INTRODUCTION

Liem (1970) made an extensive morphological study of the Old World tree frogs and established a taxonomy to split members of the family Rhacophoridae into ten genera. Since then several new suprageneric classifications have been proposed by subsequent authors (e.g., Channing, 1989; Dubois, 1981, 1992; Bossuyt and Dubois, 2001), and many new species have been added to some genera (e.g., Rhacophorus, Polypedates, and Philautus: Inger et al., 1999; Manamendra-Arachchi and Pethiyagoda, 2001; Orlov et al., 2001; Ziegler und Köhler, 2001). The genus Chirixalus Boulenger, 1893, however, is limited in the range of distribution and small in the number of species included, as shown by the fact that few new members have been added to the genus (Glaw et al., 1998).

In Vietnam, extensive field surveys by many groups of herpetologists have resulted in the discovery of new frog taxa of different lineages including rhacophorid genera Philautus and Rhacophorus (Inger et al., 1999; Ohler et al., 2000; Orlov et al., 2001; Ziegler und Köhler, 2001). Many undescribed species, however, are actually still present in this country, and during our survey of the central region in 1997, we collected a pair of specimens of a rhacoporid distinctly different from the species hitherto known to occur in Vietnam (Bourret, 1942; Inger et al., 1999). Although allocations of a species into a known genus of rhacophorids include severe taxonomic problems (see Discussion), we assign these specimens to the genus Chirixalus and describe them below as a new species.

MATERIALS AND METHODS

The junior author conducted field work at Vu Quang Nature Reserve, Rao Cua District, Ha Tinh Province, central Vietnam in June of 1997. Specimens were fixed in 10% formalin and later preserved in 70% ethanol. For preserved specimens, we took 18 body measurements (Table 1), mainly following Matsui (1984), to the nearest 0.1 mm with dial calipers under a binocular dissecting microscope. Webbing formula is that of Myers and Duellman (1982). We prepared radiographs to examine gross osteology, and made slight dissections to examine gonads. For comparisons, we examined museum specimens, including types, of most described taxa of the genus Chirixalus and allied genera (Appendix 1).

Table 1

Measurements in mm. SVL: snout-vent length; HL: head length (= from tip of snout to hind border of the angle of jaw, mea-sured not parallel with the median line); SL: snout length; EL: eye length; T-EL: tympanum-eye length; TD: tympanum diameter; HW: head width; IND: internarial distance; IOD: interorbital distance; UEW: upper eyelid width; FLL: forelimb length; 3FDW: third finger disk diameter; HLL: hindlimb length; TL: tibia length; FL: foot length (= from proximal end of inner metatarsal tubercle to tip of fourth toe); 4HDW: fourth toe disk diameter; 1TL: first toe length (= from distal end of inner metatarsal tubercle to tip of first toe); IMTL: inner meta-tarsal tubercle length.

SYSTEMATICS

Chirixalus ananjevae sp. nov. (Figs. 1, 2)

Fig. 1

Dorsal (A) and ventral (B) views of male holotype (KUHE 31963).

Fig. 2

Ventral views of hand (A) and foot (B) of male holotype (KUHE 31963). Scale bar indicates 5 mm.

DISCUSSION

Dubois (1981) reexamined Liem's (1970) results and split the genus Philautus Gistel, 1848, from all other members of Rhacophoridae and erected a monotypic subfamily Philautinae. He, however, later relegated the family Rhacophoridae to a subfamily of Ranidae and recognized three tribes (Dubois, 1992). By this action, Philautinae (Dubois, 1981) was changed to the tribe Philautini, which encompasses three subgenera. Regardless of this taxonomic change, the genus Philautus is defined by unique characters in internal and external morphology (see Liem, 1970 for details). One unique generic feature of external morphology is the absence or slight development of web between two outer metatarsals (Dubois, 1981; Manthey and Grossmann, 1997). The genus, however, is more clearly characterized by its unique reproductive habit of direct development from large yolky egg (Dubois, 1992, 1999). Unlike Philautus, the present new species has a web which extends between two outer metatarsals. Moreover the female possesses relatively large number of small eggs that are different from eggs of species with direct development.

The other six east Asian genera of Rhacophoridae (Buergeria, Chirixalus, Nyctixalus, Polypedates, Rhacophorus, and Theloderma), once grouped into the subfamily Rhacophorinae (Dubois, 1981), were divided into two tribes (Buergerini and Rhacophorini) including five genera (the genus Polypedates was synonymized with Rhacophorus: Dubois, 1992). The tribe Buergeriini (Channing, 1989) is monotypic, including the genus Buergeria which inhabits streams and has long hindlimbs with broad webs between toes (Maeda and Matsui, 1999). From the poor development of webs on toes and short hindlimb, the present new species is not considered to be a member of Buergeria, although we have no information of its breeding habits. Further, diplasiocoelous vertebral column and presence of vomerine teeth in Buergeria preclude the present new species from the genus.

Of the genera of the tribe Rhacophorini, the present new species is not assigned to Nyctixalus, because the pupil is not vertical. The new species is included neither in Rhacophorus (including Polypedates) nor in Theloderma, because the tip of finger bones is not Y-shaped (Manthey and Grossmann, 1997; Fei, 1999).

The remaining genus, Chirixalus, was proposed by Boulenger (1893:340) on the basis of C. doriae from Myanmar. Among diagnoses of the genus given by him, the lack of vomerine teeth and presence of fingers with the two inner opposed to the two outer seem to be unique. However, the former feature is currently considered to be variable within a genus of rhacophorids such as Philautus and Rhacophorus (Liem, 1970). The latter feature has been repeatedly listed as a diagnostic character of the genus, and Liem (1970) also noted it without actual character analyses. One important problem made by Liem (1970) is that he added in this genus C. eiffingeri (Boettger, 1895) from the Ryukyu Archipelago, Japan, and Taiwan, which actually lacks opposable fingers.

As if reflecting this discrepancy, C. eiffingeri has recently been separated from Chirixalus to represent a monotypic genus Kurixalus by Ye, Fei, and Dubois (Fei, 1999:383). This genus was defined by several morphological characteristics of adults, including weak development of vomerine teeth, non-opposable inner and outer fingers, and extremely developed pollex, but more important feature of Kurixalus is unique morphology and ecology of larvae that are related to oviphagy.

Their classification (Ye, Fei, and Dubois in Fei, 1999), however, includes a problem, because a sister species of C. eifingeri, C. idiootocus Kuramoto and Wang, 1987, from Taiwan, was not included in the genus Kurixalus, but was moved to Philautus without any reason. These two species are generally very close morphologically and differ from each other chiefly in the condition of pollex in adults, and in larval biology, with C. idiootocus having normal, non-enlarged pollex and normally feeding larvae (Kuramoto and Wang, 1987). They are also estimated to form a genetically unique lineage (Richards and Moore, 1998) and are suggested to be placed together in Kurixalus, (Wilkinson et al., 2002).

The present new species from Vietnam is closer to C. eiffingeri and C. idiootocus than any other rhacophorids in external morphology and might also be included in Kurixalus. However, it is premature to recognize Kurixalus as a good genus distinct from Chirixalus because of two reasons. First, the genus Kurixalus, after inclusion of C. idiootocus and C. ananjevae, needs definition by clear synapomorphic characters, but this has never been done. Second, the genus Chirixalus itself is now considered to be polyphyleic (Wilkinson et al., 2002), but no taxonomic assignment has been made as yet. Therefore, setting the problems of generic definition aside, it would at present be better to place the present new species in the genus Chirixalus (sensu lato) so as to avoid further taxonomic confusion until more extensive taxonomic revision of Chirixalus is made.

The present new species, C. ananjevae is similar to C. eiffingeri and C. idiootocus in the absence of opposable inner and outer fingers. However, it clearly differs from C. eiffingeri in the absence of vomerine teeth or development of pollex, and from C. idiootocus in less pointed snout, absence of vomerine teeth, and larger body. Additionally, C. idiootocus has white granules around anus and on limbs, and dark markings on sides of body, none of which are present in C. ananjevae. The remaining members of the genus (Glaw et al., 1998) are reported to have opposable fingers, and are easily distinguished by external appearance from C. ananjevae. Chirixalus doriae Boulenger, 1893 and C. vittatus (Boulenger, 1887) from Myanmar to China, including Vietnam, C. hansenae (Cochran, 1927) from Thailand, C. laevis (Smith, 1924) from Vietnam, and C. nongkhorensis (Cochran, 1927) from Myanmar and Thailand, have more slender body with smaller head and more elongate trunk, as well as smoother back than C. ananjevae. The color and pattern of markings on dorsum are also much different. Chirixalus simus Annandale, 1915, from Assam, India, is similar to C. ananjevae in rugose dorsal skin and dorsal coloration, but is distinctly different from it in its smaller body size (22 mm SVL), shorter snout, longer hind-limb, and yellowish ventrum with clear markings. Another Indian species, C. dudhwaensis Ray, 1992, from Uttar Pradesh is very similar to C. simus (Ray, 1992; Das, personal communication). Neither do we consider here several other species transferred from other genera to Chirixalus by Dubois and Ohler (2001), because their list (Dubois and Ohler, 2001: Table 6) includes some misidentifications (e.g., Philautus romeri Smith, 1953 from Hong Kong, which actually exhibits direct development and should remain placed in Philautus).

It is very interesting biogeographically, if Vietnamese C. ananjevae is actually phylogenetically close to two Taiwanese species, because no other amphibians exhibit such a distribution pattern. However, another rhacophorid genus, Buergeria, occurs from Japan to Taiwan and in Hainan Island (Maeda and Matsui, 1999), and this distribution may support the close relationships of Vietnamese and Taiwanese Chirixalus. In order to ascertain their relationships, future biochemical analyses would play a great role, although many problems have been only partially resolved so far by the molecular approach as shown by low supports of many clades shown by Wilkinson et al. (2002). Along with further biochemical information, data for natural history, especially for breeding habits, are also badly needed.