Wijayarana, Arifin & Chan & Smart & Hertwig & Smith & Iskandar & Haas, 2021

Arifin, Umilaela, Chan, Kin Onn, Smart, Utpal, Hertwig, Stefan T., Smith, Eric N., Iskandar, Djoko T. & Haas, Alexander, 2021, Revisiting the phylogenetic predicament of the genus Huia (Amphibia: Ranidae) using molecular data and tadpole morphology, Zoological Journal of the Linnean Society 193, pp. 673-699 : 687-690

GENUS WIJAYARANA GEN. NOV.

( FIG. 6B–F)

Z o o b a n k r e g i s t r a t i o n: l s i d:z o o b a n k.o r g:a c t: 4A7EC577-93DC-4A0D-851D-82B277E060E3

Type species: Huia sumatrana Yang, 1991 , Fieldiana Zoology 63: 31. Holotype: FMNH 209222 , an adult male, by original designation.

Type locality: Bukit Lawang Forest Reserve , Bohorok, Sumatra (3°31′N, 98°8′E), Indonesia. Collected by Harold K. Voris, 22 July 1976 .

Etymology: Wijayarana is a compound of the words ‘wijaya’ ( Vijaya , in Sanskrit, meaning victory) and ‘rana’ (Latin for frog, feminine). In this context, Wijaya alludes to the Sriwijaya (or Sri Vijaya ) empire based in Palembang (Sumatra) between the 7 th and 14 th centuries CE, with primarily maritime realms (i.e. a thalassocracy). Sriwijaya became one of the most powerful and expansive kingdoms of Indonesia, spanning across most of Sumatra, Java and the Malay Peninsula. With the exception of the Malay Peninsula, these landmasses also represent the geographical range of the genus.

Common name: Wijaya cascade frogs (English); Kongkang Jeram Wijaya (Bahasa Indonesia).

Included species: Wijayarana javana ( Yang, 1991) comb. nov., Wijayarana masonii (Boulenger, 1884) comb. nov., Wijayarana melasma (Stuart & Chanard, 2005) comb. nov., Wijayarana modigliani (Doria, Salvidio, & Tavano, 1999) comb. nov. and Wijayarana sumatrana ( Yang, 1991) , comb. nov.

Phylogenetic definition and content: Wijayarana is a crown clade consisting of lineages from Indonesia (Sumatra and Java) and, provisionally, Thailand and Laos (see Remarks). Species previously assigned to H. masonii , H. sumatrana and H. javana (resurrected herein) are transferred to the new genus Wijayarana . The phylogenetic position of frogs previously recognized as H. melasma is still uncertain, but based on the results from this study we assign it provisionally to the genus Wijayarana . We also tentatively assign frogs previously known as H. modigliani to the genus Wijayarana . However, owing to ambiguous diagnostic characters and the absence of molecular samples from the type locality, the validity of this species is still in question (for more details, see Remarks).

Diagnosis: Larvae gastromyzophorous; upper beak M-shaped, lower beak V-shaped, both undivided; LTRF 6–9((3–5)–(6–9))/6–10(1); an oblique cluster of glands immediately below the spiracular tube, running upward and backward absent; infraorbital, postorbital, prespiracular and posterior mid-lateral glands present; abdominal glands absent, except for few individuals; glands on the tail fin present. Adults medium in size, body slender; snout pointed; tympanum not depressed, clear, framed by dark Π- shaped marking; supra-, post-tympanic and dorsolateral folds present; ratio of tibia length to body length> 70%; outer metatarsal tubercle present or absent; inner metatarsal tubercle present; Finger I ≥ Finger II; length of metacarpal of Finger I subequal to Finger II (see Supporting Information, Fig. S4); width of crossbar terminal phalanx of Finger III <50% length (Supporting Information, Fig. S4); male with nuptial pad and paired lateral vocal sacs; humeral gland absent. In life, head and dorsal body generally brown, with or without scattered white markings; lips with dark markings.

Geographical range: Wijayarana has been recorded from Java and Sumatra, Indonesia. This genus might also (provisionally) occur in Thailand and Laos,

depending on the status of W. melasma (see remarks under that species).

Remarks: Yang (1991) suggested that the length of the metacarpal of Finger I for W. javana is 61% of Finger II and for W. sumatrana the length of the metacarpal of Finger I is 69% of Finger II. However, according to our findings, the length of the metacarpal of of Finger I for Wijayarana is subequal to Finger II (see Supporting Information, Fig. S4).

Yang (1991) described W. javana (as Huia javana Yang, 1991 ) without mentioning W. masonii ( Rana masonii Boulenger, 1884 ) in his description. Iskandar (1998) synonymized W. javana with W. masonii based on morphological similarities between both species. Subsequently, Manthey & Denzer (2014) pointed out that W. javana differs from W. masonii by the presence of a small outer metatarsal tubercle (vs. absent) and relative length of Finger I = Finger II (vs. Finger I> Finger II). Thus, Manthey & Denzer (2014) proposed that both W. masonii and W. javana should be considered preliminarily as valid species. This study supports the hypothesis of Manthey & Denzer (2014) by providing the following evidence: (1) sample MZB. Amph.22371 was collected from Cibodas, the type locality of W. javana (clade F) and was inferred as a distinct and divergent lineage; and (2) morphological examination of adult specimens of clade F (N = 7) matched the original description of W. javana . Thus, we support the proposal by Manthey & Denzer (2014) to consider W. javana as a valid species. Furthermore, we are confident that clade E does not represent W. masonii , because an outer metatarsal tubercle was observed in the specimens of this clade (N = 10). It is also important to note that clade E is more closely related to the true W. sumatrana (clade C–D) than to W. javana (clade F). Thus, clade E might constitute a new, undescribed species ( Huia sp. 3 ) that is conspecific with neither W. javana nor W. masonii . These results also indicate that the true W. masonii has yet to be sampled, and its identity remains dubious. Additionally, given that we resurrected W. javana in this study, several localities that were mentioned in previous studies (e.g. Iskandar, 1998; Kurniati, 2003; Kusrini, 2013) as part of the geographical distribution of W. masonii should be re-evaluated.

We assign W. sumatrana as the type species of Wijayarana because: (1) the identity of W. masonii is still unclear, although it is the oldest available name within species recognized under Wijayarana ; (2) W. sumatrana is confirmed based on the abovementioned evidence; (3) W. sumatrana was described at the same time as W. javana , but given that the presence of another species in Java remains unclear, we prefer not to select any recognized species from this island as the type species for the genus.

We tentatively place the species modigliani in Wijayarana solely based on its distribution in Sumatra. Unfortunately, no tissue samples are available to corroborate this. Manthey & Denzer (2014) pointed out several ambiguous diagnostic characters for W. modigliani that were originally proposed by Doria et al. (1999), which lead to confusion in species identification (for example, by Kurniati, 2009). However, given that Huia is now restricted to Borneo and that all confirmed Wijayarana ( W. sumatrana and W. javana ) and W. masonii (to be confirmed) are restricted to Java and Sumatra, we find it reasonable and most parsimonious to place W. modigliani tentatively within Wijayarana pending new evidence.

We also tentatively place the species melasma in the genus Wijayarana . This species was previously known only from western and northern Thailand (Stuart & Chan-Ard, 2005; Frost, 2019), but has been confirmed by our study also to occur in Laos (FMNH 271377 was collected from Luang Namtha Prov, Vieng Phou Kha District). In both our ML and BI trees (including ML and BI trees after phasing the nuclear DNA alignments; see Fig. 2; Supporting Information Fig. S2), W. melasma is sister to Wijayarana clades from Java and Sumatra. The position of melasma has never been stable in previous studies ( Fig. 1), including ours (e.g. Fig. 2), despite obtaining the longest concatenated dataset from both mitochondrial and nuclear markers, with a low percentage of missing data. However, its genetic divergence from all other closely related lineages was high (> 10.00%) and consistent with divergence observed between other genera. Wijayarana melasma is morphologically more similar to other Wijayarana species than to H. cavitympanum ; for example, its body is slender (vs. body stout in Huia ), the snout is pointed (vs. rounded in Huia ) and the tympanum is not depressed (vs. deep cavity in Huia ). On the contrary, we also notice several morphological characters of adult W. melasma (Stuart & Chan-ard, 2005) that differ from Wijayarana , e.g. the pineal body is not visible (vs. visible in Wijayarana ), and the tympanum not transparent and not encased by a dark Π- shaped marking (vs. clear tympanum and encased by Π- shaped marking in Wijayarana ); see Table 3. More importantly, tadpoles of W. melasma have never been collected; hence, it remains to be determined whether this species even has gastromyzophorous tadpoles. As such, the taxonomic and phylogenetic placement of W. melasma remains an open question that will be likely to require additional molecular and/or tadpole data from expanded geographical sampling in Thailand / Laos to resolve.

IMPLICATIONS OF THE AMENDMENTS

We are fully aware that our taxonomic proposal to resolve paraphyly of frogs previously assigned as Huia might be deemed controversial, especially considering the low support of nodes for unstable branches representing the sister relationship among the newly erected genus Wijayarana and W. melasma , Huia , Meristogenys , Sumaterana and Clinotarsus .

All available studies, including our study, despite some differences in sampling, converge to similar conclusions, i.e. there is a problem with Huia / Clinotarsus . There is some ambiguity (low nodal support) in the region where species of these taxa connect. Our sequence lengths were much longer than those used by Stuart (2008), Wiens et al. (2009), Chan & Brown (2017) and Arifin et al. (2018) and contained fewer gaps than those of Pyron & Wiens (2011) but still could not resolve the nodes in question. Moreover, even after phasing the alignments of the two nuclear loci in this study for the abovementioned problematic nodes, the nodes still had low support (see Fig. 2; Supporting Information Figs S1, S 2). Also, according to the SH test, the ML tree topology presented in this study is accepted to represent the phylogenetic relationships among Wijayarana in Java and Sumatra, W. melasma , Huia , Sumaterana , Clinotarsus and Meristogenys . Thus, we believe that these poorly supported nodes are probably attributable to rapid radiation at the time of origin rather than data deficiency ( Whitfield & Lockhart, 2007; Whitfield & Kjer, 2008). If we are correct, the effort to solve this problem probably requires highthroughput molecular data, which is a venture that is beyond the scope of the present project.

CONCLUSION

This study resolved the long-standing systematic problems of Huia using new genetic sequences and morphological data from tadpoles. We demonstrated the efficacy of wide geographical sampling (especially at type localities) of adults and larvae and the integration of multiple sources of data to inform taxonomic classification. This study also confirmed that ranids with gastromyzophorous tadpoles can be differentiated by the shape of the jaw sheaths.

For Huia s.s., larval morphology was particularly informative and can be an important source of data for species delimitation. This study resolved certain taxonomic issues, but also raised a number of important questions regarding the validity of W. modigliani , the true identity of W. masonii , the phylogenetic and taxonomic placement of W. melasma and the hidden diversity within Wijayarana and Huia , especially in the region of Indonesia, including Kalimantan in Borneo. Another missing piece of evidence is that the tadpoles of W. melasma and W. modigliani remain unknown ( Manthey & Denzer, 2014). More fieldwork in the region is needed to resolve the persisting problems completely and elucidate the full extent of diversity within this unique group of frogs.