Redescription of Dracovermis occidentalis (Digenea: Liolopidae) infecting American alligator, Alligator mississippiensis from the Bon-Secour River (Mobile–Tensaw River Delta, Alabama, USA) and a revised phylogeny for Liolopidae

We examined several American alligators, Alligator mississippiensis (Daudin, 1802) (Crocodilia: Alligatoridae) from Louisiana, Alabama, and South Carolina in August 2022. The intestine of one alligator from Alabama was infected by Dracovermis occidentalis Brooks and Overstreet, 1978 (Platyhelminthes: Digenea: Liolopidae Odhner, 1912), a seldom collected and incompletely described trematode that lacks a representative nucleotide sequence. Liolopidae comprises 5 genera and 15 species: Liolope spp. infect giant salamanders; Helicotrema spp. infect turtles and lizards; Harmotrema spp. infect snakes; Paraharmotrema spp. infect turtles; and Dracovermis spp. infect crocodilians. Based on our study of the newly collected specimens and the holotype of D. occidentalis, we redescribe D. occidentalis, correct errors in its original description, and provide an updated phylogeny for Liolopidae that, for the first time, includes Dracovermis Brooks and Overstreet, 1978. Our specimens were identified as D. occidentalis by having testes in the posterior 1/3 of the body, a pretesticular cirrus sac, a spined and eversible cirrus, a bipartite seminal vesicle, and a post-acetabular vitellarium. A phylogenetic analysis of the D1–D3 domains of the nuclear large subunit ribosomal DNA (28S) recovered Liolopidae as monophyletic; however, low taxon sampling in the group precludes hypothesis-testing about liolopid-vertebrate cophyly. This is the first collection for morphological study of the type species for Dracovermis since the genus was proposed 46 years ago, the first record of a liolopid from Alabama, and the first phylogenetic analysis that includes Dracovermis.


Introduction
The American alligator, Alligator mississippiensis (Daudin, 1802) (Crocodilia: Alligatoridae), hereafter "alligator", is the largest predatory reptile in North America.Historically abundant and ranging across coastal/estuarine waters and semi-inland riverine systems throughout the southeastern USA, alligators were hunted/poached for their skin and nearly driven to extinction in the early 1900s.They remained considered threatened as late as the 1960s but proper management has allowed alligator populations to recover relatively rapidly, with the US Fish and Wildlife Service in 1979 downlisting the alligator from CITES Appendix I to Appendix II; allowing for some legal international trade of alligator skins (Ashley and Caldwell 2013;Elsey and Woodward 2010;Jelden et al. 2014;PIJAC 2013;Thorbjarnarson et al. 1992).Regarding captive stocks of alligators, a survey conducted in 2013 revealed that > 100 commercial farms in the US maintained ~ 790,000 alligators, with 20,000 hatchlings being produced from these facilities annually.Based on these facts, along with the cultural and conservation value of alligators in the wild, as well as the commercial value of the large standing stock of alligators in aquaculture ponds, we assert that infectious diseases and parasites are relevant to alligator population health and captive husbandry/maintenance.This is in addition to the general interest in parasites of crocodilians regarding parasite-host cophyly and crocodilian natural history.
Although alligators are among the most iconic and wellstudied reptiles in North America regarding their life history and general biology, studies on the taxonomy and life cycles of their parasites have been relatively limited.This could be in part due to the difficulty in obtaining permission to sample alligators legally, logistical challenges associated with dissecting and processing large alligators in the field, as well as the technical challenges of extracting, handling, and fixing live, minute trematodes in the field.During a statesponsored alligator hunt in coastal Alabama, we necropsied fresh-killed alligators during the hunter registration of alligator carcasses with natural resource agency personnel.From those collections, we herein provide a redescription for Dracovermis occidentalis Brooks and Overstreet 1978 and place it in a phylogenetic analysis for the first time using the large subunit ribosomal DNA (28S) to test relationships and monophyly of the Liolopidae Odhner, 1912.

Materials and methods
Alligators were sampled at a coastal registration station (30°40′19.3"N87°56′06.6"W)operated by the Alabama Division of Wildlife & Freshwater Fisheries for the annual Alabama alligator season hunt during August 2022.We encountered D. occidentalis infecting the intestine of 1 alligator.The infected male alligator (Tag 1329) measured 2.52 m long and was caught by a baited line in the Bon Secour River, Alabama, by Jarrod Pettie.The digestive tract was excised intact, sliced longitudinally to expose the lumen, rinsed in saline, and decanted and placed into acrylic settling columns before the sediment was observed in a petri dish under the dissecting scope and examined for live flukes.
Trematodes intended for morphology as whole-mounts were observed microscopically and fixed following Dutton et al. (2022).Whole mounts were examined and illustrated using an Olympus BX53 microscope (Olympus Corporation, Shinjuku City, Tokyo, Japan) equipped with differential interference contrast, measured using an ocular micrometer, and illustrated using a drawing tube.Measurements are reported in micrometers (μm) as the range followed by the mean, + / − standard deviation, and sample size in parentheses.Measurements of the holotype (USNM 1370158) are provided in brackets.Vouchers were deposited in the National Museum of Natural History's Invertebrate Zoology Collection (Smithsonian Institution, USNM Collection Nos. 1718009, 1718010, 1718012-1718015).Classification and anatomical terms for liolopids follow Brooks and Overstreet (1978) and Dutton et al (2022).
The 28S phylogenetic analysis included the single sequence from the current study and the other taxa used in Dutton et al. (2022).Sequences were aligned with the multiple alignment tool using fast Fourier transform (MAFFT) (Katoh and Standley 2013) and trimmed to the length of the shortest sequence presented herein (1,237 base pairs [bp]) in Geneious Prime Software version 2023.0.4 (Geneious Corp., Auckland, New Zealand).Aligned sequences were reformatted and exported from.fastato.phy to run a maximum likelihood tree (ML).The ML was inferred with IQTREE v.1.16.12 (Nguyen et al. 2015).Substitution model testing was done with ModelFinder (Kalyaanamoorthy et al. 2017) as implemented in IQTREE.After model testing, tree inference was done using best-fitting substitution models (Chernomor et al. 2016).Default tree search parameters were used, except perturbation strength was set to 0.2, and 500 iterations had to be unsuccessful to stop the tree search.Tree inference was performed 20 times with only the tree with the best log-likelihood score reported.Support for relationships was measured with 1000 ultrafast bootstrap replicates (UFBoot) (Hoang et al. 2018).The inferred phylogenetic tree was visualized using FigTree v1.4.4 (Rambaut et al. 2018) and further edited with Adobe Illustrator (Adobe Systems) for visualization purposes.With only 4 sequences available for the 28S, 2 sequences available for the 18S and ITS2, and 1 sequence for the CO1 we have chosen not to run the individual trees for each region but continue to create a genetic library for future use.
Other localities: Horn Island, Jackson County, Mississippi, USA; Chambers, Jefferson, and Liberty counties in southeast Texas, USA; Bon Secour River (present study), Alabama, USA.

Taxonomic remarks
Our specimens were identified as D. occidentalis by having testes in the posterior 1/3 of the body, a pretesticular cirrus sac, a spined and eversible cirrus, a bipartite seminal vesicle, and a post-acetabular vitellarium.While not all of these features were clearly described in its original description, we confirmed the presence of these features in our newly collected specimens and the holotype.Noteworthy also is that our specimens of D. occidentalis were collected adjacent to the type locality for this species in the northern Gulf of Mexico.Brooks and Overstreet's (1978) specimens were contracted upon fixation, whereas ours were heat-killed and formalin-fixed, and we herein correct several errors related to D. occidentalis in Brooks and Overstreet (1978) that were due to the poor quality of the types.First, Brooks and Overstreet (1978) stated that the body surface of Dracovermis spp.lacks spines; however, in our specimens of D. occidentalis we observed spines scattered about the ventral surface of the forebody.These spines were absent or indistinct in the holotype of D. occidentalis, perhaps related to poor specimen condition.Second, the body shape of Dracovermis spp. is elongate, having equally broadly rounded anterior and posterior ends.This pattern is obscured in the original published description of D. occidentalis because of the poor quality of the types.The body shape of D. occidentalis depicted in Brooks and Overstreet (1978) is misleading in that the body appears compact and pyriform (similar to many other trematodes), not elongate and linguiform like in other heat-killed specimens of Liolopidae.Our heat-killed and formalin fixed specimens of D. occidentalis demonstrate the natural habitus of the species.Third, the ventral sucker is relatively thin and delicate with muscle fibers visible along the periphery of the sucker.Brooks and Overstreet (1978) erroneously stylized the ventral sucker as being thick-walled and extensively muscular, when in fact, the sucker appears similar to that of other liolopids.Fourth, the metraterm in heat-killed and formalin-fixed specimens is straight and strongly muscular.The holotype of D. occidentalis is strongly contracted such that the metraterm presents as a compressed coil, whereas in heat-killed specimens the metraterm is straight because the specimen is not contracted.Brooks and Overstreet (1978) diagnosed D. occidentalis as having a "folded metraterm."Fifth, the vitellarium in our specimens of D. occidentalis is patchily distributed, having gaps among clusters of vitelline follicles.Brooks and Overstreet (1978) highly stylized the vitellarium as a contiguous and evenly distributed field.Although typical for some trematode groups, it is taxonomically significant that the vitellarium of D. occidentalis is patchy (not evenly distributed) and can be confluent posteriorly or not confluent posteriorly.Lastly, we were unable to measure the esophagus and oötype in the type specimen due to the contracted nature of the holotype specimen.Dutton et al. (2022) omitted the following species from their table listing records of liolopids: Harmotrema indica Chattopadhyaya, 1970 from "Enhydrina schistosa" (= Hydrophis schistosus Daudin, 1803) (Squamata: Elapidae) from off of the coast of Bombay, Mumbai, India (Chattopadhyaya, 1970); Harmotrema microrchis Bhutta and Khan 1975 from Gavialis gangeticus (Gmelin, 1789) (Crocodilia: Gavialidae) from the Sutlej River, Pakistan (Bhutta and Khan 1975); and Harmotrema linguiforme Wang 1987 from Hydrophis cyanocinctus Daudin, 1803 (Squamata: Elapidae) from Fujian, China (Wang 1987).

Discussion
We accept 3 nominal species of Dracovermis: D. occidentalis, D. brayi Brooks andOverstreet 1978, andD. rudolphii (Tubangui andMasilungan, 1936) Brooks andOverstreet 1978. We regard D. nicolli (Mehra, 1936) Brooks and Overstreet 1978 as incertae sedis because it has a pyriform, relatively sharply-tapering anterior body end (vs.the accepted species of Dracovermis all having body ends that are equally rounded) and a vitellarium that extends from the cecal bifurcation to the posterior body end (vs.distributing from the level of the acetabulum or from slightly anterior to the acetabulum to the posterior body end in the accepted species of Dracovermis).It also has a diminutive oral sucker rather than the well-developed, bowl-shaped oral sucker of accepted Dracovermis spp.Based on the combination of these features, D. nicolli could represent a new genus of Liolopidae.Dutton et al. (2022) stated that previous studies lacked adequate taxon sampling to assess liolopid phylogenetic relationships and liolopid-vertebrate cophyly.That situation has not changed because currently only 4 species (1 from each genus) have been sequenced (Fig. 3).The tree recovered herein again is equivocal regarding vertebrate-liolopid cophyly because Dracovermis is sister to the liolopids infecting snakes and amphibians.
We examined adult alligators, 20 from eastern South Carolina (Crawl Creek, Santee River), 8 from southern Alabama (Mobile Bay), 5 from Pascagoula River (Mississippi) and Mississippi Sound (northern Gulf of Mexico) (SAB, unpublished data) as well as 15 juvenile alligators from the type locality for D. occidentalis (Rockefeller Wildlife Refuge, Cameron Parish, Grand Chenier, Louisiana).Of those, only one adult alligator from Mobile Bay was infected.Scott et al. (1997), the only other report of D. occidentalis since 1978, documented infections (prevalence = 1 of 25; intensity = 2) of D. occidentalis in adult alligators from the Trinity River, Texas.Their analysis showed that infection by D. occidentalis was significantly higher among adult alligators (no juvenile alligator was infected).The present publication is the first to examine a large number of juvenile alligators for a liolopid infection.Our data and that of Scott et al. (1997) indicate that juvenile alligators lack infections of D.
occidentalis.We suspect that this could be related to a dietary shift of adult alligators eating large aquatic vertebrates (e.g., fishes, reptiles, mammals) (Delany 1990;Delany et al. 1999;Platt et al. 1990;Scott et al. 1997;Taylor 1986;Wolfe et al. 1987) that could be the second intermediate host for D. occidentalis.Regardless of the mechanism(s), our results clearly show that alligators from different areas have different parasites and that certain parasite species can be rare across the known geographic range of the alligator.