Revision of the systematics of the Polystomoidinae (Platyhelminthes, Monogenea, Polystomatidae) with redefinition of Polystomoides Ward, 1917 and Uteropolystomoides Tinsley, 2017

Polystomatids are platyhelminth parasites that infect mainly amphibians and freshwater turtles. For more than seven decades, chelonian polystomes were classified into three genera according to the number of hamuli, i.e. absent for Neopolystoma, one pair for Polystomoidella and two pairs for Polystomoides. Following re-examination of morphological characters, seven new genera were erected the past six years, namely Apaloneotrema, Aussietrema, Fornixtrema, Manotrema, Pleurodirotrema, Uropolystomoides and Uteropolystomoides. However, the polyphyly of Neopolystoma and Polystomoides on the one hand, and the nested position of Uteropolystomoides within a clade encompassing all Neopolystoma and Polystomoides spp. on the other, still raised questions about the validity of these genera. We therefore re-examined several types, paratypes and voucher specimens, and investigated the molecular phylogeny of polystomes sampled from the oral cavity of North American turtles to re-evaluate their systematic status. We show that all Polystomoides Ward, 1917, sensu Du Preez et al., 2022, Neopolystoma Price, 1939, sensu Du Preez et al., 2022 and Uteropolystomoides Tinsley, 2017 species, display vaginae that are peripheral and extend well beyond the intestine. We thus reassign all species of the clade to Polystomoides and propose nine new combinations; however, although Uteropolystomoides is nested within this clade, based on its unique morphological features, we propose to keep it as a valid taxon. Polystomoides as redefined herein groups all polystome species infecting either the oral cavity or the urinary bladder of cryptodires, with peripheral vaginae and with or without two pairs of small hamuli. Uteropolystomoides nelsoni (Du Preez & Van Rooyen 2015), originally described from Pseudemys nelsoni Carr is now regarded as Uteropolystomoides multifalx (Stunkard, 1924) n. comb. infecting three distinct Pseudemys species of North America.


Introduction
The Neodermata, a clade comprising only parasitic platyhelminths, contains three well-defined groups of flatworms, the Digenea, the Cestoda and the Monogenea. While the monophyly of the Monogenea is still being debated [23,27,31,34], the monophyly of the two subclasses Polyonchoinea and Heteronchoinea has been widely accepted [3-5, 24-26, 29, 31]. Monogeneans of both subclasses are mainly ectoparasites of gills and skin of Chondrichthyes and Actinopterygii fishes, which may account for more than 25,000 species [9,47]. However, fewer than 250 monogenean species deviated from the norm as they are parasites of semi-aquatic tetrapods, mainly amphibians and chelonians. They are classified into three families of the Polyonchoinea, namely the Gyrodactylidae, the Lagarocotylidae and the Iagotrematidae, and into a single family of the Heteronchoinea, the Polystomatidae sensu Sinnappah et al. [37]. The Polystomatidae comprises just more than 200 species, infecting anurans, salamanders and caecilians of the Amphibia; freshwater turtles of the Testudines; the common hippopotamus, Hippopotamus amphibius Linnaeus of the Mammalia; but also a fish, i.e. the Australian lungfish, Neoceratodus forsteri Krefft of the Dipnoi. Polystome species are classified into 32 genera, of which 20 occur specifically within amphibian hosts, 10 are recognized in chelonians, and one each are reported from the common hippopotamus and Australian lungfish, respectively.
Polystomes of frogs and chelonians were first described as Polystoma Zeder, 1800, and more than a century later a new subgenus Polystomoides Ward, 1917 was created to account for chelonian polystomes. Polystomoides, being found in the mouth, esophagus, nasal cavities or urinary bladder of its host, was described as having a haptor with two pairs of large hooks, the outer pair being larger than the inner one, a single testis, a short uterus containing usually a single egg and vitellaria extending into the posterior part of the body. Vaginae and eyes are absent in adults. Polystomoides was raised later to genus rank by Ozaki [32] who pointed out the absence of a uterus. Besides Polystomoides, Price [36] created two new genera for chelonian polystomes, namely Polystomoidella Price, 1939 being found in the urinary bladder of its host and differing from Polystomoides by having a single pair of large haptoral hooks, and Neopolystoma Price, 1939, being found in the urinary bladder and nostrils of its host and differing from Polystomoides and Polystomoidella by the absence of large haptoral hooks. Strelkov [38] first reported Neopolystoma from conjunctival sacs of turtles. Tinsley and Tinsley [43], based on phylogenetic studies by Héritier et al. [11], created a new genus Uropolysto-moides Tinsley & Tinsley 2016 to account for all Polystomoides species occurring in the urinary bladder of their African, Asian, and Australian hosts. Uropolystomoides spp. differ from Polystomoides spp. of the oral cavity by the size of hamulus 1, being always bigger than the sucker diameter, which was originally mentioned in Knoepffler and Combes [18]. Tinsley [42] [43] by the presence of a uterus containing several eggs but also by a massive genital bulb encompassing a great number of genital spines. Du Preez and Verneau [8], based on the most comprehensive phylogeny of chelonian polystomes, created three new genera to account for all polystomes of the conjunctival sacs, namely Aussietrema Du Preez & Verneau 2020, Fornixtrema Du Preez & Verneau 2020, and Apaloneotrema Du Preez & Verneau 2020. Aussietrema is mainly characterized by a spherical ovary and egg, Fornixtrema by a separate egg-cell-maturation-chamber and fusiform to diamond-shaped egg with acute tips, and Apaloneotrema by a large fusiform egg with rounded tips. Finally, Du Preez et al. [7], following a revision of South American and Australian polystomes infecting specifically turtles of the Pleurodira suborder, described two new genera that are both restricted to South America and Australia, respectively. Though these two genera share vaginae that are latero-ventral and positioned in line with the anterior margin of testis, Manotrema Du Preez, Domingues & Verneau 2022 of South American pleurodires differs from Pleurodirotrema Du Preez, Domingues & Verneau 2022 of Australian pleurodires by the presence of two pairs of small hamuli with very deep cuts between handle and guard and a haptor with deep incisions between suckers.
While Bayesian trees inferred from phylogenetic analyses of the four concatenated genes 12S, 18S, 28S and COI [8,11] indicate that the two genera, i.e. Polystomoides sensu Du Preez et al. [7] and Neopolystoma sensu Du Preez et al. [7], are each polyphyletic, all Polystomoides and Neopolystoma species fall into a robust lineage, including U. nelsoni of Pseudemys nelsoni Carr. Therefore, one may question the possibility of finding specific morphological characters for this clade. In this paper, we studied polystome samples collected from North American chelonians, type and paratype slides borrowed from the Parasitic Worm Collection, National Museum, Bloemfontein, South Africa, and voucher slides stored in the private collection of the second author (LdP) to revise the classification of these two genera. We also investigated the molecular phylogeny of polystomes sampled from the oral cavity of North American turtles, including specimens of Polystomoides multifalx (Stunkard, 1924) collected from Pseudemys floridana (Le Conte) and Pseudemys concinna (Le Conte) of Florida, in order to determine the validity of the genus Uteropolystomoides.

Ethics
Ethical clearance for this study was obtained from the North-West University Animal Care ethics committee (Ethical clearance no. NWU-00256-17A5).

Turtle sampling and polystome collection
The fieldwork procedures used to collect freshwater turtles were detailed in Du Preez and Verneau [8]. To summarize, turtles were captured in a number of water bodies in North Carolina and Florida, USA using baited traps that were left overnight (Table 1). Captured animals were kept individually in plastic containers at room temperature for two to three days and screened on a daily basis for the presence of polystome eggs following the procedure detailed in Verneau et al. [45]. Polystome eggs collected were preserved in ethanol 75% for further molecular analyses. Depending on the intensity of infection, based on the number of eggs released per host individual, a few animals were euthanized with a lethal injection of a concentrated buffered MS222 (ethyl-4-aminobenzoate) solution. They were then dissected and polystomes were retrieved from the urinary bladder, oral cavity, and/or conjunctival sacs. Polystomes were removed according to the procedure reported in Du Preez and Verneau [8].

Collection of polystomes using a non-lethal method
Because killing of animals collected from the Ichetucknee River in Ichetucknee Springs State Park of Florida was not allowed, specimens of P. concinna that released polystome eggs were examined by swabbing the mouth and pharyngeal pouches. The turtle was held with the head facing upwards and the mouth held open with a small hook made from wire (Fig. 1A). A dry 120 mm wooden stem cotton swab was gently lowered down the mouth into the pharyngeal region while slowly rotating the swab. The technique was successful, and three parasites were retrieved from three distinct specimens (Fig. 1B, Table 1) with no adverse effect on the hosts. Parasites were heat killed and stored for further analysis. Some were fixed slightly flattened under coverslip pressure, while others were fixed directly either in 10% neutral buffered formalin for permanent mounts, in Bouin's fixative [15] for histology or in molecular grade 70% ethanol for genetics.

Morphological analyses
In 2004, LdP visited the United States National Parasite Collection in Beltsville, Maryland for a research visit and studied the entire polystome collection. A loan of voucher and paratype specimens was approved and specimens were studied and photographed in South Africa. Type and paratype slides borrowed from the Parasitic Worm Collection, National Museum, Bloemfontein, South Africa were also examined as well as voucher slides stored in the second author's collections ( Table 2). All slides were of whole-mounted stained specimens. While the main focus was on species belonging to Neopolystoma, Polystomoides, and Uteropolystomoides, representatives of Fornixtrema and Uropolystomoides were also examined. Polystomes infecting P. nelsoni and P. concinna were morphologically examined, measured, and photographed using a Nikon AZ100M microscope (Nikon, Netherlands) fitted with 0.5X, 1X and 4X objectives as well as a Nikon U3 digital camera. Measurements were captured using the Nikon NIS software. Small features were examined, measured, and photographed using a Zeiss Imager Axio10 compound microscope (Zeiss, Germany) fitted with a Zeiss Axio cam 305 camera (Zeiss, Germany) and Zeiss Zen Blue elements (Zeiss, Germany) software. Measurements were based on ten specimens each from P. nelsoni and P. concinna, all collected near Gainesville, Alachua County, FL, USA. Morphological examination focussed on body size, relative size of the haptor, genital bulb diameter, number of genital spines, position of the vaginae in relation to body width and length, position of ovary, position of testis, presence and size of hamuli and haptoral sucker diameter.

Molecular experiments
DNA extractions were performed with 150 μL of Chelex 10% and Proteinase K 1 mg/mL, following the protocol reported in Héritier et al. [11], from several eggs and worms collected from distinct host species and areas of North Carolina and Florida (Table 3). For the PCR, we followed the amplification procedure of Héritier et al. [11] for the two genes of interest COI and 28S. COI was amplified in one round, either with primers L-CO1p/H-Cox1p2 or L-CO1p/H-Cox1R whose sequences are reported in Littlewood et al. [22] and Héritier et al. [11]. The partial 28SrRNA gene was, however, amplified in two rounds with the combination of primers LSU5′/IR16 and IF15/LSU3′ whose sequences are reported in Verneau et al. [44] and Héritier et al. [11]. The procedure we followed for gene amplification was identical regardless of the combination of primers and gene of interest: one initial step of 5 0 at 95°C for long denaturation; 30 cycles of 1 0 at 95°C for denaturation, 1′ at 48°C for annealing and 1′ at 72°C for elongation; one final step of 10 0 at 72°C for terminal elongation. PCR reactions were run twice and independently in a final volume of 25 µL comprising Buffer 1x, MgCl 2 1.5 mM, dNTPs 0.2 mM, primers 0.4 mM, GoTaq Polymerase 0.75 unit (Promega, France) and DNA (2 µL). PCR products were then pooled and sent to Gen-oScreen (Lille, France) for purification and sequencing with their respective forward and reverse PCR primers. Finally, we used Geneious software (Saint Joseph, MO, USA) to check chromatograms, and to read and edit resulting sequences. New sequences were deposited in GenBank with accession numbers OP784895, OP793140 to OP793161 and OP793434 to OP793461 for COI, and OP795734 to OP795746 and OP795805 to OP795807 for 28S. Phylogenetic and distance analyses within polystomes of the pharyngeal cavity New COI and 28S sequences, after primer trimming, were first aligned independently using Clustal W implemented in MEGA version 7 [19] under default parameters [41]. Only those characterizing polystomes of the oral cavity were kept at this stage. All these sequences were subsequently aligned with other COI and 28S sequences of distinct polystomes species retrieved from GenBank (Table 4). These sequences characterized polystomes of the oral cavity with the exception of Fornixtrema palpebrae (Strekov, 1950) of the conjunctival sacs and Polystomoidella whartoni Price, 1939 and Uropolystomoides malayi (Rohde, 1963) of the urinary bladder, that were used for outgroup comparisons after Du Preez and Verneau [8].
In the final COI and 28S alignments, when identical sequences were found from the sequencing of eggs and/or worm, a single sequence was kept for each distinct haplotype.
The COI phylogenetic analysis was conducted on a data set comprising 64 haplotypes and 396 characters which was considered a single partition. A GTR + I + G model was selected following the Akaike Information Criterion (AIC) implemented in Modeltest 3.06 [35]. Six types of substitutions and invariablegamma rates with four gamma rate categories were therefore applied. On the contrary, the 28S phylogenetic analysis was conducted on a data set comprising 15 haplotypes and 1,370 characters also considered as a single partition. A GTR + G model was selected following the AIC, with six types of substitutions and gamma rates with four gamma rate categories. The Bayesian analyses were run using MrBayes 3.04b [14], with four chains running for one million generations and sampled every 100 cycles. The Bayesian consensus trees were drawn after removing the first 1000 trees (10%) as the burn-in phase and viewed with TreeView version 1.6 [33].
Corrected pairwise distances were calculated for COI sequences using the Kimura 2-parameter model, while the total number of differences was estimated for partial 28S in MEGA version 7 [17]. Species delimitation was discussed in the light of the COI threshold defined for polystomes [12].
After examination of newly collected specimens, as well as types and paratypes of Neopolystoma, Polystomoides, and Uteropolystomoides spp. borrowed from museum collections, no obvious morphological character was evidenced supporting the clustering of these three genera into a clade with the exception of the vaginae that are peripheral (Fig. 2). Following a thorough study of all the drawings published in the literature for chelonian polystomes (see Morrison and Du Preez [30] for a review), this character is found in all species of the genera. It also characterizes all species of Fornixtrema and some polystome species of Uropolystomoides infecting specifically cryptodire turtles.

Systematics of Uteropolystomoides, a monotypic genus infecting Pseudemys spp.
Measurements obtained from the 10 polystomes collected from Pseudemys nelsoni ( Table 5, column 1) and the 10 collected from P. concinna (Table 5, column 2), showed an overlap indicating that all specimens belong to a single species. We therefore combined the measurements from the two polystome samples into a single set of data with their range, mean, and standard deviation (Table 5, column 3).
In the molecular study, we obtained 55 COI sequences including 16 new haplotypes (H145 to H160) and 13 28S sequences including two new haplotypes (Hnuc36 and Hnuc37). The resulting Bayesian consensus trees for COI and 28S are depicted in Figures 3 and 4, respectively. The COI tree shows 12 well-resolved lineages that each likely reflect a distinct parasite species. All COI haplotypes characterizing polystomes of Pseudemys spp. cluster in a single clade being strongly supported by Bayesian posterior probabilities. The 28S tree also shows 12 well-differentiated species, including U. nelsoni (Hnuc20) which shares the same haplotype with polystomes collected from P. concinna and P. floridana (see Table 3).
The Kimura-2 parameter distances for COI vary from 0.003 to 0.016 within polystomes collected from Pseudemys nelsoni (H43), P. concinna (H145 to H148), and P. floridana (H145, H147). The distance, however, varies from 0.110 to 0.180 between these parasites and their closest relatives. Additionally, a single 28S haplotype (Hnuc20) was reported for all polystomes collected from Pseudemys spp. That haplotype has seven mutations that differ from Hnuc6, Hnuc7, and Hnuc 21, which characterize P. oris Paul, 1938, P. soredensis Héritier, Verneau, Smith, Coetzer & Du Preez, 2018, and P. scriptanus Héritier, Verneau, Smith, Coetzer & Du Preez, 2018, respectively. On the contrary, two differences were observed in the 28S between P. scriptanus and Polystomoides sp2 of Trachemys scripta (Thunberg), between P. oris and P. soredensis and between P. ocellatum (Rudolphi, 1819) and Polystomoides sp1 of Emys orbicularis (Linnaeus). According to the threshold designed by Héritier et al. [12] within chelonian polystomes, that was set to 3.4% of COI genetic divergence, and to the high degree      of 28S divergence between Hnuc20 and Hnuc19 (14 mutations), which characterizes the sister species of U. nelsoni, we suggest that all specimens collected from Pseudemys spp. belong to the same polystome species. This conclusion is strengthened by the existence of the same 28S haplotype within those polystomes.

Systematics revision of Polystomoides
All the Neopolystoma, Polystomoides, and Uteropolystomoides spp. show similar morphology with vaginae that are peripheral and extend well beyond the intestine. Though this morphological characteristic is also found within Fornixtrema and some species of Uropolystomoides, Fornixtrema differs from these species by the shape of the egg and infection site, i.e. the conjunctival sacs, while Uropolystomoides differs by the shape of its first pair of hamuli. For these reasons, we propose the generic name Polystomoides for the entire clade after excluding Uteropolystomoides (see below). According to the principle of priority in the International Code of Zoological Nomenclature, article 23 [16], Polystomoides has priority over Neopolystoma. As a result, we reassign nine species, previously attributed to Neopolystoma, to Polystomoides, and propose the following new combinations, namely P. aspidonectis (MacCallum, 1918) n. comb., P. cayensis (Du Preez, Badets, Héritier & Verneau, 2017) n. comb., P. cyclovitellum (Caballero, Zerecero & Grocott, 1956) n. comb., P. domitilae (Caballero, 1938) n. comb., P. euzeti (Combes & Ktari, 1976) n. comb., P. exhamatum (Ozaki, 1935) n. comb., P. orbiculare (Stunkard, 1916) n. comb., P. rugosa (MacCallum, 1918) n. comb., and P. terrapenis (Harwood, 1932) n. comb. It did not escape our attention that the type-species of Neopolystoma, Neopolystoma orbiculare (Stunkard, 1916), was nested in the clade (see Du Preez and Verneau [8]) but not the type-species of Polystomoides, i.e. Polystomoides coronatum (Leidy, 1888). Unfortunately, we could not sample the latter species because the identity of its type-host was fueled by ambiguity (see below). Nevertheless, in our estimation, based on the information available at present, P. coronatum should be attributed to this clade.
Polystomoides was originally created as a subgenus of Polystoma Zeder, 1800 by Ward [46] who designated Polystoma coronatum Leidy, 1888 as the type species. Ward (1917) based his subgenus chiefly on the presence of "a short uterus containing only a single egg". Subsequently, Polystomoides was raised to the genus rank by Ozaki [32]. From 1939 until recently, the generic circumscription of Polystomoides was altered several times, and several species of Polystomoides were transferred to Neopolystoma, Uropolystomoides, Uteropolystomoides, and Manotrema on the basis of one character or a combination of characters [7,36,42,43]. The type-species of Polystomoides, P. coronatum, was originally described by Leidy (1888) from a North American host turtle whose identity, "a common food terrapin", was vague. Leidy [21] described it poorly and did not include any figures. Polystomoides coronatum was redescribed thoroughly and figured by Stunkard [39] from its type-specimen (No. USNM 1315426) and allegedly collected (quoting Stunkard) from Emys palustris Leidy, 1887 (now Trachemys terrapen (Bonnaterre, 1789)) and Emys rugosa Duméril & Bibron, 1835 (now Trachemys decussata (Gray, 1831)) (Stunkard, 1917). The genus Polystomoides, as redefined herein, groups only polystomes infecting either the oral cavity or the urinary bladder of cryptodires, with or without two pairs of small hamuli and some peripheral vaginae.

Uteropolystomoides, a valid taxon?
Uteropolystomoides, as its generic name indicates, is characterized by the possession of a uterus containing a few eggs (up to 12 eggs in the present study). This feature was not found in Polystomoides or any other chelonian polystomes which possess an oötype where a single egg is often retained. The uterus is sacciform and pre-ovarian. Based on the phylogenetic relationship of polystomes infecting anurans, it was shown that Polystoma, the most widespread polystome genus, could represent a polyphyletic group, including a subgroup of species infecting specifically Asian frogs of India, China and Japan [1,44]. By investigating the morphology of these species more in depth, Chaabane et al. [6] found some specific characters of these taxa that were used for describing a new genus, i.e. Indopolystoma, Chaabane, Verneau & Du Preez 2019 within the Polystomatidae. On the contrary, given the phylogenetic position of Metapolystoma which is nested within Polystoma, Bentz et al. [2] considered that Metapolystoma might be not valid. However, based on the morphology and life cycle of the monophyletic Metapolystoma, Landman et al. [20] concluded that this genus should be kept as a valid taxon within the Polystomatidae. Although we follow a cladistic approach in general to name groups and although Uteropolystomoides is nested in the Polystomoides clade, we propose to retain Uteropolystomoides as a valid genus based on its unique morphological characteristics.

Revision of Uteropolystomoides outlines
Polystomoides multifalx, originally described as Polystoma multifalx Stunkard, 1924 from the pharyngeal region of Pseudemys floridana from central Florida (USA), was the first    (Stunkard, 1924) that were, for some of them, collected from Pseudemys concinna (Le Conte), for the others, from P. floridana (Le Conte) (see Table 3 for more details).

Uropolystomoides malayi_Hnuc32
Polystomoidella  . Bayesian tree inferred from the analysis of 28S sequences. Numbers at nodes indicate Bayesian Posterior Probabilities (BPP). Only BPP 0.95 are indicated. Scale bar reflects expected changes per site. * designates Hnuc20 haplotype that also characterizes specimens of Polystomoides multifalx (Stunkard, 1924) (see Table 3 for more details). chelonian polystome known to have a huge genital bulb bearing numerous long spines in excess of 100 (120-124) [40]. Stunkard [40] mentioned that the number of genital spines of this species was three times greater than in any other known polystomes at the time. Based on samples from the mouth of Pseudemys hieroglyphica Boulenger (now Pseudemys concinna) from Oklahoma (USA), Harwood [10] distinguished Polystomoides stunkardi Harwood, 1931 from P. multifalx by the fewer genital spines, the smaller size of the genital bulb and testis, and the arrangement of haptoral suckers. From a morphological comparison between a set of specimens collected by Mr. Macintosh from P. floridana from southern Florida and vouchers of P. stunkardi from P. concinna from Oklahoma, Price [36] proposed the conspecificity of P. stunkardi with P. multifalx. However, Tinsley [42] concluded that U. nelsoni, P. multifalx, and P. stunkardi may form a coherent group of apparently related species. Based on morphological observations and measurements of samples  (Stunkard, 1924 collected from P. concinna and P. nelsoni (Table 5), we were unable to distinguish polystomes collected from both host species. Moreover, the genetic data indicated that polystome samples collected from the three distinct host species, namely P. concinna, P. floridana, and P. nelsoni, belong to the same polystome species. We therefore agree with Price [36], and consider that the specimens collected from P. concinna from the Ichetucknee River of Florida and those collected from P. nelsoni are conspecific with P. multifalx. We thus propose to consider a single species, namely Uteropolystomoides multifalx (Stunkard, 1924)   Mehlis' glands large, surrounding the base of the oötype. Uterus, spherical sac like, containing up to 12 ovoid, operculate eggs. Of the 20 specimens, five had no eggs, four had 1, one had 2, two had 3, two had 4, two had 6, one had 7, one had 8 and two had 12. Eggs 137-269 (232) long, 137-193 (169) wide. No intrauterine development. Two lateral vaginae at the level of the ovary very prominent and big, 353-860 (565) long, bearing multiple marginal openings formed by branching vaginal canal. Vitellaria extended throughout most of body, except the ovary, uterus and genital bulb, and not entering the haptor. Stretching in between haptoral suckers, surrounding the female reproductive organs. Genito-intestinal canal, posterior to ovary. Testis 342-892 (545) long, 425-778 (632) wide, spherical, dense equatorial to post-equatorial. Vas deferens widens anteriorly to form the semen vesicle, narrowing towards genital bulb, opening in common genital opening. Genital pore opening ventral, directly posterior to intestinal ceca bifurcation, situated 18-24% (21%) of total length from most anterior point, genital bulb muscular, very big 438-847 (650) in diameter, surrounded by glandular cells, armed with a genital crown with 118-136 (125) genital spines (Fig. 5B), 83-98 (93) long. Two pairs of small hamuli (Fig. 5C) between posterior-most haptoral suckers with deep cut between handle and guard, handle 105-175 (137) long; guard 86-167 (121) long; hook 59-86 (70) long. Marginal hooklets placed as for other polystomes: pairs one and two between hamuli, marginal hooklet pairs three to five embedded in suckers, pairs six to eight between anterior suckers. Marginal hooklet pairs one 25-30 (28) long and hooklet pairs two to eight 24-29 (27) long.

Conclusion
Following our investigations on morphological and molecular characters on the one hand, and based on the most updated phylogeny of polystomes infecting turtles on the other [8], we now consider nine genera within chelonian polystomes. According to the literature related to the taxonomy and systematics of polystomes, Apaloneotrema is a monotypic genus which infects the conjunctival sacs of cryptodire restricted to the Nearctic realm; Aussietrema comprises four species infecting the conjunctival sacs of pleurodires restricted to the Australian realm; Fornixtrema comprises seven species infecting the conjunctival sacs of cryptodires of the Indomalayan, Nearctic, Neotropical and Palearctic realms; Manotrema comprises three species infecting the urinary bladder of pleurodires restricted to the Neotropical realm; Pleurodirotrema comprises four species infecting the urinary bladder and the oral cavity of pleurodires restricted to the Australian realm; Polystomoidella comprises three species infecting the urinary bladder of cryptodires restricted to the Nearctic realm; Polystomoides comprises 29 species infecting the urinary bladder and the oral cavity of cryptodires distributed in the Nearctic, Neotropical and Palearctic realms; Uropolystomoides comprises 13 species infecting the urinary bladder of both pleurodires and cryptodires that are distributed in the Ethiopian and Australian realms, respectively on the one hand and in the Indomalayan realm on the other; Uteropolystomoides is a monotypic genus which infects the oral cavity of cryptodires restricted in the Nearctic realm. Regarding the distribution of polystome genera across chelonians and geographical areas, all genera with the exception of Uropolystomoides are restricted to a single group of turtles (pleurodires versus cryptodires), and usually found in a single or a few biogeographic realms. If future studies on the morphology of Uropolystomoides spp. split polystomes infecting pleurodires from those infecting cryptodires [7], it could demonstrate a correlation between historical biogeography of pleurodires and cryptodires and the diversification of polystomes. This deserves to be studied more in depth from a