000; Wu¨ ster et al., the second and anteriorly bordering the pre- 2001). Furthermore, Barker and Barker (1994) ocular); single preocular; single supraocular; 4/ and O’Shea (1996) reported behavioral differ- 4 postoculars; 6/6 anterior temporals; 7/7 ences to Leiopython albertisii. The placement of posterior temporals; one pair of parietals about specimens from the highlands of the Owen two-thirds of the length of the elongate prefron- Stanley Range around Garaina and Kokoda into tals, separated by two small rhombic interpar- this species is ambiguous because of their ietals at the median line; parietals not in contact morphological differences from all other spec- with the postoculars. Whitish postocular spot is imens examined. Therefore, I recommend fur- absent; stomach contains rat. According to the ther collecting and research on specimens from size and body color of this specimen, it can be the highlands of the Owen Stanley Range. assumed that it is only a few months old. Scales smooth with 274 ventrals, anal entire, and 68 Leiopython bennettorum Hoser 2000 subcaudals mostly paired. Figure 6E, F Redescription of Paratypes.—BPBM 3890: head Synonymy.—Leiopython albertisii bennetti Hoser length 55 mm; 13/13 supralabials, first to third 2000 (need of emendation) pitted, fifth to seventh entering the orbit; 17/17 Holotype.—BPBM 5452, a juvenile female infralabials, 8–14/8–14 pitted in a skin fold; 2/2 specimen from near Wau, PNG collected by A. loreals; single preocular; single supraocular; 4/3 C. Ziegler on 13 June 1967. postoculars; 5/5 anterior temporals; 5/7 poste- Paratypes.—BPBM 3890, a male specimen rior temporals; 263 ventrals and 62 subcaudals. from near Wau, PNG, collected by O. R. Wilkes AMNH R-142857 (former BPBM 5137): supra- on or around 1 May 1966 at an elevation of labials 14/13, first two pitted, fifth to seventh about 1,150 m. The snake is skinned out except entering the orbit, with a lateral depression at the for the head and tail. fifth scale; infralabials 16/16; loreals 2/3; pre- Another specimen designated as paratype by oculars 1/1; postoculars 3/4; supraoculars 1/1; Hoser (2000) from near Wau, originally labeled one pair of parietals, not in median contact BPBM 5137 was donated to the AMNH (AMNH (presumably separated by interparietal, but dam- REVISION OF THE GENUS LEIOPYTHON HUBRECHT 1879 659 aged), left parietal meets uppermost postocular; District, CAS 118905 and CAS 118907. The latter elongate anterior temporals; dorsal midbody had an everted hemipenis. scales 54; ventrals 271; anal entire; subcaudals 68. Diagnosis.—Leiopython fredparkeri differs from Color in Life.—Brownish-black dorsally with a Leiopython albertisii and Leiopython biakensis in violet gleam, fading to lighter reddish-brown the absence of the whitish postocular spot, and 2 laterally, although it appears darker than in in higher infralabial counts (KW-test: x 1 5 2 Leiopython albertisii and to white ventrolaterally 28.22, P , 0.001 and x 1 5 5.06, P , 0.05). It and ventrally. The dorsal surface of the head is further differs from the former species by lower 2 slightly darker than the dorsum. The supra- and ventral counts (KW-test: x 1 5 7.58, P , 0.01). infralabials are white with the typical black This species is distinguishable from Leiopython markings on the anterior edge. The chin is bennettorum by loreal, prefrontal, and lower 2 white. average postocular scale counts (KW-test: x 1 Color in Preservative.—The body is greyish- 5 4.28, P , 0.05), from Leiopython huonensis by brown with a violet gleam fading to lighter the absence of the whitish postocular spot, and yellowish-brown laterally. The ventrals are very from the latter species and Leiopython hoserae in light yellowish. The dorsal surface of the head is having two pairs of parietals. black, supra- and infralabials are white with the Description of Holotype.—Total length about typical black marking on the anterior edge of 1,620 mm; 13/13 supralabials, first and second the labial scales. The chin is white to off-white. pitted, fifth and sixth entering the orbit, having Variation.—A second pair of prefrontals and a lateral depression; 17/17 infralabials with 8– higher loreal scale count was found in all speci- 13 pitted; single loreal and preocular; 3/3 mens examined as well as on a picture of a live postoculars; 1/1 subocular separating the sev- specimen from Wau. Three specimens (BPBM enth supralabial from the orbit; single pair of 3277, 3890 and CAS 139590) had two pairs of loreal prefrontals; two pairs of parietals, anterior ones scales. In AMNH R-142857, a small elongate scale larger and in median contact, but do not meet borders the posterior loreal and the anterior the postoculars, posterior ones elongate and prefrontal laterally on the right side, whereas the separated by an elongate interparietal at the posterior prefrontal scale on the left side is larger. median; two smaller interparietals bordering In BPBM 5452 (holotype of Leiopython bennettorum), anterior parietals posteriorly and the elongate threelorealscaleswerepresentoneachside.A interparietal anteriorly; dorsal midbody scale larger anterior loreal scale and two posterior rows 50; ventrals 266; single anal scale; sub- smaller scales border each other laterally. caudals 65, mostly paired. Distribution and Ecology.—Leiopython bennet- Description of the Paratypes.—CAS 118905: torum only occurs at Wau at an elevation total length about 1,800 mm, head length about between 1,050 and 1,400 m. The climate at 70 mm; 13/12 supralabials, first and second Wau is dry and only moderately seasonal. pitted, 6–7/5–6 entering the orbit, having a Annual rainfall is between 1,500 and 2,000 mm lateral depression; 17/16 infralabials with 8–13/ and relative humidity falls to 60% at midafter- 7–12 pitted; single loreal and preocular; 2/2 noon throughout the year (McAlpine et al., postoculars; 1/1 subocular separating the 8/7 1983). McPhee (1988) stated that at Wau peak supralabial from the orbit; single pair of breeding season in rodents is earlier than found prefrontals; two pairs of parietals, anterior ones elsewhere in PNG. This report is interesting, as larger and in median contact, but do not meet rodents are the main prey of Leiopython (Mc- the postoculars, posterior ones elongate and Dowell, 1975; Shine and Slip, 1990; Barker and separated by an elongate interparietal at the Barker, 1994). According to Shine and Madsen median; two smaller interparietals bordering (1997), there is a causal connection between anterior parietals posteriorly and the elongate prey abundance and reproductive output of interparietal anteriorly; maximum dorsal scale Liasis fuscus Peters 1873. This might lead to the rows 49; ventrals 266; single anal scale; sub- assumption that Leiopython bennettorum has a caudals 70, mostly paired. shift in timing of reproduction, but this CAS 118907: total length about 1,690 mm, assumption is speculative and further research head about 58 mm; 12/13 supralabials (12th on the ecology of this species is needed. and 13th fused on the left side), first to third pitted, fifth to seventh entering the orbit, having Leiopython fredparkeri sp. nov. a lateral depression at fifth and sixth; 17/17 Figure 6G, H infralabials with 8–13/8–13 pitted; single loreal Holotype.—A large male (everted hemipenes) and preocular; 3/3 postoculars; two pairs of collected by F. Parker 7 July 1967 at Karimui, parietals, anterior ones larger and in median Chimbu District, CAS 118906. contact anteriorly, but posteriorly divided by a Paratypes.—Two large males collected by F. small triangular interparietal that meets an Parker on 7 July 1967 at Karimui, Chimbu elongate interparietal anteriorly separating the 660 WULF D. SCHLEIP posterior parietals at the median; maximum Remarks.—Besides separation from other spe- dorsal midbody scale rows 49; ventrals 266; cies in morphometric space, this population is single anal scale; subcaudals 61, mostly paired. also geographically isolated. The morphologi- Color in Life.—Parker (1982) did not explicitly cal, geographical, and ecological factors seem to mention the coloration of live specimens from justify assigning species rank to this population, Karimui, but according to D. G. Barker (pers. but this assignment should be subject to future comm.), these specimens are very dark in color. studies on a genetic basis. Color in Preservative.—Body and head are rich Leiopython huonensis sp. nov. brown or brownish with a violet gleam, the Figure 6I, J head was only slightly darker than the body. Body color fades to yellowish ventrolaterally Holotype.—An adult female from an area and yellowish ventrally. Only one specimen about 16 km west of Lae, collected by L. van from Karimui (AMNH R-98861) had a black Royen in 1964, AMNH R-95535. Labeled: ‘‘Gift dorsal surface of the head. to H. M. van Deusen’’; SVL 940 mm; tail Variation.—In three specimens (AMNH R- 155 mm; head 37 mm. 98861, CAS 118908, and MCZ R-140801), the Paratype.—A semiadult male collected by H. posterior parietal scales were fused at the M. van Deusen and S. O. Grierson (7th median line. Additionally, two of these speci- Archibold Expedition) on 20 September 1964 mens (AMNH R-98861 and CAS 118908) also at Finschhafen, Seboagisung cave on tunnel had enlarged temporals, laterally bordering ceiling at an altitude of about 152 m, labeled: both, the anterior and posterior parietal scales. ‘‘Bl[ack].H[ea]d[ed] Python Huon Peninsula Three specimens (23%, N 5 13; CAS 118905, Finschhafen ca. 152 m’’ stored at the AMNH 118906 and 118909) had only two supralabials R-95532. SVL 666 mm; tail 110 mm; head entering the eye, but with a subocular separat- 28 mm. ing the eighth supralabial from the eye. Further Diagnosis.—This species differs from Leiopy- three specimens MCZ R-130464, MCZ R-140802, thon albertisii, Leiopython biakensis, and Leiopy- and MCZ R-140804 with only two supralabials thon fredparkeri in having only one pair of parie- entering the eye on one side. A specimen from tals followed by small, irregular scales (Fig. 2C, 6I). It can be distinguished from Leiopython the Bomai Tive Plateau (MCZ R-130464) bennettorum by the lower number of loreal and showed two loreal scales on one side, and one prefrontal scales as well as a lower average very large loreal on the other side. Additionally, number of postoculars (KW-test: x2 5 3.95, P , a small scale laterally bordering the prefrontal 1 0.05) and from Leiopython hoserae by the scale and the two loreal scales were present. arrangement posterior to the parietal scales Distribution and Ecology.—Little is known showing small, irregular scales. It also differs about the distributional extent of this species 2 from Leiopython albertisii (KW-test: x 1 5 10.31, in the Simbu Province, and specimens were P , 0.001) and from Leiopython hoserae (KW-test: hitherto only collected from the Tive Plateau, 2 x 1 5 5.35, P , 0.05) by lower average ventral Bomai, at an altitude of 1,100 m and from scale counts. Additionally, it can be distin- ‘‘Karimui’’ at an elevation up to 1,500 m. guished from the latter taxon by the presence Although no precise data of the collecting of the whitish postocular spot. locality are available, it may be assumed that Description of the Holotype.—Supralabials 12/ this species was collected from the Karimui 12, first two pitted, fifth to seventh entering the Basin. This Basin forms a plateau at an elevation orbit with a lateral depression on the fifth scale; between 1,000 and 1,500 m, surrounded by a infralabials 16/16, with 7–13 pitted; loreals 1/1; ring of mountains, which isolates the more postoculars 3/3 with whitish spot; supraoculars lowland fauna of the basin from others and is 1/1; one pair of prefrontals; one pair of parietals home of several endemic bird species (Dia- in median contact and meet uppermost post- mond, 1967, 1972). The climate of this area is oculars; anterior temporals 5/4; posterior tem- moderately seasonal (McAlpine et al., 1983) porals 5/5; midbody scale rows 48; ventrals 269; with mean annual rainfall of about 3,330 mm anal entire; subcaudals 73. (Diamond, 1972). According to McAlpine et al. Description of the Paratype.—AMNH R-95532: (1983), the mean maximum daily temperature at supralabials 12/12, first two pitted, fifth to this elevation (1,000–1,500 m) is about 25uC. The seventh entering the orbit with a lateral depres- habitat consists of evergreen forests, grasslands, sion at the fifth scale; infralabials 15/15, with 8– and palm swamps. 12 pitted; loreals 1/1; postoculars 3/3 with Etymology.—This species is named in honor of whitish spot; supraoculars 2/1 (left divided into the collector of these and many other speci- two scales); one pair of prefrontals; one pair of mens, the herpetologist Fred Parker, Towns- parietals in median contact but do not meet ville, Australia. uppermost postoculars; anterior temporals 4/4; REVISION OF THE GENUS LEIOPYTHON HUBRECHT 1879 661 posterior temporals: 5/5; dorsal midbody scale Leiopython biakensis sp. nov. rows 49; ventrals 282; anal entire; subcaudals 78. Holotype.—A large female specimen at the Color in Preservative.—Specimens in preserva- Museum of Natural History, Leiden, The tion are light brownish dorsally only slightly Netherlands, RMNH 10193, collected at Biak, fading in color ventrolaterally but yellowish Schouten Islands, 1952–1953, donated by Fleet ventrally. The dorsal surface of the head is Air Arm Royal Netherlands Navy. black; the supra and infralabials are off-white Paratype.—A large male specimen, RMNH with the typical black markings. The chin is off- 10194, collected at the Oregon Trail, Biak, white as well. Schouten Islands by Legion Quartermaster Distribution and Ecology.—Leiopython huonensis Talens, Royal Netherlands Navy; snout–vent occurs at almost sea level around Finschhafen length 1,550 mm, tail 220 mm. and Lae at the southern coast of the Huon Diagnosis.—This species differs from Leiopy- Peninsula. One specimen (MCZ R-145252) was thon albertisii in having only two labials entering found east of the Ramu River at Dumpu in the the orbit (100%, N 5 2) and in lower ventral Morobe Province. The Finesterre Mountains to scale counts (271 6 1.41; range 5 270–272, N 5 the north and the Markham and Ramu Rivers to 2) than found in specimens from the western the west and southwest geographically separate part of Papua (278 6 2.49; range 5 274–283, N 5 this species from others, although, these rivers 13; KW-test: x2 5 4.99, P , 0.05; see also might not be insuperable barriers. Colgan et al. 1 Brongersma, 1956). However, ventral scales (1993) pointed out that the peninsula is a zoogeographic barrier in lowland mammals counts are in the range of specimens from and Rawlins and Donnellan (2003) considered PNG. It differs from Leiopython fredparkeri in the it as ‘‘barrier to gene flow.’’ As the Huon head scale arrangement, with this species Peninsula originated from an oceanic island arc having large and wide posterior parietals, in late Pliocene (Zweifel, 1980; Norris and whereas Leiopython fredparkeri shows two slen- Musser, 2001), a high level of endemism could der elongate posterior parietal scales (Fig. 2). It be expected in specimens. Nevertheless, Mc- further differs from the latter species by the Dowell (1984) did not find endemism in snakes presence of whitish postocular spots, and addi- at the Huon Peninsula. tionally from Leiopython huonensis and Leiopy- McAlpine et al. (1983) stated that the Huon thon hoserae in having two pairs of parietals. It is Peninsula area is highly seasonal showing a also distinguishable from Leiopython bennettorum ‘‘reverse’’ seasonality in having a wet season in the number of prefrontals and loreals. between May and August, rather than between Description of Holotype.—Total length 1,670 December and April, as found in most other mm, tail 255 mm; supralabials 12/11, first to parts of the country. Annual rainfall exceeds third pitted, fifth and sixth entering the orbit; 4400 mm (McAlpine et al., 1983; Macfarlane, infralabials 14/15 with 6–12/7–13 pitted; single 1999). Recent studies (e.g., Python reticulatus: loreal and 2/2 preoculars, the upper large, the Shine et al., 1998, 1999; Liasis fuscus: Brown and lower very small triangular shield that is Shine, 2002; Brown et al., 2002; Shine, 2003) have wedged between labials and the large preocu- revealed that tropical snakes can reproduce lar; single loreal and supraocular; two pairs of seasonally. This is strictly true in captive held large parietals both in median contact, a small specimens of Leiopython albertisii and Leiopython rhombic interparietal is wedged between ante- hoserae in the northern hemisphere (pers. obs.). rior and posterior parietals (for parietal struc- This evidence may lead to the conclusion that ture, see also Brongersma, 1956: fig. 1a), left Leiopython huonensis is subject to a dramatic shift posterior parietal divided into two scales (see in timing of reproduction, because of prey Brongersma, 1956: fig. 1a); whitish spot behind availability and ‘‘cost’’ of reproduction (Shine, the eye; maximum dorsal scale rows 45, ventrals 2003) as found in Liasis fuscus Peters from 272, subcaudals 70 mostly paired (Brongersma, Queensland and the Northern Territory (Shine 1956). and Slip, 1990), but this assumption is highly Description of Paratype.—Single preocular on speculative and further studies are needed. each side; 12 supralabials, only fifth and sixth Etymology.—Named after the locality where entering the orbit, first to third pitted; infra- this species is found, the Huon Peninsula at the labials 16/16, with seventh to 13th pitted; east coast of PNG. whitish spot behind the eye; midbody scale Remarks.—Specific rank was assigned to this rows 47, ventrals 270, anal entire, subcaudals 65. population because of diagnosable morpholog- Color in Preservative.—The color of the dorsum ical differences from other species, further is light purplish-brown fading to yellowish because of different ecological and environmen- laterally and pale yellow ventrally. The color tal conditions and an assumed allopatric distri- of the dorsal surface of the head is purplish- bution. black. 662 WULF D. SCHLEIP

Distribution.—This species is endemic to the midbody rows, 171 ventrals, 73 subcaudals; island of Biak, Indonesia. small pale spot on each preocular and on right Etymology.—Named after the island where uppermost postocular (contra McDowell, 1975), this species was found. but not as clearly as found in New Guinea Remarks.—This allopatric population shows specimens of Leiopython albertisii; large spurs on little, but diagnosable morphological, differenc- the left and right side of the cloacae. es to other species. Brongersma (1956) assumed Color in Preservative.—Body color is brownish- this population to form an incipient race. violet fading to yellowish ventrolaterally and to Because of the geographic distance to the off-white ventrally. The dorsal surface of the mainland populations of Leiopython albertisii,it head is black, the supra- and infralabials are is unlikely that gene flow occurs among these white with the typical black marking, and the populations. Hence, this population is consid- chin is off-white. ered reproductively isolated (sensu Wiens, Distribution.—According to Hoser (2000), this 2004), and, in accordance with Frost and Hillis species is endemic to Mussau Island, St. (1990) and based on the ESC (sensu Frost and Matthias Group, Bismarck Archipelago, PNG. Kluge, 1994), the assignment of specific rank to Remarks.—The original description ‘‘Leiopy- this population seems justified. Unfortunately, thon albertisii barkeri’’ does not meet the require- no adequate photograph was available from this ments of the code (article 13.1.1; ICZN, 2000). species. Hoser (2000) did not specify characters that are able to differentiate this taxon from others. In ‘‘Leiopython albertisii barkeri’’ Hoser 2000 addition, the name would have required emen- nomen nudum dation because of naming it after two persons Figure 6K, L (see also Wu¨ ster et al., 2001). Hoser (2000) states Holotype.—A semi adult female from Boliu, that this taxon is distinguishable from Leiopy- Mussau Island, St. Matthias group, Bismarck thon albertisii by allopatry, by analysis of Archipelago collected by the Danish Noona Dan mitochondrial DNA, and by lower ventral Expedition from 16 January to 18 February 1962. counts, although in the range of Leiopython Stored at the ZMUC, collection number R-5444. albertisii. Yet, Hoser (2000) had failed to provide Total length of the specimen was 1,020 mm evidence for these statements. ‘‘Analysis of (head length 50 mm; SVL 950 mm tail 170 mm). mitochondrial DNA’’ (Hoser, 2000:19) is not a Paratypes.—A semi adult male from Boliu, character as stipulated by article 13.1.1 of the Mussau Island, St. Matthias group, Bismarck International Code of Zoological Nomenclature Archipelago collected by the Danish Noona Dan (ICZN, 2000), but simply the proposed use of a Expedition during 16 January and 18 February method. Nowhere does Hoser (2000) provide 1962. Stored at the ZMUC, collection number R- evidence of mtDNA differentiation of the 5445. Total length of the specimen was 910 mm Mussau population. Allopatric distribution it- (SVL 770 mm + tail 140 mm). self may separate the Mussau Island population Redescription of the Holotype.—Supralabials geographically, but it is highly questionable if 13/13, first two pitted, fifth to seventh entering this alone is able to distinguish a taxon from the orbit with lateral depression at the fifth another, regardless of the underlying species scale; infralabials 15/16, with 7–12 pitted; concept. Nonetheless, multivariate morphomet- loreals 1/1; supraoculars 1/1; postoculars 3/4; ric analysis was not able to separate this one pair of prefrontals; preoculars 1/1 (+ a very population from Leiopython albertisii in morpho- small lower one), two pairs of parietals, anterior metric space (see Fig. 3), and therefore, it is ones in medial contact but do not meet considered conspecific with this taxon. More- uppermost postoculars; 3 interparietals; anterior over, because of the inadequate diagnosis, the temporals 5/5; posterior temporals: 4/3; 44 name is considered unavailable and is regarded midbody scale rows, 267 ventrals, 72 subcau- as nomen nudum. The description and pictures dals. are herein provided pro forma for future work Redescription of the Paratypes.—Supralabials on the genus. 13/13, first two pitted, fifth to seventh entering the orbit with lateral depression at the fifth scale; infralabials 16/16, with seventh to 12th DISCUSSION pitted; loreals 1/1; supraoculars 1/1; postocu- The results of this study support the hypoth- lars 3/4; one pair of prefrontals; preoculars 1/2 esis that the genus Leiopython comprises several (a second very small lower one), two pairs of taxa. Apart from the taxon albertisii, morpho- parietals (contra McDowell, 1975), anterior ones logical analysis uncovered another two new in medial contact but do not meet uppermost species from the mainland (Leiopython huonensis postoculars; single interparietal; anterior tem- sp. nov. and Leiopython fredparkeri sp. nov.) and porals 5/5; posterior temporals: 5/4; 45 dorsal one new island species (Leiopython biakensis sp. REVISION OF THE GENUS LEIOPYTHON HUBRECHT 1879 663 nov.) from Biak Island. The recently recognized This study has shown that the genus Leiopy- species Leiopython hoserae and Leiopython bennet- thon is more complex than previously assumed, torum (Hoser, 2000) could be verified. Addi- but little is known about the biology and tional evidence from molecular data also sup- ecology of White-Lipped Pythons. Further ports the separation of Leiopython hoserae from collecting is necessary to determine the distri- Leiopython albertisii, and with a genetic distance butional extent of Leiopython albertisii in the of about 10% between both species, both are Yapen-Waropen and Nabire regencies, Papua clearly distinct from each other. However, and of Leiopython hoserae at the southern coast of specimens found in the Enga Province appear Papua, westwards to the Fak Fak regency. As to be morphologically similar to Leiopython indicated in this study, future research on a huonensis sp. nov., although of remote distribu- genetic basis might reveal the existence of even tion (Fig. 5), but this population is herein not more species. recognized taxonomically, because of the small sample size and inadequate data. The two Acknowledgments.—I am deeply indebted to island populations from the Mussau and W. Bo¨hme (ZFMK) for signing loan requests Emirau islands were morphologically indistin- and for giving me the opportunity to examined guishable from Leiopython albertisii and, there- material in his lab, and to W. Wu¨ ster (UWB) fore, are considered conspecific with this spe- and B. I. Crother (SELU) for their assistance and cies. Yet, their actual taxonomic status remains helpful comments on several versions of the unclear. According to De Boer and Duffels manuscript that helped to improve the scientific (1996), the Bismarck Archipelago, which com- quality. I would like to thank M. O’Shea and D. prises these two islands, originated from the G. Barker (VPI) for readily sharing their great same island arc as central New Guinea and the knowledge on the genus. I am very grateful to Huon peninsula. Therefore, and because of the the curators and staff who sent specimens for distance to the mainland of PNG, it can be loan, namely C. Kishinami (BPBM), J. Vindum assumed that these populations are either part (CAS), G. Ko¨hler and M. Laudahn (SMF), and of a relict fauna (McDowell, 1975) that have not M. Andersen (ZMUC) and to the curators and diverged morphologically from the mainland staff who kindly invited me to examine material populations (sibling species) or were introduced in their holdings and to use their labs, namely L. by humans. Future studies on a genetic basis Ford (formerly AMNH), N. Gilmore (ANSP), C. might clear this ambiguity. McCarthy (BMNH), R. McDiarmid and S. Gotte Besides morphological differentiation, the (USNM), J. Rosado (MCZ), and M. S. Hoog- species recognized are geographically isolated moed (formerly RMNH, now retired). I would from each other by either geographical barriers further like to thank M. J. Mahoney (formerly (e.g., mountain ranges and rivers) or altitude. MVZ) and P. Ustach (formerly UTA) for taking Therefore, the taxonomic arrangement present- pictures and counting scales, and to A. Kluge ed is in accordance with the ESC (sensu Frost and G. Schneider (UMICH) as well as to J. B. and Kluge, 1994) in that these populations are Ladonski (FMNH) for sending pictures of diagnosable allopatric populations and, hence, specimens in their holdings for the comparison are considered separate lineages. According to study. I would further like to thank D. Dittmann Wiens (2004), allopatric populations do not (LSUMZ) for tracking back mtDNA sequences necessarily need to have undergone speciation from the GenBank to specimens, H. Cogger for mechanisms other than geographical isolation verifying locality data, resulting from his itself. Wiens (2004) argues that strict allopatry participation in the alpha-Helix expedition can be considered reproductive isolation be- (1969), and D. J. Williams and C. Pook for their cause of a lack of gene flow among populations. support. I would also like to thank the associate However, it would be impractical to consider editor and the two anonymous reviewers for populations as separate lineages without evi- their excellent comments that helped to greatly dence for their differentiation in some way improve the manuscript. I am thankful to all of (Frost and Hillis, 1990; Frost and Kluge, 1994; these people for supporting me, an amateur Wiens, 2004). Finally, I agree with Frost and herpetologist. Hillis (1990) for the given reasons, that diag- nosable allopatric populations should be con- LITERATURE CITED sidered as species rather than as subspecies. The two species Leiopython hoserae and Leiopy- ALLISON, A. 2007. Introduction to the fauna of Papua. In A. J. Marshall and B. M. Beehler (eds.), The thon fredparkeri sp. nov. exceed other species of Ecology of Papua. Vol. 1, pp. 479–494. The Ecology the genus in average body size (McDowell, of Indonesia Series. Periplus Editions, Singapore. 1975; Parker, 1982; Barker and Barker, 1994; AUSTIN, C. C. 2000. Molecular phylogeny and historical O’Shea, 1996) and egg-size in clutches (Parker, biogeography of Pacific island boas (Candoia). 1982; Barker and Barker, 1994). Copeia 2000:341–352. 664 WULF D. SCHLEIP

———. 2006. Checklist and comments on the terres- EHMANN, H. 1992. The Encyclopedia of Australian trial reptile fauna of Kau wildlife area, Papua New Animals: Reptiles. National Photographic Index of Guinea. Herpetological Review 37:167–170. Australian Wildlife. Angus and Robertson, Syd- BARKER,D.G.,AND T. M. BARKER. 1994. Leiopy- ney, New South Wales, Australia. thon albertisii. In D. G. Barker and T. M. Barker FORSMAN, A. 1991. Adaptive variation in head size in (eds.), Pythons of the World, Vol. 1, Australia, Vipera berus L. populations. Biological Journal of pp. 24–26. Advanced Vivarium Systems Inc., the Linnean Society 43:281–296. Lakeside, CA. FROST, D. R., AND D. M. HILLIS. 1990. Species in concept BOULENGER, G. A. 1893. Catalogue of the snakes in the and practice: Herpetological applications. Herpe- British Museum (Natural History). Vol. I. British tologica 46:87–104. Museum (Natural History), London. FROST, D. R., AND A. G. KLUGE. 1994. A Consideration of ———. 1898. An account of the reptiles and batrachi- epistemology in systematic biology with special ans collected by Dr. L. Loria in British New reference to species. Cladistics 10:259–294. Guinea. Annali del Museo Civico de Storia GOWER, J. C. 1966. Some distance properties of latent Naturale di Genova (2) 18:694–710. root and vector methods used in multivariate BRONGERSMA, L. D. 1953. Notes on New Guinea reptiles analysis. Biometrika 53:325–338. and amphibians II. Proceedings of the Koninklijke HAAS DE, C. P. J. 1950. Checklist of the snakes of the Nederlandsche Akademie van Wetenschappen, Indo-Australian archipelago (Reptiles: Ophidia). Amsterdam (C) 56:317–325. Treubia 20:511–625. ———. 1956. Notes on New Guinea reptiles and HARVEY, M. B., D. G. BARKER,L.K.AMMERMAN, AND P. T. amphibians IV. Proceedings of the Koninklijke CHIPPINDALE. 2000. Systematics of pythons of the Nederlandsche Akademie van Wetenschappen, Morelia amethistina complex (Serpentes: Boidae) Amsterdam (C) 59:599–610. with the description of three new species. Herpe- BROWN, G. P., AND R. SHINE. 2002. Influence of weather tological Monographs 14:139–185. conditions on activity of tropical snakes. Austral HEADS, M. 2001. Birds of paradise, biogeography and Ecology 27:596–605. ecology in New Guinea: a review. Journal of BROWN, G. P., R. SHINE, AND T. MADSEN. 2002. Responses Biogeography 28:893–925. of three sympatric snake species to tropical ———. 2002. Birds of paradise, vicariance biogeogra- seasonality in northern Australia. Journal of phy and terrane tectonics in New Guinea. Journal Tropical Ecology 18:549–568. of Biogeography 29:261–284. CAMPBELL, B. N. 1997. Hic Sunt Serpentes—Molecular HOSER, R. T. 2000. A revision of the Australasian Phylogenetics and the Boidae (Serpentes: Booidea). Pythons. Ophidia Review 1:7–27. Unpubl. Ph.D. thesis, Queen’s University, King- HUBRECHT, A. A. W. 1879. Notes III. On a new genus ston, Ontario, Canada. and species of Pythonidae from Salawatti. Notes COGGER, H. G., E. E. CAMERON, AND H. M. COGGER. 1983. from the Leyden Museum 14–15. Zoological catalogue of Australia. Vol. 1. Amphib- INTERNATIONAL COMMISSION ON ZOOLOGICAL NOMENCLA- ia and Reptilia. Bureau of Flora and Fauna TURE. 2000. International Code of Zoological No- Canberra. Australian Government Publishing Ser- menclature, 4th ed. International Trust for Zoolog- vice, Canberra, Australia Capital Territory, Aus- ical Nomenclature, Natural History Museum, tralia. London. COLGAN, D., T. F. FLANNERY,J.TRIMBLE, AND K. ALPIN. KEND, B. A. 1997. Leiopython albertisii. In G. W. Schuett 1993. Electrophoretic and morphological analysis and L. W. Schuett (eds.), Pythons of Australia, of the systematics of the Phalanger orientalis pp. 151–152. Canyonlands Publishing Group, (Marsupiaalia) species complex in Papua New Provo, UT. Guinea and the Solomon Islands. Australian KLUGE, A. G. 1993. Aspidites and the phylogeny of the Journal of Zoology 41:355–378. Pythoninae snakes. Records of the Australian CRISP, M. D., S. LAFFAN,H.P.LINDER, AND A. MONRO. Museum Supplement 19:1–77. 2001. Endemism in the Australian flora. Journal of KOOPMAN,K.F.1982.ResultsoftheArchibold Biogeography 28:183–198. expedition No. 109. Bats from eastern Papua and DE BOER, A. J., AND J. P. DUFFELS. 1996. Historical the east Papua Islands. American Museum Novi- biogeography of the cicadas of Wallacea, New tatis 2747:1–34. Guinea and the West Pacific: a geotectonic expla- KOVACH, W. L. 1999. MVSP—A multivariate statistical nation. Palaeogeography, Palaeoclimatology, Pa- package for Windows, Vers. 3.1. Kovach Comput- laeoecology 124:153–177. ing Services, Pentraeth, Wales. DIAMOND, J. M. 1967. New subspecies and records of KRAUS, F., AND A. ALLISON. 2002. A new species of birds from the Karimui basin, New Guinea. Xenobatrachus (Anura: Microhylidae) from north- American Museum Novitates 2284:1–17. ern Papua New Guinea. Herpetologica 58:56– ———. 1972. Avifauna of the Karimui basin. In 66. R. A. Paynter, Jr. (ed.), Avifauna of the Eastern KUMAR, S., K. TAMURA, AND M. NEI. 2004. MEGA3: Highlands of New Guinea, pp. 53–56. Nuttall Integrated software for molecular evolutionary Ornithological Club, Harvard University, Cam- genetics analysis and sequence alignment. Brief- bridge, MA. ings in Bioinformatics 5:150–163. DOWLING, H. G. 1951. A proposed standard system of LEVITON, A. E., R. H. GIBBS JR., E. HEAL, AND C. E. counting ventrals in snakes. British Journal of DAWSON. 1985. Standards in herpetology and Herpetology 1:97–99. ichthyology. Part I. Standard symbolic codes for REVISION OF THE GENUS LEIOPYTHON HUBRECHT 1879 665

institutional resource collections in Herpetology POSADA, D., AND K. A. CRANDALL. 1998. Modeltest: and Ichthyology. Copeia 1985:802–832. testing the model of DNA substitution. Bioinfor- LOVERIDGE,A.1948.NewGuineanreptilesand matics 14:817–818. amphibians in the Museum of Comparative PRENTICE, M. L., AND G. S. HOPE. 2007. Climate of Zoology and United States National Museum. Papua. In A. J. Marshall and B. M. Beehler (eds.), Bulletin of the Museum of Comparative Zoology, The Ecology of Papua. Vol. 1, pp. 479–494. Ecology Harvard 101:305–430. of Indonesia Series. Periplus Editions, Singapore. MADSEN, T., AND R. SHINE. 2000. Silver spoons and RAWLINGS, L. H., AND S. C. DONNELLAN. 2003. Phylogeo- snake body sizes: prey availability early in life graphic analysis of the Green Python, Morelia influences long-term growth rates of free-ranging viridis, reveals cryptic diversity. Molecular Phylo- pythons. Journal of Animal Ecology 69:952–958. genetics and Evolution 27:36–44. MACFARLANE, D. 1999. Country pasture profiles— RAWLINGS, L. H., D. G. BARKER, AND S. C. DONNELLAN. Papua New Guinea. Food and Agriculture 2004. Phylogenetic relationships of the Australo- Organization of the United Nations (FAO). Cited Papuan Liasis pythons (Reptilia: Macrostomata), date. http://www.fao.org/AG/Agp/agpc/doc/ based on mitochondrial DNA. Australian Journal Counprof/southpacific/png.htm (March 2008). of Zoology 52:215–227. MCALPINE, J. R., G. KEIG, AND R. FALLS III. 1983. Climate RAWLINGS, L. H., RABOSKY, D. L., DONNELLAN, S. C., AND of Papua New Guinea, CSIRO. Australian National M. N. HUTCHINSON. 2008. Python phylogenetics: University Press, Canberra, Australia Capital inference from morphology and mitochondrial Territory, Australia. DNA. Biological Journal of the Linnean Society MCDIARMID, R. W., J. A. CAMPBELL, AND T. A. TOURE´ . 93:603–619. 1999. Leiopython albertisii. In R. W. McDiarmid, J. A. SHINE, R. 2003. Reproductive strategies in snakes. Campbell, and T. A. Toure´ (eds.), Snake Species of Proceedings Biological Sciences of the Royal the World. A Taxonomic and Geographic Refer- Society 270:995–1004. ence Vol. 1, pp. 166–167. Herpetologist’s League, SHINE, R., AND T. MADSEN. 1997. Prey abundance and Washington, DC. predator reproduction: rats and pythons on a tropical Australian floodplain. Ecology 78:1078– MCDOWELL, S. B. 1975. A catalogue of the snakes of New Guinea and the Solomon’s, with special 1086. referencetothoseintheBerniceP.Bishop SHINE, R., AND D. J. SLIP. 1990. Biological aspects of the museum. Part II. Journal of Herpetology 9:1–79. adaptive radiation of Australasian pythons (Ser- pentes: Boidae). Herpetologica 46:283–290. ———. 1984. The Snakes of the Huon peninsula, Papua New Guinea. American Museum Novitates SHINE, R., P. S. HARLOW,J.S.KEOGH, AND BOEADI. 1998. The allometry of life-history traits: insights from a 2775:1–28. study of giant snakes (Python reticulatus). Journal MCPHEE, E. C. 1988. Ecology and diet of some rodents of Zoology (London) 244:405–414. from the lower montane region of Papua New SHINE, R., AMBARIYANTO,P.S.HARLOW, AND MUMPUNI. Guinea. Australian Wildlife Research 15:91–102. 1999. Reticulated pythons in Sumatra: biology, NORRIS, C. A., AND G. G. MUSSER. 2001. Systematic harvesting and sustainability. Biological Conser- revision within the Phalanger orientalis complex vation 87:349–357. (Diprotodontia, Phalangeridae): a third species of STIMSON, A. F., AND S. B. MCDOWELL. 1986. Liasis Gray, lowland gray cuscus from New Guinea and 1842 (Reptilia, Serpentes): proposed designation of Australia. American Museum Novitates 3356:1–20. Liasis mackloti Dumeril and Bibron 1844 as type O’SHEA, M. 1996. Snakes of Papua New Guinea. species Z.N.(S)2439. Bulletin of Zoological Nomen- Independent Publishing, Port Moresby, Papua clature 43:330–334. New Guinea. STULL, O. G. 1935. A checklist of the family Boidae. ———. 2007. Australasia and the Pacific Ocean Proceedings Boston Society of Natural History islands: White-Lipped Pythons. In M. O’Shea 40:387–408. (ed.), Boas and Pythons of the World, pp. 125–128. UNDERWOOD, G., AND A. F. STIMSON. 1990. A classifica- Princeton University Press, Princeton, NJ. tion of pythons (Serpentes, Pythoninae). Journal of O’SHEA, M., SPARKLAND, R. G., AND I. H. BIGILALE. 2004. Zoology London 221:565–603. First record for the genus Antaresia (Squamata: WHITTIER, J., C. MACROKANIS, AND R. T. MASON. 2000. Pythonidae) from Papua New Guinea. Herpeto- Morphology of the Brown Tree Snake, Boiga logical Review 35:225–227. irregularis, with a comparison of native and PAIJMANS, K. 1976. Part II. Vegetation. In K. Paijmans extralimital populations. Australian Journal of (ed.), New Guinea Vegetation, pp. 23–105. CSIRO, Zoology 48:357–367. Australian National University Press, Canberra, WIENS, J. J. 2004. What is speciation and how should Australia Capital Territory, Australia. we study it? American Naturalist 163:914–923. PARKER, F. 1982. The snakes of the western province. WILLIAMS, D. J., W. WU¨ STER, AND B. G. FRY. 2006. The Wildlife in Papua New Guinea No. 82/1. Depart- good, the bad and the ugly: Australian snake ment of Lands and Environment, Konedobu, taxonomists and a history of the taxonomy of Papua New Guinea. Australia’s venomous snakes. Toxicon 48:919–930. PETERS, W., AND G. DORIA. 1878. Catalogo dei retilli e WORRELL,E.1961.Herpetologicalnamechanges. dei batraci raccolti da O. Beccari, L. M. D’Alberts e Western Australian Naturalist 8:18–27. A. A. Bruijn. nella sotto-regione Austro-Malese. WU¨ STER, W., B. BUSH,J.S.KEOGH, M. O’SHEA, AND R. Annali del Museo Civico de Storia Naturale di SHINE. 2001. Taxonomic contributions in the ‘‘am- Genova ser. 1:401–403, plate III, Fig. 2. ateur’’ literature: comments on recent descriptions 666 WULF D. SCHLEIP

of new genera and species by Raymond Hoser. ZWEIFEL, R. G. 1980. Frogs and lizards from the Huon Litteratura Serpentium 21:67–79, 86–91. peninsula, Papua New Guinea. Bulletin of the YOSHIOKA, K. 2002. KyPlot—a user-orientated tool for American Museum of Natural History 165:390–434. statistical data analysis and visualization, Compu- tational Statistics 17:425–451. Accepted: 31 March 2008.

APPENDIX 1 SOUTH1, (N 5 18): Central Province, PNG: Boroko: Material Examined.—Specimens are organized in USNM 195611 (19 miles north of Brown River Road, OTU’s as included given in Table 2. Only photo- near ford); USNM 195609 (Brown River Road, at ford graphs were available from specimens marked with *. approximately 2 miles south of Brown River); Madew: Photographs and scale counts and measurements BMNH 1908.10.14.6 (St. Joseph River, Brit. New provided by curatorial staff were available for Guinea); Port Moresby: MCZ R-145941*, MCZ R- specimens marked with **. 150792*, USNM 213396–213397 (Mc Donald’s Corner); NORTH1, (N 5 13): Salawati Island: ANSP 25095, near Port Moresby: AMNH R-103636 (Hiknumu RMNH 4796 (holotype Leiopython gracilis Hubrecht Plantation, 3 miles north 22 m east of Port Moresby, 1879, also see Brongersma, 1953); Sorong: ANSP 520 m), CAS 118910 (Laloki River); Kokoda (highlands 26096, RMNH 9653 (mainland, see also Brongersma of the Owen Stanley Range): MCZ R-84333; Middle- 1953); Fak Fak: RMNH 9654 1–3 (three specimens, see town: MCZ R-59091; USNM 213398–213399 (Nazareth also Brongersma 1953); Manokwari: RMNH 9655 1–2 Mission, on road to, adjacent to Laloki River); Sogeri: (two specimens, see also Brongersma 1953), MSNG CAS 118939, MCZ R-149696*; Milne Bay Province, 29990* (syntype Leiopython albertisii, Peters and Doria Alotau: road-kill specimen** (without collection num- 1878: fig 2a); without exact locality data (from pet ber); Popondetta: MCZ R140775 (head and tail only); trade) ZFMK 20331–20332, SFM 72616. Other material: BMNH 97.12.10.108 (labeled ‘‘British NORTH2, (N 5 10): Eastern Jayapura: Joka: BPBM New Guinea’’). 2305 and RMNH 10192 (Lake Sentani, female); Sanduan SOUTH2, (N 5 4): Western Province: Abam: USNM (former West Sepik) Province: Mt. Somoro AMNH R- 195754; Boze: BMNH 1986.1174 (Binaturi River, 1000006 (7 miles east of Lumi, 2400–4650 ft, head and Southern Trans-Fly region); Wipim: AMNH R- half of body), Aitape: MCZ R-48614; East Sepik 107150 (holotype of Leiopython hoserae Hoser 2000); Province: Miliom: AMNH R-100005; Wewak: AMNH Oslobip: MCZ R-129393. R-75027, R-75028; Madang Province: Errima: BMNH HUONP, (N 5 15): Lae: AMNH R-66756, R-95534– 1922.11.24.30 (Astrolabe Bay); Ramu River Delta: R-95535, R-103869, R-107148, R-115056–R-115057, BMNH 1926.5.31.4 (female); Madang: AMNH R-107149. MCZ R-145942–R-145944; Finschhafen: AMNH R- ENGA, (N 5 2): Wabag: CAS 103384 (listed as 99532–R-99533, USNM 118950 (south of Langemak «Southern Highlands District»), CAS 135211 (Jimi Bay, inland about 1 mile from a point on the shore Region, Sau River, 50 miles northwest of Banz) listed midway between Nasing Alater on Bugaim River and as ‘‘Western Highlands District’’), 2,500 ft. Kawalansam on Buka Creek); Tigrain: BMNH WAU, (N 5 5): AMNH R-142857 (skin, former 1922.11.24.29 (Adler River, Huon Gulf); Ramu River BPBM 5137), BPBM 3277, 3890 Wau, (vicinity of) Big (near Dampu): MCZ R-145252. Wau Creek, 1,150 m; 5452 Wau (Bishop Museum Field BIAK, (N 5 2): RMNH 10193 (female) –10194 Station), 1,250 m; CAS 139590. (Oregon Trail, male, head damaged, see also Bron- BULOLO, (N 5 1): MVZ 74917**. gersma 1956). KARIM: (N 5 13), Simbu Province, PNG; Karimui: MUSS, (N 5 2): R5444–R5445 (also see McDowell, AMNH R-98861 (also see McDowell, 1975), CAS 1975). 118905–118909; MCZ R-115582, MCZ R-123872, MCZ EMIR, (N 5 3): MVZ 40847–40849**. R-140801–R140804 (labeled as Liasis fuscus, Karimui, UNKNOWN, (N 5 2): BMNH 97.12.10.108 (labeled 3500 ft); Bomai Tive Plateau: MCZ R-130464. ‘‘British New Guinea’’). REVISION OF THE GENUS LEIOPYTHON HUBRECHT 1879 667

APPENDIX 2. Mitochondrial DNA sequences used in this study. Lengths of base pairs (bp) are provided and references are given in the Literature Cited section. Locality data were not given in the GenBank records but were backtracked to the tissue samples in museum collections and verified by the authorities given in the references.

GenBank accession Name/Locality number Length (bp) Locality Reference Merauke AF241406 715 Assumed from Merauke, Harvey et al. (2000) Irian Jaya (D. G. Barker, pers. comm.) Western Province U69835 1114 Western Prov., PNG, Campbell, B.N. (Thesis, Mawatta; Alpha-Helix 1997) Expedition 1969, tissue LSUMZ 10505 Milne Bay EU179542 831 Gurney, Milne Bay Province, this study Province PNG Madang 1 U69836 642 Madang (not West New Campbell, B.N. (Thesis, Britain Prov., Rabaul, PNG). 1997), backtracked by Alpha- Helix Expedition Donna Dittmann, locality 1969. Maiwara at Man Road, data corrected by H. tissue LSUMZ 10626 Cogger. Madang 2 EU183230 836 Madang Province, PNG this study Outgroup AF241399 715 Morelia amethistina Harvey et al. (2000) Outgroup AF241404 715 Morelia amethistina Harvey et al. (2000)