Parasaturnius maurepasi n. gen. et n. comb. (Digena: Bunocotylidae) from the stomach of the silver mullet, Mugil curema (Perciformes: Mugilidae) in coastal lagoons of northern Yucatán, Mexico

Bunocotylid trematodes represent a group of 149 species with a rather complex taxonomic history. The current concept of the subfamily only includes three genera, Bunocotyle, Saturnius, and Robinia. Specimens of a bunocotylid were collected from the silver mullet, M. curema, from a coastal lagoon of Yucatán and identified as belonging to Saturnius. Further detailed morphological study revealed they corresponded to S. maurepasi, a species previously reported from the stripped mullet, Mugil cephalus in Mississippi, USA. Specimens were sequenced for the LSU of nuclear ribosomal RNA gene (28S) to test their phylogenetic position. We discovered that they do not belong in Saturnius since they nest as an independent lineage which is the sister taxa of a clade formed by Robinia, and Saturnius + Bunocotyle; additionally, the new genus exhibits high genetic divergence (10-12%) with respect to species allocated in the other bunocotylid genera. The species S. maurepasi was then transferred to the new genus as Parasaturnius maurepasi n. gen., n. comb. that was created to accommodate it, and was redescribed based on newly sampled specimens.

During a survey on the parasites of marine and estuarine fishes of the Yucatán Peninsula, specimens of the silver mullet, Mugil curema (Valenciennes) were collected in La Carbonera coastal lagoon, in Yucatán.Specimens of a bunocotylid trematode were obtained from the stomachs of their hosts.After comparing morphologically with the original descriptions of members of the family, we identified our specimens as Saturnius maurepasi, with a slight morphological variation.However, while conducting a molecular phylogenetic analysis of the large subunit of the ribosomal gene (28S rDNA), we unexpectedly discovered that our specimens formed a separate monophyletic clade from Saturnius spp., suggesting that they represented an undescribed genus.Here, we present the diagnosis of the genus Parasaturnius n. gen. to include S. maurepasi as a new combination, using morphological and molecular evidence, and we discuss the interrelationships within Bunocotylidae.

Host collection and morphological analysis
A total of 10 individuals of M. curema (8-10 cm) were collected in May 2022 in La Carbonera coastal lagoon, Yucatán State, Mexico (21° 08′ 1.5′′ N, 90° 07′ 55.9′′ W) using cast nets; silver mullets were kept alive and examined for helminths a few hours after capture.Individual fish were euthanized by spinal severance (pithing) following the procedures accepted by the American Veterinary Medical Association (AVMA, 2020), dissected, and immediately examined under a stereomicroscope.Bunocotylids were recovered from the stomach lining of seven of the 10 host examined, fixed in hot distilled water, and preserved in 100% ethanol for morphological and molecular analyses.
Some unflatten specimens were post-fixed in hot formalin to harden the tegument.Specimens were dehydrated through graded alcohol series, stained with Mayer's paracarmine (Merck, Darmstadt, Germany), cleared with methyl salicylate, and mounted on microscope slides with Canada balsam.Mounted specimens were examined under a bright field Nikon DS-Ri1 microscope.Measurements were taken using Nikon NIS Elements microscope software (Nikon) and are given in micrometres (μm).Drawings were made with Adobe Illustrator 25.4.1 (Adobe, Inc).Vouchers were deposited in the Colección Nacional de Helmintos (CNHE), Instituto de Biología, Universidad Nacional Autónoma de México, Mexico City.
For scanning electron microscopy (SEM), specimens were dehydrated in a graded ethanol series, critical point dried, and sputter coated with gold.Then, specimens were examined with a Hitachi Stereoscan Model SU1510 scanning electron microscope at 15 kV at the Laboratorio de Microscopia y Fotografía de la Biodiversidad, Instituto de Biología, Universidad Nacional Autónoma de México.
The two newly obtained sequences were aligned with data from other members of Hemiuroidea downloaded from the GenBank dataset, plus one species from Azygiidae used as outgroup for rooting the tree (see Table 1).The final alignment consisted of 35 sequences with 1,326 nucleotides.Alignments were Vol:.( 1234567890) built using the software Clustal W (Thompson et al., 1997) with default parameters and adjusted manually with the Mesquite software (Maddison & Maddison, 2011).Phylogenetic analyses were performed using Maximum Likelihood (ML) and Bayesian Inference (BI) methods.ML was carried out with RAxML version 7.0.4(Silvestro & Michalak, 2011) and Bayesian Inference analyses were run with MrBayes version 3.2.7 (Ronquist et al., 2012) using the online interface CIPRES (Cyberinfrastructure for Phylogenetic Research) Science Gateway v3.3 (Miller et al., 2010).The best model was estimated with the Akaike information criterion (AIC) using the jModel Test version 0.1.1 program (Posada, 2008), which predicted the best model for the 28S dataset to be GTR + I + G. Nodal ML support was achieved through 1000 bootstrap replicates.The BI analyses included Markov Chain Monte Carlo (MCMC) searches of two simultaneous runs for 10 million generations, with sampling every 1000 generations, a heating parameter value of 0.2, and the first 25% of the sampled trees were discarded.Trees were drawn using FigTree program v.1.4.4 (Rambaut, 2010).Uncorrected P distances were obtained in MEGA version 6 (Tamura et al., 2011).Ventral sucker strongly muscular, sub-spherical, anterior to mid-body.Prepharynx absent; pharynx subspherical.Oesophagus short.'Drüsenmagen' present.Caeca with constrictions at septa levels.Testes 2, smooth, in tandem, located at hindbody.Seminal vesicle large, saccular antero-dorsal to ventral sucker, larger than sinus-sac.Sinus-sac small elongate-oval, contains short muscular hermaphroditic duct.Pars prostatica vesicular, small, difficult to distinguish, enters sinus-sac at its base.Genital pore at level of anterior septum.Ovary transverse, oval, in anterior end of last pseudosegment, ventral to caeca, smooth to irregular, contiguous with vitellarium.Mehlis' gland and uterine seminal receptacle present.Laurer's canal not observed.Vitellarium compact, smooth to irregular, elongate-oval, larger than ovary.Uterus thin-walled, extends posterior to vitellarium.Metraterm short.Eggs numerous.Excretory pore wide, terminal or subterminal; vesicle Y-shaped.

Class
Etymology.The genus Parasaturnius n. gen.refers to its resemblance with Saturnius Manter, 1969, and uses the Greek prefix para (meaning resemble).

Remarks
Parasaturnius n. gen.can be differentiated from Robinia and Bunocotyle by the presence of pseudosegments (transverse fibrous septa) in the fore-and hindbody.Furthermore, the genus Robinia possesses fine striations, symmetrical testes, 11-15 muscular lobes on the oral sucker, and vestigial ecsome.The new genus differs further from Bunocotyle by the lack of cyclocoel.Parasaturnius n. gen. is morphologically very similar to Saturnius since both genera present pseudosegments.Nonetheless, the new genus can be distinguished by the presence of four pairs of small papillae surrounding the aperture of the oral sucker.
Parasaturnius maurepasi (Overstreet, 1977) (Fig 1d).Two flanges, one located at oral sucker level, and the second at or slightly posterior to ventral sucker, mound-shaped.Oral sucker aperture surrounded by 8 small papillae arranged in four pairs (Fig 1d).Ventral sucker located between the second and third pseudosegment.Forebody 25.4−35.8% of body length.Large single cells in most segments, lateral to body.Pharynx contiguous; prepharynx not observed.Oesophagus thick walled, smaller than pharynx.Caeca with constrictions in pseudosegments, terminating near posterior end, "Drüsenmagen" just posterior to oesophagus.
Testes 2, smooth or irregular, in the last two segments, in tandem, between the caeca.Sinus sac small containing eversible hermaphroditic duct; metraterm short.Seminal vesicle large, elongate, longer than sinus sac, extending to anterior border of ventral sucker.Pars prostatica vesicular, small, difficult to observe, enters sinus-sac as its base.Genital pore median at or near level of anterior septum.
Ovary transversely elongate, pos-testicular, in last major segment, ventral, smooth to irregular, in middle of caeca.Mehlis' gland conspicuous in some specimens.Vitellarium smooth, 7.9−12.6 of body length, contiguous with ovary.Uterus filling most of partitions, extending posterior to vitellarium but not reaching posterior end.Eggs numerous, small.Excretory vesicle consisting of posterior muscular sac within the last pseudosegment.Pore terminal.

Remarks
The specimens found in Mugil curema from La Carbonera coastal lagoon were identified as Parasaturnius maurepasi n. gen.n. comb.based on morphology, host and geographical distribution as in the original description (Overstreet, 1977) (Table 2).In addition, following the key to species proposed by Blasco-Costa et al. (2008), our specimens fall in S. maurepasi since they possess seven pseudosegments separated by six transverse fibrous septa.However, we found some slight morphological differences with respect to the original description, as the papillae surrounding the oral sucker, since we were able to obtain SEM photomicrographs of the sampled specimens.Moreover, the metrical data suggest specimens found in the present study in Yucatán are overall slightly smaller than the specimens described by Overstreet (1977).

Molecular data and phylogenetic analysis
The 28S data set included 35 sequences and comprised 1, 326 nucleotides.The alignment (trimmed to the shortest sequence) included one sequence of Azygiidae used as outgroup.The phylogenetic analyses inferred with ML and BI recovered similar topologies (Fig. 2).The analyses show that the superfamily Hemiuroidea is monophyletic.Within each family, Hemiuridae, Isoparorchiidae, and Derogenidae were recovered as monophyletic, with strong nodal support (1/100; 1/97; 0.99/94).The remaining three families analyzed were not recovered as monophyletic, e. g., Gonocercidae, Lecithasteridae and Bunocotylidae.Bunocotylids were separated in two independent clades, one formed by three subfamilies, Hysterolecithinae, Quadrifoliovariinae and Opisthadeninae, with low support (0.54/48).The second major clade was formed only by Bunocotylinae, with strong support (1/100).Within Bunocotylinae, the two newly generated sequences of Parasaturnius maurepasi n. gen.n. comb.were nested as the sister group of the three genera contained in the subfamily, e.g., Robinia, Bunocotyle, and Saturnius, with strong support (1/100) (Fig. 2).

Discussion
The two newly sequenced specimens of Parasaturnius maurepasi n. gen.n. comb.were recovered as the sister group of the three previously mentioned genera, with the same topology from previous studies, e.g., Robinia as sister of Saturnius and Bunocotyle (Pankov et al., 2006;Marzoug et al., 2014).Interestingly, even though Saturnius and Parasaturnius n. gen.are morphologically very similar, molecularly they are not closely related and exhibit large genetic divergence of the gene 28S.Saturnius gibsoni (KJ010542) and S. minutus (DQ354366) are sister taxa to Bunocotyle progenetica (DQ354365), with less than 3% of genetic divergence, which is quite low if we consider that divergence between Robinia aurata and the three species previously mentioned varies from 6.6 to 6.8% (Table 3).Furthermore, the genetic divergence between Parasaturnius maurepasi n. gen.n. comb.and the four species was even higher, and ranged from 10.7-12%, which suggest that presence of pseudosegments in Saturnius and the new genus are homoplasies in these two genera.Interestingly, these high genetic divergence values between genera are like those reported among other genera of hemiuroid trematodes (Pantoja & Kudlai, 2022).
The original description of P. maurepasi n. gen.n. comb.was published by Overstreet (1977) (as Saturnius maurepasi) from the stomach of Mugil cephalus in Ocean Springs, Mississippi, USA.After the original record, the species has been reported across the Atlantic coast of the Americas, as a parasite of M. liza in Colombia and Brazil (Romero & Galeano, 1981;Fernandes & Goulart, 1992;Knoff et al., 1997).The species is reported for the first time in Mexico, as a parasite of Mugil curema in La Carbonera coastal lagoon, representing new host and new locality records for this parasite.In the same  Overstreet (1977) and Fischtal (1977), did not know about each other´s paper describing a new species of Saturnius published in the same year.Furthermore, S. belizensis was described from only two specimens.In the last taxonomic review of the  2008) studied a paratype of this species and noticed that S. belizensis exhibits five pseudogements.Even though our specimens are metrically more similar to S. belizensis, the number of pseudosegments in our specimens are seven, similar to that reported by Blasco-Costa et al. (2008) for S. maurepasi.Testing the validity of S. belizensis will require the generation of 28S DNA sequences from the type locality, and a through morphological study of a larger number of specimens to corroborate the number of pseudosegments.In addition to that, the schematic drawings of S. maurepasi from M. liza in Brazil shows they possess six pseudosegments; based on host association and geographical distribution, these specimens could indeed represent a separate lineage.Therefore, a detailed morphological study and sequence data are required to corroborate the status of these reports (Fernandes & Goulart, 1992).
Based on the evidence discussed above, Parasaturnius maurepasi n. gen.n. comb. is here recognized based on molecular data and a detailed morphological analysis.Still, sampling specimens of other valid species is required to be analyzed under a molecular approach, to clarify the species and family status.It is necessary to study further Saturnius segmentatus, the type species, which is the only species allocated in the genus Saturnius exhibiting morphological traits similar to those of Robinia, such as the muscular and conspicuous papillae on the oral sucker.In addition, the cyclocoel is a morphological trait of Bunocotyle, however this character is also found in four species of Saturnius, i.e., S. gibsoni, S. minutus, S. dimitrovi and S. overstreeti.In this sense, it would be also necessary to obtain sequence data of Saturnius spp.lacking cyclocoel and Bunocotyle spp. to test the evolution of morphological traits in Bunocotylinae.
Comments on the taxonomic status of Bunocotylidae.
Using molecular data, Sokolov et al. (2018) and Atopkin et al. (2020) resurrected the family Bunocotylidae, and considered it contained the subfamilies Bunocotylinae Opisthadeninae, Hysterolecithinae and Quadrifoliovariinae.However, two recent studies provided evidence in favor of the paraphyly of Bunocotylidae (Louvard et al., 2022;Faltýnkova et al., 2022), although these studies did not include all the genera considered by Atopkin et al. (2020) as members of the family.In our study, after including all the genera of bunocotylids following the concept by Atopkin et al. (2020) we found no conclusive evidence on the interrelationships among members of the families Bunocotylidae, Lecithasteridae and Hemiuridae.The phylogenetic tree yielded a basal polytomy and it might be premature to consider that the genera Machidatrema León-Règagnon, 1998, Hysterolecithoides Yamaguti, 1934, Bilacinia Manter, 1969, Quadrifoliovarium Yamaguti, 1965, Opisthadena Linton 1910, and Genolinea Manter, 1925 (allocated in Hysterolecithinae, Opisthadeninae, Quadrifoliovariinae) belong to Bunocotylidae.Therefore, more sequence data is required to resolve the systematic interrelationships among this hemiuroid group.

Figure 2 .
Figure 2. Consensus Bayesian Inference and Maximum Likelihood trees of genera of Hemiuroids inferred with 28S rRNA sequence data; numbers near internal nodes show posterior probabilities (BI) and ML bootstrap clade frequencies.Opis=

Table 1
Sequences of 28S from GenBank used for phylogenetic analysis in the present study

Table 2
Comparative metrical data for Parasaturnius maurepasi from America in different hosts plus S. belizensis Vol:. (1234567890)

Table 2
(continued) =Estimated from the published drawing.Mean value in parenthesis.BL body length, NPSG, Number of pseudosegments, BWat-VSF Body width at ventral sucker flange, BWatVS Body width at ventral sucker, FO forebody, LPL Last pseudosegment length, LPW Last pseudosegment width, OSL oral sucker length, OSW oral sucker width, VSL ventral sucker length, VSW ventral sucker width, FatVSL Flange at ventral sucker length, FatVSW Flange at ventral sucker width, PHL pharynx length, PHW pharynx width, OEL oesophagus length, SSL Sinus sac length, SSW Sinus sac length width, SVL seminal vesicle length, SVW seminal vesicle width, PPL pars prostatica length, PPW pars prostatica width, ATL anterior testis length, ATW anterior testis width, PTL posterior testis length, PTW posterior testis width, OL ovary length, OW ovary width, VL Vitellarium length, VW Vitellarium width, PVS postvitelline space, EL eggs length, EW eggs width, BW/BL (%) body width as a percentage of body length, Sucker length ratio SLR (vsl/osl) ventral sucker length as a proportion of oral sucker length, Sucker width ratio SWR (vsw/osw) ventral sucker width as a proportion of oral sucker width, VSW/BWVS (%) ventral sucker width as a percentage of body width at the level of the anterior margin of the ventral sucker, FO/BL (%) forebody as a percentage of body length, VL/BL (%) vitellarium length as a percentage of body length, MFL/VSL (%) muscular flange length as a percentage of ventral sucker length, MFW/VSW (%) muscular flange width as a percentage of the width of the ventral sucker, MFW/BWVS (%) muscular flange width as a percentage of body width at the anterior border of the ventral sucker, MFW/MFL (%) muscular flange width as a percentage of length, LPL/BL (%) last pseudosegment length as a percentage of body length, LPW/BW (%) last pseudosegment width as a percentage of body width, LPW/LPL (%) last pseudosegment width as a percentage of its length Fischtal (1977)ion, a morphologically similar species has been reported, i.e., S. belizensis.Fischtal (1977)described S. belizensis as a parasite of M. curema from Belize, a locality relatively close to La Carbonera, Yucatán, Mexico.It is noteworthy that most probably both,

Table 3
Pairwise nucleotide sequence comparisons between taxa of Bunocotylidae for the aligned 28S rDNA sequences(1,326 nt)In bold is represented the genetic intraspecific divergence Vol.: (0123456789)