Gloydius hindukushensis sp. nov.

Common names.

Hindu Kush Pitviper (English), ہندوکش گھڑے والا سانپ (Hindu Kush gaday wali afee, Urdu)

Type material.

Holotype. • NHMW 41993, an adult female from Kumrat Valley, Upper Dir District, Khyber Pakhtunkhwa Province, Pakistan, 35.5649°N, 72.1958°E, 2360 m altitude, found freshly killed by locals on 17 September 2020 by Daniel Jablonski (formerly CUHC 10085; Fig. 15). Paratype. • CUHC 10088, a juvenile female, collected on 17 September 2020 by Daniel Jablonski, other data as for the holotype.

Description of the holotype.

Adult female, indicated by the absence of hemipenes; 1 / 1 loreal, slightly wider than high; nasal partially divided below naris; 3 / 3 elongated preoculars, the lowest is the narrowest; 1 / 1 supraocular; 3 / 3 postoculars, upper reaching onto top of head and touching the parietal, the lower can be described as postsubocular, because it extends under the eye, up to the level of the posterior border of the third supralabial; rostral scale wider than high; two internasals wider than long; two prefrontals, slightly longer than wide; frontal bell-shaped, as long as wide; 3 / 3 anterior temporals, upper two significantly smaller than the lower one, and 3 / 3 posterior temporals, the two upper of similar size, the lower one slightly larger and in contact with the second last supralabial; supralabials 7 / 7, second last three times higher than the previous one, in contact with lower posterior temporal, 3 rd / 3 rd in contact with the eye; pit opening shorter than horizontal diameter of eye and encircled by 3 / 3 scales; 18 circum-pileus scales; 9 / 9 sublabials, first four in contact with anterior submaxillars; two pairs of submaxillars, anterior nearly twice as wide and 80 % longer than posterior; followed by five rows of paired gular scales, increasing in size posteriorly; dorsal scales in 23-21 - 19 rows, strongly keeled, except the outer row, which is only very slightly keeled, paired apical pits on some dorsals, very weakly developed; four preventrals; 155 ventrals; cloacal plate entire; 38 / 39 paired subcaudal scales. Body compact, subcylindrical; tail short (TaL / TL 0.122); SVL 578 mm; TaL 80 mm, head length measured from tip of snout to posterior border of parietals 17.9 mm, head length measured from tip of snout to posterior edge of mandible 27.0 mm, head width 15.9 mm.

Dorsal scale reduction formula.

Dentition.

Maxillary bone with two posteriorly curved fangs on each side. Both teeth are not connected to the maxilla on the right side. Behind the main fangs are 6 / 7 replacement fangs at different growth stages. Main fang 5.27 mm in length, i. e., 25.5 % of skull length. Discharge orifice 1.14 mm in length, i. e., 21.6 % of fang length. Palatine bone with 4 / 4 posteriorly curved teeth slightly decreasing in size posteriorly. Teeth III and IV loose on left, teeth II and IV on right side. Lateral to each palatine tooth is a single replacement tooth at different growth stages. Pterygoid bone with 10 / 12 posteriorly curved teeth, shorter than the palatine tooth, all nearly the same size. Tooth IX loose on left, teeth II, IV, VI, VII, IX and XII loose on right side. Pterygoid bone broken between tooth III and IV on left, as well as behind the tooth row on right side. The posterior 75.0 % of the pterygoid bone is without teeth. Mandibular bone with 14 / 14 posteriorly curved teeth gradually decreasing in size posteriorly. The first two teeth closer together than the rest. Medial to each mandibular tooth are up to two replacement teeth in different growth stages. Teeth I, IV, VI, VIII, X, XII, and XIV loose on left side. Teeth I, III, V, VII, VIII, X, XI, and XIII loose on right side. Splenial 67.2 % of length of angular, latter overlaps splenial with 18.3 % of its total length. The total length of splenial-angular complex spans 33.4 % of the mandibular bone. The dental is 38.5 % as long as the mandibular bone. The complete skull of the holotype NHMW 41993 is presented in Fig. 16.

Colouration and pattern.

Colouration after approximately five years preservation in ~ 70 % ethanol was recorded as follows: Dorsal ground colour Drab-Gray (Colour 256), with 37 Sepia (279) coloured dorsolateral roundish blotches which are four to five dorsal scales wide and show a pale Drab-Gray (256) centre; most blotches open ventrolaterally or the lower outer edges disintegrate and form a ventrolateral row of small spots; below that a row of small, irregular rhombic or x-shaped Sepia (279) coloured and Cream White (52) bordered spots runs along the outer edge of the ventrals; the dorsolateral blotches merge vertebrally and form an irregular, sometimes interrupted zigzag line; dorsal tail with ten irregular blotches in the same colouration as on the body; dorsal head Sepia (279) coloured; lateral head colour Pale Buff (1) to Cream White (52) with dense Sepia (279) mottling, most intense in the loreal and upper temporal region; a wide Olive-Brown (278) postocular stripe from the posterior border of the eye to the posterior edge of the mandible, whose lower edge is bordered by a dark Sepia (279), wavy narrow line; lower parts of posterior supralabials Pale Buff (1) with a few Dark Grayish Brown (284) speckles; lower labials with small triangle shaped Sepia (279) coloured markings on their sutures; neck with two wavy lateral Sepia (279) stripes and a shorter elongated spot in the centre; venter with a colour gradient that becomes darker towards the tail; throat Pale Neutral Gray (296) coloured, submaxillars and gulars mottled with Dark Grayish Brown (284); in the further course, venter ground colour changes gradually to Light Neutral Gray (297) and shows heavy Sepia (279) mottling; tail tip Medium Fawn (258) coloured. Examples of variation in the colouration and pattern of live individuals; see Fig. 17.

Variation.

The paratype and other examined material agree well with the holotype in general appearance. For differences based on sexual dimorphism, morphometrics and scalation we refer to Tables 3, 4, 6.

Sum formula of dorsal scale reduction in females.

Variation in dentition unknown.

Variation in life colouration and pattern.

Dorsal ground colour varies from Drab-Gray (Colour 256) to Medium Fawn (257) with 37–44 Ground Cinnamon (270) to Olive-Brown (278) coloured dorsolateral blotches on body and 10–12 on tail; the lighter centres of the body blotches are set off from the darker edge with fine white speckles.

Etymology.

The specific epithet hindukushensis is derived from the Hindu Kush mountain range, where the new species was discovered. The name is a Latinised adjective meaning " from the Hindu Kush ". The term is of Persian origin, with a debated etymology that reflects the region’s role as a barrier between Central and South Asia. The Hindu Kush mountain range, spanning Afghanistan and Pakistan, forms a major biogeographic divide, separating distinct faunas on its eastern and western slopes. This pattern aligns with our findings of a distinct lineage within the G. himalayanus complex, whose distribution is shaped by mountain ridges and river valleys. The new species is distributed across the Pakistani portion of the Hindu Kush west of the Indus River. This study represents the first application of this epithet to a reptile species and highlights the region’s ecological and evolutionary distinctiveness, which supports other range-restricted or endemic species, such as Paradactylodon mustersi (Smith, 1940), Chrysopaa sternosignata (Murray, 1885), Altiphylax levitoni (Golubev & Szczerbak, 1979), or Laudakia nuristanica (Anderson & Leviton, 1969).

Distribution.

Gloydius hindukushensis sp. nov. is only known from a few localities between the border with Afghanistan and the Indus River in the south-eastern and eastern foothills of the Hindu Kush Mountains in Pakistan (own data and Sclater 1891; McMahon 1899; Wall 1911; Minton 1966; Gloyd and Conant 1990; Jamal et al. 2018). A combination of genetic and morphological data confirms the species from several localities in the Pakistani districts of Chitral, Swat, Upper and Lower Dir in the Khyber Pakhtunkhwa Province and from the districts of Gilgit and Diamer in the Gilgit-Baltistan Province (see Fig. 9, Suppl. material 1: table SS 3).

These distribution points suggest a range confined to the eastern slopes of the Hindu Kush, and while the species’ presence in eastern Afghanistan remains unconfirmed, it has been previously considered likely (Gloyd and Conant 1990). The estimated extent of occurrence (EOO) is ca 31,000 km 2, although actual occupancy is expected to be smaller due to limited known localities and the species’ apparent geographical isolation.

Habitat and ecology.

Wall (1911) described the snake (classified as Ancistrodon himalayanus) as common at Madak Lasht in Chitral where he found numerous specimens on the ground between rocks where they hide in beds of needles from various conifers e. g., Picea morinda, Abies webbiana and Cedrus deodara. Minton (1966) reported the Hindu Kush Pitviper (classified as Agkistrodon himalayanus) as plentiful at Liakot village (Swat) where it was found on rocky lightly wooded hillsides. Recently, verified locality records place this species at elevations ranging from approximately 1660 m (Torwal, Swat District, Pakistan) to 2888 m (Dog Dara, Upper Dir District, Pakistan). Other significant localities include the type locality in the Kumrat Valley (2360 m) and areas near Lal Qilla (2446–2700 m). This elevational range situates G. hindukushensis sp. nov. within montane coniferous and mixed forest zones, encompassing temperate pine forests, alpine scrub, and open rocky terrain interspersed with grassy cover and scattered shrubs (Fig. 17). These habitats are characterised by cool, humid conditions, with heavy snowfall in winter and a monsoon-influenced wet season, providing a suitable microclimate and ecological niche for this high-altitude viper.

Conservation.

An early account by McMahon (1899), referring to the species under the name Halys himlayanus, noted its abundance in the wooded tracts of the Gilgit District. However, its restricted geographic distribution in the Pakistani part of the Hindu Kush, high-altitude ecological specialisation, and apparent habitat specificity suggest that the species may be particularly susceptible to the impacts of climate change, anthropogenic habitat alteration, and possible range contraction. In Pakistan, populations historically attributed to G. himalayanus are frequently persecuted and killed by local communities due to fear or a lack of basic knowledge of snake behaviour, a threat that also led to the death of the holotype specimen described in this study. These combined pressures underscore the need for targeted conservation measures and systematic field surveys to evaluate population status, distribution boundaries, and prevailing threats.