Encotyllabe spari do not live in Brazil: Morphological and molecular evidence conrm two new species of Encotyllabe (Monogenea: Capsalidae) from Brazil

all species one closely suggesting host specicity, but other species, specically Encotyllabe 1934, have reported at from 19 belonging to nine families in two and and Concerning Brazilian of spari (but seven host species belonging to four and two have as for this The aim of this study

The status of those provisional species must be con rmed following redescriptions based on specimens from the type host and locality due to the expected host speci city of Capsalids [12] as well as the appreciable geographic variation observed in some species of monogeneans [13,14] The most reliable taxonomic criteria for mature specimens of Encotyllabe are, among others, body shape, relative size of several organs, shape and position of testes, distance between the center of the ovary and center of the testis, penis shape and length of pedicel when extended [3][4][5][6][7][8][9]. However, multivariate analysis of proportional measurements standardized by total length as well as molecular studies will strongly improve the taxonomy of Encotyllabe [3,15].
Since Encotyllabe spari have been reported from Brazilian sh belonging to seven species in four families and two orders (Perciformes and Scorpaeniformes), we searched for this species in P. pagrus and O. ruber, the most common reported hosts in Brazil, in order to con rm their identity. Fresh material allowed us to perform molecular as well as morphological and morphometric analyses. Our results indicated that specimens previously reported as E. spari in P. pagrus and O. ruber belong to two new species. These are described and differentiated below.

Sample collection and processing
During 2016 and aperiodically, 26 specimens of Pagrus pagrus and 9 specimens of Orthopristis ruber were obtained from local shermen off Cabo Frio, Rio de Janeiro, Brazil. In the laboratory, sh were dissected; gills and pharyngeal plates were removed and examined under a stereomicroscope for monogeneans. Thirty-four specimens of two species of Encotyllabe obtained from P. pagrus (n = 16) and O. ruber (n = 18) were selected for morphological and molecular analyses.
Some worms were xed in 4% neutral buffered formaldehyde and then transferred and stored in 70% ethanol for further morphological studies (light microscopy). Selected specimens from each of the two host species were transferred to 96% ethanol for DNA analysis.
Population descriptors, prevalence and mean intensity [27] were recorded.

Morphological and statistical analysis
Fixed specimens were stained with carmine, cleared with clove oil (Sigma-Aldrich, Taufkirchen, Germany) and mounted in Eukitts ® (O. Kindler GmBH, Freiburg, Germany). Drawings were made with the aid of an Olympus BX53 microscope (Olympus Corporation, Tokyo, Japan) equipped with a drawing tube. Measurements are given in micrometers unless otherwise stated with a range followed by the mean in parentheses. Type-material was submitted to the Helminthological Collection Instituto Oswaldo Cruz (CHIOC) -Rio de Janeiro -Brazil.
To comply with the regulations set out in article 8.5 (version 2012) of the International Code of Zoological Nomenclature (ICZN), details of the paper have been submitted to ZooBank. The LSID (Life Science Identi er) is

Morphometric analysis
Multivariate morphometric analyses were performed on 16 specimens from O. ruber and nine from P. pagrus. Principal component analysis (PCA) was used for proportional morphometric measurements [28]. Because of the expected correlation between the size of different organs and body length (BL, excluding peduncle), we used proportion rather than raw measurement [3]. The proportion measurements were as follows: (1) maximum body width/BL, (2) length of the peduncle/BL, (3) diameter of haptor/BL, (4) diameter of the anterior attachment organs/BL, (5) pharynx width/BL, (6) length of the ovary/BL, (7) ovary width/BL, (8) length of the testes/BL, (9) width of the testes/BL, (10) length of the large hamulus/BL, (11) length of the small hamulus/BL, and (12) length of the marginal hooks/BL. PCA was performed using the Statistic 7.0 software (Statsoft Inc., Tulsa, Oklahoma).
For each gene, a database was constructed in FASTA format (new generate sequences + sequences of Encotyllabe spp. from GenBank) and aligned with Clustal X [34]. Then, a visual inspection was performed with ProSeq v.2.91 [33] in order to edit the length of the nal data set.
BI analyses were executed with the following parameters: nst = 6, rates = invgamma according to the evolutionary model determined by jModeltest 0.1.10 for each gene; that is, TVM+ G for LSU rRNA and TIM1+I for cox1 and replaced by GTR+G+I for MrBayes. The analysis was performed for 5,000,000 generations. Analyses included two runs of four chains and sampling every 100 generations. Support for nodes in the BI tree topology was obtained by posterior probability burn-in of the initial 25% of samples. Visual inspection of log likelihood scores against generation time was performed in TRACER v.1.7 [38]. Statistical support for ML analyses was performed with 1,000 bootstraps. The trees were visualized and edited in FigTree v.1.4.4 [38]. The pairwise p-distances and numbers of nucleotide differences between Encotyllabe species were calculated using MEGA v6 [39] (Table 2). Two species of Neobenedenia (Capsalidae) were used as outgroups (   (Fig.1 B). Uterus extends anterolaterally along the posterior wall of the penis bulb. Ootype not observed, apparently hidden by Mehlis' gland, slender uterus opens at genital pore. Vaginal pore on ventral side at level of vitelline reservoir; ducts not observed. Vitelline reservoir preovarian on the left side. Vitelline follicles extensive laterally and median elds, from penis to base of peduncle. Eggs not observed.

Differential diagnosis
Encotyllabe yamagutii n. sp. resembles those species with testes side by side, located anterior to the midlevel of the body, and a peduncle larger than the haptor diameter, namely, E. spari and E. antofagastensis. The main differences between the new species and the abovementioned species are the shape of the penis with a projection near the genital pore in the new species and the body size, with E. yamagutii being smaller (1.94 mm) than E. spari (3.14 mm) and E. antofagastensis (2.43 mm). Moreover, the vitelline receptacle is de nitively preovarian in E. spari and E. antofagastensis but anterolateral in the new species. Molecular analysis of concatenated genes shows that E. antofagastensis and E. yamagutii are well discriminated species (Fig. 3). Unfortunately, molecular data for E. spari are not available in GenBank, The type host for E. spari is Sparus microcephalus (Sparidae), but it is also found in two non-related hosts in the Inner Sea (Japan): the haemulid Plectorhynchus pictus and serranid Epinephelus akaara. The type host for E. antofagastensis is the haemulid Anisotremus scapularis from the southeastern Paci c. The new species is also a parasite of a sparid (Pagrus pagrus) but from the coast of Brazil.
Examination of specimens from P. pagrus identi ed by several authors as E. spari (CHIOC 34531) has revealed that they in fact belong to the new species.  (Fig. 2B). Uterus extends anterolaterally along the posterior wall of the penis. Vaginal pore on ventral side of the vitelline reservoir; ducts not observed. Vitelline reservoir preovarian, sinistral. Vitelline follicles extensive laterally and in median elds, from the pharynx to the base of peduncle. Eggs pyramidal, with 4 long and twisted laments (Fig. 2C).

Differential diagnosis
Only two species of Encotyllabe have been described with testes smaller than the ovary, namely, E. embiotocae and E. caranxi. Of those, E. caranxi is the longest species described in the genus (11.26 mm); the type host is a carangid (Caranx lutescens) from the Great Barrier (Australia), whereas the type host for E. haemuli is a haemulid from the coast of Brazil. The relationship between large and small haptoral hooks in E. embiotocae varies between 6.8:1 on average, whereas this value reaches 10:1 in the new species, with the smaller hooks in the new species being proportionally smaller.
Specimens of Encotyllabe previously identi ed as E. spari from Anisotremus virginicus, Conodon nobilis and Haemulon sciurus deposited in CHIOC and identi ed as E. spari in fact belong to the new species. Specimens identi ed as E. spari [17] also belong to E. haemuli. Figure 3 presents the plot of specimens in the bidimensional axis of the PCA. The rst and second components explain 59.5% of the total variance. The rst component explains 37.3% of the variance and was mainly associated with the proportional morphometric measurements of the ovary width/BL, attachment organs average/BL, testes length/BL, testes width/BL, large hamulus length/BL, and marginal hooks length/BL. The second component explains 22.2% of the variance and was associated with haptor diameter/BL and body width/BL. The new species, E. yamagutii and E. haemuli, showed some overlap but were clearly differentiated.

Molecular and phylogenetic analyses.
For the LSU rRNA, ve sequences were obtained, three from E. haemuli (852 bp) and two from E. yamagutii (865 bp). The intraspeci c genetic variability for both species was 0%. Four sequences for cox1 were obtained from E. haemuli (437 bp long). Concatenated analysis based on ML and BI produced trees with similar topology (Fig. 4). Each of the new species was statistically supported in independent clades (ML>90; BI>0.9). Table 2 indicates pairwise genetic divergence for LSU rRNA and the cox1 gene.

Discussion
In general, monogeneans are among the most host-speci c parasites and may be the most host-speci c of all sh parasites [40]. When broad host speci city, i.e., many hosts for the same species, of Monogenea is studied under molecular scrutiny, that broad speci city is questioned. For instance, the "cosmopolitan" capsalid Neobenedenia melleni, recorded from more than 100 host species in ve orders, may be a complex of species, as suggested by molecular studies [12], and the related Benedenia seriolae, recorded as a parasite of natural populations of Seriola spp. in Japan, Australia, and Chile as well as in farmed conditions around the world, is also a species complex, as demonstrated by molecular evidence [41.42]. When the hosts for the herein recognized species of Encotyllabe were analyzed, an interesting picture became evident: ve species (E. anisotremi, E. carangis, E. cheilodactyli, E. lutjani, and E. xiamenensis) have been recorded from one host species [3,5,44], and two species have been recorded from two different but related host species. E. caranxi was found in three species of Caranx and one of the related Pseudocaranx (Carangidae) [5], and E. embiotocae was found to be a parasite of Cymatogaster aggregata and Amphistichus argenteus (Embiotocidae) [4]. Exceptions are E. caballeroi, reported as a parasite of three species of Lethrinus, Gymnocranius audleyi (Lethrinidae) and Scolopis monogramma and Scolopis sp. (Nemipteridae) from Australia, New Caledonia, the Philippines and Vietnam [5,45,46]; E. kuwaitensis, a parasite of Caranx sexfasciatus, Caranx sp. (Carangidae) and Plectorhinchus schotaf (Haemulidae) from the Arabian Gulf and the Mediterranean sea [6,47]; and E. souzalimae, which has been reported from two nonrelated hosts from Brazil: Trichiurus lepturus (Trichiuridae) and Thyrsitops lepidopoides (Gempylidae) [7,24). E. pagrosomi has been recorded from seven host species belonging to three families from the Galapagos Islands, Mexico (Paci c coast), Australia, Venezuela and Peru [5,48,49]. Finally, E. spari has been recorded from at least 25 host species in nine families and two orders in Japan, the Arabian Gulf, Brazil, Vietnam, Venezuela, Argentina and the Mediterranean sea [5-7, 15-26, 47, 49, 50, 51]. With regard to the Brazilian records, E. spari (but also E. cf. spari) has been recorded from at least four species of Haemulidae, Conodon nobilis, Anisotremus surinamensis, Haemulun sciurus and Orthropristis ruber; one species of Sparidae, Pagrus pagrus (Sparidae); two species of Sciaenidae (Menticirrhus americanus, Micropogonias furnieri); and one species of Dactylopteridae (Dactylopterus volitans) [15,16,20,22,25,26] (Table 3). Our data challenge the low host speci city of Encotyllabe from Brazil; worms from the Sparid Pagrus and those from the Haemulid Orthopristis ruber belong to different species, as demonstrated by molecular tools (Fig. 3). Surprisingly, the two species from Haemulidae (E. haemuli and E. antofagastensis) form a well-supported clade, presenting additional evidence of high host speci city.
The taxonomic status of members of Encotyllabe is not easy to clarify, as traditional taxonomy is based on body shape, the relative sizes of several organs, the shape and relative position of the testes, the penis shape, the extension of the vitellaria, and the size and shape of the anchors as well as the relative distances between different organs [3][4][5][6][7][8][9]. Additional characteristics, such as the relative position of the testes, have also been suggested, but multivariate analysis (i.e., principal component analysis) of proportions of different organs in relation to body size, rather the analysis or comparison of raw measurements, will be an adequate tool to discriminate species in this genus [3] as well as in the Monopisthocotylea Acanthocotyle [43], as suggested by our results (Fig. 4). Our results con rming the host speci city of Encotyllabe suggest that previous records of E. spari in hosts other than Pagrus pagrus and Orthopristis ruber could represent new species. This nding also applies to other species, particularly E. spari, E. pagrosomi and E caballeroi with nonrelated hosts and from different geographical localities.

Declarations
Ethics approval and consent to participate: This study was conducted under the protocol of the Ethical Commission of the Universidad de Antofagasta, Antofagasta, Chile.   Egorova (2000) [5].