Bioacoustics of Raorchestes garo

Acoustics. The calls of R. garo were recorded at Tura Peak, Meghalaya on 27 May 2022 at an ambient temperature of 28.2 ° C and at Lakhicherra, Assam on 18 April 2019 at an ambient temperature of 28.8 ° C between 18: 30 hrs and 19: 30 hrs. The call description is based on 42 calls from two individuals ( WII-ADA 1495 and WII-ADA 601 ). The calls are single type, non-pulsatile (Fig. 18 ), initially emitted as a single call without a group, and gradually emitted in groups (4.72 ± 2.22 calls per call group; 2–13 calls / call group). The mean call duration is 28.57 ± 2.74 ms (20–32 ms) with a call rise time of 1 ms and call fall time of 27.55 ± 2.71 ms (19–31 ms). The interval between calls is 154.31 ± 7.02 ms (144–170 ms) when emitted in groups. The mean dominant frequency of the call is 2860.84 ± 49.55 Hz (2799.3–2971.6 Hz). A detailed comparison of the advertisement calls with those of other congeners is presented in Table 2.

Remarks. Naveen et al. (2024) reported rediscovery of “ R. garo ” based on five specimens ( SACON VA 809 , 163, 305, 306, 308) collected from an unnamed locality with geographical coordinate given as “ 25.51°N , 90.38°E , 895 m ” in Garo Hills, Meghalaya and one specimen ( SACON VA 129 ) from Daribokgre Community Reserve ( 25.47297°N , 90.3148°E , 1200 m ), that they claimed as a “ topotype ”. However, the 16 S gene sequence data in Naveen et al. (2024) comes from a specimen ( SACON VA 809 ) collected approximately 15 km east of Tura. Morphological description of the alleged “ topotype ” ( SACON VA 809 ) do not match the original description in terms of dorsal colour pattern. The original description of R. garo mentions an hourglass pattern on dorsum, but this was absent in SACON VA 809 . Although Naveen et al. (2024) mentioned that two of the specimens such as VA 305 and VA 308 have hourglass patterns, they did not generate molecular data for these two specimens . The holotype ( ZSIC 19187 ) of R. garo was originally collected by S. W. Kemp from “ above Tura ” which is the ridge / peak immediately above Tura (see Paiva 1919 ). We therefore argue that without justification, Naveen et al. (2024) considered SACON VA 809 as a topotype of R. garo , and provided a misleading phylogenetic status of R. garo . The DNA sequence ( 16 S ) of the “ R. garo ” topotypical specimen ( SACON VA 809 ) submitted by Naveen et al. (2024) did not cluster with our topotypical samples of R. garo ( WII-ADA 1493 , WII-ADA 1494 , WII-ADA 1499 ) collected from Tura Peak (= above Tura) ( 25.50302°N ; 90.23853°E ; elevation 1030 m a. s. l.; Figs 2 , 3 ). The genetic divergence of the true topotypical samples of R. garo and the specimen of “ R. garo ” of Naveen et al. (2024) is 6.5–7.0 % in 16 S gene (Table S 7 A – C). Naveen et al. (2024) stated that R. garo is phylogenetically allied to R. shillongensis . However, our phylogenetic analyses indicate R. shillongensis ( topotype ; WII-ADA 1460 ) as distinct from R. garo ( topotype ; WII-ADA 1493 , WII-ADA 1494 , WII-ADA 1499 ; Figs 2 , 3 ). Our phylogenetic analyses are strongly supported by acoustic characteristics of the two species (see under respective description and comparison). This suggests that SACON VA 809 reported by Naveen et al. (2024) is a misidentification of R. garo and it represents an undescribed lineage of Raorchestes . Thus, we consider SACON VA 809 as an invalid topotype of R. garo based on accurate type locality and the phylogenetic results of this study. “ Philautus namdaphaensis ” was originally described by Sarkar and Sanyal (1985) from Farmbase Camp (at 350 m a. s. l.), Namdapha National Park, Arunachal Pradesh based on three adult males collected by S. Biswas. Later, Sengupta et al. (2000) reported this species from Garbhanga Reserve Forest, Assam , and Mathew and Sen (2009) , and Sen et al. (2013) reported it from Nokrek Biosphere Reserve, Meghalaya . On 9 May 2023 we collected two specimens of bush frogs from Haldibari ( 27.52453°N ; 96.39913°E ; elevation 500 m a. s. l.), Namdapha Tiger Reserve, Arunachal Pradesh (Fig. 15 U ) on the north bank of the Noa-Dihing River. These two specimens conform to the original description of “ Philautus namdaphaensis ” based on the following characters: 1) mid dorsal line from snout to vent, 2) limbs with crossbars and a line originating at vent reaching to heel through dorsal aspect of thigh and tibia, 3) white bar on interorbital space, 4) dark brown blotches on groin and lateral aspects of thighs. “ Raorchestes manipurensis ” was described by Mathew and Sen (2009) from Leimatak, Manipur based on a single specimen. So far, this species is only known from the type locality. On 25 July 2022 we collected a topotypical specimen ( WII-ADA 1645 ; Fig. 17 H ) from Leimatak ( 24.59375°N ; 93.66256°E ; elevation 480 m a. sl.) that conforms to the original description. This individual matches the original description by the following set of characters: 1) head wider than length, 2) distinct canthus rostralis, 3) vomerine teeth absent, 4) a faint hourglass shaped patch covering head from inter-upper eyelid space and back (dark in holotype ), posterior ends of this patch directing towards groin; 5) brown and pale white patches on flank near groin. Our phylogenetic analyses involving topotypical samples of “ P. namdaphaensis ” and “ R. manipurensis ” resulted with them nested with the topotypical samples of R. garo (Figs 2 , 3 ). The topotypical samples of R. garo have genetic divergences of 0.0–0.6 % with topotypical samples of “ P. namdaphaensis ” and 0.4–1.1 % with the topotypical sample of “ R. manipurensis ” in the 16 S gene (Table S 7 A – C). Morphological examinations of the topotypes of the three species ( R. garo , “ P. namdaphaensis ” and “ R. manipurensis ”) and examinations of the type series of “ P. namdaphaensis ” did not reveal any strong differentiating characters to separate these three species. Morphological and molecular studies of the newly collected material from different localities in northeast India suggests R. garo is a polymorphic species (see under redescription section above). We have recorded individuals of R. garo from Tura peak that resemble the typical morphotype of “ P. namdaphaensis ” (Fig. 16 A – H ). Mathew and Sen (2008) also reported “ P. namdaphaensis ” from Nokrek Biosphere Reserve, Meghalaya probably based on this character. The colour and pattern of “ R. manipurensis ” provided in the original description and the photograph of the holotype by Mathew and Sen (2009) resemble the typical morphotype of R. garo from the type locality. Therefore, these are unreliable and cannot be used as diagnostic characters. Based on molecular data and morphological characters, we propose to treat “ Philautus namdaphaensis ” and “ R. manipurensis ” as junior subjective synonyms of R. garo . “ Raorchestes cangyuanensis ” was described from Cangyuan, China ( Wu et al. 2019 ). Subsequently, Lalronunga et al. (2021) reported this species from Hmuifang, Mizoram and referred to a previous record of R. longchuanensis from Satchari National Park, Bangladesh ( Al-Razi et al. 2020 a ) as “ R. cangyuanensis ”. Rahman et al. (2022) reported the species from northeast Bangladesh . Based on the results of their phylogenetic study, Naveen et al. (2024) suggested “ R. cangyuanensis ” as a junior subjective synonym of R. kempiae . However, in our current phylogenetic analyses, “ R. cangyuanensis ” is nested with the topotypical samples of R. garo (Fig. 3 ) collected in this study (see comments under R. kempiae ). The genetic divergence between R. garo collected for this study and “ R. cangyuanensis ” samples is 0.0–0.9 % in the 16 S gene (Table S 7 A – C). Thus, we consider “ R. cangyuanensis ” as a junior subjective synonym of R. garo and state that the previous records of “ R. cangyuanensis ” from Mizoram should be considered as R. garo .

Distribution and natural history. This species is distributed in northeast India from western part of Meghalaya to the eastern part of Arunachal Pradesh in the Namdapha Tiger Reserve. This species is also distributed in the Barail Hills of Assam and Nagaland , in Manipur , and in Mizoram . It is mostly associated with lowland evergreen forests (Fig. 19 A ). This species was so far known from 50–1480 m a. s. l. across hills on the southern slopes of Brahmaputra Valley, India . Al-Razi et al. (2020 a ) reported this species as R. longchuanensis from Satchari National Park in Bangladesh . Wu et al. (2019) reported this species as R. cangyuanensis from Cangyuan, China . Lalronunga et al. (2021) reported from Mizoram as R. cangyuanensis . Choudhury et al. (2001) reported this species from Garbhanga Reserve Forest and Kulsi Reserve Forest in Assam, and Ao et al. (2003) reported it from Dzulake, Nagaland which needs further examination to confirm the records. Calling males were recorded during April – June from perches 1–2 metres above the ground and from roadside forest areas dominated by bamboo thickets (Fig. 20 A, B ). We also recorded the species calling from the leaves of a small tree about five metres above the ground and from a bamboo thicket at a height of six metres above the ground in Charoikhullen, Manipur . We observed a single female individual of this species on roadside vegetation in Teirei, Dampa Tiger Reserve, Mizoram in September 2021 .