Parasites of pufferfish, Lagocephalus spp. and Torquigener flavimaculosus of the Israeli Mediterranean: A new case of Lessepsian endoparasites

With the opening of the Suez Canal as a link between the Red Sea and the Mediterranean Sea in 1869, the biogeographical event of the Lessepsian migration has been starting. Aided by beneficial conditions in the new habitat, almost 500 marine species have immigrated and often established themselves in the Mediterranean Sea, including several pufferfish species, with all of them extending their range and becoming important components of the local fauna. The parasitic fauna of these pufferfish has scarcely been examined in the Mediterranean Sea or in their native range, which provides the opportunity to study host-parasite interaction in a new habitat. The present study describes the parasitic fauna in four alien invasive pufferfish species (Lagocephalus guentheri, L. sceleratus, L. suezensis, and Torquigener flavimaculosus) of various sizes and ages on the Israeli Mediterranean coast. The parasite fauna of these species was diverse (Maculifer dayawanensis Digenea; Calliterarhynchus gracilis, Nybelinia africana and Tetraphyllidea larvae Cestoda; Hysterothylacium reliquens, Hysterothylacium sp. and Raphidascaris sp. Nematoda; Trachellobdella lubrica Hirudinea and Caligus fugu and Taeniacanthus lagocephali Copepoda) and consisted of mostly generalist species, most likely acquired in the new habitat, and specialist copepod ectoparasites, having co-invaded with the pufferfish. Additionally, the oioxenic opecoelid digenean Maculifer dayawanensis was found in two pufferfish species. The genus was previously only known from the Indo-Pacific Ocean, representing the eighth reported case of a Lessepsian endoparasite so far. Our results suggest a change in parasite fauna to native Mediterranean species in the pufferfish like previously reported in other Lessepsian migrant predatory fish species and a wider spread of co-invasion of fish endoparasites to the Mediterranean Sea than previously assumed. The study also provides several new host records and the first report for parasites in T. flavimaculosus.


Introduction
The opening of the Suez Canal in 1869 brought down several million years of geographical isolation and enabled the immigration of Red Sea species to the Mediterranean Sea (Por, 1978;Galil et al., 2018), This 'Lessepsian migration' caused a dramatic change in the fauna of the Levantine Basin and, to a lesser degree, the Mediterranean Sea in general (Por, 1978;Galil, 2007a, b). By 2020, 400-500 species from the Red Sea could be confirmed in the Mediterranean Sea, and although the rate is slowing down compared to previous decades, the number is still growing (Zenetos et al., 2012;Galil et al., 2020). Among the alien invasive species are five pufferfish, which originally inhabited the Red Sea and the Indo-Pacific Ocean (Golani et al., 2021). The species are the diamond-back puffer (Lagocephalus guentheri Miranda Ribeiro, 1915) first reported from the Dodecanese Islands (Greece) (Sanzo, 1930), the Suez puffer (Lagocephalus suezensis E. Clark & Gohar, 1953) first reported from Lebanon (Mouneimne, 1977), the yellow-spotted puffer (Torquigener flavimaculosus Hardy & Randall, 1983), first reported in Israel (Golani, 1987), the silver-cheeked toadfish (Lagocephalus sceleratus (Gmelin, 1789)) reported simultaneously from Israel (Golani and Levy, 2005) and Turkey (Akyol et al., 2005) and the rare and diminutive bathydemersal spiny blaasop (Tylerius spinosissimus (Regan, 1908)), reported from the Island of Rhodes (Greece) (Corsini et al., 2005). All of the species have extended their range after introduction, in particular L. sceleratus, which is now considered both a massive threat and a significant nuisance to the local ecosystem and human population alike (Streftaris and Zenetos, 2006;Kalogirou et al., 2010Kalogirou et al., , 2012Nader et al., 2012;Kalogirou, 2013).
The knowledge on pufferfish parasites is mostly limited to reports of cultured fish from Japan (Ogawa and Inouye, 1997;Ogawa and Yokoyama, 1998) and the Hawaiian Islands in the Central Pacific (Palm and Bray, 2014). Reports however often focus more on the recorded parasites than their host (Ikeda et al., 2006;Ito et al., 2006;Mejia-Madrid and Aguirre-Macedo, 2011), and very few comprehensive studies on the parasite fauna of wild living species were done (Fajer-Ávila et al., 2004;Á lvarez-Borrego and Fajer-Ávila, 2006;Beveridge et al., 2014;Prasad and Radhakrishnan, 2010). For the genus Lagocephalus Swainson, 1839 some comprehensive publications exist (e. g. El-Lamie and Abdel-Mawla, 2012; Prasad and Radhakrishnan, 2010;Bakopoulos et al., 2017), but on the species level, the inaccurate distinction between closely related and similar-looking pufferfish may complicate the issue on taxonomic status (Matsuura et al., 2011;Farrag et al., 2015;Giusti et al., 2019).
The partial loss (strip off) of the native specialized parasite fauna is considered part of the success of invasive species (enemy release hypothesis) (Torchin et al., 2002;Colautti et al., 2004;Hänfling, 2007;Vignon et al., 2009;MacColl and Chapman, 2010). Many factors, however, contribute to the success of Lessepsian migrant fish (Arndt and Schembri, 2015), which also tend to accumulate a new, rich generalist parasite fauna in their newly colonized habitats (Merella et al., 2016;Boussellaa et al., 2018). The Lessepsian migration of parasites has been recorded first in 1971/1972 for ectoparasites on fish. Both monogeneans on rabbitfish (Siganidae) (Paperna, 1972;Ktari and Ktari, 1974) and copepods on grey mullets (Mugilidae) (Botros, 1971) have been reported to live on their native hosts in the newly colonized habitat, thus having survived the adverse salinity, oxygen, and temperature conditions (Por, 1978) of the Suez Canal. Findings of ectoparasitic isopods (Trilles and Bariche, 2006), monogeneans (Pasternak et al., 2007), and copepods  immigrating with their hosts have increased in later years. Unlike ectoparasites, Lessepsian endoparasites are protected inside their hosts from adverse external conditions during the passage of the Suez Canal. Because of their heteroxenous life cycles, they may lack the obligatory intermediate hosts to close them in the new habitat and thus have a lower chance of establishing themselves. So far, three species of Myxozoan parasites have immigrated and established from the Red Sea (Diamant, 2010), and four digenean species have been reported (Fischthal, 1980;Merella et al., 2016). The most recent reports are Allolepidapedon fistulariae Yamaguti, 1940 andNeoallolepidapedon hawaiiense Yamaguti, 1965, are parasitizing in the blue-spotted cornetfish Fistularia commersonii Rüppell, 1838, another Lessepsian migrant fish. Both parasite species, well known from their host in the native habitat, so far however lack reports of larval stages in the Mediterranean Sea and may have simply co-invaded as adults with individual fish from the Red Sea or may just be at a starting point of an establishment (Pais et al., 2007;Merella et al., 2010Merella et al., , 2016. The other two reported species are an old single report of Thulinia microrchis (Yamaguti, 1934) Bray, Cribb & Barker, 1993 in the marbled spinefoot Siganus rivulatus Forsskål & Niebuhr, 1775 and the taxon inquirendum larval form 'Monilicaecum ventricosum ' Yamaguti, 1942 in the brushtooth lizardfish Saurida undosquamis (J. Richardson, 1848) (Fischthal, 1980). None of these four digeneans have been found in Lessepsian pufferfish or any native Mediterranean fish species.
Three studies exist on pufferfish parasites from the Mediterranean Sea and Suez Canal (El-Lamie and Abdel-Mawla, 2012;Ö zak et al., 2012;Bakopoulos et al., 2017). At present, of the Lessepsian migrant species, only L. sceleratus has a comprehensive report of parasites from both the native and invaded range. For L. guentheri, reliable data is only available from the Mediterranean Sea and the Suez Canal (El-Lamie and Abdel-Mawla, 2012;Ö zak et al., 2012). In the case of L. suezensis, the only reported parasite is the invasive caligid copepod in the Mediterranean Sea (Özak et al., 2012) (Overview in Table 1). For T. flavimaculosus, no reports exist at all. This present study aims to describe and provide new information on the parasite fauna of the Mediterranean invasive pufferfish by examining L. guentheri, L. sceleratus, L. suezensis, and T. flavimaculosus. It also aims to provide the first report for parasites in T. flavimaculosus throughout its distribution. The qualitative and quantitative assemblage of pufferfish may finally provide new insights into their role as possible carriers of potentially dangerous invasive parasite species.

Study area and data collection
All the fish examined in this study were collected between the 3 rd of June 2019 and the 1 st of June 2020 (Table 3). The samplings were carried out by multiple methods (gill nets, trawl nets, fishing rods, strand findings) with the aid of anglers, artisanal and commercial fishermen, in depths of 1-80 m over sandy and rocky bottoms. The collected fish were either immediately dissected, kept at 4 • C for a maximum of 48 h, or frozen at − 20 • C until dissection at the ichthyological laboratory at the Israeli Oceanographic and Limnological Research Institute (IOLR).

Studied species and dissection
A total of 139 specimens were examined (n = 35 specimens of each Lagocephalus sceleratus, L. suezensis, and Torquigener flavimaculosus, n = 34 specimens of L. guentheri). Dissections were performed following standard procedures (Cribb and Bray, 2010;Palm and Bray, 2014;Klimpel et al., 2019). First, basic morphometric data, total length (L T ), standard length (L S ), and total weight (W T ) were taken. Hereafter, the body surface and openings (eyes, skin, fins, gills, nostrils, anus, and mouth cavity) were investigated macroscopically. For dissection, gill arches and eyes were removed first, afterwards, the body cavity was opened and the internal organs (liver, gall bladder, digestive tract, gonads, swim bladder, and heart) were removed. The organs were opened and inspected macroscopically and with a stereomicroscope (Zeiss Stemi 508 0.63-5.0, Carl Zeiss Microscopy GmbH, Jena, Germany). All found parasites were mechanicaly cleaned of stomach content and transferred into buffered saline solution (0.9% NaCl) and morphologically gross sorted. Parasites were transferred and stored in 96% EtOH for molecular identification or 80% EtOH for morphology. The stomach content was also identified to the lowest possible taxonomic level.

Molecular identification of cestodes and nematodes
Total genomic DNA was extracted using a Qiagen® DNeasy Blood & Tissue Kit (QIAGEN GmbH, Hilden, Germany) following the manufacturer's protocol. In the case of very high parasite counts, a representative subsample, as described in Santos at al. (2017), was analysed. Extracted DNA was used to amplify the ITS1-5.8s-ITS2 sequence region (for nematodes) and the 28S rRNA gene (for cestodes) For PCR protocols see Table 2. The samples were purified with Qiagen QIAqick® PCR Purification Kit (QIAGEN GmbH, Hilden, Germany) and sent to SEQLAB (Microsynth SEQLAB GmbH, Göttingen, Germany) for sequencing. The obtained sequences were edited with MEGA X ( (Kumar et al., 2018) and DNA Baser V4 Software (Heracle BioSoft SRL, Arges, Romania). The results were compared to the GenBank™ gene database and selected sequences were additionally uploaded to the GenBank™ for reference. Voucher specimens were deposited in the 'Zoologische Lehrsammlung' in Berlin, Germany, for further research.

Statistical analyses
The prevalence (P), mean intensity (mI), intensity (I), and abundance (A) of each parasite species were calculated, according to Bush et al. (1997). The parasite community composition was calculated by an abundance-based altered core-/satellite-species concept as used by Zander et al. (2000). The ratio of ecto-versus endoparasites (E/E) was calculated according to Palm and Bray (2014). Differences in parasite communities between the pufferfish species were compared by creating a Bray-Curtis similarity measure with PRIMER 7 software (Primer-e, Massey University, Albany, Auckland, New Zealand) and running ANOSIM (Analysis of Similarities), SIMPER (Analysis of Similarity Percentages) analysis and an MDS (nonmetric multi-dimensional scaling plot).
The adult Digenea were found in the digestive tract lumen of L. guentheri and T. flavimaculosus, with a prevalence and intensity of 11.8% and 1-9 for L. guentheri and 2.9% and 1 for T. flavimaculosus.
Total length (L T ) of Digenean (n = 2) was 1772-2225 μm, body width 895-915 μm, but when alive, distinctly more elongated (L T > 2x body width). The oral sucker was subterminal, a diameter of 110-251 μm. The following prepharynx was surrounded by a prominent, muscular postoral ring. The pharynx length x width was 123 μm × 110 μm. The ventral sucker was closely located to mid-body and larger than the oral sucker, with a diameter of 290-425 μm. Vitelline follicles measured 22 μm × 21 μm in diameter, reaching from the pharynx to the end posterior of the body cavity. Pigment granules were present, and sparsely distributed throughout the body. Testes were located opposite each other, measuring 160 μm × 114 μm and 146 μm × 116 μm. The ovary overlapped the right testis dorsally and was located posterodorsally to the ventral sucker. Eggs were oval and operculated, length x width was 26-43 μm x 14-30 μm (Fig. 1). The presence of the muscular post-oral ring around the prepharynx, body length over 2x body width, and sparse pigment vesicles throughout the body identified the Digenea as genus Maculifer Nicoll, 1915 according to Martin et al. (2018). The body Table 3 Catch data, location, and measurements of the examined specimens of the four pufferfish species at the Israeli Mediterranean coast.   (2), cirrus sack (3), ovary dorsally behind left testes (4) and opposite testes (5), as well as vitelline follicles reaching up to the pharynx. Follicles extend over the whole body and are partially omitted in the drawing for the sake of clarity.
size, size, and position of the measured internal organs either pointed towards M. indicus or M. dayawanensis since the exact position of the genital pore in relation to the intestinal bifurcation was not recognizable. The extent of the vitelline follicles from the pharynx, not just from the intestinal furcation to the posterior end of the body and its hosts identified as Maculifer dayawanensis Sheng and Tong, 1990 according to Sheng and Tong (1990) and Madhavi and Bray (2018). Specimens are deposited in the collection of the Museum für Naturkunde Berlin (catalogue nos. ZMB E.7672; ZMB E.7673). All three found cestode species were plerocercoid larval stages from two families, Trypanorhyncha and Tetraphyllidea. Two species of Trypanorhyncha were encysted on various organs, the Tetraphyllidea were free inside the digestive system. The first species was found encysted on the intestines (mesenteries, swim bladder) surrounded by connective tissue capsules in the body cavity in L. sceleratus. Prevalence was 22.9% and intensity 1-3. Freed from the blastocysts (n = 3), L T was 3.4-10.2 mm with a transparent long and slender body (width 618.7-1000.8 μm at pars bothrialis). Scolex is long, slender, and well-developed. Two opposing posteriorly slightly notched bothridia are with length of 529.1-839.0 μm. Four long tentacles (length 1081.4-1934.5 μm; basal width 25.2-63.7 μm) are bearing hollow hooks. Tentacle sheaths are as spirals inside the scolex, ending in four elongated (length 484.6-1934.5 μm; width 131.9-224.4 μm) bulbs. The appendix behind the pars bulbosa is long and slender. All measurements and general morphological traits were consistent with the description by Palm (2004). The morphology, morphometric measures, and shape of the tentacle sheaths identified the specimens as plerocerci of Callitetrarhynchus gracilis. Blasting on the GenBank against 28S rRNA sequence (n = 1, accession no. OM867573) also gave 100% Query Cover and 98.06-100% Identity for C. gracilis (accession nos. FJ572957.1; DQ642758.1-DQ642759.1; MG694210.1; MN488532.1).
The second species was encysted on the stomach wall of L. guentheri.
P was 2.9% and intensity was one individual. L T was 766.0 μm, width 457.6 μm at pars bothrialis. A compact, well-developed scolex with four opposing triangular bothridia (length 379.5 μm). Four straight and short tentacles with solid hooks (length 272.8 μm; basal width 29.7 μm), retracted in straight sheathes, ending in four short bulbs (length 231.8 μm). Appendix behind the pars bulbosa short. The measurements fitted the ones by Palm (2004) for plerocercoids of Nybelinia africana Dollfus, 1960. The small size, morphological features, and morphometric measurements identified it as a plerocercoid of N. africana. The third cestode were plerocercoids in the digestive tracts of Lagocephalus spp. Prevalence and intensity ranged from 11.4% and from one to nine individuals in L. suezensis to 50.0% and from one to 200 individuals in L. guentheri. L T was 0.2-2 mm, with the bodies being transparent and brittle, showing a diverse body shape, with several distinct morphotypes. Four bothria were present on the scolex, sometimes sitting on peduncles, sometimes retracted into the scolex or completely absent and a central apical sucker was present on the very top of the scolex. The morphological features were identical to the characteristics described in Euzet (1994) and used for his diagnosis of Tetraphyllidea Carus, 1863. The four bothria in combination with the apical sucker and the location of the plerocercoids attributed the cestodes to the order Tetraphyllidea.
Among the nematodes, three species were present. The first species was present in all four pufferfish species, living freely inside the intestine. Only two specimens were found enrolled on the liver and gills in L. suezensis. Prevalence and intensity ranged from 31.4% and from one to eleven individuals in L. suezensis to 88.6% and from one to 80 parasites in L. sceleratus. Larval stages (L3/L4) were present in all pufferfish, adults only in larger specimens of L. sceleratus and only a single adult male in L. guentheri. L T was 2-40 mm, with a long slender body reaching its greatest width near the middle. The cuticle was smooth, with no ornaments on the conical tail. Both posteriorly directed ventricular appendix and anteriorly directed internal caecum were present. Adults with three pronounced lips and spicules (males), or visible ovaries (females). Measurements and morphological features fell within the description of Deardorff and Overstreet (1981) provided for Hysterothylacium reliquens (Norris & Overstreet, 1975) Deardorff and Overstreet, 1981. Blasting on the GenBank™ against ITS1-5.8s-ITS2 sequence regions (n = 5, accession nos. OM888581-OM888585) resulted in a 94-100% Query Cover and 99.8-100% Identity for deposited vouchers of H. reliquens from the Mediterranean Sea (accession nos. MF062507.1-MF062509.1), Arabian Gulf (accession nos. KX786286.1-KX786293.1) and the Atlantic US-coast (accession nos. MF668856.1; MF668873.1; MF668876.1; MF668880.1) and confirmed the identity of H. reliquens.
The second nematode was indistinguishable from H. reliquens larvae and could only be identified through molecular identification after blasting. The host was L. guentheri, with a prevalence of 5.8% and an intensity of two individuals. Only two specimens could be identified (accession nos. OM888586-OM888587). Query Cover was 87-88% and identity 100% for an unidentified larval form of the genus Hysterothylacium Ward & Magath, 1917 from the Mediterranean Sea (accession nos. MT365529.1-MT365537.1) described by Guardone et al. (2020).
The third nematode was only present in L. sceleratus, with a prevalence of 2.9% and intensity of two parasites, with two specimens found. All measured specimens were incomplete, so no L T was measurable. The body of the nematode was very elongated and slender, reaching the greatest width near the middle. A posteriorly directed ventricular appendix was present, and intestinal caecum was absent. Length of ventricle was 136-162 μm, width of ventricle was 97-100.6 μm, and oesophagus length was 1488-2009 μm. Three prominent hooked lips were located on the head. One measured specimen was an adult female, having fully developed ovaries and a gonopore. The shape of the lips, the presence of a posteriorly directed ventricular appendix, the absence of an intestinal caecum, and the measurements defined the parasite as belonging to the genus Raphidascaris Railliet & Henry, 1915 as described by Moravec (1998). Specimen are stored in the institute's collection.
The single Hirudinea specimen was found on the skin of L. sceleratus under the pectoral fin, with a prevalence of 2.9% and an intensity of one individual. L T was 3.5 cm and the body was of brown-green color. The anterior part of the body was distinctly slenderer and of a contrasting yellow in color. On the oral sucker was one pair of eyes. The body greatly widened towards the rear third, at the location of a pair of contractive vesicles, and then tapered down into a small posterior sucker without ocelli. The skin was smooth, except for 14 pairs of contracting lateral vacuoles on the side of the body. The size, presence of vesicles and vacuoles, coloration and present morphological features distinguished the leech as Trachelobdella lubrica (Grube, 1840) according to the description given by Ben Ahmed et al. (2015) and Noël and Sittler (2021).
Two copepod species were free-living on the gill filaments of the pufferfish. The first copepod was present on Lagocephalus spp. Prevalence and intensity ranged from 20.0% and from one to two individuals in L. suezensis to 38.2% and from one to eight individuals in L. guentheri. Mostly adults and rarely chalimus stages were present. The genus could be identified as Caligus Müller, 1785 by the presence of two suction organs (lunnules) on the base of the antennas in combination with a visible single nauplius eye on top of the cephalosome. Measurements for adult ♀ (n = 4) were: 3.1-3.5 mm for L T , cephalothorax length x width 1.0-1.3 mm x 1-1.2 mm. The length of the genital complex was 1.3-1.6 mm, 1.36-1.90 times as long as the two-segmented abdomen (0.7-1.0 mm), while the first segment of the abdomen was 2.11-2.59 times as long as the second. The maxillipeds showed a well-developed, tapering process proximally on the medial margin. The measurements, traits, and hosts are consistent with the description for Caligus fugu Yamaguti andYamasu, 1959 by Boxshall and.
The second copepod exclusively inhabited L. guentheri. Prevalence was 88.2% and intensity was from one to 42 individuals. Egg-bearing females were predominantly present, whereas males were significantly fewer. Measurements for ♀ (n = 2) were: L T 2.3-2.7 mm, cephalothorax length x width 0.5-0.7 mm x 0.7-0.8 mm, 22-26% of L T . Body threesegmented, with segments almost as wide as cephalothorax, abdomen with four segments. Caudal ramus length x width 66-74 μm × 40 μm, 1.7-1.9 times longer than wide. Four rows of spinulae on the dorsoventral surface of the genital segment, with a fifth close to the base of the caudal ramus. The diagnosis for Taeniacanthus lagocephali Pearse, 1952 by Dojiri and Cressey (1987) fitted the measured specimens. The measurements, distinct three free thoracic segments equal in width to the cephalothorax, armature at the anal segment, and its presence solely on L. guentheri, a close relative of the originally described hosts, confirmed the species as T. lagocephali.

Differences in parasite assemblages
The differences in parasite assemblage between the pufferfish species were mostly determined by the most common and abundant parasites (H. reliquens, Tetraphyllidea, protists, and T. lagocephali). ANOSIM analysis showed that the parasite assemblies of each of the four species differed from the others, but with some overlaps for each species (R = 0.48; p = 0.001-0.015). SIMPER analysis showed that assemblies were the most similar between T. flavimaculosus and L. sceleratus (Average dissimilarity = 61.4%), followed by T. flavimaculosus and L. suezensis (70.1%) being equally different from L. suezensis and L. sceleratus (70.3%). The differences between L. scleratus and L. guenteri were larger (79.5%) and the biggest dissimilarity was found between T. flavimaculosus and L. guentheri (83.7%) as well as L. suezensis and L. guentheri (84.5%). Other factors such as size, size classes, or catch date failed to yield significant differences. An MDS plot (Fig. 2) also supported the results by showing a clustering in dependence of species with a sufficiently low-stress level (0.15). All infected L. guentheri formed a distinct cluster, as well as L. sceleratus. L. suenzensis clustered partially with L. sceleratus, whereas T. flavimaculosus showed the biggest overlap with L. suezensis.

Discussion
The study was focused on parasites of the invasive pufferfish from Israeli Mediterranena Sea, with emphasis on potential Lessepsian migrant parasites. The first report of parasites in Torquigener flavimaculosus was additionally presented. Six out of seven newly reported Mediterranean native parasites represented new host records for the studied pufferfish. The opecoelid Digenea Maculifer dayawanensis, previously only recorded from the South-Chinese Sea, could be reported as a new Lessepsian endoparasite, found in the digestive tract of Lagocephalus guentheri and T. flavimaculosus. In addition, we confirmed the two previously reported Lessepsian copepods.

Digenea
The opecoelid Digenea Maculifer dayawanensis was either a rare or a satellite parasite found in the pufferfish, with a prevalence (2.9-11.8%) similar to previous reports of Indo-pacific pufferfish (11.0%) (Martin et al., 2018). Only mature specimens were present in L. guentheri and T. flavimaculosus. This record represents the fifth report of a macroscopic Lessepsian endoparasite from the Mediterranean Sea. All previous reports were also about Digenea in fish, have been found exclusively as adults or could not be properly identified (invalid larval stages) (Fischthal, 1980;Merella et al., 2010Merella et al., , 2016. In the case of the adult Digenea, a possible co-invasion with the hosts from the Red Sea could not be excluded (Pais et al., 2007;Merella et al., 2010Merella et al., , 2016. For the genus Maculifer, only adults have been reported from the Indo-Pacific (Nicoll, 1915;Cribb, 2005;Liu et al., 2010;Martin et al., 2018). Since M. dayawanensis was both present in the non-migratory T. flavimaculosus and the more migratory and habitat-changing L. guentheri, it is so far the strongest example of a Lessepsian endoparasite successfully establishing itself. The parasite was not found by El-Lamie and Abdel-Mawla (2012), who dissected 294 specimens of L. guentheri. Accordingly, M. dayawanensis either may have entered the Mediterranean Sea after 2012 or at an unknown point in the past, most likely coinciding with the immigration of L. guentheri in 1930 (Sanzo, 1930). Another possible introduction pathway is ballast water of ships instead of direct immigration, given the extremely long distance between the first reported location, Daya Bay in the South-Chinese Sea (Sheng and Tong, 1990), and the Israeli Mediterranean Sea. The Suez Canal is a major global shipping corridor, and rules for ballast water treatment only were implemented in 2017 (Galil et al., 2018). Ballast sediments in particular have not been studied as thoroughly as ballast water, are not always filtered, and may provide a suitable habitat for the first hosts, infected snails, releasing them during ballast water change (Gollasch et al., 2019). At least one opecoelid Digenea is reported to have a shortened life cycle, omitting intermediate and even final hosts in its ontogenesis and leaving the primary intermediate host already as adults (Barger and Esch, 2000). Both infected pufferfish species consumed gastropods, which are possible first intermediate hosts for Maculifer spp. A possible shortened life cycle would also explain the absence of encysted larval stages in the pufferfish stomachs from ingestion, as described for a closely related genus of Digenea (Cable, 1956). This study provides L. guentheri and T. flavimaculosus as new host records for M. dayawanensis.

Copepoda
The copepod Caligus fugu was among either the secondary or satellite parasites, and the presence of egg-carrying females and chalimus stages proved a successful reproduction in Israeli waters. This parasitic copepod was previously reported from L. guentheri and L. suezensis in the Mediterranean Sea and the Suez Canal, and other pufferfish species in the Indo-Pacific (mostly the genus Takifugu ). The prevalence was slightly lower than previously described for the Mediterranean Sea (34.7-80.0%) and similar to the Suez Canal (34.6%) (El-Lamie and Abdel-Mawla, 2012;Ö zak et al., 2012). Only Lagocephalus spp. were infected, while the sand-bound T. flavimaculosus (Farrag et al., 2016) either did not come into contact with the parasite or was no suitable host. The presence of C. fugu on L. sceleratus either presented a possible successful host change after its original immigration, or it was previously just unreported from this species. This study The copepod Taeniacanthus lagocephali was one of the common parasites of L. guentheri. The prevalence was similar to previous reports from the Mediterranean Sea (94%) and significantly higher than in the Suez Canal (34.6%). The presence of egg-carrying females confirmed the successful reproduction of this species in Israeli waters, as previously reported in Turkey and the Suez Canal (El-Lamie and Abdel-Mawla, 2012;Ö zak et al., 2012). The infection pathway was identical, based on the habitat preference, to C. fugu (Farrag et al., 2016).

Cestoda
All cestodes found were from specialized parasite families (Trypanorhyncha and Tetraphyllidea) with multiple-host life cycles. Pufferfish are particularly susceptible to trypanorhynch cestodes, showing a high diversity of these parasites in their native habitats, but also for other cestodes (Palm and Bray, 2014). The trypanorhynch cestode Callitetrarhynchus gracilis was among the satellite parasites and only present in larger specimens of L. sceleratus (L T ≥ 45 cm). The scolex was well developed; distinguishing L. sceleratus as a third intermediate host according to Palm (2004). The cestode has been reported from over 140 intermediate hosts, including pufferfish (Palm, 2004;Beveridge et al., 2014). Small clupeids (sardines), engraulids (anchovies), or carangids (jacks) are second intermediate hosts, larger predatory fish are third intermediate hosts, and carcharinid sharks are the final hosts (Palm, 2004). Since large L. sceleratus fed on clupeids, carangids, and engraulids in the Mediterranean Sea, the infection pathway was accomplished by consumption of second intermediate hosts. C. gracilis has been globally reported, including in the Mediterranean and Red Sea (Palm, 2004;El-Sayed Abdou and Palm, 2008;Abdelsalam et al., 2016). This study identified L. sceleratus as a new intermediate host.
The trypanorhynch cestode Nybelinia africana was a rare parasite and only a single specimen was found in L. guentheri. The specimen had a fully developed scolex typical for plerocercoids in a third intermediate host (Palm, 2004). The second intermediate hosts are either small fish or euphausid shrimp ("krill"), a larger fish is the third intermediate host, and a large predatory fish or cephalopod is the fourth intermediate or paratenic host. The final hosts are a diverse group of elasmobranchs (Palm, 2004). Fish, including goatfish, and cephalopods were important prey items for L. guentheri in this study and are documented as intermediate hosts for this parasite (Palm, 2004). The genus Nybelinia Poche, 1926 is widely distributed and N. africana has been reported from the Mediterranean Sea, Atlantic, and the Indian Ocean surrounding Africa (Palm, 2004;Merella et al., 2010Merella et al., , 2016. The genus has previously been reported in L. sceleratus (Beveridge et al., 2014) and N. africana also infects the Lessepsian predatory fish Fistularia comersonii in the Mediterranean Sea, with a much higher prevalence (82.4-100.0%) (Merella et al., 2016). This study provided L. guentheri as a new host for N. africana and therefore was the first report for this species in a tetraodontid fish.
Larval Tetraphyllidea were either common or satellite species infecting Lagocephalus spp. Tetraphyllidea were reported to infect L. guentheri in the Suez Canal (El-Lamie and Abdel-Mawla, 2012), having a much lower prevalence than in the present study (prevalence of 6.1%). Several distinct morphotypes were present, possibly representing several species infecting the pufferfish. Tetraphyllidea have life cycles involving multiple hosts, including copepods, small fish, squid, and larger fish as paratenic hosts, with elasmobranchs as final hosts (Dick et al., 2006;Pardo-Gandarillas et al., 2009). They are reported globally, including the Mediterranean Sea (Keser et al., 2007;Pardo-Gandarillas et al., 2009;Felizardo et al., 2010). The high numbers of up to 200 plerocercoids in the pufferfish assigned them as paratenic hosts. The sand-bound T. flavimaculosus was not infected, so Lagocephalus spp. may have got infected in the Posidonia oceanica meadows they inhabit (Kalogirou et al., 2010(Kalogirou et al., , 2012Kalogirou, 2013;Farrag et al., 2016). The comparatively higher prevalence and intensity of the infection in L. guentheri could be attributed to its earlier switch and a stronger emphasis on fish (several confirmed small intermediate Mediterranean hosts (Radujković and Š undić, 2014)) and cephalopods (suggested important intermediate hosts of Mediterranean shark tapeworms (Tedesco et al., 2020)) in its diet, thereby accumulating more plerocercoids over time. L. sceleratus and L. suezensis were new records for the Tetraphyllidea as paratenic hosts.

Nematoda
The nematodes belonged to the Raphisascarididae, maturing in teleost fish. Hysterothylacium reliquens was present in all four pufferfish species. It was one of the common or a secondary parasite. While adult nematodes were only present in L. sceleratus and one single specimen was found in L. guentheri, marking them as final hosts. The two smaller pufferfish species only acted as intermediate hosts. The entire life cycle and first hosts of H. reliquens are unknown, but larval stages are described in various fish, cephalopods, and decapod crustaceans (Deardorff and Overstreet, 1981). For the related H. aduncum, small crustaceans and polychaetes are known as the first and intermediate host (Køie, 1993;González, 1998). All pufferfish species in this study fed on crustaceans. The high prevalence and intensity in L. sceleratus could be attributed to the high percentage of crustaceans in its diet. H. reliquens is reported globally (Deardorff and Overstreet, 1981;Zhao et al., 2016), including the Mediterranean Sea (El-Sayed et al., 2004;Simsek et al., 2018). Reported hosts include pufferfish in the Atlantic Ocean (Rodríguez et al., 2019). Sequence analyses revealed high similarity (Query Cover 94-100% and identity 99.8-100%) with various vouchers from the Atlantic Ocean, Arabian Gulf and the Mediterranean Sea, confirming this nematode as globally distributed and not as a Lessepsian migrant species. This study identified L. guentheri, L. sceleratus, L. suezensis, and T. flavimaculosus as new hosts of H. reliquens.
The second nematode also belonged to the genus Hysterothylacium, being different from H. reliquens. The sequence analysis showed a high similarity to a Hysterothylacium sp. larval form reported from octopus (Eledone spp.) in the Mediterranean Sea with a similarly low prevalence (6.7%) by Guardone et al. (2020). The sole presence and low prevalence of larvae in L. guentheri could be attributed to the high percentage of cephalopods in its diet. L. guentheri thereby is a paratenic host for Hysterothylacium sp. Taxonomic status of this nematode as a native parasite to the Mediterranean Sea could also be confirmed by molecular markers. This study presented L. guentheri as a new host for this parasite and shed light on the life cycle of this not yet fully identified taxon.
The third nematode belonged to the genus Raphidascaris and was a rare parasite of L. sceleratus. The presence of an adult female specimen identified L. sceleratus as a final host of this parasite. Raphidascaris sp., R. mediterraneus and R. macrouri, were reported from predatory fish in the Mediterranean Sea (Dallarés et al., 2014;Pérez-i-García et al., 2015). Since the isolated parasite specimens lacked significant morphological features, identification to species level was impossible. The full life cycle of this genus is only known for one species, with invertebrates as the first hosts, with small fish as the secondary or paratenic hosts, and with large predatory fish as final hosts (Anderson, 2000). Pufferfish are known as final hosts to the genus Raphidacaris in the Pacific Ocean (Moravec and Justine, 2020). Therefore, the found Raphidascaris may theoretically be a Lessepsian migrant. This study demonstrated the second report of the genus Raphidascaris in Tetraodontidae and also described L. sceleratus as a new host.

Hirudinea
The marine leech Trachellobdella lubrica was a rare parasite of L. sceleratus. The species has a circumglobal distribution in warm and temperate waters, including the Arabian and the Mediterranean Sea (Saglam et al., 2003;Ö ktener and Utevsky, 2010), and prefers shallow waters (Noël and Sittler, 2021). Trachelobdella lubrica parasitizes on a wide range of hosts of all sizes (Epshtein, 1973) and has been reported to parasitize another Lessepsian migrant, the cornetfish Fistularia commersonii (Merella et al., 2016). This study identified L. sceleratus as a new host for this species and the first report of a T. lubrica parasitizing a member of the Tetraodontidae.

Overall parasite community
The comparison of our study with the available parasite assemblies of other pufferfish (both free-living and captive) showed some key differences. Monogenea, important ectoparasites of pufferfish in the Indo-Pacific Ocean, were conspicuously absent and the diversity of Digenea was also significantly lower (Ogawa and Inouye, 1997;Ogawa and Yokoyama, 1998;Fajer-Ávila et al., 2004;Á lvarez-Borrego and Fajer-Ávila, 2006;Prasad and Radhakrishnan, 2010). The same differences occurred in the data from Hawaii (Palm and Bray, 2014), where Monogenea were also rare, but the Digenea were much more diverse. Of the studied species, only L. sceleratus has been known to be infected by Monogenea in Chinese waters (Jianying et al., 2003;Hung et al., 2020). Monogenea belong to the most successful Lessepsian parasites (Paperna, 1972;Ktari and Ktari, 1974;Pasternak et al., 2007). Their absence in pufferfish from the Mediterranean Sea suggests either a successful strip off by the Suez Canal passage, or their general lack in the Red Sea, from which no records were available.  Bakopoulos et al., 2017), the diversity of native parasites was much higher, while the Lessepsian copepods could be confirmed. In comparison to Hawaii (Palm and Bray 2014), a loss of parasite diversity for the whole family of Tetraodontidae in the Mediterranean Sea is also suggested by the E/E ratio differences of this study. The combined E/E ratio was higher (0.5 vs. 0.11) (Palm and Bray, 2014), indicating the loss of endoparasites compared to the Indo-Pacific. At the species level, however, compared to the studied Hawaiian members of the same pufferfish genera (Lagocephalus lagocephalus (L.) and Torquigener hypselogeneion (Bleeker, 1852)), the number of found parasite species per pufferfish species was higher (2-7 vs. 2-3) in average, while the E/E ratios per species were similar (0-0.67 vs. 0-1), typical for a diverse ecosystem (Rückert et al., 2009). The shared parasites between the Mediterranean Sea and Hawaii were copepods, most likely specialists, and generalistic Rhaphidascarididae nematodes of the genus Hysterothylacium Ward & Magath, 1917. These shared parasites are further supporting the strip off of native parasites during the transit of the Suez Canal, except for very few oioxenic species, and the acquisition of a rich fauna of generalists in the Mediterranean Sea. Differences in parasite communities between the four pufferfish species could be explained by their different habitat preferences and prey spectrum (Kalogirou et al., 2010(Kalogirou et al., , 2012Kalogirou, 2013;Farrag et al., 2016). As Fig. 2 illustrates, L. guentheri was separated from the others. Although, the general ecology of the investigated pufferfish species in the Mediterranean Sea is unknown, L. guentheri is known to be habitat changing and migratory. Additionally, it had the most diverse invertebrate diet and five different teleost prey taxa were found. The other three pufferfish species showed an overlapping pattern, these taxa are known to be less migratory and showed lower variety in invertebrate prey taxa.
The absence of Anisakis sp., a potentially zoonotic nematode, could be explained by unsuitable conditions for the first host in the Levantine Basin, despite its presence in the rest of the Mediterranean Sea (Anderson, 2000;Mayzaud et al., 2000) and the difference in identified Hysterothylacium species by previously neglected molecular identification.
This study has significantly increased our knowledge of the parasite assemblages of four Lessepsian pufferfish in the Mediterranean Sea, on both local and global levels. We found three species of oioxenic Lessepsian parasites. Two copepods could be confirmed from previous studies (El-Lamie and Abdel-Mawla, 2012;Ö zak et al., 2012). The Digenea M. dayawanensis is a new, previously unreported, Lessepsian endoparasite and has possibly become established in the Mediterranean Sea by now. The findings of this study showed similarities to the parasite community of the cornetfish F. commersonii, another large Lessepsian predatory fish, including shared parasites (Merella et al., 2016). Both pufferfish and cornetfish lost most or all of their presumed ectoparasites (enemy release hypothesis) from the Red Sea, and retained endoparasites during the transit through the Suez Canal, which can be considered one factor for their success as alien species (Arndt and Schembri, 2015;Filiz et al., 2017). Pufferfish and cornetfish (Merella et al., 2016), in turn, also acquired a diverse new assembly of native Mediterranean, mostly endoparasitic, generalists.
As the fishes we sampled at availability by different sampling methods, sites and seasons, no spatial nor temporal analysis could be conducted. However, as visibile in Table 3, a host species sampled twice in winter shows the highest parasite species richness, and another host species that was sampled eight times (within nine month) hosted seven parasite species. Therefore, no further seasonal interpretation of found parasite community seems to be feasible.
All studies conducted on the parasites of Lessepsian pufferfish, including this one, are so far limited to the Southern Mediterranean Sea (Levantine and the Aegean Sea) and the Suez Canal (El-Lamie and Abdel-Mawla, 2012;Ö zak et al., 2012;Bakopoulos et al., 2017). Pufferfish, especially L. sceleratus, however, have extended their range to the western parts of the Mediterranean Sea and into the Black Sea (Golani et al., 2021). While neither this study nor the others could find any transfer of possible zoonotic parasites to commercially important species or the introduction of dangerous Lessepsian parasites, this issue should be addressed for the rest of the Mediterranean Sea. Particular emphasis should be placed upon a possible ongoing co-invasion of pufferfish and parasites through the Suez Canal, to detect potentially dangerous species. Additional research on the parasites of juvenile pufferfish would help to better understand their ecological niche and competition with native species. Finally, research on the rare spiny blaasop (Tylerius spinosissimus) could shed more light on the little-known deepwater parasite communities and possible Lessepsian deepwater parasites. Due to the irregular sampling, the sample size in this study varied between years and locality, and many specimens were examined after being frozen for some time. Therefore, it was possible that some parasites (ectoparasites) were displaced or migrated during transportation or preservation, as was the case for Trachelobdella lubrica. Pufferfish in the Mediterranean Sea are not commercially fished, so opportunistic sampling was the only approach that could be adopted for the study, as it had been done before for other non-commercial marine species with decent results (Santoro et al., 2010;Aznar et al., 2012: Merella et al., 2016.

Declaration of competing interest
None.

Acknowledgments
We are very grateful to the sport/professional fishermen and the staff of the IOLR who collaborated and made available the specimens of the pufferfish We are also grateful to Dr. Rodney A. Bray for his help and advice to identify the Digenea. We thank the associate editor and anonymous reviewers for providing valuable and constructive comments on our manuscript. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.