Introduction

Two species of kraits (Genus Bungarus) are known to inhabit the Hong Kong Special Administrative Region; the many-banded and banded kraits, Bungarus multicinctus and B. fasciatus, respectively (Karsen et al., 1998). While B. fasciatus is easily identifiable by its black and yellow bands, the white and black bands exhibited by B. multicinctus closely resemble those of other congeners with overlapping geographic distributions (Xie et al., 2018; Chen et al., 2021). Beyond Hong Kong, B. multicinctus has been believed to be widely distributed across Myanmar, Laos, Thailand, Vietnam, as well as Southern China (Shan et al., 2016; Uetz and Hošek, 2017). Overlapping in distribution is a closely related congener (Tan and Tan, 2015), B. candidus, which is mainly restricted to Southeast Asia (Nguyen et al., 2017; Uetz and Hošek, 2017).

In a recent study, Xie et al. (2018) genetically and morphologically re-identified several specimens of B. multicinctus collected from Southern and Southwestern China as B. candidus, including one from Hong Kong. Since then, however, Chen et al. (2021) suggested that B. candidus records in Southern and Southwestern China (Xie et al., 2018) and B. multicinctus records in Thailand, Vietnam, and Laos (Uetz and Hošek, 2017) be revised to represent B. wanghaotingi. This latter species is known to occur in Myanmar and Southwestern China, and is closely related to B. candidus and B. multicinctus (Leviton et al., 2003; Uetz and Hošek, 2017; Chen et al., 2021). Together, the three species are suggested to form a complex distinct from other congeners (Kuch, 2007; Chen et al., 2021).

Debates surrounding the discrimination of B. multincinctus as distinct from B. candidus and other congeners, such as India's B. caeruleus, are at least as old as the 19^th century (Blyth, 1860; Boulenger, 1896; Stejneger, 1907). The identification of black-and-white kraits in Hong Kong as B. multicinctus was accordingly founded on assumed differences in geographic distributions and the number of bands displayed by these congeners (Stejneger, 1907; Herklots, 1933). Yet this assumption pre-dates and potentially disregards the initial description of B. wangahaotingi as a species (Pope, 1928). Therefore, with no molecular work conducted on Hong Kong's B. multicinctus prior to the inclusion of a single specimen in the study by Xie et al. (2018) initially identifying it as B. candidus, the tenability of this distinction is worth exploring through phylogenetic and further morphological analyses. In this study we apply such methods to (1) determine whether B. wanhaotingi or B. candidus occur in Hong Kong, and (2) if B. multicinctus is present as well.

Materials and Methods

We obtained specimens previously assumed to be B. multicinctus from across Hong Kong through donations by Kadoorie Farm and Botanic Garden (KFBG) and the Hong Kong Agriculture, Fisheries, and Conservation Department (AFCD). These included tissue samples as well as whole individuals that were caught locally, had arrived at KFBG via the Wild Snake Rescue Project ( www.kfbg.org/en/wild-snake-rescue-project), and which would have been dead on arrival or euthanized. We also collected dead specimens opportunistically throughout Hong Kong. We received 11 tissue samples, 10 of which were donated by KFBG, and one collected opportunistically. We received seven whole individuals, of which four were donated by KFBG, one by AFCD, and two collected opportunistically. The whole individuals were mostly collected from across Hong Kong's New Territories, including Hoi Ha, Shing Mun, Tai Po Kau, Kau Lung Hang Village, and Pat Sin Leng, with the exception of one collected in Shek O, Hong Kong Island. These 18 total specimens range in their initial date of collection from September 1998 to June 2021 (see Appendix). We deposited all whole individual specimens at Lingnan University's Natural History Collection.

DNA extractions and phylogenetic analyses

We extracted DNA from the samples using a QIAGEN DNeasy Blood & Tissue Kit (QIAGEN, Hilden, Germany) according to the manufacturer's protocols. We measured DNA yields using a Nanodrop One Microvolume UV-Vis Spectrophotometer. We amplified a 355 bp region of Cytochrome B (Cyt b) using primers H15149B (5′-CCCCTCAGAATGATA TTTGTCTCA-3′) and L14841 (5′-ATCCAAC ATCTCAGCATGATGAAA-3′) designed by Kocher et al. (1989), and previously applied by Wong et al. (2004) to identify confiscated snake meat samples in Hong Kong. We performed polymerase chain reaction in 20 µl reactions containing 1 ng DNA, 1× Biotechrabbit^TM Lyo Hot Start PCR Master Mix and 1 mM of each of the primers. We programmed initial template denaturation at 95°C for 2 min followed by 30 cycles of 95°C for 30 s; 50°C for 20 s and 72°C for 45 s, and then a final elongation step at 72°C for 5 min. We visually confirmed PCR amplifications on a 2% agarose gel using 5 µl of PCR product. We purified PCR products using a QIAGEN QIAquick PCR purification kit (QIAGEN, Hilden, Germany) and submitted purified reactions to The Centre for PanorOmic Sciences (CPOS) at the Li Ka Shing Faculty of Medicine, The University of Hong Kong, for sequencing.

Table 1.

Morphological traits suggested to identify three, black-and-white, closely related Bungarus species overlapping in geographic distributions.

We visually assessed sequence chromatograms, generated consensus sequences for all our samples, and used the global alignment algorithm to align all consensus sequences in Geneious 10.0.2 (Kearse et al., 2012; Biomatters Ltd., Auckland, New Zealand). We obtained Cyt b consensus sequence entries for B. candidus, B. wanghaotingi, B. multicinctus and B. suzhenae from the GenBank database. We chose to include B. suzhenae because it was recently described as a species, its distribution spanning from Northern Myanmar to Southwestern China overlaps with that of B. multicinctus and B. wanghaotingi, and it forms a group sister to the complex formed by the other three species (Chen et al., 2021).

We used the Neighbour-Joining method (Saitou and Nei, 1987; Dantas et al., 2021) to construct a phylogenetic tree including our sequences obtained in Hong Kong together with the Genbank entries, as well as a voucher specimen of Naja atra as the outgroup. We generated a consensus tree inferred from 1000 replicates (Felsenstein, 1985) using the Maximum Likelihood method together with the Tamura 3-parameter model (Tamura, 1992) to obtain bootstrap values above a 70% cutoff as support for the tree. We estimated evolutionary distances between sequences using the pdistance method, which estimates the proportion of nucleotide sites differing between sequences, and conducted all phylogenetic analyses in MEGA X (Kumar et al., 2018).

Results

Phylogenetic analyses

We successfully obtained consensus sequences from 16 specimens, while two of our samples (KFBG23 and AFCD01) from whole individuals failed to yield adequate sequences. Phylogenetically, two samples grouped with B. multicinctus, and 14 with B. wanghaotingi (Fig. 1). None of our samples grouped with either B. candidus, or the recently described B. suzhenae (Fig. 1). The evolutionary distances between our two B. multicinctus samples and 14 B. wanghaotingi samples ranged from 2.6% to 2.9% for KFBG12, and from 3.3% to 3.6% for KFBG23. Distances between our two B. multicinctus samples and the B. candidus sequences from GenBank was 2.9% for KFBG12 and 3.6% for KFBG23. Distances between our 14 B. wanghaotingi samples and the GenBank B. candidus sequences ranged from 2.3% to 2.6%.

Fig. 1.

Phylogenetic tree generated with the neighbor-joining method for black-and-white krait specimens collected in Hong Kong (bolded), together with genetic sequences for four potentially matching Bungarus species obtained from the GenBank database. Numbers at nodes are bootstrap values above a cutoff value of 70% and extracted from the maximum likelihood consensus tree. A voucher specimen of Naja atra collected in Hong Kong serves as the outgroup (dashed line), while the scale represents branch length.

Morphological examinations

We examined the morphological traits of seven individuals, including two females, one male, and four juveniles for which we could not determine the sex (Table 2). Overall, specimens displayed more traits unique to B. multicinctus, as described by Chen et al. (2021), despite three individuals phylogenetically grouping with B. wanghaotingi. Of those latter three, all morphological traits measured for KFBG11 and LINGU-HERP216, and all but one for LINGU-HERP845, matched those attributed to B. multicinctus (Tables 1 and 2). Of the specimens phylogenetically grouping with B. multicinctus, four and three morphological traits matched those of this species for KFBG23 and KFBG12, respectively. KFBG11 and LINGU-HERP216, our only adult-sized whole individual specimens, displayed similar morphological traits, all consistent with B. multicinctus, yet both phylogenetically grouped with B. wanghaotingi. The number of ventral scales for KFBG11 matches all three species, and the uniform black head and necks in both KFBG11 and LINGU-HERP216 agrees with B. wanghaotingi and B. multicinctus.

Table 2.

Morphological traits and genetic identification results for whole black-and-white krait specimens collected in Hong Kong. Traits closely matching those described for B. candidus, B. wanghaotingi and B. multicintus are respectively indicated by an asterisk (*), dagger (†), and double dagger (‡).

All specimens displayed a number of white bands on the body exceeding the maximum of 33 previously described for B. wanghaotingi, and within the range given for B. multicinctus. The ventral surface on the tails of all specimens were covered in black bands and patches (Fig. 2); another trait attributed to B. multicinctus. The number of ventral scales matched those of B. multicinctus alone for three specimens (KFBG13, KFBG23 and LINGUHERP216). Head neck patterns exhibited by specimens were varied (Fig. 2). Two specimens phylogenetically grouping with B. wanghaotingi displayed uniform black head and necks, a trait attributed to adults of both B. wanghaotingi and B. multicinctus. The ventral surface of the body was white with scattered brown pigments, a trait attributed to B. multicinctus, for two specimens phylogenetically grouping with B. wanghaotingi (KFBG11 and LINGU-HERP216), and one grouping with B. multicinctus (KFBG23). Yet the ventral surface of the body was immaculate white, a trait attributed to B. candidus and B. wanghaotingi, for the other four specimens, one of which phylogenetically grouped with B. multicinctus (KFBG 12), and another with B. wanghaotingi (LINGU-HERP845; Fig. 2).

Discussion

We genetically identified both B. multicinctus and B. wanghaotingi among specimens collected in Hong Kong. We further found these specimens to display morphological traits mostly matching those described for B. multincinctus and, to a lesser extent, B. wanghaotingi (Chen et al., 2021). Following the recent re-classification of Chinese B. candidus specimens as B. wanghaotingi (Xie et al., 2018; Chen et al., 2021), we found no genetic and little morphological evidence of B. candidus occurring in Hong Kong.

Fig. 2.

Morphology of two Bungarus specimens, LINGU-HERP216 (left column) and KFBG12 (right column) collected from Hong Kong, that phylogenetically grouped with B. wanghaotingi and B. multicinctus, respectively. A and B: dorsal view. C and D: ventral view. E and F: lateral view of head and neck.

While B. multicinctus was already known to occur in Hong Kong, the new identification of B. wanghaotingi in the region we present in this study is of potential medical significance. As members of the elapid family, Bungarus species carry extremely potent venom with the capacity to be lethal to humans (Tan and Ponnudurai, 1990; Yanoshita et al., 2006; Chanhome et al., 2009). Envenomation by snake bite from these snakes is a serious public health concern in Asia, and often occur at night when victims are sleeping due to the nocturnal hunting behaviour of these snakes (Alirol et al., 2010; Warrell, 2010). In Hong Kong, victims of bites by B. multicinctus, which produces some of the most potent snake venoms in the world (Gopalkrishnakone and Chou, 1992; Yin et al., 2020), are typically administered an anti-venom specifically designed for the species (Mong et al., 2017). Likewise, patients bitten by B. candidus, which also involves severe and potentially lethal consequences (Trinh et al., 2010; Chaisakul et al., 2017; Charoenpitakchai et al., 2018), are treated with species-specific anti-venoms in Thailand (Leeprasert and Kaojarern, 2007).

While venoms within this genus do share similarities, those of B. multicinctus and B. candidus differ in some biological properties (Kuch et al., 2003; Yanoshita et al., 2006; Tan and Tan, 2015), such as their respective contents of alkaline phosphomonoesterase, L-amino acid oxidase, acetylcholinesterase and hyaluronidase (Tan and Ponnudurai, 1990). As a lesser known and seldom studied species, the properties of B. wanghaotingi venom should be further studied to understand possible differences with those of its congeners. Given the shorter evolutionary distance between B. wanghaotingi and B. candidus (Figure 1; Chen et al., 2021), anti-venom designed for B. candidus venom could be more effective for B. wanghaotingi bites than for those by B. multicinctus. In the case of Hong Kong, an awareness of nuances specifically between B. wanghaotingi and B. multicinctus venoms could provide better medical support for envenomed victims (Laxme et al., 2019). Of critical importance, treating patients bitten by one species using the anti-venom for another, despite being closely related, could prove to be ineffective (Laxme et al., 2019). This has been documented for one B. candidus bite victim in Thailand unsuccessfully being administered anti-venom derived from a Taiwanese B. multicinctus (Warrell et al., 1983; Tan and Tan, 2015). Further complicating the issue are potential intraspecific differences in the effects of Bungarus venoms depending on geographic locality (Chaisakul et al., 2017; Charoenpitakchai et al., 2018; Laxme et al., 2019; Hia et al., 2020; Oh et al., 2021).

These issues frame venomics research as a priority from the perspective of public health, which will be instrumental for ensuring adequate treatment of future victims of bites by these two species. However, our genetic evidence for the occurrence of both B. wanghaotingi and B. multicinctus in Hong Kong, as well as the possible shortcomings for morphological distinctions (Chen et al., 2021), require further work to be substantiated. Although we received only one adult male specimen, hemipenial morphology is suggested to be a distinctive feature for B. wanghaotingi and B. multicinctus (Chen et al., 2021), and should be a focal point for future studies. Accordingly, the lack of distinctive morphological traits among our specimens could be due to the fact that many were juveniles.

Beyond its medical importance, B. multicinctus carries economic value in Asia, where it is often traded for consumption (Wong et al., 2009; Ji et al., 2020), and is known to have been heavily exported from mainland China (Zhou and Jiang, 2005). Distinguishing B. wanghaotingi and B. multicinctus in Hong Kong, as well as in mainland China, would be crucial for adequately monitoring the trade and conservation statuses of these two species. The consumption of snakes in Hong Kong sees many additional non-native species regularly imported (Wong et al., 2009), raising the possibility of accidental releases. This could have resulted in a relatively recent arrival of B. wanghaotingi in Hong Kong, and explain the lapse in early descriptions of the species alongside B. multicinctus records (Stejneger, 1907; Herklots, 1933). Conversely, the opposite scenario is also possible, with B. wanghaotingi already present and B. multicinctus introduced.

Yet evolutionary processes resulting in the distributional overlap of black-and-white krait species deserve further exploration (Kuch et al., 2005). Identifying the geological time and geographic location of speciation would be a critical step in elucidating points of divergence within the Bungarus candidus/multicinctus/wanghaotingi complex. Currently our study provides foundational groundwork for future molecular, morphological and ecological investigations on a larger regional scale. The inconsistency in morphological features distinguishing the two species in our results highlight the need for larger sample sizes and additional methods for resolving the taxonomy of this species complex. Given the limited number of specimens from Southern China examined by Chen et al. (2021), future work identifying and differentiating the morphology, evolutionary history and distribution of Bungarus species across the region is necessary. This would be instrumental in decisively identifying each species, potentially supporting our current phylogenetic results. Ultimately, distinguishing B. wanghaotingi and B. multicinctus within Hong Kong should be prioritized for economic, public health, and conservation reasons.