Helminth communities of endemic cyprinoids of the Apennine Peninsula, with remarks on ectoparasitic monogeneans, and a description of four new Dactylogyrus Diesing, 1850 species

The fauna of the Apennine Peninsula is, in comparison to other southern European peninsulas, relatively species-poor regarding the number of endemic cyprinoid species. Nonetheless, the recent introduction of non-native species has significantly increased the total number of freshwater species in this region. Such invasive species may represent a threat to the native fauna, associated among other things with the introduction of non-native parasites with their original hosts.In the present study, we investigated endemic cyprinoid species for the presence of helminth parasites. A total of 36 ectoparasitic monogenean species and five endoparasitic helminth species were collected from ten cyprinoid species in five localities in northern Italy. Out of 20 Dactylogyrus species (gill monogeneans specific to cyprinoids), four were identified as new to science and herein described: Dactylogyrus opertus n. sp. and Dactylogyrus sagittarius n. sp. from Telestes muticellus, Dactylogyrus conchatus n. sp. from T. muticellus and Protochondrostoma genei, and Dactylogyrus globulatus n. sp. from Chondrostoma soetta. All new Dactylogyrus species appear to be endemic to the Apennine Peninsula; however, they share a common evolutionary history with the endemic Dactylogyrus parasitizing cyprinoids of the Balkans. This common origin of cyprinoid-specific parasites supports a historical connection between these two (currently separated) geographical regions.


Introduction
The composition of a parasite community is the result of multiple factors influencing parasitic organisms and their hosts (Poulin, 2007). Besides environmental factors (Galaktionov and Bustnes, 1999;Marcogliese, 2001;Maestri et al., 2017;Clark et al., 2018), the biology of a given host usually has a great impact on the species richness of parasite communities and the abundance of individual parasite species (or in combination with the environment, Berkhout et al., 2020). In general, the species richness of parasites is intimately linked with the species diversity of their hosts (Krasnov et al., 2004;Hechinger and Lafferty, 2005;Thieltges et al., 2011;Kamiya et al., 2014); however, both the range and spatial pattern of distribution of a given host species may affect parasite assemblages, as the resource availability varies across the whole distribution range of the host (Clark et al., 2018;Berkhout et al., 2020). Studying parasite communities and the phylogenetic relationships between parasite species (especially parasite species exhibiting high host-specificity) can shed light on the evolutionary history and historical dispersion of their hosts (e.g. Nieberding et al., 2004;Nieberding and Olivieri, 2007;Verneau et al., 2009). Moreover, an unusual recent distribution of host-specific parasites may serve as a good indicator of the human-induced introduction of invasive species into non-native regions (Benovics et al., 2018;Šimková et al., 2018;Wilson et al., 2019).
The Apennine Peninsula is one of the regions in Europe which has experienced numerous introductions of non-native species over the last few decades (Gherardi et al., 2008;Nocita et al., 2017;Volta et al., 2018). According to Gherardi et al. (2008), approximately 46% of freshwater fish species are non-native and their presence in this peninsula was induced by humans. Moreover, anthropogenic modifications of habitats significantly altered the natural abundance and composition of native fish fauna (Bianco, 1995). Gherardi et al. (2008) also hypothesized that with recent trends in introduction (either intentional to add commercially 'more attractive' fish into Italian rivers, or unintentional via ballast water exchange), the local fauna will become even more homogenized in the future. Currently, the northern Apennine Peninsula is characterized by two main ichthyogeographic districts: Padano-Venetian and Tuscano-Latium; each reflecting the recent distribution of particular endemic species (Bianco, 1990(Bianco, , 1995. The Padano-Venetian district covers the northern Adriatic basin and its faunal elements are shared with the north-western Balkans. The area of this district basically corresponds to the drainage of the Po River during the Last Glacial Maximum, when the sea level drastically regressed (Pielou, 1979;Bianco, 1990;Waelbroeck et al., 2002). During that time, freshwater river streams facilitated the mixing of faunas between two currently isolated regions (i.e. the north-western part of the Balkan, and the north-eastern part of the Apennine peninsulas) (Waelbroeck et al., 2002;Stefani et al., 2004). The Tuscano-Latium district spreads out along the Tyrrhenian slope of central Italy, between the River Arno in the north and the River Tiber in the south, and reflects the distribution of endemic Squalius lucumonis and encompasses almost exclusively endemic species (Bianco, 1990;Kottelat and Freyhof, 2007).
The Cyprinoidei [recently elevated Cyprinidae, and divided into the Leuciscidae and Cyprinidae families, following Stout et al. (2016), Schönhuth et al. (2018) and Tan and Armbruster (2018)] is the most diversified freshwater fish group in the Apennine Peninsula represented by 15 species, of which 12 are endemic. The distributions of three out of five species native to the Padano-Venetian district [namely, Barbus plebejus, Rutilus aula (Bonaparte, 1841) and Squalius squalus] extend up to the north-western Balkans (for the distribution ranges of species, see Kottelat and Freyhof, 2007). The other two species native to this district [i.e. Chondrostoma soetta and Protochondrostoma geneithe latter monotypic genus was only recently erected by Robalo et al. (2007)] are present only in the Apennine Peninsula. In contrast, Telestes muticellus represents a species with a relatively wide distribution range covering both the ichthyogeographical Padano-Venetian and Tuscano-Latium districts. Nonetheless, T. muticellus actually appears to represent a species complex that contains taxa with large genetic diversity (Stefani et al., 2004), putatively linked with the fragmentation of populations into refuges during the last glaciation cycles (Marchetto et al., 2010).
The distribution of the fish fauna, and especially cyprinoids, is well known in the Apennine Peninsula; however, knowledge about their parasite diversity is scarce. In light of the previous discussion, there are two major reasons to investigate the diversity of parasites in a given region: (1) parasites can represent a threat to native fauna (especially co-introduced parasite species), and (2) parasites can serve as a useful tool for investigating the evolutionary history of their hosts. Therefore, in the present study, we aimed to study the species diversity of parasitic helminths and investigate whether the diversity and distribution of the local parasite fauna of cyprinoid fish have been influenced by the introductions of non-native fish species into this region; i.e. that parasite assemblages also include alien parasite species. Moreover, we expect that native endemic cyprinoid fish will also harbour endemic parasite species (especially highly diversified and host-specific monogeneans), and by investigating the phylogenetic relationships of endemic parasites to congeners from different European regions, we will shed new light on the evolutionary history of their hosts.

Parasite sampling, fixation and identification
During the summers of 2015 and 2018, a total of 92 fish specimens belonging to ten species were collected from five localities in northern Italy (Fig. 1) and examined for metazoan parasites. In addition, ten specimens of B. plebejus and six specimens of S. squalus were collected by a local fisherman from the uncertain site on Po River (Table 1). Fish species were determined using Kottelat and Freyhof (2007). Prior to parasitological dissection, live fish were kept in aerated holding barrels with river water from the collection site. Fishes were paralysed and then sacrificed by severing the spinal cord. All applicable institutional, national and international guidelines for the care and use of animals were followed. The study was approved by the Animal Care and Use Committee of the Faculty of Science, Masaryk University in Brno (Czech Republic).
Fish were dissected using the standard method described in Ergens and Lom (1970) and individual organs were observed under a stereomicroscope. Monogenean parasites were collected from fish skin, fins, gills and head cavities. Eyes, brain and muscles were checked for the presence of metacercariae of digeneans. Intestine and abdominal internal organs (i.e. gallbladder, gas bladder, liver and spleen) were checked for the presence of acanthocephalans, cestodes, digeneans and nematodes.
Monogenean parasites (Dactylogyrus spp., Gyrodactylus spp. and Paradiplozoon spp.) were mounted on slides (completely flattened under coverslip pressure) and fixed in a mixture of glycerine and ammonium picrate (Malmberg, 1957). For Dactylogyrus, at least five specimens of each species were cut in half, and one part (usually the half with the attachment organ) was preserved in 96% ethanol for further molecular analyses. The remaining body part was mounted on a slide for species determination by means of morphology. The body half used for the morphological study was later deposited (one per species) as a hologenophore [see Pleijel et al. (2008) for terminology] in the helminthological collection of the Institute of Parasitology of the Czech Academy of Sciences in České Budějovice (IPCAS). Species determination was performed according to the size and shape of the sclerotized elements of the attachment organ and copulatory organs following Pugachev et al. (2009). Endoparasitic helminths were removed from hosts and preserved in 70% ethanol. Prior to fixation in Canada balsam, cestodes and digeneans were stained in iron acetocarmine, following the protocol of Georgiev et al. (1986). Species determination followed Kuchta et al. (2008) and Gibson et al. (2002), respectively. Nematodes were mounted on slides and covered in a mixture of glycerine and water (in the ratio 3:7), and cleared by gradually increasing the volume of glycerol (according to Moravec, 2013). The species determination of nematode parasites followed Moravec (2013). Basic epidemiological data, i.e. prevalence, mean abundance, and minimum and maximum intensities of infection, were calculated for each parasite species (Table 2) according to Bush et al. (1997). Prevalence, as the percentage of fish infected by a given parasite species, and mean abundance, as the mean number of parasite specimens per individual host taking into account both infected and uninfected hosts, were calculated. Following the suggestion of Rózsa et al.  LocID, codes of collection sites corresponding to the codes in Fig. 1; N, number of processed fish specimens; NSP, number of parasite species; dashes (-) indicate that no parasite species of the given taxa was found. Potentially alien species are marked by asterisks (following Galli et al., 2007).
(2000) for interpreting epidemiological data, a confidence interval at the level of 95% was calculated for mean abundance.

DNA extraction, amplification and sequencing
The extraction of genomic DNA from Dactylogyrus parasites was performed using a commercially produced extraction kit (DNeasy Blood & Tissue Kit, Qiagen, Hilden, Germany) following the manufacturer's protocol. Two genomic DNA regions were amplified, the first one covering the partial gene coding for 18S rRNA, the complete ITS1 region and the partial gene coding for 5.8S rRNA (hereinafter abbreviated as 18S, ITS1 and 5.8S), and the second one covering the partial gene coding for 28S rRNA (hereinafter abbreviated as 28S). The first region was amplified using the primers S1 (forward, 5 ′ -ATTCCGATAACGAACGAGACT-3 ′ ) and IR8 (reverse, 5 ′ -GCTAGCTGCGTTCTTCATCGA-3 ′ ), which anneal to the segments of DNA coding for 18S and 5.8S rRNA, respectively (Šimková et al., 2003). The amplification reaction followed the protocol optimized by Benovics et al. (2018). The latter region was amplified using the primers C1 (forward, 5 ′ -ACCCGCT GAATTTAAGCA-3 ′ ) and D2 (reverse, 5 ′ -TGGTCCGTGTTTC AAGAC-3 ′ ) (Hassouna et al., 1984), following the protocol of Benovics et al. (2020). The PCR products (∼1000 for 18S, ITS1 and 5.8S, and ∼800 bp for partial 28S) were checked on 1% agarose gel and purified using the ExoSAP-IT kit (Ecoli, Bratislava, Slovakia) following the standard protocol. The purified products were directly sequenced using the same primers as for PCR and BigDye Terminator Cycle Sequencing kit (Applied Biosystems, Prague, Czech Republic). Sequencing was performed on an ABI 3130 Genetic Analyzer (Applied Biosystems).

Morphometric data and species description
The mounted monogeneans (or their parts) were studied using an Olympus BX 61 microscope equipped with phase-contrast optics. Drawings were made with the aid of a drawing attachment and edited with a graphic tablet compatible with Adobe Illustrator and Adobe Photoshop. Measurements were taken using digital image analysis (Stream Motion, version 1.9.2) and are given in micrometres [mean followed by the range and number of specimens measured (n) in parentheses]. The measurement scheme for the sclerotized structures of Dactylogyrus is shown in Fig. 2. The numbering of hook pairs (in Roman numerals) follows that recommended by Mizelle (1936). The male copulatory organ is abbreviated as MCO.
The parasitological material was deposited in the Helminthological collection of the Institute of Parasitology, Academy of Sciences of the Czech Republic, České Budějovice, Czech Republic (IPCAS). The material was remounted in Canada balsam (according to Ergens, 1969)

Phylogenetic analyses
To infer the phylogenetic positions of new Dactylogyrus species, three genetic markers were selected for phylogenetic analyses; partial genes coding for 18S rRNA and 28S rRNA, and the ITS1 region. In addition to the newly obtained sequences, sequences corresponding to already described Dactylogyrus species were retrieved from GenBank (see accession numbers in the Supplementary data, Table S1). The Dactylogyrus species for phylogenetic analyses were selected in order to investigate the phylogenetic relationships between species from two southern European peninsulas with a historical connection (i.e. the Apennine and Balkan peninsulas), and their relationships to Dactylogyrus species parasitizing cyprinoids with a wide distribution range in Europe, especially in the geographically proximal central Europe. For Dactylogyrus species found on cyprinoids of the Apennine Peninsula and previously reported on cyprinoid species in other European regions (the Balkans or Central Europe, Šimková et al., 2004;Benovics et al., 2017Benovics et al., , 2018Řehulková et al., 2020), already published DNA sequences were retrieved from GenBank. DNA sequences were concatenated and aligned using the fast Fourier transform algorithm implemented in MAFFT (Katoh et al., 2002) and subsequently manually trimmed to unify the lengths of all used sequences. The final alignment was built from 42 Dactylogyrus species (38 previously described, and 4 new) of various cyprinoid taxa from the Apennine and Balkan peninsulas, and central Europe. Following the phylogenetic reconstruction of Benovics et al. (2018), three common Dactylogyrus species of Carassius gibelio (Bloch, 1782) (namely D. anchoratus, D. formosus and D. vastator) were used as the outgroup for rooting the phylogenetic trees. The aligned dataset was treated as partitioned, each partition associated with an individual genetic marker, i.e. 18S, 28S, ITS1. General time-reversible model was applied to each partition. Phylogenetic analyses were computed by means of maximum likelihood (ML) and Bayesian inference (BI) methods, using RaxML v 8.2.11 (Stamatakis, 2006(Stamatakis, , 2014 and MrBayes v 3.2.6 (Ronquist et al., 2012), respectively. Internal node support for the ML tree was assessed by running 1000 bootstrap pseudoreplicates. Two parallel runs, each with four Markov chains, were executed for the BI analysis and run for 10 7 generations. Trees were sampled every 10 2 generations and the first 30% of the resulting trees were discarded as initial burn-in after checking convergence of runs in Tracer v 1.7.1 (Rambaut et al., 2018). Posterior probabilities were calculated as the frequencies of samples recovering particular clades.

Diversity of helminth parasites
A total of 41 helminth species were collected and identified (Table 2), with D. vistulae as the most prevalent parasite species, and G. lomi as the most abundant species. Parasites communities with the greatest species richness were harboured by S. lucumonis from the Torrente Cerfone locality (11 helminth species). In contrast, only three helminth species were collected from C. soetta (Carmagnola). The basic epidemiological parameters, i.e. prevalence, mean abundance with a confidence interval at the level of 0.95, and minimum and maximum intensity of infection, are presented for each parasite species in Table 2. In general, monogeneans exhibited the highest species diversity. Thirty-one known monogenean species and four newly-described species were found on the gills and in the nasal cavity of fish. Concerning Dactylogyrus, out of 20 species, D. vistulae was reported on the widest range of host species (collected from six host species) exhibiting the highest prevalence and highest mean abundance on S. squalus. The second most abundant monogenean genus, Gyrodactylus, was represented by 14 species, of which G. gobii and G. laevis were collected from the highest number of host species (four species). Only one diplozoid species (Paradiplozoon megan) was found, parasitizing all four species collected from Torrente Cerfone, with the highest prevalence on S. lucumonis. However, P. megan achieved the highest mean abundance and intensity of infection on P. genei. Only one species of eye fluke (metacercariae of Diplostomum spathaceum) was found on two specimens of T. muticellus from Torrente Cerfone with moderate abundance. No other digeneans were detected during this survey. Intestinal helminths were represented by three species of nematodes (Molnaria intestinalis, Pseudocapillaria tormentosa and Rhabdochona hellichi) and a single cestode species belonging to the monotypic genus Bathybothrium (Bathybothrium rectangulum) collected from two Barbus species (B. caninus and B. tyberinus). However, only a single specimen of B. rectangulum was collected from B. tyberinus.

Morphological and molecular characterization of the new species
The new Dactylogyrus species described in this study were classified into two morphological groups according to Pugachev et al. (2009) on the basis of the morphology of the sclerotized parts of the reproductive organsthe MCO and the vagina.
Dactylogyrus ergensi, collected from Chondrostoma nasus in the Danube River in Hungary, was described by Molnár (1964). Since then, specimens from several distant localities displaying the same basic morphology but also high morphometrical variability in their sclerotized structures have been assigned to this species (Pugachev et al., 2009); we consider these specimens likely representing the members of more than one species. The morphology of the sclerotized structures of the present new species was compared with that of D. ergensi using an original description of the latter species (Molnár, 1964).
The last of the four Dactylogyrus species described in this study (i.e. D. opertus n. sp.) was assigned to the 'nanus' morphological group, specifically to the subgroup of species displaying the following features: an MCO comprising a simple copulatory tube, smoothly curved all along its length from the initial part (or more curved distally), and terminally enclosed with a markedly enlarged distal part (a sheath) of the accessory piece; a vagina present as a short or medium-size bent cylindrical tube, closed with a cap (i.e. the subgroup including Dactylogyrus nanus Dogiel et Bychowsky, 1934;D. nanoides Gussev, 1966;and D. suecicus Nybelin, 1937 Representative DNA sequences: A nucleotide sequence of partial gene for 28S rRNA (789 bp long; MW443032), and nucleotide sequences representing a fragment (961 bp long; MW443035) including partial gene for 18S rRNA (487 bp), and the ITS1 region (474 bp). No intraspecific variability was found (seven specimens were analysed).
Etymology: The specific name is derived from Latin (globus = ball, sphere, globe; treated as an adjective) and refers to the shape of the distal part of the MCO accessory piece which seems to be rolled into a ball.
Remarks: Dactylogyrus globulatus n. sp. is similar to Dactylogyrus species belonging to the 'ergensi' morphological group (Pugachev et al., 2009) (see above), including D. conchatus n. sp. and D. sagittarius n. sp. The new species most closely resembles D. ergensi and D. conchatus n. sp.the species possessing a basically cross-shaped ventral bar. While the posterior arm ventral bar is reduced in the new species and D. conchatus n. sp., it is apparently more developed in D. ergensi [according to the drawing of Molnár (1964)], separating the new species, and also D. conchatus n. sp., from all other species of the 'ergensi' group.
On the basis of the comparison with D. ergensi, originally described (and depicted) by Molnár (1964), the present new species can be distinguished from the latter by the following characters: (i) dorsal anchor with shaft evenly curved to point, and The differentiation of D. globulatus n. sp. from D. conchatus n. sp. and D. sagittarius n. sp. is provided in the differential diagnosis for the latter two new species.
Etymology: The specific epithet is derived from Latin (concha = mollusc shell; treated as an adjective) and reflects the snail shell appearance of the distal potion of the MCO accessory piece.
According to the original description (and depiction) of D. ergensi by Molnár (1964), D. conchatus n. sp. can be distinguished from the latter species by the following characters: Dactylogyrus conchatus n. sp. can be separated morphologically from D. globulatus n. sp. on the basis of the following characters: (i) dorsal anchor possesses well-differentiated point (vs less differentiated point, with a slightly recurved tip in D. globulatus n. sp.), (ii) dorsal bar is more slender, (iii) anterior arm of ventral bar is widened terminally. Moreover, the D. conchatus n. sp. is also divergent from other congeners at the molecular level.
Differences between D. conchatus n. sp. and D. sagittarius n. sp. are specified in the differential diagnosis for the latter species.
Representative DNA sequences: A nucleotide sequence of partial gene for 28S rRNA (789 bp long; MW443034) and nucleotide sequences representing a fragment (966 bp long; MW443037) including partial gene for 18S rRNA (487 bp), and the ITS1 region (479 bp). No intraspecific variability was found (six specimens were analysed).
Etymology: The specific name, a noun, is from Latin (Sagittarius = the archer armed with arrow), referring to the new species having an arrow-shaped vagina.
Remarks: Dactylogyrus sagittarius n. sp. resembles Dactylogyrus species belonging to the 'ergensi' morphological group (Pugachev et al., 2009) (see above), including D. globulatus n. sp. and D. conchatus n. sp. The presence of a three-armed ventral bar, together with the absence of a membranous skirt around the margin of the copulatory tube base, allows this species to be distinguished from D. ergensi, D. globulatus n. sp. and D. conchatus n. sp. Dactylogyrus sagittarius n. sp. differs from all the species of the 'ergensi' group by the shape of the dorsal anchor (i.e. an anchor having an elongate evenly curved shaft and well-differentiated but very short point).
Etymology: The specific name is derived from Latin (opertus = hidden, secret) and refers to the distal part of the copulatory tube enclosed by (covered with) an accessory piece.
Remarks: Dactylogyrus opertus n. sp. was initially included in the study by Benovics et al. (2020) as Dactylogyrus sp. 7. Dactylogyrus opertus n. sp. was classified into the 'nanus' morphological group, showing a close resemblance to D. nanus, D. nanoides and D. suecicus (Pugachev et al., 2009) (see above). However, the new species differs from all species of the 'nanus' group mentioned above by possessing a cross-shaped ventral bar [the ventral bar is three-armed in D. nanus, D. nanoides and D. suecicus (the anterior arm divided into horn-like ends in D. suecicus)]. Further, the D. opertus n. sp. accessory piece of the MCO in D. opertus n. sp. is distinctive, having a bluntly finished terminal part, enlarged to form a sheath directed backwards and enclosing (covering) the distal part of the copulatory tube (in D. nanus, D. nanoides and D. suecicus, the distal part of the accessory piece is simply enlarged to form a sheath supporting The presence of a cross-shaped ventral bar makes D. opertus n. sp. similar to other species of the 'nanus' group, i.e. to D. rutili Glaeser, 1965 andD. distinguendus Nybelin, 1937. However, these two species, in contrast to the new species, belong to the species of the 'nanus' group having a proximally sharply curved copulatory tube and an elongate (S-or C-shaped) vagina (Pugachev et al., 2009). While in D. opertus n. sp., the posterior arm of the ventral bar is of similar length to (or longer than) the transverse arm in D. opertus n. sp., the ventral bar in D. rutili and D. distinguendus has a short posterior arm in D. rutili and D. distinguendus (i.e. a posterior arm shorter than the transverse arm; Pugachev et al., 2009). Dactylogyrus opertus n. sp. also resembles D. sandai Řehulková, Benovics et Šimková, 2020, a species recently described from Balkan cyprinoids and displaying close similarity with respect to the basic morphology of both the haptoral and reproductive structures (Řehulková et al., 2020). The anchors of the new species possess slightly more elongated points, and roots differ even more in their lengths. The anterior arm of the connective ventral bar in D. opertus n. sp. is also wider in comparison to D. sandai. Moreover, the terminal part of the MCO of D. opertus n. sp. forms a sheath pointing backwards and covering the distal part of the copulatory tube, while a similar structure in D. sandai does not enclose the copulatory tube but points forward.

Phylogenetic position of the newly described Dactylogyrus species
The final concatenated alignment of the DNA sequences of the 42 Dactylogyrus species studied here spanned 1787 unambiguously aligned nucleotide positions (446 bp for 18S, 704 bp for 28S and 637 bp for ITS1). Both phylogenetic analyses (ML and BI) yielded trees with identical topologies, differing only in their node support values (see tree from BI in Fig. 7).
The phylogenetic analyses revealed that the newly-described Dactylogyrus species are associated with two phylogenetic lineages. Dactylogyrus opertus n. sp. is phylogenetically close to the recently described D. sandaian endemic species of the Balkan Peninsulaand nested within a clade containing the morphologically similar species D. nanus, D. rutili and D. suecicus. All other new species, i.e. D. conchatus n. sp., D. globulatus n. sp. and D. sagittarius n. sp., are nested within the phylogenetic lineage that includes species with the 'ergensi' type of copulatory organ (i.e. D. ergensi, D. dirigerus, D. caucasicus and D. tissensis).

Discussion
The species richness of a parasite community in fish usually reflects the effects of numerous abiotic factors related to the environment, the biology of the host and the diversity of the communities of syntopic organisms serving as intermediate hosts for many fish endoparasites (e.g. Gregory, 1990;Landsberg et al., 1998;Poulin, 2007). One of the less obvious factors is the introduction of non-native fish into a new region, as introduced fish may also serve as vectors for the co-introduction of non-native parasites (especially monogeneans, e.g. Havlátová et al., 2015;Benovics et al., 2017Benovics et al., , 2018Benovics et al., , 2020Šimková et al., 2018;Kmentová et al., 2019;Wilson et al., 2019), which may represent a substantial threat to native host populations (Lymbery et al., 2014). Since current Italian freshwater fauna includes numerous non-native species (Bianco, 1995), it is tempting to expect that non-native parasites, co-introduced with their non-native host species, will also be present on endemic cyprinoids.
In the present study, we investigated ten species of endemic Italian cyprinoids from six sites in the northern part of the Apennine peninsula for their helminth parasites. The highest parasite diversity was observed among monogeneans, which are a group of ectoparasitic platyhelminths usually parasitizing either on the gills (Gyrodactylus, Dactylogyrus and Paradiplozoon), or on the skin (Gyrodactylus) of fish. A total of 20 Dactylogyrus species were identified, exceeding the number of species reported in the checklist of Galli et al. (2007). Herein, we reported 15 Dactylogyrus species that were not previously documented in Italy. Historically, Italian cyprinoids have not been comprehensively investigated for the presence of monogenean parasites. Before 2002, only eight monogenean species had been reported (Molnár and Ghittino, 1977;Bona et al., 1995). Some efforts to investigate the diversity of monogeneans came after the studies of Galli et al. (2002Galli et al. ( , 2005Galli et al. ( , 2007; however, endemic species were tackled only marginally, since studies were focused on  Benovics et al. (2017)]. All three parasite species are host-specific to C. carpio and C. gibelio across their whole distribution range (e.g. Moravec, 2001;Galli et al., 2002;Jalali and Barzegar, 2005;Šimková et al., 2006, 2013Ling et al., 2016). Their presence on B. plebejus, an endemic species of the Padano-Venetian ichthyogeographical district (sensu Bianco, 1990), may be a result of the introduction of the abovementioned alien species into the Apennines and the subsequent host-switch to endemic and phylogenetically close Barbus (all host species belonging to Cyprinidae). The presence of non-native D. vastator on the endemic Balkan Aulopyge huegelii Heckel, 1843 was previously reported by Benovics et al. (2017) and also explained by the host switch from C. carpio or C. gibelio. However, the alien Dactylogyrus species were not observed on B. tyberinus; therefore, we can postulate that host-switching occurs only in water bodies where congeneric Barbus lives in sympatry with introduced cyprinids, such as the Po River, where C. carpio and C. gibelio are commonly distributed (Galli et al., 2007;Kottelat and Freyhof, 2007).
The most common Dactylogyrus species on endemic cyprinoids was D. vistulae (present on 60% of investigated species), which is considered as a true generalist species with the widest host range in Europe (Moravec, 2001;Šimková et al., 2006;Benovics et al., 2018Benovics et al., , 2020. Herein, we present B. plebejus as a new host of D. vistulae, increasing its host range to 34 cyprinoid species. In addition to already known Dactylogyrus species, we reported and described four new species from Apennine leuciscids. While D. globulatus n. sp., D. opertus n. sp. and D. sagittarius n. sp. appear to be host-specific to C. soetta (D. globulatus n. sp.) and T. muticellus (the latter two species), D. conchatus n. sp. was reported from two leuciscid species -T. muticellus and P. genei. The morphological similarity of D. globulatus n. sp., D. conchatus n. sp. and D. sagittarius n. sp. in the shape of their copulatory organs, and the fact that the only difference among them is in the shape of their haptoral elements suggest their common evolutionary origin (illustrations of the sclerotized structures directly taken from hologenophores are shown for comparison in Fig. 8). These endemic species most likely share a common ancestor with other species of the 'ergensi' morphological group, and evolved in the Apennines by subsequent intrahost duplication on T. muticellus (or its ancestor). The common origin of these three species and other species from the 'ergensi' morphological group (i.e. D. caucasicus, D. dirigerus, D. ergensi and D. tissensis) is also supported by our phylogenetic analyses. Endemic species most likely diverged and underwent speciation after the dispersion of their ancestor into the Apennine Peninsula. The differences in the haptoral elements might be adaptations for attaching to different substrates on different parts of the gill arch (i.e. microhabitats) of the same host species after intra-host speciation [morphological adaptations in the forms of monogenean haptors were suggested by Rohde (1979) and Šimková et al. (2000, 2002)]. A different evolutionary origin may be postulated for D. opertus n. sp., which more resembles, on the basis of morphology, species parasitizing on Squalius, e.g. D. nanus, D. nanoides and D. suecicus (see Pugachev et al., 2009 for morphotypes). Dactylogyrus opertus n. sp. [initially included as Dactylogyrus sp. 7 in the phylogenetic study by Benovics et al. (2020)] was found to be phylogenetically close to recently described D. sandai [identical with Dactylogyrus sp. 6 included in the study of Benovics et al. (2020)], a species from Telestes in the Balkans. The study of Benovics et al. (2020) also included a third species from Telestes (Dactylogyrus sp. 8 from T. montenigrinus), which has not yet been described; according to their study, all three Dactylogyrus species have a common evolutionary origin. Moreover, the present study confirmed the close morphological similarity between D. sandai and D. opertus n. sp. We can assume that these two species (probably also Dactylogyrus sp. 8a species of clade 6 in Fig. 1, Benovics et al., 2020) are genus-specific to Telestes (more specifically, they parasitize T. karsticus and T. muticellus, respectively). The two Telestes species are recently distributed over the Padano-Venetian ichthyogeographic district and, according to Buj et al. (2017), they share a common ancestor in the Balkans, which dispersed into the northern part of the peninsula and subsequently diverged into the Apennines. Therefore, we can assume that the evolutionary histories of host-specific D. sandai and D. opertus n. sp. are intimately linked with their Telestes hosts and that these Dactylogyrus species originated from cospeciation.
A total of 14 Gyrodactylus species were reported in this study, of which the most prevalent were G. katharineri and G. laevis. Both species exhibit a rather wide host range without obvious specificity to some hosts (each of them was reported from four Apennine cyprinoid species). They can be harboured by cyprinids, leuciscids and also gobiids (e.g. Gutiérrez-Galindo and Lacasa-Millán, 2001;Moravec, 2001;Harris et al., 2004;Djikanovic et al., 2012). Gyrodactylus cf. markewitschi was isolated from Barbus caninus from the Malezzo River. This species resembles G. markewitschi, which appears to be specific to the western Palearctic Barbus (Kakacheva-Avramova, 1973; Moravec, 2001;Harris et al., 2004), with some accidental infections reported from percids (Angelescu, 1974), or carp (Hussain et al., 2007). Therefore, G. cf. markewitschi might represent a species derived from G. markewitschi, endemic for the Apennine Peninsula. Unfortunately, to validate our hypothesis of the evolutionary proximity of these two species, molecular data for G. markewitschi would be required; however, these are still missing. The only locality where no Gyrodactylus parasites were recorded was Carmagnola, comprising gravel pits of relatively recent origin, filled with water from the Po River. Potential changes in abiotic factors from the river stream to the gravel pits (e.g. water current or temperature) may putatively have influenced the populations of gyrodactylid parasites and facilitated their reduction. For example, water temperature has a significant influence on parasite communities; therefore, the different water temperatures at the different sites in this study most probably influenced gyrodactylid communities to different degrees. Temperature directly affects the reproduction and survival time of Gyrodactylus (Jansen and Bakke, 1991;Bakke et al., 2007); during summer months, the intensity of gyrodactylids on hosts tends to decline (e.g. Koskivaara et al., 1991;Blažek et al., 2008). The water in gravel pits is captured from the Po River but, due to its stationariness, reaches a higher temperature than the running water of the River, which (considering studies above) might impact the parasite communities.
This study provides evidence that phylogenetically-close cyprinoid species from the same locality share generalist monogenean species. For example, all leuciscid species from Torrente Cerfone (i.e. P. genei, R. rubilio, S. lucumonis, S. squalus and T. muticellus) were parasitized by D. vistulae. Moreover, all abovementioned leuciscids, except T. muticellus, were also hosts of P. megan, suggesting that this diplozoid species does not exhibit host specificity in the investigated region, but rather a geographical specificity, which is in concordance with the previous reports of P. megan in cyprinoids (Benovics et al., 2021).
The fauna of intestinal helminths in the herein investigated cyprinoid species was species poorer in comparison to ectoparasitic monogeneans, and included only one species of eye fluke (metacercariae of D. spathaceum), three species of nematodes and one species of cestode. Among them, the most remarkable record is B. rectangulum, a tapeworm species from both investigated Barbus spp. from two geographically distant localities. While this representative of the monotypic genus Bathybothrium is relatively common in Barbinae distributed in the Palearctic region (Lühe, 1910;Joyeux and Baer, 1936;Protasova, 1977;Moravec and Amin, 1978;Van Maren, 1979;Dubinina, 1987;Scholz, 1989;Moravec, 2001), prior to this date there was no evidence of this cestode species in the Apennine Peninsula. The common host of B. rectangulum is B. barbus (e.g. Scholz and Moravec, 1996;Moravec, 2001), the Barbus species with the widest distribution range in Europe (Kottelat and Freyhof, 2007). Bathybothrium rectangulum reached relatively high prevalence and abundance in B. barbus (e.g. Laimgruber et al., 2005;Chunchukova and Kirin, 2018), and this widely distributed host species may have served as a potential vector for the distribution of B. rectangulum into different European regions. A more thorough ichthyoparasitological investigation in other southern European Peninsulas may potentially reveal other new hosts for this remarkable parasite species.

Conclusions
Even though the Apennine Peninsula represents a region of high interest regarding ichthyological research, from the parasitological point of view, it remains underexplored, especially with respect to the parasites of endemic fish species. After more than a decade, we present the first study focused on the parasite fauna of endemic cyprinoids in northern Italy, and described four new monogenean species. We also found that local parasite communities appear to be influenced by the continual introduction of nonnative fish into the region, potentially threatening native endemic species that are already undergoing reductions in their population sizes.
Supplementary material. The supplementary material for this article can be found at https://doi.org/10.1017/S0031182021000615.
Data. The data supporting the conclusions of this study are included in this article. The type-material of the new species described in this study was deposited in the Helminthological Collection of the Institute of Parasitology, Czech Academy of Sciences, České Budĕjovice, Czech Republic under the accession number IPCAS M-753-M-756. The newly generated sequences were submitted to the GenBank database under accession numbers MW443031-MW443037.