A New Limnonectes (Anura: Dicroglossidae) from Southern Thailand

Simple Summary New species of frogs continue to be discovered at a rapid rate in Southeast Asia, often as a result of reexamining populations of geographically widespread species using new molecular and bioacoustic tools. Here, we show that members of the fanged frog genus Limnonectes from Ko Pha-ngan, Ko Samui, and Ko Lanta Yai Islands in southern Thailand can be distinguished from the morphologically similar species Limnonectes doriae in molecular, advertisement call, morphometric, and qualitative morphological characters. On the basis of these multiple lines of evidence, we describe the insular populations in southern Thailand as a new species, Limnonectes pseudodoriae sp. nov. The new species occurs near small streams in low to mid-elevation forests and breeds in terrestrial nests consisting of moist, clay depressions in which the eggs and larvae develop. Abstract A new species in the dicroglossid frog genus Limnonectes is described from Ko Pha-ngan, Ko Samui, and Ko Lanta Yai Islands in southern Thailand. Males of Limnonectes pseudodoriae sp. nov. lack a caruncle on top of the head and very closely resemble L. doriae (Boulenger, 1887) from Myanmar and western and southern Thailand. However, the new species is distinguished from L. doriae and its congeners using an integrative taxonomic approach of morphology, mitochondrial DNA, and bioacoustics. Limnonectes pseudodoriae sp. nov. differs from L. doriae and its congeners by having a unique combination of morphological characters, including body size; skin texture of the interorbital region, dorsum, and shank; toe webbing; relative size of the inner metatarsal tubercle; and coloration of the tympanum, venter, and ova. The advertisement call of the new species is also readily differentiated from that of L. doriae in temporal parameters. Limnonectes pseudodoriae sp. nov. is highly divergent in mitochondrial DNA from L. doriae and its congeners, but its phylogenetic position within the genus is not resolved. The natural history of the new species is presented, and the geographic range of L. doriae in Thailand is clarified.


Introduction
The dicroglossid frog genus Limnonectes Fitzinger, 1843, contains 76 recognized species and is widely distributed in East and Southeast Asia [1]. Most species in the genus are sexually dimorphic, with males having larger body size, enlarged heads with hypertrophied jaw musculature, and enlarged odontoid processes or 'fangs" on the lower jaw relative to females. Males of some Limnonectes species have a swollen or cap-like structure on the top of the head or "caruncle" [2][3][4][5][6][7][8]. Members of this genus also tend to exhibit a great deal of morphological similarity [9,10]. Many new species of Limnonectes have been discovered and described during the last decade, including in Thailand and adjacent regions. Limnonectes lauhachindai Aowphol, Rujirawan, Taksintum, Chuaynkern, and Stuart, 2015, from northeastern Thailand morphologically resembled L. gyldenstolpei (Andersson, 1916) and L. dabanus ; L. coffeatus Phimmachak, Sivongxay, Seateun, Yodthong,

Morphological Measurements
Measurements were taken to the nearest 0.1 mm with digital Vernier calipers (Mitutoyo CD-6" ASX Digimatic Caliper, Japan) and a Nikon SMZ 745 Zoom Stereomicroscope. Fifteen characters were measured on adult specimens following Aowphol et al. [ Sex and maturity were determined by internal examination of gonads and the presence of secondary sexual characteristics (e.g., enlarged odontoid processes, hypertrophied heads, and/or vocal sac openings). The toe webbing formula follows that used by Savage and Heyer [35]. Qualitative characters were measured following Aowphol et al. [8]. Thirty specimens of L. doriae were examined from the holdings of ZMKU and FMNH (Data S1; data from FMNH specimens were taken by Somphouthone Phimmachak). Comparative data for males of caruncle-lacking Limnonectes species were obtained from their original and expanded descriptions [16,18,[36][37][38][39][40].
All morphological measurements (except SVL) were corrected using the allometric equation X adj = X ± β(SVL ± SVL mean ), where X adj = adjusted value; X = measured value; β = unstandardized regression coefficient of each clade (= species); SVL = measured snout-vent-length; SVLmean = overall average SVL of a given clade (= species) [41][42][43]. Specimens were assigned to group (= species) based on their mtDNA assignment (below). Principal component analysis (PCA) was performed separately by sex using FactoMineR and factoextra package [44,45] in R program v.3.6.1 [46] to assess morphometric differences between groups. Normality of data distribution was tested through the Shapiro-Wilk test and homogeneity of variance was tested by F-test. Morphological differences between species were tested separately by sex using t-test (for equal variance) and Welch t-test (for unequal variance) if the data conforms normality, and Mann-Whitney U test was used for the non-normal distributed data at a significance level of p ≤ 0.05. All statistical analyses were performed in R [46].

Bioacoustic Analyses
Call recordings of L. doriae were made at Mae Hong Son Province (Mueang Mae Hong Son District, Mok Champae Sub-district) and of Limnonectes sp. at Surat Thani Province (Ko Pha-ngan Island) during June-July 2018. Calls of individual males were recorded at night (1900-2330 h) under natural conditions using a Tascam DR-40 Linear PCM Recorder and a Røde NTG-2 condenser shotgun microphone using uncompressed WAV format at a sampling rate of 44.1 kHz and 16-bit resolution. During recording, the microphone was held 20-30 cm from the calling individual. Each recorded male was subsequently collected and preserved as a voucher specimen. Ambient air temperature was measured to the nearest 1 • C immediately after each recording (within approximately 50 cm of the location of the calling frog) using a Kestrel 4000 Weather Meter.
We analyzed nine calls from one male of L. doriae from Mae Hong Son Province and 180 calls (60 calls per individual) of three males of Limnonectes sp. from Surat Thani Province. The acoustic analyses were performed in Raven Pro 64 1.5 for Mac (Cornell Lab of Ornithology, Ithaca, NY, USA). The audio spectrograms were achieved using Hanning windows function (frame width with a 512-point fast Fourier transform, 50% overlap, 86.1 Hz resolution) in the spectrum section plot. Spectrogram figures were generated with seewave package version 2.1.4 [47] in R.
Temporal and spectral parameters measured were modified from Cocroft and Ryan [48] and Thomas et al. [49] as follows: Dominant call frequency (the frequency of maximum amplitude measured from a power spectrum generated using Raven's selection spectrum function over the duration of the entire call); call duration (the time from beginning to end of one call); intercall interval (duration of the interval between consecutive calls); call rate (the total number of calls-1 divided by the time from beginning of first call to beginning of last call); note (the number of notes in a call); note duration (time between the onset and the offset of the middle note in a call); and internote interval (the time interval of the middle notes in a call); note rate (the number of note per second, the total number of notes-1 divided by the time from beginning of first note to beginning of last note). We used call structure description and the terminology of pulses, notes, and calls as defined by Köhler et al. [50] and Emmrich et al. [51]. Comparative advertisement call characters for L. dabanus, L. lauhachindai, and L. savan were taken from the literature [7,8,17].

Phylogenetic Relationships
The aligned dataset contained 2380 mtDNA characters from 71 individuals of Limnonectes species and two individuals of the outgroup species. The standard deviation of Animals 2021, 11, 566 5 of 28 split frequencies was 0.002506 among the two BI runs, and the ESS of all parameters were ≥14,783, indicating that the two runs had been sufficiently sampled and converged. The maximum likelihood value of the best ML tree was lnL = −13,015.988. The 50% majority rule consensus tree from BI analysis and the best ML tree had slightly different topologies, most notably in the phylogenetic position of L. kohchangae (Figures 1 and 2, respectively). However, in both analyses, the new species represented a well-supported monophyletic group (1.0 PP, 100 BS) within a clade containing L. coffeatus, L. doriae, L. hascheanus, L. kohchangae, L. limborgi, L. macrognathus, and L. plicatellus. Although the exact sister relationship of the new species was unresolved, the new species represented a deeply divergent mtDNA lineage that was distinct from other described species.

Morphology
PCA of male specimens revealed morphometric differences with no overlap on a plot of the first two axes between the new species and L. doriae ( Figure 3A). The first two principal components (PC) accounted for a cumulative 63.9% of the total variation and loaded most heavily for EYE adj , IND adj , SHK adj , TGH adj , HND adj , FTL adj , IML adj , TMP adj , and TMP/EYE along PC1 (40.6% of the total variation), for HDL adj , HDW adj , SNT adj , IOD adj , and IMW adj along PC2 (23.2% of the total variation; Table 3). These results indicated that males of the new species had larger tympana (TMP adj , TMP/EYE), but smaller eye (EYE adj ), internarial distance (IND adj ), forelimbs (HND adj ), hindlimbs (SHK adj , THG adj , FTL adj ), and inner metatarsal tubercles (IML adj ) than did males of L. doriae based on scores of the first two axes ( Figure 3A). PCA of female specimens also indicated morphometric differences with no overlap on a plot of the first two axes between the new species and L. doriae ( Figure 3B). The first two PC accounted for a cumulative 81.0% of the total variation and loaded most heavily for the most characters except TMP adj and TMP/EYE along PC1 (66.1% of the total variation), for TMP adj , and TMP/EYE along PC2 (14.9% of the total variation; Table 3). These results indicated females of the new species had smaller body size (SVL), heads (HDL adj , HDW adj ), snout length (SNT adj ), internasal distance (IND adj ), eye (EYE adj ), interorbital distance (IOD adj ), forelimbs (HND adj , LAL adj ), hindlimbs (SHK adj , TGH adj , FTL adj ), and inner metatarsal tubercles (IML adj , IMW adj ) did than females of L. doriae based on scores of the first two axes ( Figure 3B). Animals 2021, 11, x 12 of 30  Statistical comparisons show that most morphological characters were significantly different between the new species and L. doriae at p < 0.05-0.001 (Table 4). Males of these two species were significantly different in SVL, EYE adj , HND adj , and TMP adj having p < 0.05, and the characters IND adj , SHK adj , TGH adj , FTL adj , IML adj , IMW adj , and TMP/EYE with p < 0.001 (Table 4). Females were also significantly different in TMP/EYE having p < 0.05, and the most characters, except TMP adj , and TMP/EYE with p < 0.001 (Table 4). Comparisons of morphometric measurements of adult males and females are given in Tables S2 and S3.

Systematic Account
The samples of Limnonectes from the three localities in southern Thailand differed from L. doriae and congeners in mtDNA, morphology, and advertisement call. On the basis of these three corroborated lines of evidence, the southern Thailand Limnonectes is hypothesized to be a distinct species, and is described as new as follows.

Systematic Account
The samples of Limnonectes from the three localities in southern Thailand differed from L. doriae and congeners in mtDNA, morphology, and advertisement call. On the basis of these three corroborated lines of evidence, the southern Thailand Limnonectes is hypothesized to be a distinct species, and is described as new as follows.
Limnonectes pseudodoriae sp. nov.    Etymology: The specific name pseudodoriae is a noun in apposition. The Greek word pseudo means "false" or "resembling", in reference to the morphological similarity between the new species and L. doriae. Etymology: The specific name pseudodoriae is a noun in apposition. The Greek word pseudo means "false" or "resembling", in reference to the morphological similarity between the new species and L. doriae.

Diagnosis
Limnonectes pseudodoriae sp. nov. is assigned to the genus Limnonectes on the basis of its inferred phylogenetic position (Figures 1 and 2), the presence of fang-like odontoid processes on the lower jaw [6,10], and enlarged heads with hypertrophied jaw musculature in adult males [6]. This species can be distinguished from its congeners by having the following combination of characters: (1)  (2) males lacking cephalic caruncles, but having broadened convex skin with densely grained translucent spinules on the interorbital region from the level of the posterior margin of the eyes to the level of the posterior margin of the tympanum; (3) males with enlarged heads with hypertrophied jaw musculature; (4) two enlarged odontoid processes, triangular, sharp-tipped, angled posteriorly on anterior margin of the lower jaw, larger in adult males than females; (5) tympanum dark brown with yellow mottling, horizontal diameter in adult males greater than that of the eye, in juveniles and females less than that of the eye; (6) dorsum with enlarged, oval tubercles, and slightly elongated tubercles, not arranged in rows; (7) dorsal surface of shank covered with small, distinct, moderately dense, homogenously-sized tubercles tipped with translucent spinules; (8) chin and throat yellowish white with dark brown or gray mottling (9) chest, belly, ventral surface of upper forelimb and groin yellowish white; (10) ventral surfaces of forelimbs and hindlimbs orange-yellow; (11) Toe V with raised, but unmovable dermal ridge along outer margin; (12) inner metatarsal tubercle length approximately 35% that of Toe I; (13) toe webbing formula: I0−2II0−2½III0−(2½-3)IV(2½-3)−0V; and (14) ova with pigmented poles. (Thai).

Diagnosis
Limnonectes pseudodoriae sp. nov. is assigned to the genus Limnonectes on the basis of its inferred phylogenetic position (Figures 1 and 2), the presence of fang-like odontoid processes on the lower jaw [6,10], and enlarged heads with hypertrophied jaw musculature in adult males [6]. This species can be distinguished from its congeners by having the following combination of characters: (1)  (2) males lacking cephalic caruncles, but having broadened convex skin with densely grained translucent spinules on the interorbital region from the level of the posterior margin of the eyes to the level of the posterior margin of the tympanum; (3) males with enlarged heads with hypertrophied jaw musculature; (4) two enlarged odontoid processes, triangular, sharp-tipped, angled posteriorly on anterior margin of the lower jaw, larger in adult males than females; (5) tympanum dark brown with yellow mottling, horizontal diameter in adult males greater than that of the eye, in juveniles and females less than that of the eye; (6) dorsum with enlarged, oval tubercles, and slightly elongated tubercles, not arranged in rows; (7) dorsal surface of shank covered with small, distinct, moderately dense, homogenously-sized tubercles tipped with translucent spinules; (8) chin and throat yellowish white with dark brown or gray mottling (9) chest, belly, ventral surface of upper forelimb and groin yellowish white; (10) ventral surfaces of forelimbs and hindlimbs orange-yellow; (11) Toe V with raised, but unmovable dermal ridge along outer margin; (12)

Description of Holotype
Habitus moderately stocky; body tapering to groin. Head broad and depressed; head width slightly greater than head length. Snout obtusely pointed in dorsal view, rounded in profile, projecting well beyond lower jaw in profile; nostril dorsolateral, below canthus, much closer to tip of snout than eye; internarial distance less than interorbital distance; canthus rostralis indistinct, rounded, slightly constricted behind nostrils; lores concave, oblique; eye diameter 51% of snout length, interorbital distance greater than upper eyelid width; pineal ocellus visible; tympanum round, not elevated from side of head, annulus visible, tympanum diameter approximately 149% eye diameter and greater than distance between tympanum and eye; small, slit-like vocal sac openings on anterior floor of mouth near lateral margin of tongue; vomerine teeth on two short oblique ridges, equal in distance to each other as choanae, beginning on anterior inner edge of choanae and extending toward back; two enlarged odontoid processes, triangular, angled posteriorly on anterior margin of lower jaws; median triangular protuberance at mandibular symphysis.
Forelimbs moderately robust. Fingers moderately slender, without webbing; tips of finger rounded, weakly expanded into discs; relative finger lengths II = IV < I < III; subarticular tubercles rounded, distinct, one on Finger I−II, two on Finger III−IV; thenar tubercle oval, distinct; two distinct oval palmar tubercles in contact at base of Finger III−IV; nuptial pad absent.
Hindlimb moderately robust. Toe moderately slender; tips of toes rounded, expanded into small discs; relative toe lengths I < II < V < III < IV; toe webbing formula I0−2II0−2 1 2 III0−3IV3−0V, webbing on Toe I to base of disc, on preaxial side of Toe II to level of subarticular tubercle and continuing as a fringe to base of tip, on proximal side of Toe II to base of tip, on preaxial side of Toe III to midway between distal and proximal subarticular tubercle and continuing as a fringe to base of tip, on postaxial side of Toe III to base of tip, on preaxial and postaxial sides of Toe IV to level of middle subarticular tubercle and continuing as a fringe to base of tip, and on Toe V to base of tip; raised but immovable dermal ridge along the outer side of the Toe V from base of tip to level of outer metatarsal tubercle; subarticular tubercles prominent; distinct fold on distal two-thirds of tarsus; distinct, elongate, oval, inner metatarsal tubercle, length approximately 35% distance between tip of Toe I and tubercle; no outer metatarsal tubercle.
Skin on snout and side of head shagreened; upper eyelid covered with rounded tubercles; dorsum shagreened covered with enlarged, oval tubercles, and slightly elongated tubercles, most tubercles condensed near flank and lower back; tubercles tipped with single, translucent spinules; caruncle absent, but broadened convex skin with densely grained translucent spinules on interorbital region from level of the posterior margin of the eyes to level of posterior margin of the tympanum; tubercular ridges from posterior margin of eyelid to occiput forming indistinct W-shape; dorsolateral fold absent; distinct supratympanic ridge from posterior corner of eye to axilla; skin on dorsal surfaces of forelimbs and thigh shagreened; small, distinct, moderately dense, homogenously-sized tubercles tipped with translucent spinules on dorsal surfaces of shank; skin on throat wrinkled, that on remaining ventral surfaces smooth; rictal gland absent.

Color of Holotype in Life
Dorsum dark orange brown; W-shaped tubercular ridge and tubercles on upper back with scattered dark brown mottling; snout beige with scattered dark brown mottling; dark brown band between eyes with beige posterior margin; broadened convex skin on interorbital region dark brown; dark brown band under canthus and supratympanic fold, extending from nostril to upper half of tympanum; tympanum dark brown with yellow mottling; upper surfaces of flank dark orange brown with darker mottling; lower surfaces of flank dark yellow. Upper lip orange brown with white mottling alternating with broad dark brown bands, lower lip dark brown with white mottling. Iris bronze with black, vermiform mottling and horizontal bars forming +-shaped marking over eye. Dorsal surfaces of limb with indistinct dark brown crossbands. Chin creamy white and throat yellowish, with gray-brown mottling; chest creamy white; ventral surfaces of upper arm, belly, and upper groin with yellowish white, ventral surfaces of lower arm and hindlimb with yellow orange; ventral surfaces of hands and feet uniformly dark brown ( Figure 5).

Coloration of Holotype in Preservative
Dorsum nearly uniformly brown with indistinct, scattered dark brown mottling, lips with indistinct dark brown broad bands, and dorsal surfaces of limbs with indistinct dark brown cross-bands. Tympanum and forelimbs with lighter brown than remaining dorsum. Interorbital band distinct dark brown. Tubercles on dorsum encircled with dark brown. Chin and throat with dark brown mottling, belly cream, groin yellowish white, and ventral part of hands and feet uniformly dark brown ( Figure 6A). than males; have narrower heads in dorsal view than males ( Figure 6); have relatively smaller tympana than males, with the horizontal diameter of tympanum less than that of eye; have smaller and shorter odontoid processes than males; have less distinct and fewer tubercles on dorsum, flank, and posterior region of shank than males; lack wrinkled skin on the throat; have less mottling on the throat and chin than males; and have ova with pigmented poles.

Morphological Variation
Females lack broadened convex skin on interorbital region; have smaller body size than males; have narrower heads in dorsal view than males ( Figure 6); have relatively smaller tympana than males, with the horizontal diameter of tympanum less than that of eye; have smaller and shorter odontoid processes than males; have less distinct and fewer tubercles on dorsum, flank, and posterior region of shank than males; lack wrinkled skin on the throat; have less mottling on the throat and chin than males; and have ova with pigmented poles.
The largest male in the type series is paratype ZMKU AM 01578 having SVL of 48.2 mm, the second largest is paratype ZMKU AM 01579 having SVL of 48.1 mm, and the third largest is the holotype (Table S2). The degree of toe webbing on the preaxial and postaxial sides of Toe IV varies in some specimens by reaching to midway between the distal and proximal subarticular tubercles before continuing as a fringe to base of tip. Some specimens have more or less mottling on the dorsum, less mottling on the throat, more distinct crossbands on the dorsal surfaces of limbs, and less distinct lip bands than the holotype (Figure 7). Morphological measurements of the type series are summarized in Tables S2 and S3. postaxial sides of Toe IV varies in some specimens by reaching to midway between the distal and proximal subarticular tubercles before continuing as a fringe to base of tip. Some specimens have more or less mottling on the dorsum, less mottling on the throat, more distinct crossbands on the dorsal surfaces of limbs, and less distinct lip bands than the holotype (Figure 7). Morphological measurements of the type series are summarized in Tables S2 and S3.

Habitat, Distribution, and Natural History
Limnonectes pseudodoriae sp. nov. is currently known from only three localities in southern Thailand: Ko Pha-ngan Island (type locality) and Ko Samui Island in the Gulf of Thailand, and Ko Lanta Yai Island in the Andaman Sea (Figure 8). The new species occurs in hill and semi-evergreen forests from 44-322 m elevation, including secondary forests surrounded by agricultural lands (orchards and rubber plantation) and human dwellings. Specimens (n = 37) were collected at night (1900-2230 h) during the early-mid rainy season (May-July) on the forest floor (35.1%; n = 13); in small (≤8 cm wide) clay holes (8.1%; n = 3) associated with a permanent, small (1-4 m wide) stream; on rocky and sandy banks (37.8%; n = 14) of a small, rocky stream; in puddles (16.2%; n = 6); or on small (≤15 cm tall) shrubs (2.7%; n = 1).
At Ko Pha-ngan Island, the holotype was found while calling on the stream bank of Pheang Waterfall ( Figure 9A) during a drizzle rain on 17 July 2018. Other calling males (ZMKU AM 01565, ZMKU AM 01568) and females (not collected) were found in a shallow puddle that was covered with grass in the same area, but these calls were not recorded. Advertisement calls were recorded from two males in a small clay hole under a clump of ferns (ZMKU AM 01579-80) and one male (ZMKU AM 01581) in a small, shallow (<5 cm deep), flowing stream with gravel and clay that was covered with grasses, weeds, and leaf litter during a drizzle rain ( Figure 10) in an orchard near the forest edge of Mueang Mak Mine on 18 July 2018. More than 15 adult males were calling in the same area (ca. 4 m around), but not collected. All collected female specimens were gravid, with many bicolored eggs visible inside the abdominal cavity through the belly skin. Other sympatric frog species found at this locality included Duttaphrynus melanostictus (Schneider, 1799), Fejervarya limnocharis (Gravenhorst, 1829), and L. blythii (Boulenger, 1920).  Table S1. Shaded symbols indicate morphological data only, Shaded symbols with white center dot indicate molecular data only, and shaded symbols with black center dots indicate both molecular and morphological data were studied.
At Ko Pha-ngan Island, the holotype was found while calling on the stream bank of Pheang Waterfall ( Figure 9A) during a drizzle rain on 17 July 2018. Other calling males (ZMKU AM 01565, ZMKU AM 01568) and females (not collected) were found in a shallow puddle that was covered with grass in the same area, but these calls were not recorded. Advertisement calls were recorded from two males in a small clay hole under a clump of ferns (ZMKU AM 01579-80) and one male (ZMKU AM 01581) in a small, shallow (<5 cm deep), flowing stream with gravel and clay that was covered with grasses, weeds, and leaf litter during a drizzle rain ( Figure 10) in an orchard near the forest edge of Mueang Mak Mine on 18 July 2018. More than 15 adult males were calling in the same area (ca. 4 m around), but not collected. All collected female specimens were gravid, with many bicolored eggs visible inside the abdominal cavity through the belly skin. Other sympatric frog species found at this locality included Duttaphrynus melanostictus (Schneider, 1799), Fejervarya limnocharis (Gravenhorst, 1829), and L. blythii (Boulenger, 1920).  Table S1. Shaded symbols indicate morphological data only, Shaded symbols with white center dot indicate molecular data only, and shaded symbols with black center dots indicate both molecular and morphological data were studied.
At Ko Samui Island, the oviposition site of L. pseudodoriae sp. nov. was found at Song Reau Waterfall on 20 July 2018 ( Figure 9B). The eggs were deposited in terrestrial nests on the forest floor, under boulders next to a small, shallow, flowing stream (ca. 1 m wide) at the edge of secondary forest (Figure 11). This site was approximately 5 m away from the main waterfall stream. The egg clutches were laid in a moist, shallow, small clay hole that was covered with leaf litter. We observed eight isolated nests in the same area (1 m wide) at an ambient temperature of 28.1 • C and relative humidity 84.4%. The nests were approximately 5-7 cm in diameter and 3-5 cm deep. The nests contained 17, 18, 19, 25, and 30 eggs (not collected), and 12, 16, and 21 tadpoles (not collected) in the moist clay without free water ( Figure 11B,C) or only a small volume of fluid ( Figure 11D). The recently oviposited eggs were bicolor, with the animal hemisphere dark and the vegetal hemisphere white in color. Each egg was enclosed in a large volume of a clear jelly layer that melded with that of other eggs. When the egg develops, the dark pigmentation is more evenly distributed at its surface ( Figure 11B,C). Tadpoles remained relatively motionless and did not exhibit apparent feeding behavior during their observation at night ( Figure 11D). Moreover, we also found one calling male (ZMKU AM 01583), two non-calling adult males (ZMKU AM 01582, ZMKU AM 01584), and other unvouchered males surrounding the nests (≤1 m from the nest; Figure 11E). All collected female specimens were gravid, with many bicolored eggs visible inside the abdominal cavity through the belly skin. Other sympatric frog species found at this locality included L. blythii and Sylvirana nigrovittata (Blyth, 1856).  At Ko Samui Island, the oviposition site of L. pseudodoriae sp. nov. was found at Song Reau Waterfall on 20 July 2018 ( Figure 9B). The eggs were deposited in terrestrial nests on the forest floor, under boulders next to a small, shallow, flowing stream (ca. 1 m wide) at the edge of secondary forest ( Figure 11). This site was approximately 5 m away from the main waterfall stream. The egg clutches were laid in a moist, shallow, small clay hole that was covered with leaf litter. We observed eight isolated nests in the same area (1 m wide) at an ambient temperature of 28.1 °C and relative humidity 84.4%. The nests were approximately 5-7 cm in diameter and 3-5 cm deep. The nests contained 17,18,19,25, and 30 eggs (not collected), and 12, 16, and 21 tadpoles (not collected) in the moist clay without free water ( Figure 11B,C) or only a small volume of fluid ( Figure 11D). The recently oviposited eggs were bicolor, with the animal hemisphere dark and the vegetal hemisphere white in color. Each egg was enclosed in a large volume of a clear jelly layer that melded with that of other eggs. When the egg develops, the dark pigmentation is more evenly distributed at its surface ( Figure 11B,C). Tadpoles remained relatively motionless and did not exhibit apparent feeding behavior during their observation at night ( Figure 11D). Moreover, we also found one calling male (ZMKU AM 01583), two non-calling adult males

Discussion
The new species morphologically resembles L. doriae, but detailed comparisons of qualitative morphological characters, morphometrics, and statistical analyses revealed clear differences between them. These differences in morphology were corroborated by the finding that the new species represents a deeply divergent mtDNA lineage (both in genetic p-distances and in the molecular phylogeny) compared to its congeners. Last, the new species was readily differentiated from L. doriae in the temporal parameters of the advertisement calls. Therefore, the congruence in mtDNA, morphology, and advertisement calls supported the validity of L. pseudodoriae sp. nov. The phylogenetic analysis in this study revealed that the new species belongs to a clade containing L. coffeatus, L. doriae, L. hascheanus, L. limborgi, L. macrognathus, and L. plicatellus, but unfortunately, the exact sister taxon relationship of the new species was not resolved. These phylogenetic relationships were also unresolved in other studies (e.g., [17]), despite our addition here of the ND3 gene to the more commonly used 16S gene. Additional sequence data (nuclear and/or additional mitochondrial markers) are probably needed to better resolve these relationships.
In Thailand, L. doriae is patchily distributed in mountainous habitats from the northwestern region (Tak Province) to the southern region (Chumphon Province or probably through to Trang Province) of the country [14,15]. In this study, we extended the geographic range of L. doriae northward into Mae Hong Son Province. Some authors have suggested that the southern limits of the geographic range of L. doriae extend throughout southern Thailand and into Peninsular Malaysia [1,[11][12][13]. Limnonectes pseudodoriae sp. nov. is currently known only from three islands in southern Thailand, Ko Pha-ngan and Ko Samui Islands on the Gulf of Thailand coast, and Ko Lanta Yai Island on the Andaman Sea coast. At those localities, we did not find L. doriae. The presence of L. pseudodoriae sp. nov. on the mainland, and the details of its geographic separation from L. doriae, remain unknown. It is possible that records of L. "doriae" from extreme southern Thailand and Peninsular Malaysia are actually L. pseudodoriae sp. nov., and that the two species are separated by the Khlong Mauri Fault south of the Isthmus of Kra [63,64], a known historical barrier to dispersal and stimulant for vicariant speciation in many organisms (e.g., [65-

Discussion
The new species morphologically resembles L. doriae, but detailed comparisons of qualitative morphological characters, morphometrics, and statistical analyses revealed clear differences between them. These differences in morphology were corroborated by the finding that the new species represents a deeply divergent mtDNA lineage (both in genetic p-distances and in the molecular phylogeny) compared to its congeners. Last, the new species was readily differentiated from L. doriae in the temporal parameters of the advertisement calls. Therefore, the congruence in mtDNA, morphology, and advertisement calls supported the validity of L. pseudodoriae sp. nov. The phylogenetic analysis in this study revealed that the new species belongs to a clade containing L. coffeatus, L. doriae, L. hascheanus, L. limborgi, L. macrognathus, and L. plicatellus, but unfortunately, the exact sister taxon relationship of the new species was not resolved. These phylogenetic relationships were also unresolved in other studies (e.g., [17]), despite our addition here of the ND3 gene to the more commonly used 16S gene. Additional sequence data (nuclear and/or additional mitochondrial markers) are probably needed to better resolve these relationships.
In Thailand, L. doriae is patchily distributed in mountainous habitats from the northwestern region (Tak Province) to the southern region (Chumphon Province or probably through to Trang Province) of the country [14,15]. In this study, we extended the geographic range of L. doriae northward into Mae Hong Son Province. Some authors have suggested that the southern limits of the geographic range of L. doriae extend throughout southern Thailand and into Peninsular Malaysia [1,[11][12][13]. Limnonectes pseudodoriae sp. nov. is currently known only from three islands in southern Thailand, Ko Pha-ngan and Ko Samui Islands on the Gulf of Thailand coast, and Ko Lanta Yai Island on the Andaman Sea coast. At those localities, we did not find L. doriae. The presence of L. pseudodoriae sp. nov. on the mainland, and the details of its geographic separation from L. doriae, remain unknown. It is possible that records of L. "doriae" from extreme southern Thailand and Peninsular Malaysia are actually L. pseudodoriae sp. nov., and that the two species are separated by the Khlong Mauri Fault south of the Isthmus of Kra [63,64], a known historical barrier to dispersal and stimulant for vicariant speciation in many organisms (e.g., [65][66][67]). Additional field sampling for L. doriae and L. pseudodoriae sp. nov., in south-ern Thailand, southern Myanmar, and Peninsular Malaysia, as well as re-evaluation of existing museum specimens, is needed to determine the actual geographic ranges of these two species. Moreover, additional field work is needed to understand the full breadth of the ecological requirements of L. pseudodoriae sp. nov. to effectively conserve the new species.

Conclusions
We described a new species of the fanged frog genus Limnonectes (Amphibia, Anura, Dicroglossidae) from southern Thailand, Ko Pha-ngan and Ko Samui Islands (the Gulf of Thailand coast) and Ko Lanta Yai Island (the Andaman Sea coast), using an integrative taxonomic approach based on multiple lines of evidence. Limnonectes pseudodoriae sp. nov. can be differentiated from the most morphologically similar species, L. doriae, and its congeners on the basis of mitochondrial DNA (16S rRNA and ND3), qualitative and quantitative morphology, and advertisement calls. However, the exact sister relationship of the new species remains unresolved, and additional sequence data may be required to resolve these relationships. Limnonectes pseudodoriae sp. nov. is known only from a relatively small geographic area in southern Thailand, where it inhabits mountain streams at mid to low elevations similar to many other fanged frogs. The new species deposits eggs in terrestrial nests near streams that consist of moist, shallow, clay holes covered with leaf litter, and the tadpoles develop in the moist clay of the nest, without free water or in only a small volume of fluid. Additional field sampling is needed to understand the extent of the geographic range and natural history of the new species.
Supplementary Materials: The following are available online at https://www.mdpi.com/2076-2 615/11/2/566/s1, Data S1: Comparative specimens examined, Table S1: Specimens of Limnonectes used in (A) molecular and/or (B) morphological analyses in this study, Table S2: Morphological measurements (mm) of adult male specimens of Limnonectes pseudodoriae sp. nov. and L. doriae. Data are given as mean and standard deviation, followed by range in parentheses, Table S3: Morphological measurements (mm) of adult female specimens of Limnonectes pseudodoriae sp. nov. and L. doriae. Data are given as mean and standard deviation, followed by range in parentheses, and Table S4: Call parameters of three male paratypes of Limnonectes pseudodoriae sp. nov. Parameter values are given as mean ± SD and ranges in parentheses.
Author Contributions: Conceptualization, S.Y. and A.A.; methodology: All authors; data curation, data processing, and formal analysis, S.Y., A.R. and A.A.; validation and investigation, A.A. and B.L.S.; writing original draft preparation, S.Y., A.R. and A.A.; writing review and editing, A.A. and B.L.S. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.