Tarentola bibronii Smith, 1846: Illustr. Zool. S. Afr. Pl. L, fig. 1 and 2 accompanying unnumbered text pages. Lectotytpe: BMNH 1946.8.26.20 (collector Andrew Smith) here designated (see Comments below). Paralectotypes: BMNH 1946.8.26.21–28 (collector Andrew Smith). Type locality: “Southern Africa,” here restricted to Southern Africa (South Africa or Namibia) south of –26.4°S and west of 26.4°E.

Pachydactylus bibronii Smith (1849: unpaginated index and errata slip).

Homodactylus bibronii (part) Gray (1865:612).

Pachydactylus bibroni (part) Boulenger (1910:460).

Pachydactylus bibronii var. stellatus Werner, 1910:309. Lectotype: NMW 17995:4 (collector Leonhard Schultze, 1903–1905; from the collection of Franz Werner) here designated (see Comments below). Type locality: “Groß-Namaland,” [= Great Namaqualand] Namibia. On the basis of the lectotype locality, the type locality is here restricted to “Bethanien” [Karas Region, Namibia]. Paralectotypes: See Comments below.

Pachydactylus stellatus Hewitt (1911:43).

Pachydactylus bibroni stellatus Hewitt (1927:401).

Pachydactylus bibronii bibronii (part) FitzSimons (1935a:527).

Pachydactylus bibronii stellatus FitzSimons (1935a:528).

Pachydactylus bibronii turneri (part) Parker (1936:129).

Pachydactylus bibronii var. turneri (part) FitzSimons (1943:109).

Chondrodactylus bibronii Bauer and Lamb (2005:117).

Diagnosis. A large Chondrodactylus (SVL to 108 mm, TM 18185) bearing prominent subdigital lamellae. Body very robust, somewhat depressed, habitus most similar to C. turneri among its congeners. Head large, subtriangular, both wide and high, area behind orbits squarish, with nearly parallel lateral sides (Fig. 5A), in contrast to the wide rectangular parietal table of C. fitzsimonsi and more rounded shape of other congeners; snout short and rounded with a shallow midline concavity. Canthus rostralis moderately well-developed, scales on snout and canthal region smooth, domed, equal to or larger than those on parietal region but smaller than those on occiput, which are heterogeneous, strongly keeled and stellate, with prominent striae radiating from the central keel. Circumauricular scales prominent and strongly keeled to mucronate. Scales between posterior rim of orbit and ear greatly enlarged, oblong, with prominent keels. Chin and gular scales minute and granular (Fig. 7A), approximately 5 contained within half the diameter of a paravertebral dorsal tubercle. Dorsal tubercles large, round, and very strongly keeled or mucronate (Fig. 6B), largest in paravertebral position just posterior of midbody, becoming mucronate on flanks (Fig. 6A) and in lumbar region; tubercles in 14–16 longitudinal rows (several shorter rows continue onto the flanks, but only near the midbody), tubercles within a single row usually separated by granular scales from each other, but tubercles of adjacent longitudinal rows often in touch with one another. Vertebral midline covered by a series of small keeled tubercles alternating with even smaller smooth scales, each several times the size of intertubercular granules (Fig. 6B). Scales on dorsal surfaces of thighs, shanks, upper arms, and forearms mucronate. Tail approximately equal to or slightly greater than SVL, strongly verticillate, each whorl at tail-base bearing 6–8 enlarged, raised mucronate tubercles; tubercles per whorl decreasing distally. Across the body as a whole, the scalation of C. bibronii is typically more heavily sculptured than in its congeners giving it a distinctly rugose appearance that contrasts strongly with the button-scaled morphology of western clade C. laevigatus, the only toe-padded congener with which it is sympatric.

Dorsal coloration usually light to mid-brown or grayish with moderately well-developed to bold dark brown dorsal crossbars, especially dark on the nape and shoulders. Basic pattern similar to congeners with nape, shoulder, mid-body, mid-abdomen, and hip bands, which may appear as wide bands, each becoming paler anteriorly or as a series of chevrons or zigzag lines formed only by the darker posterior border of each band. Bright white markings, when present, typically immediately posterior to dark bands. Tail banded, with 7–8 dark bands fading ventrolaterally (Fig. 9). Iris bronze to coppery.

Distribution. Chondrodactylus bibronii has a temperate distribution, occurring from the southern Eastern and Western Cape Provinces of South Africa, northward through most of the Northern Cape and the far western Free State, as well as sporadically along the western periphery of North West Province. In Namibia it occurs south of the Kuiseb River, although it is absent from the southern sand sea except for some rocky isolates. It enters extreme southwestern Botswana along the Nossob River and near the junction of the Nossob and Molopo River (Fig. 4, bottom). It is broadly sympatric and even syntopic with C. laevigatus in southern Namibia (Methuen and Hewitt, 1914) and the Richtersveld of the far northern Northern Cape (Bauer and Branch, “2001” 2003). Although its range approaches that of C. turneri in North West Province, South Africa, there are no known examples of sympatry or parapatry.

Comments. The syntypes of Tarentola bibronii Smith, 1846 are without specific locality. Although the majority of Smith's period in South Africa (1821–1837) was spent within the bounds of the then Cape Colony, which would have harbored only C. bibronii, his expedition to Namaqualand in 1828 would have brought him into the extreme southern limits of the range of C. laevigatus. In 1831 on his expedition to Natal, he reached Zululand and Delagoa Bay at the southern limit of C. turneri and on his expedition to the interior of South Africa (1834–1835) any northeastern localities visited after departing Kuruman (Kirby, 1965) would have all been in the range of C. turneri. Boulenger (1885) listed Smith's types but did not provide the precise number of specimens present. At the time of Boulenger's writing Pachydactylus laevigatus had not yet been described and Homodactylus turneri was regarded by him as a synonym of Tarentola bibronii. Later, FitzSimons (1937) reported on the status of Smith's type material in London and Edinburgh. As of 1935 he recorded 9 surviving specimens in the type series. Following World War II these were reregistered as BMNH 1946.8.26.20–28, and these specimens were examined by one of us (AMB) in 2016. As reported by FitzSimons (1937), of the 9 specimens 5 are adult, 1 is a subadult, and 3 are juveniles. FitzSimons found one of the adults to be a good match for the larger animal in Smith's (1846) plate L, fig. 1 and considered it “quite probably the type.” However, FitzSimons did not indicate which specimen this was, nor did he, or any previous or subsequent author formally designate a lectotype for the taxon (Uetz et al., 2019).

Smith's (1846) illustration shows an adult and a juvenile. The latter is not rendered in sufficient detail to determine much. The adult, however, clearly has strongly keeled tubercles, a feature not found in Namaqualand C. laevigatus. The written description, while relatively detailed, is insufficient to unambiguously distinguish C. bibronii from C. turneri, although the reference to the chin and gular scales as “minute” is certainly consistent with the former species. Smith's illustrated adult has an original tail. This characterizes only 2 of the adult syntypes, BMNH 1946.8.26.20 and BMNH 1946.8.26.22. Both Specimens unambiguously exhibit the diagnostic minute chin and gular scales of C. bibronii. In contrast, the other members of the type series are referable to C. turneri. Given that the syntype series is composite, we here designate a lectotype to stabilize the established use of the name Chondrodactylus bibronii for the primarily western South African species. BMNH 1946.8.26.20 is in good condition (Fig. 9), with the original tail broken, but present with the specimen. BMNH 1946.8.26.22 is in poor to fair condition with the tail still attached, but with the body discolored and the skin somewhat pulpy with the tubercles mostly flattened. It is likely that the latter specimen is that figured by Smith because the tail of the former is distinctive in its prominent tubercles, whereas that illustrated is more typical and similar to BMNH 1946.8.26.22. Given the somewhat deteriorated state of the other specimen, we here select BMNH 1946.8.26.20 as the lectotype of Tarentola bibronii. Although it is highly likely that the specimen was collected within the confines of the Cape Colony as it existed in Smith's time, the possibility that it was collected beyond the bounds of the colony remains, and there even exists the possibility that it was collected outside of the present Republic of South Africa because Smith's journey to Namaqualand crossed the Orange (Gariep) River into extreme southern Namibia. Thus, it is possible to restrict the type locality only slightly from the original “Southern Africa” to Southern Africa (South Africa or Namibia) south of –26.4°S and west of 26.4°E.

Werner (1910) described Pachydactylus bibronii var. stellatus as a subspecies from Great Namaqualand, southern Namibia, recognizing it as distinct from both P. laevigatus and the nominotypical form. Parker (1936) subsequently synonymized P. b. stellatus with P. turneri based on their shared widespread stellate tubercles. Parker assigned Namibian material from 4 localities to P. turneri. In fact, 3 of these (Otjosongombe, Otavifontein, and Lake Otjikoto) represent a clade of C. laevigatus that is characterized by particularly spinose scalation (see C. laevigatus species account), whereas the fourth, Maltahöhe, supports some of the most northerly populations of C. bibronii (Fig. 7A). Indeed, all southern Namibian Chondrodactylus that have stellate tubercles anywhere on the body are, in fact, referable to C. bibronii. Thus, Werner's (1910) types from Great Namaqualand are also C. bibronii and the synonymization of P. stellatus with P. turneri by Parker (1936) and followed by FitzSimons (1943) and Loveridge (1947) was in error. This name is here allocated to the synonymy of C. bibronii (Smith, 1846). This has been confirmed by examination of NMW 17995.4 (Gemel et al., 2019), which is unambiguously referable to this species (Fig. 10). In order prevent future instability should another member of the original syntype series be found to be referable to C. laevigatus, we here designate NMW 17995.4 as the lectotype of Pachydactylus bibronii stellatus. The identification of some of Franz Werner's type material can be difficult because he was not based at a museum himself, and largely described material housed in a diversity of collections (other than that in Vienna, from which he was barred until 1919; see Adler, 1989) or material from his own large private collection, portions of which were sold to institutions around the world. Delisle et al. (2013, 2016) listed BMNH 1923.3.16.7 (collector Leonhard Schultze) and NMW 17995:1–2 as syntypes. However, the BMNH registers indicate no specimen with the registration number listed and no corresponding specimen could be located among either the type or non-type material in London (P. Campbell, pers. comm. 1 June 2021). The solution seems to be that 23.3.16.7 was the Werner collection number associated with NMW 17995:3 and 17995:4. NMW 17995:1, 17995:2, and 17995:3 were all collected by Schltze and at least the last two came from the Werner collection, but only 17995:4 is listed as part of the type series by Gemel et al. (2019). It is certainly possible that these additional specimens, all of which bear only the locality “Deutsch SW-Afrika,” may be paralectotypes.

The chin scale/paravertebral dorsal tubercle ratio is sufficient to unambiguously identify specimens to Chondrodactylus bibronii. Intraspecific divergences in C. bibronii are particularly shallow, with no evidence of taxonomically relevant variation. Benyr (1995) illustrated examples of variation in the species and we illustrate the very minor variation seen across the range of the species (Fig. 8).

Pachydactylus laevigatus Schmidt (1933:5, pl. I, second from left).

Pachydactylus laeviegatus tessellatus FitzSimons, 1938:172, fig. 6. Holotype: TM 17202 (collector V. F. M. FitzSimons). Type locality: “Kamanyab” [= Kamanjab], Kunene Region] Namibia. See Mashinini and Mahlangu (2013) for additional data on the type series.

Pachydactylus laevigatus laevigatus (part) FitzSimons (1943:109).

Pachydactylus laevigatus fitzsimonsi Loveridge, 1947:400. Nomen substitutum (see below).

Pachydactylus fitzsimonsi Benyr (1995:50); Branch (1998:255).

Chondrodactylus fitzsimonsi Bauer and Lamb (2005:117).

Diagnosis. A moderate-sized Chondrodactylus (SVL to ≥89.3 mm SVL; CAS 176273; Bauer et al., 1993) bearing prominent subdigital lamellae. Body robust, somewhat depressed; head large, triangular, very broad across adductor musculature and angle of jaws (approximately as broad as long), snout moderately elongate, rounded (Fig. 5C), canthus rostralis relatively prominent, interorbital region weakly concave. Crown of head wide, flat; parietal table rectangular. Chin and gular scales enlarged, juxtaposed, either rounded or polygonal, a row of 5 chin scales approximately twice the width of a paravertebral dorsal tubercle (Fig. 7C). Dorsal head scales large, flattened (chiefly in midline) weakly domed (laterally), never keeled, larger on loreal region than on crown, largest above ears and across occiput. Anterior margin of ear bearing 3–4 enlarged conical tubercles. No discrete rows of enlarged dorsal tubercles. Dorsal scales large, heterogeneous in size, flat, and juxtaposed, rounded to polygonal, in some cases forming a virtual pavement of juxtaposed scales; interstitial granules absent or rare in northern specimens. Dorsal scales rounded, slightly raised but flat-topped in southern populations, never bearing a keel or mucro; interstitial granules often present (Fig. 6C). Scales on thighs flattened to weakly conical, smooth, not keeled or mucronate. Tail weakly (Fig. 11C) to strongly (Fig. 11D) verticillate, each whorl at tail-base bearing six enlarged conical (dorsal) to mucronate (lateral) tubercles, less prominent than in congeners; tubercles per whorl decreasing distally.

Dorsal coloration either dark, with almost black cross-barring or pale, oatmeal-colored with darker dorsal markings. Typically with a dark band on nape ventrolaterally coalescent (or nearly so) with a wide band across shoulders, the space between enclosing a lighter brown area. Additional broad bands at midbody, above mid-abdomen, and across hips, posterior border of each band darkest, anterior border from bold to diffuse, intermediate area paler; in some specimens only the dark posterior borders of these markings remain in adults. Small dark spots present or absent within pale spaces between body bands. White scales, when present, adjacent to darkest posterior borders of dorsal bands. Tail with alternating light and dark bands. Light bands more- or-less uniform; dark bands with bold anterior and especially posterior borders and fading to intermediate brown centrally. Approximately 7–8 dark bands on original tails. Pale lines from nostril through eye and on to upper temporal region less pronounced and more diffuse than in congeners. Iris bronze to coppery.

Distribution. Chondrodactylus fitzsimonsi is restricted to the western regions of northern Namibia and adjacent southern Angola (Fig. 12, right). In Namibe Province, Angola it is found south of Moçâmedes (formerly Namibe) to the Namibian border in Iona National Park (Ceríaco et al., 2016; Marques et al., 2018) with a single record from Ongueria, just above the escarpment in Huíla Province (Laurent, 1964). The distribution commonly reported for the species (e.g., Branch, 1998) extends southward in the Kunene Region of Namibia at least as far as the Grootberg Pass and the type locality at Kamanjab. A phenotypically distinctive morph has a far western distribution extending from the northern Kaokoveld southward to the west of the Brandberg and thence to Henties Bay in the Erongo Region. The species is absent from the northern Namib dunefields and in much of its range is associated with boulder landscapes.

Comments. Pachydactylus laevigatus fitzsimonsi Loveridge, 1947 was proposed to replace the name P. l. tessellatus FitzSimons, 1938, which was preoccupied by P. tessellatus (Werner, 1910), which is currently regarded as a synonym of Pachydactylus capensis (Smith, 1846). Schmidt's (1933) plate image of “P. laevigatus” from Pico Azevedo reveals that his specimens were, in fact, C. fitzsimonsi, and we have subsequently confirmed this through examination of the relevant specimen (CM 5621). Benyr (1995) treated C. fitzsimonsi as a full species in his unpublished thesis, and this was followed by Branch (1998) and subsequent authors. The identification of C. fitzsimonsi has generally been nonproblematic because of its restricted range and highly distinctive morphology.

There is deep divergence within C. fitzsimonsi. A chiefly Erongo Region clade from near the Brandberg and in near coastal regions north of Henties Bay is sister to remaining populations. This includes specimens from Gaias Spring in the southwestern Kunene Region, adjacent to the Brandberg. A more widely distributed clade (the chiefly Kunene subclade) occurs in near sympatry with the Erongo clade (within 2 km of each other near Gaias) but extends northward to the Angolan border and inland to the top of the escarpment. A single sequenced specimen from Henties Bay falling within this subclade may represent a translocated specimen. Chondrodactylus fitzsimonsi from the area of Gaias are morphologically identical regardless of clade membership and differ from specimens from the type locality and other more inland and northerly areas (see Fig. 11). The morphological and genetic variation within C. fitzsimonsi will be the subject of a subsequent study. The most common nuclear haplotype of C. fitzsimonsi is shared by both morphs of the species; however, 2 C. fitzsimonsi haplotypes found in the southwest of the distribution are close to haplotypes of C. pulitzerae (Heinz, 2011). Along the western edge of the species' distribution occasional specimens with head shapes and dorsal scalation intermediate between C. fitzsimonsi and either C. pulitzerae or C. laevigatus, with which they may be sympatric or even syntopic, are encountered.

Homodactylus bibroni Bocage (1867a:220).

Pachydactylus bibronii (part) Boulenger (1885:201).

Pachydactylus bibroni (part) Boulenger (1910:460).

Pachydactylus bibronii pulitzeræ Schmidt (1933:6, pl. 1, far left). Holotype: CM 5619 (collectors R. and L. Boulton). Type locality: “Pico Azevedo,” Namibe Province, Angola (McCoy and Richmond, 1966). The 2 paratypes are FMNH 18478 [formerly CM 5620] (Marx, 1958) and MCZ R39728 [formerly CM 5622] (Barbour and Loveridge, 1946).

Pachydactylus bibroni pulitzerae Parker (1936:129).

Pachydactylus laeviegatus laeviagatus (part) Loveridge (1947:398).

Pachydactylus bibronii pulitzerae (part) Loveridge (1947:403).

Pachydactylus bibronii turneri (part) Loveridge (1947:405).

Pachydactylus laevigatus pulitzerae Benyr (1995:50).

Pachydactylus turneri (part) Branch (1998:254).

Pachydactylus turneri pulitzerae Bauer (“1999” 2000:56).

Chondrodactylus pulitzerae Heinz (2011:55), Ceríaco et al. (2014:670).

Chondrodactylus cf. pulitzerae (part) Conradie et al. (2016:24).

Diagnosis. A large Chondrodactylus (to ≥102.2 mm SVL; PEM R21610) bearing prominent subdigital lamellae. Body robust and somewhat depressed. Head large and triangular, not as broad as long (usually <90% broad as long), inflection at ear relatively angular when viewed from above, snout more pointed than in congeners, elongate, canthus rostralis relatively well-developed, loreal region somewhat inflated, interorbital region strongly concave. Scales from parietal region forward small (smallest medially), smooth to weakly keeled, contrasting strongly with the very large stellate scales on the occiput and nape. Unlike congeners, most dorsal head scales, except those of snout, are separated from one another by minute granules (Fig. 5B). Enlarged tubercles of nape becoming more conical to mucronate laterally. Prominent spiny tubercles over ears. Chin and gular scales small and granular, approximately 5 chin scales contained within the diameter of a single paravertebral dorsal tubercle. Dorsal tubercles large, oval, and weakly to strongly keeled, usually separated by smaller granular scales, becoming conical to mucronate on flanks. A “naked” mid-vertebral line, at least on the occiput, nape and shoulders, although often extending well down the trunk or to the tail base, several granular scale rows wide and appearing at a distance as a pale mid-vertebral stripe (Fig. 6D). This may be clearly evident along the entire trunk (Fig. 13A), or part thereof (Fig. 13B, C) or may be relatively subtle (Fig. 13D) but is always present. A similar “naked” area, although always limited to the nape is present in some C. turneri. 16–18 regular to irregular longitudinal rows of relatively flattened, smooth to keeled, but rarely stellate, oval to rounded, dorsal tubercles. Trunk tubercles almost always separated from one another in all directions by small granules. Tubercles on dorsum of thigh very large, flattened to weakly inflated, becoming keeled or mucronate on shank. Tail distinctly verticillate, each whorl at tail-base bearing 6–8 enlarged, though not strongly projecting, keeled or conical (dorsal) to strongly mucronate (lateral) tubercles; tubercles per whorl decreasing distally.

Dorsal coloration usually buff to light brown with indistinct to moderately well-developed reddish-brown to dark brown dorsal crossbars, especially anteriorly. Basic pattern similar to congeners, with nape, shoulder, mid-body, mid-abdomen, and hip bands. White tubercles, when present, typically immediately posterior of dark bands. Pale interspaces in between darker bands may form a regular pattern of oval markings connected by the pale middorsal granular line (Fig. 13A). Thick pale line from snout to dorsal portion of eye generally distinct and bordered above and below by thinner dark lines. Tail banded, with 8–9 dark bands fading laterally; boundaries between pale and dark bands usually marked by complete or incomplete dark brown edges; some darker bands may be reduced to middorsal blotches.

Distribution. The species is known from southwestern and western Angola as far north as Capanda in Malanje Province (Ceríaco et al., 2014) and Luanda and Cacuaco, Luanda Province (Ceríaco et al., 2017; Marques et al., 2018; Fig. 12, left). Both the Capanda record and one from Huambo Province are above the Angolan Escarpment. Loveridge (1947) followed by Mertens (1955, 1971) tentatively assigned material (e.g., MCZ R43401) from the Erongo Mountains in Namibia to this taxon, although he suggested they might alternatively represent an undescribed species. In fact, only C. laevigatus occurs in this region. Ceríaco et al. (2016) noted the presence of C. pulitzerae in far northwestern Namibia but did not provide details. It occurs from the border at the Kunene River south as far as Sesfontein (entire Namibian distribution in the Kunene Region) with a single locality further south at “10 km N of the Hunkab River” (but unknown where along the river, TM 52910–11). Throughout most of its range in Namibia it is sympatric with C. laevigatus and C. fitzsimonsi, although its easternmost occurrences in both Angola and Namibia are out of the range of the latter species.

Comments. Prior to its description by Schmidt (1933), specimens referable to this taxon were assigned to P. bibronii (Bocage, 1867a, b, 1887a, b, 1895; Boulenger, 1885; Mertens, 1926), based on overall similarity and the presence of mostly strongly keeled or mucronate scales across the dorsum. Parker (1936) and later authors (Mertens, 1937, 1938; Barbour and Loveridge, 1946; Loveridge, 1947; Hellmich, 1957a, b; Marx, 1958; Laurent, 1964) accepted Schmidt's trinomial and regarded P. b. pulitzerae as an Angolan endemic (but see comments about Erongo populations above), although Monard (1937) used only the binomen C. bibronii and Loveridge (1947) used not only Schmidt's name but also P. l. laevigatus and P. b. turneri in referring to some Angolan specimens. Pachydactylus bibronii pulitzerae was subsequently referred to by Mertens (1955, 1971) and Wermuth (1965) but otherwise the name then went largely unused in the published literature for 5 decades. It was regarded as a valid subspecies of P. laevigatus by Benyr (1995) and as a full species of Chondrodactylus by Heinz (2011) in their respective unpublished theses. Ceríaco et al. (2014) first used C. pulitzerae as a specifically valid name on the basis of Heinz's (2011) molecular data, and it has since been used consistently (e.g., Ceríaco et al., 2016, 2017; Conradie et al., 2016; Heinicke et al., 2017; Marques et al., 2018; Branch et al., 2019).

Specimens genetically sampled fall into a more northern clade, extending northward of the Giraul River in Namibe Province, Angola, and a southern clade occurring southward into Namibia. The genetic identity of a specimen from Quiçama (CAS 263109) with one from Cambambe (PEM R21611), 150 km distant, and their high similarity to specimens from northern Namibe, approximately 450 km away, along with the sporadic distribution of C. pulitzerae north of Lobito, raises the possibility that some northern records may represent introductions, although Bocage (1895) stated that the species (as Pachydactylus bibronii) was common south of the Kwanza River. More extensive sampling is needed both north and east of Namibe Province to determine whether the range is contiguous and whether there is genetic substructure consistent with geography.

Homodactylus turneri Gray, 1864: Proc. Zool. Soc. London 1864:59, pl. 9, fig. 2. Lectotype: BMNH 1946.8.26.7 (collector J. Kirk), here designated (see Comments below). Paralectotypes: BMNH 1946.8.26.8, 64.1.9.10–13, 64.1.9.15 (collector J. Kirk). Type locality: “southeastern Africa,” (restricted to “Tette” [=Tete], Mozambique fide Loveridge 1947:405; see Comments below).

Pachydactylus Bibronii (part) Peters (1865:457).

Homodactylus bibronii (part) Gray (1865:612).

Pachydactylus bibronii bibronii (part) FitzSimons (1935b:336).

Pachydactylus bibronii turneri (part) Parker (1936:129).

Pachydactylus bibronii var. turneri (part) FitzSimons (1943:109).

Pachydactylus bibronii turneri (part) Loveridge (1947:405).

Pachydactylus laevigatus turneri Benyr (1995:50)

Pachydactylus turneri Branch (1998:254).

Pachydactylus turneri turneri Griffin (2003:30).

Chondrodactylus turneri Bauer and Lamb (2005:117).

Diagnosis. A moderately sized Chondrodactylus (to ≥95 mm SVL, MCZ R190407) bearing prominent subdigital lamellae. Body robust and somewhat depressed. Head large, relatively deep, and subtriangular, nearly as broad as long, inflection at ear gently curved when viewed from above (contrasting with C. pulitzerae), snout typically shorter and broader than in congeners, canthus rostralis moderately developed more variable than in congeners; loreal region moderately to strongly inflated, interorbital region flattened to weakly concave. Tubercles on occipital region very large and keeled to stellate, becoming smaller on the crown and interorbital region and slightly larger again on the dorsum of the snout; interorbital and especially snout scales distinctly domed, most bearing weakly defined keels. Most anterior dorsal head tubercles in contact with one another, whereas tubercles of the occiput and nape usually well-separated from one another by tiny granules (Fig. 5D). Tubercles around ear heterogeneous, generally less massive than in other congeners. Chin and gular scales small and granular, becoming progressively smaller postero-medially (Fig. 7B), approximately 5 chin scales contained within the diameter of a single paravertebral dorsal tubercle. Dorsal tubercles large, oval to rounded, strongly keeled, variably bearing small peripheral pustules or short to long radiating ridges in a stellate pattern, becoming smaller and more conical to mucronate on flanks. Trunk tubercles usually well-separated by smaller granular scales, forming 14–18, usually very regular longitudinal rows of enlarged tubercles (Fig. 14). In addition, mid-vertebral line with much smaller, rounded, keeled tubercles separated from one another by alternating pairs of paravertrebral keeled tubercles intermediate in size between the tubercles of the mid-dorsal and more lateral tubercle rows (Fig. 6E). Tubercles on dorsum and postaxial surface of thigh and shank large, somewhat flattened, keeled or stellate. Scales on upper arm non-tuberculate, imbricating, becoming tubercular on forearm, keeled to mucronate, but much smaller than tubercles of shank. Tail distinctly verticillate, each whorl at tail-base bearing six (eight close to tail base) enlarged, though not strongly projecting, keeled or conical (dorsal) to strongly mucronate (lateral) tubercles; tubercles per whorl decreasing to 4 then 2 on distal portion of tail.

Dorsal coloration buff to light to medium brown, sometimes with reddish or grayish tones with indistinct to moderately well-developed dark brown to almost black dorsal crossbars, especially prominent anteriorly. Basic pattern similar to congeners, with nape, shoulder, mid-body, mid-abdomen, and hip bands. White tubercles, when present, typically immediately posterior to or within dark bands. Tail banded, boldly or obscured, with 8–10 dark bands fading laterally; boundaries between pale and dark bands usually marked by complete or incomplete dark brown edges; some darker bands may be reduced to middorsal blotches.

Additionally, C. turneri exhibits a genetic autapomorphy—a rearrangement of the genes coding for transfer RNAs downstream of ND2 (see Discussion), which is not only unique among its congeners, but also among all gekkotans.

Distribution. Chondrodactylus turneri occupies most of the southeastern margin of the distribution of the genus (Fig. 16). Its range in South Africa and Eswatini (formerly Swaziland) has been presented by Branch (2014; only records east of 25°E apply to this species, others are actually C. laevigatus). It occurs in northern KwaZulu-Natal (i.e., Zululand), central and eastern Eswatini, throughout most of Mpumalanga and Limpopo (except parts of Sekhukuneland), in northern Gauteng and in northeastern North West Province. In Botswana C. turneri occurs along the southeast margins of the country and in Zimbabwe it occupies the southern and central parts of the country. Nearly all Chondrodactylus records in southern Mozambique are also likely referable to this species, although this is based on biogeographic grounds and weak support from morphology. The precise dividing line between C. laevigatus and C. turneri in Zimbabwe and central and northern Mozambique remains unclear but, based on morphology and limited genetic sampling, the division in the former country may roughly correspond to the division between the Zambezi and Limpopo drainages.

Sampling in Mozambique is uneven and genetic material is derived only from Nampula in the north of the country. Specimens from throughout the northern provinces of Niassa, Cabo Delgado, and Nampula provinces all appear to be C. laevigatus. Specimens from south of Beira are conspecific with those in Masvingo and southern Manicaland in Zimbabwe and with those in the South African lowveld (i.e., C. turneri). This is supported by the relatively contiguous lowland habitat and, to the extent possible, in the absence of unambiguous diagnostic features, by morphology. Specimens from the type locality of Homodactylus turneri Gray, 1864 at Tete and from lowland localities downstream of there are also almost certainly referable to this species. However, specimens from closer to Cahora Bassa in western Tete Province and from northern portions of Manhica and Sofala provinces are problematic. Geographically, they lie between C. laevigatus from the Vumbas in Zimbabwe (genetically confirmed) to the southwest and Malawi (morphologically determined) and northern Mozambique to the northeast. However, the Zambezi Valley provides a low-elevation riparian corridor deep into the inland of Tete Province, with Tete itself lying at only 140 m. Elevation may explain some aspect of the distribution boundary between C. laevigatus and C. turneri, but it is clearly not the sole factor because C. laevigatus occurs at quite low elevations in some places, whereas C. turneri is present in Bulawayo and the Waterberg Massif of Limpopo, both >1,000 m in elevation.

Comments. The published type locality for C. turneri (Gray, 1864) is “SouthEastern Africa,” interpreted to be Tette (=Tete), Mozambique by Loveridge (1947) and entered as such in the British Museum of Natural History (now The Natural History Museum, London) re-registration catalogue. In fact, Gray (1864) explicitly listed Tette (now Tete) as a locality for only some of the specimens collected by Kirk. The rest of the specimens, including Homodactylus turneri, are associated only with the general locality “South-Eastern Africa,” which could be anywhere within the area visited by John Kirk on the Zambezi Expedition during the period 1858–1863. Although Kirk spent a good deal of time at Tete and in the Shire Highlands in what is today Malawi, he also collected specimens below Tete on the Lower Zambezi, including near Sena (Senna) and the mouth of the Shire River. In addition, during the early period of the expedition, Kirk was stationed in the Zambezi Delta and on numerous occasions during the expedition he returned to the river mouth. Other biological specimens returned by Kirk have the localities Quellimane (now Quelimane) and Gorongoza (Günther, 1864; Hill, 1922; Dritsas, 2005). When locality data were available for Kirk's material, he apparently did provide it to the scientists who were working with the material (Dritsas, 2005). The absence of a specific locality suggests that either the specimens represented something widespread in the area covered by Kirk's journeys or that the data were not recorded or were subsequently lost. It must also be remembered that Kirk's primary natural history interest was in botany (Hill, 1922), so it is perhaps not surprising that herpetological material, largely collected incidentally to plants, might not be as carefully recorded. Based on a lack of explicit locality data accompanying the syntypes of H. turneri, we think it likely that the type series of H. turneri may have originated from multiple sites in the lower Zambezi Valley.

All 7 syntypes of Homodactylus turneri are in good condition, nonetheless, their specific identity is still difficult to establish with certainty. All appear to be referable to Chondrodactylus turneri as recognized here on the basis of ≥1 morphological traits; however, there is overlap with C. laevigatus in each of these traits. Given the extreme similarity of these two congeners and the uncertainty surrounding their precise distributional limits (as well as the extent of the area from which the types were collected), we consider it prudent to select a lectotype to stabilize the current use of the name C. turneri. We here designate BMNH 1946.8.26.7 (Fig. 15) as the lectotype of Homodactylus turneri on the basis that it is characterized by the greatest number of C. turneri features, specifically, head deep, snout short and broad, lores inflated, and relatively extensive exposed granular skin on occiput, nape, and anterior midline of the back.

This species was synonymized with Pachydactylus bibronii by Peters (1865) and, for the majority of the period since, it has remained in the synonymy of this species (see synonymies of FitzSimons, 1943 and Loveridge, 1947). Gray (1865) himself, noted that he had somehow overlooked Smith's (1846) description and illustration of Tarentola bibronii and sank H. turneri into its synonymy; however, he retained the generic name Homodactylus, believing the species to be generically distinct. Gray (1865) further acknowledged that Peters, who had examined Smith's types, had pointed out the synonymy. Loveridge (1947), largely following Parker (1936) and FitzSimons (1943), recognized it as a subspecies of C. bibronii but included P. bibronii stellatus in its synonymy and his concept of the taxon included specimens from Angola and Namibia currently allocated to C. laevigatus as well as true C. bibronii from the Free State and Northern Cape provinces of South Africa. Benyr (1995), however, identified a reliable scale character (the size of the anterior chin scales relative to the paravertebral dorsal tubercles) that unambiguously separated P. bibronii from what he considered to be P. laevigatus, with 3 subspecies, the nominotypical form in the Northern Cape and western Namibia, P. l. pulitzerae in Angola, and P. l. turneri in the east. Benyr (1995), however, had overlooked that Homodactylus turneri Gray, 1864 had priority over Pachydactylus laevigatus Fischer, 1888 and thus, as corrected by Branch (1998) and discussed by Lamb and Bauer (2002), the taxon names became P. t. turneri, P. t. pulitzerae, and P. t. laevigatus. Since the transfer of the P. bibronii group to Chondrodactylus the majority of the literature on Southern African reptiles has applied the name C. turneri to scansorial Chondrodactylus (other than C. bibronii) distributed from Kenya to South Africa (e.g., Branch, 2014).

The genetic distinctiveness of C. laevigatus from C. turneri sensu stricto was demonstrated by Heinz (2011) in his unpublished thesis; however, the two names were not used in the published literature as specifically valid to refer to different taxa until several years later (see C. laevigatus account), although without explicit justification. Phenotypically, C. laevigatus from the east of its range strongly resembles C. turneri; thus, unambiguous identification of specimens from areas of parapatry, from Botswana and Zimbabwe through East Africa, can be difficult, especially in the case of juveniles.

Pachydactylus capensis Peters (1854:615). [previously included in the synonymy of C. turneri, e.g., Loveridge (1947)].

Platydactylus (Pachydactylus) Bibronii Peters (1862:15). [previously included in the synonymy of C. turneri, e.g., Loveridge (1947)].

Pachydactylus Bibronii Peters (1869:139).

Pachydactylus laevigatus Fischer, 1888:15, pl. 2, fig. 3. Syntypes: BMNH 1946.8.26.1–2 (formerly BMNH 89.12.16.9–10; collector J. Steingröver). Type locality: “bei Aus und auf dem Wege nach Bethanien” [= near Aus and on the way to Bethanien], Karas Region, Namibia.

Pachydactylus bibroni laevigatus Methuen and Hewitt (1914:129, fig. 14).

Pachydactylus stellatus Schmidt (1933:5).

Pachydactylus bibronii laevigatus FitzSimons (1935a:527).

Pachydactylus bibronii bibronii (part) FitzSimons (1935b:336).

Pachydactylus laevigatus laevigatus FitzSimons (1938:172).

Pachydactylus bibronii turneri (part) Parker (1936:129).

Pachydactylus bibronii pulitzerae (part) Loveridge (1947:403).

Pachydactylus bibronii pulitzerae Hellmich (1957b:49)

Pachydactylus turneri (part) Branch (1998:254).

Pachydactylus turneri laevigatus Griffin (2003:30).

Chondrodactylus turneri (part) Bauer and Lamb (2005:117)

Chondrodactylus turneri laevigatus Bauer et al. (2006:90).

Chondrodactylus cf. pulitzerae (part) Conradie et al. (2016:24).

Chondrodactylus laevigatus Ceríaco et al. (2014:670).

Diagnosis. A large Chondrodactylus (to ≥100 mm SVL, MCZ R190191) bearing prominent subdigital lamellae. Body robust and somewhat depressed. Head large, relatively depressed, and subtriangular, not as broad as long but proportionally wider across adductor region than C. pulitzerae; inflection at ear gently curved when viewed from above, snout somewhat longer and more acuminate than in C. turneri, canthus rostralis moderately developed, loreal region weakly to moderately inflated, interorbital region flattened to weakly concave or with a narrow furrow. Tubercles on occiput and posterior half of crown large very large and keeled to stellate, becoming smaller on the anterior crown and interorbital region and slightly larger again on the dorsum of the snout; interorbital and especially snout scales flattened or domed, keelless or with only weakly defined keels. Most anterior dorsal head tubercles in contact with one another, whereas tubercles of the crown and nape may be separated from one another by tiny granules (Fig. 5E–H), although typically not as widely spaces as in C. turneri. Tubercles anterior and dorsal to ear heterogeneous, generally large and bearing a prominent longitudinal keel. Chin and gular scales small and granular, becoming progressively smaller postero-medially (condition similar to Fig. 7B), approximately 5 chin scales contained within the diameter of a single paravertebral dorsal tubercle. Dorsal tubercles large, oval to rounded, from unkeeled to strongly keeled, if strongly keeled then variably bearing small peripheral radiating ridges, although not as strongly stellate as in C. bibronii, becoming smaller and more conical on flanks. Trunk tubercles usually well-separated by smaller granular scales, forming 14–20 (usually 18 in the western clade, 14 in the eastern clade, and 16–18 in the Kgaligadi clade), slightly irregular to very regular longitudinal rows of enlarged tubercles (Fig. 17). In addition, mid-vertebral line with a mixture of granules and smaller, rounded, unkeeled to moderately keeled tubercles (single or in pairs; Fig. 6F–H). Tubercles on dorsum and postaxial surface of thigh and shank large, somewhat flattened, to keeled or mucronate. Scales on upper arm non-tuberculate, flattened, imbricating, becoming tubercular on forearm, conical to mucronate, but smaller and less prominent than tubercles of shank. Tail distinctly verticillate, each whorl at tail-base bearing 6 (8 close to tail base) enlarged, keeled (proximal whorls) or conical to strongly mucronate tubercles; tubercles per whorl decreasing to 4 then 2 on distal portion of tail.

Dorsal coloration buff to light to medium brown, sometimes with reddish or grayish tones with indistinct to well-developed dark brown to almost black dorsal crossbars, especially prominent anteriorly. Basic pattern similar to congeners, with nape, shoulder, mid-body, mid-abdomen, and hip bands. White tubercles, when present, typically immediately posterior to or within dark bands. Tail banded, boldly or obscured, with 8–10 dark bands fading laterally; boundaries between pale and dark bands usually marked by complete or incomplete dark brown edges; some darker bands may be reduced to middorsal blotches.

Variation. Individuals of the western clade of C. laevigatus are highly distinctive and easily distinguished from all other congeners, including members of the “Kgaligadi” and Eastern clades. They are generally more depressed overall (Fig. 17A–B) and have the characteristic button-scale type tubercles, which are typically raised but not, or only weakly, keeled. This clade extends from the southernmost limits of the species to the Otjozondjupa Region of northcentral Namibia. The “Kgaligadi” clade extends northward from the Otjozondjupa and Erongo regions to southern Angola and eastward through the Caprivi Strip and northern Kgaligadi, and the Eastern clade continues from western Zimbabwe and central Zambia to East Africa. We could identify no diagnostic characters distinguishing members of these two clades from one another. Members of these clades usually have the dorsal tubercles more strongly keeled or conical (Fig. 17D) and often bearing transverse stria or having a stellate pattern of keels on at least some scales (particularly on the occiput and nape). Especially strongly keeled specimens of the “Kgaligadi” clade occur in populations near Uis, Takauasa (e.g., NMNW 210), and in the Grootfontein/Tsumeb region of northern Namibia (Fig. 17E). The Eastern clade, likewise, expresses a spectrum of rugosities from moderate to extensive (Figs. 17E, F, H, G and Fig. 18 in order of increasingly large and strongly textured tubercles). The Western clade button-scaled geckos, which share much of their range with the very rugose C. bibronii may be exhibiting character displacement in areas of sympatry. The greatest difficulty in the identification of Chondrodactylus species from each other is distinguishing the “Kgaligadi” and Eastern clade C. laevigatus from C. turneri. This is especially difficult in the broad area from central Zimbabwe to central Mozambique in which they likely occur in parapatry or potentially even sympatry (unverified).

Distribution. Chondrodactylus laevigatus has by far the broadest range of any member of the genus (Fig. 16). Its distribution in South Africa has been documented (as C. turneri [part]) by Branch (2014) and is limited to northern portions of the Northern Cape Province from near Springbok, east along the Orange River as far as Upington, with scattered records near the Molopo and Nossob rivers in the Mier Kalahari and Steinkopf in Namaqualand. It is probably more-or-less continuously distributed throughout this area except for sandy areas that offer no suitable refuge sites, but is most frequently encountered from Lekkersing northward to the Namibian border and in rocky areas along the south shore of the Orange River, where it is syntopic with C. bibronii (Methuen and Hewitt, 1914), even occupying refuges beneath the same rock slabs (Bauer and Branch, “2001” 2003). Virtually the whole of Namibia is occupied by C. laevigatus, with the exception of the Namib sand seas and portions of the Kalahari that are devoid of rocks and trees. Nonetheless, isolated rocky outcrops surrounded by sand for many kilometers may support populations of C. laevigatus. Verified records in Angola are few (Marques et al., 2018) but include records in Cunene (Monard, 1937), Huíla (Baptista et al., 2019; Butler et al., 2019) and Cuando Cubango (Conradie et al., 2016 [as C. cf. pulitzerae]) provinces, all above the Angolan escarpment. It is likely distributed widely across the southern Angolan Plateau, although the verified records are too few to estimate its northern limits. A record by Schmidt (1933) from Pico Azevedo in the lowlands of Namibe is referable to C. fitzsimonsi (Marques et al., 2018); however, we here report several records in coastal Namibe Province (Table 1).

To the east of Namibia and Angola, the distribution of C. laevigatus is less precisely known. It occurs throughout suitable habitats in western and northern Botswana, western and northern Zimbabwe, through Zambia and diagonally northeastward as far as southern Kenya. No examples of sympatry with C. turneri are known with certainty, but its ability to co-occur with C. bibronii and C. fitzsimonsi suggests that this may be possible, perhaps in regions of central Zimbabwe or along the course of the Zambezi River in Mozambique. The eastward extension of C. laevigatus roughly follows the Zambezi Valley and in Zambia is mostly bracketed by the Zambezi and Luangwa rivers (Simbotwe and Mubamba, 1993). Sampling in central Mozambique is poor, but existing records suggest that lowland records south of the Pungwe River are C. turneri, whereas those north of the Ligonha River are C. laevigatus. The distribution of C. laevigatus in Tanzania and Kenya appears to be disjunct and scattered (Loveridge, 1928, 1947, 1951, 1955, 1957a). This may be an artifact of collection effort or bias or it may reflect limited suitable habitats, or perhaps competition with large-bodied Hemidactylus spp. and Elasmodactylus spp.

An isolated record of a Chondrodactylus sp. from Rwanda has often been reported or plotted, albeit with some speculation (de Witte, 1941; Loveridge, 1947, 1957b; Spawls et al., 2006, 2018) and stems from a report by Sternfeld (1912) based on a specimen from the “Vulkangebiet.” The single specimen, ZMB 24330 (Fig. 18), was collected by Reinhard Houy on the 1911 expedition led by Hans Heinrich Josef Meyer to the area between Lake Kivu and Lake Victoria (Meyer, 1913). The Museum für Naturkunde catalogue entry for the specimen lists only “Dt. Ost Afrika.” Vulkangegiet, the published locality, is more specific but still vague, referring to the area to the northeast of Lake Kivu, in northern Rwanda and adjacent Nord-Kivu Province, Democratic Republic of Congo. This record has never been verified by additional vouchers or by records with specific localities and the species has not been mentioned in more recent regional literature (e.g., Laurent, 1956). In the absence of confirmation of the occurrence of Chondrodactylus in the region and in light of the unlikely disjunction from the otherwise more-or-less contiguous distribution of the genus, we regard this record as dubious and consider both the Democratic Republic of Congo and Rwanda to be outside of the range of the genus. ZMB 24330 is typical of Chondrodactylus laevigatus from East Africa (see below) and the expedition would have passed through the known range of this species from Kilamatinde to Morogo in east-central Tanzania on its way back to the coast. The mid-sized chin scales (Fig. 18B) and slightly concave snout profile (Fig. 18C) are characteristic of C. laevigatus in general, whereas the very large body tubercles (Fig. 18A) are typical of the East African clade specifically.

The exact patterns of distribution of C. turneri and C. laevigatus in Zimbabwe, central Mozambique, and southernmost Malawi are uncertain. The characters used herein to distinguish eastern (highly tuberculate) laevigatus from turneri were not yet determined more than a decade ago when this project began. Thus, some specimens were examined before what ultimately proved to be the most diagnostically valuable characters were identified. Further, not all specimens for which we have locality data (i.e., are plotted in Fig. 16) have been examined by us. This applies particularly to specimens from the Natural History Museum of Zimbabwe. From genotyped individuals, we know that C. laevigatus extends across northern Zimbabwe from Lake Kariba to the former Elim Mission in the Vumba Mountains, whereas C. turneri occurs from Bulawayo east to southern Masvingo and Manicaland provinces. Chondrodactylus laevigatus likely occupies the higher elevations throughout the Eastern Highlands of Zimbabwe, but we are currently unsure of the allocation of the populations of the Zimbabwean Midlands. Likewise, we are uncertain where a switch to (or an area of sympatry with) C. turneri might occur between the Chimanimani Mountains and the adjacent lowlands. Likewise, the identity of populations between the lower end of Lake Kariba and the vicinity of Tete and between the Lower Zambezi and the Shire Highlands of Malawi remains uncertain.

Comments. The description of Pachydactylus laevigatus was by J.G. Fischer, who was the volunteer in charge of the herpetological collections of the Hamburg Museum at the time (Adler, 2007). Fischer (1888) mentioned only 2 specimens of this species received from Steingröver, one of which was illustrated. When Fischer died the following year, his collection was purchased by the British Museum from “Madame Fischer.” The collection included a number of other Namibian specimens corresponding to material cited in the same paper. Fischer's (1888) illustration of the dorsum of one of his syntypes matches exactly the color pattern of BMNH 1946.12.26.1 (Fig. 19), as does its regenerated tail and size (89 mm SVL), confirming its identity as a type. Details of the second specimen also correspond well to BMNH 1946.12.26.2. Both syntypes are unambiguously assignable to the same taxon and cannot be mistaken for any other congener.

The type locality of Pachydactylus laevigatus Fischer, 1888 lies within the distribution of the southwestern clade and members of this clade are easily diagnosable on morphological grounds, being characterized by the button-scaled morphology, with tubercular keels and mucrones lacking. In both remaining subclades, body scalation is more heterogeneous and some populations are characterized by an extremely acuminate tuberculation, similar to that of C. turneri (see above). Given the lack of ambiguity regarding the identity of the syntypes and the fact that no syntypes are unaccounted for, there is no compelling reason to designate a lectotype in this instance.

Prior to the description of P. laevigatus, specimens of this species were referred to P. bibronii (e.g., Boulenger, 1885). As the specific epithet implies, typical C. laevigatus are characterized by their relatively smooth (so-called “button-scaled”) tubercles, which easily distinguish them from C. bibronii, which have much more strongly keeled or mucronate tubercles. However, although this scale morphology is typical in the western half of Namibia and in adjacent areas of South Africa, the phenotype of C. laevigatus becomes decidedly more “bibronii-like” from the Namibian Kalahari, Kavango, and Zambezi Regions eastwards. As a consequence, eastern P. laevigatus until 1995 had frequently been assigned to P. bibronii. Benyr (1995), however, identified a reliable scale character (the size of the gular scales relative to the paravertebral dorsal tubercles) that unambiguously separated P. bibronii from what he considered to be P. laevigatus, with 3 subspecies—the nominotypical form in the Northern Cape and western Namibia, P. l. pulitzerae in Angola, and P. l. turneri in the east of the subcontinent. Benyr (1995), however, had overlooked that Homodactylus turneri Gray, 1864 had temporal priority over Pachydactylus laevigatus Fischer, 1888. Branch (1998), however, who was cognizant of this, recognized a monotypic P. turneri for all geckos in the P. bibronii group exclusive of P. bibronii itself and the distinctive C. fitzsimonsi from northwestern Namibia. Although Branch (1998) provided no justification for the resurrection of P. turneri, the rationale was subsequently explained by Lamb and Bauer (2002), and the taxon names subsequently recognized in light of Benyr's (1995) work were P. t. turneri, P. t. pulitzerae, and P. t. laevigatus. Most subsequent authors did not use trinomials; therefore, the name Pachydactylus turneri (and, after 2005, Chondrodactylus turneri) was widely applied sensu Branch (1998) until the genetic distinctiveness of C. laevigatus and C. pulitzerae from C. turneri sensu stricto was explicitly demonstrated by Heinz (2011) in his unpublished thesis. However, C. turneri and C. laevigatus were not recognized in their current sense in the published systematic literature until Ceríaco et al. (2014) and later Heinicke et al. (2017), although again without explicit justification. Marques et al. (2018) likewise used C. laevigatus as specifically distinct from C. turneri, citing the unpublished phylogeny of Heinz (2011).

Heinz (2011) reported shared nuclear haplotypes amongst individuals from across Namibia, indicating gene flow across the various morphotypes of C. laevigatus. Likewise, he found similar haplotypes represented in northeastern Zimbabwean Chondrodactylus that we here interpret as C. laevigatus, whereas these haplotypes differed markedly from those in southeastern Zimbabwe and eastern South Africa, all of which also share a rearrangement in ND2 and are here interpreted as C. turneri. In part because of these shared haplotypes, we choose to recognize C. laevigatus as a single species. However, an argument could be made for its subdivision into 3 putative species, based on the deep mitochondrial divergence between the main subclades and their strong support (BS 100%). Under such a scenario C. laevigatus sensu stricto, the western clade, would be easily diagnosable, but the difficulties of distinguishing the more eastern forms from C. turneri would still remain.