A morphological and molecular study of Hydrodynastes gigas (Serpentes, Dipsadidae), a widespread species from South America

Background Studies with integrative approaches (based on different lines of evidence) are fundamental for understanding the diversity of organisms. Different data sources can improve the understanding of the taxonomy and evolution of snakes. We used this integrative approach to verify the taxonomic status of Hydrodynastes gigas (Duméril, Bibron & Duméril, 1854), given its wide distribution throughout South America, including the validity of the recently described Hydrodynastes melanogigas Franco, Fernandes & Bentim, 2007. Methods We performed a phylogenetic analysis of Bayesian Inference with mtDNA 16S and Cytb, and nuDNA Cmos and NT3 concatenated (1,902 bp). In addition, we performed traditional morphometric analyses, meristic, hemipenis morphology and coloration pattern of H. gigas and H. melanogigas. Results According to molecular and morphological characters, H. gigas is widely distributed throughout South America. We found no evidence to support that H. gigas and H. melanogigas species are distinct lineages, therefore, H. melanogigas is a junior synonym of H. gigas. Thus, the melanic pattern of H. melanogigas is the result of a polymorphism of H. gigas. Melanic populations of H. gigas can be found in the Tocantins-Araguaia basin.


INTRODUCTION
Species are considered lineages with distinct evolutionary histories (De Queiroz, 2007). Taxonomic studies are traditionally based on morphological descriptors to delimit species (e.g., Franco et al., 2017;França et al., 2018;Meneses-Pelayo & Passos, 2019). However, in many cases, species are difficult to delimit due to the limited number of Given the wide distribution of Hydrodynastes gigas in South America, a level of intraspecific variation throughout its populations is to be expected, with some potential to represent still undescribed cryptic species. Indeed, Franco, Fernandes & Bentim (2007) considered one of these populations as a distinct species, describing H. melanogigas mainly through its differential color pattern and pointed out its similarity with H. gigas on meristic and hemipenial characteristics. Therefore, the distinction between H. melanogigas and H. gigas rests mainly on its melanistic color pattern and on its inferred allopatric distribution with H. gigas. The present study aims to evaluate the taxonomic validity of Hydrodynastes gigas and Hydrodynastes melanogigas using an integrative taxonomic approach inferred by molecular and morphological data.

MATERIALS & METHODS
We evaluated the taxonomic status of Hydrodynastes gigas and Hydrodynastes melanogigas by sequencing two mitochondrial and two nuclear genes for 32 individuals belonging to the two species. We further analyzed the external morphology of 186 specimens of H. gigas and H. melanogigas.
We used PartitionFinder 2 to identify partitioning schemes and the most appropriate nucleotide replacement models (Lanfear et al., 2016). According to our concatenated alignment, we found five partitions evaluated by BIC (Table 2). For phylogenetic analysis, we used the Bayesian inference implemented in MrBayes v3.2.6 (Ronquist & Huelsenbeck, 2003) using the substitution models generated by PartitionFinder. We ran two independent runs of four Markov chains for 20 million generations sampling every 5,000 generations and discarding 25% as burn-in. We evaluated the stability of the analysis in Tracer v1.6, ensuring that all ESS values were above 200 (Rambaut et al., 2014). We calculated the divergence between sequences (p-distance) in Mega v10.0.5 (Tamura et al., 2013).

Morphological analysis
We examined 144 specimens of Hydrodynastes gigas and 42 specimens of H. melanogigas (Fig. 1B Franco, Fernandes & Bentim (2007) described Hydrodynastes melanogigas based on 17 specimens collected in the municipalities of Palmas (type locality), Porto Nacional, and Lajeado, which are all located in the state of Tocantins, Brazil. Unfortunately, most of the type series was lost in the 2010 fire that occurred at the Instituto Butantan. Currently, only three individuals remain from the type series: two at the proper Institute in São Paulo (paratypes IBSP 65978 and IBSP 66387) and one at the Museu Nacional do Rio de Janeiro (paratype MNRJ 15101). From the remaining type-series, we analyzed all the remaining individuals.
We examined 14 meristic characters (Table 3) and eight morphometric ones (Table 4), in addition to the coloration pattern and morphology of the hemipenis. Sex was determined by the presence or absence of hemipenes through a ventral incision at the base of the tail. We measured individuals with an electronic caliper (0.01 mm) and a flexible ruler (1 mm), on their right side whenever possible. In order to test morphometric differences between H. gigas and H. melanogigas, we conducted a principal component analysis (PCA) and took the first two principal components of the ordination to create a MANOVA. We   ran this analysis with adult males and females separately, and performed the analysis in R software (R Core Team, 2014) using the package Vegan (Oksanen et al., 2007). We followed the terminology of Dowling (1951) for counting the ventral scales, and Peters (1964) and Vanzolini, Ramos-Costa & Vitt (1980) for pholidosis. We surveyed the geographic coordinates of the data catalogs of zoological collections using Google Earth software.

Hemipenial morphology
We prepared a hemipenis from a topotype of Hydrodynastes melanogigas and 19 from H. gigas from different localities in the Amazon, East Brazil and La Plata hydrobasins (Appendix S1). Whenever possible, we prepared the hemipenes on the right side according to the technique originally described by Manzani & Abe (1988), as modified by Pesantes (1994), Zaher (1999), and Zaher & Prudente (2003). We stained the external calcareous structures with alizarin red, as suggested by Nunes et al. (2012), for a better visualization of microstructures in the surface of the organ. Terminology follows Dowling & Savage (1960), Zaher (1999) and Myers & Cadle (2003).

Molecular approach
We recovered the genus Hydrodynastes as monophyletic, and the topology of the concatenated gene tree showed two strongly supported clades with posterior probability ( pp = 1.00). Our concatenated dataset tree grouped Hydrodynastes melanogigas within H. gigas (Fig. 2). The uncorrected p-distance for both the mtDNA 16S and Cytb showed low genetic differences (0.01% and 0.2%, respectively) between the lineages of H. gigas and H. melanogigas. However, the genetic differences between H. gigas and H. bicinctus were 0.43% for 16S and 13% for Cytb (

Notes.
In parenthesis the sampled number (n). SVL, snout-vent length (from the tip of the snout to the cloaca); TL, tail length; HL, head length (from the tip of the snout to the quadratemandibular articulation); HW, head width (length of the widest part of head); DN, distance between nostrils (maximum distance between the nostrils); EN, distance between eye and nostril; ED, eye diameter; HH, head height (maximum distance between the base of the mandible and the parietal surface.

Morphological approach
We found overlap in all meristic and morphometric characteristics between Hydrodynastes gigas and H. melanogigas (Tables 3 and 4). The first principal components from both PCA analysis (males and females) recovered 99% of variation, and through the MANOVA of males (F = 2.2949; p = 0.1109) and females (F = 0.3463; p = 0.7095) we found no significant morphometric differences between both species (Fig. 3). In addition, we observed gradient levels of melanism in H. melanogigas (Figs. 4A-4L). We examined fully melanic specimens (Fig. 4A) to specimens with clear visible bands along the body (Fig. 4K). We also observed that some H. gigas individuals from the Amazon, La Plata and Tocantins-Araguaia basins present darker coloration overlapping the gradient of melanism found in H. melanogigas (Figs. 5A-5L). We did not find any morphological characteristics that differentiate the two species. We did not observe coloration patterns within or between the populations of Hydrodynastes gigas (Figs. 6A-6R). We observed ontogenetic variation in the color pattern of all populations analyzed, with no distinction between males and females. Juveniles in the early stages of life have well-defined rounded dark spots all over their backs to the end of their tails; these spots are outlined by a lighter line, while in adults rounded spots may or may not be well defined, and may not present a clear outline (Figs. 7A-7J). Furthermore, we identified two young males of H. gigas (CHUNB 22053, Figs. 7I-7J); CHUNB 22068) from the type locality of H. melanogigas, as well as 18 more specimens  sulcate and assulcate faces. The sulcus spermaticus bifurcates in the proximal region of the hemipenis body and each branch extends centrolinearly until it reaches the proximal region of the lobes, in which they follow a centrifugal position that ends at the lateral region of the tip of each. The capitulum, formed by papillate calyces and spikes, extends over most of the surface of the lobes, except in the region of the assulcate face that is occupied by two parallel rows of papillated and conspicuous body calyces that extend to the distal region of the hemipenis body, where they converge on the lobular crest and continues to the middle portion of the hemipenis. We detected low intraspecific variation among Hydrodynastes gigas populations. Some hemipenes showed little size variation in lobes and body calyces, varying from slightly visible to conspicuous in the hemipenial body.

DISCUSSION
Hydrodynastes gigas is widely distributed throughout South America, occurring with low genetic variability throughout most of its extension range. Although the genetic structure of widely distributed species can be easily influenced by natural barriers (Patton, Da Silva  Malcolm, 1994;Pellegrino et al., 2005;Rocha et al., 2015), this clearly is not the case for the genus Hydrodynastes (see Murta-Fonseca, Franco & Fernandes, 2015).
Here, we used an integrative taxonomic approach and adopted the species concept of one lineage with distinct evolutionary histories (De Queiroz, 2007), to test the taxonomic validity of Hydrodynastes melanogigas. Our results did not separate H. gigas and H. melanogigas based on molecular, meristic, morphometric and hemipenial characters. The hemipenis of Hydrodynastes melanogigas analyzed showed no differences from the hemipenis of H. gigas (n = 19). In their description of H. melanogigas, Franco, Fernandes & Bentim (2007) also pointed out its similarity with H. gigas based on meristic characters and hemipenis morphology. The only superficial distinction that remains between these two taxa is the presence of melanism in the latter. Geographic or regional melanism has already been reported for several populations of Squamata (Pearse & Pogson, 2000;King, 2003;Bernardo et al., 2012). In addition, polychromatism can be a bias within taxonomy, if the revision and/or description of species does not take into account the organisms throughout their whole distribution (Bernardo et al., 2012;Ruane et al., 2018;Mângia et al., 2020). In fact, Franco, Fernandes & Bentim (2007) carried out an analysis covering almost the entire distribution of H. gigas, however we found degrees of melanism in some populations that were not identified by other authors. The variation of melanism found in H. gigas and the melanistic gradient observed in H. melanogigas (Figs. 4 and 5), along with genetic support, suggests that H. melanogigas is not a distinct species but rather a melanic population of H. gigas. The distribution of Hydrodynastes melanogigas without sympatry with H. gigas in the Tocantins-Araguaia basin was an important factor for its description (Franco, Fernandes & Bentim, 2007). However, in this study, we analyzed two juveniles of H. gigas from the type locality of H. melanogigas (CHUNB 22053, Figs. 7I-Figs. 7J;CHUNB 22068). All specimens analyzed by Franco, Fernandes & Bentim (2007) and herein were adults or juveniles and no neonates were observed. Therefore, we do not know whether the specimens considered as H. melanogigas could have been born melanic or if melanism occurs during their ontogeny. Still, some studies suggest that thermal melanism is associated with latitude and high altitudes, i.e., relatively cold environments (Capula, Luiselli & Monney, 1995), which does not agree with the present case. More studies are needed to confirm the adaptive meaning of melanism through studies of thermal biology. Therefore, due to the lack of any characteristic that sustain these two taxa as distinct species and their low genetic distance (0.04% 16S and 0.2% Cytb), we consider H. melanogigas Franco, Fernandez & Bentim, 2007 as a junior synonym of H. gigas (Duméril, Bibron & Duméril, 1854).

Systematic account
Hydrodynastes gigas (Duméril, Bibron & Duméril, 1854) -5), which only presents a cranial picture. Overall, Wallach, Williams & Boundy (2014, p. 339) indicate that the types would be MNHN 2493 a-c, but according to Uetz et al. (2019), it would be an individual labeled MNHN 3623. Due to this conflicting information, we contacted the curator of the Herpetological Collection at the Muséum National d'Histoire Naturelle de Paris (Dr. N. Vidal) who confirmed that there is only one type specimen, a skin labeled MNHN 3623 (Fig. 9), deposited in the collection, and that the other two specimens appear to be lost . Since Duméril, Bibron & Duméril (1854) did not select any specific specimen from their type series, we therefore, designate the specimen MNHN 3623 as the lectotype of Hydrodynastes gigas.
Description of the lectotype MNHN 3623 (Fig. 9): Adult of undetermined sex; SVL 1570 mm; TL 540 mm; HL 62 mm; HW 35 mm; DN 10 mm; two internasals; nasal divided; one loreal; one preocular; three suboculars; two postoculars; temporal 2 + 2/2 + 2 ; nine supralabials, none contacting the orbit; eleven infralabials, first to sixth contacting chin shields on the right side and first to fifth on the left side; two pairs of chin shields; dorsal 19/16/15 scales, smooth; two apical pits; ventral 153; subcaudals 74, paired and cloacal scale single. Color of the preserved lectotype (ethanol 70%) (Fig. 9): Head brown with black 'U' shaped spot at the end of the parietal scale; post-ocular stripe that extends longitudinally on each side; supralabial brown with the last four scales stained black; infralabials and chin shields cream; dorsum of body brown with dark rounded spots that extend to the end of the tail; ventral body cream with three black stripes to the middle of the body.
Diagnosis: Hydrodynastes gigas is distinguished from its congener H. bicinctus by the following combination of characters: dorsal scales normally 19/19/15; ventral scales in males 150-168 and in females 152-172; subcaudal scales in males 58-88 and in females 49-84; maxillary teeth 15-17; two apical pits in the dorsal scales; post-ocular stripe that extends longitudinally (on each side); ventral body with three lines of continuous spots up to the middle of the body.
Variation: All variation in morphometric and meristic data are presented in Tables 3  and 4. Regarding coloration patterns, a considerable degree of color variation can be found in H. gigas (Figs. 4A-4L; 5A-5L; 6A-6R and 7A-7J) dorsum ranging from yellow to dark brown or completely dark in melanic populations; rounded spots on the dorsum may vary in shape and size, in some individuals they may be hollow or filled; in neonates the dark rounded spots are well defined throughout the dorsum until the end of the tail, and these spots are outlined by a lighter line; darks spot in the shape of 'V' or 'U' at the end of the parietal scale, only visible in non-melanic populations; ventral body cream/brown with three black stripes that usually go up to the middle of the body, rarely to the cloaca, some individuals do not have these stripes, in melanic populations the belly is dark without spots or when present these spots are continuous on the sides. Distribution: Hydrodynastes gigas is widely distributed throughout South America, east of the Andes, occurring in Amazon, East Brazil, La Plata, North Brazil, Northeast South America, Parnaiba and Tocantins-Araguaia hydrobasins (Fig. 10).

CONCLUSIONS
Our results did not separate H. gigas and H. melanogigas based on molecular, meristic, morphometric and hemipenial characters. Therefore, the melanistic pattern of Hydrodynastes melanogigas is characterized here as the result of polymorphism within H. gigas. Although our integrative approach helped elucidate the taxonomic status of H. melanogigas, we believe future, multi-loci phylogeographic studies are needed in order to better understand the evolutionary history of the populations belonging to the two remaining species H. gigas and H. bicinctus.
• Diego J. Santana conceived and designed the experiments, analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the paper, and approved the final draft.

DNA Deposition
The following information was supplied regarding the deposition of DNA sequences: The 16S sequences are available at Genbank: MT192267-MT192299; MT215327-MT215335. The Cytb sequences are available at Genbank: MT224972-MT225014.
The Cmos sequences are available at Genbank: MT328065-MT328096. The NT3 sequences are available at Genbank: MT328097-MT328134 and MT424769. In addition, all information is available in Table 1.

Data Availability
The following information was supplied regarding data availability: Summarized morphometric and meristic data are available in Tables 3 and 4. Location data and zoological collections are available in Appendix S1. The data shows overlap in all meristic and morphometric characteristics between Hydrodynastes gigas and H. melanogigas.

Supplemental Information
Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.10073#supplemental-information.