Tympanocryptis cephalus Günther, 1867

Doughty, Paul, Kealley, Luke, Shoo, Luke P. & Melville, Jane, 2015, Revision of the Western Australian pebble-mimic dragon species-group (Tympanocryptis cephalus: Reptilia: Agamidae), Zootaxa 4039 (1), pp. 85-117 : 93-97

Tympanocryptis cephalus Günther, 1867

Coastal pebble-mimic dragons Figs. 5 A, 6, 7

Lectotype. BMNH 1946.8.12.78, from Nicol (Nickol) Bay (= Karratha), Western Australia ( Fig. 6 ). Lectotype designation by Cogger et al. (1983).

Diagnosis. Distinguished from other Tympanocryptis by the following combination of character states: presence of two pre-cloacal pores, lack of longitudinal stripes on the dorsum, presence of enlarged scales with raised spines arranged in transverse rows, snout straight or convex, scales on snout with low keels, rostral width ~3–4 times height, weakly defined row of enlarged scales at anterior and dorsal edge of thigh, scales on dorsal surface of thigh not aligned, ventrals with low keels, and irregular dark blotches along midline of dorsum.

Description. A small to medium-sized (to 64 mm SVL), moderately rotund dragon; head small with angular snout; moderately short neck; moderate gracile limbs, ArmL%SVL—0.193 (0.014), LegL%SVL—0.256 (0.020); gracile digits; short tail. Head small, HeadL%SVL—0.311 (0.017), HeadW%SVL—0.241 (0.014), HeadD%SVL—0.159 (0.01); neck ~3/4 of widest part of head; snout moderate, SnoutL%HeadL—0.253 (0.018), convex when viewed in profile, narrowing to blunt tip; canthus well-defined and angular, forming continuous line with projecting brow ridge; nostril located below canthus in enlarged scale, opening projecting dorsally and posteriorly; eye moderate, EyeL%HeadL—0.257 (0.023); eyes with laterally-projecting scaly eyelids forming a weak fringe, not projecting past brow when viewed dorsally; tympana covered with fine scales, partly encircled by scattered enlarged scales with low spines; rostral scale ~3–4 times wider than tall; scales on snout with low keels, not tending to align; scales on crown slightly larger with lower keels; scales on back of head small with variablysized low scattered spines; 11–15 supralabial scales, above which are 3–4 rows of scales with low horizontallyorientated keels continuing to temporal region, uppermost row slightly more prominent and forming edge of eye socket; loosely defined cluster of enlarged spines at posterior and lateral corners of head; mental twice as long as wide; 12–15 infralabials with low keels; 4–5 rows below infralabials of scales with low keels parallel to angle of jaw, gulars kite-shaped with low keels; prominent gular fold.

Body dorsoventrally compressed, ovoid in dorsal view with widest part 2–3 times wider than neck and pelvis; TrunkL%SVL—0.453 (0.037); dorsum with heterogeneous scales in size and shape; largest scales with sharp spines angled 10–30° posteriorly and ~5–7 times larger than smallest scales; large dorsal scales with spines tending to occur in transversely-aligned clusters of ~4–6; smaller dorsal scales rugose or with weak, low keels, arranged in loose whorls around clusters of large, spiny scales; enlarged dorsolateral edge of pelvis always with a cluster of enlarged spines at posterior edge where skin is fused to bone; ventral scales homogeneous, approximately half the size of large dorsal scales, kite-shaped and arranged in diagonal rows, median keel low with posterior spine at most only slightly protruding past posterior edge of scale.

Limbs covered in elongate kite or teardrop-shaped scales with prominent low keels, spine usually protruding beyond distal edge of scale; arms and legs moderately long, ArmL%SVL—0.193 (0.014), LegL%SVL—0.256 (0.020); scales on dorsal surface of upper arm large with keels tending to align; keels of dorsal scales on lower arm aligned to form lines that extend to hand and fingers; keels of ventral scales mostly aligned forming lines along the length of the arm to palmar surfaces; scales on underside of digit with two rows of spiny lamellae; claw long and recurved, lower portion terminating with circular opening, upper portion continuing past ultimate lamellae to form sharp claw; finger length: 4> 3> 2 = 5> 1. Scales on legs kite or teardrop-shaped; tops of upper and lower leg with large heterogeneous non-aligned scales, otherwise with low keels that align and extend to feet and toes; scales at insertion of limb to body small and lacking keels, abrupt transition from dorsal to posterior edge of leg (large to small scales); scales on underside of toes as for fingers; toe length: 4 >> 3> 2 = 5> 1.

Pre-cloacal pores 2, set among 3–5 scales, positioned anterior to distal edges of cloaca, midway between anterior and posterior edge of leg; tail relatively short and thin with blunt tip, TailL%SVL—1.231 (0.134); scattered enlarged scales aligning along most of the length of the tail to terminus; lateral surfaces of tail base with short protruding spines.

Coloration and pattern. In life (based on WAM R165248; Fig. 5 A), ground color light to medium brown, vertebral zone of dorsum with 5 tan irregularly-shaped blotches, series continuing to tail as 10–12 pale irregular bands that alternate with ground color, anterior edge of pale bands edged with black; limbs and digits weakly banded; sides of head below eye and naris paler than top of head, with 3–4 darker lines originating from eye to supralabial scales; nuchal region dark brown with three discrete pale bands. In preservative, the warmer hues tend to be lost, leaving a light brown ground color with pale tubercles in higher contrast ( Fig. 7 ); by-catch individuals from glycol invertebrate pits tend to be a dark reddish-brown; WAM R165248 ( Fig. 5 A) and WAM R 161047 appear to have vestiges of longitudinal lines; ventral surfaces pale, occasionally with reddish hue, ventral surface of tail lacking dark bands.

Habitat. Very little data exists on this species, but WAM specimen records indicate its occurrence on red rocky loams or clayey substrates, with Triodia, Snakewood ( Acacia xiphophylla ), or annual grasses as the vegetation type. The Pilbara Biodiversity Survey data also indicated a preference for clayey substrates ( Doughty et al. 2011).

Distribution. Confined to a coastal strip near Karratha and Roebourne in the Pilbara region, Western Australia ( Fig. 1 ). The westernmost record is from Mardie Pool, with the easternmost record 200 km distant at Balla Balla Creek (~ 50 km east of Roebourne). Records occur from the coast and inland to about 40 km. Thus the total range of this species at current estimate is approximately 5,000 km 2.

Etymology. The word cephalus is a Latinized version of the Greek cephalos, meaning ‘head’. Presumably this refers to the distinctive short head of this this species relative to other Australian agamid lizards at the time of description, as no explicit etymology was given. We maintain usage of cephalus as a noun in apposition, in contrast to occasional use of cephala (e.g. Storr 1964, 1982).

Comparisons with other species. Tympanocryptis cephalus is most likely to be confused with T. gigas , T. fortescuensis sp. nov., and T. diabolicus sp. nov., as these species all occur in the Pilbara or Gascoyne regions. Tympanocryptis cephalus is distinguished from T. gigas by possessing enlarged dorsal scales in short transverse rows on the dorsum (versus scattered spines not arranged in rows), scales on snout with low keels (versus rugose or with feeble keels), ventrals keeled (versus smooth), and smaller average body size.

From T. diabolicus sp. nov. and T. fortescuensis sp. nov., T. cephalus is distinguished by generally longer transverse rows of enlarged scales on dorsum (5–7 versus 2–5), slightly wider rostral (3.5 versus 3.0 times wider than high), enlarged row of scales at front of thigh not forming a conspicuous ridge, and usually possessing a series of dark blotches along the midline; further distinguished from T. diabolicus sp. nov. by scales on snout possessing low keels (versus rugose scales or feebly keeled) and scales on upper thigh not aligned.

From the more geographically distant T. pseudopsephos sp. nov., T. cephalus is distinguished by straight or convex snout (versus concave), low keels on snout (versus rugose or feebly keeled scales), rostral scale 3.5 times wider than high (versus 2 times), enlarged row of scales at front of thigh not forming a conspicuous ridge, keels on dorsal surface of upper arm not aligned, ventrals with low keels and spine protruding past posterior edge of scale (versus smooth with no spines), and more brownish coloration with vertebral series of irregular dark blotches.

Remarks. A theme during the history of this taxon is that most treatments and field guides focused on the more widely distributed Goldfields form, and indeed similar forms from South Australia and Queensland as well. The two syntypes of T. cephalus were collected from Nicol (= Nickol) Bay, where the town of Karratha is today, by F. Duboulay. For many years, the two syntypes were the only specimens from this region, and they were located at the Natural History Museum in London. Mitchell (1948) had no access to specimens from the Pilbara, but had photographs of the syntypes supplied by the Natural History Museum and provided some comments on these. Storr (1964) had access to only a single specimen (WAM R12495 from Mardie), and nearly 20 years later, only eight further specimens (but representing all three Pilbara species) were available to Storr (1982) for his taxonomic comments.

The lectotype ( Fig. 6 ) is in agreement with specimens from the Pilbara coast by possessing dorsal blotching along the midline and low keels on the scales of the snout and ventrum, and ridge of scales on leading edge of thigh not well defined. The genetic analyses clearly indicate the coastal Pilbara population is the basal split within Western Australian T. cephalus senso lato.

As mentioned above, very few specimens of this species were available for study to Mitchell and Storr. Based on records in the WAM collections, only five specimens of true T. cephalus were available up to 1985. Since then, only one further specimen had been collected until the WA Department of Environment & Conservation’s (now Parks & Wildlife) Pilbara Biodiversity Survey (PBS) from 2002–2007 ( McKenzie et al. 2009; Doughty et al. 2011). This survey employed many pitfall trap arrays, and for the first time sufficient material was captured to provide the tissue samples necessary for the analysis of Shoo et al. (2008) and this study, as well as the specimens examined for morphology in this study (the same applies to the other two Pilbara species described below). This is perhaps due to the habits and predator evasion tactics employed by these species. Owing to the need to remain motionless to avoid predators individuals are not likely to move when a human approaches and so they remain undetected. In contrast, short movements are likely to occur regularly for thermoregulation and foraging, and when pitfall traps with low mesh fencing are employed, capture rates can be relatively high.

Figures

FIGURE 1. Distribution map of the Tympanocryptis cephalus species-group in Western Australia.

FIGURE 5. Photographs in life of Tympanocryptis cephalus species-group members (left to right, top to bottom): a) T. cephalus (photo—G. Harold), b) T. gigas (M. Peterson), c) T. diabolicus sp. nov. (M. Peterson), d) T. fortescuensis sp. nov. (G. Harold), e) T. pseudopsephos sp. nov. — plain form (G. Harold), f) T. pseudopsephos sp. nov. — with charcoal wash on dorsum (G. Harold).

FIGURE 6. Lectotype of Tympanocryptis cephalus (BMNH 1946.8.12.78) (scale bar = 1 cm).

FIGURE 7. Variation in Tympanocryptis cephalus (scale bar = 1 cm).