<Unknown Taxon>

Sadlier, Ross A., Smith, Sarah A. & Bauer, Aaron M., 2006, A New Genus for the New Caledonian Scincid Lizard Lygosoma euryotis Werner, 1909, and the Description of a New Species, Records of the Australian Museum 58 (1), pp. 19-28 : 23-27

Celatiscincus similis Sadlier, Smith & Bauer, n.sp.

Figs. 2–5

Type material. HOLOTYPE: MNHN 2003.1006 (formerly AMS R153524); New Caledonia, Province Nord,Tsiba, c. 3 km north of Ouaco, 20°48'14"S 164°28'19"E; collected by R.A. Sadlier, M. Christy & M. Thurlow, 25.xi.1998. PARATYPES: AMS R153504–23, R153525, same location and date of collection as holotype; AMS R153559–66, CAS 226168–69, 226170–72, MCZ R183656, same location and collectors as holotype, 26.xi.1998; AMS R153585–86, CAS 226173, New Caledonia, Province Nord, Massif Ouazangou-Taom, Mt Taom, 20°45'35"S 164°32'51"E, collected by R.A. Sadlier, M. Christy & M. Thurlow, 26.xi.1998; AMS R164185, R164188, New Caledonia, Province Nord, Massif Ouazangou-Taom, Mt Taom summit, 20°46'52"S 164°34'46"E, collected by R.A. Sadlier & G.M. Shea, 7.xii.2004.

Additional material. In addition to the type material the following specimens referable to Celatiscincus similis were examined but do not form part of the type series or description: AMS R161164–65, R161184, Massif Ouazangou-Taom, Gomen Mine Mt Taom, 20°46'36"S 164° 33'44"E; QM J37441 Ouaième via Hienghène, 20°37'S 164°51'E; AMNH 62685 Tao 20°33'S 164°48'E—note the specimens from Ouaième (Sadlier,1986) and Tao (Bauer & Sadlier, 1994) were originally reported as C. euryotis.

Diagnosis. See diagnosis for Celatiscincus euryotis above.

Etymology. The epithet similis is Latin for similar or resembling, and refers to the great overall similarity of the new species to C. euryotis.

Description. Based on the type series of 41 specimens (32.0–45.5 mm SVL), comprising 16 adult males, 19 adult females, and 6 subadults.

Measurements. Maximum SVL of males 43.0 mm, females 45.5 mm; distance from axilla to groin 48.8–58.8% of SVL (¯x = 53.9%, n = 41); distance from forelimb to snout 37.4– 43.4% of SVL (¯x = 40.6%, n = 41); hindlimb length 40.0– 48.7% of SVL (¯x = 4.0%, n = 35); tail length 167.6% of SVL for adult individual with most complete tail.

Scalation ( Fig. 3 ): prefrontals moderately large and moderately to widely separated; frontoparietals fused; interparietal distinct; parietals each bordered by a single large upper secondary temporal and two (rarely three) equal sized scales no larger than the surrounding dorsal scales; nasals well separated; loreals two, anterior narrowly contacting labials, occasionally present as a semilunar scale failing to contact the labials; supraciliaries usually 7 (95%), rarely 6 (3.8%) or 8; upper labials usually 7; a complete subocular row between preocular and pretemporal scales; primary temporal single; lower secondary temporal single; tertiary temporals two; postlabials two; lower eyelid with an obvious, centrally located semi-transparent disc.

Midbody scale rows 30–32 (¯x = 31.0±1.00, n = 41); paravertebral scales 49–56 (¯x = 52.4±1.75, n = 41)—no sexual dimorphism between 13 males and 15 females (¯x 52.38 vs 52.25, t 34 = 0.202, P<0.99); scales on top of fourth finger 8–11 (¯x = 9.5±0.61, n = 39); lamellae beneath fourth finger 15–20 (¯x = 17.2±1.02, n = 39); scales on top of fourth toe 12–15 (¯x = 13.3±0.63, n = 35); lamellae beneath fourth toe 30–40 (¯x = 35.4±1.87, n = 35).

Osteology: presacral vertebrae 29 (n = 37)–30 (n = 2); postsacral vertebrae 40 (specimen with most complete tail showing no evidence of regeneration); phalangeal formula for manus and pes 2.3.4.5.3 and 2.3.4.5.4 respectively.

Colouration: There is marked sexual dimorphism between adult males and females, with subadults and juveniles similar in colouration to adult females.

Adult males ( Fig. 4 ): dorsal and lateral surface tan-light brown or mid brown, with a few scattered dark flecks on the anterior and upper lateral surface, and a continuous, narrow, dark stripe across the snout in some specimens. In life with a bold russet flush to the lateral, dorsal, and ventral areas at the base of the tail and hindlimbs, and a dull to moderate orange to yellow flush to the posterior half of the abdomen. Throat and chest pale and without an obvious pattern of dark transverse markings on the outer chinshields and throat scales, and a variably defined, narrow, dark, longitudinal streak on the throat and chest.

Adult females ( Fig. 4 ): body usually two-toned. Dorsal surface of the body a dull light-brown or mid-brown. Lateral surface light-mid grey with scattered dark flecking overall, boldly marked individuals with these darker markings covering most of the lateral surface and contrasting markedly with the paler dorsal colour. Dark markings on dorsal surface occasionally aligned in longitudinal rows in the paravertebral region, and as a concentration of longitudinal flecks along dorsolateral margin, tending to form a more continuous, narrow, dark stripe along the temporal margin and across the snout in some specimens. A narrow, white, hip stripe (dark-edged above) extends along basal portion of tail and over the hindlimbs, continuing as a pale, narrow upper edge (paler than hip stripe) to the dark dorsolateral markings, variably extending forward to near the temporal region. In life with a bold orange to russet flush to the lateral, dorsal, and ventral areas of the entire tail, remainder of underside of body and hindlimbs white.

Subadults—overall body colouration usually darker with dark lateral flecking extending over most of the lateral surface, otherwise subadult males have a colour pattern typical of adult females and juveniles including a weak to obvious pale hip stripe and a continuous concentration of dark flecks along the dorsolateral margin in the shoulder region.

Note—The features of colouration along the dorsolateral margin mentioned by Sadlier (1986) for the specimen from Ouaième on the northeast coast are similar to those seen in some of the more boldly marked female specimens from the northwest coast.

Reproduction: The series of specimens collected at the very end of the dry season in late November 1998 contained four gravid females which consistently showed an egg laying mode of reproduction with a clutch size of 2 shelled oviducal eggs, one in each oviduct. Three other adult females collected at this time had 2 enlarged yolked ovarian follicles (one in each oviduct), and one individual 3 (one in the left oviduct and 2 in the right).

Distribution and habitat. The species is known from two locations on the far northwest coast and two on the far northeast coast ( Fig. 2 ).

The two sites in the northwest of the main island, the Ouazangou-Taom Massif and Tsiba just north of Ouaco, are approximately 15 km apart. In this region the species was collected at both low (Tsiba 220 m; Mt Toam 275 m) and moderately high (Mt Toam 850 & 1000 m) elevation. At low elevations it was collected in remnant patches of low closed vegetation, usually located in steep gullies, on ultramafic ranges. They were observed active amongst leaf litter in sunlit patches. These patches were surrounded by extensive area of maquis shrubland, the heath-like vegetation endemic to this soil type. At higher elevations it was collected in maquis/forest at 850 m, and near the summit of Mt Toam in low moderately dense maquis (1000 m). It is worth noting that individuals collected at low elevation sites appeared to have low levels of heat tolerance in comparison to locally sympatric skink species (Caledoniscincus austrocaledonicus and Caledoniscincus haplorhinus) of similar size. The species occurrence in the more open maquis habitat at higher elevations is most likely due to the more moderate conditions present at higher elevations in temperature and humidity.

There is no precise information regarding the site of collection or habitat for the two specimens from Ouaième and Tao on the northeast coast of the main island.Opportunistic field research by the authors in this region over a number of years has not located further specimens from either area or any surrounding areas searched. However, the habitat searched in this region has nearly always been closed humid forest, and it is possible the species is restricted to a different habitat type which has not yet been sampled, perhaps similar to that occupied on the northwest coast.

The species occurrence on both the northwest and northeast coast could suggest a wider distribution in the north of the island. However, extensive field research in recent times covering a wide range of habitats in the northwest from Sommet Poum south to Massif du Boulinda (Whitaker Consultants, in prep.) has failed to locate it at any site other than Mt Taom and nearby Tsiba.

Conservation status

Celatiscincus similis satisfies the criteria for inclusion into one of the threatened species categories under a modified IUCN classification system (see Sadlier & Bauer, 2003). It is currently known from only four locations, and from what is known of the two populations on the northwest coast the species is at risk from a range of threats.

The preferred habitat of C. similis on the northwest coast is severely fragmented, and the size of the patches of remnant closed vegetation on the ultramafic massifs is very small, particularly at lower elevations—the extent of suitable habitat at Tsiba ( Fig. 5 ) is estimated at about 1 hectare. These isolated sub-populations are undergoing a decline in extent of occurrence as a result of the impact of burning of adjacent maquis shrubland, and from activities associated with mining ( Fig. 6 ). The extent of the species’ overall distribution on the western side of the island may ultimately prove to be much broader (i.e. extend further south and north in suitable habitat on ultramafic soils), but its area of occupancy (by virtue of the patchy and relictual nature of its preferred habitat) will most likely remain small.

Due to the fragmented and relictual nature of its distribution, the threats posed by the presence or potential for mining (some sites), and the impact of frequent firing, C. similis could be conservatively ranked as Vulnerable under the modified IUCN classification system (Sadlier & Bauer, 2003) used for the New Caledonian lizard fauna (area of occupancy <500 km 2, severely fragmented i.e. isolated sub-populations with a reduced probability of colonization, if once extinct; number of locations <5; continuing decline in area and extent of occurrence as indicated by a continuing decline in quality of habitat at some sites). It could be placed at a higher level of threat if continued declines in the area, extent and/or quality of habitat in the species very narrow range persist. Targeted field research is required, particularly in the northeast coast of the main island to fully assess the species’ distribution and habitat preferences in the region. Assessing the species status on the northeast coast of the main island is problematic due to a lack of precise information on the species preferred habitat in this region. It is probably similarly restricted in its habitat preferences, even though ultramafic soils do not occur in this region (Paris, 1981; Sautter, 1981).

Celatiscincus euryotis is known only from two sites on the Isle of Pines, one a remnant of closed forest near Waa Mé Bay on the west side of the island, and the second at Vao. The main island of the Ile des Pins has a total area of 152 km 2, and the likely area of suitable habitat (excluding maquis shrubland and introduced plantations) is considerably less than this. It had previously been regarded as Secure by Bauer & Sadlier (2000) but given its extremely limited distribution it has been reassessed and conservatively ranked as Vulnerable (Sadlier & Bauer, 2003). The species distribution on the island and the threats likely to impact on the species need to be established to fully evaluate its conservation status.

Relationships

Our interpretations as to the generic distinctiveness of Celatiscincus and the specific distinction between C. euryotis and C. similis are based on a combination of evidence from the morphological information and DNA sequence data.

Intergeneric Relationships: Both ND2 and Rag-1 support the monophyly of the two species of Celatiscincus. Mitochondrial ND2 haplotypes of C. euryotis from two individuals from the Isle of Pines and five individuals (from two localities) of C. similis show a high level of sequence divergence between populations (12.5%), and no or very low levels of divergence within each population (southern: 0, northern: 0.16%). This level of genetic differentiation suggests the two widely allopatric populations represent two distinct evolutionary lineages, and is comparable with between-species divergence in other genera (e.g., Marmorosphax 12.1%, Caledoniscincus 12.9%). No single scalation character will allow unequivocal assignment to one species or the other. However, significant differences were detected between the two species which showed a trend towards C. similis having: fewer midbody scale rows (¯x 30.9 vs 33.1, t 67 = 7.92, P<0.01); more scales on top of fourth finger (¯x 9.5 vs 8.9, t 64 = -4.58, P<0.01); more lamellae beneath fourth finger (¯x 17.2 vs 16.2, t 63 = -4.60, P<0.01); more scales on top of fourth toe (¯x 13.3 vs 12.75, t 59 = -3.60, P<0.01); and more lamellae beneath fourth toe (¯x 35.4 vs 32.2, t 58 = -7.10, P<0.01)—see Table 1 . Further, there were significant differences in the modal number of presacral vertebrae between the two species, but again these were not unequivocal.

We recognize that an arbitrary degree of genetic differentiation alone is an inappropriate measure of specific distinctness (Ferguson, 2002; Sites & Marshall, 2003), however, we here use such measures in support of diagnostic features of colour and osteology outlined in the diagnoses. While colour may be regarded by some as too highly labile to serve alone as an adequate diagnostic feature, the relatively large samples of C. euryotis and C. similis demonstrate the consistency of this feature. Minor differences in colour, scalation, and proportion are also diagnostic of some other New Caledonian skinks that show similar levels of genetic differentiation between species (i.e. certain Caledoniscincus and Marmorosphax), including instances of broad sympatry between sister or “sibling” species (Bauer & Sadlier, 2000).

Based on both intensive and extensive sampling by the authors over a period of more than 20 years, the large geographic gap separating the two species of Celatiscincus from one another appears to be real.This disjunction is certainly great enough to completely isolate the populations from one another.As such, the two species, diagnostic differences aside, are certainly evolving as separate lineages on independent trajectories and would in any case meet the criteria for phylogenetic species (Frost et al., 1992; Grismer, 1999; McKitrick & Zink, 1988). This disjunction remains enigmatic. No other sister taxa pair of New Caledonian reptiles exhibits this particular distribution pattern, or such a degree of geographic separation. It is difficult to surmise what historical factors may have contributed to such a pattern. The geological and ecological dissimilarity of the areas where the two species occur argues against their ever having shared a common continuous range, and their apparent intolerance of heat and exposure suggest that they are not likely to have achieved their current distribution through dispersal. It is perhaps most likely that Celatiscincus was once a more speciose and widely distributed genus and that differential extinction of geographically intervening species may have resulted in the pattern observed today.

Intrageneric Relationships: A phylogenetic analysis of all New Caledonian skink taxa using the mitochondrial gene ND2 (514 bp) and the nuclear gene Rag-1 (830 bp) suggests that most described New Caledonian endemic lygosomine genera are strongly supported monophyletic lineages, but reveals little regarding intergeneric relationships (Smith et al., in prep.). Celatiscincus is one of many generic level lineages with long

.

independent histories that form a basal polytomy within Australo-Pacific Eugongylus group skinks. There is no evidence to suggest that Celatiscincus is more closely related to Marmorosphax or Sigaloseps than to other New Caledonian skink genera.As such, we regard the genetic data as supporting our morphological interpretation of the distinctiveness of Celatiscincus. Although we remain ignorant of the position of Celatiscincus within the New Caledonian Eugongylus group, we believe that the erection of a new genus serves an important purpose: to highlight Celatiscincus as a distinctive, monophyletic lineage that must be considered separately from all others in future phylogenetic analyses.

A revised diagnosis for Marmorosphax Sadlier

The transfer of Lygosoma euryotis Werner to the newly erected genus Celatiscincus Sadlier, Smith and Bauer necessitates a redefinition of Marmorosphax Sadlier. Also the recent acquisition of a gravid specimen of Marmorosphax montana that contains developing embryos allows a live bearing mode of reproduction to tentatively be added to the suite of apomorphic characters diagnosing the genus.

The revised diagnosis below for Marmorosphax provides a suite of apomorphic character states to diagnose this genus from all other genera in the Eugongylus group of skinks.

Figures

Fig. 2. Distribution of Celatiscincus euryotis and Celatiscincus similis n.sp. in New Caledonia.

Fig. 3. Dorsal (upper), lateral (middle), and ventral (lower) views of the headshields of the holotype of Celatiscincus similis n.sp. (MNHN 2003.1006).

Fig. 4. Celatiscincus similis n.sp., an adult male (upper—AMS R153524) and adult female (lower—AMS R153525) from Tsiba, northwest New Caledonia.

Fig. 5. Tsiba near Ouaco, the small population of Celatiscincus similis n.sp. at this site was located in low closed vegetation in the centre of the photo

Fig. 6. Ultramafic ranges south of Pouembout in northwestern New Caledonia, showing impact of strip mining on the environment.