(cicak batik; batik bent-toed gecko) (Figs. 1 A, B; 2; 3 A; 4 A)
Holotype. ITB. DTI 2805, an adult female with original tail from Longkoga Stream, Bualemo, Mount Tompotika, Balantak Mountains (between 00° 40 '05.1"– 00° 40 ' 12.7 "S; 123 °06' 41.7 "– 123 °06' 39.2 "E; alt: 951–1002 m asl), Desa Trans Tanah Merah, Kecamatan Bualemo, Kabupaten Banggai, Propinsi Sulawesi Tengah, Sulawesi Island, Indonesia collected by A. Rachmansah and Umilaela on 20 May 2009.
Paratypes. DTI 2784, DTI 2801, DTI 2804, DTI 2785, DTI 2803, DTI 2802, same data as for the holotype, collected 19–20 May 2009.
Diagnosis. A large form of Cyrtodactylus with SVL reaching 115 mm in adult females, males slightly smaller, up to 110 mm, tail 108–120 % of SVL; body robust, limbs medium length; digits long; single pair of postmentals in contact posteriorly, isolating triangular mental from chin shields; dorsum with 23–26 transverse rows of slightly keeled trihedral tubercles, slightly larger than adjacent dorsal scales giving a generally smooth appearance; 48–57 smooth, round, juxtaposed ventral scales between distinct ventrolateral folds; no precloacal groove, no precloacal or femoral pores, no enlarged femoral scales; distinctly enlarged precloacal scale patch; 7–10 transversely expanded lamellae proximal to basal inflection of 4 th toe, 10–16 narrow lamellae distal to inflection. Underside of the hemipenal bulge of tail base bearing approximately 30 rows of small postcloacal scales, followed by approximately five rows of slightly enlarged, rectangular subcaudals followed by transversely expanded subcaudals.
Etymology. The specific epithet is used as a noun in apposition, originating from the specific Indonesian pattern of traditional “ batik ” cloth that is especially well known on Java. The dorsal pattern of the new species is similar to that of traditional batik cloth.
Holotype description. An adult female, SVL 103.2 mm (104.8 mm measured prior to fixation after being euthanized), TL 115.1 mm (fresh 118.2 mm, TL/SVL ratio 1.08–1.20). Head moderately long (HL/SVL ratio 0.29), wide (HW/HL ratio 0.60), moderately depressed (HH/HL ratio 0.34), distinct from neck. A raised, rounded supraorbital ridge continuous with canthus rostralis. Distinct frontoparietal depressions posterior to each supraorbital prominence. Lores weakly convex anteriorly, mildly depressed posteriorly; separated from anterior palpebrals and orbits by deep lacrimal grooves. Dorsal surface of snout anteriorly swollen above nostrils. Lacrimal groove met orthogonally by midpalpebral depression continuing around orbit. Small, raised extension of skin comprised of superciliaries and distalmost rows of palpebrals extending to the circumference of the orbit. Supraorbital scales uniform, lacking small tubercles. Superciliaries large, composed of two rows of overlapping scales. These scales are relative long, uniform, height ranged from 0.9 mm at the sides to 1.2 mm above the center of eye; forming a crenulated, erect rim around eye, rounded and smooth, without distinct keel, 16–21 rows of interorbital tubercles across narrowest point of frontal bone. Snout relatively short (SL/HL ratio 0.40); longer than eye diameter (OD/SL ratio 0.59). Scales of snout round, granular, uniform in size, rostrum scales smooth, without tubercles; tiny tubercles present on medial palpebral and interorbital regions. Head tubercles pointed, symmetrical, larger posteriorly, attaining maximum size at occiput. Scales of rostrum granular, regular in size (as in snout region). Eyes moderate (OD/HL ratio 0.23). Auricular openings erect, egg shaped, narrower dorsally and rounded ventrally; openings large (EaL/HL ratio 0.14); eye to ear distance slightly greater than diameter of eye (EaEy/OD ratio 1.1). Rostral 50 % deeper (2.1 mm) than wide (4.2 mm) at narrowest point, 69 % as deep (2.9 mm) as wide (4.6 mm) at longest point; incompletely divided by dorsal Y– shaped rostral groove; two supranasals, anteriormost pair separated by pair of median postinternasals and two internasal scales; rostral in contact with first supralabial, anteriormost supranasal, median postinternasal and internasal. Nares oval, oriented laterally, in contact with two supranasals, first supralabial; 3 / 3 postnasals, about the same size of head or body granular scales; Mental triangular in shape, nearly twice as wide as deep; single pair of postmentals contacting posteriorly about half their length. Postmentals bordered laterally by first infralabials, the anterior tips extend to the suture between mental and first sublabials; posteriorly by two slightly enlarged chin shields (adjacent to second infralabial), 2 intermediate sized and seven small irregularly shaped, granular gular scales. Throat covered with uniform granular scales. Nine supralabials to midpoint of orbit (13 to angle of jaw). Eight infralabials to midpoint of orbit (12 on the left and 11 on the right to angle of jaw). Body slender (TrL/SVL ratio 0.49); indistinct ventrolateral folds with about a dozen rounded, white tubercles. Dorsum between forelimb insertion and caudal region characterized by small granular scales interspersed with irregularly spaced, keeled unicarinate tubercles. Anterior side of tubercles rounded, rising gradually, defined by a single pronounced median, rounded keel; posterior sides convex, steeply sloping. From forelimb insertion to frontal region, tubercles progressively decrease in size, becoming round, non-keeled on head. A total of 36 paravertebral tubercles between forelimb and hind limb insertions are present. Ventral scales smooth, juxtaposed, regular in size; ventrals larger in diameter than both dorsal granular scales and some dorsal tubercles; gulars small, granular, uniform; 49 scales between dorsolateral folds across midsection of body. No precloacal pores or precloacal groove. No femoral pores or enlarged femoral scales, but enlarged precloacal scales present. Scales adjacent the precloacal area and on the lower parts of the femur become abruptly smaller compared to those at the precloacal region. Limbs medium in length, moderately robust; forearms shorter than hind limbs; forearm short (FaL/SVL ratio 0.17); tibia short (CrL/ SVL ratio 0.18); suprabrachials and prebrachials larger than scales of adjacent dorsum, granular, tuberculate, much smaller in postbrachial region proximal to elbow; infrabrachials small, granular, about equal to gulars; postantebrachials granular at elbow; numerous tuberculation present on tibial surfaces, except infratibial surfaces; supraantetibials regular, granular proximally, becoming large, continuous, unchanged in size with supracarpals and supradigital lamellae; infraantebrachial squamation similar to supraantebrachials, though tubercles not present; suprafemorals and prefemorals similar to supratibials, scales regular, smooth, tubercles numerous, well differentiated compared to neighboring granules; infrafemorals small, similar to granules on adjacent body; no enlarged infrafemoral or postfemoral scale series, squamation similar to adjacent interfemorals; ventrals, supratibials and pretibials small, granular, as in ventral femoral regions; unlike supracarpal squamation, supratarsals consistently small, granular, met abruptly by enlarged supradigital lamellae; infratarsals similar in size proximally, but becoming enlarged, subimbricate distally. Digits long, strongly inflected at basal interphalangeal joints. Claws large (maximum length of 1.9 mm), surrounded by elongate, deeply notched distal-most subdigital lamella, one enlarged supradigital scale, and a smaller (about half size) lateral scale. Subdigital lamellae elongate, narrow distal to first interphalangeal joint. Subdigital lamellae proximal to first interphalangeal joint wider than long, swollen, padlike, especially at interphalangeal joint. Counts of subdigital lamellae on manus I: 15 / 15; II: 18 / 19; III: 20 / 20; IV: 21 / 22; V: 19 / 19, on pes I: 15 / 14; II: 19 / 19; III: 23 / 23; IV: 24 / 26; V: 23 / 24; webbing absent. Relative lengths of digits on manus: IV> III> II> I> V, on pes: IV> V> III> II> I. Tail relatively long, (TL/SVL ratio 1.11); portion of tail subrectangular at base with regularly–spaced strongly trihedral, keeled tubercles more sparsely distributed than tubercles of dorsum; three prominent, enlarged post–cloacal spurs on each side of vent; subcaudals arranged in several rows of small, narrow rectangular scales followed by enlarged median subcaudal plates variable in size, largest 3.9 mm wide, 1.2 mm long; most approximately half this size.
Holotype coloration. Overall dorsal appearance uniform velvety black, tubercles have the same color as background hence not visible on photographs except on the flanks. Four pairs of overlapping “ ><” shaped irregular yellow transverse bands between nape and base of tail and 10 similar marks on the tail, the first three “ >< ” shaped marks on the tail are more or less similar to the dorsal pattern, the remainder less distinct in form with yellow spots and crosses. The areas within the overlapping “ ><” shaped marks are lighter compared to the dorsum. A yellow line borders the posterior margin of the head. Limbs with irregular yellow bands or spots at various angles; distinct yellow bars at the finger-carpal joint; head coloration slightly lighter than dorsum, faintly marbled with yellow spots which are variable in size, a yellow line running along superciliaries to occiput enclosing parietal region of head and posterior part of canthus rostralis; upper labials scales generally lighter with some yellow spots along the upper border; rostral as dark as body coloration; nape dark; eyes essentially black, iris greenish metallic during daylight (see Fig 1 B); lateral surfaces similar to dorsum but with yellow tubercles, sparsely arranged on the flanks, strongly contrasted with velvety black base color; venter and undersides of limbs uniformly blackish, ventral scales with numerous fine purple flecks covering otherwise pale scales.
Variation in the paratypes. The head varies slightly in length (HL/SVL ratio 0.29–0.32), width (HW/HL ratio 0.59–0.65), and moderately depressed (HH/HL ratio 0.27–0.34); interorbital tubercles across narrowest point of frontal bone varies between 16–21 rows. Snout relatively short (SL/HL ratio 0.39–0.43); longer than eye diameter (OD/SL ratio 0.56–0.58). Eyes moderate (OD/HL ratio 0.22–0.25). Auricular openings is large (EaL/HL ratio 0.12–0.14); eye to ear distance slightly greater than diameter of eye (EaEy /OD ratio 1.1–1.3). Rostral varies from 43–53 % deeper (1.8–2.1 mm) than wide (4.2–4.6 mm) at narrowest point, 67–88 % as deep (2.9–3.7 mm) as wide (4.6–5.6 mm) at longest point. Anteriormost pair of SuN separated by a pair of median post InNs and two to three InNs. PM bordered laterally by first and/or second InLs, the anterior tips extend to the suture between mental and first InLs; posteriorly by two slightly enlarged chin shields (adjacent to second InL) and two to eight intermediate sized and up to seven small irregularly shaped, granular gular scales. Supralabials varies from nine or ten to midpoint of orbit (13–15 to angle of jaw). Body slender (TrL/SVL ratio 0.39–0.49). A total of 33–40 PVTs between forelimb and hind limb insertions are present. Ventrals ranges from 48–57 scales between dorsolateral folds across midsection of body. The limbs are medium in length, forearm short (FaL/SVL ratio 0.15–0.17); tibia short (CrL/ SVL ratio 0.17–0.19); Variations of subdigital lamellae on manus I: 13–16; II: 18–20; III: 20–23; IV: 21–23; V: 18- 22, on pes I: 14–16; II: 19–21; III: 23–25; IV: 24– 27; V: 20–24; webbing absent. Tail relatively long, (tail length/ SVL ratio 1.08–1.20). For other detailed measurements and detailed count of the whole type series see Table 1.
Regenerated tail completely round; caudals of re–grown tail extremely reduced, tubercles absent; subcaudals of regenerated part of tail in proportion to original subcaudals, but reduced in size and shorter. Scales of regrown tail similar in shape, lightly colored, mottled, lacking tubercles and the “ >< ” shaped marks (see Fig 1 B).
Secondary sex characteristics. From specimens on hand, the male attains a size nearly as large as females, and can be distinguished by the presence of bulging hemipenes at the base of the tail, whereas females have a slen- der tail base. Both sexes lack precloacal and femoral pores, and hence these are not useful for determining sex of this species.
Ecological notes. All specimens were found on vegetation in undisturbed primary forest, from 1.5 to 3 meters above ground and more than 50 meters from the closest stream (Longkoga) in primary forest. The trees format the collecting sites were lianas and small trees with trunks less than 40 cm in diameter. The smaller species, C. jellesmae was not found in the same site, but at other camp sites it occurred approximately 1–2 meters above the ground on tree trunks and smaller vegetation.
Comparison between species. Cyrtodactylus batik and C. wallacei have a similar coloration pattern (Fig 2 A, B), but the basic coloration of C. batik is black with yellow (cream) transverse stripes, sharply demarcated with tubercles having the same coloration as the background except on the flanks. In C. wallacei, the dorsal ground coloration is reddish brown with weakly defined lighter stripes made up by light colored tubercles on the dark background. The scales and tubercles of C. batik are smaller compared to those of C. wallacei. Consequently, every detail in number of scales, head scales, tubercles as well as number of lamellae under finger and toes are higher in C. batik, except for the number of first and second toe lamellae, which are higher in C. wallacei (see Table 2). Despite similarity in size, the banding pattern, velvety black dorsum, differences in number of tubercles, and number of ventral scales distinguish it from the very similar and geographically close C. wallacei. The smaller size and light brown dorsal coloration with sharply defined blotching pattern distinguish C. jellesmae from C. batik. These three Sulawesi species are similar in lacking precloacal and femoral pores. Based on examination of gravid females, C. jellesmae is a species complex with at least two different body size groups (Table 2). The smaller form is widely distributed, has fragmented subcaudals and shares the presence of tubercles along the ventrolateral fold with C. spinosus and C. batik. The larger form is restricted to southeast Sulawesi and is characterized by the absence of enlarged tubercles along the ventrolateral fold, but has broad transverse subcaudals as in C. batik. Otherwise both forms currently identified as C. jellesmae have similar dorsal coloration (aside for other small forms with different dorsal blotching pattern and not included in the analysis).
Cyrtodactylus batik is distinguishable from all but eight congeneric species by the absence of a precloacal groove, and precloacal and femoral pores. The first two (three if the large form identified as C. jellesmae is counted as a separate species) occur in Sulawesi and have been discussed previously. The six remaining congeners may be distinguished from C. batik by the following characteristics: tiny dorsal conical tubercles, a black dorsal coloration and large size (maximum SVL 113 mm) distinguish C. batik from C. laevigatus (Darevsky, 1964; presence of an enlarged femoral scale row forming a distinct boundary and separating smaller posterior femorals distinguish C. paradoxus (Darevsky & Szczerbak, 1997); presence of enlarged femoral scales, smaller SVL and dark bands or blotches contrasting with a light gray background distinguish C. semenanjungensis (Grismer & Leong, 2005); presence of a quadrangular rostral bordered by fewer scales and with a single median cleft, smaller size, and heavy yellow spotting on arms, dorsum and, most distinctively, along ventrolateral folds and labial regions distinguish C. sermowaiensis (de Rooij, 1915); presence of an extremely enlarged femoral and precloacal scale series, slightly smaller size and distinct reddish-orange and black bandings distinguish C. thirakhupti (Pauwels et al.). Cyrtodactylus malayanus (de Rooij, 1915) and C. consobrinus (Peters, 1871) are two species from Borneo, about equal in size, that differ dramatically from the Sulawesi forms by their banded color pattern and in having enlarged femoral scales. Precloacal and femoral pores may be present in some individuals, but these, in combination with different dorsal coloration, set them apart from all Sulawesi forms. Cyrtodactylus spinosus and Cyrtodactylus sp. from Sulawesi Barat (see Appendix 1) both have precloacal pores, while C. fumosus has both precloacal and femoral pores in a continuous series. The lack of spines along the ventrolateral flanks and head, as well as size and blotching pattern distinguish C. batik from C. spinosus.
Sulawesi. *) Data on C. wallacei and C. spinosus were extracted from the original description. n.a = data not available.
In the Lesser Sunda Islands three other species have been described: C. darmandvillei (Weber, 1890), C. gordongekkoi (Das, 1993) and C. wetariensis (Dunn, 1927). Each has precloacal and femoral pores, hence they are easily distinguished from C. batik. From Maluku, three species have been recognized, C. halmahericus (Mertens, 1929) from Halmahera and Seram; C. deveti (Brongersma, 1948) from Morotai Island and C. nuaulu (Oliver et al., 2009) from Seram Island. As all three species have precloacal and femoral pores, they are easily distinguished from C. batik. Regarding coloration differences, C. deveti has bold bars on the dorsum and C. nuaulu is ornamented with elongated bands along paravertebral area, hence these are easily distinguishable from C. batik. Cyrtodactylus halmahericus has a number of poorly defined bands and hence is also easily distinguishable from C. batik.
It is interesting that among the eight poreless species mentioned above, four occur in the Wallacean region, three are present in Southeast Asia and one in New Guinea. The absence of precloacal pores, femoral pores and an enlarged femoral scale series in C. batik, C. wallacei, C. jellesmae complex, C. laevigatus and C. sermowaiensis suggests these species may be closely related. Three species are Sulawesi endemics. In contrast, C. laevigatus is endemic to Lesser Sunda Islands while C. sermowaiensis is restricted to the north coast of New Guinea. Nothing is known about the evolutionary process towards pore loss or other frequently-used diagnostic morphological characters in Cyrtodactylus. Change in body size and the wide transverse subcaudal scales are plausible paths of evolutionary change. Several other species without precloacal and femoral pores have been found in Sumatra, but these forms comprise individuals with completely different coloration (Iskandar et al., in prep.). We suspect that the Sulawesi species can be distinguished in a phylogenetic framework, as half of the Sulawesi species lack pores at the precloacal and femoral region. We have no basis to suggest that the Papuan C. sermowaiensis and C. laevigatus from the Lesser Sunda islands represent a single phylogenetic entity together with those from Sulawesi as both have a completely different coloration compared to the Sulawesi species and are of the same general small size and coloration as the Sunda shelf forms. We therefore hypothesize that Cyrtodactylus batik, C. wallacei and C. jellesmae complex form a monophyletic group marked by evolutionary shifts in body size and modification of subcaudal scales. Although C. jellesmae has distinct body coloration; its coloration pattern reflects the same basic structure as those of C. wallacei and C. batik. The light dorsal patterns are much wider and with poorly defined margins in C. jellesmae as compared to C. batik, in which they are very narrow and straightly demarcated (Fig. 4). The body coloration pattern of C. wallacei can be considered the most extreme case. It is very similar to that of C. batik, but the light parts are chiefly confined around enlarged dorsal tubercles (Fig, 3).
Remarks. Sulawesi has been isolated for at least 25 million yr and has also undergone a complex suite of geological processes which occurred throughout the Tertiary (Hall 1996, 1998, 2001). Paleogeographic studies indicate that Sulawesi’s current topography is the result of a complex suite of geological processes that resulted in at least seven oceanic islands that were initially isolated but later joined together as a single more or less continuous tract of land (Moss & Wilson 1998; Hall 2002, 2008). The herpetofauna of Sulawesi is depauperate in terms of number genera and highly endemic in species compared to the adjacent regions of Sundaland (Whitten & Whitten 1992; Whitten et al. 2000; How & Kitchener 1997; Iskandar & Tjan 1996; Iskandar & Colijn 2000, 2001; Inger & Voris 2001; McGuire et al. 2007). Similar levels of faunal impoverishment and high species endemism are documented for mammals (Musser 1987; Shekelle et al. 2008) and invertebrates (Gressit 1961; Vane-Wright 1991; Whitten et al. 1987), indicating the pattern observed in reptiles and amphibians is likely real and not merely an artifact of the limited attention biologists have paid to Sulawesi’s herpetofauna in comparison to other regions in Southeast Asia. Wallacean and Australian Cyrtodactylus are likely of western origin as the bulk of the species diversity in the genus is Sundaic, with more than 90 % of described species occuring west of Wallace’s line. While hypotheses regarding the interspecific relationships of Sulawesi Cyrtodactylus are currently under study, the discovery of C. batik is biogeographically significant as it is the second large Cyrtodactylus to be found on Sulawesi after C. wallacei. The present study reveals that the large C. jellesmae could well be a new form distinguished by its size and by differences in the nature of subcaudal scales and other characters (Table 2). Judging from the general distribution pattern of most vertebrates and the high external similarities between C. batik and C. wallacei, we suspect that the Kabaena population might well represent a distinct form as it does not match the microcontinental geological history of Sulawesi (Hayden et al. 2008; Evans et al. 2003 a, b; McGuire et al. 2007). Disjunct distribution is also known in Eutropis grandis (Howard et al., 2007). Being practically isolated, Mount Tompotika is also the type locality of several endemic species such as Luperosaurus iskandari (Brown et al., 2000), an undescribed Occidozyga as well as an Oreophryne (Iskandar et al., in press.). In the neighboring area, several Calamaria have been described (Inger & Marx, 1965; Koch et al., 2009). The finding of this new species is thus expected and provides further evidence of the high conservation significance of the Balantak Mountains.