A new species of the genus Leptolalax (Anura: Megophryidae) from southern Vietnam

We describe a new species of megophryid frog from Phu Yen Province in southern Vietnam. Leptolalax macrops sp. nov. is distinguished from its congeners by a combination of the following morphological attributes: (1) body size medium (SVL 28.0–29.3 mm in three adult males, 30.3 mm in single adult female); (2) supra-axillary glands present, creamy white; ventrolateral glands indistinct; (3) tympanum externally distinct; (4) dorsal skin roughly granular with larger tubercles, dermal ridges on dorsum absent; (5) rudimentary webbing present between fingers I–II and II–III; rudimentary webbing between all toes; fingers and toes without dermal fringes; (6) in life ventral surface greyish-violet with white speckling; (7) supratympanic fold distinct, dark brown in life; (8) iris bicolored, typically golden in upper half, fading to golden green in lower half; (9) tibia short (TbL/SVL 0.44–0.45 in males); and (10) eyes large and protuberant (ED/SVL 0.15–0.16 in males). From all congeners for which comparable sequences are available, the new species differs markedly in the 16S rRNA mitochondrial gene sequence (P-distance>5.7%). The new species is currently known only from montane evergreen tropical forests of Song Hinh District, Phu Yen Province, and M’Drak District of Dak Lak Province at elevations of 470–630 m a.s.l.. We suggest the new species should be considered as Data Deficient following the IUCN’s Red List categories. We also report a previously unknown Leptolalax mtDNA lineage from an evergreen tropical forest in the Hoa Thinh District of Phu Yen Province, which may also represent an undescribed species.


INTRODUCTION
Members of the genus Leptolalax Dubois, 1983 (Megophryidae Bonaparte, 1850) are widely distributed from northeastern India and southern China southward to the Southeast Asian mainland and Borneo. Knowledge about Leptolalax species diversity has strikingly increased in recent decades, from only four in 1983 (Dubois, 1983) to 53 recognized species at present, 31 of which (~60% of total species) have been described in the last 10 years (Frost, 2017). In Vietnam, the number of Leptolalax species has increased remarkably from six (Nguyen et al., 2009) to 23 species (Rowley et al., 2016(Rowley et al., , 2017a(Rowley et al., , 2017b within the last decade. However, considering the high morphological similarity of many species within the genus (Rowley et al., 2016) and the poor level of biological exploration of many parts of Indochina, additional taxa likely

Morphological characters
Morphological data were recorded from preserved specimens. Measurements were taken using a digital caliper to the nearest 0.1 mm; the morphometrics of adults and character terminology follow Mahony (2011), Mahony et al. (2013), . Morphometric abbreviations are: snout to vent length (SVL); head width (HW); head length (HL); eye diameter (ED); tympanum diameter (TYD); eye to tympanum distance (E-T); snout length, measured from snout tip to the anterior corner of the eye (E-S); eye to nare distance (E-N); nare to snout distance (N-S); interorbital distance, measured as the narrowest distance between upper eyelids (IO); internarial distance (IN); upper eyelid width (ELW); forearm length (FAL); hand length (HAL); first finger length (FIL); second finger length (FIIL); third finger length (FIIIL); fourth finger length (FIVL); tibia length (TbL); femur length (FeL); foot length (FOL); tibiotarsal articulation to tip of fourth toe distance (TFOL); and inner metatarsal tubercle length (IMT). Additionally, for description of the type series we measured the distance between anterior orbital borders (IFE); distance between posterior orbital borders (IBE); and length of toes I-V (TI-VL). All measurements were taken on the right side of the examined specimen. Sex was determined by gonadal inspection following dissection.
Sexes were separated for subsequent comparisons among samples. One-way analysis of variance (ANOVA) and Duncan's post-hoc tests were used for morphometric comparisons. A significance level of 95% was used of all statistical tests.

DNA isolation, PCR, and sequencing
Total DNA was extracted from muscle tissue using standard phenol-chloroform extraction (Hillis et al., 1996), followed by isopropanol precipitation. We amplified a 454-474-bp length fragment of the 16S rRNA mitochondrial gene, which has been successfully applied for DNA-identification of cryptic diversity within the genus Leptolalax (Poyarkov et al., 2015a;Rowley et al., 2015aRowley et al., , 2016Rowley et al., , 2017aRowley et al., , 2017b. The 16S rRNA-gene fragment was amplified using ScreenMix-HS (Evrogen, Russia) following the manufacturer's instructions. The PCR contained 6 µL of ScreenMix-HS, 21 µL of water, 0.9 µL of each primer at a concentration of 10 pmol/µL, and 1.2 µL of template DNA at a concentration up to ca. 100 ng DNA/µL in a 30 µL reaction volume.
The PCR conditions followed Poyarkov et al. (2015b). The amplification protocols included: 94 • C for 5 min of initial denaturation; followed with 35 cycles of denaturation at 94 • C for 1 min, annealing at 55 • C for 1 min, and extension at 72 • C for 1 min; and a final extension at 72 • C for 10 min. The obtained PCR products were loaded onto 1% agarose gels and visualized in the presence of ethidium bromide in a Dark Reader Transilluminator (Clare Chemical, USA). If distinct bands were produced, they were sent to Evrogen (Moscow, Russia) for subsequent purification and sequencing in both directions. The obtained sequences were checked by eye using chromatogram editor software DNA Baser v4.20.0; primer sequences were removed, and the edited sequences were submitted to GenBank under the accession numbers MG787987-MG787993 (Table 1).

Phylogenetic analyses
For phylogenetic analyses of the L. applebyi species group, we used 32 published sequences of 16S rRNA (Poyarkov et al., 2015a;Rowley et al., 2015aRowley et al., , 2016 and seven newly obtained sequences of Leptolalax sp. from Phu Yen Province (Table  1). In total, a dataset of 39 ingroup sequences was used for the analyses. Sequences of L. ventripunctatus, L. bourreti, L. pluvialis, L. firthi, and L. pictus, representing different species groups within Leptolalax, were used as outgroup taxa following Rowley et al. (2016).
Sequences of 44 specimens of Leptolalax representatives, with a total length of up to 1 046 bp, were included in the final alignment and subjected to phylogenetic analyses. Sequences were initially aligned using ClustalW (Thompson et al., 1997) in Bioedit 7.0.5 (Hall, 1999) with default parameters. Mean uncorrected genetic distances (P-distances) between sequences and species were calculated using MEGA 7.0 (Kumar et al., 2016). PartitionFinder v.1.1.0 (Lanfear et al., 2012) was applied to estimate the optimal evolutionary models used for dataset analysis. The best-fitting model was the GTR+I+G model of DNA evolution, as suggested by the Akaike Information Criterion (AIC), corrected Akaike Information Criterion (AICc), and Bayesian Information Criterion (BIC).
The matrilineal genealogy was inferred using Bayesian inference (BI) and maximum likelihood (ML) algorithms. The BI analyses were conducted in MrBayes v.3.1.2 (Huelsenbeck & Ronquist, 2001;Ronquist & Huelsenbeck, 2003); Metropolis-coupled Markov chain Monte Carlo (MCMCMC) analyses were run with one cold chain and three heated chains for ten million generations and sampled every 1 000 generations. Five independent MCMCMC runs were performed and 1 000 trees were discarded as burn-in. Confidence in topology was assessed by posterior probability (BI PP, Huelsenbeck & Ronquist, 2001). The ML analyses were conducted using Treefinder (Jobb et al., 2004) and confidence in node topology was tested by non-parametric bootstrapping with 1 000 replicates (ML BS, Felsenstein, 1985). We a priori regarded tree nodes with bootstrap (ML BS) values of 70% or greater and Bayesian posterior probabilities (BI PP) values over 0.95 as sufficiently resolved (Felsenstein, 2004;Hillis & Bull, 1993;Huelsenbeck & Hillis, 1993). The ML BS values between 70% and 50% (BI PP between 0.95 and 0.90) were treated as tendencies and nodes with ML BS values below 50% (BI PP below 0.90) were regarded as unresolved.

Sequence variation
The 16S rRNA dataset contained 39 ingroup and five outgroup Leptolalax sequences. The final alignment consisted of 1 075 sites, with 617 conserved sites and 413 variable sites, 139 of which were parsimony-informative; the transition-transversion bias (R) was estimated as 2.14 (all data given for ingroup only). Substitution rates were estimated under the General Time Reversible (GTR) model (+I+G). Nucleotide frequencies were A=31.23%, T=24.47%, C=24.27%, and G=20.03%.

Phylogenetic relationships
Phylogenetic analysis results of the 16S rRNA gene fragment are shown in Figure 2. The ML and BI phylogenetic analyses showed essentially similar topologies, which only differed slightly in associations at poorly supported basal nodes.
In general, the topology of the BI cladogram was consistent with results reported in previous work (Poyarkov et al., 2015a;Rowley et al., 2015aRowley et al., , 2016, suggesting monophyly of the L. applebyi species group (node support values 1.0/99, hereafter given for BI PP/ML BS, respectively) and the presence of two major lineages within it. Clade I encompassed three species inhabiting the Tay Nguyen (Kon Tum) Plateau in central Vietnam and northeastern Cambodia: namely, L. applebyi, L. ardens, and L. melicus (Figure 2). Clade II comprised the remaining L. applebyi group species from the Langbian (Da Lat) Plateau of the Southern Annamite Mountains. Phylogenetic relationships within Clade II were not sufficiently resolved: there was a tendency toward a more distant position for L. pyrrhops and L. maculosus, with the remaining lineages forming a monophyletic group (0.95/70). The two newly discovered populations of Leptolalax sp. from Phu Yen Province formed two independent mtDNA matrilines: that is, the Hon Den Mt. lineage and Hoa Thinh lineage. The sequence of Leptolalax sp. from Dak Lak Province (indicated as "molecular lineage 7" in Rowley et al., 2015a) shared the same mtDNA haplotype as the Leptolalax sp. population from Hon Den Mt., Phu Yen Province, suggesting that these two populations are conspecific.

Genetic distances
The uncorrected P-distances among and within the 16S rRNA gene fragment sequences of the studied Leptolalax species are shown in Table 2. The observed interspecific distances within the L. applebyi group members ranged from 4.4% (between L. kalonensis and L. pallidus) to 10.3% (between L. applebyi and L. pyrrhops) of substitutions. The uncorrected genetic P-distances in the ingroup and outgroup comparisons partly overlapped: genetic distances between the L. applebyi group members versus the Leptolalax taxa outgroup ranged from 9.2% (between L. maculosus and L. pluvialis) to 15.4% (between L. kalonensis and L. firthi).
The newly discovered population of Leptolalax sp. from Hon Den Mt., Song Hinh District, was clearly distinct from all other group members in the examined 16S rRNA fragment sequences and appeared to be most closely related to L. bidoupensis from the eastern edges of Langbian Plateau (Lam Dong and Khanh Hoa provinces) and to a Leptolalax sp. population from Hoa Thinh, Tay Hoa District (Phu Yen Province) (P-distance=5.7% for both comparisons). The Leptolalax sp. population from Hoa Thinh was genetically closer to L. bidoupensis and L. pallidus, with a P-distance value of 4.5% (both species from eastern Langbian Plateau).
The observed pairwise divergence in 16S rRNA was greater than that usually seen among species of anurans (Vences et al., 2005a(Vences et al., , 2005bVieites et al., 2009) and was higher than distances between some other recognized species of the 1L. applebyi group (e.g., 4.4% between L. pallidus and L. kalonensis and 4.2% between L. ardens and L. melicus) ( Table 2).
Intraspecific genetic P-distances were 0.0% in the Leptolalax sp. population from Hoa Thinh, and 0.1% in the Leptolalax sp. from Hon Den Mt.; the five examined specimens of the latter species of Leptolalax from Dak Lak and Phu Yen provinces had two haplotypes of the 16S rRNA gene fragment.

Taxonomy
Our molecular data clearly indicated that the two recently discovered populations of Leptolalax sp. from Song Hinh (Hon Den Mt.) and Tay Hoa (Hoa Thinh) districts of Phu Yen Province belong to two independent mtDNA lineages, clearly distinct from each other and from the remaining members of the L. applebyi species group. Despite geographical proximity (~30 km between Hon Den Mt. and Hoa Thinh), these two localities cradle distinct species of Leptolalax, and both appear to be new to science. These two potentially new species were assigned to the Langbian Plateau clade of the L. applebyi species group and appear to be closely related to L. pallidus, L. kalonensis, and L. bidoupensis. At the same time, the population of Hon Den Mt. appears to be conspecific to a Leptolalax sp. found in the eastern part of Dak Lak Province (~30 km between localities).
Lacking enough material for morphological comparisons, we tentatively indicate the Leptolalax sp. population of Hoa Thinh (Tay Hoa District) as a candidate new species sensu Vieites et al. (2009); further morphological and acoustic studies are necessary to clarify its taxonomic status. Based on genetic differentiation, phylogenetic analyses of a 16S rRNA fragment of mtDNA, and analyses of diagnostic morphological characters (see below in "Comparisons"), the population of Leptolalax from Hon Den Mt. in Phu Yen Province of southern Vietnam clearly represents a new species, which we describe as follows.
Etymology: Specific epithet "macrops" is a noun in the nominative case, derived from Greek "macros" for "large" and "ops" for "eye", in reference to its comparatively large eye size.

Recommended vernacular names:
We recommend "Big-eyed Litter Frog" as the common English name of the new species and the common name in Vietnamese as "Cóc mày mát to".

Description of holotype:
Medium-sized Leptolalax specimen (SVL 28.0 mm); body and head in good state of preservation, fingers and toes partially dehydrated due to ethanol preservation ( Figure 3A,B). Left thigh of holotype damaged, skin on ventral surface of thigh dissected, with a significant portion of femoral muscle removed, dissection length ca. 10 mm. Belly also dissected medially, dissection length ca. 9 mm, testes can be seen through dissection.
Head: Head flattened, slightly longer than wide (HW/HL 92.7%), top of head weakly concave; snout short (E-S/HL 34.2%), slightly projecting beyond margin of lower jaw; slightly truncated in dorsal view ( Figure 3A), obtusely rounded in ventral view ( Figure 3B), gently sloping and rounded in profile ( Figure 3C); nostril ovoid, oblique, slightly closer to tip of snout than to eye ( Figure 3C; N-S/E-N 72.1%); canthus rostralis distinct, bluntly rounded; loreal region slightly concave; eyes very large (ED/HL 40.3%), eye diameter greater than snout length (ED/E-S 117.8%), notably protuberant in dorsal view in life (eyeballs depressed down in preserved holotype, Figure  3C); pupil vertical, diamond-shaped; tympanum distinct, round with vertical diameter equal to horizontal diameter; tympanum small, less than half eye diameter (TyD/ED 44.0%); tympanic rim indistinct, not elevated relative to skin of temporal region; pineal ocellus absent; vomerine teeth absent; vocal sac gular, vocal sac openings small, oval, and slit-like, located laterally in corners of mouth floor; tongue long, wide, with free posterior end, heart-shaped with shallow medial notch at posterior tip; supratympanic fold well-developed forming distinct glandular ridge, running from posterior corner of eye posteriorly toward dorsal edge of tympanum, gently curving ventrally toward axilla, bearing several flat tubercles ( Figure 3C).
Forelimbs: Forelimbs thin, slender; finger tips in life rounded, but appear slightly enlarged and truncate in preservative due to partial dehydration, finger tips approximately same width as distal finger articulation; relative finger lengths: IV=I<II<III; nuptial pad indistinct; subarticular tubercles absent, replaced with low dermal ridges prominent on fingers II-IV; inner metacarpal tubercle large, fused with outer one, forming single bulging callous structure, prominent on palmar surface (maximal length 1.3 mm); border between inner and outer metacarpal tubercles indistinct; fingers in life lack dermal fringing, basal webbing present between fingers I and II and fingers II and III, absent between fingers III and IV (Figures 3D, 4A).
Hindlimbs: Hindlimbs slender, short, tibia less than half snout-vent length (TbL/SVL 44.5%); tibiotarsal articulation of adpressed limb reaching eye-level; toe tips round in life, slightly truncate in ethanol preservative due to partial dehydration; relative toe lengths: I<V<II<III<IV; subarticular tubercles absent, replaced by dermal ridges, distinct on all toes and continuing to metatarsus of toes III-V; inner metatarsal tubercle large, oval-shaped, nearly two times longer than wide (IMT to width of inner metatarsal tubercle ratio 186.4%), outer metatarsal tubercle absent; toes without lateral dermal fringes; basal webbing present between all five toes, webbing well-developed between toes I and II, II and III, and III and IV (reaching level of proximal finger articulation), and somewhat reduced between toes IV and V ( Figures 3E, 4B).

Skin texture and skin glands:
Skin on entire dorsum roughly granular, covered in tubercles of varying sizes, smaller dorsolaterally; upper eyelids with numerous small rounded tubercles (flattened in preservative, Figure 3C), snout smooth; ventral skin smooth; pectoral gland distinct in preservative and in life, round, located near axilla, 0.9 mm in diameter ( Figure  3B); femoral gland oval, small, 0.7 mm in diameter, located on posteroventral surface of thigh, approximately five times closer to knee than to vent; supra-axillary gland present, protuberant, creamy white, located in axillary region dorsally from insertion of forelimb, 0.9 mm in diameter; ventrolateral glands indistinct.

Coloration in life:
Dorsal surfaces of head and trunk dark brownish-grey with indistinct dark brown blotches scattered on posterior part of dorsum and between eyes; interorbital region with dark bar with indistinct edges; several light brown blotches of irregular shape and indistinct edges on anterior part of upper eyelids, scapular region, and sacrum. Dorsal surfaces of forelimbs and hindlimbs brownish-grey, elbows and upper arms dorsally much lighter with coppery orange background. Dark brown line running along canthus rostralis through eye, and continuing below supratympanic fold, terminating above axilla, encompassing nare, loreal region but not tympanum; tympanum lighter than surrounding skin of temporal region. Faint transverse dark brown bars on dorsal surface of thighs, tibia, tarsus, lower arms, fingers, and toes. Small indistinct dark brown blotches on flanks. Tiny whitish flecks scattered on dorsolateral sides of body from groin to axilla. Belly and chest greyish-violet with rare white speckling on entire ventral surface, including throat, arms, and legs. Supra-axillary gland creamy white; femoral glands whitish; pectoral glands white. Iris bright orange-gold with greenish tint in lower half and fine black reticulations throughout. Iris periphery lined with black. Sclera light yellowish-green.

Coloration in preservative:
In preservative, coloration of holotype significantly faded to light brown on dorsum and flanks, with slightly paler limbs and beige on ventral sides ( Figure  3B); dark markings on dorsal surfaces brownish, dark banding on dorsal surface of tibiotarsus, antebrachium, hands, and feet well-discernable ( Figure 3A). Elbows and upper arms pale brown. White speckles on ventral surface not discernable. Macroglands creamy white.
Variation: All individuals in the type series were generally similar in morphology and body proportions; measurements of the type series are shown in Table 3 and representative photograph of male paratype in life is shown in Figure 5. Eyes were notably protuberant in living specimens ( Figure 5). All specimens showed certain variation in darker brown patterns on dorsum and dark bands on shanks, forearms, hands, and feet. The single known female (PYU DTD-509) was slightly larger (SVL 30.3 mm) than the holotype and two paratype males. Skin texture appeared to be much less tuberculate in preservative (Figure 3)      Leptolalax macrops sp. nov. is most similar to members of the L. applebyi species group inhabiting the Central Highlands of central and southern Vietnam and the northeastern part of Cambodia, including L. applebyi, L. ardens, L. bidoupensis, L. kalonensis, L. maculosus, L. melicus, L. pallidus, L. pyrrhops, and L. tadungensis. Superficially, the new species resembles L. pyrrhops, another medium-sized member of the L. applebyi species group with large eyes, distributed in the western part of Langbian Plateau (Lam Dong Province) (Poyarkov et al., 2015a). Comparisons of the new species with members of the L. applebyi species group are thus appropriate.

Conservation status:
To date, the new species is only known from a small montane area on the border of Dak Lak, Khanh Hoa, and Phu Yen provinces. It is likely that the range of Leptolalax macrops sp. nov. is quite narrow. The species probably inhabits Ea So Nature Reserve (Dak Lak Province); however, additional research in this area is needed. The new species appears to require closed evergreen forest along the streams where it occurs. Areas of low to middle elevation montane tropical forest are greatly endangered in the southern coastal areas of Vietnam, including Phu Yen Province. Given the available information, we suggest the species should be considered as Data Deficient following the IUCN's Red List categories (IUCN, 2001) until the distribution and habitat requirements of the new species are more fully documented.  Frogs were recorded along a cascading stream.

DISCUSSION
Our molecular data revealed hidden diversity of the L. applebyi species group, with additional herpetological surveys in mountain areas of Indochina possibly leading to the discovery of further new lineages and species of Leptolalax. Our finding brings the number of recognized species of the genus Leptolalax to 54, and the number of Leptolalax species known from Vietnam to 24.
The herpetofauna of the Phu Yen Province is poorly studied compared to the adjacent provinces of Dak Lak and Khanh Hoa. Nguyen et al. (2009) Do et al. (2017b) reviewed the available data on amphibian species found in Phu Yen Province, and added eight new provincial records and listed 33 species of amphibians for the province. The present paper describes a new species of Leptolalax from the Song Hinh District. We also recorded a previously undescribed lineage of Leptolalax sp. from Tay Hoa District of Phu Yen Province; however, further morphological and molecular research is required to clarify the taxonomic status of this population.
Tropical forests are greatly endangered throughout Southeast Asia, including Vietnam.
Compared with the hard-to-access montane tropical forests in the Annamite Mountains, evergreen tropical forests in lowland and foothill areas of the southern coastal region of Vietnam are more endangered; most areas of lowland tropical forest are already destroyed due to logging and other human activities (De Koninck, 1999;Laurance, 2007;Meijer, 1973;Meyfroidt & Lambin, 2008). However, despite their accessibility, the many remaining patches of tropical forest could cradle still unknown biodiversity, which makes the need for biological exploration in this region even more urgent.

ADDENDUM
During the revision process of the present manuscript, a new paper by Nguyen et al. (2018)  We were unable to include L. rowleyae in the comparisons section or phylogenetic analysis in the present manuscript; however, it is markedly distinct from the new species in a number of morphological attributes: by smaller body size: SVL 23.4-25.4 mm in males and 27-27.8 mm in females (vs. SVL 28.0-29.3 mm in adult males and 30.3 mm in single adult female of the new species); by pinkish milk-white to light brown ventral surface with numerous white speckles (vs. greyish-violet ventral surface with rare white speckling in the new species); and by much smaller eyes, ED/SVL 0.08-0.11 in males (vs. ED/SVL 0.15-0.16 in males of the new species) (data from Nguyen et al., 2018). The new species can also be distinguished from L. rowleyae by deep divergence in the 16S rRNA mtDNA gene (P-distance 12.60%) and phylogenetic position (the new species is mentioned as "Leptolalax sp." in the work of Nguyen et al., 2018: Figure 1).
In addition, a recently accepted manuscript by Chen et al. (2018) (published online on 10 March 2018) provides a novel multi-locus phylogenetic hypothesis for the genus Leptolalax, describing the latter as a synonym of the genus Leptobrachella Smith, 1925. Due to the simultaneous review period of the present paper and the work of Chen et al. (2018), we were unable to implement the new taxonomy at the stage of submission and reviewing process. We suggest that the new species Leptolalax macrops sp. nov. should hereafter be referred to as Leptobrachella macrops Duong, Do, Ngo, Nguyen & Poyarkov to reflect the revised taxonomy.