A new species of shieldtail snake (Squamata: Uropeltidae: Uropeltis ) from the Bengaluru uplands, India

: A new species of shieldtail snake, Uropeltis jerdoni , is here described based on eight specimens from Devarayana Durga and Nandi Durga that are under-researched hills near Bengaluru in southern India. The new species is a member of the Uropeltis ceylanica group that can be distinguished from related taxa as follows: a truncate and flattened caudal shield with a circumscribed concave disc; part of rostral visible from above subequal to its distance from frontal; rostral partially separating nasal scales; 17: 17: 17 dorsal scale rows; 140–148 ventral scales; 7–9 pairs of subcaudals; dark blackish-grey above, powdered with minute yellow specks, yellow lateral stripes on neck and tail; ventrolateral region with yellow mottling; venter black. This new species is currently known only from two ranges Devarayana Durga and Nandi Durga but judging by the presence of similar, adjacent massifs, is hypothesized to be present in nearby hillocks surrounding Bengaluru City.


INTRODUCTION
Snakes of the family Uropeltidae Müller, 1832 are an under-researched group of small and unassuming fossorial snakes from the Indian subcontinent (Beddome 1886;Smith 1943;Rajendran 1985;Whitaker & Captain 2004;Wallach et al. 2014).The genus Uropeltis Cuvier, 1829 currently consists of 25 species occurring in the hills of peninsular India (Pyron et al. 2016;Jins et al. 2018;Ganesh & Achyuthan 2020).The first species in this genus that was described was Uropeltis ceylanica Cuvier, 1829, a species that is currently considered to be a complex (Gower et al. 2008;Ganesh et al. 2014) and with an erroneous type locality 'Ceylon' that is outside the known distribution of the species (see Pyron et al. 2016).Of late, two species U. bicatenata (Günther, 1864) and U. shorttii (Beddome, 1863) previously considered invalid were resurrected and one species U. madurensis (Beddome, 1878) that was previously considered a subspecies, was elevated to a full species rank (see Gower et al. 2008;Ganesh et al. 2014).In recent times, two new species Uropeltis bhupathyi Jins, Sampaio, Gower, 2018 and Uropeltis rajendrani Ganesh & Achyuthan, 2020 were described from the Western Ghats and the Eastern Ghats, respectively (Jins et al. 2018;Ganesh & Achyuthan 2020).
These snakes, owing to their naturally-patchy distribution and high beta diversity, that is a diversified multi-species assemblage of fauna constituted by each having a small, typically non-overlapping distribution range, resulting in turn-over among hill ranges (sensu Socolar et al. 2016), were hypothesized to be potential model organisms for evolutionary studies in the Indian peninsula (Cadle et al. 1990;Bossuyt et al. 2004;Ganesh 2015;Pyron et al. 2016).Molecular phylogenetic studies reveal that this genus of snakes radiated rapidly and recently during early Miocene, some 20 million years ago (Cyriac & Kothandaramiah 2017).Despite these works, the fact is that our current understanding of the diversity and distribution of the genus Uropeltis remains incomplete.Here, we describe a new species of Uropeltis representing an innominate population from a locality that is previously-unsampled for shieldtail snakes, a hill-dominated region situated around Bengaluru City that is recently recognised as an important area for herpetological diversity and endemism (see Agarwal et al. 2019).

MATERIALS & METHODS
The current work is based on our examination of 39 preserved specimens (representing 16 congeners) and the type specimens as well as live uncollected specimens of the new species that is described herein.During our expeditions in the uplands of Bengaluru, we came across three specimens (two dead, one alive) that we could assign to the genus Uropeltis sensu Pyron et al. (2016) in lacking mental grove, supraocular, postocular or temporal scales and having a dorso-ventrally depressed tail with a scaly caudal disc.The road kill specimens were noticed having apparent breakage of certain scales in the ventral aspect and indentations in parts of their trunk suggesting a run over by a small vehicle.During our perusal of uropeltid collections in the Bombay Natural History Society Museum, we came across six historical specimens that fully match with the new species.We photographed the subjects using high resolution digital cameras.We scored morphological details like scalation, measurements and colour patterns with the help of magnifying hand lenses (5 X optical zoom).We measured the preserved specimens using vernier calipers (LC 0.1 mm) except for snout-vent length that was measured with a standard measuring tape (LC 1 mm).We followed Smith (1943) for definition and terminology of morphological characters, except for ventral scales for which Gower & Ablett (2006) counting method was followed.Symmetrical head scalation values were given in left / right order.Dentition characters were scored by counting one half (lateral side) of both the upper jaw (maxillary) and the lower jaw (mandibular/ dentary).Teeth were counted by manually opening the preserved specimen's mouth and inserting a cotton plug.Counts were done viewing through a Celestron 20-200 X zoom magnification illuminated microscope.A linear incision in the subcaudal was done on the preserved specimens to check for genitalia.Comparisons and differential diagnosis are provided based on the series of preserved voucher specimens in collections that we examined (see Appendix 1) and also on our perusal of original description papers and subsequent taxonomic treatises (see literature cited).The new species belongs to Smith's (1943) Group II, in having an obliquely truncate tail, terminating in a thickened, circumscribed, concave caudal disc covered with multicarinate scales (see Smith 1943).Comparisons are presented as differential diagnosis, following the pattern in works on the genus Uropeltis by Ganesh et al. (2014) and Ganesh & Achyuthan (2020).Museum abbreviations are as follows: CSPT-Chennai Snake Park Trust, Chennai, India; Holotype: BNHS 3562, adult female, a rather intact roadkill near village, coll.KGP and NSA in January 2020.
Etymology: Patronym named in genitive singular case, honouring Thomas Caverhill Jerdon (1811-1872), a pioneering English naturalist who described some of the earliest reptiles from the Bengaluru uplands.
Diagnosis: A species of Uropeltis known from the Bengaluru uplands, characterized by the following combination of characters: caudal shield truncate, with a distinct thickened circumscribed concave disc; part of rostral visible from above subequal to its distance from frontal; rostral scale partially separating nasal scales; snout fairly pointed, subovoid; eye diameter 3/4 th that of ocular shield; supralabials 4; infralabials 3-4; dorsal scale rows 17:17:17; ventral scales 140-148; subcaudal scales 7-9 pairs; dorsum dark blackish-grey overall with minute yellow speckling; an yellow stripe on either sides on neck and tail; ventrolateral region distinctly mottled with yellow; venter uniform dark blackish-grey, rarely with a few yellow dots.

Description of holotype
Habitus: A fairly small but thick-set and robust shieldtail snake; forebody mildly thicker than the rest of trunk; head not evident, narrower than neck; snout fairly pointed in profile, subovoid; eyes large, ¾ the size of ocular scale; tail with a distinctive flat, thickened, circumscribed disc.
Measurements in mm: snout to vent length 186.2; tail length 11.1; maximum body width 7.1; head length 7.7; head width 4.6; head depth 3.8; internarial distance 1.26; interorbital distance 2.81 at the front of the ocular and 3.41 at the rear of the ocular; eye to snout tip distance 3.67; eye to lip distance 1.12; snout-parietal distance 3.0; posterior end of rostral to posterior end of parietal distance 5.14; tail shield length 9.91; tail shield width 6.12; tail shield depth 3.67; parietal scale length 3.32; parietal scale width 2.71; frontal scale length 1.45; frontal scale width 1.78; ocular scale length 1.41; prefrontal scale length 1.17; midbody ventral scale width 1.33; midbody basal coastal scale width 2.31.Scalation: Rostral visible from above, partly dividing nasals, anteriorly, but posteriorly in contact with one another, behind rostal; part of rostral visible from above subequal to its distance from frontal; nostril piercing nasal, pointed towards rostrum and first supralabial; nasals slightly smaller than prefrontal; ocular scale slightly smaller than nasal and/ or prefrontal; frontal hexagonal, longer than broad; broader anteriorly, posteriorly produced towards a tapering point, wedged deeper within the midline contact of parietals; parietals large, as large as distance between snout tip and anterior end of frontal; ocular scale separating contact between prefrontal and parietal; ocular, in contact with frontal; supralabials 4/4, first supralabial the smallest of all head scales, rectangular; second supralabial subequal to rostral, higher than broad; third supralabial broader than high, higher anteriorly, shorter posteriorly; fourth supralabial the largest, subequal to frontal, smaller than parietal; infralabials 3/4, first infralabial large, second the largest, third/ fourth ones small and elongate; scales overall smooth and glossy, imbricate, cycloid; dorsal scale rows: 17: 17: 17; ventrals 146, 1.5 times as wide as adjacent scale rows; anal scale paired, subequal to ventral scale but larger than subcaudal scale; subcaudals 9, paired; caudal scales across length of tail shield 8; caudal scales across width of tail shield 4; caudals scales with 2-5 keels per scale; tail shield ending with two projecting spurs.
Colouration: Dorsum lustrous dark blackish grey overall; anterior end (head, neck) with a brownish tinge, while the posterior end (tail shield) steely bluish-black; dorsum with very fine, scarce yellow powdering all along the trunk from head to near pre-cloacal region; tail and tail shield devoid of yellow patterning above; a distinct pair of yellow ventrolateral stripes from snout tip till tail shield; the yellow stripes rather evident from infralabials till neck, from where onwards the yellow colouration becomes restricted only to scale borders of the last rows of coastal scales; the central part of coastal scales and

J TT
almost whole of the ventral scales totally black, rarely with any yellow intrusions; thick yellow stripes along subcaudals that widen and meet across the anal shield; tongue dark reddish-brown, darker at the tips; iris black.
Dentition: On each side of the jaw, eight maxillary (upper jaw) teeth and five mandibular or dentary (lower jaw) teeth present; teeth conical, visibly curved inwards, uniform in size throughout, except for the two frontmost teeth that are slightly smaller; diastema absent.
Distribution and Natural History: Uropeltis jerdoni sp.nov. is a poorly-known snake, as this is a so-far unsampled population about which published literature has not dealt with (see Pyron et al. 2016 and references therein).Though Wall had collected this species from "Nandydug" (=Nandi Durga), historical literature during or after Wall's time (e.g., Smith 1943), never stated the occurrence of any uropeltids near about Bengaluru, except U. ellioti that belongs a different species group.The holotype and one paratype were roadkills recently collected from the Ghat road of a hill fort temple -Devarayana Durga.These snakes would have probably been killed the previous night by vehicle plying on the ghat road.A live uncollected adult of unknown sex, measuring about 250 mm total length was sighted in an earthworm farm at the type locality.The snake was dug out from underneath the soil surface by the J TT workmen when we authors (PKG and NSA) were present to document biodiversity.The snake was inoffensive and tried to dig underground when exposed and during photography.It had blue-black dorsum; ventrolateral yellow reticulations; black venter; concave, circumscribed tail disc; scale rows 17:17:17; 143 ventrals; paired anal scale; nine pairs of subcaudals, thereby matching in morphology with the preserved specimens.To the best of our knowledge, the only other uropeltid snake sympatric with the new species is Uropeltis cf.ellioti (Gray, 1858), a distinctly reddish-brown coloured species with evident, convexly-rounded tail shield (Group I of Smith 1943) having a large yellow spot on tail tip (also see Whitaker & Captain 2004;Pyron et al. 2016).The distribution range of Uropeltis jerdoni sp.nov. is a mix of deciduous vegetation distributed within a sprawl of predominantly rocky boulder-dominated hilly terrain (Boraiah & Fathima 1970;Bhaskar & Kushalappa 1995), currently known from two peaks north of Bengaluru -Devarayana Durga and Nandi Durga that are 40 airline km apart.Uropeltis jerdoni sp.nov. is hypothesized to be a primarily nocturnal, worm-eating, viviparous, fossorial snake that is particularly active during rain, like most members of its family (Rajendran 1985).
Comparisons and differential diagnosis: The new species is here compared with the 25 recognized species of Uropeltis from India (see Pyron et al. 2016;Jins et al. 2018;Ganesh & Achyuthan 2020).By having an obliquely truncate tail terminating in a thickened, circumscribed, concave caudal disc covered with multicarinate scales, Uropeltis jerdoni sp.nov.differs from the following 14 species: U. bhupathyi, U. ellioti, U. nitida, U. ocellata, U. dindigalensis, U. beddomei, U. macroryncha, U. woodmasoni (Group-I tail shield of Smith 1943), U. grandis, U. maculata, U. petersi, U. liura, U. pulneyensis (Group-III tail shield of Smith 1943).Further, Uropeltis jerdoni sp.nov.also differs from the remaining congeners (after Gower et al. 2008;Ganesh et al. 2014;Ganesh & Achyuthan 2020) with a thickened, circumscribed, caudal shield categorized under Smith's (1943) Group II A & B as follows (only opposing suite of character states listed): U. arcticeps (southern Western Ghats): dorsal scales lacking a clearly defined yellow scale border; ventral scale counts much lower (127-128); U. bicatenata (northern Western Ghats): yellowish scalloping chain-like pattern across both sides of the body; ventral scale count 130-141; U. broughami (southern Western Ghats): 19 midbody scalerows; rostral scale much produced and ridged with a dorsal keel; dorsum brown with distinct small, yellowblack-edged transverse ocelli; ventral scale counts much higher (181-230); U. ceylanica s. auct.(Western Ghats): anterior dorsum without distinct yellow spots; venter lacking a clearly defined brownish scale border; ventral scale counts much lower (119-146;130 in holotype -Gower et al. 2008); U. macrolepis complex (northern Western Ghats): 15 midbody scalerows; lower ventral scale counts (120-140); dorsum blackish-brown with yellow broken spots forming zig-zag crossbars or annuli or a pair of distinct, thick, yellowish-orange paravertebral stripes extending across most of the body except near neck, where there are two large orange spots; U. madurensis (southern Western Ghats): snout much more rounded in profile; body colour rich brown, dorsal scales with a clearly defined yellow scale border throughout the back, giving a yellow-reticulated appearance; no ventrolateral yellow reticulations, but ventrals with large alternating yellow blotches; ventral scale count 144-157; U. myhendrae (southern Western Ghats): dorsum with brownish-black body, each scale with yellowish posterior border forming more or less complete band or annuli; part of rostral visible from above distinctly longer than its distance from frontal; ventral scales 139-156; U. phipsoni (northern Western Ghats): a pair of yellowish lateral streaks along both sides of the body; part of rostral visible from above distinctly longer than its distance from the frontal; ventral scales 138-157; U. rajendrani (southern Eastern Ghats): ventrals 146-158; rounded snout profile; body deep ochre brown; presence of yellow colouration in the ventral scales; part of rostral visible from above, not much longer than its distance from frontal; U. rubromaculata (southern Western Ghats): presence of two large red caudal spots; much lower ventral counts (127-136); U. rubrolineata (southern Western Ghats): presence of two ventrolateral red stripes; much higher ventral counts (165-172); U. shorttii (southern Eastern Ghats, allopatric): dorsal body brownish or bluish-black, with distinct yellowish annuli or crossbars; ventral scales 137-156.

DISCUSSION
Uropeltis jerdoni sp.nov. is the 26 th species of Uropetlis to be described.Recent descriptions of Uropeltis were either from the Western Ghats (Jins et al. 2018) or the Eastern Ghats (Ganesh & Achyuthan 2020).But in the present case, Uropeltis jerdoni sp.nov. is described from the intervening region -the Bengaluru uplands, that is flanked by both the Western and the Eastern Ghats on either sides.In fact, the only species of shieldtail snake known from regions in India apart from the Western Ghats and the Eastern Ghats, is the apparently 'widespread' U. ellioti (Gray, 1858) reported from most of the hilly areas across the Indian peninsula (Smith 1943;Whitaker & Captain 2004).Thus, Uropeltis jerdoni sp.nov. is a previously unsampled new species of shieldtail snake that has not been reported in literature under any incorrect names.This is in contrast to U. bhupathyi that was long-thought to be and misreported in literature as U. ellioti (see Jins et al. 2018).
Uropeltis jerdoni sp.nov. is described based on two, recently preserved, female road-kills (holotype and paratopotype), six historically-collected specimens and one uncollected, unsexed, live individual (referred J TT material).These materials originate from two, nearby (40 airline km apart) hill ranges -Devarayana Durga (type locality) and Nandi Durga.Shieldtail snakes, especially the diverse genus Uropeltis is a radiation of cryptic species (Cyriac & Kodandaramiah 2017), with each of the constituent species displaying very subtle morphological variations (Gower et al. 2008;Ganesh et al. 2014;Jins et al. 2018;Ganesh & Achyuthan 2020) and occupying small, allopatric geographic ranges (Pyron et al. 2013;Ganesh 2015).In the case of U. jerdoni sp.nov.its nearest related congeners are U. shorttii of Shevaroys that is 200 airline km south off Devarayana Durga-Nandi Durga and U. ceylanica s. lat. of the equally-distant Malnad part of the Western Ghats.
The localities where Uropeltis jerdoni sp.nov.has been recorded, the Bengaluru uplands, is poorly inventoried for biodiversity, especially herpetofauna.T.C. Jerdon was perhaps the foremost naturalist to explore the area in and around Bengaluru, when he described a new gecko Cnemaspis mysoriensis (Jerdon, 1853), over 165 years ago.Uropeltis jerdoni sp.nov. is a humble tribute to his pioneering efforts to inventory and describe the reptiles of Bengaluru.In recent times, five more new reptiles were described from places near Bengaluru-Hemidactylus graniticolus from Harohalli (Agarwal et al. 2011), Hemidactylus whitakeri from Kodalagurki (Mirza et al. 2018), Cyrtodactylus srilekhae from Thathaguni (Agarwal 2016), Hemiphyllodactylus jnana from Kodigehalli (Agarwal et al. 2019), and a snake Lycodon deccanensis from the same Devarayana Durga (Ganesh et al. 2020).These works well indicate that further explorations around Bengaluru would reveal further reptile diversity, endemism and novelties.As for the genus Uropeltis, the recent taxonomic research and increase in diversity hints that more studies in this and other genera of uropeltid snakes will add to the growing body of literature on their increased taxonomic diversity.
Profiles of Holotype BNHS 3562 of Uropeltis jerdoni sp.nov.: A-head left view | B-head right view | C-head ventral view | Dhead dorsal view | E-tail ventral view | F-tail dorsal view | G-close up of ventral scales.© N.S.Achyuthan.
J TT