Taxonomic Review of the Genus Herpetoreas (Serpentes: Natricidae), with the Description of a New Species from Tibet, China

: The taxonomic history of the Amphiesma sensu lato has long been confused, and this complex was recently divided into three genera, i.e., Amphiesma Dum é ril, Bibron, and Dum é ril, 1854 sensu stricto , Hebius Thompson, 1913, and Herpetoreas Günther, 1860. Being the least known genus, Herpetoreas is reviewed herein through an integrative taxonomic approach. Our results indicate that specimens previously referred to Hebius parallelus (Boulenger, 1890) from M ê dog, Tibet, China, represent a new species. We describe here this new species and refer it to the genus Herpetoreas ; therefore, we remove Hebius parallelus from the Chinese herpetofauna. A diagnostic key to all the species of Herpetoreas is also provided. Furthermore, we re-evaluate the diagnostic characters of the three genera formerly confused with Amphiesma , namely, Amphiesma , Hebius , and Herpetoreas . We provide a key to these three morphologically similar genera. We also emphasize the importance of the maxillary teeth and hemipenial morphology in the generic diagnosis in the family Natricidae.

Maximum likelihood (ML) analysis was conducted in RAxML v8.2.10 [31] under the best-fit substitution model (GTRGAMMA) based on the AIC criterion, as implemented in PartitionFinder 2 [32]. Bootstrap proportions (BSP) were investigated with 1000 bootstrap Diversity 2022, 14, 79 5 of 38 replicates using the fast-bootstrapping algorithm, otherwise under default parameters. Bayesian inference (BI) analyses were performed in MrBayes 3.2.6 [33], two Markov chain Monte Carlo (MCMC) initiated from random trees and run for 1 × 10 7 generations and sampled every 100 generations, the first 25% of trees were discarded as "burn-in". Convergence of MCMC chains and effective sampling size (ESS) were checked in Tracer 1.6 [34], ensuring a sample size > 200 for all parameters.

Phylogenetic Analyses
The final alignment of the three gene dataset consisted of 2149 bp, and includes 1074 bp from Cyt b, 503 bp from 16S, and 572 bp from CMOS. GTR+G was selected as the best-fit model for both ML and BI analysis. For phylogenetic analyses, a majority rule consensus tree inferred from BI was consistent with the ML tree, and the topologies of our results are largely agreeing with previous analysis (Figure 1) [3,8,10,13]. Among our sample, three genera of Amphiesma sensu lato were recovered as strongly supported clades within the family Natricidae, that were highly diverged with long basal branch lengths. The genus Herpetoreas was recovered as monophyletic with current samplings (BSP = 97 / BPP = 1.00). Although interspecific relationships among congeners within Herpetoreas were not fully resolved, the final consensus tree yielded high support-  Among our sample, three genera of Amphiesma sensu lato were recovered as strongly supported clades within the family Natricidae, that were highly diverged with long basal branch lengths. The genus Herpetoreas was recovered as monophyletic with current samplings (BSP = 97/BPP = 1.00). Although interspecific relationships among congeners within Herpetoreas were not fully resolved, the final consensus tree yielded high supportive values (BSP > 70; BPP > 0. 95) for key nodes concerning the relationship between the Tibetan population with recognized congeners. The Tibetan population was recovered as a highly diverged monophyletic clade (BSP = 100/BPP = 1.00) within Herpetoreas, which was sister to Herpetoreas burbrinki (BSP = 92/BPP = 1.00). The sequences considered to be Herpetoreas cf. parallelus (KIZ 06681) by Che et al. [13] showed no genetic divergence to our newly sequenced Tibetan population (marked in red), and we consider them to be conspecific. The sample of Hebius clerki (CAS 215036, marked in blue) from Nujiang County, Yunnan Province, China, previously misidentified as Hebius parallelus, was nested within the genus Hebius ( Figure 1; Table 1).
Although the Tibetan population was previously identified as Hebius parallelus, morphological comparisons indicate that the Tibetan population is easily distinguished by a suit of characters against Hebius parallelus (see Comparison below). Furthermore, the Tibetan population is recovered to be a distinct evolutionary lineage of the genus Herpetoreas (Figure 1), and morphologically differs from all known species of this genus (Table 2). Therefore, we describe the Mêdog population of Herpetoreas as a new species herein. On the generic level, we review the taxonomic history of Herpetoreas. Moreover, in despite of the fact that phylogenetic definitions of Amphiesma sensu lato have been clarified [3], the morphological difference of Amphiesma sensu lato still remains unknown. Based on morphological comparisons of specimens from Amphiesma sensu lato, we provide diagnostic characters and a diagnostic key to the genera Amphiesma, Hebius, and Herpetoreas.   [37] as "Maxillary teeth are 21 in continuous series, last 15 enlarged", which is a highly isolated condition in the genus Herpetoreas, further confirmation is needed. 2 Given in Das et al. [8] as "a small gap but without diastema", whereas we consider that the "small gap" is homologous with diastema, thus referred both of them as "diastema" in this work.  scale rows of Her. platyceps are feebly keeled only on five to seven mid-dorsal scale rows without exception. However, Her. platyceps specimens from Gyirong County, Tibet, China are distinctly differ from descriptions provided by Malnate [35] in having stronger keeled dorsal scales along the anterior body ( Figure 3B). Further phylogenetic work is recommended to incorporate the topotype of Her. platyceps to clarify whether this variation is caused by cryptic diversity or intraspecific variation. In addition, see comments under Her. sieboldii.   19-19-17; weekly keeled only on five to seven mid-dorsal scale rows, namely, on the 3rd-9th to 5th-9th dorsal scale rows plus the vertebral row; (4) ventrals 180-234; (5) cloacal plate and subcaudals divided; (6) subcaudals 78-99; (7) supralabials 8, 3rd, 4th and 5th entering orbit; (8) maxillary teeth 18-22, slightly enlarged posteriorly, last two distinctly enlarged, separated from anterior teeth by small diastema; (9) hemipenis short and thin, shallowly bilobed, spinous throughout with single basal hook; (10) sulcus spermaticus single, centripetal, extends to the tip of inner right lobe; (11) the upper edge of the supralabials bordered by dark postocular streak; (12) dorsal color highly variable, dorsum frequently speckled with small dark spots; (13) venter yellowish-beige, each ventral decorated with a pair of dark spots at lateral edges or not.
Comments. Although Herpetoreas platyceps has been widely recorded across southern Asia, few studies have focused on the cryptic diversity of this highly variable species. For example, previously considered a junior synonym of Her. platyceps, the type species of Herpetoreas, Her. sieboldii is morphologically similar to Her. platyceps [6,35]. Malnate [35] revised and re-described both Her. sieboldii and Her. platyceps, indicating that the dorsal scale rows of Her. platyceps are feebly keeled only on five to seven mid-dorsal scale rows without exception. However, Her. platyceps specimens from Gyirong County, Tibet, China are distinctly differ from descriptions provided by Malnate [35] in having stronger keeled dorsal scales along the anterior body ( Figure 3B). Further phylogenetic work is recommended to incorporate the topotype of Her. platyceps to clarify whether this variation is caused by cryptic diversity or intraspecific variation. In addition, see comments under Her. sieboldii.    19-19-17; all keeled but some may not be keeled in outermost rows; (4) ventrals 168-216; (5) cloacal plate and subcaudals divided; (6) subcaudals 81-111; (7) supralabials 8, 3rd, 4th and 5th entering orbit; (8) maxillary teeth 19-23, slightly enlarged posteriorly, last two distinctly enlarged, separated from anterior teeth by small diastema; (9) hemipenis short and thin, shallowly bilobed, spinous throughout with single basal hook; (10) sulcus spermaticus single, centripetal, extends to the tip of lobe; (11) a light, dark-bordered crescent extends from the last supralabial up and back over the nape; (12) dorsum frequently speckled with small light spots; (13) venter often darkened posteriorly or usually with a lateral series of dark streaks.
Comments. The genus Herpetoreas has long been regarded as a junior synonym of Amphiesma [1], of which the species Her. sieboldii has long been synonymized with Her. platyceps [6,84]. Although Malnate [35] revalidated sieboldii as a separate species and provided distinctive characters between Her. sieboldii and Her. platyceps, the number of ventrals in Her. sieboldii listed by Malnate [35] is distinctly lower than its original description (168-207 vs. 216) [38], thus further pholidosis confirmation is needed especially for the holotype. Guo et al. [3] revalidated the genus Herpetoreas to accommodate the following three species: Her. sieboldii, Her. platyceps, and Her. burbrinki. However, no genetic data are available for Her. sieboldii to our knowledge [7,8,13]. Further study is also recommended to clarify the relationships between Her. sieboldii and Her. platyceps. In addition, see comments under Her. platyceps.
David et al. [11] reported three specimens (CAS 177474, 177672-177673,) of Her. sieboldii from Zhangmu, Xizang Autonomous Region (Tibet), China. However, no morphological data was provided and the species is previously not listed from Chinese herpetofauna [98]. The distribution record from Myanmar provided by Malnate [35] was based on a single specimen (BNHS 80.4) from Shan State (Figure 2), which was considered to be highly isolated. We question the Burmese record of Her. sieboldii pending further confimation. available for Her. sieboldii to our knowledge [7,8,13]. Further study is also recommended to clarify the relationships between Her. sieboldii and Her. platyceps. In addition, see comments under Her. platyceps. David et al. [11] reported three specimens (CAS 177474, 177672-177673,) of Her. sieboldii from Zhangmu, Xizang Autonomous Region (Tibet), China. However, no morphological data was provided and the species is previously not listed from Chinese herpetofauna [98]. The distribution record from Myanmar provided by Malnate [35] was based on a single specimen (BNHS 80.4) from Shan State (Figure 2), which was considered to be highly isolated. We question the Burmese record of Her. sieboldii pending further confimation.     (7) supralabials 9, 4th and 5th entering orbit; (8) maxillary teeth 13-21, slightly enlarged posteriorly, last two distinctly enlarged, separated from anterior teeth by small diastema; (9) hemipenis short and thin, shallowly bilobed, spinous throughout with single basal hook; (10) sulcus spermaticus single; (11) dorsum dark brown above with paler and darker spots; (12) venter yellowish-beige, each ventral decorated with a pair of dark spots at lateral edges that are connected to form a continuous ventrolateral line.

Herpetoreas xenura
Comments. Herpetoreas burbrinki is still known only from two specimens to our best knowledge [37]. The holotype of Her. burbrinki (YBU 071128) was firstly reported as Amphiesma cf. craspedogaster from Tibet, China by Guo et al. [119], and its measurement provided by Guo et al. [3] is consistent with Guo et al. [119], i.e., SVL 495 mm, TaL 130 mm. However, we re-examined the holotype of Her. burbrinki (YBU 071128) and found our observations are mostly in agreement with previous studies for most of characters except for a longer SVL (527 mm vs. 495 mm) and a distinctly longer tail (TaL 230 mm vs. 130 mm; TaL/TL 0.304 vs 0.208) [3,119]. The measurements of the holotype of Her. burbrinki provided by the current study were based on our data, which were also similar to the newly reported female specimen Tal/TL 0.304 vs. 0.285 (ANU 20210006) [37]. (7) supralabials 8-9, 3rd-5th or 4th-6th entering orbit; (8) maxillary teeth 20-21, slightly enlarged posteriorly, last two distinctly enlarged, separated from anterior teeth by small diastema; (9) hemipenis short and thin, shallowly bilobed, spinous throughout with a single basal hook; (10) sulcus spermaticus single, centripetal, extends to crotch; (11) a light, dark-bordered crescent extending from the last supralabial up and back over the nape; (12) dorsum reddish-brown in life, speckled with short dark stripes; (13) venter reddishorange in life, each ventral decorated with dark spots at lateral edge (Figure 7). Position of the reduction to 6 suprascale rows around the tail at 20th SC. Range, in number of subcaudals spanned, of the portion of the tail with 6 supracaudal scale rows from 20th-44th SC, 25 subcaudals in total; range of the length of tail with 4 supracaudal scale rows: 47th-87th SC, 41 subcaudals in total; ratio: length with 4 supracaudal scale rows/length of the portion of tail with 6 supracaudal scale rows: 1.64. supralabials 8-9, 3rd-5th or 4th-6th entering orbit; (8) maxillary teeth 20-21, slightly enlarged posteriorly, last two distinctly enlarged, separated from anterior teeth by small diastema; (9) hemipenis short and thin, shallowly bilobed, spinous throughout with a single basal hook; (10) sulcus spermaticus single, centripetal, extends to crotch; (11) a light, darkbordered crescent extending from the last supralabial up and back over the nape; (12) dorsum reddish-brown in life, speckled with short dark stripes; (13) venter reddish-orange in life, each ventral decorated with dark spots at lateral edge (Figure 7). Dorsal side of head scales complete including 2 internasals, 2 prefrontals, 2 supraoculars, 1 frontal, and 2 parietals. Rostral wider than high, width approximately twice as long as high, visible from above; nasals subpentagonal, slightly elongate, about 1.2 times longer than high, completely divided, anterior part slightly larger than the posterior one; nostrils located on the anterior part of nasals; internasals subtriangular, in broad contact with each other, not in contact with loreal, about equal length to width, distinctly narrowing anteriorly, the width of the posterior margin approximately twice as long as the anterior margin; 2 prefrontals, subhexagonal, slightly broader than long, about 1.3 times longer than internasals; prefrontal sutures about equal length to internasal sutures; frontal pentagonal, elongate, length 1.45 times longer than width, about twice as long as prefrontal; supraocular 1/1, rectangular, elongate, narrowed anteriorly, about 2.1 times longer than width; parietals large, length 1.95 times longer than width, 1.5 times longer than frontal; parietals in broad contact with each other, parietal suture subequal to length of frontal. Loreal 1/1, subquadrangle, small, about equal in length to depth, in contact with 2nd-3rd supralabials, not entering orbit; preoculars 1/1, depth about 1.7 times longer than width, about two times higher than loreal; postoculars 3/3, size decreasing from top to bottom; supralabials 9/8, 2nd-3rd in contact with the loreal, 4th-6th entering orbit on the left side, 3rd-5th entering orbit on the right side, penultimate one largest; temporals in 3 rows, elongate, variable and asymmetric, anterior temporals 2 (left) and (1 + 1/1) (right), broadly in contact with supralabials 7-8 on the left side, supralabials 6-7 on the right side, followed by 2 + 2 (left) and 1 + 2 (right) posterior temporals; mental subtriangular, width 2.5 times longer than high; 10/9 infralabials, first pair in contact with each other after mental, 1st-5th/1st-4th in contact with anterior chin shields; chin shields pairs two; posterior chin shields 1.3 times longer than anterior ones, separated from each other by small scales; mental groove apparent ( Figure 8). Maxillary teeth 19 + 2, slightly enlarged posteriorly, last two distinctly enlarged, separate from anterior maxillary teeth by a small diastema, not grooved.
The left side of hemipenis is missing, the description of hemipenis in retracted condition is based on the right side. The organ is shallowly bilobed, retracted hemipenis extending to SC 7 with its crotch extending to SC 6; the fork point of m. retractor penis magnus extending to SC 8, origin of m. retractor penis magnus at level of SC 21.
Coloration in preservative. Dorsal surface of head overall creamish-beige, yellowish-brown or blackish-brown, usually more distinctly paler on anterior half of head (Figure 9). Upper surfaces of internasals, anterior sides of prefrontals, and central part of frontal yellowish-beige, sparsely decorated with irregular blackish-brown spots. Upper cephalic scales irregularly and incompletely edged with blackish brown; these dark outlines of dorsal head scales usually become more distinct on their posterior margins. Lateral sides of parietals distinctly darker, almost forming two thick lateral stripes along the outer margins. Parietals speckled with dark brown, a pair of elliptic beige spots symmetrically present on inner sides of corresponding scales, just separated by parietal sutures. A six-scales-long, two-scales-wide beige sagittal line extends just behind parietal suture. Maxillary teeth 19 + 2, slightly enlarged posteriorly, last two distinctly enlarged, separate from anterior maxillary teeth by a small diastema, not grooved.
The left side of hemipenis is missing, the description of hemipenis in retracted condition is based on the right side. The organ is shallowly bilobed, retracted hemipenis extending to SC 7 with its crotch extending to SC 6; the fork point of m. retractor penis magnus extending to SC 8, origin of m. retractor penis magnus at level of SC 21.
Coloration in preservative. Dorsal surface of head overall creamish-beige, yellowishbrown or blackish-brown, usually more distinctly paler on anterior half of head (Figure 9). Upper surfaces of internasals, anterior sides of prefrontals, and central part of frontal yellowish-beige, sparsely decorated with irregular blackish-brown spots. Upper cephalic scales irregularly and incompletely edged with blackish brown; these dark outlines of dorsal head scales usually become more distinct on their posterior margins. Lateral sides of parietals distinctly darker, almost forming two thick lateral stripes along the outer margins. Parietals speckled with dark brown, a pair of elliptic beige spots symmetrically present on inner sides of corresponding scales, just separated by parietal sutures. A six-scales-long, two-scales-wide beige sagittal line extends just behind parietal suture. Coloration of sides of the head distinctly separated by a narrow, oblique, dark streak, which originates from posterior margins of nasals, goes through upper edges of loreal and preocular, extends from middle postocular downwards to last supralabials until corner of mouth. Coloration above this dark streak as upper surface of head, whereas that below uniformly creamishbeige. Posterior edges of supralabials 1-3, upper margins of 4th-6th (left) or 3rd-5th (right) supralabials bordered by darker stripes. Posterior bottom corners of penultimate supralabials with a dark spot. Chin and throat uniformly creamish-yellow; anterior margin of mental and posterior corners of last two infralabials with faint, darker edges.
Dorsal surface of body umber or dark brown. Some dorsal scales are slightly darker, each dorsal scale more or less scattered with paler spots. A conspicuous, pale (cream or pale yellow), V or Y-shaped chevron extends onto neck, starting on each side from behind last supralabial and reaching upper surface of neck, pointing backwards, not in contact with beige sagittal line just behind parietal suture. Venter uniformly creamish-yellow, with outer edges of ventral scales dark brown or blackish-brown, in contact with the short streaks of lower edges of 1st DSR. Ventral surface of subcaudals uniformly colored as venter ( Figure 9).
Coloration in life. In life, dorsal surface of head and neck are bright yellowish brown; the background color of supralabials, loreals, and lower part of preoculars is yellowish off-white; posterior corner of eye streak is bordered with reddish orange (Figures 7 and 8). The sagittal line just behind the parietal suture and the chevrons along each side of neck are bright reddish orange, slightly paler than the background color of neck, barely visible in general view. The dorsal surface of body is yellowish-brown, brownish-red, or reddishochre, gradually tending to become darker towards the posterior part of the body, sparsely speckled with small dark blotches and spots, especially on the outer margins of dorsal scales. The background color of the lateral surface is darker than dorsal body surface, giving a dark maroon hue; caudal scales are narrowly edged with dark grey, forming several continuous or not, dark zigzag streaks on the posterior part of the tail.
The ventral surface of head is creamish-white, speckled with dark, grey, and beige, especially on the posterior margin of the infralabials. The first preventral is as throat, whereas the second preventral is decorated with bright reddish-orange. The ventral surface of body is glossy bright jacinth, reddish-orange or even scarlet posteriorly, distinctly separated from ventral head surface from the first ventral scale and throat. Ventral scales are ornated with a pair of more or less conspicuous blackish-brown spots on both sides, these spots progressively become larger and more distinct posteriorly and do not connect with each other. The ventral part of the tail is the same as the venter, whereas the subcaudals are more or less decorated with lateral blackish spots and nearly immaculate towards tail tip.
In situ, the hemipenis extends up to SC 6-7 with its crotch extending to SC 4-6; the fork point of m. retractor penis magnus extends to SC 7-8, the origin of m. retractor penis magnus invariably at level of SC 21.
In addition, Herpetoreas tpser sp. nov. may be confused with Hebius clerki; it can be distinguished from the latter species by the presence of a diastema before the enlarged maxillary teeth (vs. absent in Heb. clerki), a different ventral color in life (ventral surface of body and tail bright jacinth, reddish-orange to scarlet vs. bright yellow to white). For the morphological differences with reference to other members of the genera Amphiesma and Hebius, see the discussion below.
Etymology. The specific epithet tpser (pronounced as "/tIps@r/") is the acronym of "Tibetan Plateau Scientific Expedition and Research" of China, which greatly promoted the scientific research of the Qinghai-Tibet Plateau. All specimens of Herpetoreas tpser sp. nov. were collected during The First/The Second Tibetan Plateau Scientific Expedition and Research.
Comments. Herpetoreas tpser sp. nov. was previously misidentified as Hebius parallelus. Although the Tibetan record of Heb. parallelus has been clarified in this work, other records outside China still need to be studied. For example, another specimen (KSC 414) reported from Nagaland, northeastern India largely agrees with Herpetoreas tpser sp. nov. but differs from the latter in its strongly colored and patterned dorsal coloration (see Ao et al. [123]: Figures 1 and 2), which should be identified as Heb. clerki [11]. Furthermore, Kramer [72] doubted the occurrence of Heb. parallelus west of 88 • E and supposed that the Nepalese records might be questionable. The description of Heb. cf. parallelus from Nepal provided by Schleich et al. [82] is similar to Herpetoreas tpser sp. nov. except for having a slightly higher number of ventrals (163-172 vs. 153-167) and a pair of parietal spots (vs. absent). Thus, the systematics of the genus Herpetoreas still remain poorly known, and its biodiversity probably has been underestimated.

Morphological Characters of the Genera Previously Confused with Amphiesma
The genus Herpetoreas has long been considered as a synonym of the genus Amphiesma, mainly due to its morphological similarities [1,3,14]. Guo et al. [3] divided the genus Amphiesma auctorum into three genera-Amphiesma, Hebius, and Herpetoreas based on their phylogenetic results. The morphological diagnoses of these three genera have long remained to be elusive [3], which broadly caused taxonomic confusions [7,8]. Although the diagnostic key provided by Ren et al. [2] includes three genera of the former genus Amphiesma sensu lato, it could be only applied to species known from China and Vietnam. More importantly, it has been a challenging task to elucidate a morphological diagnosis of these three morphologically similar and speciose genera, especially for their conservativeness of pholidosis characters and general appearance. For example, these three genera previously confused with Amphiesma are similar to each other in having the following key generic diagnosis in the family Natricidae [124,125]: (1) head moderately distinct from neck; (2) nostrils situated and directed laterally or dorsolaterally; (3) eyes relatively moderate to large; (4) prefrontals mostly paired; (5) common dorsal scale rows, as 19-19-17; (6) largely overlapped ventral scale counts, as 118-158 in Amphiesma, both min. and max. in A. stolatum [6]; as 101-187 in Hebius, min. in Heb. viperinus [126], max. in Heb. arquus [127]; as 136-234 in Herpetoreas, min. in Her. pealii, max. in Her. platyceps (Table 2). Thus, the general morphological features cannot distinguish Amphiesma sensu lato.
Hemipenial morphology has been widely applied for diagnosis in the family Natricidae, especially at a genus level [1,24,128]. Malnate [1]: 47 also highlighted the effectiveness of hemipenial morphology and maxillary teeth in generic identifications of natricid taxa. We compared pholidosis, hemipenial morphology, and maxillary teeth of Amphiesma sensu lato, and found that different genera could be readily distinguished by a suit of characters including (1) the number of supralabials in contact with the nasals; (2) the diastema before the distinctly enlarged posterior maxillary teeth; (3) the direction of progression of the sulcus spermaticus; (4) the position of the distal end of sulcus spermaticus; (5) the relative size of the basal hook of the hemipenis; (6) the degree of heterogeneity between the left and right lobe of the hemipenis; (7) the conditions of apical naked areas of the hemipenis (protruding or not); and (8) the visibility of apical naked area from asulcate surface ( Figure 11; Table 4). The left lobe is defined to be the left lobe when face to the sulcus side of hemipenis in some studies [128,129], the orientation of hemipenis is followed Zhang et al. [24] in this study.
Specifically, the genus Amphiesma is the most morphologically distinct genus among the three different genera of Amphiesma sensu lato, which could be also reflected by its distant related phylogenetic relationship with reference to both Hebius and Herpetoreas (Figure 1, also see Deepak et al. [10]). The genus Amphiesma differs from both Hebius and Herpetoreas in having (1) only the 1st supralabial in contact with nasal vs. 1st-2nd supralabials in contact with nasal; (2) sulcus spermaticus extends to the middle of crotch vs. extends to the base or the tip of inner side of right lobe; (3) basal hook slightly higher than the adjacent spines vs. basal hook at least two times higher than the adjacent spines; (4) hemipenis sparsely ornamented with spines vs. densely ornamented with spines ( Figure 11; Table 4). Furthermore, the genus Amphiesma differs from Hebius in having a diastema before the distinctly enlarged diastema vs. absent. Although the hemipenial morphology is currently unknow for A. monticolum, a second member of Amphiesma [10], the character that 1st supralabial in contact with nasal is still capable of identification.
Secondly, Herpetoreas is different from Hebius in the morphology of maxillary teeth and hemipenis. Herpetoreas differs from Hebius in having the presence of diastema before the distinctly enlarged posterior maxillary teeth vs. absent in Hebius. The morphology of dentition has been considered as key diagnosis in the taxonomic history of natricid snake since Malnate [1], and this character is proved to be conservative within a given genus, such as Fowlea [130,131], Rhabdophis (only except for some ecological specialists) [132], and Opisthotropis [125]. Thus, the dentition is considered to be of significant value in distinguishing Herpetoreas from Hebius.
Additionally, although it has been broadly demonstrated that hemipenial morphology is a key morphological diagnosis in the taxonomy of Reptilia [23,133,134], very few are known in the genus Hebius and Herpetoreas [8,24,128,129]. Das et al. [8] briefly compared the hemipenial morphology within the genus Herpetoreas, while their differences among Amphiesma sensu lato remained unknown. Nevertheless, based on our current samplings, Hebius and Herpetoreas could also be distinguished from each other by hemipenial morphology. The morphology of hemipenis in Hebius has been depicted as shallowly bilobed, densely spinose, with nude crotch and a single sulcus spermaticus mostly extending to the base (sometimes the tip) of short lobes [24,128,129]. The apical naked area in the genus Hebius is highly developed, which extends proximally onto the asulcate surface (e.g., in Heb. venningi, see Cadle [128]: 9) or protrudes from crotch to different degree (e.g., in Heb. optatus, see Zhang et al. [24]: 41, Figure 52B and Appendix C, Figure A1H). Although this apical naked area has been previously recorded, few attentions have been given to the distinct hemipenial character, which is rare in other Asian natricid taxa and regarded to be significant [128,133,134]. In the current study, two distinct morphs of hemipenis are defined in the genus Hebius, based on the degree of crotch protrusion and the shape of lobes, which includes (1) morph I: apical naked area slightly protruded with relatively long and slender lobes ( Figure 13C2); (2) morph II: apical naked area distinctly protruded with relatively short and vestigial lobes ( Figures 11B and 13(C3,C4), Appendix C). The genus Herpetoreas, in contrast to Hebius, differs from all the two Hebius morphs in having less developed apical nude area, which is invisible from asulcate surface vs. well-developed apical nude area, which is visible from the asulcate surface in Hebius ( Figure 13).
Consequently, we emphasize the importance of hemipenial morphology and maxillary teeth in generic identifications of the Family Natricidae, due to the conservative evolution of pholidosis characters in natricid snakes [1]. Nevertheless, Hebius is the most specious genus in Natricidae (~47 species) [99,135], and morphological data of more species are also needed to improve the diagnosis.

Generic Assignment of Some Hebius Snakes
After the taxonomic revision of Amphiesma sensu lato, Hebius parallelus was considered as a member of Hebius based on molecular phylogenetic results [3]. David et al. [11] subsequently confined Heb. parallelus to northeastern India only, and revalidated Amphiesma clerki to the genus Hebius to accommodate populations from the Eastern Himalayas, northern Myanmar, and western Yunnan, China. Therefore, the sample of Heb. parallelus (CAS 215036) from Yunnan, China used in Guo et al. [3] would actually represent Hebius clerki ( Figure 1; Table 1). Recently, Che et al. [13] reconstructed the phylogenetic tree of the genus Herpetoreas with inclusion of sample of Hebius cf. parallelus from Mêdog, Tibet, China and transferred the species to Herpetoreas. However, our results indicated that the taxon identified as Hebius cf. parallelus from Mêdog, Tibet, China is a misidentification of the new species Herpetoreas tpser described above.
We cannot here ascertain the generic position of Hebius parallelus (Tropidonotus parallelus Boulenger, 1890: 345. Type locality, by virtue of lectotype designation: "Sikhim", i.e., now Sikkim, India). The lectotype was designated by Kramer [72]: NHMUK 1946.1.13.53, adult male; collected and deposited by Sir Joseph Dalton Hooker. No genetic sequence of this rare species is available. However, we noticed that Heb. parallelus is similar to species currently referred to Herpetoreas in having (1) a similar number of maxillary teeth, 21-22 vs. 13-23 min. in Her. pealii, max. in Her. xenura, (2) posterior teeth distinctly enlarged, (3) the presence of diastema before the enlarged maxillary teeth. Nevertheless, we here refrain from referring Hebius parallelus to the genus Herpetoreas pending the availability of phylogenetic analyses.
Since the diagnostic characters of species previously referred to Amphiesma sensu lato have long been unclear, the generic assignments of several Hebius taxa to the genus Hebius were recently found to be inappropriate [7,8]. Moreover, Guo et al. [3] also rightly suggested that the genus Herpetoreas may contain additional species then referred to Hebius. Particularly, the generic allocation of a newly described species, namely, Hebius lacrima Purkayastha and David, 2019, needs to be further ascertained. Hebius lacrima agrees with the diagnosis of Herpetoreas in having "24 gradually enlarged maxillary teeth, followed, with a diastema, by 3 distinctly enlarged posterior teeth", a major difference from the genus Hebius [113] (Table 4; Figure 13B). The morphological differences provided by Purkayastha et al. [113] also cannot distinguish it from Herpetoreas when one takes all present members of the latter genus into account, including (1) ventrals scales counts 147 in Heb. lacrima vs. 136-234 (min. in Her. pealii; max. in Her. platyceps), and tail length ratio TaL/TL, 0.301 in Heb. lacrima vs. 0.227-0.317 (min. in Her. pealii; max. in Her. tpser). The distribution range of Hebius lacrima also falls within the known distribution range of Herpetoreas. Therefore, we question the generic assignment of Hebius lacrima, and recommend further molecular phylogenetic analyses to clarify the taxonomic confusions still present among species of the former genus Amphiesma sensu lato.