Unregulated introduced fish (Perca fluviatilis Linnaeus, 1758) is host to zoonotic parasites in a small Mediterranean island

Introduced species have a major impact on freshwater ecosystems, particularly on islands. Numerous fish species have been introduced in Corsica (Mediterranean island, southern France) as part of planned programs or clandestinely. The introduction of non-native freshwater fish species can have a range of impacts on the recipient ecosystem, including through the co-introduction of its pathogens. A sample of introduced perch Perca fluviatilis Linnaeus, 1758 from the artificial reservoir of Padula was examined following a report of parasites by an angler. The analyses revealed the occurrence of Eustrongylides sp. (Nematoda) and Clinostomum complanatum (Digenea), two zoonotic parasites in P. fluviatilis. Both parasites are reported for the first time in France. Eustrongylides sp. and C. complanatum may have been introduced with their fish intermediate hosts or through their final bird hosts. The occurrence of the two parasites raises concerns from both a veterinary and human health perspective as they can use a wide range of amphibians as intermediate hosts and can be acquired in humans through the consumption of raw or undercooked fish. Supplementary Information The online version contains supplementary material available at 10.1007/s00436-024-08264-4.


Introduction
Southern and Western Europe are a hotspot for biological invasions, with more than a quarter of total fish diversity being non-native species (Leprieur et al. 2008).Corsica is a small island in the Northwestern Mediterranean off Southern France and Western Italy, home to a particular fish fauna composed of only four native species and naturally lacking the Cypriniformes reported from the Ibero-Franco-Italian region (Roule 1933;Changeux 1998).The fish diversity in Corsica has however been artificially increased by the manmediated introductions of more than 20 non-native fish species, the majority of them for recreational fishing purposes (Roche and Mattei 1997;Roché 2001).The first introduced species was the mosquitofish Gambusia holbrooki Girard, 1859 in an attempt to control mosquitoes (malaria control).Then, the brook trout Salvelinus fontinalis (Mitchill, 1814) and domestic brown trout Salmo trutta Linnaeus, 1758 were introduced in several mountain lakes.Afterwards, several species, e.g., the rudd Scardinius erythrophthalmus (Linnaeus, 1758), the carp Cyprinus carpio Linnaeus, 1758, the pikeperch Sander lucioperca (Linnaeus, 1758), the tench Tinca tinca (Linnaeus, 1758), and the roach Rutilus rutilus (Linnaeus, 1758), were introduced into several artificial reservoirs and spread by anglers to rivers and others reservoirs (Roché and Mattei 1997;Roché 2001).The rainbow trout Oncorhynchus mykiss (Walbaum, 1792) is regularly released in reservoirs and lower watercourses.The study of parasites carried by non-native hosts is of great importance as they may infect native host species, a phenomenon termed parasite spillover, as well as being transmitted to humans in the case of zoonotic parasites (Daszak et al. 2000;Prenter et al. 2004;Chai et al. 2005;Lymbery et al. 2014;Chalkowski et al. 2018;Eiras et al. 2018).
The European perch Perca fluviatilis Linnaeus, 1758 was introduced in Corsica outside the framework of any planned program and from an unknown source.Its presence was first noted in 1984 during the first emptying of the Uspidali reservoir.The species was then transferred to three secondcategory reservoirs in the eastern plain of Corsica (namely Peri, Alzitone, and Tepe-Rosse) and now forms sustainable populations in artificial habitats, i.e., several reservoirs (the four aforementioned ones plus Codole, Tolla, and Padula) and disused gravel quarries (Roché and Mattei 1997;Roché 2001;Fleury and Le Mesle 2022).In such habitats, this species is caught by recreational anglers and regularly consumed.Recently, the occurrence of live parasites was reported by recreational anglers in freshly caught P. fluviatilis from the Padula reservoir.There is no record of transfer of P. fluviatilis to Padula reservoir and its introduction in this ecosystem is believed to be the work of individual recreational anglers.In this reservoir, P. fluviatilis is known to occur since at least 2020.Parasites of P. fluviatilis have never been examined in Corsica.
The aim of this study was to determine whether the European perch population of the Padula reservoir can host zoonotic parasites, using morphological and molecular data.Their origin and potential impact on local fauna and human health are discussed.

Material and methods
Padula Reservoir (0.25 km 2 surface area, 15 m maximum depth) (Fig. 1) is an artificial reservoir on the Furmicaiola stream, located in the north of the island at an altitude of 60 m above sea level, and administered by the Office d'Equipement Hydraulique de Corse (OEHC-Corsica water resources office).It was commissioned in 1992 and serves the purposes of irrigation and drinking water supply (Colonna 2021).To our knowledge, the fish population in Padula consists of P. fluviatilis, S. lucioperca, C. carpio, S. erythrophthalmus, the largemouth black bass Micropterus salmoides (Lacepède, 1802), and a few rainbow trout O. mykiss.Amphibians are also reported from the reservoir, including two endemic Urodela: the Corsican mountain newt Euproctus montanus (Savi, 1838) and the Corsican fire salamander Salamandra corsica Savi, 1838 according to the National Inventory of Natural Heritage (Inventaire du Patrimoine Naturel -INPN) database managed by the Muséum national d'Histoire naturelle (MNHN) and the French Agency of Biodiversity (OFB) (https:// inpn.mnhn.fr/ espece/ cd_ nom/ 59; https:// inpn.mnhn.fr/ espece/ cd_ nom/ 701817).Several water and piscivorous birds visit the reservoir, e.g., the grey heron Ardea cinerea Linnaeus, 1758, the little egret Egretta garzetta (Linnaeus, 1766), and the black-crowned black heron Nycticorax nycticorax (Linnaeus, 1758) (Table 1).
Ten specimens of P. fluviatilis (total length 208 ± 36 (116-248 mm), total weight 123 ± 47 (16-172 g)) were sampled by a recreational angler on 07/06/2023.Fish were brought back dead to the laboratory in an insulated bag by the angler, through the Angling Federation of Corsica (FDAAPPMA2).Fish were dissected and the body surface, opercular cavity, abdominal cavity, liver, swimbladder, and muscle tissue were thoroughly examined.However, neither the digestion (McGladdery 1986) nor the incubation (Shamsi and Suthar 2016) methods were used.As a result, an eventual underestimation of parasite abundance cannot be excluded, especially for Clinostomum, which has been reported in the skull and brain of its fish hosts (Aghlmandi et al. 2018;Shamsi et al. 2021).Parasites were recovered and preserved in either 70% or 96% ethanol for further optic and molecular identification.For scanning electron microscopic (SEM) examination, several randomly selected specimens were fixed in 2.5% glutaraldehyde (in 0.1 M sodium cacodylate buffer at pH 7.2), dehydrated through a graded ethanol series (30%, 50%, 70%, 90%, and 100%), then dried in an Emitech K850 (Quorum Technologies, Laughton, UK) critical point dryer using CO 2 .After mounting, the specimens were coated with platinum in a Q150T-ES sputter coater (Quorum Technologies, Laughton, UK) before examination with a Regulus 8230 scanning electron microscope (Hitachi, Tokyo, Japan) at an accelerating voltage of 2 kV.Parasite indices were calculated following the terminology of Bush et al. (1997).
Molecular identifications were done on randomly selected individuals in order to support the morphological determinations.DNA extraction, amplification, and Sanger sequencing were performed by Eurofins Genomics (Ebersberg, Germany).Both Nematoda and Digenea parasites were identified using the nuclear internal transcribed spacer (ITS) rDNA (ITS1, 5.8S, and ITS2) with the same primers 81_f 5′-GTA ACA AGG TTT CCG TAG GTGAA-3′ (Gustinelli et al. 2010) and ITS2.S_r 5′-CCT GGT TAG TTT CTT TTC CTC CGC -3′ (Matthews and Cribb 1998), following Mazzone et al. (2019).For Nematoda, the mitochondrial cytochrome oxidase subunit 1 (COI) marker was also amplified using H7005 5′-ACNACR TAR TANGTR TCR TG-3′ and L6625 5′-TGR TTY TTYGGNCAYCC-3′ primers (Hafner et al. 1994).DNA amplification was performed by PCR in a final 20 µl volume containing 10 µl of GoTAQ-HotStart Green MasterMix from Promega (Madison WI, USA), 4.5 µl of each of the two primers at 10 pM, and 1 µl of extracted DNA.After denaturation for 2 min at 95 °C, the PCR was run for 35 cycles of 1 min, 95 °C; 30 s, 50 °C; 1 min 30 s, 72 °C, before a final extension for 10 min at 72 °C and conserved at 4 °C on a Biometra Tadvanced thermocycler (Biometra GmbH, Göttingen, Germany).Forward and reverse sequences were assembled and cleaned with Geneious Prime ® 2020.2.4 (https:// www.genei ous.com).Obtained sequences were deposited in GenBank (accession numbers from PP888044-PP888052, PP888083-PP888092 and PP887879-PP887887), and were first blasted through the NCBI platform (https:// blast.ncbi.nlm.nih.gov/) in order to have a preliminary identification.When several sequences were available in GenBank, molecular analyses were carried out.Sequences were aligned using MEGA X (Kumar et al. 2018) and the MUS-CLE algorithm (Edgar 2004) with a molecular dataframe of reference from previous studies (Lazarova et al. 2006;Caffara et al. 2011;Xiong et al. 2013;Sereno-Uribe et al. 2013;Maleki et al. 2018;Li et al. 2018;Nitta and Ishikawa 2019;Locke et al. 2019;Pekmezci and Bolukbas 2021;Youssefi et al. 2023) (Supplementary data).JModeltest v.2.1.1 (Darriba et al. 2012) was used to estimate the best evolution model for the Bayesian phylogenetic inference analyses selected under the Bayesian Information Criterion (GTR + I for COI of Nematoda and HKY + G for ITS of Digenea).The percentage of divergence between sequences Fig. 1 Localization of Corsica in the Mediterranean Sea and of the sampling site.Diamonds are the major cities in Corsica, the red asterisk is the sampling locality (Padula Reservoir), perch icons are the localities in which the occurrence of Perca fluviatilis has been reported according to Roché and Mattei (1997), Roché (2001), as well as Fleury and Le Mesle ( 2022) (p-distances) was calculated using MEGA X.The phylogenetic tree was constructed with MrBayes v.3.2.6 (Ronquist et al. 2012).Two independent analyses were run for 10 million generations, sampling every 200 generations.The convergence of the two analyses was checked and the tree obtained is a consensus with 10% of the trees discarded as burn-in.Two phylogenetic trees were then reconstructed on the ITS marker (145 sequences, 915 bp) for Digenea and COI marker (14 sequences, 385 bp) for Nematoda.Median-joining networks were constructed for COI and ITS datasets from respectively Nematoda and Digenea using Network v.4.6 (Bandelt et al. 1999).A maximum parsimony algorithm was applied with the criterion "frequency > 1" to simplify the complex branching scheme and generate networks representing the most parsimonious relationships.Genetic diversity indices (haplotype diversity, number of polymorphic sites, and number of haplotypes) were calculated using DnaSP V6 (Rozas et al. 2017).

Results
Following complete parasitological examination, two different parasites were recovered from the sampled P. fluviatilis: Nematoda were morphologically identified as larvae of Eustrongylides sp.(Nematoda: Dioctophymatidae) and Digenea as metacercariae of Clinostomum (Trematoda: Clinostomidae).

Eustrongylides
Eustrongylides larvae were found coiled in or moving through the muscles, in the swim bladder, in the abdominal cavity, and encysted in the liver.They were bright to dark red in color (Fig. 2a, b, c, d).The individuals observed in SEM showed 12 cephalic papillae arranged in 2 circles of 6 papillae each (2 lateral, 2 subdorsal, and 2 subventral with lateral papillae slightly more anterior than subdorsal and subventral papillae, Fig. 2e).Papillae of the inner circle had narrow bases and spike-like apices (Fig. 2f) and papillae of outer circle had broad bases and nipple-like apices (Fig. 2g).The anus was terminal (Fig. 2h).
The phylogenetic tree supported the identification of Corsican specimens as members of the genus Eustrongylides.However, our COI sequences obtained from P. fluviatilis in Corsica clustered with a specimen (GenBank accession number MK013341) recovered from a pikeperch S. lucioperca (Linnaeus, 1758) sampled in Derbent Dam Lake, Turkey (Pekmezci and Bolukbas 2021) and attributed to Eustrongylides excisus Jägerskiöld, 1909 (Fig. 3a).
Our sequences are grouped in five haplotypes and two haplogroups with an average divergence of 2% (Fig. 3b).These results are also in accordance with the ITS identification using BLAST where our sequences matched at 100% with the only available sequence in GenBank with a species name (KU963206) as E. excisus from a Northern pike Esox lucius Linnaeus, 1758 from Freidoonkenar, south of the Caspian Sea (Mazandaran province, Iran) (Youssefi et al. 2023).
The observed prevalence was 90% as only the smallest P. fluviatilis (the only one being around 12 cm long) was found to be uninfected.The mean intensity (minimum-maximum) for Eustrongylides spp. was 29.6 ± 22.7 (3-64).

Clinostomum complanatum
Digenea metacercariae were determined as Clinostomum following Caffara et al. (2011).Clinostomum metacercariae were mainly distributed in the muscle of its hosts (75.4%) and in the opercular cavity and on gill arches (20.3%), and to a lesser extent in the abdominal cavity (2.9%) and the swimbladder (1.4%).SEM observation of excysted Clinostomum metacercariae showed an elongated body, wider in the gonadal region, with a terminal excretory pore (Fig. 4a).The oral sucker was small and surrounded by a prominent oral collar (Fig. 4b).The tegument was covered in spines along the body, sharper in the oral region (Fig. 4c) and more rounded midbody (Fig. 4d).
The ITS phylogenetic tree includes our sequences from Corsica within the Clinostomum complanatum (Rudolphi, 1814) group with an average intrinsic distance of 0.2% (Fig. 5a).The haplotype network highlights that our sequences share the same haplotype (H1) as Italian, Romanian, Iranian, and Chinese specimens (Fig. 5b) and thus seems widely distributed in Eurasia.There was a total Morphological and molecular data corroborate the identification of Digenea parasites as C. complanatum.

Discussion
While molecular phylogeny enabled us to identify C. complatanatum with confidence, the lack of consensus between morphological and molecular data prevents a definitive species diagnosis in the case of Eustrongylides Nematoda.The outer circle papillae being larger than the inner circle papillae and the shape of these papillae are characteristics of Eustrongylides tubifex (Nitzsch in Rudolphi, 1819) according to the key to Eustrongylides third-stage larvae presented in Moravec (1994), while E. excisus shows similar dimensions for inner and outer circle papillae (Measures 1988;Moravec 1994;Gupta 2018;Mazzone et al. 2019).Ambiguity in morphological features may have led to confusion, especially for larval stages which have been reported not to display features allowing a reliable morphological species identification (Moravec and Nagasawa 2018;Mazzone et al. 2019).The non-exhaustiveness of molecular databases is a further issue encountered when studying Eustrongylides as the vast majority of available sequences are identified up to the genus level for these organisms.As previously noted by Shamsi et al. (2023a, b), there is no sequence available for Eustrongylides prevalence in Padula reservoir (90%) is among the highest in P. fluviatilis in Europe, where reported prevalence ranges from 0.6% for E. tubifex and 13.9% for E. excisus in Bulgaria (Lake Srebarna) and 3.3%, 6.8%, and 10.0% for Eustrongylides spp. in Italy (Lake Bracciano, Trasimeno, and San Michele, respectively) to 72.0-100.0%and 94.0-100.0%for E. excisus in Ukraine (Zaporizhya Reservoir) and Moldova (Prut-Dniester interriveran), respectively (Branciari et al. 2016;Menconi et al. 2020b;Honcharov et al. 2022;Maggio et al. 2024).There does not seem to be a clear relation between Eustrongylides prevalence and the type of habitat.However, the identification of Eustrongylides larvae can be problematic and species-specific prevalence variations are not to be excluded.
The distribution of C. complanatum metacercariae mostly in the muscle and gill cavity and arches is consistent with previous reports of this parasite from fish hosts (e.g., Wang et al. 2017;Menconi et al. 2020a).The prevalence of infection was higher than that reported from Italy (14.3-21.4%,Fig. 4 Clinostomum complanatum scanning electron microscopy.a Whole individual with the oral sucker (Os) and collar, ventral sucker (Vs), and excretory pore (Ep) clearly visible.b Cephalic region with the oral sucker (Os), oral collar (Oc), and sensory papillae distributed around the oral sucker.c Tegumental spines in the anterior region and d in the midbody region Lake Endine), Turkey (13.1% and 53.8%, Lake Sığırcı and Gala, respectively), and Czech Republic (15.0%,Morava river basin) (Kadlec et al. 2003;Çolak 2013;Soylu 2014), as well as infection intensities.As already noted by Menconi et al. (2020a), the higher prevalence observed by Soylu (2014) in Lake Gala, compared to other localities, may be attributed to its position on the avian migration route, and thus the attraction of numerous final hosts.In the same way, numerous piscivorous birds may be attracted to Padula reservoir by the nearby Biguglia lagoon, a RAMSAR wetland situated on a bird migration route (MTES 2020).
As Corsica is a Mediterranean island situated on bird migration routes (Bruderer and Liechti 1999;Jourdain et al. 2007;Maggini et al. 2020), the possibility that Eustrongylides sp. and C. complanatum may have been transported with migrating birds before finding suitable fish hosts rather than being co-introduced with piscine hosts cannot be ruled out.It has been reported, for example, that Phalacrocorax carbo has possibly played a key role in the recent expansion of Eustrongylides sp. in Central and Northern Italy (Castiglione et al. 2023).However, the hypothesis of co-introduction through infected fish is also probable.Further investigations are needed to resolve this issue, if indeed it can be resolved.Furthermore, the possibility of an introduction via the first intermediate hosts cannot be ruled out.However, the lack of data concerning eventual introductions of Lymnaeidae and Oligochaeta in Corsica prevents from properly assessing this possibility.
Additionally, infection with Clinostomum or Eustrongylides were recorded in numerous amphibian species (e.g.Yildirimhan et al. 2012;Caffara et al. 2014;León-Règagnon 2019).Seven amphibian species are present in Corsica, among which three endemics.A perspective would be to conduct a parasitological examination of these organisms in order to determine if Clinostomum complanatum and/or Eustrongylides sp. could have been transferred to local amphibians.Indeed, it would be of interest as these organisms may constitute an additional reservoir for these zoonotic parasites.

Risk of food-borne Eustrongylidosis and Clinostomiasis in human
Both parasites are zoonotic as several human cases have been reported, mostly in eastern Asia (Japan, Korea) for Clinostomum and in North America (USA) for Eustrongylides (e.g., Gunby 1982;Wittner et al. 1989;Isobe et al. 1994;Chung et al. 1995;Narr et al. 1996;Hara et al. 2014;Kim et al. 2019).The infection route in human for both parasites is the consumption of either live or raw fish: Eustrongylides infections were caused either by the consumption of live bait by recreational anglers or sashimi made from unfrozen fresh fish (Gunby 1982;Wittner et al. 1989;Narr et al. 1996) and C. complanatum infections by the consumption of raw freshwater or brackish water fish (Isobe et al. 1994;Chung et al. 1995;Hara et al. 2014;Kim et al. 2019).Given the capacity of both C. complanatum and Eustrongylides spp. to infect humans, we recommend that the occurrence of these parasites in the fish of Padula reservoir should be monitored, as prevalence can change over time as was shown in Trasimeno lake, Italy (Franceschini et al. 2022).The cccurrence of the two parasites could only be analyzed in P. fluviatilis but Nematoda matching the description of Eustrongylides were also reported by a recreational angler from S. lucioperca but not from M. salmoides, both species occurring in Padula reservoir.Perca fluviatilis and S. lucioperca are prized by fishermen for their tasty meat and are commonly consumed (Fleury and Le Mesle 2022).Both the local population and tourists should be warned to avoid the raw consumption of both P. fluviatilis and S. lucioperca.In addition, the monitoring should be extended to the other reservoirs where these fish are known to occur.The occurrence of final bird hosts is another component that should be taken into account while monitoring both parasites as the population growth of the P. carbo is believed to have had a major impact on the apparent expansion of Eustrongylides sp. in Italy (Castiglione et al. 2023).

Conclusion
The parasites of the non-native P. fluviatilis were studied for the first time in the French Mediterranean island of Corsica, following the report of a recreational angler.This led to the first observation of two zoonotic parasites, the Nematoda Eustrongylides sp. and the Digenea C. complanatum.The present study is thus an example of how crucial it is that academic research and citizens communicate and work together.Clinostomum complanatum was reliably identified through the use of molecular and morphological study.
Eustrongylides could not be identified at the species level, mainly because molecular databases do not cover all of this genus' species.It is highly likely that the life cycle of both C. complanatum and Eustrongylides can be completed in Corsica as intermediate and final hots seem to be available on the island.The past introductions of numerous non-native species in Corsica may facilitate the completion of these life cycles as well as the maintenance of these parasites in Corsica.The occurrence of C. complanatum and Eustrongylides sp. is concerning from both a veterinary and human health perspective as these parasites can use a wide range of amphibians as intermediate hosts and can be acquired in humans through the consumption of raw or undercooked fish.Even if the consumption of freshly caught wild fish sushi or sashimi is not reported to be a common practice in Corsica, recreational anglers eating their catch and medical practitioners should remain vigilant.Because they potentially interact with numerous components of their ecosystem, in addition to being zoonotic, both Eustrongylides and C. complanatum are an illustration of how interconnected are human, animal, and ecosystem health, as described in the One Health framework.

Fig. 2
Fig. 2 Eustrongylides optic and scanning electron microscopy.a Coiled in and b free and moving through P. fluviatilis muscle.c Exiting a cyst in P. fluviatilis liver.d Exiting through the skin upon its host's death.e Cephalic end showing both inner (arrowheads) and outer (arrows) papillae circles and the mouth.f Spike-like papillae of the inner circle with narrow base (Nb) and spike-like apex (Sa).g Nipple-like papillae of the outer circle with broad base (Bb) and nipple-like apex (Na).h Posterior end showing the terminal anus (A)

Fig. 3
Fig. 3 Phylogenetic tree by Bayesian inference with a GTR + I model based on COI data (385 bp) tree of Eustrongylides, identifying 9 specimens sampled in Padula reservoir, Corsica.Numbers at nodes correspond to posterior probability values.Corresponding host species are indicated when available.No sequences were available for E. tubifex (a).Haplotypes network obtained for Eustrongylides, on the 10 sequences generated in the present study and retrieved from GenBank.Circle size is proportional to the observed haplotype frequencies and red points represent hypothetical haplotypes.Colors highlight specimens' origin (b)

E
. tubifex and very few for E. excisus, which prevents any definite conclusion regarding the identification of the Corsican specimens.The simultaneous study of morphological features and DNA sequences of several Eustrongylides species could enable clarification of this situation.The grouping of the sequences obtained in the present study in two distinct haplogroups (H1 and H2 and H3, H4, and H5) hints at the hypothesis of two origins for this parasite, and possibly two distinct species.The occurrence of both Eustrongylides and C. complanatum are reported here for the first time in France and must be recorded in the French taxonomic register (TAXREF) of the INPN database (TAXREF 2024).

Table 1
Final bird host species reported for Eustrongylides spp.and Clinostomum complanatum from the literature, with information on their presence in Corsica.The presence of species is indicated according to the INPN database