A new arboreal Lepidodactylus (Reptilia: Gekkonidae) from Espiritu Santo Island, Vanuatu: from egg to holotype

Zootaxa 1918: 26 38 (2008) www.mapress.com/zootaxa/ Copyright 2008 Magnolia Press ISSN 1175-5326 (print edition) ZOOTAXA ISSN 1175-5334 (online edition) A new arboreal Lepidodactylus (Reptilia: Gekkonidae) from Espiritu Santo Island, Vanuatu: from egg to holotype IVAN INEICH Muséum national d Histoire naturelle, Département Systématique & Evolution USM 602. Taxonomie & Collections (Reptiles) CP 30, 25 rue Cuvier, F-75005 Paris, France.E-mail ineich@mnhn.fr Abstract Lepidodactylus buleli is a new gecko from Espiritu Santo Island, central Vanuatu, Melanesia. This species most closely resembles L. pulcher from Admiralty Islands (Papua New Guinea) by its elevated number of midbody scale rows, and belongs to Brown and Parker s (1977) Group II based on having an entire terminal scansor on all digits followed by two or three divided or deeply notched subterminal scansors on digits II V. It differs from other Group II species in having the following combination of character states: a high number of scales around midbody, relatively weak dilation of digits, slight webbing of toes III and IV, cloacal spurs, and an original coloration. This species lives in myrmecophilous plants hanging high on trees in the deep primary forest on the west dry coast of Espiritu Santo Island. It also is the first lizard known from a holotype collected as an egg in the field, hatched in captivity, and grown until reaching adult size, allowing its description. Key words: Reptilia, Gekkonidae, Lepidodactylus buleli new species, Espiritu Santo, SANTO 2006 expedition; Vanuatu, Melanesia, myrmecophilous plants French abstract Lepidodactylus buleli est une nouvelle espèce de gecko de l île Espiritu Santo, Vanuatu central, Mélanésie. Elle se rapproche de L. pulcher des îles Amirautés (Papouasie Nouvelle-Guinée) par plusieurs caractères de son écaillure, notamment le nombre élevé de rangs d écailles autour du corps. Elle appartient au Groupe II défini par Brown et Parker (1977) car elle possède une lamelle digitale terminale entière à tous les doigts, suivie par deux ou trois lamelles subterminales divisées ou fortement encochées aux doigts II-V. Elle se différentie cependant des autres espèces du Groupe II par la combinaison des caractères suivants: nombre élevé d écailles autour du milieu du corps, faible dilatation des doigts et orteils, faible palmure des orteils III et IV, présence d éperons cloacaux, un dessin dorsal original à la base de la queue et une coloration jaune citron des labiales. Cette espèce vit dans ou à proximité de plantes épiphytes myrmécophiles suspendues en hauteur dans les arbres de la forêt primaire sur la côte occidentale sèche d Espiritu Santo. C est aussi le premier lézard décrit à partir d un holotype rapporté depuis le terrain sous la forme d un œuf, éclos en captivité par la suite, puis élevé jusqu à la taille adulte permettant la description de l espèce. Introduction Tropical Pacific island herpetofaunas are remarkable by some of their radiations, mostly of two genera: the skink genus Emoia Gray (see Brown 1991), and the gekkonid genus Lepidodactylus Fitzinger. The latter is a group of small geckos and, except for the widely ranging unisexual-bisexual Lepidodactylus lugubris complex (including at least L. lugubris (Duméril & Bibron, 1836), L. moestus (Peters, 1867), L. woodfordi Bou- 26 Accepted by S. Carranza: 1 Oct. 2008; published: 29 Oct. 2008

lenger, 1887b and an undescribed bisexual species of Eastern Polynesia (Ineich 1987, Ota et al. 1995, Radtkey et al. 1995, Ineich unpubl.), each species has a relatively small distribution limited to a few islands in Southeast Asia and Oceania (Brown & Parker 1977, Brown & Alcala 1978, Bauer & Henle 1994, Ota et al. 1995, Zug et al. 2003, Buden 2007). Brown and Parker (1977) and Brown and Alcala (1978) originally recognized 18 to 20 species in the genus, while Ota et al. (1995) later recognized 24 species. This number has grown to 32 recognized species today (Uetz & Hallermann 2008). Note, however, that L. pumilus Cope, 1869, was considered insertae sedis by Bauer and Henle (1994: 226) and its type locality, NW Australia, seems not in accordance with the present distribution of the other members of the genus. Species are arranged into three well diagnosed but informal phenetic groups (Groups I, II, and III), mostly based on their finger and toe lamellae morphology, toe webbing and tail shape. No study has tested the monophyly of these groups or, for that matter, the monophyly of the species currently assigned to Lepidodactylus [Kluge (1968) provided a set of diagnostic traits for the genus but unfortunately did not address monophyly]. Note, however, that intermediates between these three phenetic groups exist, probably as individual variants and not with such intermediate characters stabilised at specific level: e.g., intermediate between groups I and II, L. ranauensis Ota and Hikida, 1988 (see Ota & Hikida 1988) or intermediate between groups II and III, L. lugubris from French Polynesia (Ineich unpubl.). The most widespread species, L. lugubris, is in fact a unisexual-bisexual species complex with unclear taxonomy and several complex hybridisation events giving rise to numerous widespread and local diploid or triploid clones, and also to local sterile intersexual individuals (Ineich 1987, 1999, Saint Girons & Ineich 1992). It is classically considered as being commensal with humans, with a range extending far beyond the boundaries of the region, including the Indian Ocean to the west, and Central America to the east. Lepidodactylus lugubris was long considered as the only species of Group III (L. lugubris and relatives) outside the Philippines, but several additional species of that group have been described or revalidated recently (Ota 1987, Ota et al. 1995). One or more Group III species occurs on almost every island in Oceania and they are considered to be the most derived taxa of the genus. The other species, all bisexuals, are distributed in Southeast Asia and the tropical Pacific islands, with especially high species abundance in the Philippines (e.g., Ota & Crombie 1989), New Guinea and Melanesia in general (Brown & Parker 1977, Brown & Alcala 1978, Pernetta & Black 1983, Bauer & Henle 1994). Their current distribution likely derives from natural and certainly human-assisted dispersal for at least some species (Ineich 1999). The ten species in Group I (L. pumilus and relatives) and the eleven species in Group II (L. guppyi and relatives) exhibit restricted ranges, often to a single island or small group of islands. Since 1977, three more Group I species, L. browni Pernetta and Black, 1983 from Papua New Guinea, L. euaensis Gibbons and Brown, 1988 from Tonga and L. flaviocularis Brown, McCoy and Rodda, 1992 from the Solomon Islands, have been described. In the same way, since 1977, six additional Group II species have been revalidated or described: L. intermedius Darevsky, 1964 and L. lombocensis Mertens, 1929 from the Lesser Sunda Archipelago, Indonesia (see Ota et al. 2000); L. paurolepis Ota, Fisher, Ineich and Case, 1995 from the Ngerukewid Group of the Palau Islands; L. vanuatuensis Ota, Fisher, Ineich, Case, Radtkey and Zug, 1998 from several Vanuatu islands; L. tepukapili Zug, Watling, Alefaio, Alefaio and Ludescher, 2003 from Funafuti Atoll, Tuvalu; and L. oligoporus Buden, 2007 from Namoluk Atoll, Mortlock Islands, Chuuk State, Federated States of Micronesia. Dispersal in Lepidodactylus occurred probably in the distant past and well before human colonization of the tropical Pacific; several endemic species occur irregularly in the area. Sea level lowstands related to glaciation events and typhoons certainly facilitated dispersion, island colonization and also hybridisation through contact of separated taxa giving birth to unisexual lineages. Most species are morphologically well differentiated and largely forest residents, thereby suggesting long periods of isolation and relictual status. During and following the SANTO 2006 Expedition (see http://www.santo2006.org), I found two specimens of an undescribed Lepidodactylus from Espiritu Santo Island. One was accidentally cut in half with machete in the field by Bruno Corbara when examining interior of an epiphytic plant and preserved immedi- A NEW FOREST LEPIDODACTYLUS FROM VANUATU Zootaxa 1918 2008 Magnolia Press 27

ately by me. The second specimen was taken in an epiphytic plant as an egg and transported to Paris, hatched with the help of gecko breeders Francis Girard and Thomas Lefranc, grown to adult size, and long observed alive before preservation in 70% ethanol. Both specimens differ from all other Group II members by a combination of colouration, morphological and meristic characters and are herein described as a new species. Material and methods Measurements were taken to the nearest 0.01 mm with dial calipers. Each specimen was sexed on the basis of the presence (male) or absence (female) of preanal and femoral pores, and developed cloacal spurs. The specimen grown in captivity was considered to be an adult male because he had distinct swelling in the cloacal region and developed cloacal spurs, whereas the machete-cut specimen was an adult female having two wellformed eggs visible in its accidentally opened abdominal cavity. Both Vanuatu specimens possess the characteristics of Brown and Parker s (1977) Group II (i.e. undivided terminal scansors on all digits; deep notch or division in two or three subterminal scansors on digits II-V; nondepression of tail), but were clearly distinct from the recently described Group II species, Lepidodactylus vanuatuensis from Vanuatu (Ota et al. 1998). I therefore compare them with all known Group II species (Brown & Parker 1977, Ota et al. 1995, 1998, Zug et al. 2003, Buden 2007). While I derived most data from the literature (Ota et al. 1998, Zug et al. 2003, Buden 2007), I examined comparative specimens of L. vanuatuensis from MNHN collections and additional collections from the Santo 2006 Expedition. It was not necessary to examine type specimens (particularly those of L. guppyi and L. pulcher) since differences between the two Santo specimens and other known species were evident from published descriptions. Terminology for descriptions of mensural and scalation characters follows that of Ota et al. (1995, 1998) and Zug et al. (2003), except that several counts were made for particular scale characters like midbody scale rows number. Morphometric characters (in mm) considered are: SVL: snout-vent length; HL: head length; HW: head width (distance between angles of gape; this distance avoids possible error introduced by measurement of the broadening of the head caused by the jaw musculature); HH: head height (dorsoventral distance from the top of head to the underside of the jaw at the transverse plane intersecting the angle of jaws); NBB: scales bordering nostril; SEL: snout-eye length; NEL: naris-eye length; OD: orbit diameter (eye diameter or length of other authors, although they measure anteroposterior axis length of orbit); EEL: eye-ear length; ERL: ear opening length; IND: internasal distance; IOD: interorbital distance (transverse distance between the anterodorsal corners of left and right orbits); SW: snout width (internasal distance of other authors); SFL: snout-forelimb length; TL: trunk length (axilla-groin length; distance between the posterior edge of the forelimb insertion (axilla) to the anterior edge of the hindlimb insertion (inguen)); TIVW: toe IV width; TIVL: toe IV length; TL: tail length; TW: tail width; TD: tail depth. Meristic characters considered are: CNS: circumnasal scales (number of scales abutting naris, exclusive of rostral and first infralabial; CS: chin (Zug et al. 2003; secondary postmentals) scales (number of scales transected by straight line from left to right 3 rd -4 th infralabial sutures); PMP: postmental ( PosMen of Zug et al. 2003; primary postmentals) scales: number of scales touching mental and infralabials from left to right 3 rd -4 th infralabial sutures; ROS rostral size: width to height proportion in 0.5 intervals (1, 1.5, 2, ); ROC: rostral cleft (crease): absence or presence of midline cleft or crease; SNS: snout scales: number of scales between left and right nares and touching rostral; UL: upper labials; LL: lower labials; MEN mental size (width to height proportion scored as for ROS); NILC: naris-infralabial contact (naris abuts or separated from first infralabial); NRC: naris-rostral contact (naris abuts or separated from rostral by scale); IOS: scales between midpoints of upper margins of orbits; MSR: rows of scales around midbody; FIS: subdigital scansors on left-right finger I (lamella is wider than deep and contacts the marginal scales; fragmented proximal scales are excluded; FIIS: subdigital scansors on left-right finger II; FIIIS: subdigital scansors on left-right finger III; FIVS: subdigital scansors on left-right finger IV; FVS: sub- 28 Zootaxa 1918 2008 Magnolia Press INEICH

digital scansors on left-right finger V; TIS: subdigital scansors on left-right toe I; TIIS: subdigital scansors on left-right toe II; TIIIS: subdigital scansors on left-right toe III; TIVS: subdigital scansors on left-right toe IV; TVS: subdigital scansors on left-right toe V; SCS: subcaudal scales: size of the median subcaudal scales relative to the dorsal caudal scales (score as for ROS); CL: crush length (length of tibia from knee to heel); SCBS: subcaudal scales size relative to belly scales; CS: cloacal spurs: scored as absent, single, or double; WEB: webbing: relative amount of webbing, scored as fours states: 0, none between the 2 nd and 3 rd digit of hindfoot; 1, slight, basal ¼ of 2 nd digit s length; 2, moderate, 1 / 4 to 1 / 3 ; 3, strong, more than 1 / 3 ; STR: number of scales touching rostral between right and left nares. Spellings for island names are those of Motteler (1986). Sample sizes of adult specimens for most Group II species are statistically inadequate [L. shebae (Brown & Tanner, 1949) is known only by its holotype], thus comparison data are only indicative. Many voucher specimens were not carefully prepared, thus resulting in bent or folded specimens or parts thereof. Thus significant variation often results from poor preparation or difficulties in checking some characters or scale counts. To increase measure precision, I sometimes used the mean of several different counts, e.g., for MSR. Results Lepidodactylus buleli, new species Figures 1 5 Holotype. MNHN 2008.0004, an adult male from the forest about two walking hours east of Penaoru Village(14.95894 S, 166.63290 E), collected as an egg on 14 November 2006 by Ivan Ineich. It was euthanized with sodium pentobarbitol injection in March 2008 after numerous photographs were taken. A sample of tissue was preserved from the ventral side of its right thigh. Paratypes.- None. Etymology. The specific epithet buleli is given by the author as a reference to a personal and private story and has no particular signification related to the species, its characteristics, geographical origin or biology. It is a noun in apposition and not formed from the masculine personal name Bulel in the genitive singular case. Diagnosis. An intermediated-sized Lepidodactylus, body slightly depressed dorso-ventrally, limbs overlapping when adpressed. Two or three deep notched or divided subterminal scansors of second through fifth digits of the fore- and hindfeet (thus belonging to Group II of Brown and Parker 1977). It differs from other Group II members by the following combination of characters: moderately dilated digital pads of pes and manus, lamellae covering most of elongate toe length, presence of one enlarged cloacal spur followed by 2 3 smaller spurs on each side (fig. 5), a short and wide snout (fig. 1), an elevated number of midbody scale rows, uniform lemon-yellow upper- and lower labials and a yellowish head area mixed with purplish dark brown particularly between eye and in the tympanic region (fig. 4), whitish chin and throat (fig. 3), and a distinctive anchor-like yellowish to cream dorsal pattern at the base of tail (fig. 5). Description of holotype (measurements in mm, after preservation). Snout-vent length 37.5; head length 9.06; head width 6.28; head height 4.68; jaws length from posterior border to snout end left/right 6.44/6.75; snout-eye length 4.33; naris-eye length 3.07; naris circular about 0.48 0.40; orbit diameter 2.26; eye-ear length 3.43; snout width (internasal distance) 1.82; ear opening length width 1.22 0.80; interorbital width 3.80; snout-forelimb length 14.44; trunk length 17.37; length of hind limb 16 (62.8 % of axilla-groin distance); length of fore limb 11 ; crus length 4.69; tail length 40 (entire); tail width 3.59; tail depth 2.93. Snout rounded (fig. 4), rather short and wide; snout scales (scales touching rostral between left and right nares) 5; supranasals separated by three scales in contact with rostral; rostral entering nares, broader than high, 2.03 0.92 (width about 2.2 times height); no rostral cleft; nares bordered by five scales: one postnasal, A NEW FOREST LEPIDODACTYLUS FROM VANUATU Zootaxa 1918 2008 Magnolia Press 29

and two supranasals (in total three nasals), one rostral, and first supralabial; five scales touching rostral between left and right nares; 39 40 interorbital scales (several counts); 11 left and 10 right supralabials, 8th below centre of eye; 10 left and 10 right infralabials; mental scale distinct, triangular, its anterior width nearly equals to midline length (0.91 1.08); 17 chin scales and 10 postmentals. Mental followed posteriorly by a pair of enlarged plates also in contact with infralabial I on each side, themselves followed laterally on each side by another enlarged plate, slightly shorter, in contact with infralabial I or its suture with infralabial II, and with infralabial II. FIGURE 1. Lepidodactylus buleli sp. nov., general view of the alive holotype (MNHN 2008.0004). FIGURE 2. Lepidodactylus buleli sp. nov., dorsal view of the preserved holotype (MNHN 2008.0004). Body slightly depressed; 147 rows of scales (3 counts: 144, 146, 152) around midbody, grading into granular scales on lower lateral surfaces; dorsal and lateral scales granular, without enlarged tubercles, and in juxtaposition; ventral scales almost flat, cycloid, 2 3 times larger than dorsal scales; limbs well developed (figs. 1 3); digital lamellae ventrally covering nearly all of forefoot digits and digits I II of hindfoot, about ¾ of 30 Zootaxa 1918 2008 Magnolia Press INEICH

digits III V of hindfoot; all digits of fore- and hindfoot clawed except the first; ultimate claw-bearing and penultimate phalanges of fore and hind digits raised above pad and extending over distal edge of pad; foreand hindfoot webbing slight (less than basal 20% of 2 nd digit s length); breadth of fourth toe about 28 31 % of its length; breadth of fourth finger 31 32 % of its length; left/right scansors covering underside of toes to base, 8/8 under first toe, 10/10 under second toe, 13/12 under third toe, 14/13 under fourth toe, 10/10 under fifth toe; left/right scansors covering underside of fingers to base, 8/8 under first finger, 10/11 under second finger, 10/11 under third finger, 14/14 under fourth finger, 6/5 under fifth finger; fourth toe lamellae: one entire, two divided, ten entire and 1 2 enlarged but not touching the toe border laterally; percentage of digit and toe covered by entire lamellae (right side only) respectively 80 90 % and 75 80 %; precloacal and femoral pore rows absent, as are enlarged adjacent scales. FIGURE 3. Lepidodactylus buleli sp. nov., ventral view of the preserved holotype (MNHN 2008.0004). Tail entire, subcylindrical throughout length, gradually tapering to a blunt tip; lateral margins without spines or skin flanges; scales on tail annulate, squarish or rectangular, larger ventrally than dorsally, subcaudal scales about 1.5 times belly scales; base of tail distinctly swollen by hemipenes; one large and about two smaller blunt cloacal spurs on each side. Color in life (based on laboratory observation and numerous color photographs of holotype taken at different times). Coloration is very variable depending on the physiological and psychological situation of the animal or its following response to light intensity. Dorsal ground color is variable from dark, purple brownish to pinkish, with distinct more or less circular vivid black spots in seven medio-dorsal paired rows from neck to tail base (fig. 1). The first two pairs of middorsal black spots are more widely separated between them than the five following pairs. The first pair, anteriorly, medium in relative size, is located just in front of anterior members. The second anterior pair is the largest, nearly circular, and located just behind anterior members. The third pair, like the four following pairs, is more centred against the mid-back, thus narrower. Pairs 4 and 5 are identical to pair 3. Pair 6 is located at the level of posterior limbs. The last (7th) pair is located just behind posterior members and has also a similar size than former pair, but is followed anteriorly by oblique darker lines giving the whole an open V-shape. That black pair is followed by a darker brownish square patch. Those seven pairs of dark black spots have also sometimes irregular lateral black smaller spots. If interpreted as the A NEW FOREST LEPIDODACTYLUS FROM VANUATU Zootaxa 1918 2008 Magnolia Press 31

W-chevron considered by Ineich (1988) for the Lepidodactylus lugubris unisexual-bisexual species complex s dorsal pattern, black spots pairs 1 and 2 are on lateral tops of the W, whereas pairs 3 to 7 are located on both basal points of the W. The oblique dark lines following black spots of pair 7 correspond to the W lateral lines running from bottom to top. FIGURE 4. Lepidodactylus buleli sp. nov., head of the alive holotype (MNHN 2008.0004) showing the star-shaped dark markings around eye. Anterior and posterior of pair 7 is a lighter shape representing a shape like an anchor (fig. 5). That shape was almost yellowish or cream white in live holotype, depending on the gecko s state. The anchor began posterior to black spots pair 6, at the level of the first light annuli on tail base and the round hook part of the anchor is located at the second light tail dorsal ring. That shape is clearly visible and may constitute a good diagnostic character for the species, in combination with other characters. Dorsum with 5 lighter more or less square-shaped purplish zones medially between pairs 2 and 7 of black spots. Flanks are always darker than dorsum, covered with darker marks and some whiter spots like ocelli. Each ocellus covers about 20 granules. In total, on the right side five to six ocelli are visible. Each dorsal black spots pair from 2 to 5 is laterally followed on flanks by a brownish rectangular dark brown mark. Venter is cream white with small light brown spots more developed on sides. Supra- and infralabials are brightly lemon-yellow and without any darker mark. Eye is not clearly encircled by a lighter coloration, but the whole peri-ocular area including labials is lemon-yellowish, as is the tympanum area, but that yellowish coloration is mostly developed in the posterior head part (fig. 4). Eye is surrounded by 5 dark reddish brown radiating bands (fig. 4). Eyes are brownish orange with reticulations. Upper head ground colour is sometimes dark brown, purplish like the back (when animal is not too darkened), or even more or less yellowish. A first median V shaped dark mark, pointing posterior, can be seen medially on the snout in front of eyes. Tail is annulated with cream and dark rings. Note however that annuli can be more or less evident according to the gecko owing to physiological and psychological state. There are 11 dark dorsal tail annulations and 12 lighter cream annulations posterior to tail base. Light rings are about 2/ 32 Zootaxa 1918 2008 Magnolia Press INEICH

3 to 3/4 as wide as dark ones. Ventrally tail is lightly orange coloured. Throat is whitish to yellowish. Mouth interior colour was not checked. Colour of holotype in preservative. Colour of the preserved holotype is very reduced and retains only a small fraction of coloration in life (fig. 2). On upper head the snout anterior opened V mark is still discernable. The interocular dark band also is reduced but still visible. Between eyes there is just a small black spot conserved and another one in the neck area. The dark band in front of the eyes is visible. Middorsal black spot pairs are variably visible. Pair 1 is weak, pair 2 very evident and still heavy black, pair 3 is visible with lateral dark expansions, the same for pair 4 with even larger lateral expansions, another large lateral dark expansion is visible between dorsal black pairs 4 and 5 and a last one dorso-ventrally at pair 5 level. Between black spot pairs 3 to 5 one supplementary black mark can be seen. Pair 6 has reduced lateral dark expansions and pair 7 has clearly line prolongations forming together an opened V shaped black mark pointing posteriorly at tail base. Dorsal coloration is whitish with some brownish patches. Tail is dorsally brownish and upper thighs white. Ventrally, tail is darker than venter but still not dark except its end (fig. 3). Anchor-shape mark is still discernable but no more as evident. Venter is white with fine speckles mostly on sides and more evident in the area between anterior and posterior members (fig. 3); each speckle covering 1 to 2 ventral scales. Throat is weakly speckled. MNHN 2006.0544 (female) also present typical disseminated minute speckles on venter and throat. FIGURE 5. Lepidodactylus buleli sp. nov., holotype (MNHN 2008.0004). The anchor-shaped typical marking at the base of tail is cream yellowish coloured. Cloacal spurs are clearly evident. A NEW FOREST LEPIDODACTYLUS FROM VANUATU Zootaxa 1918 2008 Magnolia Press 33

FIGURE 6. Inside of an opened myrmecophilous plant showing one of both adhesive eggs deposited there by a gravid female (Penaoru Forest, Espiritu Santo) in communal egg laying sites. Distribution. The species is known presently only by two specimens from its type locality, primary forest at about 630 m elevation, east of the Village of Penaoru, on the west dry coast of Espiritu Santo, Vanuatu. Natural history. The egg giving birth to holotype was taken in a myrmecophilous plant (Hydnophytum sp., Rubiaceae) located at about 22 meters high on a gymnosperm tree trunk (Decussocarpus sp.) in primary dry forest at 630 m elevation. The epiphytic plant had an ellipsoidal shape and a size of about 30 25 cm and had numerous holes inside. It was collected by Bruno Corbara and his team using a professional climber. The egg was later hatched in captivity in Paris and grown until adult size by the gecko breeder Francis Girard. A gravid female ready to lay its eggs was also collected on the same epiphytic plant. The eggs of Gehyra vorax Girard (non adhesive) were found in the same plant, just beneath L. buleli adhesive eggs (fig. 6). Interactions between ants living in myrmecophilous plants and L. buleli are unknown. In any case, large areas of such plants are not occupied by ants, thus leaving enough space for geckos. Co-existence of geckos inside myrmecophilous plants on Santo will be reviewed in another paper (Corbara & Ineich in prep.). Lepidodactylus buleli is clearly an arboreal species, according to its collection site, above 20 m high on a tree, but also to its resting position with curled tail and its habits, to easily jump from hand or from support to the soil when handled. The holotype often vocalized when kept captive and disturbed. Reproduction. The female was gravid on 14 November 2006, and its both eggs were ready to be laid according to their size and shell state. We measured 16 eggs referable to L. buleli collected inside an epiphytic myrmecophilous plant from Penaoru. Comparison with 15 eggs of L. vanuatuensis shows them to be significant larger (Table 1). 34 Zootaxa 1918 2008 Magnolia Press INEICH

TABLE 1. Egg size comparison (mm) between L. vanuatuensis (Tasmate Village, west coast of Espiritu Santo) and L. buleli (Penaoru Forest, west coast of Espiritu Santo). L. vanuatuensis L. buleli Number of eggs 15 16 Mean width (range) Mean height (range) 5.95 ± 0.17 5.69 6.20 8.46 ± 0.41 7.61 9.29 7.46 ± 0.30 6.88 7.95 10.36 ± 0.64 9.04 11.47 Comparison. Lepidodactylus buleli belongs to Group II as defined by Brown and Parker (1977). It can be easily distinguished from the sympatric L. vanuatuensis by its narrower toes and fingers, lamellae extending more to the bases of fingers and toes, a more elongate and ovoid tympanum (more circular by L. vanuatuensis), less elongate and more massive head, presence of two enlarged median postmentals in contact with mental plate, grey brownish speckled venter, smaller head granules, and particular colouration (see above). When regarding number of scale rows at midbody, L. buleli, with the highest number of such scales (147), can be easily distinguished from most Group II species: L. gardineri Boulenger, 1897 (103 118), L. guppyi Boulenger, 1884 (110 138), L. intermedius (121), L. lombocensis (110 112), L. novaeguineae Brown and Parker, 1977 (108 125), L. shebae (91), and L. tepukapili (105 118) (see Table 2). TABLE 2. Summary of selected scalation and morphometric data of adult Group II Lepidodactylus. Abbreviations are defined in Material and Methods. Ranges of the data are presented when possible; all values are for both sexes mixed, except males only for cloacal spurs. Data after Brown & Parker (1977), Ota & Hikida (1988), Ota et al. (1995, 1998), Zug et al. (2003), Buden (2007). Lepidodactylus MSR TIVS IOS UL LL WEB CS L. buleli 147 13 14 39 40 10 to 11 10 to 11 less 20 % + L. gardineri 103 to 118 11 to 17 29 to 39 / 31 to 35 11 / 8 to 10? less 20 % + L.guppyi 117 to 138 / 110 to 133 11 to 13 / 9 to 13 29 to 39 / 33 to 37 11 13 / 8 11 10 to 13 25 to 33 % + L. intermedius 121 8? 10 to 11? 20% + L. lombocensis 110 to 112 8 to 9? 9 to 10? less 20 % + L. novaeguineae 124 to 132 / 108 to 125 11 to 16 / 10 to 16 29 to 39 / 32 to 39 9 to 12 / 8 to 10 11 25 to 33 % + L.oligoporus 130 to 134 15 to 19 34 to 35 10 to 12 10 to 12 25 to 33 % no L. paurolepis 98 to 110 12 to 14 29 to 39 / 31 to 32 11 to 12 9 to 11 less 20 % + L. pulcher 143 to 167 17 to 20 39 to 46 10 to 12 12 25 to 33 % + L. shebae 91 11 27 10 9 20 to 33 %? L. tepukapili 100 to 118 12 to 15 29 to 35 7 to 9 8 to 9 less 20 % + L. vanuatuensis 99 to 114 / 91 to 118 10 to 12 35 to 38 / 31 to 35 10 to 12 / 9 to 10 10 to 11 less 20 % + By its scalation, L. buleli is most similar to L. pulcher Boulenger, 1885, from Wild Island off the northwest coast of Manua Island (type locality), and Plot and Lengendrowa Islands on the southeast coast, Admiralty Islands, Papua New Guinea. Records from the Huon Peninsula need confirmation (Bauer & Henle 1994: 162). With that species it shares an elevated number of midbody scale rows and IOS (Table 2). It can however be distinguished from L. pulcher by its lower count of fourth-toe lamellae, its slight toe webbing, the lack of typical heavy spotted head pattern (Boulenger 1885) and presence of dark medio-dorsal black spots even present in preserved specimens (see Plate XIII, fig. 5 of Boulenger 1887a illustrating L. pulcher), shorter and narrower head and its colouration; also by the presence of several cloacal spurs vs. only one in L. pulcher (see A NEW FOREST LEPIDODACTYLUS FROM VANUATU Zootaxa 1918 2008 Magnolia Press 35

Ota et al. 1995). It also differs from L. guppyi by several scalation characters (tabl. 2) and its more flattened and less elongated body shape, colour pattern and interorbital dark band (see pl. 19 20 in McCoy 2006). Discussion A second specimen of L. buleli was accidentally cut with machete when opening an epiphytic myrmecophilous plant on 14 November 2006. That specimen (MNHN 2006.0544) is now in two separate parts and in very bad state. I have not designate it as paratype, but it clearly belongs to the same species as holotype. Its bad state does not allow its use as comparison specimen to determine variation within the species. Tissue was taken from that specimen too. Herpetofaunal elements of Vanuatu have several origins and attempts have been made to explain geological events on the basis of such a composite faunal assemblage (Gibbons 1985, Bauer 1988). Past counterclockwise movement of the Vanuatu archipelago alone does not explain the diversity of its herpetofaunal composition. A separate origin of three island groups of the archipelago as presently considered should not be excluded: (1) northern small islands with strong Solomon affinities (Banks and Torres), (2) central large islands with numerous Pacific widespread species, and (3) southern isolated smaller islands allowing local speciation from rather southern elements (Loyalty affinities) such as Futuna and Anatom Islands. Some species have a clear New Caledonian (more likely Loyalty) origin like Caledoniscincus atropunctatus, Gehyra vorax, Nactus pelagicus, Candoia bibroni or Laticauda frontalis (closely related to Laticauda saintgironsi); others have a Solomon origin, e.g. E. atrocostata, E. cyanogaster, E. nigra, Nactus multicarinatus or Gekko vittatus; a Micronesian origin, e.g. Perochirus guentheri (with some doubts, however, regarding some Tongan fossils tentatively attributed to that genus; see Pregill 1993); while others are clearly southern Vanuatu endemics, probably also with Loyalty affinities, such as Emoia aneityumensis, E. erronan, and E. nigromarginata, most likely also E. sanfordi extending later further north of the archipelago reaching Banks and Torres Islands. Some Vanuatu species are also widespread southern Pacific island species with unclear origins like Lipinia noctua (Asiatic?) or Gehyra oceanica, while others are recently introduced species like the frog Litoria aurea, the gecko Hemidactylus frenatus, and the typhlopid snake Ramphotyphlops braminus. Lepidodactylus vanuatuensis was considered as a species with Solomon affinities according to its low cytochrome-b sequence difference compared with L. guppyi (see Ota et al. 1998). However, that species also closely resembles L. gardineri from Rotuma Island, although the latter species was not tested for its cytochrome-b sequences. If L. vanuatuensis has Solomon affinities it is curious that it is not found on northern Vanuatu, e.g. on Banks and Torres Islands (L. vanuatuensis is only reported from Santo and south of it, no further north). It could rather be related to L. gardineri, having northern/north-eastern Fijian affinities, but that hypothesis has to be tested through molecular studies. Strongest affinities of L. buleli with L. pulcher, an Admiralty Islands endemic, are difficult to explain and even to accept. However, such affinities could also be possible for several other Vanuatu species like Emoia impar and perhaps E. nigra and E. caeruleocauda. Emoia impar is clearly not a Solomon affinity species (contrary to E. cyanura) and its origin is located around Bismarck Archipelago. Vanuatu populations attributed to E. impar certainly correspond to an undescribed species (Bruna et al. 1996), thus showing its long presence on that island group. That blue tailed Emoia lineage clearly has Bismarck/Admiralty affinities. In any event, no Vanuatu or Solomon Group II Lepidodactylus species shares stronger scalation affinities with L. buleli than L. pulcher. Another possible hypothesis is that an undiscovered Group II species related to L. buleli is present somewhere in the Solomons. Nevertheless, L. buleli has no related species in Vanuatu, and L. lugubris (Group III species) and L. vanuatuensis have different and no close affinities with L. buleli. Establishing a molecular phylogeny for the genera Lepidodactylus and Emoia should now be a priority for Pacific island herpetologists. 36 Zootaxa 1918 2008 Magnolia Press INEICH

Finally this study also shows how delicate gecko colouration can be in species description. The holotype of L. buleli can have very different colour patterns in life and in preservative (compare figs. 1 & 2). Both patterns could easily be interpreted as belonging not just to different individuals of the same species, but to different species. However, the basal ground black pattern still persists in life (independent of stress) or in preservative, but its analysis is delicate and needs careful observation and comparison. Live and preserved colour pattern are still important in Lepidodactylus species descriptions, but it has also to be completed with a shape and scalation analyses. Acknowledgments Author wish to thank Philippe Bouchet, Bruno Corbara, Philippe Keith, and Olivier Pascal for their support and considerable organisation work for the success of SANTO 2006 Expedition. Bruno Corbara and his team allowed me to collect some eggs from the myrmecophilous plants they studied. Ms. Donna Kalfatak of the Environment Unit of the Republic of Vanuatu provided valuable support to conduct our research in Vanuatu and to obtain permits. Olivier Gargominy and Benoît Fontaine provided appreciated companionship during fieldwork, and everyday support when drinking evening kava. Special thanks to Francis Girard and Thomas Lefranc for keeping and hatching eggs from all my field trips; their help was considerable. This paper is dedicated to my ni Van field team friends from the Village of Tasmate on the west coast of Santo: William, Walter, Tessy and Ismaël. Literature cited Bauer, A. M. (1988) Hypothesis: a geological basis for some herpetological disjunctions in the southwest Pacific, with special reference to Vanuatu. Herpetological Journal, 1(1988), 259 263. Bauer, A.M. & K. Henle (1994) Familia Gekkonidae (Reptilia, Sauria). Part I Australia and Oceania. Das Tierreich, The Animal Kingdom, Teilband 109. Verlag Walter de Gruyter, Berlin, New York, I XIII + 306 pp. Boulenger, G.A. (1884) Diagnoses of new reptiles and batrachians from the Solomon Islands, collected and presented to the British Museum by H. B. Guppy, Esq. Proceedings of the Zoological Society of London, 1884, 210 213. Boulenger, G.A. (1885) Catalogue of the Lizards in the British Museum (Natural History). Second edition. Volume I, Geckonidae, Eublepharidae, Uroplatidae, Pygopodidae, Agamidae. London (reprinted 1965), Wheldon & Wesley, Ltd. and Verlag J. Cramer, i xii + 436 pp. Boulenger, G.A. (1887a) Catalogue of the Lizards in the British Museum (Natural History). Second edition. Volume III, Lacertidae, Gerrhosauridae, Scincidae, Anelytropidae, Dibamidae, Chamaeleontidae. London (reprinted 1965), Wheldon & Wesley, Ltd. And Verlag J. Cramer, i xii + 575 pp. + 40 pls. Boulenger, G.A. (1887b) Second contribution to the herpetology of the Solomon Islands. Proceedings of the Zoological Society of London, 1887, 333 338. Boulenger G.A. (1897) On the reptiles of Rotuma Island, Polynesia. Annals and Magazine of Natural History, Ser. 6, 20, 306 307. Brown, W.C. (1991) Lizards of the genus Emoia (Scincidae) with observations on their evolution and biogeography. Memoirs of the California Academy of Sciences, 15, i vi + 1 94. Brown, W.C. & A.C. Alcala (1978) Philippine lizards of the family Gekkonidae. Silliman University Natural Science Monograph Series 1, Silliman University Press, Dumaguete City, Philippines, i vi + 1 146. Brown, W.C., McCoy, M. & G.H. Rodda (1992) A new Lepidodactylus (Reptilia: Gekkonidae) from Guadalcanal Island, Solomons. Proceedings of the Biological Society of Washington, 105(3), 440 442. Brown, W.C. & F. Parker (1977) Lizards of the genus Lepidodactylus (Gekkonidae) from the Indo-Australian Archipelago and the islands of the Pacific, with descriptions of new species. Proceedings of the California Academy of Sciences, Fourth Series, 41(8), 253 265. Brown, W.C. & V.M. Tanner (1949) Rediscovery of the genus Pseudogekko, with description of a new species from the Solomon Islands. Great Basin Naturalist, 9, 41 45. Bruna, E. M., Fisher, R.N. & T.J. Case (1996) Morphological and genetic evolution appear decoupled in Pacific skinks (Squamata: Scincidae: Emoia). Proceedings of the Royal Society of London B, 263, 681 688. A NEW FOREST LEPIDODACTYLUS FROM VANUATU Zootaxa 1918 2008 Magnolia Press 37

Buden, D.W. (2007) A new species of the genus Lepidodactylus Fitzinger (Squamata: Gekkonidae) from the Mortlock Islands, Chuuk State, Federated States of Micronesia. Pacific Science, 61(3), 407 414. Cope, E.D. (1869) Observations on Reptiles of the old world. Proceedings of the Academy of natural sciences of Philadelphia, 1868, 316 323. Darevsky, I.S. (1964) Two new species of gekkonid lizards from the Komodo Island in Lesser Sundas Archipelago. Zoologischer Anzeiger, 173, 169 174. Duméril, A.M.C. & G. Bibron (1836) Erpétologie générale ou histoire naturelle complète des reptiles. Librairie Encyclopédique de Roret, Paris. Vol. 3, i iv + 517 pp. Gibbons, J. H. (1985) The biogeography and evolution of Pacific Island reptiles and amphibians. In: Grigg, G., Shine, R. & Ehmann, H. (Eds.), Biology of Australasian frogs and reptiles. Royal zoological Society of New South Wales, pp. 125 142. Gibbons, J.R.H. & W.C. Brown (1988) A new Lepidodactylus (Gekkonidae) from Eua Island, Tonga. Herpetologica, 22(3), 356 360. Ineich, I. (1987) Recherches sur le peuplement et l'évolution des Reptiles terrestres de Polynésie française. Ph.D. Thesis, Université des Sciences et Techniques du Languedoc, Montpellier, 515 pp. Ineich, I. (1988) Mise en évidence d'un complexe unisexué-bisexué chez le gecko Lepidodactylus lugubris (Sauria, Lacertilia) en Polynésie française. Compte-Rendus de l Académie des Sciences Paris, 307(série III), 271 277. Ineich, I. (1999) Spatio-temporal analysis of the unisexual-bisexual Lepidodactylus lugubris complex (Reptilia, Gekkonidae). In: H. Ota (Ed.), Tropical Island Herpetofauna : Origin, Current Diversity and Conservation. Elsevier, Developments in Animal and Veterinary Sciences, 29, pp. 199 228 (353 pp.). Kluge, A.G. (1968) Phylogenetic relationships of the gekkonid lizards genera Lepidodactylus FITZINGER, Hemiphyllodactylus BLEEKER, and Pseudogekko TAYLOR. Philippine Journal of Sciences, 95[1966], 331 352. McCoy, M. (2006) Reptiles of the Solomon Islands. Pensoft Publ., Sofia (Bulgaria), Moscow, 147 pp. + 64 unumb. pl. Mertens, R. (1929) Zwei neue Haftzeher aus dem Indo-Australischen Archipel (Rept.). Senckenbergiana, 11, 237 241. Motteler, L.S. (1986) Pacific island names. A map and name guide to the New Pacific. Bishop Museum Miscellaneous Publications, 34, 1 91. Ota, H. (1987) A new species of Lepidodactylus (Gekkonidae: Reptilia) from Lanyu Island, Taiwan. Copeia, 1987(1), 164 169. Ota, H. & R.I. Crombie (1989) A new lizard of the genus Lepidodactylus (Reptilia : Gekkonidae) from Batan Island, Philippines. Proceedings of the Biological Society of Washington, 102(3), 559 567. Ota, H., Darevsky, I.S., Ineich, I. & S. Yamashiro (2000) Reevaluation of the taxonomic status of two Lepidodactylus species (Squamata: Gekkonidae) from the Lesser Sunda Archipelago, Indonesia. Copeia, 2000(4), 1109 1113. Ota, H., Fisher, R.N., Ineich, I. & T.J. Case (1995) Geckos of the genus Lepidodactylus (Squamata: Reptilia) in Micronesia: description of a new species and reevaluation of the status of Gecko moestus Peters, 1867. Copeia, 1995(1), 183 195. Ota, H., Fisher, R.N., Ineich, I., Case, T.J., Radtkey, R.R. & G.R. Zug (1998) A new Lepidodactylus (Squamata: Gekkonidae) from Vanuatu. Herpetologica, 54(3), 325 332. Ota, H. & T. Hikida (1988) A new species of Lepidodactylus (Sauria : Gekkonidae) from Sabah, Malaysia. Copeia, 1988(3), 616 621. Pernetta, J.C. & D. Black (1983) Species of gecko (Lepidodactylus) in the Port Moresby area, with the description of a new species. Journal of Herpetology, 17(2), 121 128. Peters, W. (1867) Herpetologische Notizen. Monatsberichte der Akademie für Wissenschaften zu Berlin, 1867, 13 37. Pregill, G. K. (1993) Fossil lizards from the late Quaternary of 'Eua, Tonga. Pacific Science, 47, 101 114. Radtkey, R.R., Donnellan, S.C., Fisher, R.N., Moritz, C., Hanley, K.A. & T.J. Case (1995) When species collide: the origin and spread of an asexual species of gecko. Proceedings of the Royal Society of London, 259, 145 152. Saint Girons, H. & I. Ineich (1992) Histology of the reproductive tract of hybrids between gonochoristic males and parthenogenetic females of Lepidodactylus lugubris (Duméril and Bibron, 1836) in French Polynesia (Reptilia, Gekkonidae). Journal of Morphology, 212, 55 64. Uetz, P. & J. Hallermann (2008) The TIGR Reptile database. Available from http://www.reptile-database.org, www.tigr.org (accessed 7 March 2008). Zug, G.R., Watling, D., Alefaio, T., Alefaio, S. & C. Ludescher (2003) A new gecko (Reptilia: Squamata: Genus Lepidodactylus) from Tuvalu, South-central Pacific. Proceedings of the Biological Society of Washington, 116(1), 38 46. 38 Zootaxa 1918 2008 Magnolia Press INEICH