Bioacoustics of Raorchestes kempiae

Comments on taxonomic status. Boulenger (1919) described Raorchestes kempiae based on a single specimen collected by A. Kemp from “ above Tura ”, Garo Hills. The type locality refers to the ridge / peak immediately above Tura (see Paiva 1919 ). It is possible that this specimen was collected in 1917, which is evident from Paiva (1919) and Brunetti (1918) that A. Kemp was part of their expedition. However, there was no allocated voucher number available for the type in the original description. Chanda (1994) mentioned a possible location of the type specimen as the “ British Museum, (Natural History), London, U. K. ” which was followed by Bossuyt and Dubois (2001) and repeated in Frost (2025) . Chanda et al. (2000) did not provide the presence of the type specimen of “ Ixalus kempiae ” or “ Philautus kempiae ” in the collection of the National Zoological Collection of Zoological Survey of India ( ZSI ), Kolkata (Calcutta). Recently Naveen et al. (2024) provided a photograph of a specimen in ZSI as the “ holotype ” ( ZSI 18859 ) of R. kempiae . As there was no published literature that mentions the presence of a holotype with the voucher number “ ZSI 18859 ” prior to Naveen et al. (2024) , we were unaware of this specimen during our examination of type specimens of Raorchestes / Philautus housed in ZSI . It was lost and found recently in the ZSI collections ( https://zsicollections.in/specimen/ZSI0000002582 ). We later examined this specimen. On 27 May 2022 , we encountered a few individuals of bush frogs (Fig. 21 ) in Tura Peak ( 25.502825°N ; 90.23921°E ; elevation 1060 m a. s. l.), West Garo Hills District of Meghalaya syntopic with R. garo but having a distinctive call pattern. These individuals resemble the original description of Boulenger (1919) by the following morphological characters: 1) dark bar on interorbital space, 2) crossbars on limbs, vertical bars on upper jaw, tubercles on dorsal surface of head and dorsum, 3) barely visible tympanum (although mentioned as “ tympanum hidden ” in original description), 4) indistinct canthus rostralis, and 5) brown marbling on throat and abdomen. Following exceptions were observed in the newly collected materials: 1) snout shape rounded or sub-oval in dorsal view (rounded in holotype ), 2) nostril equidistant between eye and snout tip or closer to snout tip than eye (nostril equidistant between eye and snout tip in holotype ), 3) concave stripes on dorsum diffused or not visible. For taxonomic stability, we provided an expanded description and determined the phylogenetic position based on specimens collected from the type locality. Thus, based on morphological similarity with the original description of R. kempiae , we refer to this population from “ above Tura ” as topotypic material of the species.

Remarks. Naveen et al. (2024) provided a phylogenetic status of “ R. kempiae ” based on a specimen ( SACON VA 806 ) collected from Mikadogre Community Reserve ( 25.433660°N , 90.398981°E ,), South Garo Hills District, Meghalaya at an elevation of 174 m . This collection locality is approximately 18 km southeast of the original type locality of “ above Tura ” which is located at an elevation of 1000–1200 m a. s. l (see Paiva 1919 ). We argue that Naveen et al. (2024) overlooked the original description by Boulenger (1919) and did not provide a justification for their treatment of the material from Mikadogre as topotypes of R. kempiae . In our phylogenetic analyses, the “ R. kempiae ” of Naveen et al. (2024) did not cluster with the topotypical R. kempiae collected from Tura peak, instead it is nested within the topotypical samples of R. garo (Figs 2 , 3 ) we collected from the same locality. Thus, our robust phylogenetic analyses suggest that the “ R. kempiae ” of Naveen et al. (2024) should be referred as R. garo . The photograph of an uncollected male with white bar on head provided on page 366 of Naveen et al. (2024) as “ R. kempiae ” falls within the morphological variation in R. garo across its range (see morphological variation section under R. garo and Figs 16 , 17 ). In the present study, we found that the topotypical morph of R. kempiae corresponds to the original publication of Boulenger (1919) in having a blackish bar on the inter-upper eyelid space which is not found in R. garo (see variation in R. garo , Figs 16 , 17 ). Although the type locality of R. kempiae and R. garo are the same and the two species are found in sympatry, they can be clearly distinguished based on their acoustic characteristics (Fig. 18 ). Furthermore, in our PCA the type specimens of R. garo and R. kempiae clustered with its respective topotypes collected in this study (Fig. 6 E ). Mistaken identity of R. kempiae by Naveen et al. (2024) was subsequently repeated in Naveen et al. (in press ) and Warjri et al. (2025) . In Naveen et al. (in press ) another “ topotype ” of R. kempiae was reported to be collected near Tura ( 25.522648°N , 90.185480°E ) at an elevation of 210 m a. s. l. Interestingly, Naveen et al. (in press ) did not include their previous sequences of “ R. kempiae ” provided in Naveen et al. (2024) without any explanation. Additionally, Naveen et al. (in press ) provided three specimens ( RSNFM 13–15) under a single GenBank accession number ( PV 061647 ) in the phylogenetic figure (“ Figure 1 ”) and “ Table 1 ”. Furthermore, the locations of these specimens in the text and their reference in the phylogenetic tree do not correspond. However, in our phylogenetic analyses all these sequences of the claimed “ topotype ” of R. kempiae and sequences from Mizoram previously referred as “ R. cangyuanensis ” were clustered with topotype of R. garo sensu stricto . Naveen et al. (in press ) synonymised R. manipurensis based on specimens ( MZMU 2179 and MZMU 2029 ) collected from Ralruwng (Larong) ca. 50 km southeast of the actual type locality, which they claimed to be the topotype . Similarly, they synonymised “ P. namdaphaensis ” based on a specimen ( V/ APRC /A-317 ) collected from Gibbons Land, Namdapha Tiger Reserve which is located on the southern slope of the Noa-Dihing river. Another specimen ( V/ APRC /A-563 ) collected from Magoring village in the Itanagar Wildlife Sanctuary in Arunachal Pradesh on the northern slope of the Brahmaputra River referred to as “ near topotype ” of “ P. namdaphaensis ” without any justification or molecular evidence. In contrast, Naveen et al. (in press ) used “ P. namdaphaensis ” sample ( ON 493540 ) available in GenBank with the voucher number MZMU-V 1277 but did not use either of the two specimens in their phylogenetic analysis. However, in our study, sequences of “ R. manipurensis ” and “ P. namdaphaensis ” of Naveen et al. (in press ) and actual topotype of these two species collected in this study were clustered with topotype of R. garo sensu stricto . Thus, we suggest “ R. manipurensis ” and “ P. namdaphaensis ” are junior synonym of R. garo and not a synonym of R. kempiae . Naveen et al. (2024) described a new species, “ Raorchestes asakgrensis ” from Eman Asakgre Community Reserve ( 25.36788°N ; 90.54344°E ; 174 m ) of Meghalaya State . In our phylogenetic analyses, the specimen of “ R. asakgrensis ” ( SACON VA 805 ) is nested with R. kempiae (but not the “ R. kempiae ” of Naveen et al. (2024) ) (Figs 2 , 3 ). The genetic divergence of “ R. asakgrensis ” and specimens of R. kempiae collected from Tura Peak is 0.7–1.7 % in the 16 S gene (Table S 7 A – C). Thus, we consider “ R. asakgrensis ” as a junior subjective synonym of R. kempiae . We consider this taxonomic chaos created by Naveen et al. (2024) is an artefact of i) sampling gaps involving topotypic material of R. garo and R. kempiae , ii) not assigning materials according to the original morphological description of Boulenger (1919) , and iii) providing single DNA sequence data each from the materials collected far away from the original type locality.

Acoustics. The calls of R. kempiae were recorded from three localities: Tura Peak on 27 May 2022 at an ambient temperature of 28.3 ° C; in the Forest Rest House Campus, Tura on 24 May 2022 at an ambient temperature of 28.0 ° C and in Daribokgre on 25 May, 2022 at an ambient temperature of 28.0 ° C. All calls were recorded between 18: 30 hrs and 19: 30 hrs. the description of the calls is based on analyses of 40 calls from three different individuals ( WII-ADA 1464 , WII-ADA 1476 and WII-ADA 1497 ). The calls of R. kempiae are single type, pulsatile and emitted mostly at regular intervals and not in groups (Fig. 18 ). The mean call duration is 149.29 ± 43.92 ms (67–200 ms) with 3.43 ± 0.59 pulses per call (2–4 pulses per call) and at a pulse rate of 19.21 ± 2.92 pulses / sec (18.52–21.74 pulses / sec). The call rise time is 1 ms and the fall time is 146 ± 46.88 ms (64–199 ms). The pulses are not closely packed; the mean pulse duration is 21.52 ± 8.09 ms (9–37 ms) and the mean pulse period is 54.55 ± 5.06 ms (45–63 ms). The mean inter-call interval is 833.97 ± 298.61 ms (398–1695 ms). The mean dominant frequency is 3306.41 ± 101.26 Hz (3057.7–3488.4 Hz). A detailed comparison of acoustic characters with that of its congeners is given in Table 2.

Distribution and natural history. From our current sampling it appears that R. kempiae has a patchy distribution encompassing Tura peak, Nokrek Biosphere Reserve in Garo Hills and Nongkhyllem Wildlife Sanctuary of Meghalaya State and Athibung in Nagaland State within an elevation range of 230–1140 m a. s. l. (Fig. 19 B ). Mathew and Sen (2009) reported this species from Sohra, East Khasi Hills, Meghalaya . Naveen et al. (2024) reported this species from Nongkhyllem as “ R. asakgrensis ”. However, we recover the population from Nongkhyllem Wildlife Sanctuary as R. kempiae sensu stricto (see remark below). We recorded R. kempiae on shrubs at a perch height of approximately one metre, inside semi-evergreen forest as well as near settlements (Fig. 20 A, B ). We observed calling individuals of R. kempiae and R. garo in the same habitat type at Tura peak and Daribokgre.