High parasite diversity in a neglected host: larval trematodes of Bithynia tentaculata in Central Europe

Bithynids snails are a widespread group of molluscs in European freshwater systems. However, not much information is available on trematode communities from molluscs of this family. Here, we investigate the trematode diversity of Bithynia tentaculata , based on molecular and morphological data. A total of 682 snails from the rivers Lippe and Rhine in North Rhine-Westphalia, Germany, and 121 B . tentaculata from Curonian Lagoon, Lithuania were screened for infections with digeneans. In total, B. tentaculata showed a trematode prevalence of 12.9% and 14%, respectively. The phylogenetic analyses based on 55 novel sequences for 36 isolates demonstrated a high diversity of digeneans. Analyses of the molecular and morphological data revealed a species-rich trematode fauna, comprising 20 species, belonging to ten families. Interestingly, the larval trematode community of B. tentaculata shows little overlap with the well-studied trematode fauna of lymnaeids and planorbids, and some of the detected species ( Echinochasmus beleocephalus and E. coaxatus ) constitute first records for B. tentaculata in Central Europe. Our study revealed an abundant, diverse and distinct trematode fauna in B . tentaculata , which highlights the need for further research on this so far understudied host – parasite system. Therefore, we might currently be underestimating the ecological roles of several parasite communities of non-pulmonate snail host families in European fresh waters. (41); width: 58 82 (71) vs. 37 56 (45)], pharynx [length: 17 22 (19) vs. 19 – 31 (23); width: 14 – 22 (17) vs. 25 – 31 (28)] and larger tail [length: 454 – 495 (476) vs. 324 – 342 (336); width: 72 – 82 (76) vs. 43 – 56 (49)]. The above comparisons demonstrate that Psilostomidae gen. sp. 1 may represent a species of the genus Psilotrema , but is not conspecific with P . oligoon , P . simillimum or P . spiculigerum . Further identification of Psilostomidae gen.


Introduction
With about 25,000 species and a cosmopolitan distribution, digenetic trematodes constitute one of the most diverse and ubiquitous groups of parasites on the planet (Esch et al., 2002). Despite their complex life history with a wide variety of vertebrate definitive hosts, including fish, amphibians, reptiles, birds and mammals, this group shares a unifying character: practically all species require molluscs (usually gastropods) as first intermediate hosts. Due to their complex interaction with their hosts and their wide distribution and abundance, trematodes have been studied in a wide range of ecological contexts. For example, trematodes have been shown to make up a large proportion of an ecosystem's biomass (Kuris et al., 2008;Preston et al., 2013;Soldánová et al., 2016), contribute significantly to the energy flow within ecosystem (Thieltges et al., 2008), function as structuring forces in food webs (Lafferty et al., 2008;Thieltges et al., 2013) and can affect host populations by influencing host mortality, fecundity, growth and behaviour (Mouritsen & Jensen, 1994;Marcogliese, 2004;Lagrue & Poulin, 2008;Rosenkranz et al., 2018). Moreover, they can serve as useful bioindicators to assess environmental conditions and changes due to their intricate life cycles (e.g. Lafferty, 1997;Huspeni & Lafferty, 2004;Vidal-Martínez et al., 2010;Shea et al., 2012;Nachev & Sures, 2016). Altogether, there is increasing awareness that trematodes are important ecosystem components that require our attention in order to fully understand the complex interactions and dynamics in ecosystems.
In freshwater systems, snails of the families Lymnaeidae and Planorbidae play a key role in the life cycle of trematodes. In Europe, members of both families serve as important first intermediate hosts to a wide variety of digenean trematodes, with 87 and 92 described species, respectively, which accounts for more than 85% of the described trematode species from gastropod hosts from this region (Faltýnková et al., 2016;Schwelm et al., 2018). Both groups are well-studied model host-parasite systems in terms of their diversity, ecological function and their role as infectious agents (e.g. Faltýnková & Haas, 2006;Faltýnková et al., 2007Faltýnková et al., , 2008Faltýnková et al., , 2016Soldánová et al., 2010Soldánová et al., , 2013Soldánová et al., , 2017Novobilský et al., 2014;Horák et al., 2015;Selbach et al., 2015a, b). Consequently, detailed identification keys (Faltýnková et al., 2007(Faltýnková et al., , 2008Selbach et al., 2014) as well as accessible molecular vouchers (e.g. Georgieva et al., 2013Georgieva et al., , 2014Selbach et al., 2014Selbach et al., , 2015bZikmundová et al., 2014;Soldánová et al., 2017) are available for these parasite taxa, which enable studies to accurately identify trematodes and asses their ecological role. Moreover, the life cycles of many trematodes are known in detail (Brown et al., 2011 and references therein), which allows conclusions to be drawn about the presence or absence of free-living organisms in ecosystems (Byers et al., 2010).
In contrast to the detailed knowledge about lymnaeid and planorbid host-trematode systems, the role of bithyniid snails and other non-pulmonate snails, such as Hydrobiidae, Melanopsidae, Neritidae, Valvatidae and Viviparidae, as first intermediate hosts for trematodes in Central Europe remains largely unexplored. Although snails of these families are known to host digenean trematodes and these data were included in some faunistic surveys (e.g. Cichy et al., 2011;Faltýnková et al., 2016), these studies were mostly focused on lymnaeid and planorbid hosts. The faucet snail, Bithynia tentaculata, which is common and widespread throughout Europe, has established itself as a non-indigenous species in North America (Mills et al., 1993;Duggan et al., 2003;Bachtel et al., 2019). Bithynia tentaculata is highly tolerant towards salinity and temporal droughts and occurs in most waterbodies throughout Europe (Glöer, 2002;Welter-Schultes, 2012). With 32 trematode species according to Cichy et al. (2011) and 14 species according to Faltýnková et al. (2016) known from B. tentaculata, it represents the most species-rich snail-parasite assemblage among the group of non-pulmonate freshwater snails (formerly known as 'Prosobranchia') in Europe. However, due to the lack of focussed faunistic studies on trematode communities in this host species, with the exception of individual studies investigating selected parasite groups (e.g. Serbina, 2005;Kudlai et al., 2015Kudlai et al., , 2017, the number of trematode species known from B. tentaculata may well grossly underestimate the true diversity of this hostparasite system. A serious obstacle for most freshwater ecologists and parasitologists encountering trematodes of the snail families Bithynidae, Hydrobiidae, Melanopsidae, Neritidae, Valvatidae and Viviparidae is the lack of morphological and molecular information for species identification. Unlike for planorbid and lymnaeid snails, there are no keys for larval trematodes parasitizing snails from these families, and morphological descriptions are often restricted to adult stages (e.g. Gibson et al., 2002;Tkach et al., 2003;Jones et al., 2005;Bray et al., 2008;Besprozvannykh et al., 2017;Kudlai et al., 2017). Moreover, existing literature is often not available in English (e.g. Našincová, 1992;Ataev et al., 2002;Serbina, 2005;Besprozvannykh, 2009), which also exacerbates the investigation of this parasite-host system. These obstacles lead to a further bias towards well-studied species, such as Lymnaea stagnalis, Radix spp. and Planorbarius corneus, while other snail species continue to remain overlooked and avoided in the assessments of the ecological role of trematodes. It is, therefore, important to characterize the trematode fauna in B. tentaculata, and thus facilitate further studies on the ecological role of this host-parasite system, as is possible for lymnaeid and planorbid snails. Moreover, some of the trematodes utilizing Bithynia spp. are important pathogens of wildlife that can have serious impacts on migrating birds (e.g. Herrmann & Sorensen, 2009;Roy & St-Louis, 2017;Bachtel et al., 2019), which further highlights the need for a better understanding of this host and its parasite fauna.
Here, we assess the diversity of the larval trematodes of B. tentaculata in Central Europe and provide molecular and morphological reference material to fill this gap in our knowledge. With this study, we also hope to draw more attention to this essential and largely overlooked parasite-host system and promote further studies on this group.
All snails were collected with strainers or hand-picked from sediments, stones, macrophytes and floating vegetation from the riverside or along the littoral zone of the pond. In the laboratory, snails were placed in individual cups with filtered river water at 20°C and exposed to a light source to induce the emergence of cercariae. Each cup was screened for the presence of cercariae three times over three consecutive days after sampling under a stereomicroscope. Snails that did not shed cercariae during this time period were dissected and examined for prepatent infections (rediae/sporocysts). To obtain isolates for molecular analyses, cercariae, rediae and sporocysts were pooled from one single infected snail host and fixed in molecular-grade ethanol. Additionally, cercariae were fixed in 4% formaldehyde solution for measurements of fixed material. For documentation and measurements of the snail hosts, photomicrographs of the snail shell were taken with a Keyence VHX5000 microscope (Osaka, Japan). Foot tissue from infected snails was fixed in molecular-grade ethanol for molecular analysis and identification.

Morphological analyses
Larval stages were preliminarily identified under an Olympus BX51 microscope (Tokyo, Japan) using morphological descriptions of Našincová (1992) and Bykhovskaya-Pavlovskaya & Kulakova (1971) and other relevant publications (e.g. Heinemann, 1937;Zdun, 1961;Našincová & Scholz, 1994;Kudlai et al., 2015). Preliminary identification was made to the family or genus level. Morphology of cercariae was studied on live and fixed specimens. Series of photomicrographs were taken for collected isolates with an Olympus UC30 digital camera (Tokyo, Japan) for measurements and further identification. Measurements were taken from the digital images using cellSens 1.16 Life Science image software (https://www.olympus-lifescience. com/en/software/cellsens). Measurements are in micrometres (μm) and are presented as a range, followed by a mean in parentheses.

Molecular sequencing
DNA isolation was performed following a modified salt precipitation protocol after Sunnucks & Hales (1996) and Grabner et al. (2015). To each sample, 600 µl TNES Buffer and 10 µl proteinase K solution were added. Trematode samples were incubated at 50°C for two to three hours depending on the quantity of the sample material. Snail tissue samples were incubated overnight at 35°C. In order to precipitate the proteins, 170 µl of 5 M sodium chloride was added, followed by vortexing and centrifuging for 5 min at 20,000×g at room temperature. The supernatant was transferred into a new reaction tube and centrifuging was repeated. The supernatant was again transferred into a new reaction tube, the pellet was discarded and 800 µl of 99% ice-cold ethanol was added to the supernatant and mixed by repeated inverting. The solution was centrifuged at 20,000×g for 15 min at 4°C. In order to purify the sample, 180 µl of 70% ethanol was added after the supernatant was discarded. The sample was centrifuged for 15 min at 20,000×g at 4°C, the ethanol was discarded and the pellet air-dried. The DNA pellet was dissolved in 100 µl TE buffer.
Target gene fragments were chosen based on preliminary identification of cercariae to the family level and were amplified via polymerase chain reaction (PCR) (table 1) following the corresponding protocols (Folmer et al., 1994;Cribb et al., 1998;Galazzo et al., 2002;Kostadinova et al., 2003;Tkach et al., 2003). Tissue of snails was also used for DNA extraction and PCR amplification using the primers and protocols by Folmer et al. (1994). PCR products were purified using purification kits (GATC Biotech, Constance, Germany). The original PCR primers and the internal primers for 28S were used for sequencing (table 1). Contiguous sequences were assembled and edited in Geneious ver. 11 (https://www.geneious.com). All sequences were submitted to GenBank under accession numbers MN720141-MN720149; MN723852-MN723854; MN726941-MN726975; and MN726988-MN727001. For species identification, each sequence was compared with sequences available in GenBank by using the Basic Local Alignment Tool (BLAST).

Phylogenetic analyses
The newly generated sequences were aligned with sequences available in GenBank according to the trematode family and gene amplified (supplementary table S1). Sequences were aligned with MUSCLE (Edgar, 2004) implemented in Geneious ver. 11. A total of eight alignments for nine families were analysed. Outgroup selection was based upon the molecular phylogenies of Olson et al. (2003), Tkach et al. (2016), Kanarek et al. (2017) and Hernández-Orts et al. (2019). Phylogenetic trees for each dataset were constructed with Bayesian inference (BI) and maximum likelihood (ML) analyses on the CIPRES portal (Miller et al., 2010) and the ATGC bioinformatics platform, respectively. The Akaike Information Criterion implemented in jModelTest 2.1.1 (Guindon and Gascuel, 2003;Darriba et al. 2012) was used to determine the best-fit nucleotide substitution model for each dataset. These were the general time reversible model, with estimates of invariant sites and gamma distributed among-site rate variation (GTR + I + G) for six alignments:

General observations
A total of 12.9% of all B. tentaculata from Germany showed larval trematode infections. Snails collected in Lithuania showed an overall prevalence of 14%. Phylogenetic and BLAST analyses based on 55 novel sequences for 36 isolates recovered from B. tentaculata collected in Germany and Lithuania (table 2) demonstrated high diversity of digeneans, including 20 species belonging to ten families: Cyathocotylidae, Echinochasmidae, Lecithodendriidae, Lissorchiidae, Notocotylidae, Opecoelidae, Opisthorchiidae, Pleurogenidae, Prosthogonimidae and Psilostomidae. Six partial cox1 sequences were generated from isolates of B. tentaculata sampled in all German localities (MN720141-MN720146). The sequence difference between the isolates was 0-0.2% (1 nucleotide (nt) difference), thus confirming their conspecificity. Molecular identification of the snail isolates was achieved by comparing our sequences with sequences for B. tentaculata in GenBank. A BLAST search analysis indicated a 86% coverage and 98% of similarity with two isolates of B. tentaculata from Germany (AF445334) (Hausdorf et al., 2003) and North America (JX970605) (Wilke et al., 2013); and a 89% coverage and 92% of similarity with one isolate from Croatia (AF367643) (Wilke et al., 2001). Snails from the Lithuanian system were identified morphologically.

Molecular results
In total, four snails from three localities were infected with cercariae belonging to the family Cyathocotylidae (prevalence: River Lippe: 0.2%; River Rhine: 4%). Sequences for the partial 28S rRNA gene and entire ITS1-5.8S-ITS2 gene cluster were generated for one isolate per locality.
Both BI and ML analyses of the Cyathocotylidae based on 28S rDNA alignment included novel sequences and those retrieved from GenBank ( fig. 2a; supplementary table S1), and resulted in trees with similar topologies. Sequences for the isolates CR1 and CK2a clustered with the sequence for Cyathocotyle prussica Mühling, 1896 with a strong support. A single isolate (CR2) formed a branch basal in the clade of Cyathocotyle spp.

Gene fragment
Primer name Nucleotide sequence Source a Internal primers. PCR conditions were followed as described in the source papers.
Sequences of partial 28S rDNA obtained in this study were aligned with the available sequences for echinoschasmids (n = 15) and psilostomids (n = 13) from GenBank (supplementary table S1). Two species of the Himasthlidae, Acanthoparyphium spinulosum Johnston, 1917 and Himasthla limnodromi Didyk & Burt, 1997 were used as the outgroup based on the topologies in the phylogenetic tree of the Echinostomatoidea published by Tkach et al. (2016). Both BI and ML analyses yielded a similar topology, with two main clades representing the two families, Echinochasmidae and Psilostomidae. The newly generated sequences fell into two distinct and strongly supported clades within each family. Two isolates, ECR1 and EBR1, collected from the River Rhine were identical with the sequences for Echinochasmus coaxatus Dietz, 1909 (KT956928)  Gallus gallus from Vietnam. The interspecific divergence between Echinochasmus sp. 1 and E. milvi was 1.5% (58 nt), and between Echinochasmus sp. 2 and E. beleocephalus and E. japonicus was 0.2% (2 nt) and 0.6 (7 nt), respectively.
Six isolates (SR2, PS1R1, PS1R2, PS1K2a, PSCB and PS2R2) represented by three species fell within the clade for the Psilostomidae. One isolate (SR2) collected from the River Rhine was identical to the isolate for Sphaeridiotrema sp. ex B. tentaculata (KT956958) from Lithuania (fig. 5; supplementary table S1). Five remaining isolates formed a strongly supported clade with four of them (PS1R1, PS1R2, PS1K2a and PSCB) representing the same species, whereas the fifth isolate (PS2R2) was distinct. The sequences for these five isolates did not show affiliation to any of the psilostomid genera included in the analyses. The interspecific divergence between the two species was 1.8% (20 nt). Based on these results, both species were identified to the family level as Psilostomidae gen. sp. 1 and Psilostomidae gen. sp. 2.
Additional analyses were conducted for the Psilostomidae in order to include sequences of the three species of the genus Psilotrema Odhner, 1913 available in GenBank (supplementary table S1). The sequences for these species were not included in the main analyses due to their short length (759 nt). In these analyses, the isolates for Psilostomidae gen. sp. 1 and Psilostomidae gen. sp. 2 clustered in a clade with representatives of the genus Psilotrema, P. oschmarini, P. simillium and P. acutirostris, while Psilostomidae gen. sp. 1 appeared to be conspecific with P. oschmarini (see supplementary fig. S1). Based on this result, both species -Psilostomidae gen. sp. 1 and Psilostomidae gen. sp. 2may belong to the genus Psilotrema. However, at this stage, we refrain from identifying cercariae in our material as P. oschmarini due to the results being based on a short dataset and the lack of morphological vouchers for the sequences in GenBank.  torial, 26-33 × 23-31 (30 × 27). Oral/ventral sucker width ratio 1:0.90-1.14 (1:1.03). Prepharynx distinct, pharynx spherical, muscular, 6-10 × 5-10 (8 × 7). Caeca indistinct. Penetration glandcells numerous, on both sides posterior to oral sucker. Cystogenous gland-cells few, rounded, with rhabditiform contents. Excretory vesicle bipartite; anterior part transversely oval, at level of posterior margin of body, posterior part transversely oval, at junction of body and tail; main collecting ducts wide, dilated between mid-level of pharynx and level of posterior margin of ventral sucker, containing large dark refractile excretory granules of different size (12 on each side). Locality. River Rhine (R1), Germany. Representative DNA sequences. 28S rDNA, one replicate (MN726945).

Remarks
Cercariae of both identified species of Echinochasmus -E. coaxatus and E. bursicolahave been previously reported from B. tentaculata (Karmanova, 1973(Karmanova, , 1974. The life cycle of E. coaxatus  was described by Karmanova (1974) in the Astrakhan Nature Reserve, Russia. Cercariae were found in B. tentaculata and metacercariae were found in the freshwater fishes of the families Cyprinidae and Percidae, collected in the River Volga. Cercariae of E. coaxatus were also reported from Radix auricularia in the lake Gołdapiwo, Poland, by Wiśniewski (1957). General morphology of cercariae found in our study corresponded well to the description for cercariae of E. coaxatus by Karmanova (1974), except in the number of the refractile excretory granules per collecting duct (12 vs. 13-14, respectively). The differences in the metrical data for body [114-184 vs. 103-124 (115)] and tail lengths [123-140 vs. 72-115 (92)] may be due to the different fixation methods. Karmanova (1974) did not indicate whether the measurements were taken from live or fixed cercariae.
Cercariae of E. bursicola were previously described by Karmanova (1973) from B. tentaculata collected in the Lower Volga, Russia. Although the method of fixation was not specified, the present cercariae differ from cercariae described by Karmanova (1973)  Further identification of Echinochasmus sp. 1 and Echinochasmus sp. 2 to the species level requires the sequences of the adults from the definitive hosts, which are typically fisheating birds and rarely mammals (Tkach et al., 2016).

Description
Cercariae of Sphaeridiotrema sp. were found in one snail at Locality R2 in the River Rhine. The species of the cercariae was identified based on the results of molecular analyses. No morphological data were obtained for cercariae of this species.  Locality. River Rhine (R2), Germany. Representative DNA sequences. 28S rDNA, one replicate (MN726954).

Description
Cercariae of Psilostomidae gen. sp. 2 (fig. 5e) were found in one snail at Locality R2 in the River Rhine. Cercariae were identified based on the results of molecular analyses. No morphological data were obtained for cercariae of this species.
Sphaeridiotrema sp. in our material appeared to be conspecific to the species that was previously reported from Lithuania (Tkach et al., 2016). Both isolates were identified only to the genus level and require sequences of adults from the definitive hosts (water birds), to complete identification to the species level.
General morphology of cercariae of Psilostomidae gen. sp. 1 resemble morphology of cercariae of P. oligoon described by [454-495 (476) (49)]. The above comparisons demonstrate that Psilostomidae gen. sp. 1 may represent a species of the genus Psilotrema, but is not conspecific with P. oligoon, P. simillimum or P. spiculigerum. Further identification of Psilostomidae gen. sp. 1 and Psilostomidae gen. sp. 2 to the species level requires the sequences of the adults from the definitive hosts, which are mainly birds and mammals (Kostadinova, 2005). Locality. River Lippe (K3), Germany. Representative DNA sequences. 28S rDNA, one replicate (MN726955).

Molecular results
Cercariae of Asymphylodora sp. were found in one snail in the River Lippe (prevalence: 0.2%). Partial 28S rDNA sequences were generated from one isolate (table 2). The partial 28S rDNA sequence for Asymphylodora sp. obtained in the present study was compared to the sequences of Asymphylodora perccotti Besprozvannykh, Ermolenko & Atopkin, 2012 (FR822715 −FR822731) ex Perccottus glenii Dybowski, 1877 from Russia (Besprozvannykh et al., 2012), the only sequences for this genus currently available in GenBank. The sequence divergence between our isolate and 17 isolates for A. perccotti ranged between 2.7% and 2.8% (31−32 nt).

Description
The species of the cercariae was identified based on the results of molecular data. No morphological data were obtained for cercariae of this isolate.

Molecular results
Infection with the cercariae of the family Notocotylidae was detected in nine snails from four localities in the River Lippe (prevalence: 1.5%). Partial 28S rDNA sequences were generated for four isolates ( fig. 6; table 2) and aligned with 16 sequences for species of the Notocotylidae available in GenBank (supplementary table S1). Members of the families Opisthotrematidae, Rhabdiopoeidae and Labicolidae were used as the outgroup based on the topologies in the phylogenetic tree of the Digenea published by Olson et al. (2003). The results of phylogenetic analyses demonstrated that two isolates preliminarily identified as Notocotylus sp. (N11K0 and N12K0) clustered within a clade comprising Notocotylus spp., demonstrating the close affinity to the isolate of Notocotylus attenuatus (Rudolphi, 1809) (AF184259), the type species of the genus Notocotylus, collected from Aythya ferina in Ukraine (Tkach et al., 2001). Sequences for two isolates of Notocotylus sp. from our study were identical and differed from N. attenuatus by 0.4% (3 nt). The sequences for two other notocotylid isolates (N2K0 and N2K2b) from B. tentaculata collected in the River Lippe were identical and formed a basal branch to the clade consisting of Notocotylus spp. and Catatropis spp., albeit without support ( fig. 6). The taxonomic identity of these two isolates was not justified based on the phylogenetic analyses and we, thus, provide the identification for this species only to the family level, as Notocotylidae gen. sp. The sequence divergence between two notocotylid species recorded in our study was 2.9% (23 nt).

Description
No morphological data were obtained for cercariae of these isolates since the infections were prepatent.

Remarks
Members of the family Notocotylidae are reported to utilize lymnaeids, planorbids and a variety of other snail families in their life cycles (Filimonova, 1985). To date, six speciesnamely, N. attenuatus, N. ponticus, N. parviovatus, N. imbricatus, Notocotylus sp. and Catatropis verrucosawere reported to develop in B. tentaculata in Europe (Bock, 1982;Filimonova, 1985 and references therein;Morley et al., 2004). Further identification of the two species collected from B. tentaculata in the River Lippe to the species level requires the sequences of adult worms from the definitive hosts, which are mammals and birds (Filimonova, 1985).

Systematics
Superfamily: Allocreadioidea Looss, 1902Opecoelidae Ozaki, 1925 Molecular results Infection with cercariae of the family Opecoelidae was detected in 45 snails from four localities in the River Lippe (prevalence: 7.4%). Sequences for the partial 28S rDNA and ITS2 region were generated for six isolates ( fig. 7; table 2). Comparative sequence analyses of 28S and ITS2 datasets revealed the presence of two species of the family Opecoelidae in our material. Five sequences of partial 28S rRNA gene (table 2) were aligned with seven GenBank sequences for species of the Opecoelidae known to parasitize freshwater fish (supplementary table S1). A species of the Opecoelidae, Buticulotrema thermichthysi Bray, Waeschenbach, Dyal, Littlewood & Morand, 2014, was used as the outgroup based on the topologies in the phylogenetic tree of the Opecoelidae published by Martin et al. (2019). The results of the phylogenetic analyses demonstrated a close affinity of the four isolates (O11K2a, O1K1, O1K2b and O13K2a) with Sphaerostoma bramae (Müller, 1776) (MH161435) collected from Abramis brama in Russia (Sokolov et al., 2019). The sequences for our isolates were identical and differed from the sequence of S. bramae by 0.2% (3 nt), which is considered as interspecific variation and, thus, this species was identified as Sphaerostoma sp. The intraspecific divergence between the isolates of Sphaerostoma sp. (O11K2a, O1K1, O1K2b and O12K2a) within the ITS2 dataset was 0.2% (1 nt).
The remaining isolate (O2K2a) collected in the River Lippe, representing a second opecoelid species in our material, formed a branch basal to a clade consisted of Sphaerostoma spp. and Plagiocirrus spp. The 28S rDNA sequence of this species differed from the sequence of Sphaerostoma sp. by 5.9% (71 nt), from S. bramae by 5.6% (68 nt) and from Plagiocirrus spp. by 7.1-7.2% (85-87 nt), whereas the ITS2 sequence differed from the sequence of Sphaerostoma sp. by 4.7-4.9% (21-22 nt). Based on the results, we identified this species only to the family level as Opecoelidae gen. sp.

Description (no photomicrograph available)
No morphological data were obtained for cercariae of this isolate since the infections were prepatent.

Molecular results
Infection with cercariae belonging to the families Pleurogenidae and Prosthogonomidae was detected in nine snails from three localities (prevalence: Pleurogenidae: River Lippe, 0.5%; River Rhine, 1.3%; Prosthogonomidae: River Lippe, 0.7%; River Rhine, 1.3%). Sequences for the partial 28S rDNA (n = 9) and ITS2 region (n = 8) were generated for the isolates from all localities ( fig. 9; table 2). The 28S rDNA sequences were aligned with the sequences for pleurogenids (n = 13) and prosthogonomids (n = 5) available in GenBank. Two species of the family Microphallidae were used as the outgroup based on the topologies in the phylogenetic tree of the Microphalloidea published by Kanarek et al. (2017) (fig. 9; supplementary table S1). Both BI and ML analyses yielded similar topology with two main clades corresponding to the Pleurogenidae and Prosthogonomidae. The sequence of the isolate (PBK2b) collected from the River Lippe (K3) appeared to be identical to the sequence of Parabascus duboisi ex M. daubentonii from Ukraine (AY220618) . The isolate (PL2R1) collected from B. tentaculata in the River Rhine clustered with Leyogonimus polyoon (KY752116) from Fulica atra collected in Poland (Kanarek et al., 2017). The sequence divergence between two species was 2% (23 nt). This isolate was identified to the family level as Pleurogenidae gen. sp. 2. The two isolates (PL11K2a and PL12K2a) collected from the River Lippe (K2) clustered with pleurogenid species from the genera Brandesia, Candidotrema, Pleurogenes, Pleurogenoides and Prosotocus, and identified only to the family level as Pleurogenidae gen. sp. 1.
Sequences of the ITS2 region for pleurogenids and prosthogonomids obtained in this study were aligned with sequences of prosthogonomids available in GenBank (supplementary table S1). Sequences of the four isolates (PO1K2b, PO1R1, PO2R1 and POK2a) clustered with the sequence of P. ovatus (KP192722) from A. ferina collected in the Czech Republic (Heneberg et al., 2015). The four remaining isolates (PBK2b, PL2R1, PL11K2a and PL12K2a) identified as the members of the family Pleurogenidae based on 28S rDNA analyses clustered within a nearly supported clade ( fig. 9a).

Description
No morphological data were obtained for cercariae of this isolate since the infection was prepatent.

Remarks
The life cycle of P. duboisi is unknown and our finding is the first to report B. tentaculata serving as the first intermediate host for this species. Parabascus duboisi is known to parasitize, among other bats, those of the genera Eptesicus, Miniopterus, Myotis, Pipistrellus and Rhinolophus (Sharpilo & Iskova, 1989).

Discussion
This study examined the parasite diversity in the faucet snail B. tentaculata in Central European fresh waters. To the best of our knowledge, this is the first broad faunistic survey investigating the trematode fauna in B. tentaculata in Central Europe and providing a combined morphological and molecular dataset. The study reveals a high trematode diversity of 20 species belonging to ten families and a high overall prevalence of infection of 12.9% in this snail host.
Not surprisingly, the larval trematode community of B. tentaculata shows little overlap with that of lymnaeids and planorbids, highlighting the distinctive host-parasite associations of pulmonate and non-pulmonate snails. Only one (L. linstowi) out of 20 species has also been recorded from lymnaeid and planorbid snail hosts (Enabulele et al., 2018). Bithynids are the typical hosts of L. linstowi, and the single finding of this species in Radix sp. most probably represents an accidental infection. The species has only been reported in a single lymnaeid host, so it is reasonable to consider it as an accidental infection. Looking at the faunistic overlap of trematodes at the family level, we find a slightly different picture. Out of ten recorded families, six (Lecithodendriidae, Lissorchidae, Notocotylidae, Opecoelidae, Pleurogenidae, Psilostomidae) are also known to occur in lymnaeids and planorbids (Faltýnková et al., 2007(Faltýnková et al., , 2008(Faltýnková et al., , 2016Cichy et al., 2011;Enabulele et al., 2018). On the other hand, species of the families Cyathocotylidae, Echinochasmidae, Prosthogonomidae and Opisthorchiidae seem to be strictly host-specific to non-pulmonate freshwater molluscs as first intermediate hosts. Some species, such as E. beleocephalus and E. coaxatus, have been recorded from B. tentaculata only in Russia Journal of Helminthology 19 (Frolova, 1975;Karmanova, 1975), so our records constitute the first record for Central Europe. Echinochasmus sp. 1 and Psilostomidae gen. sp. 1 are the only species that were both detected in Germany and Lithuania. The definitive hosts of echinochasmids and psilostimids are typically birds and mammals (Kostadinova, 2005;Tkach et al., 2016). Birds are especially mobile, so the occurrence of the two species in both countries can be easily explained by seasonal migration. However, since the sampling effort in Germany was much higher (682 vs. 121 B. tentaculata), we would expect more trematode species to be present in Curonian Lagoon, which our limited survey did not detect.
Overall, this diverse and distinctive trematode community of B. tentaculata, and the high prevalence of infection, reveal the important role of this snail species as a first intermediate host for trematodes in European freshwater ecosystems. Similar to other well-studied host-parasite systems, B. tentaculata supports a parasite community that presumably fulfils vital and central ecological functions, ranging from contributing to ecosystem diversity, structuring food webs or serving as ecosystem engineers (Thomas et al., 1999;Mouritsen & Poulin, 2002;Lafferty et al., 2008;Hatcher et al., 2012;Dunne et al., 2013). Moreover, since parasites can also serve as indicators of the local diversity and trophic interactions of free-living organisms (Hechinger et al., 2007;Byers et al., 2010;Shea et al., 2012), the distinct trematode communities of B. tentaculata offer valuable insights into local habitat conditions.
One interesting example of the indication of local diversity and trophic interaction using digenean trematodes might be the detection of P. duboisi and L. linstowi. On the basis of our findings, we can infer the presence of bats in the studied habitat, as both are known to parasitize bats, e.g. the Daubenton's bat M. daubentonii (Gottschalk, 1970;Esteban et al., 2001;Tkach et al., 2003). The Daubenton's bat feeds on aquatic insects and insects with aquatic larvae, such as Lepidoptera, Diptera and Trichoptera, and it is, therefore, highly dependent on water sources. It hunts over standing or slow-moving water bodies and takes its prey from the water surface (Krapp, 2011 and references therein). Based on the finding of P. duboisi and L. linstowi, we are able to make inferences about the presence of bats at the studied habitat and the trophic relations between aquatic insects, which most probably serve as second intermediate hosts for the detected parasite species (Sharpilo & Iskova, 1989;Kudlai et al., 2015;Enabulele et al., 2018) and its final bat host.
The current study was limited by the lack of relevant sequences for many trematode families in GenBank. Consequently, a major proportion of our isolates could only be identified to the genus or family level. Moreover, the complete life cycle of many trematodes parasitizing B. tentaculata have not yet been elucidated (see Kudlai et al., 2015) and remain unclear. Such obstacles impede and exacerbate extensive studies on the diversity, the ecological role and the influence of digeneans on food webs. Therefore, it is important to extend and compile morphological data of cercariae also from non-pulmonate snails, obtain more molecular isolates of adult specimens to facilitate molecular identification and clarify the still unknown life cycles of many trematode species. The present study can be seen as an important step in compiling morphological and molecular data on the digenean parasite fauna of bithyniids. Among the 20 digenean species, we are able to present the characteristics (measurements and/or photomicrographs) for 14 taxa. With the present host-parasite list we hope to foster parasitological research on parasites of understudied snail families.
Taken together, our findings and the limitations we encountered demonstrate unambiguously that our knowledge of the studied parasite-host system remains limited and large-scale studies focussing on non-pulmonate freshwater snails are lacking. This fits the overall trend of a currently highly patchy research effort on parasite diversity, which not only prevents a full inventory of parasite biodiversity but also impedes predictions of their role in ecosystems (Jorge & Poulin,2018). Our study revealed an abundant and diverse trematode fauna in B. tentaculata, which highlights the need for further research on this host-parasite system. Therefore, we might currently be underestimating the ecological roles and impacts of parasite communities of non-pulmonate freshwater snails in European fresh waters. In order to fully comprehend the numerous and often central roles these parasites play in aquatic ecosystems, we need to better understand such understudied host-parasite systems.