Hemiurid Trematodes (Digenea: Hemiuridae) from Marine Fishes off the Coast of Rio de Janeiro, Brazil, with Novel Molecular Data

Simple Summary Brazil, with its remarkably diverse marine habitats, harbour one of the world’s richest fish diversities. Consequently, the diversity of their trematode parasites is also expected to be extremely rich. However, our current knowledge on this group of animals is incomplete and there are many unknown trematode species that await discovery and genetic characterisation. The Hemiuridae (Digenea) is the second most speciose trematode family in marine fishes from Brazil; however, to date, it remains understudied. We examined forty-three specimens of nine fish species belonging to eight families (Carangidae, Clupeidae, Haemulidae, Muraenidae, Percophidae, Pinguipedidae, Trichiuridae, and Triglidae) collected from the coastal zone off Rio de Janeiro, Brazil and found hemiurid trematodes in the stomach of 14 fishes. Using morphological and molecular analyses, we identified eight species from four genera of the family Hemiuridae. One of these species is reported in Brazil for the first time, four are reported from new fish hosts, and four were genetically characterised for the first time. Our novel data contributes to the knowledge on marine biodiversity in Brazil and will further contribute to the classification of the family Hemiuridae. Abstract Brazil is a tropical country with remarkably diverse marine habitats that harbour a rich diversity of fish. Only a small portion of this fish diversity has been investigated for parasites, and thus the diversity of their trematode parasites remains unexplored. Moreover, only 5 out of 184 known digenean trematode species of marine fish in Brazil have been genetically characterised. The Hemiuridae Looss, 1899 is the second most speciose trematode family in marine fishes from Brazil but, in many ways, it remains a neglected group. Forty-three trematode specimens from nine fish species were collected from the coastal zone off Rio de Janeiro, Brazil. Trematodes were found in the stomach of 14 specimens of 9 fish species belonging to 8 families (Carangidae, Clupeidae, Haemulidae, Muraenidae, Percophidae, Pinguipedidae, Trichiuridae, and Triglidae). Trematode specimens were studied using morphological and molecular genetic analyses. A total of eight hemiurid species from four genera, Ectenurus, Lecithochirium, Myosaccium, and Parahemiurus were identified. This paper reports on new host records for four species of hemiurids, adds a new record on the geographical distribution for one species, and provides the first DNA sequence data supplemented with the detailed description of morphology for five species. Phylogenetic analyses supported that the subfamily classifications of the Hemiuridae—based entirely on morphological characters—needs to be reconsidered, taking into account a wider range of information sources.


Introduction
Brazil is a tropical country with remarkably diverse marine habitats that harbour one of the world's highest diversities of fish [1,2]. Based on the assumption that parasite diversity is positively correlated with host diversity [3], parasites ought to be one of the (957 nt long) included 37 28S rDNA sequences of the species from the Hemiuridae; 14 sequences of 7 species generated in the present study. Alignment 2 (416 nt long) included six ITS2 sequences of Lecithochirium spp.; three sequences of three species generated in the present study. Alignment 3 (443 nt long) included eight cox1 sequences of four species of Lecithochirium; six sequences of four species generated in the present study. Pairwise genetic distances (uncorrected p-distance and no. of differences) for the three datasets were calculated in MEGA ver. 11 [32] using the following conditions: "Variance Estimation Method = None", "Model/Method = p-distance or no. of differences", "Substitutions to Include = d: Transitions + Transversions" and "Gaps/Missing Data Treatment = Pairwise deletion".
Phylogenetic relationships of the taxa in Alignment 1 were assessed using Bayesian inference (BI) and maximum likelihood (ML) analyses. Sequences of Isoparochis eurytremus (MH628315), a parasite of Silurus asotus Linnaeus from Takashima, Japan, was used as the outgroup based on the topology in the phylogenetic tree of the superfamily Hemiuroidea by Louvard et al. [13]. The analyses were conducted using the GTR + I + G model, which was predicted as the best model by the Akaike Information Criterion in jModelTest 2.1.2 [33]. BI analysis was performed using MrBayes software (ver. 3.2.3) [34] through the CIPRES Science Gateway ver. 3.3 [35] accessed on 22 September 2022. Markov Chain Monte Carlo chains were run for 10,000,000 generations, log-likelihood scores were plotted and only the final 75% of trees were used to build the consensus tree. ML analysis was performed using PhyML ver. 3.0 [36] and run on the ATGC bioinformatics platform (http://www. atgc-montpellier.fr, accessed on 22 September 2022) with a nonparametric bootstrap value of 100 pseudoreplicates.

Results
Evaluation of the specimens of hemiurid trematodes obtained in the present study via morphological and molecular methods confirmed the presence of eight trematode species in the examined fishes. Out of the 43 specimens of 9 fish species examined, 14 (32.5%) were found to be infected with at least 1 species of hemiurid and 4 (9.3%) were infected with 2 species of hemiurids. All specimens of hemiurids were detected in the fish stomach. Thirty sequences were newly generated; twenty-eight sequences for seven out of eight recorded species of trematodes: 28S rDNA (n = 14), ITS1-5.8S-ITS2 (n = 1), ITS2 (n = 3) and cox1 (n = 10). Sequences of the unidentified species of Lecithochirium were not generated. Two cox1 sequences were generated for the fish hosts, Pseudopercis numida and Trichiurus lepturus. The photographs of the fish specimens that were used to generate sequences are provided in Figure S1. Fish hosts were identified following Menezes and Figueiredo [37]. The cox1 sequences for fish obtained in the present study were compared with sequences available in GenBank. The sequence divergence between our specimen of T. lepturus (OP905634) and specimens of T. lepturus (GU702467; JX124915) collected in São Paulo, Brazil was low and ranged from 0.2 to 0.3% (1-2 nt). Thus, we consider these isolates as conspecific. The sequence divergence between our specimen of Ps. numida (OP925860) and specimens of Ps. semifasciata (JQ365526; EU074573) from São Paulo, Brazil and from Argentina was low and ranged from 0 to 1.6% (0-8 nt). Although the genetic divergence between our specimen identified as Ps. numida and specimens of Ps. semifasciata available in GenBank suggests that these specimens are conspecific, morphologically our specimens belong to Ps. numida.  Pre-oral lobe distinct, 25 long. Oral sucker muscular, well developed, subspherical or transversely oval, ventro-subterminal, 76 long, 86 wide. Prepharynx absent. Pharynx muscular, well developed, spherical or subspherical, 45 long, 45 wide. Oesophagus absent. "Drüsenmagen" present. Intestinal bifurcation just posterior to pharynx. Caeca blind, with thin walls and wide lumen, terminates in posterior body extremity. Ventral sucker muscular, well developed, spherical or subspherical, 129 long, 139 wide, larger than oral sucker (1:1.62), pre-equatorial.
Ovary median, entire, subspherical (n = 3) or transversely oval (n = 2), 48 long, 68 wide, in anterior half or in middle of hindbody, contiguous with sinistral testis. Vitellarium in two compact oblique masses, subspherical; dextral mass posterior to ovary, 37 long, 41 wide, sinistral mass lateral to ovary, 41 long, 39 wide. Juel's organ and Mehlis' gland not observed. Uterus coiled, occupies post-ovarian region. Metraterm not differentiated, terminal part of uterus passes into sinus-sac ventrally, joins male duct forming hermaphroditic duct. Eggs numerous, 25-26 × 11-13 (n = 4) (Figure 1c). Excretory vesicle not observed; excretory pore terminal. Remarks: Specimens found in the present study correspond well to the generic diagnosis of Myosaccium Montgomery, 1957 provided by Gibson [8] in having a saccular and elongate-oval seminal vesicle posterior to the middle of the ventral sucker, a vesicular pars prostatica with a muscular wall, a tubular sinus-sac enclosing the hermaphrodite duct, and vitellarium composed of two distinct masses.
Myosaccium ecaude is known as a parasite of the stomach of fishes from the family Clupeidae in the Pacific and Atlantic oceans. After the original description, the species was reported in Op. libertate and Harengula thrissima (Jordan and Gilbert) in the Chamela Bay, Jalisco, Mexico [39,40], in the type host Sa. sagax in California, USA, and Baja California, Mexico, north-eastern Pacific Ocean [41][42][43] and in H. clupeola and S. brasiliensis in Rio de Janeiro, Brazil, southwestern Atlantic Ocean [44,45].
Myosaccium ecaude was originally described as possessing filamented eggs. Overstreet [46] and León-Règagnon et al. [39] re-evaluated the type material and provided different conclusions. The first author concluded that "specimens from Sardinella anchovia do not have filaments or spines on the eggs, although a look at collapsed specimens on a fixed plane strongly suggests their presence". Additionally, the same author did not observe morphological differences between M. ecaude and the only other species of the genus, M. opisthonemae (Siddiqi and Cable, 1960) described from Op. oglinum (Lesueur) in Playa Mani, Puerto Rico, by Siddiqi and Cable [47], north-western Atlantic Ocean, and suggested that M. opisthonemae might be a small or progenetic form of M. ecaude. León-Règagnon with co-authors [39] observed the presence of filamented eggs in the type material and in their newly collected specimens from fish in Mexico. These authors suggested that the presence of filamented eggs is a strong characteristic to distinguish M. ecaude from M. opisthonemae. Thereafter, Kohn and Buhrnheim [48] collected specimens of M. ecaude in S. aurita in the same region to that of the present study and described their specimens as having filamented and spined eggs. However, later Kohn et al. [4] corrected their species identification in Kohn and Buhrnheim [48] to M. opisthonemae without further explanation.
In our material, the hologenophore possess collapsed and non-collapsed eggs. The collapsed eggs may look as if they bear filaments and spines; however, after detailed examination, we concluded that these structures are absent (Figure 1c). Therefore, we agree with Overstreet [46] who, after re-examining the type material of M. ecaude, did not report on filaments or spines on the eggs. Although Overstreet [46] did not observe morphological differences between M. ecaude and M. opisthonemae, there are morphometric differences that can be noticed (Table 1). Newly collected material of M. ecaude from S. brasiliensis in the present study was used to provide a detailed description of the species together with novel DNA sequence.
Excretory vesicle not observed; excretory pore terminal. Remarks: Specimens obtained in this study agree well with the generic diagnosis of Ectenurus Looss, 1907 provided by Gibson [8] in having a plicated body surface, a seminal vesicle divided into three portions in the anterior hindbody, the presence of a sinus-sac, a permanent sinus-organ, and a short pars prostatica connected to the seminal vesicle by a distinct, long, and occasionally convoluted aglandular duct.
In Brazil, Travassos et al. [59] reported E. lepidus from Oligoplites saurus (Bloch and Schneider) in Espírito Santo. However, based on their morphological description, Gibson and Bray [60] claimed that Travassos et al. [59] found E. virgula. Thereafter, Pereira et al. [61] reported E. virgula from the phycid U. brasiliensis (Kaup) in Rio de Janeiro. Our record of E. virgula infecting A. virginicus, D. punctatus, and Pr. punctatus collected off the Brazilian coast represent a new host record. Specimens of E. virgula collected in our study contributed to a novel detailed description of the species and to generation of DNA sequences. Description (Figure 2a,b). (Based on four paragenophores and two hologenophores; measurements of paragenophores in Table 3 and hologenophores in description): Body elongate, dorso-ventrally flattened, 993-1174 long. Maximum width close to posterior body extremity, 345-393. Tegument covered with conspicuous plications to level of posterior margin of ovary. Forebody short, 138-158, representing 13-14% of body length. Ecsoma well developed, withdrawn (n = 2), protruded (n = 3) or partially extruded (n = 1).
Excretory vesicle not observed; excretory pore terminal. Remarks: Specimens found in the present study correspond well to the generic diagnosis of Parahemiurus Vaz and Pereira, 1930 provided by Gibson [8] in having a well-developed ecsoma, the absence of an ejaculatory vesicle, in possessing a tubular sinus-sac, a vitellarium composed of two distinct oval masses, a plicated body surface, an oval seminal vesicle with a muscular wall, and a tubular pars prostatica.
Our specimens correspond in their morphology to Parahemiurus merus (Linton, 1910) described as Hemiurus merus from the stomach of Sardinella aurita (=Clupanodon pseudohispaniciis) in Tortugas, Florida, USA, north-western Atlantic Ocean by Linton [49]. Specimens corresponded in body shape, having plicated body surface to level of the vitellarium, in possessing a narrow and elongate sinus-sac, an oval seminal vesicle, and contiguous testes. However, except for the egg size (22-27 × 7-11 vs. 27 × 10) (Table 3), the maxima for all dimensions in our specimens was lower. The features and dimensions of our specimens overlap with specimens collected from Sardinella aurita Valenciennes in São Paulo, Brazil, specimens collected from Engraulis anchoita Hubbs and Marini off the Argentinean and Uruguayan coasts, southwestern Atlantic Ocean by Vaz and Pereira [63] and Timi et al. [64], specimens collected from several fish hosts of the families Carangidae, Clupeidae, Haemulidae, Merlucciidae, Pomatomidae, Salmonidae, Scorpaenidae, and Sparidae in different localities from the Atlantic, Indian, and Pacific oceans in a review published by Bray [65], and specimens collected from several fish hosts of the families Balistidae, Clupeidae, Engraulidae, and Haemulidae in Chamela Bay, Jalisco, Mexico, north-eastern Pacific Ocean by León-Règagnon et al. [39] (Table 3).
Parahemiurus merus is reported mainly from clupeid, carangid, salmonid, and engraulid fishes from the Atlantic, Indian, and Pacific oceans [65]. In Brazil, this species has been reported several times in fishes from at least 12 families (Belonidae, Carangidae, Clupeidae, Dactylopteridae, Engraulidae, Haemulidae, Ophidiidae, Phycidae, Pinguipedidae, Pomatomidae, Sciaenidae, and Sparidae) [9]. Wallet and Kohn [66] and Luque et al. [31] reported P. merus in H. clupeola, and Luque et al. [44] and Moreira et al. [45] reported it in S. brasiliensis. Benicio et al. [67] reported it in Cetengraulis edentulus (Cubier) in the same region as in our study. Newly collected specimens of P. merus have allowed us to provide a detailed description of the species and generate the first DNA sequence data.
Remarks: Specimens found in the present study correspond well to the generic diagnosis of Lecithochirium Luhe, 1901 provided by Gibson and Bray [60] and Gibson [8] in having a well-developed ecsoma, a pre-oral lobe, a tubular pars prostatica, a vitellarium of two lateral masses divided into three and four short lobes, and eggs without polar filaments. Comparative sequence analyses (see below) demonstrated that the 28S rDNA and cox1 sequences of our isolate from the percophid P. brasiliensis identified as L. floridense was identical to sequences of the same species reported in Auxis thazard (Lacépède) in Rio de Janeiro, Brazil by Pantoja et al. [7].
Lecithochirium floridense is a parasite of the stomach of a variety of marine fish species. To date, L. floridense has been reported from fishes from at least 17 families, including the present data, with the majority of records coming from the western Atlantic Ocean (Pantoja et al. [7]). In Brazil, L. floridense was reported from A. thazard in Rio de Janeiro, southwestern Atlantic Ocean. Our record of L. floridense infecting Pe. brasiliensis collected off the Brazilian coast represents a new host record for this species. We do not provide a description of the specimens of this species because a detailed morphological description of this species with DNA sequence data was provided by Pantoja et al. [7].
Excretory vesicle not observed; excretory pore terminal. Remarks: Specimens found in the present study correspond well to the generic diagnosis of Lecithochirium in characters as mentioned above. Following the key to the species-group of Lecithochirium proposed by Bray [68], our specimens belong to the 'Microstomum-group'. This diagnosis is based on the presence of a non-glandular presomatic pit, a vitellarium of compact masses with distinct and digitiform lobes, terminal genitalia of the 'musculus' type, a non-muscular seminal vesicle, and the absence of internal elevations of the ventral sucker.
In Brazil, L. microstomum has been reported from fishes of 11 families (Carangidae, Engraulidae, Gempylidae, Gerreidae, Paralichthyidae, Percophidae, Phycidae, Pinguipedidae, Sciaenidae, Scombridae, and Serranidae) [9,67]. Teixeira de Freitas and Kohn [71], Wallet and Kohn [66], Silva et al. [77,78], and Carvalho and Luque [79] reported L. microstomum from the same fish host, T. lepturus, and the same locality as in the present study. Newly collected material of L. microstomum in the present study has allowed us to provide a detailed description of the species together with DNA sequence data.
Excretory vesicle not observed; excretory pore terminal. Remarks: Specimens collected in this study possess features that fully correspond to the generic diagnosis of the genus Lecithochirium in characters as mentioned above.
According to the key to the species-group of Lecithochirium proposed by Bray [68], our specimens belong to the 'Microstomum-group' based on the presence of a non-glandular presomatic pit, the absence of internal elevations in the ventral sucker, a non-muscular seminal vesicle, terminal genitalia of the 'musculus' type, and a vitellarium represented by compact masses with distinct and digitiform lobes.
Comparing to the species in the 'Microstomum-group', our specimens could be distinguished from L. alectis in the position of the seminal vesicle (antero-dorsal to the ventral sucker vs. entirely anterior to the ventral sucker), from L. mecosaccum in possessing a short sinus-sac (116-129 vs. 170-300), and from L. antennari, L. chaetodontis, L. maomao, L. microstomum, and L. priacanthi in the position of the testes (obliquely symmetrical vs. obliquely tandem).
Morphologically, our material is most similar to the species of L. muraenae described by Manter [80], from the stomach of the Hourglass moray Muraena clepsydra Gilbert in Cape Elena, Ecuador, southeastern Pacific Ocean, in that they possess a conspicuous and non-glandular presomatic pit, symmetrical testes, a tripartite seminal vesicle, and a vitellarium consisting of two lateral masses divided into three and four short lobes. Nonetheless, specimens in our study differ from the material of Manter [80] by having smaller dimensions for all internal organs (Table 5), in the position of the genital pore (at level of intestinal bifurcation vs. posterior to intestinal bifurcation), and in possessing a straight instead of convoluted hermaphroditic duct. The variation in metrical data, in our opinion, relates to the differences in the fixation method (heat-killed fixation of our material vs. fixation under a cover glass, with the application of slight pressure of the material in Manter [80]). Considering the similarities in morphology and group of hosts (muraenids), we provisionally identify our specimens as L. muraenae.
Lecithochirium muraenae is a parasite of muraenid fishes. After the original description, the species was reported in Gymnothorax porphyreus Guichenot (=G. wieneri) in El Callao, Peru, southeastern Pacific Ocean [81]. Our record of L. muraenae infecting G. vicinus off the Brazilian coast potentially represents a new host and geographical record for this species. Additionally, we provide DNA sequence data that may further help to elucidate the taxonomic identity of our specimens.
Remarks: Specimens found in the present study correspond well to the generic diagnosis of Lecithochirium in characters as mentioned above. Our specimens collected from the haemulid A. virginicus and from the pinguipedid Ps. numida are morphologically similar to those described from A. thazard by Pantoja et al. [7] in Rio de Janeiro, Brazil. The identification of our specimens as L. synodi was confirmed via comparative sequence analyses, which demonstrated that the 28S rDNA sequence of our isolate of Lecithochirium was identical to two isolates of L. synodi reported by Pantoja et al. [7]. Their cox1 sequences differed by 0.23-0.45% (1-2 nt). Lecithochirium synodi is a parasite of the stomach of a variety of marine fish species. To date, this species has been reported from fishes from at least six families (Haemulidae, Monocanthidae, Paralichthyidae, Pinguipedidae, Scombridae, and Synodontidae), including the present data, with the majority of records coming from the western Atlantic Ocean [7]. Our record of L. synodi infecting A. virginicus and Ps. numida collected off the Brazilian coast represents two new host records for this species. We do not provide the description of the specimens of this species because a detailed morphological description of this species with DNA sequence data was recently provided by Pantoja et al. [7].
Ovary dextral, entire, transversely oval, in posterior half of hindbody, always separated from posterior testis by uterine coils, contiguous with vitellarium. Vitellarium in two lateral compact masses, divided into three and four short lobes. Juel's organ and Mehlis' gland not observed. Uterus coiled, restricted to body. Metraterm passes into sinus-sac ventrally, joins male duct just distally to ejaculatory vesicle forming hermaphroditic duct. Eggs numerous, small. Excretory vesicle not observed; excretory pore terminal.

Remarks:
The specimen found in the present study correspond well to the generic diagnosis of Lecithochirium in characters as mentioned above. Following the key to the speciesgroup of Lecithochirium proposed by Bray [68], our specimen belongs to the "Musculusgroup" based on the absence of a presomatic pit, the presence of a vitellarium comprised of compact masses with distinct and short digitiform lobes, a terminal genitalia of the "musculus" type, a non-muscular seminal vesicle, and the absence of internal elevations in the ventral sucker.
In comparison to species from the "Musculus-group", our specimen can be distinguished from all other species; from L. brevicirrus (Nicoll, 1915), L. floridense (Manter, 1934), L. medius Acena, 1941, L. microcercus (Manter, 1947) and L. musculus (Looss, 1907) in the position of the testes (not contiguous with the ventral sucker vs. contiguous or just posterior to the ventral sucker or overlapping it), from L. imocavus by having a distinctly wider body (694 vs. 250-400) and from L. monticellii (Linton, 1898) in the position of the testes (contiguous vs. separated by uterine coils).
Out of the "Musculus-group", our specimen closely resembles L. trichiuri ('Keokeogroup') described from the same host, T. lepturus (=T. haumela) in the China Sea by Gu and Shen [83]; it is similar in body shape, it possesses testes separated from the ventral sucker, and has a similar ratio of the suckers (1:2.46 vs. 1:2.4-2.9). However, our specimen lacks a presomatic pit, whereas L. trichiuri possess a presomatic pit.
Due to the presence of a single specimen in our material and difficulties in assigning this specimen to any known species, we identify it to the genus level as Lecithochirium sp.

Molecular Results
In this study, 30 novel sequences were generated for 16 isolates. The phylogenetic relationships of the studied species of the Hemiuridae were assessed based on the partial 28S rDNA sequences. Sequences of ITS1-5.8S-ITS2, ITS2, and cox1 were used to calculate the pairwise genetic distances between species and to contribute to a growing DNA sequence library for the Hemiuridae. The phylogenetic tree obtained from the BI analysis based on Alignment 1 (28S rDNA; 1071 nt) is presented in Figure 4. Pairwise genetic distances of this dataset are presented in Table S1. Novel 28S rDNA sequences (n = 14) of seven species were positioned in different clades with the members of the family Hemiuridae.
The most surprising result of our molecular analyses is that a sequence of M. ecaude (OP918123), a member of the subfamily Aphanurinae according to classifications based on morphology, clustered with a sequence of Hemiurus levinseni (MN962990) (a member of the Hemiurinae) collected from Cylichna alba (Brown) in White Sea, Russia with strong support. Two identical sequences of P. merus generated in the present study (members of the Hemiurinae) and a sequence of Hemiurus appendiculatus (Rudolphi, 1802) clustered in the same clade, albeit with low support. The sequence divergence between M. ecaude and He. levinseni, He. appendiculatus, and P. merus was 4.84-9.12% (50-79 nt). In comparison with Aphanurus mugilus Tang, 1981 (LT607807), a species of the type genus of the Aphanurinae, our sequence of M. ecaude differed by 8.32% (86 nt). This result suggests that the position of M. ecaude within the Aphanurinae requires revaluation.
Our molecular phylogenetic analyses confirmed the positions of E. virgula within the subfamily Dinurinae and Lecithochirium spp. within the subfamily Lecithochiriinae. The three sequences of E. virgula (OP918121, OP918122, and OP918126) were identical and clustered with sequences of members of the same subfamily Dinurinae, Di. euthynni (OP458333) collected from A. thazard in Rio de Janeiro, Brazil, and Di. longisinus (AY222202) collected from Coryphaena hippurus Linnaeus, Jamaica. The sequence divergence between these three species ranged from 3.21 to 3.80% (33-49 nt). All sequences generated for species of Lecithochirium in this study clustered in a strongly supported clade. The three novel sequences of L. synodi (GenBank OP918129, OP918130, and OP918132) clustered with sequences of two isolates of the same species (OP458330 and OP458331) found in A. thazard in Rio de Janeiro, Brazil and L. microstomum Chandler, 1935 (KC985235) collected from T. lepturus in the USA. All five sequences of L. synodi were identical. The three novel sequences of L. microstomum (OP918119, OP918120, and OP918127) were identical and clustered with Lecithochirium sp. (MK648288) collected from the same host, T. lepturus in Veracruz, Mexico in the same strongly supported sub-the DNA sequence divergence between E. virgula and Di. euthynni was lower than the divergence between congeneric Di. euthynni and Di. longisinus (3.27%, 34 nt vs. 3.80%, 39 nt, respectively). Morphologically, members of both genera are similar in that they have a plicated body surface, a partitioned seminal vesicle, and the presence of a sinus-organ. However, they differ in the length of the pars prostatica.
Lecithochirium is one of the most speciose genera of the family Hemiuridae. The species recognition of the genus, in our view, is often problematic due to the poor descriptions for many species and the lack of critical differentiation analysis at the time of description. Thus, the true species composition of this genus can only be assessed by new, accurate, and precise work, including detailed morphological descriptions together with the incorporation of DNA sequence data. In the present study, the genus Lecithochirium is the richest and is represented by five species. Two of these, L. floridense and L. synodi, were recently reported from A. thazard in Brazil with a morphological description and DNA sequences provided [7]. Therefore, the identification of these species was based on both morphological and comparative sequence analyses. Our findings suggest that the host spectrum for both species is wider. Although L. floridense is reported for the first time in a percophid fish and L. synodi in haemulid and pinguipedid fishes, the euroxenous nature of both species was already observed in previous studies [7,86].
For the three remaining species of Lecithochirium, we provided detailed morphological descriptions, with two of them being genetically characterised, in turn, further helping to avoid ambiguities in species delimitation. Lecithochirium microstomum is an euryxenous parasite widely spread across the tropical and subtropical Atlantic and Pacific oceans [87]. This species is the most reported trematode species in marine fishes form Brazil. So far, fishes from 11 families of 8 orders are known to be the hosts of L. microstomum in the country [9,67]. Our report of L. microstomum in a triglid fish, Pr. punctatus, increased the host spectrum for this species. This species was previously reported five times in the same host, T. lepturus, and in the same locality in Brazil as in the present study [9]. Although numerous records of L. microstomum are available, there were no associated DNA sequence data for this species in Brazil. However, one 28S rDNA sequence of L. microstomum (KC985235) from T. lepturus collected in Gulf of Mexico, Mississippi, USA, by Calhoun et al. [88] is available and our isolate clustered in the same clade with it; the intraspecific sequence divergence was considered high (1.5%; 16 nt) when compared with the intraspecific sequence divergence observed in other Lecithochirium spp. (Figure 4 and Table S1). A morphological description was not provided with the published sequence, and therefore we cannot compare the morphology of the isolates. Based on morphological features and comparisons of the metrical data, our specimens agree with the descriptions of L. microstomum. An unidentified species of Lecithochirium (GenBank number MK648288) collected from T. lepturus in Veracruz, Mexico [89]-even with the lack of a reference morphological voucher for this sequence-is likely conspecific with L. microstomum found in the present study. The difference between the 28S rDNA sequences of our isolates and the isolate from Mexico was low, 0.47% (5 nt).
Molecular data showed that L. synodi and L. microstomum are closely related, although they belong to different morphological groups, i.e., "Synodi-group" and "Microstomumgroup". Morphological similarities are also remarkable, i.e., the absence of internal elevations of the ventral sucker, a tripartite seminal vesicle, the presence of a presomatic pit, obliquely tandem testes, and testes separated from the ventral sucker. A morphological feature that separates L. synodi and L. microstomum between different groups according to the key of Bray [68] is the type of presomatic pit (glandular in the "Synodi-group" and non-glandular in the "Microstomum-group"). Our specimens of L. microstomum possess a non-glandular presomatic pit. However, León-Regàgnon and co-authors [39] examined the type material of L. microstomum and concluded that the presomatic pit is glandular in the material of Chandler [70]. Due to the genus being specious, at the present stage it is not possible to identify which features are more phylogenetically important. Detailed morphological descriptions including molecular data are only available for 5 out of more than 100 nominal species of the genus [7,68,90].
Specimens tentatively identified as L. muraenae in our study were reported for the first time in G. vicinus and in Brazil. Prior to our study, this species was only known from the Pacific Ocean [80,81]. The record of this species in the Atlantic Ocean demonstrates that its geographical distribution is possibly wider than previously reported. Lecithochirium muraenae is a stenoxenous parasite infecting muraenid fishes. DNA sequence data for L. muraenae from muraenids in the Pacific Ocean were not available for comparison.
Interestingly, isolates of Lecithochirium cf. muraenae clustered distantly from other Lecithochirium spp. recorded in this study. It formed a well-supported clade with an unidentified species of Lecithochirium (ON614672) collected from Octopus bimaculatus in Mexico [90] and Pu. cyanovitellosus (the subfamily Pulmoverminae). Although the specimens from octopuses were immature, there are a couple of morphological similarities that we could observe when comparing to specimens of Lecithochirium cf. muraenae, these are: the position of the genital pore at the level of intestinal bifurcation, a blind caeca terminating in the body, symmetrical testes positioned laterally posterior to the ventral sucker, and an ovary closely posterior to dextral or sinistral testis. Morphologically, Lecithochirium cf. muraenae and Pu. cyanovitellosus conspicuously differ in body shape (plump, piriform vs. elongate, slender), the size of the ecsoma (well-developed vs. reduced), the shape and size of the seminal vesicle (short and straight vs. long and convoluted), and the position of the testes in relation to the ventral sucker and to each other (contiguous with the ventral sucker and symmetrical vs. separated from the ventral sucker and tandem). Additionally, these species parasitise different groups of hosts (fish vs. sea-snake) and were found in different sites in the hosts (stomach vs. lung). The published sequence of Pu. cyanovitellosus was not supplemented with either morphological descriptions of the isolate or an electronic voucher, and thus no data is available for morphological comparison of the actual isolates.
Based on their morphology, species of L. muraenae and L. microstomum were placed in the same group, the "Microstomum group" [68]. However, the comparative sequence analysis suggests that these species are rather genetically distant, and thus further revaluation of morphotype groups with DNA sequence data, including data for the type-species, L. rufoviride (Rudolphi, 1819), will aid in the identification of a valid composition of each group.
The genus Myosaccium accommodates only two species which parasitise marine fishes of the Clupeidae and Carangidae, with their distribution restricted to the Atlantic and Pacific oceans [39]. Due to strong morphological similarity, Overstreet [46] suggested M. opisthonemae to be a synonym of M. ecaude. Although specimens of M. opisthonemae are smaller than those identified as M. ecaude, the main feature used to distinguish between them-the presence or absence of spines and polar filament on eggs-was demonstrated to be erroneous. Molecular genetic characterisation of smaller specimens is needed to assess the taxonomic status of M. opisthonemae. Myosaccium ecaude is a stenoxenous parasite and has previously been reported only in clupeid fishes, whereas M. opisthonemae is an euryxenous parasite and has been reported in clupeid and carangid fishes [9,16].
The relative relation of M. ecaude to He. levinseni in the clade containing the members of the subfamily Hemiurinae was rather surprising (Figure 4) because the genus Myosaccium currently belongs to the subfamily Aphanurinae. We found no convincing morphological characters, potentially justifying the move of Myosaccium into the Hemiurinae. The main feature differentiating Myosaccium and Aphanurus (type genus of the Aphanurinae) is the shape of the vitellarium (two distinct masses vs. one distinct mass, respectively), whereas Hemiurus and Parahemiurus possess vitellarium composed of two distinct masses. The taxonomic value of this character for subfamily classification is still to be confirmed.
The genus Parahemiurus is represented by 22 species [91]. In Brazil, only two species have been reported, the type-species P. merus and P. anchoviae Pereira and Vaz, 1930 [9]. Parahemiurus merus is an euryxenous parasite tending to be most frequently reported in temperate pelagic fishes and is rarely reported in the vast area covered by the tropical Indo-West Pacific region [65]. This species is the second most reported trematode species in marine fishes in Brazil. So far, fishes from 12 families of 9 orders are known to be the hosts of P. merus in Brazil [9]. Despite numerous records of P. merus, there were no DNA sequence data available for this species. Accordingly, and for the first time, we provided DNA sequences for the genus Parahemiurus and confirmed its position within the Hemiurinae based on molecular data.
According to the present and previous phylogenetic analyses of Atopkin et al. [11], Faltýnková et al. [92], and Sokolov et al. [12], it is evident that the subfamily classifications of the Hemiuridae, based entirely on morphological characters, needs to be reconsidered taking into account a wide range information sources, including morphological, molecular, and ecological data. Although the molecular library of the Hemiuridae is slowly growing, the lack of DNA sequences of numerous taxa, the limitation of data from single conservative molecular markers, and the absence of morphological vouchers of the actual isolates used to generate sequences, represents an impediment for the clarification of the taxonomic content of the family. Accordingly, such information is also imperative for understanding the value of the morphological features used to define subfamilies.

Conclusions
Based on morphological and molecular genetic analyses, we identified eight hemiurid species belonging to four genera parasitising marine fishes in Brazil. This is the first study to report phylogenetic relationships based on 28S rDNA sequences of members of three hemiurid genera, Ectenurus, Myosaccium, and Parahemiurus. These data contribute to the knowledge of the systematics and evolutionary history of the Hemiuridae which remains incomplete. In phylogenetic analysis, the aphanurines Myosaccium and Aphanurus were rendered paraphyletic. This data casts doubts over the traditional subfamily classification based on the morphology of the members from these genera, which will need to be rearranged in the future when a higher number of sequenced hemiurid taxa will be available. The record of the specimens tentatively identified as Lecithochirium muraenae increases the number of hemiurid species off the Brazilian coast. The first records of digenean trematodes in the fishes Anisotremus virginicus and Decapterus punctatus in the country, indicates the relatively poor coverage of Brazilian marine digenean fauna. Studies on marine fish trematodes using morphological and molecular datasets remain in their infancy in Brazil. Our findings highlight the importance of exploring this region, which is considered as a parasite biodiversity hotspot [93], and the need to describe digenean diversity using integrative data.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ani12233355/s1, Figure S1: Photomicrographs of the molecular vouchers of fish; Table S1: Pairwise comparisons of genetic distances of the Hemiuridae based on 28S rDNA sequences (957 nt long alignment); Table S2: Nucleotide comparison of the partial ITS2 sequences of Lecithochirium spp. based on 416 nt long alignment. Table S3: Nucleotide comparison of cox1 sequences of Lecithochirium spp. based on 443 nt long alignment.
Author Contributions: Conceptualization, C.P. and O.K.; methodology, C.P. and O.K.; data curation, C.P.; writing-original draft preparation, C.P. and O.K.; writing-review and editing, C.P. and O.K.; project administration, C.P. and O.K. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement:
This study was conducted applying national and international guidelines. All procedures involving animals were conducted in accordance with the Ethics Committee on the Use of Animals (CEUA), Institute of Veterinary, Federal Rural University of Rio de Janeiro, Brazil. Species registration for scientific research purposes was carried out at SisGen (A255046) according to Brazilian legislation on access to the biodiversity.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data generated in this study are available from the corresponding author upon request.