Description of a New Species of Amolops Cope,1865 (Amphibia:Ranidae) from Nepal and Nomenclatural Validation of Amolops nepalicus Yang,1991

Janak Raj KHATIWADA,Guocheng SHU,Bin WANG,Tian ZHAO,Feng XIE and Jianping JIANG

CAS Key Laboratory of Mountain Ecological Restoration and Bioresource Utilization &Ecological Restoration and Biodiversity Conservation Key Laboratory of Sichuan Province,Chengdu Institute of Biology,Chinese Academy of Sciences,Chengdu 610041,China

Abstract A new species,Amolops mahabharatensis sp.nov.,is described from Mahabharat range (middle mountain) of Nepal.The new species is a member of the A.marmoratus group and is distinguished from its congeners using morphological and molecular data.The new species is diagnosed by having numerous warts on dorsum and flanks,dorsal body greyish olivebrown/light brown,presence of vomerine teeth series,presence of a pair of subgular vocal sacs in males and tadpole with III:5+5/1+1:II -labial tooth row.In addition,evidenced by the phylogenetic and morphological analyses in this study,Amolops nepalicus is shown to be a valid species.A molecular phylogenetic analysis based on 16S ribosomal gene for six members of the A.marmoratus group suggested that the new species is sister to a clade containing A.nepalicus and A.panhai.The uncorrected genetic divergences between the new species and its closest congeners A.nepalicus,A.panhai and A.marmoratus based on 16S rRNA gene were 10.5%,11.6%,and 8.6% respectively.The new species showed wide distribution in the Mahabharat range of Nepal.

Keywords A.marmoratus group,mahabharat range,molecular phylogeny,taxonomy

1.Introduction

Frogs of the genusAmolops(Cope,1865)are renowned for their extraordinary diversity in fast flowing streams in South and Southeast Asia (Frost,2019).Most species ofAmolopsare habitat specialist (Khatiwada and Haugaasen,2015) and might have relatively small and mostly allopatric distributions in this range.This genus is comprised of 60 currently recognized species(Frost,2019).Amolopsare highly adapted for torrent habitat conditions (Schleich and K?stle,2002);they possess several morphological modifications such as the presence of adhesive disks on the tips of the digits with circummarginal grooves (Bijuet al.,2010) and tadpoles with gastromyzophorus adhesive disk(Altig and McDiarmid,1999).In the recent five years,more than 10 new species ofAmolopshave been recorded (Sunget al.,2016;Jianget al.,2016;Feiet al.,2017;Chanet al.,2018;Yuanet al.,2018;Lyuet al.,2019a,2019b;Phamet al.,2019;Yuet al.,2019).

Previous taxonomic and amphibian distribution studies were primarily based on morphological examination of specimens deposited in the museums (Dubois and Matsui,1983;Nanhoe and Ouboter,1987;Yang,1991;Das,1998;Shah and Tiwari,2004).Three species ofAmolopsare currently recognized in Nepal:A.formosus(Günther,1875),A.marmoratus(Blyth,1855),andA.monticola(Anderson,1871) (Frost,2019;Schleich and K?stle,2002).Apparently,A.himalayanuswas also reported from Nepal based on a tadpole (Kripalani,1961),which was later doubted by Schleich and K?stle (2002).MostAmolopspopulations in the Mahabharat ranges (between 400?2 000 m) of Nepal have been assigned toA.marmoratus(Schleich and K?stle,2002;Shah and Tiwari,2004),without any detailed morphological or molecular comparisons.Amolops marmoratusremained an unresolved and challenging species group for identification because of the high degree of morphological similarities among its members(Bossuyt and Milinkovitch,2000;Deveret al.,2012).For instance,Yang (1991) describedAmolops nepalicusfrom Arun river basin,type locality:“Sabhaya,Kbota”=Sabha Khola,eastern Nepal based on adult males and tadpole morphology.Later on,Dubois(2000;2004) revised the taxonomic status ofA.nepalicusand synonymised it withAmolops marmoratus.Since the recognition on theAmolops marmoratuscomplex has been discussed recently and the trueAmolops marmoratusis only recognized from southeastern Myanmar (Deveret al.,2012;Lyuet al.,2019b),the nomenclatural validation ofA.nepalicusand the taxonomic status on other Nepali populations previously assigned toA.marmoratusneed to be clarified.

Although a number of molecular studies onAmolopshave been published on Southeast Asian species (Jianget al.,2016;Sunget al.,2016;Lyuet al.,2019a),no integrative taxonomic studies have been used to evaluate the taxonomic position ofAmolopsspecies in Nepal.During field surveys in 2014 to 2018 in the eastern,central and western Nepal,we collected a series of adult specimens and tadpoles ofAmolopsfrogs.Detailed morphological and molecular analyses suggested that theAmolopspopulation from middle mountain region of Nepal(Mahabharat range) are genetically and morphologically distinct from previously known species of genus theAmolops.Thus,we provide a description of the Mahabharat range population ofAmolopsas a new species.Besides,this study also found that theAmolopsfrogs collected from Dobhan population was genetically distinct with other members of genusAmolopsand showed morphological similarities withAmolops nepalicusand further discuss the nomenclatural validation ofA.nepalicus.

2.Materials and Methods

Figure 1 Location of the study area showing the sampling localities in Nepal.Purple triangle refers to Sabha Khola-type locality of Amolops nepalicus,red dots refer to sampling locations of Amolops mahabharatensis sp.nov (MRP),blue dot refers to sampling location of Amolops nepalicus (DP) and black dots refer to sampling sites of Amolops formosus (KLP).

2.1.Sampling and specimen collection Field work was carried out from May to September each year from 2014 to 2018 in the eastern,central and western Nepal (Figure 1).Following the methods of (Khatiwadaet al.,2019a),amphibians were surveyed using nocturnal time constrained visual encounter surveys between 19:00 h-00:00 h.This method involved systematic walking at a slow pace,intensively searching for amphibian species by turning over stones,logs,leaf litter,tree branches,shrubs and bushes along the transects using both visual and acoustic aids.Specimens were captured by hand,photographed and were euthanized using 20% Benzocaine gel,fixed in 4%formalin for 24 hours and preserved in 75% ethanol.Tadpoles at different developmental stages were also sampled and were preserved in 70% alcohol.Sex was determined by the observation of secondary sexual characteristics,such as vocal sacs and nuptial pads in males.Females were identified by direct inspection of the enlargement of the coelomic cavity and checked the presence of ovary and eggs by minor dissections.Tissue samples were taken from thigh muscle in adults and small portion of tail muscle of tadpoles and preserved in 95%ethanol for further molecular analysis.A total of 33 individuals(one individual from Lamatar,one from Kimathanka,12 from Dobhan,two from Barahakshetra,10 from Hattibang,three from Pokhara and three individuals from Latinath,Darchula(for details see Molecular methods section) were included in the morphological examination.All morphological measurements were taken from the preserved specimens.Handling and processing of all amphibians were in accordance with the guidelines of the Department of National Park and Wildlife Conservation,Government of Nepal and Animal Care and Use Committee of Chengdu Institute of Biology,Chinese Academy of Sciences.Vouchers were deposited at the Museum collection of Central Department of Zoology,Tribhuvan University,Kathmandu,Nepal.

2.2 Morphological measurements Morphological measurements were carried out using digital calliper (to nearest 0.01 mm).Measurements descriptions of adults and tadpoles are provided in Table 1 as described by (Khatiwadaet al.,2019a).Comparative morphological data of species:A.marmoratuswas taken from Deveret al.(2012) whereasA.nepalicus,A.monticolaandA.formosuswere taken from their original descriptions Yang (1991),Anderson (1871) and Günther (1875)respectively.PreservedAmolopsspecimens were also measured in the herpetological museum of Chengdu Institute of Biology(CIB),Chinese Academy of Sciences,Chengdu,China,Central Department of Zoology,Tribuvan University (CDZ-TU) and the British Museum of Natural History (BMNH),UK.A total of 13Amolopsspecies were examined (Appendix Table S1).

2.3.Molecular methods Total genomic extraction was carried out with thigh muscle preserved in 95% ethanol using the DNeasy Tissue Kit (QIAGEN).The mitochondrial gene 16S ribosomal RNA gene (hereafter 16S) was amplified and subjected to polymerase chain reaction (PCR) with primers as described by Matsuiet al.(2006).The amplified PCR products were purified using Qiagen PCR purification kit and sequences were obtained from an ABI 3100 automated sequencer.All sequences were deposited in GenBank under accession numbers (GenBank accession numbers are available in Table 2).Available nucleotide sequences of 16S gene ofAmolopswere downloaded from NCBI GenBank database and aligned with ClustalW built into BIOEDIT Version 7.1.9 (Thompsonet al.,1997) using the default parameters.Rana dalmatinaandHylarana leptoglosawere selected as the outgroups (Stuartet al.,2010).Alignments were also checked and manually edited,if necessary.Maximum Likelihood (ML) analysis was conducted with the rapid bootstrapping algorithm using the program RAxML v8.00 (Stamatakis,2014) on the CIPRES Science Gateway server v3.2 (Milleret al.,2010).Nodal support for ML was assessed with 1000 rapid bootstrap replicates (BS).The Bayesian inference (BI) analyses and the best-fit substitution model were selected under the Bayesian Information Criterion by the program jModeltest 2.1.4 (Darribaet al.2012).The bestfit substitution model for the 16s dataset was GTR+I+G.BI analyses were conducted in the program MrBayes 3.1.2(Ronquist and Huelsenbeck,2003).Two independent runs were initiated each with four simultaneous Markov Chain Monte Carlo (MCMC) chains for 20 million generations and sampled every 1000 generations.The convergence of chains and burnin period of all runs were examined by plots of log-likelihood scores and low standard deviation of split frequencies.The first 25% generations were discarded as burn-in,and the last remaining trees were used to create a 50% majority-rule consensus tree and estimate Bayesian posterior probabilities(BPP).The graphical viewer Figtree (Rambaut,2007) was used to edit the resulting output of RAxML and MrBayes analyses.Pairwise divergences (uncorrected p-distance) between species on 16s dataset were calculated using MEGA 7 (Kumaret al.,2016).

2.4.Statistical analysis Principal Component Analysis (PCA)was used to show the overall morphological variation between the species.Prior to PCA,all morphometric data were converted into ratio values of SVL (morphological character/SVL＊100) to reduce allometric bias.Statistical analysis was performed in R package vegan (Oksanenet al.,2016) in R-software version 3.3.0(R Development Core Team,2016).

3.Results

Molecular analyses

The aligned dataset of 16S contained 620 bps including 409 variable sites and 111 parsimony informative sites (excluding outgroups).The maximum likelihood (ML) and Bayesian inference (BI) phylogenetic trees constructed from 16S DNA sequences produced identical topologies with strong node support (Figure 2).The phylogenetic analysis suggested that theAmolopsspecimens (including tadpoles from Hattibang) fromMahabharat Range Population (MRP) represented a sister taxon toAmolopsfrom Dobhan Population (DP),and clustered to theA.marmoratusspecies group and was sufficiently resolved with high bootstrap supports (BI=0.99).Whereas,Amolopsfrom Kimathanka and Lamatar Population (KLP) were clustered in theA.monticolagroup and was also supported by high bootstrap support value (BI=1).

Table 1 Morphological characters used and their measurement descriptions for adults and tadpoles.

The uncorrected genetic divergence between MRP and DP was 13%,with MRP and topotypicA.marmoratuswas 12% and MRP andA.panhaiwas 15% (Table 3).Amolopsspecies from DP was sister taxon ofA.panhaiand the uncorrected genetic divergence between species of DP andA.panhaiwas 15%whereas withA.marmoratusit was divergent by 13% (Table 3).KLP formed a distinct lineage in the BI analysis.The genetic distance between KLP with MRP and DP were 13% and 15%respectively.

Figure 2 Bayesian inference (BI) tree based on 16s gene.Values on branches of the tree are Bayesian posterior probabilities (bpp) (above)and Maximum likelihood (ML) (below).

Table 2 Samples used in molecular analysis.

(Continued Table 2)

Morphological analysis

Based on phylogenetic analysis,MRP was the sister taxon to species from DP.The result from PCA based on size-corrected values support the molecular analyses and showed the overall morphological variation between MRP,DP and KLP.For the male population,PCA extracted seven principal component axes with eigenvalues greater than one where the first two component explained 55.4% of the total variation (Table 4).The first two principal component axes radially separated MRP from DP based on hand,finger and toe length (Figure 3) in male population.Species with larger and positive scores on PC1 reflected shorter head length,head width,snout length,inter-orbital width,length of 1st,2nd,3rd and 4th fingers,length of femur,tibia,tarsus and meta-tarsus,length of 1st,2nd,3rd,4th and 5th toes while a negative scores signified larger eye diameter,width of upper eyelid,inter-narial width,tympanum diameter,length of arm and hand.The second PC with negative scores were associated with species having larger head length,internarial width,length of arm,hand and tarsus whereas positive scores were associated with species with shorter morphological traits such as head width,snout length,eye diameter,nostril to eye,width of upper eyelid,inter-orbital width,tympanum diameter,length of 1st,2nd,3rd and 4th fingers,length of femur,tibia and meta-tarsus,length of 1st,2nd,3rd,4th and 5th toes.For females,PCA extracted six principal component axes with eigenvalues greater than one and the first two components explained 56.3% of the total variation (Table 4).The first two principal component axes radially separated the female populations of MRP,DP and KLP based on length forelimbs fingers,tibia,meta-tarsus and toes (Table 4).

Table 4 Taxonomic review Based on phylogenetic and morphological results,Amolopspopulations from Nepal (MRP,DP and KLP) represented three distinct species.The results confirmed that there is substantial genetic divergence between the specimens from Mahabharat Range Population (MRP) and other knownAmolopsspecies,and support that this population is a distinctly evolving lineage,representing an undescribed species.This population also presents distinct morphological characteristics that are not observed in the closely related speciesA.marmoratus,A.panhaiandA.nepalicus.

The Dobhan Population (DP) showed the large genetic and morphological variation from MRP andA.marmoratusand possessed similar morphological features withA.nepalicusas described by Yang (1991).Hence,the population of the Dobhan lineage has been identified asA.nepalicus.Additionally,the geographical distance between topotypicA.nepalicus(Shaba Khola,Sankhuwasabha) and Dobhan,Sankhuwasava population in this study is in proximity (approximately 14 km north from the type locality) which further substantiated the nomenclature validation.Therefore,the explanations provided by the Dubois (2004) are not here considered adequate justification for a synonymisation ofA.nepalicusas aA.marmoratus.

The Kimathangka Lamatar Population (KLP) population showed the large genetic and morphological variation from MRP and DP and possessed similar morphological features withA.formosus(holotype -BMNH 1947.2.4.18) deposited at the British Museum of Natural History (Günther 1875).Hence,the population of the KLP lineage has been identified asA.formosus.Taxonomic accounts

Amolops mahabharatensissp.nov.(Figure 4 and 5)

Holotype:CDZMTU0110 (Figure 4 and 5), an adult male found resting on boulder associated with subtropical degraded forest beside wide stream (about 5 m) on Hattibang,collected from Hugdi Khola (Khola refers to stream in Nepali) Hattibang,Chitwan district,Nepal,27.7741048°N,84.6944107°E;elevation 775 m a.s.l.collected by Janak Raj Khatiwada between 19:30 and 20:00 h on 16thJuly 2016 and deposited in the collection of Central Department of Zoology,Tribuvan University,Kirtipur,Kathmandu,Nepal.

Suggested common name:Mahabharat Torrent frog

Suggested common name in Nepali:(Pahadi Pirre Paha)

Etymology:The species name is derived from the noun“Mahabharat” which refers to the foothills of Himalaya,from where this species were collected.

Paratypes:CDZMTU0114 (adult female ? Figure 6),CDZMTU0115 (adult female) and CDZMTU0111 -0113 (adult male) from the same location as the holotype and deposited in the collection of Central Department of Zoology,Tribuvan University,Kirtipur,Kathmandu,Nepal.

Figure 3 Principal Component Analysis (PCA) based on 24 morphometric traits.Prior to PCA,all morphometric data were converted to percentage value of SVL.

Figure 4 Male holotype in natural habitat A.Dorsolateral view B.Ventral view

Figure 5 Male holotype in preservative A.Dorsal view B.Ventral view C.Ventral view of hand D.Ventral view of foot.

Diagnosis:The new species is assigned to genusAmolopsbased on the molecular and morphological similarity (first finger with circummarginal and transverse grooves on disk and tadpoles with gastromyzophorous) of the specimens to those of other species in theA.marmoratusgroup.Further,the new species is assigned to the genusAmolopsdue to the presence of ventral sucker in tadpole (tadpoles were identified with molecular method);a major diagnostic feature (Inger,1966).Amolopsmahabharatensissp.nov differed from other species ofAmolopsbased on the following morphological characters:(1)male SVL (mean 37.1 mm,n=14),(2) dorsal surface granular,(3) head wider than longer,(4) distinct dorsolateral fold,(5)nuptial pad present on first finger of male (6) circummarginal grooves present on all digit discs,(7) toes fully webbed,(8) males with dual gular poaches,(9) vomerine teeth weakly developed,and (10) tadpoles with gastromyzophorous (ventral abdominal sucker) and distinct mouth parts (III:5+5/1+1:II).

Description of holotype (measurements provided in Table 5):Body size moderate (SVL 34.2 mm);head wider than long (HW 12.9,HL 9.7,HW:HL 133.5%) and flat;snout short and round in lateral view (SL:HL 50%);canthus rostralis distinct;nostril lateral;loreal region slightly concave,small eyes and convex (ED 2.8 mm);eye diameter shorter than snout length (ED:SL 58%);tympanum small (TYD:ED 46%);interorbital distance (IOD 4.4 mm),greater than width of the upper eyelid (UEW 3.4 mm),and internarial distance (ENL 1.9 mm);supratympanic fold present,extending from back of eye to shoulder;small tubercles present between the junction of supratympanic fold and upper mandible;vomerine teeth weakly developed;tongue lanceolate,deeply notched posteriorly;gular pouches present.

Arm robust (FAL=7.4 mm) and shorter than hand(HAL=10.6 mm);relative finger lengths I＜II＜IV＜III;third finger(F3=6.7) shorter than arm (FAL=7.4);finger tips dilated with oval disks with circummarginal grooves;third disk width (FD3 1.5),greater than tympanum (TYD:FD3 84%);nuptial pad on first finger,without conical spines,fingers without distinct lateral fringes,webbing absent;rounded subarticular tubercles;formula for subarticular tubercles:1,1,2,1;outer metatcarpal tubercle ovoid,flat and larger than inner tubercle,round and small.

Hindlimbs powerful and long,tibiotarsal articulation reaching the snout when hindlimb is kept parallel to the body;shank (SHL 23.2) longer than thigh (THL 20.7),tarsus (TL 10.6)and foot length (FOL 20.2);toes thin and long,relative lengths I＜II＜III＜V＜IV;toes tips with circular disks with circummarginal grooves;diameter of fourth disks (TD 4 1.7);toes fully webbed;subarticular tubercles slightly ovoid,formula 1,1,2,3,2;inner metatarsal tubercle prominent and oval;outer metatarsal poorly developed;supernumerary and plantar tubercles absent.

Skin texture and colouration in life:Skin granular on the dorsal surface of head,body,limbs,fingers,toes and flank regions,intensity of granular warts increases towards vent;flanks with white larger granules;dorsolateral fold very weakly developed with warty granules;abdomen and ventral surfaces of head,body and thighs free of granules;tympanum smooth.Dorsum greyish olive;dorsolateral and supratympanic folds light brown;thighs with black bands;toe webbing brown;ventral region white;iris pale yellow to light gold,pupil black.

Colouration in preservation:Dorsum of head and body dark brown;flanks creamy;dorsal region of thigh brown,lateral surfaces of arms and thighs light rust;dorsal granules turned to black from creamy white;transverse cross-bars on dorsal surfaces of hands,shanks,tarsus and feet turned to grey;flank and ventral surfaces of hand and thigh region turned to creamy white;ventral surfaces of head and abdomen creamy white.Webbing of the toes turned to semi-transparent grey.

Variations:Morphometric variation of type specimens is presented in Table 4.

Sexual dimorphism:Some morphological characters showed sexual dimorphism.In SVL,females were significantly larger than males;males with sub-gular vocal sac (v.s.absence in females);distinct nuptial pads (v.s.absence in females);thick,robust forearms (v.s.thin and long in females).

Larvae:Two tadpoles ofAmolops mahabharatensissp.nov.(identity confirmed by molecular analysis) on Gosner stage 32 were examined (Figure 7).In dorsal view,body oval or elliptical,posterior region wider than the anterior,total body length (TL 34.8 mm),body length (BL 16.7 mm) 48% of TL and tail length(TAL 21.2 mm) 61% of TL.Head broad,snout rounded,eyes positioned dorsolaterally,naris oval and closer to the snout,ED (1.8 mm) 11% of BL,ENL (2.0 mm) 12% of BL,SN (1.8 mm)11% of BL,SSD (8.1 mm) 48% of BL,IND (1.7 mm) 11% of BL and IOW (2.0 mm) 12% of BL.Open nostril;spiracle sinistral single,midventral,posterodorsally directed and extended as a short tube;spiracular opening oval;dextral vent tube opens at the margin of the ventral fin;intestinal coils invisible;tail musculature robust and greatly narrowing towards the tail tip;tail tip rounded;TMH (3.1 mm) 15% of TAL,UTF (0.82 mm)4% of TAL,LTF (1.0 mm) 5% of TAL,TMW (2.5 mm) 12% of TAL,MTH (2.9 mm) 14% of TAL,dorsal fin originates behind body,upper tail fin originates at first quarter of tail body junction,lower tail fin origin near the middle of tail,upper tail fin higher than lower tail fin,both fins converged at the tip.In ventral view,a large suctorial abdominal disk almost covers 80% of ventral body area.Oral disc:lips lacking papillae,three uninterrupted teeth row and five interrupted teeth row on the upper labium.The lower labium comprises of one interrupted teeth row followed by two uninterrupted teeth row.The teeth row:III:5+5/1+1:II.Beak completely black,margin serrate,about 64 serrates in upper beak and 36 in lower beak,upper beak inverted V-shaped,lower V-shaped.

Figure 6 Female-paratype in natural habitat.

Figure 7 Tadpole of Amolops mahabharatensis sp.nov.A.Stage 32 larva in dorsolateral view B.Ventral view C.Mouth showing denticles

Coloration of tadpoles in preservative:Dorsal and lateral surface of head,body and tail region dark brown,abdominal disc creamy white,abdomen light grey,ventral surface of tail creamy white.

Ecological notes:Amolops mahabharatensissp.nov.inhabits torrent streams with high canopy cover.The vegetation is subtropical mixed forest.Adults of new species were collected from rocky boulders in a fast-flowing stream with more than 85% of rock cover (Figure 8).Tadpoles were attached on the rocks below water and were sampled with the help of a fishing net placed at the bottom of the rock.The tadpoles were rubbed and detached from rock by using a plastic slipper and collected into the net.

Figure 8 A.Habitat and B.Microhabitat of Amolops mahabharatensis sp.nov.at Siraichuli Hill,Chitwan district,Nepal.

Comparisons

Based on the phylogenetic analysis,Amolops mahabharatensissp.nov.is the member ofA.marmoratusspecies group (A.afghanus,A.indoburmensis,A.marmoratus,A.nepalicusandA.panhai) and is morphologically distinct from the currently known members ofA.marmoratusgroupand all other congeneric species (Table 6).

Amolops mahabharatensissp.nov.differs fromA.afghanusby having granular dorsal surface (vs.less granular),smaller body size both in males SVL:33.9-39.1 (vs.large size,SVL:45.6-62.9)and females SVL:61.0-71.7 (vs.large size,SVL:67.7-94.1) and relatively larger tympanum (about 8.2% of SVL) (vs.smaller diameter about 5% of SVL) (Deveret al.2012).

Amolops mahabharatensissp.nov.differs fromA.indoburmanensisby having small body size both in males (vs.large size,SVL:59.0-86.0) and females (vs.large size,SVL:63.0-106.0) (,head wider than long (vs.longer than wide),finely granulated (vs.coarsely granulated in dorsum body and limbs)(Deveret al.2012).

Amolops mahabharatensissp.nov.differs fromA.marmoratusby having head wider than long (vs.longer than wide),distinct tympanum (vs.indistinct),supernumerary tubercles on finger absent (vs.supernumerary tubercles on three outer fingers).

Amolops mahabharatensissp.nov.differs fromA.nepalicusby having head wider than long (vs.longer than wide), keratodont formula of tadpole ? III:5-5/1-1:II (vs.II:4-4/1-1:II).

Amolops mahabharatensissp.nov.differs fromA.panhaiby having large body size in female,SVL:61.0-71.7 (vs.small size,SVL:47.2-57.5),head wider than long (vs.longer than wide),gular pouches present in males (vs.absent),keratodont formula of tadpole II:4-4/1-1:II (vs.III:5-5/1-1:II).

Amolops mahabharatensissp.nov.further differs fromA.formosus,A.monticolaandA.himalayanusby having granular dorsal surface (vs.smooth),circummarginal groove on first finger present (vs.absent),dorsal body greyish olive (vs.dorsum dark green) and distinctive dorsolateral fold absent (vs.present inA.monticola).

Amolops mahabharatensissp.nov.further differs from all other species ofAmolopsbased on the following characters:gular pouches present in males (vs.absent inA.albispinus,A.bellulus,A.caelumnoctis,A.chakrataensis,A.cremnobatus,A.daiyunensis,A.formosus,A.gerbillus,A.granulosus,A.hainanensis,A.jinjiangensis,A.kaulbacki,A.larutensis,A.lifanensis,A.loloensis,A.mantzorum,A.minutus,A.nidorbellus,A.nyingchiensis,A.ricketti,A.shuichengicus,A.sinensis,A.tuberodepressus,A.viridimaculatus,A.xinduqiaoandA.yatseni),head wider than long (vs.longer than wide or head inA.akhaorum,A.albispinus,A.assamensis,A.australis,A.bellulus,A.caelumnoctis,A.chunganensis,A.compotrix,A.cucae,A.daorum,A.gerbillus,A.gerutu,A.granulosus,A.hongkongensis,A.indoburmanensis,A.iriodes,A.jaunsari,A.jinjiangensis,A.kaulbacki,A.kohimaensis,A.larutensis,A.lifanensis,A.loloensis,A.marmoratus,A.mengdingensis,A.mengyangensis,A.minutus,A.monticola,A.nepalicus,A.nyingchiensis,A.panhai,A.shuichengicus,A.splendissimus,A.torrentis,A.tuberodepressus,A.viridimaculatus,A.vitreus,A.wenshanensis,A.xinduqiaoandA.yunkaiensis),distinct tympanum (vs.indistinctA.albispinus,A.gerbillus,A.granulosus,A.jinjiangensis,A.lifanensis,A.loloensis,A.marmoratus,A.shuichengicus,A.viridimaculatusandA.xinduqiao),presence of vomerine teeth (vs.absent inA.daiyunensis,A.daorum,A.hainanensis,A.hongkongensis,A.torrentis,andA.wuyiensis),absent of distinctive dorsolateral fold (vs.presentA.akhaorum,A.aniqiaoensis,A.archotaphus,A.bellulus,A.chakrataensis,A.chayuensis,A.chunganensis,A.compotrix,A.cremnobatus,A.cucae,A.daorum,A.gerbillus,A.iriodes,A.jaunsari,A.jinjiangensis,A.kohimaensis,A.mengdingensis,A.mengyangensis,A.minutus,A.nyingchiensis,A.shuichengicusv,A.vitreus,A.wenshanensisandA.xinduqiao).

Table 5 Measurements (in mm;mean±SD) of the type series of Amolops mahabharatensis sp.nov.See Material and methods for abbreviations.

Supplementary description ofAmolops nepalicus(Yang,1991),based on specimens from Dobhan population (DP)

Diagnostic keys to theAmolops nepalicus(Yang,1991)

1.Head slightly longer than wide

2.Tympanum visible

3.First and second fingers equal in length

4.Supernumerary tubercles at bases of all fingers

5.Toes fully webbed;subarticular tubercles small;inner metatarsal tubercle present

6.No dorsolateral fold;an obvious temporal fold

Comparisons

Based on the diagnostic characters of theAmolops nepalicus(Yang,1991),DP is morphologically similar withA.nepalicusand is morphologically distinct from the currently known members ofA.marmoratusgroup (A.afghanus,A.indoburmensis,A.marmoratus,A.mahabharatensissp.nov.andA.panhai)and all other congeneric species (Table 6).

Figure 9 Amolops nepalicus in preservative A.Dorsal view B.Ventral view C.Ventral view of hand D.Ventral view of foot.

Amolops nepalicusdiffers fromA.afghanusby having small body size,SVL: smaller body size both in males SVL:33.9-39.1(vs.large size,SVL:45.6-62.9) and females SVL:61.0-71.7 (vs.large size,SVL:67.7-94.1),head longer than wide (vs.wider than long),outer metatarsal tubercle (vs.absent) (Deveret al.2012).Amolops nepalicusdiffers fromA.indoburmanensisby having small body size (vs.large size),head longer than wide (vs.wider than long),outer metatarsal tubercle (vs.absent) (Deveret al.2012) (Table 6).Amolops nepalicusdiffers fromA.marmoratusby having head wider than long (vs.longer than wide),distinct tympanum (vs.indistinct).Amolops nepalicusdiffers fromA.mahabharatensissp.nov.by having longer head than wide (vs.wider than long),,keratodont formula of tadpole:II:4-4/1-1:II (vs.III:5-5/1-1:II).Amolops nepalicusdiffers fromA.panhaiby having outer metatarsal tubercle (vs.absent),keratodont formula of tadpole:II:4-4/1-1:II (vs.III:4-4/1-1:II).Amolops nepalicusfurther differs fromA.formosus,A.monticolaandA.himalayanusby having granular dorsal surface (vs.smooth),circummarginal groove on first finger present (vs.absent),and distinctive dorsolateral fold absent (vs.present inA.monticola).

Further,A.nepalicusis differ from several name bearingA.marmoratuscomplex group such asRana latopalmataby having smaller body size (vs.larger SVL;male:77-90 mm and female:90-100 mm),keratodont formula of tadpole:II:4-4/1-1:II (vs.III:5-5/1-1:II) (Boulenger 1920) and withPhilautus argusby having larger body size (vs.smaller SVL;male:27 mm),head longer than wide (vs.wider than long) (Boulenger 1920).

Holotype:Adult male (UMMZ 132063),from “Sabha Khola,Sankhuwasabha district,eastern Nepal”,collected by Douglas Lay on 02 June 1973.

Examined specimens:Specimens ofA.nepalicuswere collected from fast flowing stream of Dobhan,Sankhuwasabha district,Nepal,27.55432 °N,87.299366 °E;elevation 848 m a.s.l.collected by Janak Raj Khatiwada between 19:00 and 22:00h on 25th May 2015 and deposited in the collection of Central Department of Zoology,Tribuvan University,Kirtipur,Kathmandu,Nepal.Description:The given description is based on the collection of specimens from Dobhan,Sankhuwasabha district,eastern Nepal (Figure 9 and Table 7).

Mean body size (male:mean SVL=41.1 mm,n=7 and female:mean SVL=60.0 mm,Table 7);head relatively wide(HW 33% of SVL and HL 30% of SVL);canthus rostralis distinct;eye moderate (37% of HW);pupil round;tympanum small and round (37% of ED);inter-orbital distance (27% of HW) is greater than the inter-narial distance (34% of HW);snout flat and round (22% of HW) and is shorter than nostril to eye distance(18% of HW);width of upper eyelid (25% of HW) almost equal in length as the interorbital width (27% of HW);vomerine teeth present between choanae;dorsolateral folds indistinct.Arm shorter (LA 21% of SVL) than length of hand (LH 47% of SVL);fingers slender,free of webbing and tips with large and wide disc with circummarginal groove which are larger than those of the toes;relative length of fingers from shortest to longest F1 ≤ F2 ＜ F4 ＜ F3;subarticular tubercles distinct and bulging;,supernumerary tubercles small and round,located at bases of all fingers.Hindlimbs moderately slender;length of femur 51% of SVL;length of tibia 56% of SVL;length of tarsus 31% of SVL,length of meta-tarsus 45% of SVL.Tibiotarsal articulation reaches the snout when the hindlimb is kept parallel to the body.Toes strong and long with large disc in third and fourth digits;relative length of toes T1 ＜ T2 ＜ T5 ＜ T3＜ T4;toes fully webbed;subarticular tubercles small,round and distinct;metatarsal tubercle indistinct.Skin with few rounded tubercles scattered on back;dorsolateral fold indistinct;small tubercles on outer side of thigh;belly smooth.

Coloration (in alcohol):Light brown or white on back with dark brown spots on tubercles;darker brown and black crossbars on limbs;belly yellowish grey.Male with paired gular pouches.

Distribution:This species is known with certainty only from the Sabha Khola and Dobhan,Sankhuwasabha district,Nepal.

Ecological notes:The species inhabits torrent habitats with high canopy cover.The vegetation is subtropical mixed forest.Adults ofA.nepalicuswere collected from rocky surfaces in the fast-flowing stream with more than 90% of rock cover.

IdentificationAmolopspopulations from Kimathanka and Lamatar asAmolops formosus(Guenther,1875)

Diagnostic keys to theAmolops formosus[based on the original description detailed by Guenther (1875) and measurements of the holotype (BMNH 1947.2.4.18)].

1.Snout flat,short,rounded,with distinct canthus rostralis

2.Tympanum very small,only as large as the disk of the third finger

3.Toes fully webbed,metatarsal tubercle,indistinct

4.Skin quite smooth

5.Vomerine teeth presence between choanae

6.Dorsolateral folds indistinct

7.Dorsum green,marbled with black,the black spots enclosing a number of small whitish dots.

Comparisons

Based on the morphological characters of theAmolops formosus(holotype BMNH 1947.2.4.18)].,KLP is morphologically similar withA.formosusand is morphologically distinct from all other congeneric species (Table 6).A.formosusdiffers fromA.himalayanusby having distinct tympanum (vs.indistinct),circummarginal groove on first finger present (vs.absent),tibiotarsal articulation reaches to the snout (vs.reaches beyond the tip of the snout) and distinctive dorsolateral fold absent (vs.present inA.monticola).

Holotype:Adult male (BMNH1947.2.4.18,formerly 1874.4.17.281),from “Khasi Hills,Assam,India”,collected byThomas Caverhill Jerdon.

Examined specimens:Specimens ofA.formosuswere collected from fast flowing mountainous stream of Kimathanka,Sankhuwasabha district (27.839452° N,87.393538°E;elevation 2508 m a.s.l.) and Lamatar,Taplejung district,Nepal(27.5361° N,87.80802° E;elevation 1586 m a.s.l.) collected by Janak Raj Khatiwada between 19:00 and 22:00 h on 24th June 2015 and deposited in the collection of Central Department of Zoology,Tribuvan University,Kirtipur,Kathmandu,Nepal.

Description:The given description is based on the collection of specimens from Sankhuwasabha and Taplejung district,eastern Nepal (Figure 10 and Table.8).

Female body size (SVL=63.1-73.1 mm,n=3);head relatively wider (HW 27% of SVL and HL 17% of SVL);canthus rostralis distinct;eye of moderate (39% of HW);pupil round;tympanum small and round (3% of SVL);inter-orbital distance (8% of SVL)greater than the inter-narial distance (9% of SVL);snout flat and rounded,snout (8% of SVL) longer than nostril to eye distance (8% of SVL);width of upper eyelid (4% of SVL) almost half length to the interorbital width;presence of small granules at angle of jaws and extra tympanic area;dorsolateral folds indistinct.Arm shorter (12% of SVL) than LH (length of hand 48% of SVL);fingers muscular,free of webbing,finger tips with large and wide disc with circummarginal groove and larger than those of the toes;relative length of fingers from shortest to longest F1 ＜ F2 ＜ F4 ＜ F3;subarticular tubercles distinct and oval;elongated inner metacarpal tubercle,smaller than outer metacarpal.Tibiotarsal articulation reaches the snout when the hindlimb is kept parallel to the body.Hindlimbs muscular and slender;length of femur 48% of SVL;length of tibia 53% of SVL;length of tarsus 37% of SVL,length of meta-tarsus 21% of SVL.Toes strong and long with large disc in third and fourth digits;relative length of toes T1 ＜ T2 ＜ T5 ＜ T3 ＜ T4;toes fully webbed,subarticular tubercles oval and distinct;metatarsal tubercle indistinct.

Colouration:In life:skin smooth,dorsum green with irregular distinct chocolate coloured blotches with yellow dots.Ventral light and abdomen greenish.In alcohol:green colour faded to grey;the chocolate coloured blotches changed to brown.

Figure 10 Amolops formosus A.Adult female B.Juvenile

Distribution:This species is distributed in Assam and Sikkim,India and eastern Nepal (Günther,1875;Schleich and K?stle,2002;Dineshet al.,2009) and in this study recorded above 1500 m in the eastern Nepal.

4.Discussion

In this study we have demonstrated significant differences in molecular and morphology between three species ofAmolopscurrently distributed in the Mahabharat range of Nepal and with other closely related taxa.The population ofAmolopsfrom middle mountain range of Mahabharat range of Nepal are described as a new species (Amolops mahabharatensissp.nov.).Previously,theAmolopspopulation of the middle mountainous range of Nepal were identified asAmolops marmoratus(Dubois 2000;2004;Nanhoe &Ouboter 1987;Schleich &K?stle 2002;Shah &Tiwari 2004).However,our results show the presence of three species in the middle mountainous regions of Nepal.Although,Amolops mahabharatensissp.nov.resembled some morphological similarities withA.nepalicusandA.marmoratus,it is genetically different from topotypicA.marmoratusandA.nepalicus.Possibly,A.marmoratusdoes not occur in Nepal.Whereas,theAmolopspopulation from Kimathanka,Sankhuwasava and Lamatar,Taplejung showed similar morphological characteristics of holotype ofA.formosus(BMNH1947.2.4.18,formerly 1874.4.17.281) and identified asA.formosus(Table 7).

AllAmolopsspecies in the Mahabharat range of Nepal were originally identified asA.marmoratusbased on morphology.This method has resulted in misleading the species identification,especially for the cryptic taxa likeAmolopswhich possess high morphological similarities (Bainet al.,2003;Lyuet al.,2019b).In the recent years,with the advancement of molecular techniques,several cryptic amphibians have been identified as new species from Nepal (Khatiwadaet al.,2015;Khatiwadaet al.,2017;Mahonyet al.,2018;Khatiwadaet al.,2019a).This study has resolved previous taxonomic confusion in the genusAmolopsusing molecular phylogenetic analysis.The discovery ofA.mahabharatensissp.nov.may indicate the presence of additional cryptic diversity withinAmolopsspecies from Nepal.

Natural history and conservation:Majority of people in the survey area have a strong belief in ethnoherpetology.Ethnic communities in the mountainous areas of Nepal depend on natural resources for their livelihood (Khatiwada and Haugaasen 2015).They engage in the collection of tadpoles,adults and even eggs ofAmolopsspecies.Khatiwada and Haugaasen (2015) also reported the overharvesting ofAmolopsspecies.The amphibian population is highly threatened and at the risk of extinction by recent global environmental changes and further accelerated by anthropogenic activities like overharvesting (Chanet al.2014).Since,the richness and abundance ofAmolopsspecies is relatively low compared to other amphibian species in Nepal (Khatiwadaet al.,2019b),monitoring of the populations of edible frogs should be a conservation priority in the study area and elsewhere in Nepal.

Table 8 Measurements (in mm;mean±SD) of Amolops formosus. See Material and methods for abbreviations.

AcknowledgementsWe are thankful to Subarna Ghimire,Suman Sapkota,Purnman Shrestha,Bibas Shrestha,Bijaya Paudel,Suman Acharya and Anish Timsina for their assistance during fieldwork.We are grateful to the Department of National Park and Wildlife Conservation (DNPWC) and Department of Forest,Government of Nepal for providing necessary research permit for conducting the research.We are thankful to Benjamin Tapley,Zoological Society of London,London,UK for the measurement of holotype ofAmolops formosus,reviewing English language and also provided critical comments and suggestions that substantially changed the quality of the manuscript.We are thankful to two anonymous reviewers and Prof.S.D.Biju,University of Delhi for their insightful comments,suggestions and improvements in the final version of manuscript.This work was supported by the Second Tibetan Plateau Scientific Expedition and Research Program (STEP),Grant No.2019QZKK04020202 and National Natural Sciences Foundation of China (NSFC-31471964)granted to Jianping Jiang.Janak Raj Khatiwada was supported by Chinese Academy of Sciences President’s International Fellowship Initiative (2018PB0016).