We record a tree frog of the genus Chiromantis for the first time from outside the Southeast Asian continent and describe it as a new species, Chiromantis inexpectatus. The new species from the Malaysian state of Sabah, Borneo, is a small-sized Chiromantis (male snout-vent length ca. 22 mm), and is distinguished from all other members of the genus by the combination of the following morphological characteristics: dark stripes absent, but dark spots present on dorsum; a dark-brown lateral band present from snout tip to half of body, bordered ventrally by white stripe; third and fourth fingers less than half webbed; third finger disk wider than tympanum diameter; and inner metatarsal tubercle present. Significance of findings of this species from Borneo Island, as well as phylogeny and breeding habit of the genus Chiromantis, are briefly discussed.
The Old-World tree-frog genus Chiromantis Peters, 1854 had long been confined to several African tree frogs (Liem, 1970; Frost, 1985), but based on results of recent molecular studies on rhacophorids (Frost et al., 2006), the genus was enlarged to contain frogs formerly known as Chirixalus Boulenger, 1893 occurring from southern China through Thailand to northeastern India (e.g., Bourret, 1942; Taylor, 1962). Recently, two additional new species have been described from the Asian continent (Grismer et al., 2007; Chan et al., 2011).
Outside the continental region of Southeast Asia, no formal report of occurrence of the genus has been given, although their presence in Sumatra and Java is likely (Amir Hamidy, pers. comm.). However, our field survey in Sabah, Malaysian Borneo, revealed presence of a small rhacophorid species morphologically similar to the genus Chiromantis there. The species formed a clade on mtDNA trees with some members of the genus, and also differed from all the known congeneric members morphologically. Hence we describe it as a new species of Chiromantis, which is the first formal record of the genus outside the Southeast Asian continent.
MATERIALS AND METHODS
We obtained DNA sequence data from tissues preserved in 99% ethanol for a specimen of rhacophorid species from Sabah (BORNEENSIS [BORNEENSIS collection, Institute for Tropical Biology and Conservation, University Malaysia Sabah] 22421; GenBank accession number AB813160), Chiromantis doriae (Boulenger, 1893) from Phu Luang, Thailand (KUHE [Kyoto University, Graduate School of Human and Environmental Studies] 19301; AB813159), C. nongkhorensis (Cochran, 1927) from Song Khla Buri, Thailand (KUHE 19498; AB813158), C. xerampelina Peters, 1854 from pet trade (KUHE 46345; AB813157), C. vittatus (Boulenger, 1887) from Song Khla Buri, Thailand (KUHE 19441; AB813162), C. hansenae (Cochran, 1927) from Nam Tok Hongkaew, Thailand (KUHE 34136; AB813161), Rhacophorus gauni (Inger, 1966) from Sarawak (KUHE 53511; AB847124), R. dulitensis Boulenger, 1892 from Sabah (BORN 09097; AB847123), Feihyla kajau (Dring, 1984) from Sarawak (KUHE 53591; AB847122), Theloderma leprosum Tschudi, 1838 from Malay Peninsula (KUHE 52581; AB847128), Kurixalus appendiculatus (Günther, 1859) from Sarawak (KUHE 53614; AB847125), Philautus davidlabangi Matsui, 2009 from Sarawak (KUHE 19594; AB847127), and Ph. macroscelis (Boulenger, 1893) from Sabah (BORN 08499; AB847126). Methods for DNA extraction, and amplification and sequencing of the mtDNA fragments are the same as those reported by Kuraishi et al. (2013). We deposited the resultant sequences (1987 base pairs [bp] of partial sequences of mitochondrial 12S and 16S rRNA genes and the intervening tRNA gene) in GenBank (Accession numbers AB813157–813162, 847122– 847128, as shown above). In addition to our own new sequence data, we used GenBank data of C. rufescens (Günther, 1869) (AF 458126), Polypedates leucomystax (Gravenhorst, 1829) (AB728138), R. borneensis Matsui et al., 2013 (AB781693), Nyctixalus pictus (Peters, 1871) (DQ283133), Buergeria buergeri (Temminck and Schlegel, 1838) (AB127977), and a mantellid, Aglyptodactylus madagascariensis (Duméril, 1853) (DQ283056). For tree construction and calculation of genetic distances (uncorrected p-distance), we followed Kuraishi et al. (2013).
Specimens examined are stored in the BORNEENSIS collection. We took the following 25 body measurements to the nearest 0.05 mm with a dial caliper under a binocular microscope, following Matsui (1984, 1994): (1) snout-vent length (SVL); (2) head length (HL) from tip of snout to hind border of angle of jaw, not measured parallel with the median line; (3) snout-nostril length (S-NL); (4) nostril-eyelid length (N-EL); (5) snout length (SL); (6) eye length (EL); (7) eye diameter (ED), diameter of the exposed portion of the eyeball; (8) tympanum-eye length (T-EL); (9) tympanum diameter (TD); (10) head width (HW); (11) internarial distance (IND); (12) interorbital distance (IOD); (13) upper eyelid width (UEW); (14) forelimb length (FLL); (15) lower arm and hand length (LAL); (16) first finger length (1FL), measured from distal edge of inner palmar tubercle; (17) inner palmar tubercle length (IPTL); (18) hindlimb length (HLL); (19) thigh length (THIGH); (20) tibia length (TL); (21) foot length (FL); (22) first toe length (1TOEL); (23) inner metatarsal tubercle length (IMTL); (24) third finger disk diameter (3FDW); and (25) fourth toe disk diameter (4TDW). We followed the system of description of toe-webbing states used by Savage (1975).
The topology of the maximum-likelihood (ML) and Bayesian trees was nearly identical, and only the former is shown in Fig. 1. Support for the monophyly of the rhacophorid sp. from Sabah and six species of Chiromantis was sufficient in the Bayesian tree (Bayesian posterior probability [BPP] = 0.98), but was moderate in the ML tree (bootstrap support [BS] = 67%). Two groups were recognized in the clade, and the first group formed a fully-supported clade (BS = 100%, BPP = 1.00) and contained four Chiromantis species, in which C. doriae and C. nongkhorensis, and C. xerampelina and C. rufescens, respectively, showed sister relationships. The second group contained a fully-supported clade of C. vittatus and C. hansenae (BS = 100%, BPP = 1.00) and the rhacophorid sp. from Sabah, and the monophyly of the group was moderately supported (BS = 71%, BPP = 0.97). Although the relationships were not highly supported, these results indicate placement of the rhacophorid sp. from Sabah in the genus Chiromantis.
Uncorrected p-distances (in %) for fragment of 16S rRNA among 18 rhacophorid and a mantellid taxa compared.
Chiromantis sp. from Sabah differed genetically from the six species of Chiromantis by large genetic distances (16.1– 18.6%: Table 1), and the species also substantially differed from species of the other genera (14.7% from F. kajau to 20.7% from P. leucomystax). The distances between the six species of Chiromantis ranged from 5.7–20.9%. Further more, Chiromantis sp. from Sabah is separated morphologically from all nominal species of Chiromantis in congruent with genetic separation. Thus, we conclude the rhacophorid specimens from Sabah, Borneo as a distinct species in the genus Chiromantis and describe it as follows:
Chiromantis inexpectatus is distinguishable from its congeners by the combination of the following characters: (1) size small, SVL 22.2–22.4 mm in males, (2) dark stripes absent, but dark spots present on dorsum, (3) a dark-brown lateral band present from snout tip to half of body, bordered below by white stripe, (4) third and fourth fingers less than half webbed (III 2–2.5 IV), (5) third finger disk wider but toe disks narrower than tympanum diameter, (6) inner metatarsal tubercle present.
The specific name is a Latin adjective, referring to the fact that the occurrence of the genus on the island of Borneo, where the new species was found, was unexpected.
BORNEENSIS 22421, adult male, collected by Tomohiko Shimada on 9 March 2005 from Camel Trophy field station (4°54' N, 116°53' E; ca. 1050 m a.s.I.) of the Maliau Basin Conservation Area, Sandakan Division, Sabah, East Malaysia.
BORNEENSIS 22419, metamorphosing young, and BORNEENSIS 22420, adult male, same collection details as for holotype.
Description of holotype
Adult male; SVL 22.2 mm; head longer (HL 38.7% SVL) than wide (HW 36.9% SVL), wider than body, relatively flat; snout truncated in lateral view, length (SL 14.9% SVL) subequal to eye length (EL 14.4% SVL), sloping anteroventrally, projecting beyond mouth; canthus rostralis rounded; loreal region vertical and concave; nostril nearer to tip of snout (SNL 8.1% SVL) than to eye (N-EL 9.0% SVL); internarial distance (IND 12.4% SVL) less than interorbital distance (IOD 14.9% SVL), which in turn nearly twice as wide as upper eyelid (UEW 7.9% SVL); eye large, protuberant; tympanum distinct, subcircular, diameter (TD 5.4% SVL) three-eighths of eye length and separated from eye by one-eighth of tympanum diameter (T-EL 0.7% SVL); vomerine teeth absent; choana oval; tongue notched posteriorly; single median vocal sac; vocal slits on floor of mouth well anterior to jaw commissure on both sides.
Forelimb long (FLL 67.1% SVL); hand and forearm long (LAL 56.3% SVL) and relatively robust; finger length formula: I < II < IV < III; expanded disks each with a circummarginal groove and a transverse ventral groove; disks on third and fourth fingers (3FDW 6.1% SVL) only slightly wider than those on second (2FDW 6.0% SVL) and first (1FDW 5.9% SVL), all wider than tympanum; no webbing between first and second fingers; second and third fingers, and third and fourth fingers less than half webbed (II 3–3 III 2–2.5 IV); inner two fingers widely separated from outer two fingers (opposable); subarticular tubercle between penultimate and adjoining proximal phalange on third and fourth fingers indistinct; inner palmar tubercle flat (IPTL 5.4% SVL), outer one absent; nuptial pad absent.
Hindlimb relatively short (HLL 159.9% SVL); tibiotarsal articulation extends to anterior corner of eye when fully stretched leg adpressed to body; heels touching each other when thigh (THIGH 54.7% SVL) and tibia (TL 54.5% SVL) placed at right angle to body; foot (FL 41.9% SVL) much shorter than tibia; toe length formula I<II<III<V< IV; toes bearing expanded disks each with a circummarginal groove and a transverse ventral groove, width of fourth toe disk (4TDW 4.7% SVL) much narrower than those of finger disks; webbing formula I 2–2 II 11/2–2 III 11/2–2 IV11/2–11/2 V; subarticular tubercles oval, poorly developed; inner metatarsal tubercle small (IMTL 3.6% SVL) and flat, about onethird length of first toe (1TOEL 11.5% SVL); no outer metatarsal tubercle.
Dorsal surface nearly smooth, sparsely scattered with minute, blunt asperities between shoulder and sacral regions; supratympanic fold weak, continued as glandular fold to shoulder; skin of lower jaw posterolateral^ forming a short fold at anterior base of upper arm; ventral surface composed of flat granules; skin of ventral surface of foot smooth, with longitudinal folds.
The color changes in life: when captured, the dorsal ground color was whitish-yellow on body and pinkish on limbs, but later changed to light brown (Fig. 2). Dorsum scattered with vaguely defined small dark-brown spots; a darkbrown lateral band extending from tip of snout, through lower eyelid, upper half of tympanum, to fade one-half way down body, bordered ventrally by a narrower white stripe extending from snout tip below canthus, through lower eyelid and tympanum; venter immaculate creamy white, semitransparent.
Measurements (in mm) of types of Chiromantis inexpectatus. See text for abbreviations.
In alcohol, all dorsal and ventral surfaces faded to beige, but dark-brown and white stripes remained.
Morphometric variation is shown in Table 2. The male paratype (BORNEENSIS 22420) is generally similar with the holotype in general morphology and coloration, but has more isolated, clearer dark spots. Another paratype (BORNEENSIS 22419) is a metamorphosing juvenile with a tail stub of 0.1 mm and lacks a visible tympanum. Darkbrown lateral band is indistinct and ventral white stripe is not recognizable.
Known from the type locality, Camel Trophy of the Maliau Basin Conservation Area, Sandakan Division, State of Sabah, Malaysian Borneo.
In the type locality, Camel Trophy, two adult males and a metamorphosing juvenile of the type series were simultaneously found at night perching each on a leaf of low trees (< 1 m) extending above the surface of a shallow pool (3–5 m × 10 m). The air temperature before the time of finding was 24°C. No tadpoles were found in the pond nor eggs on overhanging leaves of trees. Early March may not be a breeding season since males were not calling, but is surely in the season of metamorphosis as evidenced by the presence of a juvenile with a degenerating tail. Frogs found together with the new species included Rhacophorus borneensis, Kurixalus appendiculatus, Polypedates macrotis (Boulenger, 1891), and Microhyla petrigena Inger and Frogner, 1979.
Chiromantis inexpectatus is distinguished from C. cherrapunjiae (Roonwal and Kripalani, 1966) from northern India by light brown or yellowish brown dorsum with dark-brown spots, dark-brown dorsolateral stripe bordered below by white stripe, and the presence of inner metatarsal tubercle (vs. dorsum pale green to dark green and inner metatarsal tubercle absent in C. cherrapunjiae: Chanda, 1994; Mathew and Sen, 2010); from C. doriae from northern India through Myanmar, Thailand, Laos, Cambodia, and Vietnam to China, by the absence of dark dorsal stripes and presence of external vocal sac (vs. dark dorsal stripe present and external vocal sac absent in C. doriae); from C. dudhwaensis (Ray, 1992) from northern India by the possession of dorsum with dark-brown spots, and third finger disk larger than tympanum (vs. dorsum with dark stripe, and disk on third finger smaller than tympanum in C. dudhwaensis); from C. hansenae from Thailand by dark-brown dorsolateral stripe bordered below by white streak on anterior half of flank, and clearly webbed second to fourth fingers (vs. dorsolateral stripe cream bordered with violet streaks on each edge to groin, and webs on hand rudimentary in C. hansenae: Taylor, 1962); from C. marginis Chan et al., 2011 from the Malay Peninsula by dorsolateral stripe dark-brown, bordered below by white streak, both of which not continuous but fading one-half way down body, webbing between third and fourth fingers not extensive (III 2–2.5 IV), and disk on third finger larger than tympanum (vs. dark-brown dorsolateral stripe bordered above by white stripe, both of which continuing to groin, extensive webbing between third and fourth fingers [III 1.5–IV], and third finger disk smaller than tympanum in C. marginis: Chan et al., 2011); and from C. nongkhorensis from Myanmar, Thailand, Laos, Cambodia, Vietnam, and Malaysia, by smaller body size, relatively small eye shorter than snout, possession of clear dorsolateral stripes and dorsum speckled with small dark-brown spots (vs. body larger, SVL 29–32 mm in males, eye longer than snout, dorsolateral stripes absent, and distinct dark dorsal markings between orbital and sacral regions in C. nongkhorensis).
The new species also differs from the following species morphologically: from C. punctatus (Wilkinson et al., 2003) from Myanmar by dorsum scattered with vaguely defined small dark-brown spots, and dark-brown dorsolateral stripe bordered ventrally by narrow white stripe extending from snout tip below canthus, through lower eyelid to one-half way down body (vs. dorsum with many clearly defined darkbrown spots, and wide white dorsolateral stripe running from snout above canthus rostralis through upper eyelid to groin, bordered ventrally by narrow dark stripe in C. punctatus); from C. samkosensis Grismer et al., 2007 from Cambodia by the dorsal ground color of yellow to light brown, distinct white streak below canthus, and possession of disk on third finger larger than tympanum (vs. ground color of dorsum lime-green, no white streak below canthus, and third finger disk slightly smaller than tympanum in C. samkosensis: Grismer et al., 2007); from C. senapatiensis (Mathew and Sen, 2009) from northern India by the absence of dark dorsal stripe and presence of clear webs on second to fourth fingers (vs. dark middorsal stripe present and webs on hand rudimentary in C. senapatiensis: Mathew and Sen, 2010); from C. shyamrupus (Chanda and Ghosh, 1989) from northern
India by the absence of dark dorsal stripes, and presence of clear webs on second to fourth fingers and inner metatarsal tubercle (vs. a dark middorsal stripe present, fingers free of web, and inner metatarsal tubercle absent in C. shyamrupus: Chanda, 1994; Mathew and Sen, 2010); from C. simus (Annandale, 1915) from northern India by the smaller body size (male SVL 22 mm) and the absence of uniform tuberculation and dark stripes on dorsum (vs. body larger, SVL 30 mm, and dorsum uniformly tuberculate with dark stripes in C. simus: Ahmed et al., 2009); and from C. vittatus from northern India through Myanmar, Thailand, Laos, Cambodia, and Vietnam to China, by the smaller body size and presence of dark-brown dorsolateral stripe bordered below by white streak on anterior half of flank (vs. size larger, males 26–28 mm in SVL, and dorsolateral stripe bordered above by white stripe running to groin in C. vittatus).
The African species, C. kelleri Boettger, 1893 from Ethiopia and Somalia through Kenya to Tanzania, C. petersii Boulenger, 1882 from Kenya and Tanzania, C. rufescens from West Africa, and C. xerampelina from Kenya, Angola and Namibia to South Africa, are all much larger (SVL 35–80 mm: Liem, 1970) than the present new species (ca. 22 mm).
The island of Borneo is famous for its high frog diversity, and as many as 155 species have been recorded by the beginning of this century (Matsui, 2006). The number of species is still increasing, and the situation for rhacophorids is no exception. Several new species have been described recently based chiefly on morphological and acoustic data (Dehling, 2008, 2010; Dehling and Grafe, 2008; Matsui, 2009), as well as molecular approaches (Matsui et al., 2013), further contributing to the discovery of the Bornean frog diversity.
Compared with these recent findings, discovery of Chiromantis from Borneo Island is biogeographically more significant, since occurrence of the genus has never been expected on the island. However, the finding is not surprising, because another rhacophorid tree frog, Theloderma horridum Boulenger, 1903, long known only from the Peninsular Thailand and Malaysia was also discovered from the island rather recently (Inger et al., 1995). Tree frogs are generally more difficult to be detected than frogs and toads of the other lineages because of their habit of spending most of their life on trees, sometimes on high crown, and can mainly be found only in the breeding seasons when they come down to lower places near the water body.
Maliau Basin Conservation Area, where the present new species was found, is situated in the southern part of Sabah, and is not easy of access. The three specimens were obtained during a joint expedition of University Malaysia Sabah (UMS) and Japan International Cooperation Agency (JICA), together with the recently described Rhacophorus borneensis, which had been confused with R. reinwardtii Schlegel, 1840 (Matsui et al., 2013), suggesting that the region has never been amply surveyed.
Our phylogenetic analyses using limited samples indicated deep genetic divergence among species of Chiromantis. Of the Asian species, C. doriae and C. nongkhorensis formed a well-supported clade with the African taxa, but the other Asian species were much divergent from it, supporting the findings of Frost et al. (2006) and Li et al. (2009). The genus is reported to be monophyletic when it is compared with the Indian genus Ghatixalus Biju et al., 2008 (Pyron and Wiens, 2011), but further analyses including species from India and those recently described from Southeast Asia are necessary to determine the validity of the monophyly of Chiromantis.
The African C. xerampelina is noted for the construction of a foamy egg mass (e.g., Noble, 1931), and C. dorae and C. nongkhorensis also lay eggs in a foamy mass (Taylor, 1962; Fei et al., 2009; M. Matsui, pers. obs.). The Indian member C. simus is also reported to lay eggs in a foamy nest (Ahmed et al., 2009). In contrast, C. vittatus lays smaller, non-foamy egg mass on leaves (Ahmed et al., 2009; Fei et al., 2009; M. Matsui pers. obs.). Variation in the mode of egg deposition is another interesting problem in understanding phylogeny and evolution in the genus Chiromantis.
The Economic Planning Unit of Malaysia (EPU: 40/200/19 SJ. 1158) and Sabah Parks (SP) granted permission for fieldwork. UMS and JICA kindly provided all the facilities for conducting the field research. T. Shimada thanks L. Apin, Y. Hashimoto, K. B. Kueh, K. Tanaka, and S. Yasuma for help during the field trip. We are particularly indebted to N. Kuraishi and K. Eto for help in the laboratory. This research was partly supported by a fellowship of Japan Society for the Promotion of Science (JSPS) to T. Shimada and a Grant-in Aid from the Monbusho through JSPS (Field Research, Nos. 15370038, 20405013 and 23405014) to M. Matsui.