Hidden tribe: A new species of Stream Toad of the genus Ansonia Stoliczka, 1870 (Anura: Bufonidae) from the poorly explored mountainous borderlands of western Thailand

The integrated results of morphological and molecular phylogenetic analyses confirmed the new species status of a recently discov­ ered population of Ansonia from Suan Phueng District, Ratchaburi Province, Thailand. Ansonia karen sp. nov. is separated from all other species of Ansonia by a unique combination of mensural, discrete morphological, and color pattern characteristics and is the sister species of A. thinthinae from Tanintharyi Division, Myanmar. This discovery fills a geographic hiatus of 350 km between it and A. kraensis from Ranong Province, Thailand. Ansonia karen sp. nov. is the newest member of a long list of range­restricted endemics having been recently discovered in the northern Tenasserim Mountain region of western Thailand and continues to underscore the unexplored nature of this region and its need for conservation.


Introduction
Stream toads of the genus Ansonia Stoliczka, 1870 com prise a distinctive clade of small anurans with relatively flat bodies and heads, and long thin limbs with slender bulbous digits that are adaptations for their scansorial, lotic life style. Species of Ansonia are generally restrict ed to rocky fast-flowing streams along the mountainous and hilly border regions of southeastern Myanmar and western and southern Thailand, southward through the ThaiMalay Peninsula to Sumatra, Borneo, and the Phil ippines (Grismer et al. 2016;Quah et al. 2019).
The general life history of their tadpoles-adhering themselves to the surfaces of rocks beneath fast-flow ing water or in the spray zones of cascades-restricts the distribution of Ansonia to riverine habitats. As such, rangerestricted endemism is characteristic of many spe cies in this genus and those with widespread distributions from multiple localities are likely to be species complex es (e.g. Sanguilla et al. 2011;Grismer et al. 2016;Gris mer et al. in prep.). This is the case for a ThaiBurmese clade of nine species of Ansonia distributed from eastern Myanmar to western and southern Thailand where each is known from only its type locality or another locality close by with confluent riverine systems (Fig. 1). We report here a new population from western Thailand found in the hilly region of Suan Phueng District in Ratchaburi Province, that fills in a notable 350 km hiatus between A. thinthinae Wilkinson, Sellas & Vindum from the Tanintharyi Na ture Reserve, Tanintharyi Division, Myanmar (Wilkinson et al. 2012) and A. kraensis Matsui, Khonsue & Nabhi tabhata from the Punyaban Waterfall, Ranong Province, Thailand (Matsui et al. 2005). A molecular analysis using 12S and 16S ribosomal RNA genes recovered this new population as the sister species of A. thinthinae, and uni variate and multivariate analyses of mensural characters and comparisons of discrete morphological and color pat tern differences among species within the Thai-Burmese clade, clearly differentiate it from all other members. Therefore, based on genetic and morphological data, we describe this new population as a new species.

Sampling
Specimens were collected in Suan Phueng District, Ratch aburi Province, Thailand by Parinya Pawangkhanant, Platon V. Yushchenko, Mali Naiduangchan, Chatmon gkon Suwannapoom, and Nikolay A. Poyarkov during several field surveys from 2016 to 2019. The loca tion of the surveyed locality and the distribution of the ThaiBurmese clade of Ansonia are shown in Figure 1. Geographic coordinates and elevation were obtained us ing a Garmin GPSMAP 60CSx and recorded in WGS 84 datum. Specimens were collected by hand, euthanized with 20% benzocaine solution, femoral muscle or liver tissue samples removed, and then fixed in 10% buffered formalin before preserving in 70% ethanol. The tissue samples were stored in 95% ethanol. Specimens and tis sues were subsequently deposited in the herpetological collections of the School of Agriculture and Natural Re sources, University of Phayao (AUP, Phayao, Thailand) and of the Zoological Museum of Moscow University (ZMMU, Moscow, Russia).
Specimen collection and animal use protocols were approved by the Institutional Ethical Committee of An imal Experimentation of the University of Phayao, Pha yao, Thailand (certificate number UP-AE61-01-04-022, issued to Chatmongkon Suwannapoom) and were strictly complacent with the ethical conditions of the Thailand Animal Welfare Act. Field work, including collection of animals in the field and specimen exportation, was autho rized by the Institute of Animals for Scientific Purpose Development (IAD), Bangkok, Thailand (permit num bers U1012052558 and UPAE5901040022, issued to Chatmongkon Suwannapoom).

Morphological data and analyses
Observations on color pattern were based on the examination of specimens in life as well as digital images of living and euthanized specimens prior to preservation. Measure ments were recorded with a Mitutoyo dial caliper under a Nikon SMZ 1500 dissecting microscope to the nearest 0.01 mm. Measure ments of adult specimens generally follow ing Wilkinson et al. (2012) were: snout-vent length (SVL, from tip of snout to vent); head length (HL, from tip of snout to hind border of angle of jaw); head width (HW, width of head at its widest point); snout width (SW, width of snout at anterior corner of eyes); snout length (SL, from anterior border of eye to tip of snout); distance from nostril to eye (DNE, from center of nostril to anterior border of eye); internarial distance (IND, dis tance between center of nares); interorbital distance (IOD, minimum distance between upper eyelids); eye diameter (ED, horizontal diameter of eye); upper eyelid width (UEW, greatest transverse width of eyelid); vertical tympanum diameter (VTD; vertical diameter of tympanum); horizontal tympanum diame ter (HTD; horizontal diameter of tympanum); tympanum to eye distance (TED, from ante rior edge of tympanum to posterior edge of eye); forearm length (FAL, from elbow to tip of third finger); hand length (HAL, from proximal edge of palmar tubercle to tip of third finger); first finger length (1FL, from distal end of inner metacarpal tubercle to tip of first finger); thigh length (THL, from vent to knee); tibia length (TIL, from knee to an kle); foot length (FL, from proximal end of outer metatarsal tubercle to tip of fourth toe); first toe length (1TL, from distal end of inner metatarsal tubercle to tip of first toe); inner metatarsal tubercle length (IMTL, greatest length of tubercle); outer metatarsal tuber cle length (OMTL, greatest length of tuber cle); third finger disk width (3FDW, maxi mal width of terminal disk on third finger); hindlimb length (HLL, length of straightened hindlimb from groin to tip of fourth toe); and forelimb length (FLL, length of straightened forelimb from axilla to tip of third finger) ( Table 1).
Measurements on a single tadpole speci men AUP02091 followed Inger (1985) and Wilkinson et al. (2012), and were taken un der a Nikon SMZ 1500 dissecting microscope with a micrometer as follows: total length, headbody length (tip of snout to insertion of tail), headbody depth, maximum headbody width, diameter of eyeball, interorbital dis tance, eye to tip of snout, internarial distance, width of oral disc, tail length, maximum tail depth, tail muscular depth, thigh length, tibia length, and foot length. Tadpole staging followed Gosner (1960), oral apparatus terminol ogy followed Altig and McDiarmid (1999), and labial tooth row formula followed Altig et al. (1998).
The morphospatial clustering among sampled individ uals from a selected subset of species and characters for which there was full coverage for each species were vi sualized using principal component analysis (PCA) from the ADEGENET package in R (Jombart et al. 2010). Only males were used in the analysis because only one female of Ansonia thinthinae was available. Characters analyzed were HL, HW, SL, ED, HTD, IND, IOD, FL, and TIL. To remove potential effects of allometry, size was normalized using the following equation: X adj =log(X )β[log(SVL)log(SVL mean )], where X adj =adjusted value; X=measured value; β=unstandardized regression coef ficient for each population; and SVL mean =overall aver age SVL of all populations (Thorpe 1975(Thorpe , 1983Turan 1999;Lleonart et al. 2000, accessible in the R package GroupStruct (available at https://github.com/chankinonn/ GroupStruct). The morphometrics of each species were normalized separately and then concatenated so as not to conflate intra-with interspecific variation (Reist 1986). All data were scaled to their standard deviation to insure they were analyzed on the basis of correlation and not covariance. For corroboration, a discriminant analysis of principal components (DAPC) was performed on the same data set. DAPC relies on scaled data calculated from its own internal PCA as a prior step to ensure that variables analyzed are not correlated and number fewer than the sample size. Dimension reduction of the DAPC prior to plotting, is accomplished by retaining the first set of PCs that account for 90-99% of the variation in the data set (Jombart and Collins 2015) as determined from a scree plot generated as part of the analysis.
Twosample t-tests of the all the scaled mensural char acters were used to search for statistically significant mean morphometric differences between the new Thai population and its sister species, A. thinthinae (see be low). Characters were subjected to an F-test to test for homogeneity of variances. Those with unequal variances were subjected to a Welch's t-test and those with equal variances were subjected to a Student t-test. All statistical analyses were performed in R [v3.4.3]. Raw and adjusted data are presented in Table S1.

Laboratory methods
For the molecular phylogenetic analyses, we extracted the total genomic DNA from ethanolpreserved femoral mus cle tissue of six specimens of the new Thai population us ing standard phenolchloroformproteinase K extraction procedures with consequent isopropanol precipitation, for a final concentration of about 1 mg/ml (protocols followed Hillis et al. (1996) and Sambrook and David (2001)). We visualized the isolated total genomic DNA in agarose electrophoresis in the presence of ethidium bromide. We measured the concentration of total DNA in 1 μl using NanoDrop 2000 (Thermo Scientific), and consequently adjusted to ca. 100 ng DNA/μL.
We amplified mtDNA fragments covering partial 16S rRNA gene sequences to obtain a 560 bplength contin uous fragment per specimen. The 16S rRNA gene has widely been applied in biodiversity surveys in amphib ians (Vences et al. 2005a(Vences et al. , 2005bVieites et al. 2009), and has been used in the most of the recent phylogenetic studies on Ansonia (e.g. Grismer et al. 2016;Matsui et al. 2007;Matsui et al. 2010;Wilkinson et al. 2012). We performed DNA amplification in 20 μl reactions using ca. 50 ng genomic DNA, 10 nMol of each primer, 15 nMol of each dNTP, 50 nMol additional MgCl 2 , Taq PCR buffer (10 mM TrisHCl, pH 8.3, 50 mM KCl, 1.1 mM MgCl 2 and 0.01% gelatine) and 1 unit of Taq DNA polymerase. Primers used in the PCR and sequencing include 16sL1 (CTGACCGTGCAAAGGTAGCGTAATCACT) and 16H1 (CTCCGGTCTGAACTCAGATCACGTAGG) (Hedges 1994). The PCR conditions included an initial denaturation step of 5 min at 94°C and 43 cycles of de naturation for 1 min at 94°C, primer annealing for 1 min with the TouchDown program from 65 to 55°C reducing 1°C every cycle, and extension for 1 min at 72°C, and a final extension step for 5 min at 72°C.
PCR products were loaded onto 1.5% agarose gels in the presence of ethidium bromide and visualized in aga rose electrophoresis. When distinct bands were produced, we purified the PCR products using 2 μl of a 1:4 dilution of ExoSapIt (Amersham) per 5 μl of PCR product prior to cycle sequencing. A 10 μl sequencing reaction included 2 μL of template, 2.5 μl of sequencing buffer, 0.8 μl of 10 pMol primer, 0.4 μl of BigDye Terminator version 3.1 Sequencing Standard (Applied Biosystems) and 4.2 μl of water. The cyclesequencing used 35 cycles of 10 sec at 96°C, 10 sec at 50°C and 4 min at 60°C. We purified the cyclesequencing products by ethanol precipitation. We carried out sequence data collection and visualization on an ABI 3730xl Automated Sequencer (Applied Biosys tems).

Genetic data and phylogenetic analyses
Ingroup samples consisted of 128 individuals represent ing 32 nominal species and included three of the six samples from the new Thai population. Outgroups used to root the tree were Rentapia hosii (Boulenger), Pelophryne brevipes (Peters), P. misera (Mocquard), and P. signata  Table 2).
Maximum Likelihood (ML) and Bayesian Inference (BI) were used to estimate phylogenetic trees. Best-fit models of evolution were determined in IQTREE (Nguy en et al. 2015) using the Bayesian information criterion (BIC) implemented in ModelFinder (Kalyaanamoorthy et al. 2017). TIM2+F+I+G4 was the best-fit model of evolution for both 12S and 16S. The ML analysis was performed using the IQ-TREE webserver (Trifinopoulos et al. 2016) with 1000 bootstrap pseudoreplicates using the ultrafast bootstrap (UFB) analysis (Minh et al. 2013;Hoang et al. 2018). The BI analysis was performed on CIPRES Science Gateway (Miller et al. 2010) using Mr Bayes v3.2.4 (Ronquist et al. 2012). Two independent runs were performed using Metropoliscoupled Markov Chain Monte Carlo (MCMCMC), each with four chains: three hot and one cold. The MCMCMC chains were run for 60,000,000 generations with the cold chain sampled every 6000 generations and the first 25% of each run be ing discarded as burnin. The posterior distribution of trees from each run was summarized using the sumt func tion in MrBayes v3.2.4 (Ronquist et al. 2012).
Stationarity was checked with Tracer v1.6 (Rambaut et al. 2014) to be sure the effective sample sizes (ESS) for all parameters were greater than 200. We considered Bayesian posterior probabilities (BPP) of 0.95 and above and ultrafast bootstrap support values (UFB) of 95 and above as strong nodal support (Huelsenbeck et al. 2001;Minh et al. 2013). Uncorrected pairwise sequence di vergences (pdistance) were calculated in MEGA v6.06 (Tamura et al. 2013) using the complete deletion option which removes gaps and missing data from the alignment prior to analysis.

Phylogenetic data
The ML and BI analyses recovered nearly identical trees (Fig. 2). Both analyses recovered a strongly support ed (UFB 100/BPP 1.00) Thai-Burmese clade of Ansonia composed of A. siamensis Kiew, A. khaochangensis Grismer, Wood, Aowopol, Cota, Grismer, Murdoch, Aguilar, and Grismer, A. pilokensis Matsui, Khouse, and Panha, A. phuketensis Matsui, Khonsue, and Panha, A. kyaiktiyoensis Quah, Grismer, Wood, Myint Kyaw Thu ra, Oaks, and Aung Lin, A. inthanon Matsui, Nabhitabha ta, Panha, A. kraensis, A. thinthinae, and the new popu lation from Suan Phueng District, Ratchaburi Province, Thailand. The Suan Phueng population was recovered as the strongly supported (100/1.00) sister species of A. thinthinae in both analyses with an uncorrected pairwise sequence divergence between them for the 16S rRNA gene of 4.1%. The only difference between the ML and BI analyses was the placement of A. pilokensis as a poor ly supported (UFB 79) sister species of A. phuketensis in the ML analysis, whereas, in the BI analysis, these two species formed an unsupported polytomy with the remaining species in the ThaiBurmese clade. Their sis ter species relationship was poorly supported in both analyses (84/0.54) in Quah et al. (2019). All other rela tionships were in complete concordance with Quah et al. (2019).

Mensural data
The new Thai population from Suan Phueng and its sis ter species Ansonia thinthinae have statistically differ ent head, body, and limb proportions (Tables 2, 3) with the former having a significantly shorter and wider head (HL and HW, respectively); shorter snout (SL); smaller eyes and tympana (ED and HTD, respectively); smaller internarial and interorbital diameters (IND and IOD, re spectively); and shorter hind limbs (TIL and FL, respec tively). The two populations are widely separated in the PCA (Fig. 2) where principal component (PC) 1 accounts for 63.8% of the variation in the dataset and loads most heavily for head width (HW), snout length (SL), limb length (FLL and HLL), and eye diameter (ED). PC 2 ac counts for an additional 10.4% of the variation and loads most heavily for internarial distance (IND) ( Table 4). The DAPC corroborates the PCA in that each population is completely separated from one another along the first dis criminant function that accounts for 99.4% of the data set ( Fig. 2). Based on all the above data, the new Thai population from Suan Phueng District is described below as a new species.

Coloration in preservative.
After five years of storage in ethanol, the warm reddish, yellowish and orange tints have significantly faded, the specimen looks dark grey ishbrown; however all major features of coloration pat tern are still welldiscernable.
Variation. Raw and adjusted mensural data of the type se ries are presented in Tables 1 and 1s, respectively. Males have smaller body sizes than females, and their SVL val ues do not overlap (male SVL = 23.2-25.6 mm, average 24.4±0.8 mm; vs. female SVL = 26.2-29.2 mm, average 28.2±1.2 mm). The members of the type series generally agree in coloration with that of the holotype (see Fig. 5). Males ZMMU A-7607 (Fig. 5A) and ZMMU A-7608 ( Fig. 5C) have generally lighter pinkish-grey coloration of ventral surfaces. The shape of ventral reticulum var ies from having dense small yellowish and blackish spots (as in male AUP-00662, Fig. 5E) to larger intermittent longitudinal black blotches with yellowish veins between them (as in males ZMMU A7607 and ZMMU A7608, Fig. 5A, C). Males AUP-00661 and AUP-00662 original ly had damaged and partially regenerated left hind limbs (Fig. 5D-E). Other morphological features showed no significant variation among the type series.
Larval morphology. Description based on AUP02091 at Gosner (1960) stage 38. Total length 16.9 mm, head body length 6.2 mm, headbody depth 2.4 mm, maxi mum headbody width 3.3 mm, diameter of eyeball 0.9 mm, interorbital distance 1.4 mm, eye to tip of snout 1.7 mm, internarial distance 1.1 mm, width of oral disc 3.0 mm, tail length 10.8 mm, maximum tail depth 2.4 mm, tail muscular depth 1.7 mm, thigh length 1.0 mm, tibia length 1.2 mm, foot length 1.7 mm. Body distinctly flat tened dorsoventrally (Fig. 6A), broadly oval-shaped in dorsal and ventral views with maximum width posterior to eyes (Fig. 6B); snout broadly rounded in dorsal view (Fig. 6B); eyes with dorsolateral orientation; nostrils lo cated closer to eyes than to tip of snout, with anterolateral orientation. Oral disc ventral, forming a sucker, compris ing ca. 93% of headbody width, not emarginate, both oral labia expanded (Fig. 6D); anterior labium slightly smaller than posterior labium, separated from tip of snout by deep groove; marginal papillae in single row across posterior labium and not discernable on anterior labium, submarginal papillae in two rows on posterior labium; black, serrated jaw sheaths, upper divided with gap of ca. same length as single sheath, lower continuous; la bial tooth (keratodont) row formula 2/3, all rows con tinued, well separated from jaw sheaths, anterior rows medially curved, slightly longer than posterior rows (Fig.  6D); tail musculature welldeveloped, tapering posteri orly to pointed tail tip; tail deepest in anterior one third of length; dorsal and ventral fins approximately equal in depth.

Tadpole coloration.
In life (Fig. 6) dorsal surfaces of body and tail dark violetbrown with numerous, golden and bronzecolored specks scattered along tail, getting denser on body anteriorly and laterally and around eyes (Fig. 6B). Laterally dark violet-gray with bright golden or metal specks on tail and body flanks; limbs dorsally with bronze specks and transverse dark bands (Fig. 6B). Ventrally semitranslucent lavendergray lacking golden specks (Fig. 6C). After three years in preservative, dor sal surfaces of body turned grayishbrown, with dense ly welldiscernable scattered brown chromatophores; ventral surfaces with very few chromatophores medially getting denser laterally; dorsal and ventral tail fins trans parent with few chromotaphores. Distribution. Ansonia karen sp. nov. is currently known only from the type locality and nearby locality in same forest stream in the environs of Khao Laem Mountain, in Suan Phueng District of Ratchaburi Province in west ern Thailand, less than 2.0 km from the international Thai-Myanmar border (Fig. 1). The new species likely inhabits the middle portion of the Northern Tenasserim Mountain range (between the Kanchanaburi and Prach uap Khiri Khan provinces), and is expected to occur in adjacent montane areas in the western part of Phetch aburi Province of Thailand and Thanintharyi Division of Myanmar.
Natural history. The new species inhabits a polydom inant montane tropical evergreen forest on Khao Laem Mountain at elevations from ca. 700 to 750 m a.s.l., where the adult specimens were observed at night perched on leaves or stones ( Fig. 7B-C) along an approximately 1-3 m wide, slow-flowing mountain stream (Fig. 7A), or be neath stones along the stream's edge. The multi-species polydominant tropical forest at the type locality has dense vegetation with tangles of giant bamboo Dendrocalamus asper (Schult.) Backer. Males were calling during our field observations from June to November throughout 2017-2019. The tadpoles of the new species were record ed in the same stream and were usually concentrated in pools under small waterfalls, hiding under gravel on the stream bottom, or sitting on the vertical surfaces of large submerged boulders to which they were attached by their oral discs (Fig. 7D).
The species of amphibians and reptiles recorded in sympatry with the new species at the type locality in clude: Leptobrachium tenasserimense Pawangkhanant, Poyarkov, Duong, Naiduangchan & Suwannapoom, L. smithi Matsui, Xenophrys cf. major (Boulenger), Leptobrachella melanoleuca (Matsui), L. fuliginosa (Matsui), Etymology. The specific name "karen" is given as a noun in apposition and refers to the name of the Karen people. Originally inhabiting wide areas in southern and southeastern Myanmar, many Karen have migrated to Thailand, having settled mostly on the Thailand-Myan mar border, including the Suan Phueng District, the type locality of the new species, due to the political turmoil during the end of XX -beginning of XXI centuries. We received significant help and assistance from the local Karen community in Suan Phueng during our field sur veys and want to thank them for their permanent support. NAP also thanks Karen Sarkisian for his support and en couragement.
Comparisons. Ansonia karen sp. nov. is most closely re lated to A. thinthinae but differs from it by being smaller, more squat and having statistically significant differences in head and limb proportions (see above and Table 2). It differs in coloration and pattern from A. thinthinae in having (as opposed to lacking) redtipped tubercles on the dorsum and flanks; lacking, as opposed to having, gular spotting; having irregularly shaped gray crossbars on the hind limbs as opposed to having regularly shaped, thin, yellowish crossbars; and the bottoms of the hands and feet being reddish-orange as opposed to black. Differenc es between, and among, other species in the ThaiBur mese clade are summarized in Table 5.

Discussion
The genus Ansonia was hypothesized to have evolved and diversified in Borneo before independently dispersing to the Philippines, Sumatra, and twice onto the ThaiMalay Peninsula (Grismer et al. 2016). The first colonization of the ThaiMalay Peninsula ~11.1 million years ago (mya), ultimately resulted in the evolution of a clade that cur rently contains at least 17 species, including the new spe cies Ansonia karen sp. nov. Diversification of this clade at its point of origin in Peninsular Malaysia was followed    by a northward expansion across the KangarPattani Line between Thailand and Peninsular Malaysia at approxi mately 6.7 mya, giving rise to the ThaiBurmese clade and further diversification into at least eight species after crossing the Isthmus of Kra farther north at approximate ly 5.5 mya (Grismer et al. 2016). Ansonia karen sp. nov. is the newest member of the Thai-Burmese species which is confined to the rugged mountainous regions west of the Chao Praya Basin of Thailand. The close geographic proximity of some of its nonsister species (e.g. A. pilokensis, A. khaochangensis, and A. phuketensis) and the discordance between the phylogenetic relationships and geographic distribution of the other Ansonia species, is indicative of the complicat ed biogeographic nature concerning the origin of these rangerestricted endemics. The discovery of Ansonia karen sp. nov. in this section of the Tenasserim Mountains is more of an expectation than a surprise in that it fills a notable hiatus of 350 km between A. thinthinae from the Tanintharyi Nature Reserve, Tanintharyi Division, Myanmar and A. kraensis from the Punyaban Waterfall, Ranong Province, Thailand (see Fig. 1; Wilkinson et al. 2012).
The northern Tenasserim Mountain region is nota ble for the recent discoveries of endemic amphibians and reptiles (Matsui 2006;Sumontha et al. 2012Sumontha et al. , 2017Wilkinson et al. 2012;Connette et al. 2017;Grismer et al. 2016Grismer et al. , 2020aGrismer et al. , 2020bGrismer et al. , 2020cMatsui et al. 2018;Pawangkhanant et al. 2018;Suwannapoom et al. 2018;Lee et al. 2019;Chomdej et al. 2020Chomdej et al. , 2021Poyarkov et al., 2020). The mountains of northern Tenasserim are recognized as one of multiple local centers of amphib ian diversity and endemism in Indochina . Phylogenetic studies further indicate that these mountains have significantly influenced the cladogenetic structure and faunal exchange between mainland Indo china and Sundaland throughout the Cenozoic (see Chen et al. 2017Chen et al. , 2018Grismer et al. 2018;Poyarkov et al. 2018;Suwannapoom et al. 2018;Gorin et al. 2020;Al Razi et al., 2021). Additional field surveys and subse quent integrative taxonomic analyses are necessary to continue expanding our knowledge of this region's ex ceptional herpetofaunal diversity and endemism in order to effectively put into place science-based conservation management programs.