Abstract

We describe a new species of kukri snake (Oligodon Fitzinger, 1826) from the limestone karst formations of Satun and Trang Provinces in southern Thailand. Phylogenetic analyses based on three mitochondrial DNA fragments (12S–16S ribosomal rRNA and cytochrome b) recover the new species within the Oligodon cinereus species complex, where it forms a deeply divergent yet poorly supported clade sister to Oligodon saiyok Sumontha et al., 2017 and another unnamed lineage currently referred to Oligodon cinereus (Günther, 1864) from southwest Myanmar. Morphologically, the new species is distinguished from all other members of the genus by the following combination of characters: ventral scales 189–193 with distinct lateral keeling; subcaudal scales 47–54, paired; anterior dorsal scale rows 17–19, with the reduction from 19 to 17 rows occurring above the 28^th–30^th ventral scale when present; maxillary teeth 8, blade-like and laterally compressed; dorsum olive–gray, plain; ventral surface white anteriorly, dark gray posteriorly; underside of tail dark gray, smeared with white. We briefly discuss the natural history and conservation status of this new species and provide observations of other kukri snakes inhabiting limestone karst habitats. Our study also incorporates genetic samples of four recently described Oligodon endemic to Thailand, all of which are recovered in the O. cinereus species complex. In agreement with previous studies, we demonstrate that species-level diversity within the O. cinereus species complex is underestimated, and additional sampling is necessary to revise this taxonomically challenging clade.

Biodiversity, molecular phylogenetics, Serpentes, Southeast Asia, systematics, taxonomy, Thai–Malay Peninsula

Oligodon Fitzinger, 1826 (family: Colubridae) is the second most speciose snake genus in the world, with the number of recognized species ranging from 87 to 90, depending on the author (Lee et al. 2023; Uetz et al. 2023; David et al. 2023). Commonly referred to as kukri snakes due to their blade-like dentition, Oligodon occur throughout tropical Asia and are especially diverse in the Indochinese region (Cambodia, Laos, Vietnam, and Thailand), where several new species have been described (Nguyen et al. 2017, 2022; Pauwels et al. 2017, 2021; Sumontha et al. 2017; Nguyen et al. 2020; David et al. 2022). Earlier authors had partitioned most Oligodon native to Indochina into two intrageneric groups based on differences in hemipenial morphology (Smith 1943; Leviton 1963). Members of the O. cinereus species group have hemipenes with a simple sulcus spermaticus and distinct ‘papillae’-like appendages visible on the retracted organ (henceforth termed ‘myoectases’, sensu Wagner 1975), whereas the O. cyclurus-taeniatus species group include members with longer hemipenes and a divided sulcus spermaticus. The presence of myoectases in the latter group varies, with species more closely related to Oligodon cyclurus (Cantor, 1839) lacking these structures and those to Oligodon taeniatus (Günther, 1861) retaining them (David et al. 2008a, 2008b; Green et al. 2010; Nguyen et al. 2020; Yushchenko et al. 2023b). Both cinereus and cyclurus-taeniatus groupings were originally conceived as phenetic classifications, though molecular phylogenies have so far recovered each as monophyletic (Green et al. 2010; Pyron et al. 2013; Nguyen et al. 2020; Yushchenko et al. 2023a).

In Thailand, four new Oligodon have been described within the past 10 years, and all are endemic to the country and have localized distributions (Pauwels et al. 2017, 2021; Sumontha et al. 2017). First, Pauwels et al. (2017) described Oligodon huahin based on a series of specimens collected near the outskirts of Kaeng Krachan National Park in Prachuap Khiri Khan Province. Then, Sumontha et al. (2017) described another species, O. saiyok based on two specimens found in limestone karst formations near Sai Yok National Park, Kanchanaburi Province. Most recently, Pauwels et al. (2021) described Oligodon phangan and Oligodon promsombuti, which were collected from two separate localities in Surat Thani Province. Three of these species (O. huahin, O. phangan and O. promsombuti) were putatively considered members of the Oligodon cyclurus-taeniatus species group based on their hemipenial morphology, which were bilobed in shape when partially everted. However, the assignment of O. phangan to this species group was considered tentative, as no male specimens were known at the time of its description. Likewise, the species group assignment of O. saiyok was not evaluated, since the hemipenis of the male holotype was left retracted inside the tail base, preventing an assessment of its shape and ornamentation. Thus, all four recently described species of Thai Oligodon were described exclusively from morphology, and their phylogenetic positions have remained untested, pending the acquisition of molecular data.

Two more studies (Yushchenko et al. 2023a; David et al. 2023) made taxonomic changes relevant to Oligodon in Thailand. Most pertain to Oligodon cinereus (Günther, 1864), a widespread species that contains multiple color morphs and junior synonyms across its range (Smith 1943; Green et al. 2010; David et al. 2011). Based on molecular data, Yushchenko et al. (2023a) found that O. cinereus was not monophyletic. They restricted ‘true’ O. cinereus to populations in southern Indochina (Cambodia, Laos, and Vietnam) and transferred the recently described Oligodon cattienensis Vassilieva et al., 2013 to its synonymy. In addition, Yushchenko et al. (2023a, 2023b) demonstrated that the hemipenial morphology of O. cinereus, and other newly named Oligodon, had been improperly described by past studies, leading earlier authors to assign taxa into incorrect species groups. While Yushchenko et al. (2023a) did not propose any other changes, their results showed that O. “cinereus” (as previously conceived) represents a composite of species-level lineages, a finding that was also hinted by earlier works (Green et al. 2010; David et al. 2011).

Subsequently, David et al. (2023) made additional re-arrangements to O. cinereus. While they agreed with the results of Yushchenko et al. (2023a), David et al. (2023) diverged from those authors by resurrecting Simotes multifasciatus Jan in Jan and Sordelli, 1865 and Simotes swinhonis Günther, 1864 (both junior synonyms of O. cinereus) to species species, and synonymizing Oligodon condaoensis Nguyen et al., 2016 with O. cinereus. The classification suggested by David et al. (2023) is not unwarranted, and in this study, we adopt their latter change (synonymizing O. condaoensis). However, in contrast to David et al. (2023), we are more conservative regarding O. multifasciatus and O. swinhonis and prefer to recognize O. “cinereus” as a single unit that encompasses both of these taxa. Our hesitation is due to the fact that O. cinereus, O. multifasciatus and O. swinhonis (sensu David et al. 2023) do not correspond with the phylogenetic clades delimited by Yushchenko et al. (2023a). In recognition of the changes proposed by David et al. (2023), the term “species complex” is used hereafter to describe the clade containing O. “cinereus” and related taxa (see Fig. 1).

Figure 1. 

Sampled localities of the Oligodon cinereus species complex used in this study, including Oligodon speleoserpens sp. nov., O. huahin, O. phangan, O. promsombuti, O. purpurascens, O. saiyok, O. joynsoni, O. inornatus, O. nagao, O. cinereus s. str., and unnamed lineages of O. “cinereus” labeled by clade. Base map created using simplemappr.net. Filled icons denote populations included in phylogenetic analyses; empty icons denote populations not included in the phylogenetic analyses; type localities are indicated by a dot at the center of an icon. Colors and locality numbers correspond to Figure 2 and Table S1. Localities. Oligodon speleoserpens sp. nov.: (1) Tham Khao Ting Cave, Trang Prov., and Tham Le Stegodon Cave, Satun Prov., Thailand; Oligodon sp.: (2) Krabi Prov., Thailand (unconfirmed locality); O. saiyok: (3) Sai Yok Distr., Kanchanaburi Prov., Thailand; O. cinereus clade 4: (4) Rakhine Yoma Mts., Rakhine St., Myanmar; O. cinereus clade 1 (sensu stricto): (5) Cat Tien NP., Dong Nai Prov., Vietnam (type locality of O. cattienensis); (6) Bu Gia Map NP, Binh Phuoc Prov., Vietnam; (7) Kirirom NP, Kampong Speu Prov., Cambodia; O. cinereus clade 2: (8) Kalaw, Shan St., Myanmar; (9) Alaungdaw Kathapa NP, Sagaing Div., Myanmar; (10) Chi Linh, Hai Duong Prov., Vietnam; (11) Tam Dao NP, Vinh Phuc Prov., Vietnam; (12) Pu Mat NP, Nghe An Prov., Vietnam; O. phangan: (13) Mingalardon, Hlawga WP, Yangon Div., Myanmar; (14) Ko Phi Phi Don Isl., Krabi Prov., Thailand; (15) Ko Pha Ngan Isl., Surat Thani Prov., Thailand; O. huahin: (16) Kaeng Krachan NP, Phetchaburi, Thailand; (17) Suan Phueng, Ratchaburi, Thailand; O. nagao and O. cinereus clade 3: (18) Hainan Isl., Hainan Prov., China; (19) Nonggang NR, Guangxi Prov., China; (20) Huu Lien, Lang Son Prov., Vietnam (type locality of O. nagao); O. joynsoni: (21) Ngao Dist., Lampang Prov., Thailand (type locality); (22) Doi Tung, Chiang Rai Prov., Thailand; (23) Xishuanhbanna NP, Yunnan Prov., China; O. promsombuti: (24) Khao Phanom Wang, Surat Thani Prov., Thailand; O. inornatus: (25) Nong Kai Ploi, Chonburi Prov., Thailand (type locality); (26) Samkos WS, Pursat Prov., Cambodia.

During herpetological surveys targeting limestone karst formations in Trang and Satun provinces, Thailand, we observed three Oligodon that were referrable to the O. cinereus species complex. A closer examination of the Trang and Satun province specimens revealed additional genetic and morphological differences that separate them from all other members of the genus. In this study, we describe these specimens as a new species and use this opportunity to investigate the phylogenetic position of the four recently described Oligodon from Thailand (O. huahin, O. phangan, O. promsombuti and O. saiyok), all of which have not been included within published molecular datasets until now.

Both the ML and BI analyses agree with previously published phylogenies of Oligodon (Che et al. 2020; Nguyen et al. 2020; Mirza et al. 2021; Qian et al. 2021; Das et al. 2022; Yushchenko et al. 2023a, 2023b; Lee et al. 2023) (Fig. 1 and Appendix S2). Members of the O. cinereus species complex were monophyletic in both analyses (Fig. 1), and were sister to a poorly supported clade (70 / 0.92, corresponding to ML UFBS / BI PP support values, respectively) containing Oligodon maculatus (Taylor, 1918), Oligodon splendidus (Günther, 1875), O. kheriensis Acharji and Ray, 1936, Oligodon lipipengi Jiang et al., 2020 and Oligodon albocinctus (Cantor, 1839), with the latter species rendered paraphyletic by O. lipipengi. In the O. cinereus species complex, we recovered four separate lineages containing specimens previously identified as O. “cinereus” that were spread out across the phylogeny.

The first split within the species complex includes a clade containing O. cinereus clade 4, the new species from Trang and Satun Provinces, Thailand, and O. saiyok. The new species was recovered sister to O. saiyok, and those two species were recovered sister to O. cinereus clade 4 (CAS 205028, locality 4; see Fig. 1) from Rakhine State, Myanmar with low support (56 / –). O. cinereus sensu stricto (O. cinereus clade 1; restricted to southern Vietnam, Laos and Cambodia) was the next lineage to split, forming a well-supported clade (100 / 1.0) sister to all remaining members of the O. cinereus complex, which include O. phangan, O. huahin, Oligodon nagao David et al., 2012, Oligodon joynsoni (Smith, 1917), O. promsombuti, and Oligodon inornatus (Boulenger, 1914), along with O. cinereus clade 2 (including samples from Myanmar and northern Vietnam, localities 8–12; see Fig. 1) and O. cinereus clade 3 (including samples from southern China, locality 18; see Fig. 1). Two populations of Oligodon from Yangon Region, Myanmar (CAS 213379, locality 13; see Fig. 1) and Phi Phi Island, Krabi Province, Thailand (AUP TS2605, locality 14; see Fig. 1) were found to be closely related to topotypic O. phangan from Pha Ngan Island, Surat Thani Province, Thailand (locality 15; see Fig. 1). A sample morphologically identified as Oligodon nagao from Guangxi Province, China (KIZ 014591; identification by Jiang et al. 2012) was grouped with two specimens of O. cinereus clade 3 from southern China (one from Hainan Island and another from an unnamed locality). Lastly, O. promsombuti was recovered within the same clade as O. joynsoni, rendering the latter paraphyletic (Fig. 1). Uncorrected pairwise genetic distances (hereafter p distances; given for 12S–16S and cyt b) between and within Oligodon in this species complex are presented in Table S4.

Figure 2. 

Phylogenetic tree of Oligodon from the analysis of 12S rRNA–16S rRNA and cyt b mitochondrial DNA gene sequences. For voucher specimen information and GenBank accession numbers see Table S1. Numbers at tree nodes correspond to ML UFBS / BI PP support values, respectively; an en-dash denotes no support. Colors of clades and locality numbers correspond to those in Figure 1.

Based on morphology, the new species from Satun and Trang provinces of Thailand resemble most members of the O. cinereus species group, especially uniform/reticulated morphotypes of O. “cinereus”. Nonetheless, the specimen has a significant genetic distance compared to other clades within the O. cinereus species complex (mean 16S rRNA divergence 2.6% from O. cinereus clade 4; mean cyt b divergence 7.7% from O. saiyok; Table S4). These divergences are larger than other, well-established sister species pairs within Oligodon, such as Oligodon chinensis (Günther, 1888) and Oligodon formosanus (Günther, 1872) (mean 12s-16s divergence 1.8%; Nguyen et al. 2020). Furthermore, the new species can be differentiated from all other members of the O. cinereus species group by multiple morphological characteristics (see Comparisons and Table S3), especially its number of ventral and subcaudal scales, lower number of maxillary teeth, and the presence of 19 anterior dorsal scale rows in all but one specimen. The combined molecular and morphological data provide strong evidence that the Oligodon from Satun and Trang represents a new species, which we describe below.

Oligodon speleoserpens sp. nov.

https://zoobank.org/37DE18D9-FD02-4DD3-9176-BD1A3655955B

Figures 3, 4, 5, Tables 1, S3

Holotype

ZMMU Re-17696 (field number ISS-128), adult male collected on 20 January 2023 at the entrance of Tham Le Stegodon Cave, Satun Province, Thailand (7.158315°N, 99.802631°E, 28 meters elevation), by Parinya Pawangkhanant and Sabira S. Idiiatullina.

Paratype

ZMMU Re-17697 (field number NAP-13128), adult female collected 24 September 2022 inside Tham Khao Ting Cave, Trang Province, Thailand (7.158315°N, 99.802631°E, 12 meters elevation), by Nikolay A. Poyarkov, Nikita S. Kliukin, Chatmongkon Suwannapoom and Parinya Pawangkhanant.

Diagnosis

O. speleoserpens sp. nov. is referred to the genus Oligodon based on the presence of enlarged blade-shaped maxillary teeth without a diastema, an elongate and subcylindrical body, and the presence of a large inflated rostral scale that blocks the internasal scales from contacting anteriorly (Wall 1923; Smith 1943; David et al. 2023). It is distinguished from all other Oligodon by the following combination of morphological characters: 1) head oblong-shaped with a truncated snout and slightly inflated rostral scale; 2) 8 maxillary teeth, with the posterior three teeth enlarged and blade-like; 3) dorsal scales in 19–17–15 rows (17–17–15 rows in one specimen); 4) reduction from 19 dorsal scale rows to 17 dorsal scale rows occurring at the 28^th–30^th ventral scale; 5) reduction from 17 dorsal scale rows to 15 dorsal scale rows occurring at the 100^th–113^th ventral scale; 6) cloacal plate entire; 7) 189–193 ventral scales (189–190 in male; 193 in female), distinctly keeled; 8) 47–54 paired subcaudal scales (47–54 in male; 48 in female); 9) 238–244 total body scales; 10) relative tail length 0.136–0.139 and subcaudal ratio 0.198–0.221; 11) 8 supralabials on either side of the head, with the fourth and fifth scales in contact with the orbit; 12) 9 infralabials on either side of the head with the first four scales in contact with the first pair of chin shields; 13) one loreal and one presubocular present; 14) 1+2 temporal scales; 15) dorsal color pattern uniform gray or grayish–brown without any markings or reticulations; 16) anterior half of ventral surface white with gray–brown irregularly shaped spots, posterior half immaculate dark gray, underside of tail splashed with white markings; 17) hemipenis bilobed with broad, awn-shaped lobes, simple sulcus spermaticus and smooth calyces.

Comparisons

We compared Oligodon speleoserpens sp. nov. with all other members of the O. cinereus species group and use Table S3 to compare it with other Oligodon native to Thailand. Its uniform gray dorsum distinguishes most members of the O. cinereus species group that have banded, cross-barred, or blotched dorsal color patterns; namely, O. albocinctus, Oligodon ancorus (Girard, 1857), Oligodon kampucheanensis Neang, Grismer and Daltry, 2012, O. lipipengi, O. nagao, Oligodon purpurascens (Schlegel, 1837), O. saiyok, and Oligodon teyniei David, Hauser, and Vogel, 2022. Exceptions include O. “cinereus”, O. huahin, O. joynsoni, O. inornatus, and O. phangan, which have variable color patterns, or are uniform/reticulated dorsally. Between these congeners, Oligodon speleoserpens sp. nov. is most easily distinguished by usually having 19 anterior scale rows (vs. no more than 15 or 17 scale rows anywhere on the body; only one specimen of Oligodon speleoserpens sp. nov. with 17 anterior scale rows) with the reduction from 19 to 17 rows occurring between ventral scales 28 and 30.

More specifically, O. speleoserpens sp. nov. is distinguished from O. huahin by having 189–193 ventrals (vs. 166–173 in males), 47–54 subcaudals (vs. 35–41 in males), 9 infralabials (vs. 7–8), dorsum dark gray (vs. dorsum uniform but lighter gray, tan or orange), and a dark gray venter without conspicuous dark spots or rectangular blotches on the posterior half (vs. venter plain white); from O. inornatus by having 189–193 ventrals (vs. 169–173), 47–54 subcaudals (vs. 31–43), 8 maxillary teeth (vs. 10–11), and a ventral coloration without conspicuous dark spots or rectangular blotches on the posterior half (vs. ventral surface with small square or rectangular spots); from O. joynsoni by having 54 subcaudals in females (vs. 37–41 in females), relative tail length (TailLR) 0.139 in females (vs. 0.125–0.129 in females), 8 maxillary teeth (vs. 11–12), and a uniform gray or grayish–brown dorsum without crossbars or reticulations (vs. light tan, ochre brown or gray, narrow irregular shaped crossbars present on dorsal surface of body and tail); from O. phangan by having 189–193 ventrals (vs. 163–166), 47–54 subcaudals (vs. 33–42), 8 maxillary teeth (vs. 12), 9 infralabials (vs. 8), presubocular present (vs. absent), dorsum uniform gray or grayish–brown (vs. dorsum uniform orange-gray with indistinct vertebral lines present), and a ventral coloration without conspicuous dark spots or rectangular blotches on the posterior half (vs. ventral surface plain white); and from O. promsombuti by having 189–193 ventrals (vs. 177), 47–54 subcaudals (vs. 40), 8 maxillary teeth (vs. 12), 9 infralabials (vs. 8), presubocular present (vs. absent), and a ventral coloration without conspicuous dark spots or rectangular blotches on the posterior half (vs. venter with large rectangular blotches across entire ventral surface). O. speleoserpens sp. nov. differs from all clades hitherto referred to as O. “cinereus” (including O. cinereus sensu stricto) by having 189–193 ventral scales (vs. no more than 186), 47–54 subcaudals (vs. no more than 45), 8 maxillary teeth (vs. 10–13), and 9 infralabials (vs. usually 7–8, rarely 9). Moreover, the posterior portion of the venter is dark and uniform colored in O. speleoserpens sp. nov., whereas uniform or reticulated colored populations of O. “cinereus” we examined (Appendix S1) have an immaculate venter with rectangular spots or blotches on the lateral edge of each ventral scale. The ventral underside of the tail, which is dark gray smeared with minimal white markings, is found in all specimens of O. speleoserpens sp. nov. and is absent in all uniformly patterned O. “cinereus”, as well as most other members of the O. cinereus species complex.

Five additional species of Oligodon outside of the O. cinereus species group may occur in close proximity with O. speleoserpens sp. nov. and are thus compared here. Oligodon speleoserpens sp. nov. differs from members of the O. cyclurus-taeniatus species group found in the same area of southern Thailand, namely Oligodon fasciolatus (Günther, 1864), Oligodon mouhoti (Boulenger, 1914) and O. taeniatus, by its uniform dark-olive dorsum without any conspicuous head markings (vs. ligher brown dorsum with a well-defined series of vertebral stripes [O. mouhoti and O. taeniatus] or series of large blotches [O. fasciolatus], in addition to the presence of conspicuous nuchal and temporal markings on the head). Next, O. speleoserpens sp. nov. most easily differs from Oligodon signatus (Günther, 1864), restricted to a few provinces at the southern end of Thailand (Pawangkhanant et al. 2021), by its uniform dorsum (vs. dark gray with a series of reddish-brown blotches), 8 supralabials (vs. 7) and two postoculars (vs. only one). Finally, O. speleoserpens sp. nov. differs from Oligodon jintakunei Pauwels et al., 2002, known from just one specimen allegedly from Krabi Province, by its uniform dorsum (vs. distinct black and white crossbars) and by possessing an entire cloacal plate (vs. divided). Additional characters distinguishing each species can be found in Table S3.

Description of the holotype

Adult male specimen in excellent condition immediately after preservation (Fig. 3). Small ventral incisions at tail base and at midbody. SVL 601 mm, TailL 95 mm (TotalL 696 mm). HeadL 16.7 mm, HeadW 11.5 mm, HeadD 8.6 mm, SnoutL 5.3 mm, EyeD 2.7 mm, EyeLip 2.6, FrontalL 5.3 mm, FrontalW 4.3, IOD 6.5, IND 4.4 mm; RostralW 4.9. TailLR 0.136, HeadW/L 0.69, SnoutL/HeadL 0.32, EyeD/SnoutL 0.51, EyeD/HeadL 0.16, FrontalL/W 1.23. Body elongated and cylindrical in cross section; head oblong, only slightly distinct from neck (Fig. 4a, b); snout narrowing in dorsal view, in lateral view slightly truncate; snout tip subterminal near mouth; eyes moderately sized relative to head, pupil round; nostrils small and subelliptical, pointed laterally; lips curving upwards posteriorly along the last supralabial; tail broad at base, tapering gradually to a blunt terminal scute (Fig. 4f).

Figure 3. 

Photographs of the holotype of Oligodon speleoserpens sp. nov. (ZMMU Re-17696, adult male) (a) dorsal and (b) ventral views in life. Photographs taken by Parinya Pawangkhanant.

Figure 4. 

Photographs of the holotype of Oligodon speleoserpens sp. nov. (ZMMU Re-17696, adult male) (a) right lateral, (b) dorsal and (c) ventral views of the head in life; (d) partially everted hemipenes; (e) venter at midbody, note the distinctly keeled ventral scales; and (f) underside of tail depicting a dark gray and white color pattern. Photographs by Parinya Pawangkhanant (a–c and e–f), and Mali Naiduangchan (d).

Rostral distinctly enlarged and inflated, wider than high, triangular in dorsal view, partially separating the anterior half of internasals (Fig. 4b); posterior rostral scale suture bordering the internasals “deep-V” shaped, creating a narrow obtuse angle (~100º); internasals subrectangular, 1.7× wider than long, anterior suture of each scale rounded and concave, both scales oriented obliquely and posteromedially in dorsal profile; prefrontals subpentagonal, 1.5× wider than long, 1.4× longer than suture dividing both internasals, 1.2× wider than internasals, anterior suture bordering each internasal slightly rounded, concave; frontal pentagonal and shield shaped, longer than wide, anterior suture bordering prefrontals straightened; frontal 2.3× longer than prefrontals; eyes placed posterior to anterior edge of frontal; angle formed by sutures producing the posterior vertex of frontal a narrow obtuse angle (~99º); supraoculars subrectangular, 1.8× longer than wide, 0.7× as long as frontal; parietals subpentagonal, 1.3× longer than wide, width of each scale 1.1× longer than parietal suture; posterior suture of parietals straightened; frontal and parietal approximately equal in length; frontal 1.4× longer than parietal suture; anterior angle formed by the sutures between the parietal/frontal and the supraocular/parietal a broad obtuse angle (~130º) with the lateral ray oriented slightly posterolaterally. Nasal scale rectangular, longer than wide, fully divided (Fig. 4a); loreal 1/1, small, squared, slightly higher than long, less than half the size of nasal; supralabials 8/8; with 4^th and 5^th supralabials in contact with orbit; 6^th supralabial largest, 1^st supralabial smallest; all supralabials in broad contact; preoculars 1/1, wider than long; presubocular 1/1, less than half the size of preocular; postoculars 2/2, uppermost postocular slightly larger; anterior temporal 1/1; posterior temporals 2/2, all scales longer than wide; infralabials 9/9 in all specimens, first pair contacting each other, 4/4 infralabials in contact with anterior chin shields; 5^th infralabial largest, 2^nd and 9^th infralabials smallest; mental subtriangular, wider than long; small, indistinct mental groove present, starting from border of 1^st pair of infralabials and terminating past posterior pair of chin shields; both pairs of chin shields subrectangular, anterior pair 1.4× longer than posterior pair.

Dorsal scale rows 19–17–15, smooth throughout, without apical pits; dorsal scale reduction from 19 rows to 17 rows occurring at 30^th ventral scale on either side of body; reduction from 17 rows to 15 rows occurring at 100^th ventral scale on either side; preventrals 2 and ventral scales 189; lateral edges of each ventral scale with discrete lateral keeling; subcaudals 54, paired; total body scales 244; subcaudal ratio 0.221; cloacal plate entire. Maxillary teeth 8, posterior two teeth enlarged, laterally compressed, blade-like. The partially everted hemipenes is bilobed and weakly calyculate (Fig. 4d). The base of the organ and most of the capitulum is bulbous and covered with small smooth calycles. The sulcus spermaticus is simple, indistinct, extending from the base of the organ and terminating at the point of bifurcation. Both apical lobes are large, nude, and rounded distally, lacking any ornamentation.

Coloration in life and preservative

In life, dorsal ground color ashy gray, margins of most dorsal scales edged with black, but distinct reticulations absent (Figs 3a, 5a); scattered small reddish–orange mottling interspaced across dorsal surface of the tail and neck. Dorsal portion of head brown (Fig. 4b); a very indistinct temporal bar on each side of the head across the supraoculars, frontal, and medial region of the parietals; a dark, hued, and indistinct nuchal chevron present at the medial region of each parietal, descending posterolaterally towards the neck; remainder of head plain, bottommost portion of supralabials light brown (Fig. 4a). Ventral surface of head white, dark brown hueing along lateralmost margins of first 5 infralabials; the anterior third of the venter is white, with long irregular gray blotches first visible along the seventh ventral scale; these blotches continue to merge until the venter is uniform and dark gray for the remaining two thirds of the body; discrete keeling present on the edges of each ventral scale light gray, forming a subtle longitudinal stripe along each side of the venter until the tail base. The underside of the tail is dark gray, with white pigment present midventrally until its posterior half, which is completely white until the last few subcaudals and the terminal scute. In preservative, coloration overall very similar, except the dorsum is darker gray.

Figure 5. 

Photographs (a) of the holotype of Oligodon speleoserpens sp. nov. (ZMMU Re-17696, adult male) in life; (b) habitat at the type locality, Tham Le Stegodon cave, Satun Province, Thailand; (c) paratype of Oligodon speleoserpens sp. nov. (ZMMU Re-17697, adult female) in life from Tham Khao Ting cave, Trang Province, Thailand; and (d) uncollected adult male specimen of Oligodon speleoserpens sp. nov. from the same locality as the paratype. Photos by Parinya Pawangkhanant (a, b), Nikolay A. Poyarkov (c), and Harry Ward–Smith (d).

Description of the paratype and variation

The adult female paratype (Fig. 5c) agrees in almost all scalation and mensural aspects of the holotype, with a few exceptions. The paratype has a small anterior portion of the dorsum with past injuries, and small ventral incisions at the tail base and midbody. Measurements of the paratype are as follows: SVL 681 mm, TailL 110 mm (TotalL 791 mm). HeadL 16.5 mm, HeadW 14.0 mm, HeadD 10.6 mm, SnoutL 7.6 mm, EyeD 2.7 mm, EyeLip 2.9, FrontalL 5.6 mm, FrontalW 4.7, IOD 8.0, IND 5.7 mm; RostralW 4.9. TailLR 0.139, HeadW/L 0.85, SnoutL/HeadL 0.46, EyeD/SnoutL 0.36, EyeD/HeadL 0.16, FrontalL/W 1.19. Body elongated, cylindrical in cross section. Head oblong, only slightly distinct from neck. Snout more elongate in lateral view than the holotype; snout tip subterminal near mouth; eyes moderately sized relative to head, pupil round; nostrils small, subelliptical, pointed in lateral view; mouth with lips curving upwards posteriorly across the last supralabial; tail gradually tapering to a blunt terminal scute.

Rostral wider than high, triangular in dorsal view; posterior scale suture of rostral bordering internasals “deep-V” shaped, creating a narrow obtuse angle (~100º, same as holotype); internasals subrectangular, 2.3× wider than long, anterior sutures rounded, concave; prefrontals subpentagonal, 2.4× wider than long, 1.3× longer than internasals, anterior sutures bordering each internasal also rounded; prefrontals 1.3× wider than internasals; frontal pentagonal, shield shaped, 1.2× longer than wide, anterior suture bordering each prefrontal straightened; frontal 1.4× longer than prefrontals; eyes placed posterior relative to the anterior edge of frontal; angle formed by suture producing the posterior vertex of frontal a narrow obtuse angle (~96º); supraoculars subrectangular, 1.4× longer than wide, 1.3× shorter than frontal; parietals subpentagonal, 1.2× longer than wide, width of each scale 1.2× longer than parietal suture, posterior suture of parietals straightened; length of parietals approximately 1.1× longer than frontal; frontal 1.1× longer than parietal suture; anterior angle formed by the sutures between the parietal/frontal and supraocular/parietal a broad obtuse angle (~131º) with its lateral ray pointing somewhat posterolaterally. Nasal scale rectangular, longer than wide, fully divided; loreal 1/1, small and square shaped, 1.2× longer than high, less than half the size of nasal; supralabials 8/8; with 4^th and 5^th supralabial in contact with orbit; 6^th supralabial largest, 1^st supralabial smallest; all supralabials in broad contact; preoculars 1/1, wider than long; presubocular 1/1, less than half the size of preocular; postoculars 2/2, uppermost postocular slightly larger in size; anterior temporal 1/1; posterior temporals 2/2; infralabials 9/9 in all specimens, first pair in contact, 4/4 infralabials in contact with anterior chin shields; 5^th infralabial largest, both 2^nd and 9^th infralabials smallest; mental subtriangular, wider than long; small, indistinct mental groove present, condition identical to holotype; chin shields subrectangular, anterior pair 1.2× longer than posterior pair.

Dorsal scale rows 19–17–15, scale ornamentation like holotype; dorsal scale reduction from 19 rows to 17 rows occurring at the 28^th ventral scale; reduction from 17 rows to 15 rows occurring at the 113^th ventral scale; preventrals 2, ventral scales 193, distinctly keeled; subcaudals 48, paired; total body scales 242; subcaudal ratio 0.198; cloacal plate entire. Maxillary teeth 8, posterior two teeth enlarged and blade-like.

In life, dorsal ground color dark olive–gray (Fig. 5c), plain, edges of each dorsal scale slightly darker, distinct reticulations absent; a very indistinct pair of vertebral stripes present from nape to vent; scattered, small reddish–orange mottling interspaced across dorsal surface of tail; tail tip dark gray. Dorsal portion of head olive–brown; indistinct ocular and temporal bar present on each side of head, both subdued, dark brown, present across the supraoculars, frontal and medial portion of the parietals; remainder of head plain olive–brown, bottommost portion of supralabials light gray. Ventral surface of head also light gray, darker hueing along lateralmost margins of the first four infralabials; anterior third of ventral surface white with small irregularly-shaped olive–gray mottling; posterior two thirds of the venter plain dark gray; keel along ventral scales forming a narrow light-gray stripe present until the vent. Ventral surface of tail dark gray, less white pigment than holotype, anterior half similar to ventral surface of body, posterior half with small white spotting at midline until tail tip. After preservation, coloration very similar but head markings completely subdued and dorsal coloration darker gray, matching the coloration of the ventral surface.

Morphological data of another male individual from the same locality as the paratype were collected by two of us (HWS and RGL; Fig. 5d) shortly after capture (see Table 1 for more information). The uncollected specimen from Tham Khao Ting cave agrees with the paratype in most features except it has 17 anterior dorsal scale rows compared to the 19 rows found in both type specimens (total dorsal scale row formula: 17–17–15). The dorsum is ashy gray with dark edges on each dorsal scale, as observed in the type series. The indistinct temporal bar and nuchal chevron present in the holotype are extremely subdued and essentially absent. The venter is white anteriorly with small gray and irregularly shaped markings along the anterior third of the body before becoming plain and dark gray along its remaining posterior two thirds. The ventral underside of the tail resembles the paratype and is entirely dark gray except for a few subdued white spots present near the posterior end of the tail.

Table 1.

Download as

CSV

XLSX

Selected morphological counts of all three specimens of Oligodon speleoserpens sp. nov. All linear measurements are in millimeters, with abbreviations matching those listed in the materials and methods section.

Character	Oligodon speleoserpens
Catalog number	ZMMU Re-17697	ZMMU Re-17696	No voucher
Sex	F	M	M
SVL	681	601	612
TailL	110	95	97
TailLR	0.139	0.136	0.137
HeadL	16.5	16.7	–
HeadW	14.0	11.5	–
HeadW/L	0.85	0.69	–
HeadD	10.6	8.6	–
RostralW	4.91	4.9	–
FrontalL	5.6	5.3	–
FrontalW	4. 7	4.3	–
SnoutL	7.3	5.3	–
EyeD	2.7	2.7	–
EyeLip	2.9	2.6	–
DSR	19-17-15	19-17-15	17-17-15
Preventrals + VEN	2+193	2+189	2+190
CP	Entire	Entire	Entire
SC	48	54	47
SL	8/8	8/8	8/8
SL-eye	4+5/4+5	4+5/4+5	4+5/4+5
IL	9/9	9/9	9/9
IL-CS	4/4	4/4	4/4
LOR	1/1	1/1	1/1
PrO	1/1	2/1	1/1
PrsO	1/1	1/1	1/1
PtO	2/2	2/2	2/2
Temporals (Ate + Pte)	1+2 / 1+2	1+2 / 1+2	1+2 / 1+2
MT	8	8	–

Etymology

The species epithet “speleoserpens” is a compound name combining the Latinzed Greek noun “spēlēum”, meaning “cave” and the Latin noun “serpens [= serpentis]”, the present active participle of “serpō” meaning “to crawl or creep”, often used in reference to snakes. This roughly translates to “cave crawler” or “cave serpent”, an allusion to both the type locality and the discovery of the paratype. We recommend the English common name “Cave Kukri Snake” and the Thai common name “ปี่แก้วควนหิน” (Ngu Pi Kaew Kuan Hin) for this species.

Distribution and natural history

Oligodon speleoserpens sp. nov. is currently known from only two localities in southeastern Peninsular Thailand: Tham Le Stegodon, Satun Province and Tham Khao Ting cave, Trang Province (see locality 1, Fig. 1). Both locations are cave formations that form part of a larger complex of limestone karst massifs along the border between Satun and Trang provinces. The holotype was discovered crawling along a vertical limestone wall just outside the entrance of Tham Le Stegodon (Fig. 5b) at 1950 hrs (local ITC time), whereas the paratype was discovered at 2200 hrs in-situ coiled within a crevice in Tham Khao Ting cave, approximately 10 meters past the entrance. The third, uncollected specimen from Tham Khao Ting cave was observed by two of us (HWS and RGL) on 19 March 2023 at 2100 hrs. It was first observed ~12–15 meters up a karst wall (Fig. 6a) with its head peeking into a small crack. It proceeded into this crack, emerging 5 minutes later with no signs of finding prey. Both of these specimens were adult males, whereas the paratype collected earlier in September 2022 was an adult female. Due to the proximity of Satun and Trang Provinces to the equator, sunset varies very little through the year, from 1810 hrs at the winter solstice to 1837 hrs at the summer solstice. All specimens were observed after sunset (1900–2300 hrs), but we assume this species is cathemeral (including both diurnal and nocturnal surface activity), as is consistent with other members of the O. cinereus species group.

Figure 6. 

Habitat photographs (a) of the karst wall above an entrance to Tham Khao Ting Cave, Trang Province, Thailand that the uncollected specimen of Oligodon speleoserpens sp. nov. was found ~12–15 m high; (b) flooded exit of the Tham Khao Ting cave; and (c) flooded habitat 50 m past the entrance of Tham Le Stegodon cave, Satun Province, Thailand. Photographs by Harry Ward–Smith.

The diet of most Oligodon is presumed to consist primarily of reptile eggs and frogs (Wall 1923; Bringsøe et al. 2020; David et al. 2023), but Meggitt (1931) and Pope (1935) both recorded several arachnid and arthropod species in the stomachs of specimens in the O. cinereus species complex (Savitzky 1983; as O. cinereus). Whether O. speleoserpens sp. nov. has a similar reptile egg-rich diet remains an open question. We observed three highly abundant gecko species present in sympatry with O. speleoserpens sp. nov. including Cnemaspis niyomwanae Grismer et al., 2010, Cyrtodactylus cf. lekaguli Grismer et al., 2012 and Gekko gecko (Linnaeus, 1758). All three lizards occupy the same karstic surfaces and utilize crevices as egg laying sites. The exploratory behavior observed in the uncollected specimen could be related to foraging for these species and their eggs. We also observed five additional generalist and widespread gecko species at the same locality, including Cyrtodactylus quadrivirgatus Taylor, 1962, Cyrtodactylus zebraicus Taylor, 1962, Gekko tokehos (Grismer et al., 2019), Gekko kuhli (Stejneger, 1902) and Hemidactylus frenatus Duméril and Bibron, 1836. Other snake species found in sympatry with O. speleoserpens sp. nov. include Ahaetulla mycterizans (Linnaeus, 1758), Ahaetulla prasina (Boie, 1827), Boiga melanota (Boulenger, 1896) (recognized at species-level fide Weinell et al. 2021), Bungarus flaviceps Reinhardt, 1843, Elaphe taeniura ridleyi (Butler, 1889), Lycodon capucinus Boie, 1827, Lycodon (Dryocalamus) davisonii (Blanford, 1878), Malayopython reticulatus (Schneider, 1801), Pareas carinatus Wagler, 1830, Trimeresurus ciliaris Idiiatullina et al., 2023, Tropidolaemus wagleri (Boie, 1827), and Xenochrophis trianguligerus (Boie, 1827). The diversity and abundance of geckos and other squamate reptiles in this karstic habitat provides numerous potential prey items for Oligodon speleoserpens sp. nov., especially in the form of eggs. No other information on its natural history or behavior is known.

Conservation status

Only three specimens of Oligodon speleoserpens sp. nov. have been documented so far, all within a span of six months. The surrounding limestone karst massif where the new species has been collected spans a total area of approximately 10.7 square kilometers. Both sites are offered protection from human development. Tham Le Stegodon is part of the Satun UNESCO Global Geopark and Tham Khao Ting is owned by the Liphang Subdistrict Administrative Organization of Trang Province. The underground river networks of both cave systems are popular tourist attractions for swimming and kayaking. We believe the current level of recreation poses minimal impact on the caves and surrounding karst habitat and is thus unlikely to threaten the conservation of Oligodon speleoserpens sp. nov. at these sites. However, future work on the ecology and behavior of this species is needed to understand whether any other human activities in this region might act as conservation threats. Owing to a lack of ecological information on this species, we suggest classifying Oligodon speleoserpens sp. nov. as “Data Deficient” based on the criteria adopted by the International Union for Conservation of Nature (IUCN) Red List of Threatened Species. This classification is based on the fact that little data exists for the new species, and that more information is needed to understand its distributional limits within southern Thailand.

The discovery of Oligodon speleoserpens sp. nov. brings the total number of recognized Oligodon species to 88 or 91 (see Introduction) and the number of species known from Thailand to 21 (but see David and Pauwels 2005; Lee et al. 2023a; Poyarkov et al. 2023; and David et al. 2023 for discussion of some species records and their validity). The surrounding karst formations of southern Thailand and the broader Nakawan Range are home to several putatively endemic reptile species (Kiew 1985; Grismer et al. 2008; Grismer et al. 2010; Grismer et al. 2012; Termprayoon et al. 2023; Grismer et al. 2023; Idiiatullina et al. 2023), including the gecko Cnemaspis niyomwanae and the newly described pit-viper Trimeresurus ciliaris. The type localities of these two species sit on the same karst massif as the type locality of O. speleoserpens sp. nov., and their precise location (Tham Khao Ting cave) is also the site where the paratype and one other specimen of O. speleoserpens sp. nov. were observed. Another endemic colubrid snake species, Lycodon cavernicolus Grismer et al., 2014, was discovered less than 64.0 km SE of the O. speleoserpens sp. nov. type locality in Gua Wang Burma cave, Perlis, Malaysia. The addition of Oligodon speleoserpens sp. nov. reinforces the importance of karst habitats as local hotspots of herpetofaunal endemism in Peninsular Thailand and suggests that many more undescribed species of reptiles restricted to these ecosystems await discovery.

Limestone karsts and their caves are well-known for supporting high levels of range-restricted herpetofauna (Clements et al. 2006; Quah et al. 2021), especially species with low-dispersal capabilities such as geckos (Ellis and Pauwels 2012; Grismer et al. 2018; Davis et al. 2019), frogs (Köhler et al. 2010; Brown et al. 2015; Grismer et al. 2017; Suwannapoom et al. 2018) and snakes (Vogel et al. 2012; Teynié et al. 2014; Grismer et al. 2014; Ruane et al. 2016; Sumontha et al. 2017; Luu et al. 2020; Sumontha et al. 2021). Before this study, three species of Oligodon were considered limestone karst specialists: O. nagao, O. promsombuti (Ian Dugdale pers. obs., February 2023) and O. saiyok (David et al. 2012; Sumontha et al. 2017; Pauwels et al. 2021). In addition, Grismer et al. (2006) and Quah et al. (2021) reported a fourth species, O. fasciolatus, from limestone caves in Thailand and Peninsular Malaysia (on the island of Pulau Langkawi). We confirm this behavior and supplement these reports with observations of O. fasciolatus in karstic landscapes in Kanchanaburi Province and Prachuap Khiri Khan Province, Thailand (HWS, pers. obs., 2022). Additionally, we add O. huahin as a species known from karst habitats, based on a single specimen (USNM 94932) collected from Sam Roi Yot, Prachuap Khiri Khan Province, Thailand, a locality that is completely comprised of limestone karst massifs. This specimen was originally identified as O. cinereus, but subsequent examination by one of us (HWS) confirms that it can be referred to O. huahin (Appendix S1). We have also observed another species, O. purpurascens, active along the edges of limestone karst formations in Phatthalung Province, Thailand. However, we do not consider O. fasciolatus, O. huahin or O. purpurascens to be limestone karst specialists, as these species are also found in mixed forests and agricultural landscapes (Tillack and Günther 2009; Das 2010; Pauwels et al. 2017). Oligodon speleoserpens sp. nov. is thus the seventh Oligodon recorded from karstic formations, and based on our observations, it appears restricted to these habitats. With the level of diversity and local endemism of geckos found in the Nakawan Range, it is not surprising that an endemic snake from a genus that normally preys on reptile eggs is also present. This discovery remains consistent with other karst dwelling Oligodon, including O. saiyok, which is sympatric with five species of endemic geckos (Bauer et al. 2003; Bauer et al. 2004; Bauer et al. 2008; Grismer et al. 2010; Panitvong et al. 2014; Sumontha et al. 2017) found in karst habitats surrounding Sai Yok, Kanchanaburi Province, Thailand. This example, and our current study, suggest both the high detectability of lizards and the beta diversity of karst endemic gecko species could be used as indicators for targeting novel Oligodon populations across peninsular Thailand, allowing researchers to refine future surveying efforts.

It is difficult to make any biogeographic assessments of O. speleoserpens sp. nov. because the broader evolutionary relationships of Oligodon remain poorly understood. Nevertheless, it is instructive to note that O. speleoserpens sp. nov., differs from most of its congeners in the O. cinereus species complex by its placement south of the Isthmus of Kra, a location that is known for its high-levels of species endemism and its function as an area of faunal turnover between Indochina and Peninsular Malaysia (Pauwels et al. 2003; de Bruyn et al. 2014; Mulcahy et al. 2018; Poyarkov et al. 2023). Although it is possible that O. speleoserpens sp. nov. represents a site-specific endemic, the continuity of similar limestone karst habitats adjacent to the type locality suggests it probably occurs at other neighboring localities. Similarly sized karst formations are found immediately east of the type locality of the new species along the leeward foothills of the broader Nakawan Range. This series of mountains extends from the Thai–Malay border near Perlis, Peninsular Malaysia to the northern edge of Trang Province, where it continues to the Gulf of Thailand as the Nakhon Si Thammarat Range. If O. speleoserpens sp. nov. is not a site-specific endemic to the karst surrounding Tham Le Stegodon and Tham Khao Ting caves, it likely inhabits adjacent habitats throughout Satun and Trang Provinces. Likewise, we also expect that additional new Oligodon in the O. cinereus species complex await discovery in southern Thailand, as has been the case for other karst-endemic geckos and pit vipers in this region (Termprayoon et al. 2023; Idiiatullina et al. 2023). Already, we are aware of two Oligodon specimens that were found in more northerly limestone karst localities on the citizen scientist website iNaturalist (www.inaturalist.org), the first from Chumphon Province (obs. 147890899) and Krabi Province (obs. 151275183). The specimen photographed from Krabi Province (see locality 2, Fig. 1) shares some features that resemble the coloration observed in O. speleoserpens sp. nov., specifically a plain, dark olive dorsum, however the remaining photos do not provide enough information to confidently identify it. Future taxonomic research and fieldwork is necessary to determine the status of these two populations. Outside of Thailand, we encountered a few sources (Green 2010; Stuart 2010; Yushchenko et al. 2023a; Uetz et al. 2023) that note the occurrence of O. cinereus in neighboring Peninsular Malaysia. However, we could not find any evidence supporting the occurrence of O. cinereus in the country (L. Lee Grismer, pers. comm. April 2023). Possibly, the statements from other authors were the result of confusion with de Rooij (1917), who reported the taxon as Oligodon violaceus (a name now synonymized with O. cyclurus, but historically confused with O. cinereus) from the island of Borneo, although that record is now considered doubtful (Stuebing et al. 2014). Until further evidence is presented, we suggest removing O. cinereus from the snake fauna of Malaysia, but do not rule out the possibility that Oligodon speleoserpens sp. nov. might eventually be detected within its geographic boundaries, as has been the case with the pitviper species T. ciliaris (Idiiatullina et al. 2023) and the re-confirmation of Cyrtodactylus zebraicus (Quah et al. 2023).

While only three specimens of O. speleoserpens sp. nov. have been observed so far, a few aspects of its natural history can be hypothesized. One feature found in all individuals of the new species is the presence of keeled ventral scales. This scalational feature has been observed in other Oligodon, including members of the O. cinereus species group (Tillack and Günther 2009), but its presence across other genus members is poorly documented. In snakes, keeled ventral scales are generally associated with arboreal and scansorial behaviors since they help maintain traction on rough, vertical surfaces. The habitats surrounding the type locality of O. speleoserpens sp. nov. contain large rock outcrops that pose as significant navigational challenges for many snakes. It seems reasonable to hypothesize that the distinctly keeled ventral scales found in this species might assist with climbing across these terranes, as evidenced by the discovery of one specimen found active high up a vertical karst wall, and the discovery of the paratype active along the chamber of Tham Thao King cave (Fig. 6a). The apparent restriction of O. speleoserpens sp. nov. to limestone karst massifs with accompanying underground caves may also suggest it is troglophilic, a behavior that is common in other snake species that utilize these types of habitats, such as E. taeniura ridleyi (Monroe, 1970; Quah et al. 2021). Nonetheless, all of these speculations can only be confirmed if additional observations of O. speleoserpens sp. nov. are made in the future.

A significant hurdle preventing a broader understanding of Oligodon systematics has been their low detectability during field surveys, which makes assessments based on large series of specimens difficult. Characters that have traditionally been used to classify Oligodon species include the shape and ornamentation of the hemipenis, number of maxillary teeth, number of dorsal scale rows, and number of body scales (ventrals and subcaudals) (Wall 1923; Smith 1943; Leviton 1963; Wagner 1975; David et al. 2008a, 2008b). These features normally diagnose taxa efficiently; however, recent studies have pointed out errors promulgated in several hemipenial descriptions of Oligodon, resulting in taxonomic confusion (Yushchenko et al. 2023a, 2023b). In this study, samples of the four recently described Thai Oligodon we included in our phylogeny (O. huahin, O. phangan, O. promsombuti and O. saiyok) were all recovered in the O. cinereus species complex, even though three were originally classified in a separate species group (Pauwels et al. 2017, 2021). The fact that these species (O. huahin and O. promsombuti) were incorrectly classified exemplify the issues related to the hemipenis raised by Yushchenko et al. (2023a, 2023b). After a re-examination of the hemipenes of O. huahin and O. promsombuti, we found that their organs are bilobed with large broad lobes, no distinct flounced calyces and an unforked sulcus spermaticus (Pauwels et al. 2017, 2021). All of these characteristics resemble the hemipenial morphology observed in O. cinereus (Yushchenko et al. 2023a). Since the molecular and morphological evidence both point to the same conclusion, we formally assign O. huahin, O. phangan, O. promsombuti and O. saiyok to the O. cinereus species group.

Outside of these taxonomic decisions, we are unable to sort out the identities of the many lineages recovered in the O. cinereus species complex, and our results raise several issues that await further study. The most glaring is the non-monophyly between O. joynsoni and O. promsombuti, and between O. nagao and O. cinereus clade 3. Both specimens of O. joynsoni sampled for genetic data agree with the morphological diagnosis of that species by having a large number of ventral scales (>190) and a color pattern consisting of dark reticulations and narrow crossbars, with one specimen (KIZ 09128) described in detail by Jiang et al. (2012). It is puzzling that such a close relationship exists between the two species, since O. joynsoni and O. promsombuti have significantly different color and scalational features and are separated from one another by a distance of ~650 km (see Table S3 and Fig. 1). However, the genetic samples obtained from the holotype of O. promsombuti consist of two short reads of 16S rRNA and cyt b respectively, and the internal node connecting the sample to the other two O. joynsoni sequences could easily be overturned if longer DNA fragments, or additional topotypic material are obtained. Similarly, O. nagao also has several color and scale charactersitics that differentiate it from populations assignable to O. cinereus clade 3, specifically a higher number of ventral scales and a distinct dorsal pattern consisting of large butterfly-shaped blotches (David et al. 2012; Jiang et al. 2012). Consequently, we are hesitant to adjust the taxonomic status of O. joynsoni, O. nagao, and O. promsombuti until additional sampling and revisionary work on the O. cinereus species complex can be conducted.

Our study also supports the existence of a paraphyletic O. “cinereus”, with more than six species nested within or between each of its lineages. The inclusion of O. huahin and O. phangan within the O. cinereus species complex significantly affects each species’ morphological diagnosis because many of their scalation features now overlap with O. “cinereus” sensu lato (Table S3). We included two topotypic samples of O. phangan from Pha Ngan Island and two specimens (AUP TS2605 and CAS 213379) from Krabi Province, Thailand and Yangon, Myanmar, respectively, that cluster within the same clade. While these latter samples would seemingly extend the distribution of O. phangan, their morphologies do not precisely match the original description of the type specimens. For example, CAS 213379 is part of a series of Oligodon collected from Yangon Region, Myanmar, and all of these specimens have a faint reticulated dorsum with dark spots along the posterior portion of the venter. As a result, we conservatively identify the two non-topotypic samples as O. cf. phangan for the time being. At a minimum, our results indicate the morphological diagnoses that would normally distinguish Oligodon from one another fail to accurately describe the lineage diversity found within O. “cinereus”. A clearer understanding of the species boundaries of the O. cinereus species complex will have to await denser sampling of genetic and morphological material.

We are indebted to Ian Dugdale (Baan Maka Nature Lodge, Phetchaburi) for sharing photographs and additional observations of Thai Oligodon, and Mali Naiduangchan (Rabbit in the Moon Foundation, Ratchaburi) for helping us photograph the hemipenis of the O. speleoserpens sp. nov. holotype. We thank the following museum collections staff for allowing us to examine specimens: David Kizirian and Lauren Vonnahme (AMNH); Ned Gilmore (ANSP); Lauren Scheinberg, Erica Ely, and Jens Vindum (CAS); Rachunliu G. Kamei, Sara Ruane, Joshua Mata and Alan Resetar (FMNH); Stevie Kennedy–Gold and Joe Martinez (MCZ); Giuliano Doria and Massimo Petri (MSNG); Patrick D. Campbell and Jeffrey W. Streicher (NHMUK); Amy L. Lathrop (ROM); Gregory L. Schneider and Daniel L. Rabosky (UMMZ); Esther Langan, Addison Wynn, Rob Wilson, George R. Zug (USNM); and Valentina F. Orlova (ZMMU). We also thank Patrick David (MNHN), Van Wallach (Cambridge, USA) and Hinrich Kaiser (Victor Valley College, USA) for their constructive reviews of the manuscript.

Fieldwork and specimen collection were approved by the Institute of Animals for Scientific Purpose Development (Bangkok, Thailand) and the Institutional Ethical Committee of Animal Experimentation of the University of Phayao (certificate no. UPAE64-02-04-005 and permit no. UPAE59-01-04-0022; both issued to CS). All fieldwork was strictly complacent with the ethical conditions of the Thailand Animal Welfare Act. This study was completed within the frameworks of research project E-1.2 of the Joint Russian-Vietnamese Tropical Science and Technology Research Centre for 2023. NAP, NSK, SSI and AVT were financially supported by the Russian Science Foundation (RSF grant No. 22-14-00037: fieldwork, specimen collection, morphological examination, molecular phylogenetic analyses, and data analyses). CS was partially supported by the Thailand Science Research and Innovation Fund and the University of Phayao 208/2567 (Unit of Excellence 2024 on the Integrative Diversity Assessment of Aquatic Animals from Thailand).