SYSTEMATICS OF
RHACOPHORUS
View in CoL
Kou, Hu & Gao (2001) described ‘
Polypedates
’
pingbianensis
(
R. pingbianensis: sensu Fei et al., 2005
) as a valid species, but Fei et al. (2005) (without comment) did not recognize it as such. According to Kou et al. (2001),
R. pingbianensis
is close to
R. omeimontis
morphologically, and the difference between them is that the former has no vocal sacs or lineae masculinae, whereas the latter has internal subgular vocal sacs and lineae masculinae. In the present study,
R. pingbianensis
and
R. omeimontis
form a monophyletic group with high support values from all analyses. Furthermore, ML and Bayesian trees demonstrate that
R. pingbianensis
is the sister taxon to
R. omeimontis
, albeit that the support values are not strong, and so it is treated here as a valid species related to
R. omeimontis
.
The distinction between
Rhacophorus
and ‘
Polypedates
’ is under debate (e.g. Liem, 1970; Dubois, 1987; Jiang et al., 1987; Fei, 1999; Fei et al., 2005). Wilkinson & Drewes (2000) suggested that a green dorsal pattern might be a character to distinguish between the two genera. In the present study, all phylogenetic analyses support a monophyletic group, in which most species of Chinese
Rhacophorus
with a green dorsal pattern are included, with the exceptions of
Rhacophorus reinwardtii (Schlegel, 1840)
and
Rhacophorus bipunctatus Ahl, 1927
(clade IV). Furthermore,
R. reinwardtii
is closely related to the group consisting of
R. bipunctatus
and
R. rhodopus
(see clade V). This suggests that the green dorsal pattern has arisen at least three times in the taxa examined here, and cannot be used to characterize a monophyletic
Rhacophorus
.
Rhacophorus reinwardtii
,
R. rhodopus
, and
R. bipunctatus
belong to the
R. reinwardtii
group of Fei (1999). Wilkinson et al. (2002) found that
R. reinwardtii
is the sister taxon of
R. bipunctatus
. However, our results indicate that
R. reinwardtii
is close to the clade containing
R. rhodopus
and
R. bipunctatus
, although the support from MP analysis is marginal. Frost (2007) united
R. rhodopus
into
R. bipunctatus
, but without discussion. According to Fei (1999),
R. bipunctatus
has a green dorsum and two black spots on the axilla, whereas the dorsum of
R. rhodopus
is red–brown, and there is only one black spot on the axilla. Furthermore, in China,
R. bipunctatus
is found only in Xizang (Tibet), whereas
R. rhodopus
has a wide distribution including Xizang (Tibet), Yunnan, Guangxi, and Hainan Provinces. In the present study, the monophyly of
R. rhodopus
is not supported.
Rhacophorus rhodopus
obviously can be divided into three lineages, and the lineage from Hainan Island is the sister group of
R. bipunctatus
(see clade V). These results indicate that the genetic structure of
R. rhodopus
is complicated; Frost’s (2007) placement of
R. rhodopus
into the synonymy of
R. bipunctatus
is not accepted here because the distinctive lineages within
R. rhodopus
may represent cryptic species, just one of which, such as the Hainan Island population, is close to
R. bipunctatus
. Unfortunately, no samples were available from Guangxi Province, or from adjacent countries such as Vietnam and Burma, so more studies will be needed to unveil the general phylogenetic structure within
R. rhodopus
.
Rhacophorus dugritei
,
R. nigropunctatus
, and
R. chenfui
belong to the
R. dugritei
group ( Fei et al., 2005). However, there is a monophyletic group consisting of
R. dugritei
,
R. minimus
, and
R. moltrechti
with strong Bayesian posterior probability, albeit that the bootstrap values of MP and ML trees for it are weak (see clade IV). So the
R. dugritei
group of Fei et al. (2005) needs re-evaluation. Additionally, five specimens from Wenshan County, Yunnan Province, form an independent clade, which is related to the group consisting of
R. reinwardtii
,
R. rhodopus
, and
R. bipunctatus
in all analyses, although the support values are not strong (see clade V). This lineage might represent an undescribed species of
Rhacophorus
View in CoL
.
Using single samples of each species, Wilkinson et al. (2002) found that
R. megacephalus
and
R. leucomystax
, which were placed into the genus ‘
Polypedates
View in CoL
’ by some herpetologists (e.g. Matsui, Seto & Utsunomiya, 1986; Wilkinson et al., 2002; Frost, 2007), are sister taxa.
Rhacophorus megacephalu
s was once included in
R. leucomystax
as a subspecies (e.g. Stejneger, 1925; Dubois, 1987). Matsui et al. (1986) resurrected
R. megacephalus
from the synonymy of
R. leucomystax
based on differences of morphology, voice, and karyotype between populations of Taiwan Island and Kalimantan. However, Rao & Yang (1996) found that the karyotypes of the
R. megacephalus
group are stable, and suggested that only
R. megacephalus
can be regarded as a subspecies of
R. leucomystax
. In the present study, a monophyletic group consisting of
R. megacephalus
and
R. leucomystax
is supported by all analyses, and there are at least three distinct clades in this
leucomystax
/
megacephalus
complex, although
R. leucomystax
is nested in
R. megacephalus
(see clade VI). Moreover, using the fragment of the 16S gene without excluding hypervariable regions, the P distances between these clades as shown in the MP tree are 4.3–6.4%, which is obviously higher than the P distances within these clades (0.4–1.6%). These indicate the inclusion of various species and the necessity for a revision of the
leucomystax
/
megacephalus
complex.
