Koch’s barking gecko

Afrikaans: Duin blafgeitjie

Ptenopus garrulus maculatus (in part) – FitzSimons (1943: 13)

Ptenopus garrulus (in part) – Brain (1962)

Ptenopus kochi Haacke, 1964: 1, pls. I–V, Haacke (1969: fig. 3a), Haacke (1975: 227)

A description of the morphology of this species is not revised here. An updated diagnosis, some natural history observations, and a formalised bioacoustic call description, are provided.

TM 28809, adult male, collected from “Gobabeb, South West Africa [Namibia], central Namib Desert (–23° 37’ South, 15° 03’ East, 408 m)”, by Wulf D. Haacke in October 1963.

TM 24993–4, collected by Charles Koch in October 1957; TM 25880–1, 25887, 25889–90, collected by Charles K. Brain in May 1959; TM 28442–6, 28448–55, 28625–7, collected by Wulf D. Haacke in October 1963. Locality: same as holotype.

See Table S1 for vouchered (1) and unvouchered photographed (16) specimens, DNA samples (16 available, 7 sequenced), and call recordings (20) included (total n = 34).

The largest Ptenopus (SVL max. 65.1, mean 61.1, n = 23) with the longest tail (TL 83% [range 63–96%] of SVL, n = 23), an overall plump appearance, extensive toe fringes and elongated fringed scales on the fingers compared to other species. It is distinguished from all other congeners by a combination of the following characters: Body and head scales finer than other species, with MBSR 187–210 (vs. generally <200 for P. garrulus and <180 for other species); dorsal colour pattern (Fig. 6) finely speckled with a few somewhat enlarged cream or yellow spots (somewhat similar to P. garrulus, vs. banded in P. carpi and P. sceletussp. nov. and large paired, oval marking in P. maculatus, P. circumsyrticussp. nov., P. kenkensessp. nov., and northern populations of P. adamanteussp. nov.); the tail, fore- and hind-foot soles are completely or partially pink and unpigmented, whereas the rest of the ventrum is white (similar to P. garrulus except for the tail, vs. immaculate white in P. maculatus, P. adamanteussp. nov., P. circumsyrticussp. nov., and black- or dark grey-speckled in P. kenkensessp. nov.); the yellow pigment in males is not limited to the throat, but extends over the sides of the snout, head, neck, and body (also the case in some P. garrulus). For a more detailed morphological description, see Haacke (1964, 1975).

In life (Fig. 6), colouration varies from bright to dull orange with yellow (especially in males) or whitish spots covering the dorsum, and differing degrees of dark brown spots or reticulations sometimes being present. The tail is spotted or lightly reticulated to somewhat barred dorsally, with a yellow or whitish tip.

Ventrally, animals are immaculate white with pinkish patches on the limbs, soles, and tail. Males have brilliant yellow throats, the colour extending across the entire ventral and dorso-ventral surface of the head and often along the lateral surfaces of the body and legs.

In preservative, the dorsal colours eventually fade to beige, brown, and grey. The brighter colours, especially yellow, eventually fade completely.

(Figs 3, 20H) Consists of 13 notes (range 11–19) uttered in rapid succession with a note rate of 7.90 s^-1 (range 5.98–10.33). Note duration is short (27 ms [range 18–40]) and relatively regular, sometimes slightly shorter as the call progresses, with note 1 duration deviance low (20% [range 6–31]). The inter-note intervals are usually short (98 ms [range 75–130]) and regular with inter-note interval range low (30% [range 14–52]). Median call density is high (0.28 [range 0.20–0.34]) and call duration long (1.6 s [range 1.2–2.4]). The basal frequency is 382 Hz (range 301–492) but very soft and may be inestimable, with harmonic bands louder towards the upper dominant frequencies, the upper at 3.3 kHz (range 3.1–3.88) and a clear lower dominant frequency around 1.4 (range 1.1–1.9) kHz, about half the upper dominant frequency. Frequency appears to remain constant throughout the notes and call, as is the case with the (human) perceived pitch. Bandwidth (90%) is difficult to estimate consistently: approximately 0.7–4.9 kHz.

This species calls more intensively and for a shorter period of time than its sympatric or parapatric congeners (P. carpi, P. maculatus and P. circumsyrticussp. nov.). Calling also commences earlier in the day than these congeners. Call period (mean 74 s) varies greatly, but can be as low as 9 seconds during peak chorus activity. Calling activity is crepuscular, commencing shortly before or at sunset, and ending at nightfall. Sporadic calls may occasionally be heard later at night, in the morning, or on overcast days (Haacke 1969).

Occurs throughout the Namib erg in the central Namib Desert, Namibia, except possibly in the southwestern extreme of this erg in the winter rainfall zone, near Lüderitz. The northern limit of its distribution is generally aligned with the northern extent of the sand sea.

The habitat of this species was initially described as the silts in Kuiseb River and the interdune plains by (Haacke 1964), and was even referred to afterwards as the “interdune barking gecko” (Branch 1998). In reality, this species rarely occurs on the interdunes and its occurrence in and around the Kuiseb River is more coincidental with sandy intrusions from the adjacent Namib erg. Its primary habitat is characterised by the dune plinth, being the base of the dune where the sand is loose, but before it slopes more steeply. “Dune plinth barking gecko” would thus be a more suitable substitute common name. Its occurrence on the interdunes is only peripheral, except in instances where the interdunes sand is very loose, similar in composition to a typical dune plinth. True interdune plains in the Namib erg are occupied primarily, or solely, by P. circumsyrticussp. nov.

Ptenopus kochi occurs parapatrically alongside P. circumsyrticussp. nov. throughout the Namib erg, and in occasional sympatry or occasional syntopy. It also occurs parapatrically with P. carpi and P. maculatus along the Kuiseb River. Some P. kochi do occur on the northern banks of the Kuiseb River, occasionally in sympatry with P. maculatus and P. circumsyrticussp. nov., on sandier patches of the gravel plains. These animals are genetically divergent from those on the southern side of the river.

Breeding generally occurs from September to December (Polakow 1997), but this species may occasionally call throughout much of a year following above-average rainfall. Rainy weather may cause calling activity to cease for days to weeks.

It has previously been characterised as being nocturnal (Haacke 1969), but is more crepuscular. This species is more frequently observed on the surface during the day (e.g., Murray and Lease 2015) than congeners, except perhaps for P. circumsyrticussp. nov.

Courtship appears to be more complex than in other species, and also more variable. Some courtship encounters witnessed by FB were similar to that of P. maculatus, with the female simply approaching the calling male and the male responding with a few calls. The female then scratches rapidly at the burrow entrance to indicate she wants to enter, after which the male retreats backwards and the female follows him inside. In other encounters, the males may exit the burrow and display their gular patch to the approaching female. Yet more variations, including mutual head bobbing and soft vocalisations, have been observed (Polakow 1997). In some cases, the female may inspect the burrow of a potential suitor, entering through one opening and leaving through another soon after, if she (presumably) finds the burrow unsuitable (Polakow 1997). Mating takes place within the burrow, with the male usually uttering something similar to the advertisement call from within the burrow, after a female has entered.

Fighting is common among closely-spaced males. In one observation (FB), a young male charged a much larger calling male at its burrow entrance, from several metres away. The two animals stood belly-to-belly on their back legs, rapidly clawing one-another with their hands for one or two seconds, before the younger male retreated. Scars from bite-marks on the body are commonly observed in both sexes, but more often in males.

The burrows of P. kochi tend to be more complex than other species, often with more than one entrance (Polakow 1997; FB observations). Males, females, and hatchlings share burrows for extended periods of time, and a male and female will generally use separate entrances to the same burrow system. See Polakow (1997) and Haacke (1964, 1978) for additional notes on P. kochi courtship and ecology.

Kalahari barking gecko / Garrulous barking gecko

Afrikaans: Kalahari blafgeitjie / Grondgeitjie

Stenodactylus garrulus Smith, 1849: Append. 6

Ptenopus garrulus (part) – Boulenger (1885: 15), Brain (1962: 3), Haacke (1964: 3)

Ptenopus garrulus garrulus – FitzSimons (1935: 524), Loveridge (1947: 31), Haacke (1975: 214), Haacke (1969: 89)

Ptenopus garrulus maculatus (part) – Mertens (1955: 51), Mertens (1971: 44)

Sir Andrew Smith did not specify the type locality of P. garrulus beyond “sandy districts in the interior of southern Africa” (Smith 1849). Smith did not name specimens in his description, although FitzSimons (1937a) later noted which specimens were the types used by Smith. Smith described type specimens as buff orange, having a spotted colour pattern and poorly or non-barred tail. Smith travelled and collected extensively in Namaqualand and briefly in southern Namibia, prior to his description of P. garrulus (Skelton 2018). These areas contain the three species as presently treated (Figs 2, 5): Ptenopus adamanteussp. nov., P. kenkensessp. nov., and P. garrulus sensu stricto. Of these, only the P. garrulus sensu stricto commonly occurs on orange, sandy substrate, with the body colour to match. The description “inland of South Africa” probably rules out the coastal and near-coastal regions occupied by P. adamanteussp. nov. Hence, the southern Kalahari in the northern or central regions of the Northern Cape Province, South Africa, most likely contain the type locality, making P. garrulus sensu stricto as presently treated, the most likely candidate for this name (see Fig. 2). We therefore elevate the subspecies ‘P. g. garrulus’ to full species, and thereby exclude all species previously referred to as ‘P. g. maculatus’, from P. garrulus sensu stricto.

The morphological characters and colour pattern of the P. garrulus type specimens from the BMNH (1946.8.23.43–49) fall within the range of P. garrulus sensu stricto in this study (except slight deviations in IOS, typically ≤46 for P. garrulus sensu stricto, for two specimens: BMNH 1946.8.23.44 IOS = 48 and 1946.8.23.46 IOS = 47). Specimen BMNH 1946.8.23.47 conforms most unambiguously to these morphological characters (Fig. 4) and typical colour pattern. Hence, we designate BMNH 1946.8.23.47 (male) as the lectotype for P. garrulus sensu stricto.

BMNH 1946.8.23.47, collected from “sandy districts in the interior of southern Africa” (probably Northern Cape, South Africa), by Andrew Smith between 1834 and 1836.

BMNH 1946.8.23.43–46, 1946.8.48–49, same collection details as the lectotype.

See Table S1 for vouchered (11) and unvouchered photographed (13) specimens, DNA samples (33 available, 20 sequenced), and call recordings (23) included (total n = 44 excluding types).

A small Ptenopus (SVL max. 53.4 mm, mean 46.6 mm, n = 22) with a moderate tail (TL 69% [range 50–86%] of SVL, n = 17) and a generally slender appearance. This species is easily distinguished from P. kochi, P. carpi and P. sceletussp. nov. by a combination of the following morphological characters: Toes extensively fringed laterally (similar to P. kochi vs. weakly fringed in P. carpi and P. sceletussp. nov.), with fringe length roughly equal to the breadth of the toe between fringes (vs. fringe length generally less than half toe breadth in P. carpi and P. sceletussp. nov.); having white pigmented ventral scales with some pink, unpigmented scales on the (hand/foot) soles (vs. pink, unpigmented patches also on the tail and limbs in P. kochi, and white entire in P. carpi and P. sceletussp. nov.); having MBSR ~182 (range 156–202, n = 31) (vs. 187–210 in P. kochi and usually <135 in P. carpi and P. sceletussp. nov.). It is further distinct from P. carpi and P. sceletussp. nov. by the nasals being more swollen and the nostrils partially covered by internal projections of the upper labials, and by more-or less speckled dorsal colour pattern (vs. banded pattern in P. carpi and P. sceletussp. nov.); from P. kochi by having fingers laterally fringed with pointed triangular scales (vs. elongated pointed scales in P. kochi).

From congeners previously included under ‘P. garrulus’, P. garrulus sensu stricto is distinguished by: (when live) patches of pink, unpigmented scales on soles (vs. white pigmented in other species and white or slightly pinkish with dark brown speckles in P. kenkensessp. nov. and P. australissp. nov.); having higher RB/RH (~1.12, range 0.95–1.24, n = 10) than P. australissp. nov. (≤0.97); having higher MBSR (~182, range 156–202, n = 31) than P. australissp. nov., P. circumsyrticussp. nov., and P. maculatus sensu stricto (usually <149); dorsal colour pattern relatively speckled, consisting of >2 longitudinal rows of white spots (rows may join to form irregular bands), including 5–7 clear (but not large) dorsolateral pairs of light spots between pectoral and pelvic girdles, with indistinct dark patches not touching the white, or dark patches absent, and indistinct or no patterning on ventro-lateral sides of the face, vs. 4–5 large and distinctive paired, light, ovoid markings interspaced by distinct dark mottled patches usually touching the light markings, and ovoid patterning on ventro-lateral portions of the face being more distinctive in P. kenkensessp. nov. (and some populations of P. adamanteussp. nov. or P. circumsyrticussp. nov.); toe fringes being generally more extensive (with fringe length roughly equal to the breadth of the toe between fringes) than other species (fringe length usually about half toe breadth), but similar to P. australissp. nov.

In life (Fig. 7), colouration varies with substrate colour, as in other species. The colour pattern tends to be speckled or juxtaposed with light, dark, and coloured scales, the lighter (cream or white) speckles usually tending to larger dots or asymmetrical bars in some locations. On the tail, light-and-dark bars are usually evident dorsally. Background colour forms, matching various substrate colours, include brick red, cream or grey, and grey-brown. Ventrally, animals are immaculate white, with faded dark grey or brown patterning around the lateral margins, and pinkish patches on the (hand/foot) palms/soles. Males have bright yellow, heart-shaped gular patches which may extend onto the lateral margins of the head and body, and anterior surfaces of the legs.

In preservative, the lighter colours fade to off-white or beige, and all the darker colours appear various shades of dark brown or grey. The brighter colours, especially yellow, fade completely.

The advertisement call (Figs 3, 20A) consists of 5.4 notes (4–6) uttered in rapid succession, with a note rate of 5.28 s^-1 (range 3.76–7.37). Note duration is short, 22 ms (range 16–31), and highly regular, with note 1 deviance 15% (range 3–27%). Inter-note intervals are short (168 ms [range 114–234]) and regular, although latter intervals can be slightly longer; inter-note interval range 14–34%. Call density relatively low (0.13 [range 0.09–0.18]). Call duration is very short, only 0.86 s (range 0.6–1.2). The basal frequency is ~450 Hz (range 373–508) but very soft and may be inestimable, with harmonic bands louder towards the dominant frequency at 4.2 kHz (range 3.9–5.1); a very slight lower peak frequency band is evident at around or just below half the dominant frequency (roughly 1.3–2.6 kHz, cannot always be reliably estimated). Frequency appears to modulate briefly up, then down by ~100 Hz with each note, or remain constant. The (human) perceived pitch tends to be higher than that of most other species. Bandwidth (90%) is difficult to estimate consistently: approximately 0.6–8.2 kHz, usually wider than other species.

During peak chorus activity, this species calls much more frequently than most other species, the call period being very short: mean 25 s, and as low as 4 s. However, calling activity is short in most localities visited by the author, usually commencing around 20 min after sunset, and lasting only 20–40 mins. Calling may continue later if the moon is above the horizon and conditions are warm, but at an extremely decreased rate. More extensive and intensive choruses, lasting throughout the night, have been recorded shortly after rain (Haacke 1969).

This is the most widespread member of the genus, occurring on sandy soils across much of arid or semi-arid (generally above 150 mm but less than 550 mm annual rainfall) southern Africa, but not in the Namib Desert or the pro-Namib (Fig. 5). It occurs on sandy flats, interdunes, and on vegetated dunes or hummocks, but usually not on steep dunes. In one location (south of Windhoek) it was found occurring on consolidated, silty soils, but this is unusual. Calling is seasonal, taking place in the austral spring, from June in the northernmost extreme, and from September or October in the southernmost extreme of its distribution.

The main calling period coincides with the breeding season (Hibbitts et al. 2012). In the northern limits of its distribution, the main breeding season appears to be between June and early September, but moves later southward (author observations). Most observations on this species were made in the southern Kalahari. Here, they breed between September and November (see Hibbitts et al. 2005, although this includes data from several Ptenopus species in addition to P. garrulus; Hibbitts et al. 2012) and possibly until late January, as hatchlings (<32 mm SVL) have been recorded beween October and May (Haacke 1975), and yolked eggs were recorded within females between late August and late January (Pianka and Huey 1978). Clutch size appears to be one egg, although a second clutch may develop shortly after the first (Hibbitts et al. 2005). One nest was found to contain three eggs, although one or two of these may have been Pachydactylus wahlbergi (Haacke 1975). Minimum recorded hatchling SVL was 23.7 mm (Haacke 1975).

Haacke (1969) observed P. garrulus calling during March at the Nosob Camp, Kgalagadi-Gemsbok National Park, South Africa, after a rainfall event, suggesting that rainfall may cause bouts of calling outside the normal breeding season. During the observed event, the geckos called throughout the night, rather than the normal calling period around sunset. A similar observation was made at the Aha Hills, Namibia (northern portion of the distribution), during early February (Predinger pers. com.). For further notes on natural history and behavioural ecology of P. garrulus, seesee Hibbitts and Whiting (2005), Hibbitts (2006), Hibbitts et al. (2007, 2012).

Damaraland barking gecko / Spotted barking gecko

Afrikaans: Damaraland blafgeitjie

Ptenopus maculatus Gray, 1866: 640, pl. 38: 1

Ptenopus garrulus (part) – Boulenger (1885: 15), Brain (1962: 314), Haacke (1964: 3)

Ptenopus garrulus maculatus (part) – Mertens (1955: 51), Haacke (1969: fig. 3d), Brain (1962: 3, 14), Haacke (1975: 221), Bauer et al. (1993: 127)

Ptenopus maculatus was described from “Damaraland”, collected by Karl (= Charles) J. Andersson. This species was synonymised under the name ‘P. garrulus’ (Boulenger 1885). Subsequently, ‘P. g. maculatus’ was resurrected at subspecies level (FitzSimons 1935), which has been regarded as valid ever since. Most of the earliest references to this subspecies, including the initial synonymisation with ‘P. garrulus’ (e.g., FitzSimons 1935, 1937b; Loveridge 1947), referred to specimens included in our treatment of P. adamanteussp. nov. and P. kenkensessp. nov., but exclude those from the type locality of ‘P. maculatus’. Mertens (1955), Haacke (1975), and several subsequent authors also referred to specimens from a broad geographic range, encompassing that of several species (current manuscript), as ‘P. g. maculatus’.

According to Andersson’s notes and the location names used at the time (see map in Andersson 1855 and 1872), “Damaraland” referred to the Erongo Region in Namibia. Notably, Andersson stayed at “Otjimbingue” [Otjimbingwe] briefly in 1851 and again in the 1860’s for several years, where this species occurs, and where it may have been collected. While staying here, a battle occurred on his property between the Damaras under chief Kamaherero, who had taken shelter at his house after fleeing their enslavement under the “Namaqua” (Nama) tribe Afrikaner, then under leadership of Christian Afrikaner (son of Jonker Afrikaner), leading to a victory by the Damara (Andersson 1875).

From the P. garrulus group species, “Damaraland” contains only P. maculatus sensu stricto (as presently treated). Additionally, the type specimens (BMNH 1946.8.23.53, 54) and a detailed drawing from the type description match the colour pattern and appearance of P. maculatus sensu stricto. Finally, the morphometric measurements of the BMNH types generally support this classification, and most unambiguously so for BMNH 1946.8.23.54 (Fig. 4). Hence, this specimen is assigned as the lectotype.

We elevate the subspecies ‘P. g. maculatus’ to full species, thereby restricting the range of P. maculatus sensu stricto to the central and northern Namib Desert, north of the Kuiseb River (Fig. 5), and excluding P. adamanteussp. nov., P. kenkensessp. nov., P. circumsyrticussp. nov., and P. australissp. nov., from P. maculatus sensu stricto

Haacke (1969) described the calls of P. maculatus sensu stricto and of P. circumsyrticussp. nov., from areas near Gobabeb (Central Namib Desert, Namibia), but was unable to link individual specimens with their calls. Haacke (1969) also described calls of P. adamanteussp. nov. as ‘P. g. maculatus’. The call of P. maculatus sensu stricto is described and distinguished in a standard taxonomic framework below.

BMNH 1946.8.23.54, collected from “South Africa, Damaraland” (Erongo Region – possibly Otjimbingwe, Namibia,), by Karl (=Charles) J. Andersson prior to 1864 (probably in 1851 or between 1855 and 1858, while travelling through and working in this area), as the lectotype for Ptenopus maculatus.

BMNH 1946.8.23.53, same collection details as the lectotype.

See Table S1 for unvouchered photographed (35) specimens, DNA samples (37 available, 12 sequenced), and call recordings (28) included (total n = 69 excluding types).

A moderately large Ptenopus (SVL max. 60 mm, mean 51.7 mm, n = 34) with a moderate tail (TL 69% [range 58–78%] of SVL, n = 21) and stout appearance. It is easily distinguished from P. kochi, P. carpi, and P. sceletussp. nov. by the following characters: Toes being intermediately fringed laterally (vs. weakly fringed in P. carpi and P. sceletussp. nov., and extensively fringed in P. kochi), with fringe length being at least half the breadth of the toe between fringes; having MBSR ~135 (range 126–146, n = 35; vs. >180 in P. kochi and usually <135 in P. carpi and P. sceletussp. nov.); by dorsal colour pattern with four to five large, paired, ovoid markings on a darker background, interspaced by dark brown wavy or broken blotches (vs. more finely spotted in P. kochi and banded in P. carpi and P. sceletussp. nov.).

From congeners previously included in ‘P. garrulus’ it is distinguished by: Higher IN/INSBB ~7.2 (range 4.9–8.8, n = 8) than P. circumsyrticussp. nov. (median 4.5, range 3.6–5.8, n = 13); higher RB/RH ~1.17 (range 0.97–1.32, n = 8) than P. australissp. nov. (<0.97) and P. adamanteussp. nov. (usually <1.1); higher INSH/NB ~1.53 (range 0.60–2.49, n = 8) than P. australissp. nov. (usually <0.80) and P. adamanteussp. nov. (usually <1.00); lower MBSR (~135, range 126–146, n = 35) than P. kenkensessp. nov. (≥158) and P. garrulus (≥156, usually >170), and somewhat lower than P. adamanteussp. nov. (median 160, range 141–185, n = 30); lower IOS ~34 (range 29–40, n = 35) than other species (usually >37); lower IOS/MBSR ~0.23 (range 0.21–0.27, n = 8) than P. circumsyrticussp. nov. (usually >0.27), P. adamanteussp. nov., and P. australissp. nov. (usually >0.25); having 1–2 internasal scales in contact with the rostral, vs. three in P. australissp. nov.; having a different colour pattern than P. garrulus (usually more finely speckled, ovoid markings being much smaller).

In life (Fig. 8), the dorsal colour pattern consists of 4–6 large, cream/yellow/white paired ovoid markings from the neck to tail base with various light, coloured and dark brown juxtaposed speckles, darker colours coalescing into ill-defined paired patches between the oval markings; white ovoid markings laterally. Background dorsal colour is usually light brown or camel with occasional bright orange, pink, or purple speckling. The ventrum is immaculate white. Males have bright yellow gular patches. In some specimens, the yellow pigment extends onto the nose or patchily onto lateral margins of the head and neck.

In preservative, the lighter colours fade to off-white or beige, and all the darker colours appear various shades of dark brown or grey. The yellow fades completely after some days or weeks in ethanol.

The advertisement call of P. maculatus (Figs 3, 20F typical call, 20H Hoanib River population) consists of 6 notes (range 4–7, sometimes 8 in Hoanib River population) uttered in rapid succession, with a note rate of 4.99 s^-1 (range 2.58–7.17). Note duration is long, 50 ms (range 26–71), with the first note being notably longer than the rest, note 1 duration deviance 138% (range 39–274); the Hoanib River population has a relatively shorter first note than the typical call. The inter-note intervals are usually short (145 ms [range 88–303]) and regular with inter-note interval range 20% (range 10–42), although the first interval may be slightly longer. Call density is higher than all other species, mean 0.36 (range 0.17–0.50). Call duration is short (1.1 s [range 0.8–1.4]). The basal frequency is 380 Hz (range 301–517) but very soft and may be inestimable, with harmonic bands louder towards the (upper) dominant frequency at 3.4 kHz (range 3.0–3.8), while a clear but softer lower frequency peak is evident at around 1.7 kHz (range 1.4–2.0), just below half the (upper) dominant frequency. Frequency appears to modulate briefly up (~0.2 kHz), then drastically down by ~0.5 kHz with the first note, and less severely down in the notes following. The first note tends to be louder, have a slightly higher dominant frequency, and a slightly higher (human) perceived pitch than the following notes. Bandwidth (90%) is difficult to estimate consistently: approximately 0.9–4.8 kHz.

Call period (mean 110 s) varies greatly, but can be as low as 10 seconds during peak chorus activity. Calling commences shortly before or at sunset and may continue throughout the night after moonrise and before moonset. Calling activity is more pronounced around the austral spring and/or summer. Rainy weather usually causes calling activity to cease for days to weeks. This species rarely calls during the day.

Ptenopus maculatus occurs on hard substrates, not on soft sand. In the southern and central portions of its range it occurs on granite-weathered gravel. The central Namib gravel plains are almost entirely granite-based gravel (Atlas of Namibia Team 2022a) and the species is widespread and common in this area. In the northern reaches of its distribution, P. maculatus occurs on silts, in patchy floodplains, riverbeds, or pans. It has not been found closer than 25–50 km, nor further than 200 km, from the coast.

In the west it occurs syntopically with P. carpi and probably with P. sceletussp. nov. which prefer the same soils. However, these species occur closer to the coast within the fog zone and the range of overlap is minimal. It occurs parapatrically with P. sceletussp. nov. in the northern portions of its range, where P. maculatus prefers silts while P. sceletussp. nov. occurs on gravels. Along the edges of the Namib erg, P. maculatus occurs parapatrically alongside P. circumsyrticussp. nov. and P. kochi (which occur on sand or sandier patches of gravel plains), or occasionally in sympatry or syntopy with these species where the substrates mix.

Like the other species, P. maculatus males call from their burrow entrance and antagonistic behaviours between males are common if two burrow entrances are within ~2 m. Aggressors will sometimes utter the advertisement call as they look for the transgressing male, possibly to elicit a response. Males will increase calling frequency if a female approaches the burrow. Males may exit the burrow and approach the female, and after a brief encounter will follow the females into the burrow. Alternatively, the male will remain in place until the female scratches at the burrow entrance; the male then reverses into the burrow and the female follows. The male will utter a muffled version of the advertisement call inside the burrow, where copulation likely commences. After about an hour, the female will emerge partially from the burrow entrance, surveying the surrounding area with her head erect for a long time. This behaviour is repeated throughout the first night after a female moves into a new burrow, and its purpose may be to memorise the location of the new burrow entrance. Males presumably move out of the burrow after the female has moved in, as males and females have not been observed living in the same burrow.

Southern barking gecko

Afrikaans: Suiderlike blafgeitjie

Ptenopus garrulus maculatus – Oelofsen and Vorster (1976), Rebelo et al. (2018)

Specimens of this species have not yet been included in any taxonomic revision of this genus, only in two publications on range extensions of ‘P. g. maculatus’. These records were further notably included in the range maps of ‘P. g. maculatus’ by Branch (1998), and of ‘P. garrulus’ by Bates et al. (2014), Telford et al. (2022), and Tolley et al. (2023).

PEM R23122, adult male, collected from farm Rooidraai, Eastern Cape Province, South Africa (–32.4645, 23.6330, 860 m a.s.l.) by Werner Conradie, Alexander Rebelo, and Philip Jordaan, on 2 November 2017.

PEM R23118–23121, adult males except for PEM R23120 (allotype), collected from farm Doringkraal, Eastern Cape, South Africa (-33.0479, 24.9611, 305 m a.s.l.), by Werner Conradie, Alexander Rebelo, and Philip Jordaan, on 31 October 2017.

See Table S1 for additional unvouchered photographed specimen (1), and call recordings (4) examined (total n = 5 excluding types).

This is the most southern Ptenopus species, occurring near the southernmost tip of Africa. Therefore, we use the specific epithet “>australis”, the Latin (masculine) adjective meaning “southern”.

The smallest Ptenopus (SVL max. 44.6 mm, mean 42.7 mm, n = 5) with the shortest tail of any Ptenopus species (TL 61% of SVL, only one specimen had full original tail, but other paratypes/holotype lost a very small portion of the tail tip and they still appear to be similarly short) and a moderately stout appearance. It is distinguished from P. kochi, P. carpi, and P. sceletussp. nov. by: Being substantially smaller; toes being intermediately fringed laterally (vs. weakly fringed in P. carpi and P. sceletussp. nov. and extensively fringed in P. kochi), with fringe length being at least half the breadth of the toe between fringes (vs. generally less than half in P. carpi and P. sceletussp. nov., and generally equal to in P. kochi); ventral surface being generally white or cream with some unpigmented and/or dark brown-speckled scales on the (hand/foot) soles (vs. substantial pink, unpigmented patches on the tail and limbs in P. kochi, and immaculate white in P. carpi and P. sceletussp. nov.); having MBSR 140–156, mean 148 (n = 5) (vs. ≥187–222 in P. kochi and usually <135 in P. carpi and P. sceletussp. nov.); a generally brown-and-cream spotted appearance with some paired light and dark markings dorsally (vs. pinkish or orange, more evenly spotted pattern in P. kochi and banded pattern in P. carpi and P. sceletussp. nov.). It is further distinct from P. carpi and P. sceletussp. nov. by the nasals being more swollen and the nostrils partially covered by internal projections of the upper labials; from P. kochi by having fingers laterally fringed with pointed triangular scales (vs. elongated pointed scales in P. kochi).

From congeners previously included in ‘P. garrulus’ it is distinguished by: A smaller internasal scale, with higher IN/INSBB (~10.6, range 7.3–11.6, n = 5) and lower INSH/NB (~0.65, range 0.57–0.97, n = 5) than P. circumsyrticussp. nov. (IN/INSBB <6, INSH/NB usually >0.97); having lower RB/RH (<1) than these congeners (usually >1, except for some P. adamanteussp. nov. individuals); having lower MBSR (~148, range 147–149, n = 5) than P. garrulus and P. kenkensessp. nov. (≥158); IOS/MBSR (~0.26, range 0.24–0.29) usually higher than P. garrulus (~0.23, range 0.20–0.25, n = 10); having 3 internasal scales in contact with the rostral (vs. usually <3 for P. garrulus and P. adamanteussp. nov., ≤2 for P. kenkensessp. nov. and P. maculatus, and only 1 in P. circumsyrticussp. nov.).

(Fig. 11). Adult male, SVL 44.63 mm with original tail short with blunt tip, 21.99 mm (49.3% of SVL). Body and head covered with minute hexagonal to round scales of a similar size, dorsally and ventrally, with scales on limbs slightly larger. Mid-ventral incision, T-shaped, for removal of liver sample. Body stout, MBSR 149, IOS 43, HL 13.48 mm, HW 8.33 mm, HH 5.68 mm, EED 3.87 mm. Upper labials 7, lower enlarged labials 7; nine granules bordering the mental. Clear but not prominent superciliary ridges tapering above mid-eye level; right eyelid partially depressed by the fixing process with skin folds on the forehead. Prenasal scales slightly swollen, separated by three small, asymmetrical internasal scales in contact with the rostral, with IN/INSBB 17.14, INSH/NB 0.65; nostrils almost entirely covered by projection from the upper nasal. Rostral is large, with RB/RH 0.96; mental deep and slightly pointed ventrally. EYE 2.58 mm, pupil vertical; ear opening is oblique (~40°), and narrow with small, slightly projecting scales at the anterior margin. Toes elongate, slightly flattened, with elongate fringed scales; small, pointed, barely projecting triangular fringes on the fingers; strong, recurved and pointed nails on fingers and toes.

In life, the paratype PEM R23121 (Fig. 10A–B) has a dorsal background colour light brown with cream and dark brown asymmetrical blotches and a few orange spots or blotches (holotype appears to have been similar, but darker with finer patterning); light and dark dorsal markings merge into poorly defined ‘bands’ on the tail hat become more defined distally; dark brown blotches extend onto lateral surfaces with slight, dark grey markings below the lateral line on the white (similar to holotype); ventrum is immaculate white except for the slight dark grey mottling laterally, which is more visible on the head and tail; slight dark speckling and some unpigmented or punkish scales on hand palm, while foot sole is densely pigmented with dark speckles. Clear, bright yellow gular patch; the bright yellow does not extend onto the lateral surfaces of the head, body, or onto the limbs. Iris is light silver.

In preservative (Fig. 9), the lighter colours have faded to off-white or beige, the darker colours remain various shades of dark brown or grey, while the orange and yellow have faded completely (on paratype PEM R23121 and presumably on the holotype and others). The eyes turned a milky blue-grey.

See Table S1 for paratype and additional material measurements and scale counts, which does not vary substantially among the types. Colour pattern varies relatively little among specimens, with darker or lighter, finer or courser colour patterns being visible. Males have bright yellow gular patches, and may have slight yellow infusions on the legs and face (Fig. 10C).

The advertisement call (Figs 3, 20B) consists of 5.8 notes (range 5–6) uttered in rapid succession, with a note rate of 6.68 s^-1 (range 5.10–7.54). Note duration is short, 21 ms (range 21–22), and highly regular, with note 1 duration deviance 6% (range 1–19%). Inter-note intervals are short (117 ms, range 107–169 ms) and regular, although the last interval can be slightly longer; inter-note interval range 29% (range 13–55%). Call density moderately low (0.18, range 0.17–0.20). Call duration shortest of all species, only 0.82 s. The basal frequency was inestimable within our sample set, with the dominant frequency at 4.4 kHz (range 4.2–4.5 kHz); a very slight lower peak frequency band may present around 1.3 kHz. Frequency appears to remain constant throughout the short note. The (human) perceived pitch tends to be higher than that of most other species. Bandwidth (90%) is difficult to estimate consistently with the recordings available.

During peak chorus activity, this species appears to call much more frequently than most other species (although sample size is small), the call period being a very short 9 seconds. The typical period of calling activity in a day is not known, but the analysed calls (during full chorus) were recorded 20–40 minutes after sunset.

This species occurs on open Karoo scrub plains with sandy or gravelly substrate (Rebelo et al. 2018), usually near small dune hummocks, south of the Great Escarpment in the Western Cape of South Africa (essentially the southern Nama Karoo). It has not been recorded in sympatry with any other species.

Very little is known about the natural history of this species, although it is assumed to be similar in many aspects to congeners, particularly to the sister species, P. garrulus, which also occurs in a similar climatic setting. So far, calls have only been recorded during late October, during the austral spring (recordings in this manuscript; Rebelo et al. 2018). The breeding period is expected to last for at least 2–3 months, as in the closely related P. garrulus and P. adamanteussp. nov. As in several other species, they call from their burrow entrances in exposed or partially concealed locations (Rebelo et al. 2018).

Diamond coast barking gecko

Afrikaans: Diamantkus blafgeitjie

Ptenopus garrulus maculatus (in part) – FitzSimons (1935: 525; Haalenberg, Luderitzbuch, Soebartsfontein); FitzSimons (1937b:160; Port Nolloth, Kamaggas), Haacke (1969: 92; Annisfontein, Daberas, Sendelingsdrif, Swartpoort), Mertens (1971: 44; southwestern localities), Haacke (1975; southwestern localities)

Ptenopus garrulus (in part) – Brain (1962: fig. 3; southwestern localities)

All coastal and near-coastal specimens in the southern Namib Desert (Namibia) and Richtersveld (South Africa) previously referred to as ‘P. g. maculatus’, are presently assignable to this species. It appears to not occur sympatrically with any other species, except potentially with P. kenkensessp. nov. on the eastern margins of its range. Northern and southern populations of this species display several notable differences, including body-size. However, phylogenetic evidence presently suggests that these populations may form a gradual cline, rather than being two distinct species.

NMNW R11390, adult male, collected from Grosse Bucht, south of Lüderitz, ||Karas Region, Namibia (–26.73379, 15.10412, 29 m a.s.l.), by Francois S. Becker and Bertha Buiswalelo on 27 September 2022.

NMNW R11391–3 (two adult males and one adult female NMNW R11391 = allotype), same collection details as the holotype; NMNW R11350, subadult female, collected from the pump house along main road, Sperrgebiet, ||Karas Region, Namibia (-26.9932, 15.3584), by Dayne Braine on 10 September 2021.

See Table S1 for unvouchered photographed specimens (21), DNA samples (17 available, 7 sequenced), and call recordings (8) included (total n = 28 excluding types).

This species is named in reference to the diamond-scattered coastline that forms its habitat, including the Sperrgebiet in Namibia and the Richtersveld in South Africa, where extensive diamond mining occurred historically and continues to this day. We use the specific epithet “adamanteus”, the Latin adjective meaning “of diamond”, framed in the male genitive to match the gender of Ptenopus.

Southern populations of this member of the genus have a small body size (SVL max. 45.1 mm, mean 39.4 mm, n = 19) with long tails (TL 77%, range 60–99%, n = 19) while that of the northern populations is moderate (up to 50.8 mm, mean 49.7 mm, n = 6), with a moderately long tail (TL 71%, range 67–76%, of SVL) and an overall lean appearance. It is distinguished from P. kochi, P. carpi, and P. sceletussp. nov. by: Toes being intermediately fringed laterally (vs. weakly fringed in P. carpi and P. sceletussp. nov. and extensively fringed in P. kochi), with fringe length being at least half the breadth of the toe between fringes (vs. generally less than half in P. carpi and P. sceletussp. nov., generally equal to in P. kochi); having MBSR 141–185 (vs. ≤131 in P. carpi and P. sceletussp. nov. and >185 in P. kochi); by having on the dorsum 4–5 paired, more-or-less symmetrical, light-coloured, ovoid markings on the body, about half the diameter of the eye or larger, interspaced by dark brown blotches (vs. more finely patterned dots in P. kochi and banded pattern in P. carpi and P. sceletussp. nov.). It is further distinct from P. carpi and P. sceletussp. nov. by the nasals being more swollen and the nostrils partially covered by internal projections of the upper labials; from P. kochi by having fingers laterally fringed with pointed triangular scales (vs. elongated pointed scales in P. kochi), and having white pigmented ventral scales on entire ventral surface (vs. pink, unpigmented patches on the tail and limbs in P. kochi).

From congeners previously included in ‘P. garrulus’ it is distinguished by: Having immaculate white pigmented scales on soles, vs. some pink, unpigmented scales on soles of P. garrulus and P. australissp. nov., and dark speckling on the soles of P. kenkensessp. nov. and P. australissp. nov.; having a usually smaller internasal scale than P. circumsyrticussp. nov., with IN/INSBB ~8 (usually >6.0, vs. usually <6.0 in P. circumsyrticussp. nov.); higher RB/RH (~1.1, usually ≥1 and <1.2) than P. australissp. nov. (usually ≤1.0), but lower than P. circumsyrticussp. nov. (usually >1.1); usually lower INSH/NB (usually <1.1) than P. circumsyrticussp. nov. (usually >1.1); having a generally higher IOS (mean 45, range 35–53, n = 30) and IOS/MBSR (mean 0.29, usually >0.24) than P. maculatus (IOS usually <37, IOS/MBSR usually <0.24); having one or two inernasal scales in contact with the rostral vs. three in P. australissp. nov.

(Fig. 11). Adult male, SVL 49.94 mm with original tail 35.20 mm (70.5% of SVL). Body and head covered with minute hexagonal to round scales of a similar size, dorsally and ventrally, with scales on limbs notably larger. Small mid-ventral incision for removal of liver sample. Body slender, MBSR 170, IOS 53, HL 15.87 mm, HW 11.19 mm, HH 6.68 mm, EED 5.89 mm. Upper labials 8, lower enlarged labials 8 with an additional small, elongate scale on the inflection of the jaw; eight granules bordering the mental. Clear but not prominent superciliary ridges tapering above mid-eye level; a single row of elongated scales around the anterior and upper margins of the eye with curved tips posteriorly, the scales being smaller and rounder along the posterior and dorsal edges. Prenasal scales swollen, separated by one small, asymmetrical hexagonal internasal scale in contact with the rostral, with IN/INSBB 6.90, INSH/NB 0.50; nostrils partially covered by projection from the upper nasal. Rostral is narrow and tall, with RB/RH 0.76, MB/RB 0.89. Eyes intermediate, EYE 3.46 mm, pupil vertical; ear opening is oblique (~45°), and narrow with small, slightly projecting scales at the anterior margin. Toes elongate, flattened, with elongate fringed scales, small, pointed, triangular fringes on the fingers; strong nails on fingers and toes, being larger and thicker on the fingers.

In life, the holotype (Fig. 12A) has a dorsal background colour of dull orange or light brown with light grey-purple, cream, and grey spots; four pairs of large, cream/white, ovoid markings on the back interspaced by diffuse, mottled, dark brown-purple patches; these markings coalesce into alternating cream and light purple/brownish bars on the tail (9 dark bars, the last two very faint). Laterally the colours and markings fade, with immaculate white ventrum. Laterally, the body has five white circular markings, dorsally outlined with light brown/purple. Yellow gular patch extensive, with additional yellow spots on the snout, laterally on the head, body and neck, and on anterior surfaces of the back legs. The iris is a light green-grey or greenish camel.

In preservative (Fig. 12), the lighter colours faded to off-white or beige and the darker colours to shades of dark brown or grey, while the immaculate white areas appear more cream. The yellow has faded completely and the iris is milky blue-grey.

Refer to Table S1 and Figure 4 for range of morphometric characters, including the paratypes. Internasal scale is typically small and asymmetrical; rostral typically narrow and tall; toe fringes relatively extensive. Dorsal colouration highly variable, matching the wide variation of substrate colours and surface textures across its range. Northern populations (Fig. 12A–D) have a similar colour pattern to the holotype, whereas the southern populations (Fig. 12E–H) have a more finely speckled or dotted appearance somewhat similar to P. garrulus. The northern populations occur more often on deflation plains with pebbles or course sand on the surface (despite finer sand beneath the surface), whereas the southern populations tend to occur in fine sandy substrates–which may account for the colour pattern difference (this appears to be the pattern across the genus). Observed colour variations include overall cream, greyish, ochre, or even brick red background colouration with speckling colour variations including brown, white, olive, purple, orange, and black. Ventrally, animals are immaculate white, although some southern populations possess black or dark grey speckling on the throat and lateral edges of the body, occasionally intruding onto the belly. Males have bright yellow gular patches which often extend onto the lateral head and body, and yellow on anterior surfaces of the hind leg.

The advertisement call (Figs 3, 20C) consists of 5 notes (range 4–6), uttered slowly at a rate of 2.81 s^-1 (range 2.64–3.15). Note duration is short (29 ms [range 24–34]) and highly consistent; note 1 duration deviance 9.2% (range -4–16). Inter-note intervals 313 ms (range 274–341), with a notably longer final interval; inter-note interval range high, 52.2% (range 36–66). Median call density is low, 0.09 (range 0.08–0.11); call duration 1.4 s (range 1.1–1.8). The basal frequency is ~415 Hz (range 370–516, but very soft and may be inestimable, with harmonic bands louder towards the (upper) dominant frequency at 4.4 kHz (range 4.1–5); usually no notable lower peak frequency peak is evident. Frequency appears to remain more or less constant throughout the note, and (human) perceived pitch is notably lower in northern than southern populations, and also coastal compared to inland populations. Bandwidth (90%) is difficult to estimate consistently: approximately 0.5–7.3 kHz.

In coastal populations this species does not call often (longer call periods and fewer evenings of calling), but inland populations are more vocal. This is probably due to higher incidence of cold, foggy, windy weather closer to the coast, in which calling activity is reduced. Call period (5–20 s, mean 14 s) is short compared to most other species during peak chorus activity. It mainly calls from about 30 minutes before sunset until nightfall, with occasional calling later at night. This species may call sporadically throughout the day, particularly under foggy or cloudy conditions.

(Fig. 5). This species occurs on small coastal dunes, sandy plains, and on deflation plains in the southern Namib Desert, or the Namaqualand-Richtersveld steppe ecoregion (see Dinerstein et al. 2017). It occurs from the edge of the intertidal zone, inland to about 50–80 km from the coast. The habitat falls within the coastal fog zone and receives rain predominantly in the austral winter.

While its breeding phenology is not well known, our sampling suggests that calling/breeding peak activity occurs from September to November, like several other species. This species often wanders on the surface throughout the night, far from its burrow, and may be active despite cold, foggy weather, similar to P. carpi and P. sceletussp. nov. The burrows are relatively complex with multiple side-tunnels, and a few egg clutches, usually two eggs but occasionally one, have been found inside burrows occupied by a female. This species occasionally closes the burrow entrance from the inside before daybreak, but may also be active diurnally, especially after or during foggy weather. Hatchlings have been observed digging their own burrows a day after hatching, and it is therefore expected that they do not remain in the parent burrows for any length of time. Males and females have not been observed sharing a burrow. This species has been observed feeding on spiders and termites, and it likely eats various arthropods.

Interdune barking gecko

Afrikaans: Duinstraat blafgeitjie

Ptenopus garrulus (in part) – Brain (1962: fig. 3; central Namib)

Ptenopus garrulus maculatus (in part) – Haacke (1969: fig. 3b–c), Haacke (1975: 225; central Namib Desert, notably Gobabeb)

Specimens of P. circumsyrticussp. nov. were not included in the earliest mentions of the name P. g. maculatus (e.g., FitzSimons 1935), but were included in several later works, including notably those of Haacke (1964, 1969, 1975). Haacke (1969) described the calls of both P. maculatus sensu stricto and P. circumsyrticussp. nov. from the area surrounding Gobabeb, under the name ‘P. g. maculatus’, noting the clear differences. However, Haacke did not manage to match the two distinct calls with particular specimens in the area where they are sympatric (on the gravel plains of the central Namib Desert), and hence was uncertain about whether the different calls reflected intraspecific or interspecific variation. It is also worth noting that, in this sympatric zone, the two species can be readily distinguished by the internasal scale in P. circumsyrticussp. nov. being large and in broad contact with the rostral, separating the nasal scales clearly, whereas P. maculatus has one or more small and rounded granules, barely separating the nasal scales. This morphological distinction is less clear in allopatric populations.

NMNW R11394, adult male, collected from the interdunes plains south of the Kuiseb River from Gobabeb Research Station, ||Karas Region, Namibia (–23.57053, 15.03618, 415 m a.s.l.), by Francois S. Becker on 6 September 2022.

NMNW R11395 (allotype, adult female), R11396 and R11622 (adult males), with the same collection details as the holotype; R11346 (adult male) and R11371 (adult female), collected near the type locality (–23.5697, 15.0388), by Francois S. Becker on 17 September 2021.

See Table S1 for vouchered (3) and unvouchered photographed specimens (13), DNA samples (16 available, 11 sequenced), and call recordings (11) included (total n = 25 excluding types).

This species is named in reference to its habitat: It occurs in and around the sand sea, on interdune plains or dune streets, and on sandy plains at the edge of the sand sea. They do not occur on the dunes themselves. Thus, we use the specific epithet “circumsyrticus”, the Latin adjective meaning “around the dune”, framed in the male genitive to match the gender of Ptenopus.

A moderately small Ptenopus (SVL max. 54.9 mm, mean 48.6 mm, n = 16) with a short tail (TL 62% [range 48–73] of SVL, n = 8) and a lean appearance. It is distinguished from P. kochi, P. carpi, and P. sceletussp. nov. by being substantially smaller; toes being intermediately fringed laterally (vs. weakly fringed in P. carpi and P. sceletussp. nov. and extensively fringed in P. kochi), with fringe length being at least half the breadth of the toe between fringes (vs. generally less than half in P. carpi and P. sceletussp. nov., generally equal to in P. kochi), although fringing can be more extensive in specimens found in looser sand, such as close to Walvis Bay or Far East Dunes; by dorsal colour pattern, characterised by paired, large, subsymmetrical, light ovoid markings interspaced by dark blotches (vs. spotted pattern in P. kochi and banded pattern in P. carpi and P. sceletussp. nov.). It is further distinct from P. carpi and P. sceletussp. nov. by the nasals being more swollen and the nostrils partially covered by internal projections of the upper labials; from P. kochi by having fingers laterally fringed with pointed triangular scales (vs. elongated pointed scales in P. kochi), having white pigmented scales on entire ventral surface (vs. pink, unpigmented patches on the tail and limbs in P. kochi), and having MBSR 108–179 (vs. >180 in P. kochi), with finer lepidosis presenting in specimens found in looser sand.

From congeners previously included in ‘P. garrulus’ it is distinguished by: The internasal scale usually being larger and broader than other species, with IN/INSBB being generally lower (median 4.5, range 3.6–5.8, but usually <4.7) than other species (usually >5.0); having higher RB/RH (median 1.27, usually >1.17) than P. adamanteussp. nov., P. australissp. nov., and P. garrulus (usually <1.17); having higher INS/NB (~1.61, usually >1.10) than P. adamanteussp. nov., P. australissp. nov., and P. garrulus (usually <1.20); having lower MBSR (median 144, usually <160) than P. kenkensessp. nov. and P. garrulus (usually >160); higher IOS/MBSR (median 0.31, usually >0.25) than P. maculatus and P. garrulus (usually ≤0.25); having only one inernasal scale in contact with the rostral, while P. australissp. nov. has three; having no pink, inpigmented scales on the soles (vs. some unpigmented scales in P. garrulus and P. australissp. nov.).

(Fig. 13). Adult male, SVL 48.75 mm with original tail 35.79 mm (73.42% SVL). Body and head covered with minute hexagonal to round scales of a similar size, dorsally and ventrally, with scales on limbs notably larger. Small mid-ventral incision for removal of liver sample. Body and head slender, MBSR 129, IOS 53, HL 14.62 mm, HW 10.99 mm, HH 5.57 mm, EED 5.81 mm. Upper and lower labials 8, six granules bordering the mental. Clear but not prominent superciliary ridges tapering above mid-eye level; a single row of elongated scales around the anterior and upper margins of the eye with curved tips posteriorly, the scales being smaller and rounder along the posterior and dorsal edges; nasal scales swollen, the prenasals separated by one large, hexagonal internasal scale in broad contact with the rostral, with IN/INSBB 4.69, INSH/NB 1.09; nostrils partially covered by projection from the upper nasal. Rostral broad, with RB/RH 1.44, MB/RB 0.81. Eyes intermediate, EYE 3.15 mm, pupil vertical. Ear opening is oblique (~45°) and narrow, with small, slightly projecting scales at the anterior margin. Toes elongate, flattened, with elongate fringed scales, small, pointed, triangular fringes on the fingers; strong nails on fingers and toes, being larger and thicker on the fingers.

In life, the holotype (Fig. 14A) has a dorsal background colour of ochre with deep orange, cream, and dark brown spots, the darker spots aggregated in clusters; three pairs of larger, cream/light yellow, ovoid markings from the neck to the back, with one unpaired marking on the left lower back and another asymmetrical pair onto the tail base; light markings interspaced by diffuse, asymmetrical, mottled or spotted, dark brown patches; the tail has 11 dark bands, unevenly spaced and somewhat diffuse, with the last being very faint; the dorsal colours and markings fade to grey on the lateral edges with clear alternating light-and-dark markings; immaculate white ventrum. Yellow gular patch with faded edges.

In preservative (Fig. 13), the lighter colours have faded to off-white or beige, and all the darker colours appear various shades of dark brown or grey with some orange being evident. The yellow has faded completely.

Refer to Table S1 and Figure 4 for range of morphometric characters, including the paratypes. Internasal scale typically large and symmetrical; rostral typically broad; toe fringes relatively extensive, notably more so in areas with looser sand. As with other species, the colour of the animals usually matches that of the local substrate, although the patterning is similar to the holotype (Fig. 14). Background colour variations include ochre to brick red. Ventrally animals are immaculate white. Males have faded to bright yellow gular patches (typically less extensive than most other species), occasionally with slight yellow on hidden dorsal surfaces of the hind leg.

The advertisement call (Figs 3, 20D) consists of 4.6 notes (range 3–7), uttered slowly at a rate of 2.44 s^-1 (range 1.57–3.68). Note duration is long (48 ms [range 31–57]) and regular (note 1 duration deviance 16% [range 3–28]), with long, highly regular intervals (inter-note interval range 13% [range 10–20]) inter-note intervals of 385 ms (range 237–587). Median call density is low (0.129 [range 0.09–0.17]) and call duration is 1.49 s (range 1.1–2). The basal frequency is ~383 Hz (range 308–474), but very soft and may be inestimable, with harmonic bands louder towards the (upper) dominant frequency at 3.8 kHz (range 3.5–4.1); no clear lower peak frequency is evident. Frequency appears to modulate up by ~0.5 kHz in the course of the note, but (human) perceived pitch is usually monotonous throughout the notes and call. Bandwidth (90%) is difficult to estimate consistently: approximately 1–4.7 kHz.

The call strongly resembles the sound of bouncing marbles and on a warm evening on the interdunes thousands can be heard calling in chorus. Call period (mean 133 s) varies greatly, but can be as brief as 16 seconds during peak chorus activity. This species tends to occur (and call) at higher densities than the sympatric congeners P. maculatus and P. kochi. It calls from shortly before sunset and calling activity may continue throughout the night as long as the moon is above the horizon, particularly on warm evenings. This species occasionally calls during the day, especially during foggy or cloudy conditions, but generally not during rainy conditions.

This species occurs on sandy soils including interdune plains within, and sandy plains around the northern and eastern edges of the Namib erg, Namibia (Fig. 5). It generally does not occur on the gravel plains north of the Kuiseb River except in a few sandier patches east of 15° longitude. This population is genetically divergent from other nearby populations in the Namib erg. It does not appear to occur along the south-eastern corner of the Namib erg, where slightly harder Arenosols are occupied by P. kenkensessp. nov., or along the southern or south-western edges of this erg, where the same soils are occupied by P. adamanteussp. nov.

It has not been found in sympatry with P. kenkensessp. nov., but overlap is likely in the southern portion of its range. It occurs parapatrically alongside P. kochi, P. maculatus, and possibly P. carpi sensu stricto (near the northern border of the Kuiseb River), as these species occupy different substrates in close geographic proximity. Occasionally, it has been found in sympatry and even syntopy with P. maculatus on the sandier patches of the gravel plains north of the Kuiseb River, and with P. kochi in the Namib erg, near the edges of the interdunes where the two species’ preferred substrates overlap. On broad interdunes, P. circumsyrticussp. nov. occurs in allopatry.

The peak calling and breeding season is in the austral spring to summer (August to November), although some calling may take place outside this period. Two eggs (dried out) have been recovered from an abandoned burrow, so clutch size can be up to two. The burrows are generally less complex than the parapatric P. kochi and P. maculatus, and appear to be inhabited solitarily. Very small juveniles have been observed in burrows of their own. While multiple natural predators for Ptenopus spp. have been observed and hypothesised (Haacke 1975), one observation of a P. circumsyrticussp. nov. being carried off by a large solifuge (le Roux pers. com. 2022) certainly constitutes a novel observation.

Common names.

Nama barking gecko

Nama: ||en||ense / ||en||enses (“||” = lateral click)

Afrikaans: Nama blafgeitjie

Ptenopus garrulus maculatus (in part) – FitzSimons (1935: 525; Kanus), FitzSimons (1937b:160; Great Fish River), Mertens (1971; southern-central Namibia localities), Haacke (1975: 225; Great Karas Mountains, Keetmanshoop)

Ptenopus garrulus (in part) – Brain (1962a: fig. 3; southeastern localities)

All Ptenopus specimens that occur on compact substrates in the northern Nama Karoo, i.e., inland in the south of Namibia and from the Northern Cape (Great Karoo), are likely to be assignable to this species.

NMNW R11389, adult male, collected from farm Goris near Giants’ Playground, ||Karas Region, Namibia (–26.4538, 18.3097, 1106 m a.s.l.), by Francois S. Becker and Bertha Buiswalelo on 27 September 2022.

NMNW R10839 (adult female, allotype) and NMNW R10841 (adult male) collected from Koës, ||Karas Region, Namibia (-25.9364, 19.0869, 1000 m a.s.l.), by Francois S. Becker, Hileni Shivolo and Sisamu Baepi on 6 October 2020; NMNW R11388, adult male, with the same collection details as the holotype.

See Table S1 for unvouchered photographed specimens (13), DNA samples (11 available, 7 sequenced), and call recordings (9) included (total n = 20 excluding types).

This species is named after the Nama name for the gecko, “||en||enses”, pronounced with two lateral clicks (produced by clicking with one’s tongue on the posterior-lateral roof of one’s mouth, similar to the typical command given for a horse to speed up), in reference to the clicking sound that it makes. The name is usually formed in the Nama female genitive, indicated by the suffix “s”, because the animal is small. This gecko occurs throughout the Nama Karoo and various places that were historically, and are currently, occupied by the Nama people. The Nama are highly familiar with its call, and believe the bite of this gecko to be extremely venomous. Nama children are taught to treat it with caution. However, the authors have been bitten by this species, and have survived.

We use the specific epithet “kenkenses”, a noun in apposition. Since the Nama symbols cannot be included in a scientific name, the phonetically similar letter “k” is used instead. However, the use of Nama lateral clicks is recommended in the pronunciation of this name.

A medium-sized Ptenopus (SVL max. 58 mm, mean 51.3 mm, n = 10) with a short tail (TL 65% [range 60–72] of SVL, n = 7) and a stout appearance. It is distinguished from P. kochi, P. carpi, and P. sceletussp. nov. by the following characters: Toes being intermediately fringed laterally (vs. weakly fringed in P. carpi and P. sceletussp. nov. and extensively fringed in P. kochi); having MBSR ~162, range 158–169 (vs. >180 in P. kochi and <135 in P. carpi and P. sceletussp. nov.); by dorsal colour pattern being characterised by large, paired, light, subsymmetrical ovoid markings interspaced with or dominated by dense, dark brown mottled patches (vs. more finely spotted pattern in P. kochi and banded pattern in P. carpi and P. sceletussp. nov.). It is further distinct from P. carpi and P. sceletussp. nov. by the nasals being more swollen and the nostrils partially covered by internal projections of the upper labials; from P. kochi by having fingers laterally fringed with pointed triangular scales (vs. elongated pointed scales in P. kochi) and having entire ventrum covered in white/cream ventral scales with some dark brown/grey specking (vs. pink, unpigmented patches on the limbs and tail of P. kochi).

From congeners previously included in ‘P. garrulus’ it is distinguished by: Having a generally smaller internasal scale with higher IN/INSBB (median 7.4, range 5.6–11.0, n = 12), than P. circumsyrticussp. nov. (usually <5.6); having a broader rostral with higher RB/RH (median 1.21, range 0.98–1.42, n = 12) than P. australissp. nov. (<0.98); having generally higher MBSR (≥147) than P. australissp. nov., P. circumsyrticussp. nov., and P. maculatus (usually <147 except some P. circumsyrticussp. nov.); having 2 or fewer internasal scales in contact with the rostral, while P. australissp. nov. has three; colour pattern distinct from P. garrulus in having 4–5 large and distinctive paired, light, ovoid markings interspaced by distinct dark mottled patches usually touching the light markings (vs. more rows of and smaller white spots and overall more speckled pattern in P. garrulus), and ovoid patterning on ventro-lateral portions of the face being more distinctive (vs. indistinct or absent in P. garrulus).

(Fig. 15). Adult male, SVL 51.41 mm with original tail 30.61 mm (59.54% SVL). Body and head covered with minute hexagonal to round scales of a similar size, dorsally and ventrally, with scales on limbs notably larger. Small mid-ventral incision for removal of liver sample. Body slender, MBSR 166, IOS 48, HL 15.97 mm, HW 11.15 mm, HH 6.85 mm, EED 5.93 mm. Upper labials 6 enlarged, with 4–5 small granules up to the angle of the jaw, lower labials 8, the posterior scale elongate and thin, and six granules bordering the mental. Clear but not prominent superciliary ridges tapering from above mid-eye level. Around the eye is a single row of elongated scales around the anterior, dorsal, and posterior margins, but more rounded and smaller along the ventral margin. Nasal scales slightly swollen, the prenasals barely separated by one small, rounded internasal scale in meagre contact with the rostral, with IN/INSBB 8.71, INSH/NB 1.03. Nostrils partially covered by projection from the upper nasal. Rostral broad, with RB/RH 1.42, MB/RB 0.75. Pupil vertical, EYE 3.46 mm. Ear opening is oblique (~45°) and narrow with small projecting scales at the anterior margin. The neck region behind the cheeks is slightly swollen with internal calcium deposits. Arms, legs, and tail stout in appearance. Toes elongate, flattened, with moderately elongate fringed scales, small, pointed, triangular fringes on the fingers; strong nails on fingers and toes, being larger and thicker on the fingers.

In life, the holotype and similar paratype NMNW R11388 (Fig. 16A and C) has a dorsal background colour of light brown with orange, cream, and beige spots, with five intermediate paired, circular, markings on the neck and back interspaced by larger dark brown, irregular markings and some smaller cream dots and dark blotches that are more-or-less symmetrically scattered across the dorsal surface; the paired cream markings merge into bands on the tail (somewhat asymmetrical in the mid-tail) interspaced by 10–11 dark bars, which are barely noticeable on the nearly white tail tip; prominent dark brown to black dorsolateral markings; ventrum is immaculate white except for some dark grey mottling laterally, on ventral leg surfaces, and substantial dark mottling on (hand/foot) palms/soles. Extensive, bright yellow gular patch with faded anterior edges and a clear but speckled posterior edge, with a large white area on the throat; the bright yellow does not extend onto the lateral surfaces of the head, body, or onto the limbs. The iris is dark beige/light yellow-brown.

In preservative (Fig. 15), the lighter colours have faded to off-white or beige, while the darker colours remain shades of dark brown or grey, and the orange and yellow have faded completely. The iris is a milky blue-grey.

Refer to Table S1 and Figure 4 for range of morphometric characters, including the paratypes. While morphometric characters are relatively similar across all specimens, the colour patterns vary substantially, usually matching the local substrate (Fig. 16). Background dorsal colouration variations include brown, light orange, beige, or red-brown; some amount of orange speckling is usually present, particularly in the western populations. Ventrally, animals are immaculate white with grey or brownish speckling clearly visible on (hand/foot) palms/soles. Males have bright yellow gular patches, while some male specimens also have yellow on the anterior or dorsal surfaces of the hind leg.

The advertisement call (Figs 3, 20E) consists of 4.6 notes (range 4–5), uttered slowly at a rate of 2.45 s^-1 (range 1.99–2.96). Note duration is the longest of all species (92 ms [range 61–154]), the first usually notably longer than those following, with note 1 duration deviance 30% [range 10–47]. Inter-note intervals are 316 ms (range 269–368) and quite regular, inter-note interval range 23.3% (range 12–42). Median call density is high, 0.29 (range 0.18–0.48). Call duration is 1.57 s (range 1.3–1.9). The basal frequency is ~399 Hz (range 366–474), but very soft may be inestimable, with harmonic bands louder towards the dominant frequency at 3.506 kHz (range 3.3–3.8); sometimes a very slight lower frequency peak is evident at around or just below half the dominant frequency (roughly 1.2–1.9 kHz, cannot always be reliably estimated). Frequency appears to modulate down by ~0.4 kHz during the course of the note, and (human) perceived pitch also seems to modulate within each note, slightly reminiscent of dripping water. Bandwidth (90%) is difficult to estimate consistently: approximately 0.5–6.9 kHz.

Calling activity for this species commences ~ 30 min before sunset and may continue late into the night if the moon is up. Call period (mean 84 s) varies considerably, but can be as low as 18 seconds during peak chorus activity. Peak calling activity occurs from September to November.

This species occurs on consolidated soils such as gravel, alluvial silt, or compacted sandy soils throughout the northern Nama Karoo: southern Namibia and probably the Northern Cape, South Africa (Fig. 5). The exposed rocky surfaces of the Nama group sedimentary formation in central-southern Namibia appears to bisect its distribution. The northeastern edge of its range is the western edge of the Kalahari Sand Sea, as it does not occur on loose sand. There is only one confirmed record from South Africa near Onseepkans, Northern Cape, but similar soils occur more broadly throughout large parts of the Northern Cape Province.

It has not been recorded in sympatry with any other species but occurs parapatrically to P. garrulus near Koës, Namibia (with P. garrulus occurring in the Kalahari Sand Sea), and in the Northern Cape, South Africa (where patches of Kalahari sand occur on more consolidated soils). It has been recorded ~40 km from the nearest confirmed P. adamanteussp. nov. record in the west, and no closer than ~60 km from the nearest confirmed P. circumsyrticussp. nov. record in the northwest of its range.

Calling (and therefore breeding) activity appears to be seasonal, probably peaking in the austral spring and early summer. This species is seldom found on the surface, except early in the evening during peak chorus activity. Burrow entrances are nearly always sealed after calling ceases, or during the day; the entrance is sealed from the inside, and it is so neatly disguised as to be almost entirely indistinguishable from the surrounding soil surface. The burrow is then re-opened before calling commences. It has been observed consuming large numbers of harvester termites (Hodotermes mossambicus), even up to 61% of its own body-weight, during sporadic events when these termites emerge en masse to forage (Bauer et al. 1989). All adult females collected by Bauer et al. (1989) in early October contained ovarian or unshelled oviductal eggs, suggesting an egg-laying period possibly later than November. One individual of P. kenkensessp. nov. has been found to be infected by a tapeworm of the genus Mesocestoides (McAllister et al. 1995).

Aggregations of this species are sometimes observed on tarred roads late at night, numbering in the hundreds on a few kilometres of road. The reason for this behaviour is unknown.

Carp’s barking gecko

Afrikaans: Carp se blafgeitjie

Ptenopus carpi Brain, 1962: plates 1b, c, e, fig. Ib, Haacke (1964: 3), Mertens (1971: 44), Haacke (1975: 230; in part)

The type locality of P. carpi is “gravel plain approximately 1 mile north of the Kuisib R. [Kuiseb River] at Gobabeb, Central Namib Desert [Erongo Region], S.W.A. [Namibia]” (-23.5456, 15.0400). Currently, no P. carpi can be found at this location. The closest location of current occurrence from Gobabeb, is 9 km to the northwest. Charles Brain’s son, Conrad Brain, who had attended the field trip, claims that they were a little lost that evening and likely the real type locality was more to the northwest, where the species currently occurs (Brain pers. comm. 2023).

Here, we split “P. carpi” into two species. The topotypical population south of the Swakop River is assigned the name P. carpi sensu stricto, while P. sceletussp. nov. is described from north of the Swakop River, below. We hereby restrict the range of P. carpi sensu stricto to between the Kuiseb and Swakop rivers. In light of this revision, the advertisement call of P. carpi sensu stricto has not been recorded or described before; Gramentz (2008) briefly described the call of P. sceletussp. nov. (based on a single specimen collected north of the Swakop River) under the name “P. carpi”.

Ptenopus carpi sensu stricto is the only member of this genus that lacks sexual dichromatism, with both sexes displaying a yellow gular patch (Fig. 20E). The observations of FB contradict the original description of P. carpi (Brain 1962) in which females, collected in the month of May, apparently had no gular patch. Haacke (1975), who collected P. carpi extensively from around the type locality, made observations corresponding with our own: That all P. carpi females possess a yellow gular patch. Haacke surmised that the presence of these patches on females may be seasonal. However, we have collected, photographed and observed multiple P. carpi sensu stricto females from across their range including the type locality and throughout the year (see Table S1; many additional observed individuals not tabled) and have not observed any female P. carpi sensu stricto without a gular patch; nor have we observed a gular patch in any P. sceletussp. nov. females. Since the gular patches fade shortly after specimen preservation (not mentioned by previous authors), we argue that Brain (1962) may have assumed that sexual dichromatism was present while describing the species based on the preserved specimens, and failed to notice the gular patches on the live females. Alternatively, the gular patch may be missing under rare circumstances, although under what circumstances we shall not surmise.

TM 25973, adult male, collected “on gravel plain approximately 1 mile north of the Kuisib R. [Kuiseb River] at Gobabeb, Central Namib Desert [in the Erongo Region], S.W.A. [Namibia]” (probably –23.5456, 15.0400), by Charles K. Brain in May 1959.

TM 25966–70, 25972, 25974–25979, 25981–86, 25990–93, 25995, 25997–8, 26207 (10 adults, 10 subadults and 6 juveniles), all from the same locality as the holotype.

See Table S1 for unvouchered photographed specimens (49), DNA samples (32 available, 9 sequenced), and call recordings (3) included (n = 51).

A large Ptenopus (SVL max. 64.7 mm, mean 53.7 mm, n = 33) with a moderate tail (TL 66.7% [range 61–72] of SVL) and lean appearance, with slender limbs and reduced toe fringes compared to other species. Preserved specimens are morphologically indistinguishable from P. sceletussp. nov., but usually have a longer snout or narrower head (than P. sceletussp. nov.), clearly visible from below (compare Figs 17E and 19E); live specimens can be distinguished by lack of sexual dichromatism, with females having a yellow gular patch as in males (Fig. 17E, vs. no female gular patch or very faint yellow in P. sceletussp. nov.), and by red or ochre iris colouration (Fig. 17F, vs. silver to brown in P. sceletussp. nov.). This species and P. sceletussp. nov. are the only members of the genus with a clearly banded dorsal colour pattern on the body and tail. For a more detailed morphological diagnosis and description, see Haacke (1975).

The advertisement call (Figs 3, 20I) consists of 10 (9–12) notes uttered in slow succession, with a note rate of 2.25 s^-1 (range 1.68–3.01). Note duration is short (26 ms [range 22–29]) and regular with note 1 duration deviance 12.8% (range 5.5–24.8). The inter-note intervals are the longest of any species (560 ms [range 323–682]) and usually irregular, the inter-note interval range being the highest of any species (84% [range 62–105]); intervals are longer at the start and sometimes end of the call, but notably shorter in the middle: The call starts slow, speeds up, and then sometimes slows again. Median call density is the lowest of all species (0.06 [range 0.05–0.08]) and call duration is very long (3.6 s [range 3.3–3.8]). The basal frequency is 392 Hz (range 310–517) but very soft and usually inestimable, with harmonic bands louder towards the dominant frequencies, with the upper dominant frequency at 3.5 kHz (3.3–3.8) and a clear lower dominant frequency (sometimes more emphasised than the upper) around 1.4 kHz (1.1–1.9), less than half the upper dominant or frequency. Frequency appears to remain constant throughout the notes and call, as with the (human) perceived pitch, which is low and monotonous compared to most other species. The call is also notably softer than most other species. Bandwidth (90%) is difficult to estimate consistently, but is broad: approximately 2–5 kHz.

The call seems to vary more than other species, with calls sometimes having as few as one to three notes, especially late at night. Call period (mean 349 s) varies greatly, but is usually much lower than other species, and has not been recorded any faster than 2 min during peak chorus activity. This species does not chorus as notably as other species.

Calling activity is crepuscular to nocturnal, calls starting well after sunset and often continuing throughout the night to some degree. One instance of a notable chorus was recorded near Walvis Bay airport on 19 April 2018. Chorusing started abruptly at ~25 min after sunset, and lasted for about 20 minutes, with very reduced calling activity continuing later into the night. Calling appears to be somewhat seasonal with a peak around April to August, coinciding with lower fog incidence. Calling activity is more pronounced on warmer evenings but may continue despite cold, foggy conditions.

As with other species, P. carpi calls from the burrow entrance, although it may also call (advertisement call) occasionally while roaming. This could not be observed directly, but a wandering gecko was often spotted in the area where a call was just heard. Having said that, wandering geckos are easier to spot than geckos in a burrow.

Ptenopus carpi occurs on hard gravel plains in the central Namib Desert, Namibia, from the northern banks of the Kuiseb River to the southern banks of the Swakop River, central Namib Desert (Fig. 5). It does not tolerate soft, sandy substrates. Ptenopus carpi occurs in the Atlantic fog band, from the high water mark to approximately 50 km inland. It is one the most common vertebrates on these coastal desert plains, but becomes increasingly rare to the east of its range. Here, it occurs syntopically with P. maculatus, and parapatrically with P. kochi and P. circumsyrticussp. nov. along the Kuiseb River (the latter species occurring in or south of the river, while P. carpi occurs north of it). Ptenopus carpi and P. kochi can occasionally be found only metres from each other, on two different substrates. Ptenopus carpi occurs parapatrically with P. sceletussp. nov. across much of the Swakop River, and sympatrically in the Rössmund area just south of the river.

In light of this revision, P. carpi has a restricted range, estimated at ~1400 km². While most of its range is presently encapsulated by statutory protected areas (Namib-Naukluft National Park and Dorob Park), on the ground management of the habitat in these areas is poor. In particular, mining and industrial activities and the degradation of the gravel plains by motorists occurs within this the protected areas. The result is that more than 80% of the species range is currently within zones encompassing active mining licenses, exclusive prospecting licenses, and reconnaissance licenses (Spatial Dimension 2024), and/or occupied by urban and industrial developments, an airport, and a military base. Considering the extensive geographic scope of these claims and properties, there are likely between four and ten threat locations according to IUCN guidelines (IUCN 2022). Given the continued decline in the extent and quality of its habitat, and the range of threats noted here, a full IUCN assessment could result in the species qualifying for a range of threat categories falling between NT and EN (B1ab).

The breeding season for P. carpi appears to be between April (earliest recorded calls) and August, with some juveniles having been found between November and February.

Ptenopus carpi does not call nearly as often as congeners (similar to P. sceletussp. nov.) but appears to spend more time wandering on the surface, even far from its burrow. It is active much later into the night than congeners, often only emerging well after dark, and its activity is not dampened by cold or foggy conditions (as in most other species, except P. sceletussp. nov. and, occasionally, P. adamanteussp. nov.).

The burrow is usually shallower and simpler than those of other species but may have two to three side-branches and typically has an escape exit (as in other species). Gravid females observed (FB) generally contain two eggs, which are presumably laid in the burrow, as in other species.

Ptenopus carpi utters two different calls: The advertisement call and the single-pulse call, also previously noticed for the closely related P. sceletussp. nov. (Gramentz 2008). Gramentz (2008), based on observations in captivity, suggested the single-pulse call may be a threat call, but FB’s observations suggest this is not the function. This call is not notably uttered in response to disturbance or the close presence of a threat, and the first individuals to utter this call were often far from the observer. This call is uttered frequently in the wild by both P. carpi and P. sceletussp. nov., usually late at night after chorusing has ceased. Individuals clearly respond to each other, with several answering the first call uttered within a second or two. It is possible that this is a spacing mechanism, or a way for individuals to keep track of the population’s general location as they wander on the surface. A squeak may be uttered if the animal is molested (as with other species) but this is not similar to the single-pulse, which is identical to a single note in the advertisement call. For more details on the natural history of P. carpi, see previous works (Brain 1962; Haacke 1975).

Skeleton Coast barking gecko

Afrikaans: Skedelkus blafgeitjie

Ptenopus carpi – Gramentz (2008)

Ptenopus carpi (in part) – Haacke (1975: 230)

Ptenopus sceletussp. nov. is closely related to P. carpi but occurs predominantly north of the Swakop River, while P. carpi only occurs to the south. A single call from one individual in captivity was previously described by Gramentz (2008) as ‘P. carpi’, although the call and its variation is herein re-assessed and fully described in a standardised taxonomic framework.

NMNW R12100, adult male, collected from NE of Swakopmund (Skeleton Coast), Erongo Region, Namibia (–22.6259, 14.5457), by Francois S. Becker on 21 April 2023.

NMNW R12101–3, adult males except for NMNW R12101 (adult female, allotype), same collection details as the holotype.

See Table S1 for vouchered (2) and unvouchered photographed (50) specimens, DNA samples (34 available, 8 sequenced), and call recordings (7) included (total n = 62 excluding types).

This species is named in reference to the Skeleton Coast, which generally refers to the coastal Namib Desert between the Swakop River and the Kunene River, encapsulating the species’ entire distribution and type locality. In addition, the white or grey dorsal surfaces of the head and feet give the gecko a ghostly or skeletal appearance (Fig. 19B). We therefore use the specific epithet “sceletus”, the Latin noun in apposition meaning “skeleton”.

A moderately large Ptenopus (SVL max. 57.1 mm, mean 52.7 mm, n = 18) with a moderate tail (TL 70% [range 62–80] of SVL, n = 12), an overall lean appearance, and comparatively slender limbs. In preserve state it is morphologically indistinguishable from P. carpi except by having a generally broader or shorter head/snout, when viewed from beneath (Fig. 19E). Live specimens are more easily distinguished, with only males possessing a bright yellow gular patch, as is typical for the genus (Fig. 19E, vs. gular patch in both sexes in P. carpi sensu stricto); by silver to brown iris colouration (Fig. 19F, vs. red or ochre in P. carpi). It is distinguished from all other congeners except for P. carpi, by a combination of the following characters: Banded dorsal colour pattern on the body and tail (vs. spotted or speckled pattern in P. garrulus and P. kochi, and large, paired, ovoid light markings interspaced by darker mottled patches in other species); weakly fringed toes, with fringe length being generally less than half the breadth of the toe between fringes (vs. more than half for other species); nostrils not covered internally by a projection from the upper nasal scale (vs. is covered partially or completely in other species); being larger than most congeners (except P. kochi), and having longer, leaner limbs.

(Fig. 18). Adult male, SVL 52.4 mm with original tail 37.2 mm (70.0% SVL). Body and head covered with minute hexagonal to round scales of a similar size, dorsally and ventrally, with scales on limbs notably larger. Body slender, MBSR 123, IOS 44, HL 16.9 mm, HW 12.3 mm, HH 7.5 mm, EED 6.6 mm. Upper labials 9 enlarged, the last being smaller at the angle of the jaw; lower labials 9 enlarged and one smaller scale posteriorly; four granules bordering the mental. Clear, but not prominent, superciliary ridges tapering from above mid-eye level. Around the eye is a single row of elongated scales around the anterior, dorsal, and posterior margins, but more rounded and smaller along the ventral margin. Nasal scales very slightly swollen, the prenasals barely separated by one tall, thin, triangular internasal scale in narrow contact with the somewhat round, septagonal rostral. Eyes large, EYE 3.9 mm. Ear opening oblique (~45°), short and relatively broad, set on the far posterior corner of the head, with small, rounded, projecting scales at the anterior margin. The neck region behind the cheeks is barely swollen with internal calcium deposits. Arms, legs, and tail lean. Toes elongate, moderately flattened and weakly fringed; small, pointed fringed scales on finger margins; strong nails on fingers and toes, being larger and thicker on the fingers.

In life, the holotype (very similar to paratype NMNW R12103; Fig. 19A) has a dorsal background colour of light yellow to cream, with light orange speckles along the spine and dark purple-brown speckles or mottles all over, which coalesce to form a clearly banded dorsal colour pattern, with three dark bands on the body (the posterior being nearly split into two bars), eight bars on the tail, and a faintly defined bar on the head between the eyes. The ventrum is immaculate white with a bright yellow gular patch that is nearly divided in the middle; some of this yellow colouration is also faintly visible on the dorsal surfaces of the limbs and snout. The iris is silver.

In preservative (Fig. 18), the lighter colours have faded to cream or off-white and the darker colours faded to various shades of grey-brown. The yellow and orange colours have completely faded. The iris is a milky blue-white.

Refer to Table S1 for range of morphometric characters, including the paratypes. Morphometric characters vary relatively little among specimens, although the colours vary somewhat (Fig. 19). The dorsal colour pattern consists of 3–4 dark brown or purplish bands on a beige, yellowish, or light orange background, densely speckled with dark brown, grey, purple and/or sometimes green; sometimes the speckling is very dense, giving the appearance that the gecko is not or barely banded; an additional six to eight dark bands occur on the tail. Banding tends to be clearer and colouration brighter in younger individuals (Fig. 19D). Males have bright yellow gular patches, sometimes partially or entirely split in the middle, while females do not, or have very faint yellow on the sides of the throat (Fig. 19E). Males also have light yellow infusions on the head, limbs, and body, while females tend to be paler overall (Fig. 19A–B); in populations further north, dorsal colouration tends to be duller (Fig. 19C). Iris colouration varies from silver to brown (Fig. 19F), with individuals in the southern populations tending to have silver eyes.

The advertisement call (Figs 3, 20J) consists of 10 (8–13) notes uttered in slow succession, with a note rate of 2.60 s^-1 (1.25–4.32). Note duration is short (28 ms [range 21–40]) and regular, note 1 duration deviance 8.7% (range 0.5–16.2). The inter-note intervals may be very long (387 ms [range 204–661]), but are shorter and more regular at the beginning, then tending to longer intervals as the call progresses: Call starts fast, then slows down. Hence the inter-note interval range is usually high (102% [range 26–163]). Median call density very low (28.3 [range 20.9–40.3]) and call duration is both long and variable (3.9 s [range 2.5–7.2]). The basal frequency is 405 Hz (range 357–516) but very soft and usually inestimable, with harmonic bands louder towards the dominant frequency. The upper dominant frequency is at 4.3 kHz (3.8–4.8), and a clear (sometimes louder) lower dominant frequency is evident at around 2.0 kHz (1.4–1.9), usually less than half the upper dominant frequency. Frequency appears to remain constant throughout the notes and call, as is the case with the (human) perceived pitch, which is low and monotonous compared to most other species. Bandwidth (90%) is difficult to estimate consistently, but is broad: approximately 0.3–6.9 kHz.

The call appears to vary more than the calls of other species, except for P. carpi, with calls uttered late at night sometimes having only one to three notes. Call period (mean 365 s) varies greatly, but is usually lower than other species, and has not been recorded any shorter than 104 s during peak chorus activity. This species does not chorus as notably as most other species. Calls can be heard from after sunset to late at night. Seasonal phenology of calling activity is not known, but calls have been recorded in May and October. Calling activity is more pronounced on warmer evenings, but may continue despite cold, foggy conditions.

As with other species, P. sceletussp. nov. calls from the burrow entrance, although it may also call occasionally while roaming. This could not be observed directly, but a wandering gecko was often spotted in the area where a call was just heard–although wandering geckos are easier to detect than geckos in a burrow.

This species occurs from the southern banks of the Swakop River, northwards in a narrow strip along the Skeleton Coast (northern Namib Desert), to the southern tip of the northern Namib erg, near the Kunene River (Namibia-Angola border; Fig. 5). Its habitat falls within the coastal fog zone and it occurs on compact, gravelly substrates.

This species occurs parapatrically with P. carpi across the Swakop River and sympatrically in the Rössmund area, just south of the river. Ptenopus sceletussp. nov. occurs in sympatry or parapatry with P. maculatus in the eastern portions of its range (P. maculatus occurs only on silts in the northern limits of its range, while P. sceletussp. nov. remains on gravels in that area).

The peak calling and breeding season of Ptenopus sceletussp. nov. is probably between May and October (span of current call recordings), and juveniles have been found in November. This is slightly later than the apparent breeding season for P. carpi, but also coincides with months of somewhat lower fog incidence.

This species does not call nearly as often as most congeners, but appears to spend more time wandering on the surface, even far from the burrows (as in P. carpi and P. adamanteussp. nov.). It is active much later into the night than most congeners, often only emerging well after dark, and its activity is not dampened by cold or foggy conditions (as in most other species).

The burrow is usually shallower and simpler than those of other species, but may have two to three side-branches and typically has an escape exit (as in other species). As in P. carpi, it utters two different calls: The advertisement call and the single-pulse call, the latter usually uttered late at night after the main calling activity has ceased, and especially during dense fog.