Rhacophorus dugritei (He, 1999)

Li, Jia-Tang, Liu, Jun, Chen, Yue-Ying, Wu, Jia-Wei, Murphy, Robert W., Zhao, Er-Mi, Wang, Yue-Zhao & Zhang, Ya-Ping, 2012, Molecular phylogeny of treefrogs in the Rhacophorus dugritei species complex (Anura: Rhacophoridae), with descriptions of two new species, Zoological Journal of the Linnean Society 165 (1), pp. 143-162 : 156-157

R. DUGRITEI SPECIES COMPLEX

The molecular phylogeny of the R. dugritei species complex is based on total nucleotide evidence. The R. dugritei species complex contains three, wellsupported lineages with relationships: (A, B)C ( Fig. 2 ).

Lineage A

Barbour (1912) describes Hyla monticola (later replaced by Hyla bambusicola ) based on specimens from Washan (=Mount Wa), Sichuan. Liu (1950) recognized the validity of R. bambusicola , and considered R. hui and R. batangensis to be junior synonyms. Subsequently, Liu & Hu (1961) considered R. bambusicola (including R. hui and R. batangensis ) to be a junior synonym of R. dugritei . Two samples of R. dugritei from the type locality of R. bambusicola (Mount Wa, Sichuan Province, China) do not cluster together, but rather are rooted as a polytomy with R. dugritei from the type locality (Baoxing County, Sichuan, China) and Hongya, Sichuan ( Fig. 2 ). Uncorrected p-distances among five samples of R. dugritei and R. bambusicola range from 0.05 to 0.37%, and these values are typical of intraspecific variation. The extent of variation is far lower than the mean value of 7.53% for species in this group. The extent of divergence should not be used to define species, yet these low values and the absence of morphological characters suggest that gene flow occurs between the localities. Therefore, R. bambusicola appears to be a junior synonym of R. dugritei , as first surmised by Liu & Hu (1961).

Rhacophorus hui is not recognized by Fei (1999) and Fei et al. (2005). Li et al. (2006) resurrect R. hui from the synonymy of R. dugritei . Recently, Yu et al. (2009) also recognized R. hui . The validity of the species remained uncertain because recent molecular analyses did not include samples of R. dugritei from the type locality ( Yu et al., 2009). Indeed, the sample of R. dugritei used by Yu et al. (2009) is geographically closer to the type locality of R. puerensis (Pu’er, Yunnan; formerly Simao, Yunnan) than it is to that of R. dugritei . We resolve R. dugritei from the type locality as sister to R. hui from its type locality. Karyotypically, a secondary constriction and Ag nuclear organizing regions (NORs) distinctly separate R. dugritei and R. hui . These transformations indicate a possible paracentric inversion in the long arm of chromosome 10 in one of the two species ( Wu & Zeng, 1994; Li et al., 2008). Morphologically, the snout coloration of R. hui is yellowish brown, whereas that of R. dugritei is green. Although the uncorrected p-distance between R. dugritei and R. hui is low (0.42–0.63%), in the interest of nomenclatorial stability we provisionally retain both as valid species pending additional evidence.

Li et al. (2006) consider R. zhaojuensis to be a junior synonym of R. hui . Our results indicate that R. hui and R. zhaojuensis from their type localities constituted a single, poorly differentiated lineage. Furthermore, the straight distance between the type locality of R. hui (Yan-wo-tang, Zhaojue County) and R. zhaojuensis (Qi-li-ba, Zhaojue County) is less than 4.5 km ( Li et al., 2006). Our molecular results supported the synonymization of R. zhaojuensis into R. hui ( Li et al., 2006) .

Lineage B

Rhacophorus hungfuensis is the sister of R. wui sp. nov. (previously known as R. cf. hui ). Together they form a sublineage with R. minimus . In turn, this sublineage is the sister group of R. hongchibaensissp . nov. (previously known as R. cf. dugritei ). In order to recognize the monophyly of R. dugritei and R. hui , the description of two new species is required.

Lineage C

This group includes the following units: R. dugritei of Yu et al. (2009); R. dugritei from Longling and Tengchong, Yunnan, and Vietnam; R. hungfuensis of Orlov et al. (2001); and R. puerensis from its type locality. Fei et al. (2005, 2009, 2010) and Yang & Rao (2008) consider R. puerensis to be conspecific with R. dugritei . Because lineage A is more closely related to lineage B than lineage C, we consider R. puerensis to be a valid species, which is consistent with Zhao et al. (2000). Furthermore, uncorrected p-distances within lineage C range from 1.69 to 2.78%. It appears that R. dugritei of Yu et al. (2009), R. dugritei from Longling and Tengchong, Yunnan, and Vietnam, and R. hungfuensis in Orlov et al. (2001) should be referred to R. puerensis . Orlov et al. (2001) and Bain & Nguyen (2004) consider that R. dugritei from the northern mountains of Vietnam might be conspecific with R. puerensis , and our analysis supports this suggestion.

Figures

Figure 2. The 50% majority rule consensus tree from a Bayesian inference analysis of partial fragments of mitochondrial (12S rRNA, valine tRNA, and 16S rRNA) and nuclear (tyrosinase and rhodopsin) genes. Black circles indicate nodes supported by Bayesian posterior probabilities (BPP) ± 0.90, and maximum likelihood bootstrap proportions (MLBS) and maximum parsimony bootstrap proportions (MPBS) ± 70%. Grey circles indicate poorly supported nodes (BPP <0.90; MLBS and MPBS <70%).