The Re-discovery of live populations of Cnemaspis tropidogaster (Boulenger, 1885) (Sauria, Gekkonidae) from Sri Lanka after 120 years Author Thasun Amarasinghe, A. A. Author Campbell, Patrick D. Author Madawala, Majintha B. Author Botejue, W. Madhava S. Author Gabadage, Dinesh E. Author Silva, Anslem De Author Suranjan Karunarathna, D. M. S. text Zootaxa 2016 4200 3 395 405 journal article 10.11646/zootaxa.4200.3.5 b81a2281-7845-4260-b950-496630402a78 1175-5326 187086 32705064-4502-4C06-B2D3-6E00FD233D8C Cnemaspis tropidogaster ( Boulenger, 1885 ) ( Figs. 1–4 ; Table 1 ) Lectotype . (designated by Manamendra-Arachchi et al . 2007 ). Male, BMNH 71.12 .14.49, 31.4 mm SVL ,“ Ceylon [= Sri Lanka ]”, Presented by W. Thwaites. Other examined material. Adult males, NMSL 5152, 30.8 mm SVL ; adult female, NMSL 5151, 30.9 mm SVL both from Maligatenna , elevation 78 m asl.; adult males, NMSL 5159, 30.5 mm SVL , NMSL 5157, 29.9 mm SVL and two adult females, NMSL 5970, 30.4 mm SVL , NMSL 5974, 30.9 mm SVL , from Pilikuttuwa , Gampaha District , elevation 52 m asl., collected by Anslem de Silva on 22nd November 2005 . Diagnosis. Cnemaspis tropidogaster differs from its Sri Lankan congeners by a combination of the following characters: paired postmentals separated medially by a small scale; each postmental bounded by 3–4 scales including medial scale; ventrals, 133–146; 21–25 scale rows across venter at midbody; 92–98 scales around midbody; dorsal scales heterogeneous; 4–6 spine-like tubercles present on flanks; ventrals including pectoral, abdominal, precloacal and tail base area, carinate and imbricate; 2–3 precloacal pores; 3–5 femoral pores (each side); no enlarged tubercles on tail base; seven supralabials to angle of jaws; 18–19 subdigital lamellae on digit IV of pes. Description (based on four adult males, including the lectotype ). 29.9–31.4 mm SVL; head moderately large, elongate, narrow, distinct from neck; snout elongate, slightly concave in lateral profile; snout-eye length greater than eye diameter; eye diameter smaller than tympanum-eye length; canthus rostralis weak, 10–11 canthus rostralis scales; interorbital distance broad; eye large, pupil rounded; ear-opening deep, taller than wide; granules on snout keeled, larger than those of occipital region; scales on interorbital, supercilium and gular region granular; rostral concave, partially divided by a medial groove and a small internasal, postero-ventrally in contact with first supralabial; nasals separated by two enlarged supranasals and a single internasal scale, not in contact with supralabial; nostrils round in shape, dorsally orientated; nasal in contact with two postnasals, the lower one in full contact with first supralabial. Mental subtriangular, fairly truncate posteriorly, as wide as long, extending posteriorly to half level of 1st infralabial, postero-laterally in contact with two enlarged postmentals; postmentals bordered posteriorly by 3–4 smooth scales including medial scale; gular scales rather rounded; one scale row separates orbit from supralabials; seven supralabials (3rd largest, 6th at mid-orbital position); seven infralabials decreasing in size towards angle of jaw; ear opening oval; scales on ventral surface of neck carinate and imbricate; 20–22 scales between anterior margin of ear opening and posterior margin of eye. Body slender, elongate; mid-dorsal granules, some fairly spinous, generally heterogeneous, intermixed with keeled, enlarged tubercles; scales on dorsum at midbody larger than those on ventral body surface at the same level; six spine-like tubercles on each flank; pectoral and abdominal scales not enlarged, strongly carinate, pointed, imbricate; approximately 23–25 ventral scales across midbody; 95–98 scales around midbody; ventro-lateral scales pointed and granular; 2–3 precloacal pores; 3–5 femoral pores (each side); scales around vent imbricate and strongly carinate, each scale with a single keel; ventrals 133–146. TABLE 1. Morphometric and meristic characters for C. tropidogaster ; NA, not applicable.

Character	male ( n = 3)	female ( n = 3)	Lectotype male
Range Mean ± SD	Range Mean ± SD
ED	1.5–1.8 1.7 ± 0.15	1.4–1.5 1.4 ± 0.1	1.6
EN	3.2–3.5 3.3 ± 0.17	3.1–3.3 3.2 ± 0.1	3.3
ES	4.3 4.3 ± 0	4.2–4.4 4.3 ± 0.1	4.1
IN	1.5 1.5 ± 0	1.4 1.4 ± 0	1.1
IO	1.0–1.2 1.1 ± 0.11	1.1 1.1 ± 0	0.8
TYE	2.8 2.8 ± 0	2.6–2.7 2.7 ± 0.1	2.9
HL	9.2–9.3 9.2 ± 0.06	8.6–8.9 8.8 ± 0.1	8.6
HW	5.2–5.4 5.3 ± 0.10	5.3–5.4 5.3 ± 0.1	5.5
SVL	29.9–30.8 30.4 ± 0.46	30.4–31.7 31.0 ± 0.6	31.4
UAL	4.2–4.5 4.3 ± 0.17	3.7–3.8 3.7 ± 0.1	3.7
LAL	4.4–4.8 4.5 ± 0.23	4.5–4.8 4.6 ± 0.1	4.4
PAL	4.4–4.8 4.6 ± 0.21	3.8–4.1 3.9 ± 0.1	3.7
FL1	1.9–2.0 1.9 ± 0.06	1.7–2.0 1.9 ± 0.1	2.2
FL2	2.0–2.3 2.2 ± 0.17	2.3–2.5 2.4 ± 0.1	2.7
FL3	2.3–2.8 2.6 ± 0.26	2.4–3.1 2.8 ± 0.4	2.9
FL4	2.6–3.0 2.8 ± 0.20	3.1 3.1 ± 0	2.9
FL5	2.0–2.6 2.3 ± 0.30	2.6–2.8 2.7 ± 0.1	2.6
FEL	5.4–5.6 5.5 ± 0.11	5.4–6.0 5.6 ± 0.3	6.4
TBL	5.7–6.5 6.0 ± 0.43	5.3–5.9 5.6 ± 0.3	6.8
FOL	5.8 5.8 ± 0	5.5–5.8 5.7 ± 0.1	6.0
TL4	3.7–3.9 3.8 ± 0.10	3.8 3.8 ± 0	3.6
SUP	7	7	7
INF	7	7	7
BET	20–22	20–21	20
CR	10–11	10–11	10
MBS	95–98	92–96	95
MVS	23–25	21–25	25
VEN	133–146	138–142	133
SDL	18–19	18–19	18
precloacal	2	NA	3
femoral	3	NA	4–5

Forelimbs moderately long; hind limbs relatively long; tibia longer than femur. Dorsal scales on both fore and hind limbs carinate; ventral scales on upper arm and lower arm carinate; scales on dorsal and ventral surface of thigh and shank carinate. Digits elongate, slender, all bearing slightly re-curved claws; subdigital lamellae entire, un-notched; subdigital lamellae on toe IV, 18–19; inter digital webbing absent; relative length of fingers: IV> III> II> V> I and of toes: IV> V> III> II> I. Caudal scales arranged in segmented whorls; tail base distinctly swollen and scales on ventral tail base smooth, intermixed with some slightly keeled scales. Dorsal scales on tail base with homogeneous scales and afterwards pointed and keeled; no enlarged tubercles on tail base; enlarged conical tubercles present on lateral side of the tail; a single conical post-cloacal spur present on each side; median row of subcaudals not comparatively enlarged, pointed, carinate, bordered by a pair of smaller flanking scales of similar size. FIGURE 1. Cnemaspis tropidogaster male NMSL 5159, ( a ) dorsal head; ( b ) lateral head; ( c ) ventral head; ( d ) dorsal body; ( e ) ventral body; ( f ) subdigital lamellae on hind limb; ( g ) lateral body; ( h ) ventral tail and pelvic area. FIGURE 2. Cnemaspis tropidogaster lectotype male (BMNH 71.12.14.49), ( a ) dorsal head; ( b ) lateral head; ( c ) ventral head; ( d ) dorsal body; ( e ) ventral body; ( f ) toes and subdigital lamellae; ( g ); lateral body ( h ) ventral tail base and pelvic area (photo courtesy of BMNH). FIGURE 3. Habitat of Cnemaspis tropidogaster at Pilikuttuwa, Gampaha District, Sri Lanka ( a ) forest foot path; ( b ) caves and rock boulders; (c) egg-nesting habitat with an adult female (not collected; Photo: W.M.S. Botejue) FIGURE 4. Current distribution pattern of Cnemaspis tropidogaster , with other lowland (wet and intermediate zones) and central highland Cnemaspis species. Type locality of each species marked with black dot centrally. Colour in life (based on Fig. 3 c). Dorsal surfaces of body, limbs and tail light brown; an oblique dark brown line with dull whitish spots in the interorbital area. A ‘W’ shaped, dark brown patch on the neck with a median dull white spot; six diffuse 'A' shaped, yellowish markings on trunk dorsum, and reddish brown band along the tail. Lateral sides of limbs and body golden yellowish, bright yellow lateral conical tubercles on tail and trunk. Canthus rostralis line present with a black stripe; two scattered, dark brown, postorbital stripes radiate from eyes; supraciliaries and nasals yellow; supralabials and infralabials with a median yellowish spot. Ventral surfaces of head, body and limbs beige to dull yellow with scattered, tiny black spots, ventral tail grey. Variation . The colouration of female specimens is similar to the males. The variation of morphometric and meristic characters are shown in Table 1 . Comparisons . Cnemaspis tropidogaster is most similar to C. silvula , C. pulchra , C. upendrai , C. clivicola , and C. pava . The diagnostic characters are listed in Manamendra-Arachchi et al . 2007 . All six species have at least some tubercles on the tail base, and C.silvula , C. pulchra , C. upendrai , and C. pava can also be distinguished by haying a distinctly enlarged median row of subcaudal scales. Other congeners from Sri Lanka have more obvious morphological characters that distinguish them from C. tropidogaster . Compared to C. tropidogaster , C. kandiana , C. scalpensis , C. podihuna , C. gemunu , C. alwisi , C. kumarasinghei , C. molligodai , C. retigalensis , C. phillipsi , C. punctata , C. latha , C. kallima , C. menikay , C. amith , and C. rammalensis all have smooth abdominal scales ( vs . keeled). Cnemaspis samanalensis has tricarinate ventrals (unicarinate). Habitat and natural history . Cnemaspis tropidogaster is only known from Maligatenna and Pilikuttuwa forested areas ( Fig. 4 ) and all specimens were found in very dry, shaded, cool crevices of large caves, rock boulders, leaf litter, tree barks, and cement walls. We observed this species from 27 different caves in the above forested areas. In April and May hatchlings were observed along with juveniles and gravid females carrying two or three eggs. Eggs were pure white (mean diameter 4.97 ± 0.03 mm ), almost completely round in shape and with a slightly flattened side which was often the side attached to the substrate or the side in between each egg. Usually 8– 12 geckos (individuals per man-hour of search) were recorded at both locations. Conservation status. The application of the IUCN Red List criteria (2013: version 10.1) shows that C. tropidogaster is Critically Endangered (CR) and it is restricted to an area of occupancy (AOO) < 1 km 2 and the extent of occurrence (EOO) < 2 km 2 in the Wet Zone [Applicable criteria is B2-b (iii)] assuming that the two sites documented here are the extent of the species’ range. However, confirmation that the species does not occur in adjacent areas should be sought before a final determination on conservation status is made. See the map ( Fig. 4 ) for known distribution.