General morphology: A small (SVL ~40 mm) and rotund Ctenophorus with an extremely blunt head (Figure 5b) and a wide neck that is only weakly defined when viewed from above (Figure 5a). The gular fold is prominent (Figure 5f,h). There is often a distinctive, continuous or broken, lateral fold/ridge from above the shoulder that runs along the dorsolateral zone, following the outline of the body, before abruptly terminating above the hind limb (Figure 3). Limbs are short and slender and the tail is short and blunt-tipped (Figure 3). Finger length 4>3>2>5>1 and toe length 4>3=5>2>1. Claws are long and recurved.

Dorsal and lateral colour: Light yellow-brown infused with rich red hues (particularly near the midline) and yellow (particularly on enlarged tubercles) (Figure 3). The crown and nape usually have some darker blotches (Figure 5a). Upper labials have a slight indication of darker mottling (Figure 3, Figure 5b,d). The zone below the dorsolateral ridge in some individuals is light to darkly pigmented in dark olive to black, which extends to the lateral surface of the tail, where it tends to break up into distinct blotches or bars (Figure 3). There are scattered dark spots on the upper surfaces of the limbs (Figure 3).

Ventral colour: Pale off-white that contrasts with the dorsal colour in the lateral regions. Gular region is pale off-white to lightly stippled with black, especially near the posterior edge. Males displayed breeding colours on their venter during the late dry season (September, Figure 5h), whereas no ventral markings were observed in the cooler mid-dry season (June, Figure 5f). In individuals with breeding colouration, a short and narrow faint grey to black triangular chest patch may be present that extends posteriorly from the gular fold to approximately the level of the axilla. The gular fold area, central venter, anterior thighs, and tail are light yellow with the outer edges of these yellow areas being a darker shade of orange. The outer venter occasionally has irregular dark mottling near the lateral edge (Figure 5h).

Pores: Total 10–16. Pre-cloacal pores 4–6, with no pores near the midline and 2 or 3 grouped together either side of the midline. Femoral pores are widely spaced in the middle to anterior of the thigh, extending across the thigh almost to the knee. All pore perforations are small and positioned in the middle of the scale (Figure 5g).

Dorsal scales: Scattered large tubercles with raised keels on the dorsal and lateral surfaces, encircled by slightly enlarged and raised scales (Figure 5c,e). However, large tubercles in the vertebral zone are not surrounded by a ring of slightly enlarged scales. Other dorsal scales are relatively flat and irregularly shaped with moderate keels that align to form longitudinal ridges. Some scales are enlarged and have more prominent keels, especially on the lower arm. Rows of keels continue from limbs onto digits. Scales on the lateral surfaces are smaller than on the dorsum, raised, and occur in regular diagonal rows. Legs have enlarged keeled scales that are more prominent and numerous on the upper thigh. Keels on the tail scales form raised ridges, as for limbs and digits, with occasional scattered enlarged tubercles on lateral and dorsal surfaces.

Ventral scales: Ventral scales are diamond-shaped and slightly overlapping. They have a low, weak keel and a posterior spine that does not protrude past the posterior edge of the scale. Gular scales are small and juxtaposed with feeble keels. Scales on the gular fold are smooth and reduced in size relative to ventral scales. Lower surfaces of arms: as for dorsal surfaces but with lower keels. Palmar and plantar scales have sharp keels. Subdigital scales have two rows of keels forming longitudinal mucronate ridges.

Head and nuchal scales: Scales on top of the head are rugose with low, irregular keels. Nuchal area has slightly larger rugose scales around the pineal gland and scale. Scales posterior to the crown/on the neck have scattered conical tubercles. Brow has two rows of scales with aligned longitudinal keels. Eyelid with ~5 rows of scales above the aperture and 6–7 scales below. Projecting scales form a fringe where the upper and lower eyelids meet. There are scattered enlarged tubercles between the eye and the tympanum and one or two moderately enlarged tubercles with horizontal keels between the eye and the corner of the mouth. The nostril is slightly below the canthus rostralis: it perforates an enlarged scale near the dorsal edge and opens dorsally. The rostral scale is small and triangular with dorsal edges that are usually concave to slightly convex. The snout is extremely blunt, nearly vertical. Posterior supralabial scales have prominent vertically-oriented keels. Infralabial scale row has horizontal keels on the ventral part of each scale, with the dorsal half smooth where the supralabial denticulate row is in contact when the mouth is closed. Scales in the parinfralabial row are rectangular with horizontal keels in the centre of each scale: the next 3–4 rows of scales are slightly enlarged with feeble keels. The mental scale is triangular with the ventral edge tapering to a blunt point and extending 1.5–2.0 times beyond the first infralabial scales.

Known distribution: Recorded sparsely from the Kalkarindji area of the Northern Territory to the Hann River area of the Kimberley in Western Australia, specifically at Kalkarindji, Kandimalal, near Hall’s Creek, and in the vicinity of Mornington Wildlife Sanctuary in the central Kimberley. Although it is currently unclear how far north, east, or west this cryptic species may occur, the northern boundaries of the Great Sandy Desert to the south and Tanami Desert to the southeast correlate with species boundaries for several other reptiles (see Pepper et al. 2011a,b; Oliver et al. 2014; Maddock et al. 2015). The extent of occurrence (EOO) is 55,600 km^2^ based on a minimum convex polygon boundary around all point records from 1979–2023.

Habitat: Consistently detected in mapped (DMIRS 2019) and observed areas of exposed bedrock with loose surface rock and a sparse to dense cover of Triodia spp. (spinifex) (Figure 6) (Storr 1981). Observed sheltering in shallow scrapes under thick surface rock – and occasionally cow manure – and within sparse to thick clumps of spinifex. Often detected at the top of ridges and rises, where exposed bedrock plates fracture into smaller rocks and tiles that provide an abundance of shelter sites. Mapped (DMIRS 2022) and observed geology and surface structure varies among regions and sites:

1) Kandimalal: the two detections were from an area with rocks and rubble that has weathered from a complex intermixture of Miocene laterite, Devonian sandstone and meteorite shale balls (Shoemaker et al. 1997). Described by Storr (1981) as a ‘lateritic plain’.

2) Halls Creek area: Mapped as the Olympio formation of metasedimentary sandstone, siltstone, mudstone and shale. Described by Storr (1981) as a ‘stony hillside’.

3) Central Kimberley Throssell Shale formation: Micaceous siltstone and shale. Most sites have a patchy to continuous cover of loose flat gravel and pebbles over skeletal soil with scattered flat or rounded boulders and cobbles. Also detected in areas with expansive broken sheets of large, flat siltstone and shale tiles (Figure 6).

4) Central Kimberley Wunaamin Miliwundi Sandstone formation: Quartz sandstone pavement with a scattering of cobbles and boulders (Figure 6).

Spinifex cover at recorded sites varies from low, sparse, and open patches to medium-sized hummocks and swathes (Figure 6). Dominant spinifex species include Triodia aeria, T. bitextura, T. epactia, T. pungens, T. intermedia and/or T. wiseana. Sarga spp. (native sorghums) may occur where soil has developed and accumulated.

Most recorded sites in the central Kimberley have a sparse to very sparse tree or low shrub layer (Figure 6), which may be dominated by Eucalyptus brevifolia, Corymbia collina, Lysiphylum cunninghamii, Grevillia pyramidalis, Acacia spp., Grevillea spp. and/or Hakea arborescens. Eucalyptus brevifolia is the dominant canopy species at the Wunaamin-Miliwundi Sandstone site and it is mapped as the dominant canopy species at the Halls Creek, Kandimalal, and Kalkarindji sites (Fox et al. 2001).

Area of occupancy: Three estimates of the potential area of occupancy of C. aurita have been mapped and derived (Figure 7) based on habitat attributes where this species has been recorded. The extent of the five mapped ecosystem types where C. aurita has been detected (dominated by Triodia spp. and E. brevifolia, Supplementary data 3) covers an area of approximately 62,000 km^2^ (black shading in Figure 7). However, there are nine additional mapped ecosystem types in the general area that are also dominated by Triodia spp. and E. brevifolia (Supplementary data 3; dark grey in Figure 7) adding another 51,000 km^2^ of inferred habitat. A further five mapped ecosystem types are dominated by Triodia spp. and they occur in rocky areas with or without E. brevifolia as a dominant or sub-dominant species (Supplementary data 3; pale grey in Figure 7). As such, C. aurita has the potential to occur across an area of 168,000 km^2^ or more in the Kimberley and Victoria-Daly regions of northwest Australia.

Behaviour: Frequently detected sheltering under rocks during cooler parts of the day. Active when temperatures are warm, basking on small rocks, cow manure, termite mounds, and at the base of spinifex hummocks. May shelter primarily within spinifex during the hottest part of the day when ground temperatures can exceed 50°C.

Uses pebble mimicry as a camouflage defence when disturbed (Figure 8). Several individuals were observed running a short distance before crouching and pulling their limbs into their body to expertly mimic a pebble. This behaviour can occur even when there is a suitable rock or spinifex clump nearby that offers shelter. Some individuals have been observed darting from the base of a spinifex clump to shelter under a rock.

Density: May occur at variable densities across its distribution. Ten individuals were detected within a 1 ha area during 0.5 person-hours of searching at one site, whereas at another site one individual was observed in a 500 ha search area during ten person-hours of searching. Two individuals were detected sheltering under the same large rock on successive days at a Throssell Shale site (Figure 8). A comprehensive search on the second day by four ecologists revealed two further individuals under different rocks within 150 m of the original capture point, and a fifth individual 1.2 km away. Three individuals were found under different rocks during three hours of comprehensive searching at a different Throssell Shale site (MB, September 2016), an approximately 200 ha area of small, flat pebbles with scattered round surface cobbles and patches of Triodia aeria.

Detectability: Cryptic and elusive. Can be difficult to detect even by experienced herpetologists. Despite many field trips by experienced herpetologists (including JM & SM), C. aurita has not been seen again at its original collection sites (Kandimalal, Halls Creek, and Kalkarindji). Numerous searches by experienced herpetologists (MB, JP et al.) at the same time of year as initial detections have failed to re-detect this species in subsequent years at multiple sites in the central Kimberley. Eight individuals were observed during targeted active searches at a site in the central Kimberley where multiple comprehensive targeted searches had failed to detect this species in previous years.