© J.A. Dołęga et al., Published by EDP Sciences, 2026

1 Introduction

Large freshwater mussels (Unionidae, hereafter referred to as naiads) require a mandatory parasitic larval phase to complete their ontogeny. Females produce tens of thousands, or even hundreds of thousands, of tiny parasitic larvae (larval length is approximately 0.05–0.45 mm), known as glochidia (Bauer, 2001). Once fully developed, glochidia are expelled by the female through the exhalant siphon into the water, where they passively float and attract a fish host to which they attach, typically to the gills or fins (Aldridge et al., 2023) where they remain encysted for several weeks in the case of Unio crassus complex (Lamand et al., 2016; Schneider et al., 2017a) or even up to 11 months in Margaritifera margaritifera (Linnaeus, 1758) (Bauer and Vogel, 1987). During this period, the larvae metamorphose and detach from the host as juvenile mussels. This phase is crucial for naiad reproduction, and a key concern is that not all fish species are capable of successfully hosting larvae of every naiad species. Due to co-evolution, each naiad species has a specific set of host species (Bauer and Wachtler, 2001). Moreover, local co-evolution and co-diversification can further specialise this relationship (Bauer and Wachtler, 2001; Douda et al., 2017; Neemuchwala et al., 2023).

From the perspective of both biodiversity and nature conservation, coevolution is particularly important when threatened naiad species depend on sensitive fish hosts, which could pose a risk of cascading extinctions (Modesto et al., 2018). This is the case for species belonging to U. crassus complex, including Unio crassus Philipsson, 1788 and Unio nanus Lamarck, 1819 – a globally endangered species (Lopes-Lima et al., 2024, Lopes-Lima and Prié, 2024) endemic to Europe, and protected by both EU (Council of the European Communities, 1992) and national regulations (Journal of Laws of the Republic of Poland, 2016) under the common name of U. crassus. One of the best-known hosts for larvae of this species complex is the European bullhead Cottus gobio Linnaeus, 1758 (Teleostei: Scorpaeniformes) (Douda et al., 2012; Lamand et al., 2016). This species is actually one of several within the C. gobio group (Freyhof et al., 2005), a complex often referred to as C. gobio sensu stricto, whose members are locally considered vulnerable (VU) (Witkowski, 2009) and are similarly, legally protected under national (Journal of Laws of the Republic of Poland, 2016) and European law (Council of the European Communities, 1992). To date, this sculpins has been tested several times for its host role in artificial infestation experiments using U. crassus complex glochidia under controlled conditions, e.g., by Douda et al. (2012), Taeubert et al. (2012a, 2012b) and Schneider et al. (2017a). The geographical origin of the fish used in these experiments suggests that, in fact, only C. gobio sensu stricto has been tested as a host to date. However, both morphological and molecular analyses have revealed that in the Oder, Vistula (Baltic Sea catchment), and Dniester (Black Sea catchment) river drainages, the sculpin species that co-occurs with the U. crassus complex is not the European bullhead C. gobio, but the Baltic sculpin Cottus microstomus Heckel, 1837 (Freyhof et al., 2005; Sideleva et al., 2018). A more recent study by Sideleva et al. (2022) shows that C. microstomus also occurs in river systems of the eastern Baltic Sea catchment, such as the Neman and Venta rivers (including the Krasnaya River).

Given the cases of strict co-evolution between naiads and their host species, it is crucial to identify the host species of U. crassus complex across its entire geographical range, particularly in relation to its threatened status. A key question concerns the role of species of the genus Cottus, which may serve as one of the most suitable hosts. In this study, we present the results of an experiment designed to assess whether C. microstomus is also a primary host for U. crassus complex larvae, as this fish species is found in the larger, eastern portion of U. crassus complex range.

2 Material and methods

2.1 Animal collection and acclimation

To test the effectiveness of C. microstomus as a host for mussel larvae, we conducted an artificial infestation using U. crassus complex glochidia under controlled conditions. As a reference species, we used the Eurasian minnow Phoxinus phoxinus (Linnaeus, 1758), which is widely used as a host in artificial infestation experiments and is known for its ability to transform U. crassus complex larvae (e.g., Taeubert et al., 2012a, 2012b; Douda et al., 2012, 2014; Schneider et al., 2017a; Geist et al., 2023). Fish and mussel sampling took place in two rivers within the Nida River system, a tributary of the Vistula River in the Baltic Sea catchment. Five gravid females belonging to U. crassus complex and 11 specimens of P. phoxinus were collected from the Warkocz River near Niestachów, while 11 specimens of C. microstomus were collected from the Mierzawa River near Sędowice (both sites are located in the Świętokrzyskie Voivodeship, southern Poland). At the time of the experiment, only the Warkocz River supported a significant population of mussels belonging to U. crassus complex, allowing for their sampling without harm to the local population. The choice of site for obtaining C. microstomus specimens was deliberate: not only it is located in terra typica of the species, but also, during the recent rediscovery and redescription of Heckel's holotype from year 1837, non-type specimens from this site were confirmed as C. microstomus (Sideleva et al., 2018). Specimens of the U. crassus complex had never been observed in the Mierzawa River, despite intensive surveys conducted since 1997 (unpublished data), thus ensuring that the collected C. microstomus specimens were naive to glochidia. A map showing the locations of animal collection sites and the cited geographic ranges of the discussed species is provided in Figure 1.

Fish were collected using a backpack electrofishing device (IUP-12A, prod. Radet, Poland) and transported to the mussel breeding facility in Krzyżanowice Średnie (Świętokrzyskie Voivodeship, southern Poland) in plastic bags filled with fresh river water and oxygen. At the site of sympatric occurrence of P. phoxinus and U. crassus, fish were inspected for visible glochidia attached to their fins or gills by gentle opening of the operculum. Only specimens appearing free of larvae were selected for the experiment. Even though C. microstomus is the only Cottus species inhabiting the Vistula river drainage, some anatomic traits were checked during fish collection: e. g. smooth dorsal profile (no hump behind the head), upper jaw reaching the anterior margin of an eye, lobular genital papilla occurrence in males and distinct narrowness of caudal peduncle (Sideleva et al., 2018). Total length of all fish was measured to the nearest 1 mm. Fish body weight data were not collected, however, considering the different allometry of C. microstomus and P. phoxinus, body mass of each specimen was calculated using the length–weight relationship equation with coefficients specific to the species or family, as determined by Froese et al. (2014). At the breeding facility, fish were placed individually in 13-litre tanks within a recirculating aquaculture system for acclimation. Each tank, equipped with a surface overflow, was constantly supplied with aerated and purified water with controlled parameters. Encystation and transformation of larvae occurred within these tanks. The entire facility was air-conditioned, maintaining a constant temperature. Following previous findings that glochidia viability and metamorphosis of larvae depend on water temperature (e.g., Zimmerman and Neves, 2002; Taeubert et al., 2014; Lamand et al., 2016; Benedict et al., 2021), and that mussel reproduction is adapted to local thermal conditions (Schneider et al., 2017b), we maintained a temperature of 16°C. This approximated natural environmental conditions while allowing efficient larval transformation and minimizing thermal stress to animals. Fish acclimation lasted for eight days.

Adult mussels (age class 5+) were located while wading in the Warkocz River, hand-collected, sexed, and inspected for fully developed, “snapping” glochidia using marsupial puncture, as described by Zając and Zając (2021). Five gravid females of U. crassus were transported in a box with fresh, aerated water and ice to the breeding facility, where they were placed in a 96-litre aerated tank. The tank was checked daily to monitor whether glochidia were released into the water, and whether the gentle opening of the valves in each specimen revealed any changes in the marsupial fill.

Fig. 1 A map of geographical ranges of discussed species and the study location; A) range of occurrence of C. microstomus: a polygon with hatched fill -spatial data was provided by IUCN (MfN Berlin, 2021b) and updated by authors to include drainages of Neman and Venta, as no available spatial data for Krasnaya river drainage was found; and C. gobio: a polygon with dotted fill -spatial data provided by IUCN: (MfN Berlin, 2021a); points indicate exact sites of collection of specimens used in the study by Lopes-Lima et al. (2024) to assess U. crassus complex genetic divergence (limited to the map range): black points - U. crassus, white points – U. nanus, grey points – sites of sympatric occurrence of both species; the red diamond indicates the area of this study; B) the sites of collection of animals in this study: 1- P. Phoxinus specimens used during artificial infestation, U. crassus complex specimens used in artificial infestation, U. crassus complex individuals used for genetic reassessment of the population inhabiting the river Warkocz; 2- C. microstomus specimens used during artificial infestation.

2.2 Infestation of fish and juvenile mussel collection

Once visible conglutinates of glochidia were observed in the tank and it was ensured that the females had emptied their marsupia, the mussels were removed from the tank. The contents were gently mixed, and all fish were simultaneously placed into the glochidia suspension to ensure a uniform infestation rate across both fish species. The fish were exposed to glochidia for 20 min. During this period, the suspension tank was continuously aerated and hand-stirred to prevent glochidia sedimentation and to stimulate fish movement. Following the exposure, each fish was carefully returned to its individual 13-litre tank within the recirculating aquaculture system. At this stage, each tank was equipped with a net (mesh size 2 mm), which was fixed horizontally approximately 3 cm above the tank bottom, to prevent the fish from foraging on juvenile mussels detaching from their gills and sinking to the bottom. Juvenile mussels were collected every 3 to 4 days. On each occasion, the individual 13-litre tank was removed from the recirculating aquaculture system, and the fish were gently transferred using a fine mesh net to a temporary holding tank. The contents of the 13-litre tank were then flushed several times through a planktonic net (mesh size 0.01 mm), which led to a small tank. The contents of this tank were subsequently passed through a sieve (mesh size 0.15 mm), and the material retained in the sieve was transferred onto a petri dish. Transformed and viable juveniles were counted using a binocular magnifier (Delta Optical, SZ-450B). After the procedure, the tank was refilled with water, and all components were restored to their original configuration to allow the procedure to be repeated.

2.3 Statistical analyses

Statistical analyses were conducted using IBM SPSS Statistics. In order to investigate differences between glochidia transformation patterns in consecutive controls between both species of fish, a generalized mixed model with repeated measures (Log-Poisson) was constructed; dependent variable: number of transformed and vital juvenile mussels at given time post infestation of fish, grouping variable: host species, continuous predictor: fish weight, random effect: fish specimen ID. The random effect was used to account for potential differences between fish specimens, which may result from earlier acquisition of immunity to glochidia, or individual condition, infestation intensity or any other non-controlled factor resulting from testing a random group of wild fish. Differences between specific levels of variables were tested using pairwise contrasts post-hoc tests and are presented in the Supplementary Materials. Since juvenile mussels were collected every 3 or 4 days, in the descriptive statistics we refer to the “potential excystation period” and “potential parasitism period”, which reflects respectively: the maximum possible time between first and last excystment of juvenile mussel from a host, and maximum possible time from infestation to excystment of the last juvenile, both determined for each host fish specimen. The mean parasitism duration was calculated based on the time each individual glochidium spent attached to the host, from initial attachment until collection of transformed juvenile mussel - these data were then pooled for each host species.

2.4 Molecular taxonomic verification of the U. crassus complex species from the Warkocz River

While much of the existing literature and legal frameworks refer to “U. crassus” as a single species, recent findings suggest that it should be regarded as a species complex. In their recent study, Lopes-Lima et al. (2024) identified 12 molecularly distinct species within what was previously recognised as “U. crassus”, two of which are likely present in our study area: U. crassus sensu stricto and U. nanus. To reassess the taxonomic identity of Unio specimens from the Warkocz River under the revised taxonomic framework of the U. crassus complex, we reanalyzed mitochondrial cytochrome c oxidase subunit I (COI) barcode sequences previously obtained by our team (Kilikowska et al., 2020). At the time of the original study, all specimens were assigned to U. crassus according to the taxonomy then in use.

Following the integrative taxonomic revision by Lopes-Lima et al. (2024), which recognized 12 geographically structured species within the U. crassus complex, we conducted a comparative molecular analysis to determine the species that inhabited the Warkocz River. Two previously published COI sequences represented two haplotypes from the Warkocz population (GenBank accession numbers KJ525912 and KJ525917; Kilikowska et al., 2020) were analyzed together with one representative COI sequence for each of the 12 species described by Lopes-Lima et al. (2021, 2024). All sequences have been downloaded from GenBank (accession numbers listed in supplementary materials, Tab. S3).

The COI sequences were aligned using the MUSCLE algorithm implemented in MEGA X v10.0.3 (Kumar et al., 2018) with default settings. All barcode sequences were trimmed to equal length, resulting in a final alignment of 614 bp. The alignment contained no gaps or ambiguous sites. To confirm the absence of pseudogenes, sequences were translated into amino acids using the invertebrate mitochondrial genetic code, and no stop codons were detected.

Uncorrected pairwise genetic distances (p-distances) were calculated in MEGA X. Codon positions included were 1st+2nd+3rd+Noncoding. All positions containing gaps and missing data were eliminated (complete deletion option). The resulting distance matrix was used to assess genetic divergence between haplotypes of U. crassus species complex from Warkocz River and reference sequences representing the newly delimited species within the U. crassus complex. All pairwise p-distance values are provided in Tab. S3 in supplementary materials.

3 Results

3.1 Artificial infestation

The Baltic sculpin C. microstomus demonstrated the ability to successfully transform glochidia released by females originating from a mixed population of U. crassus sensu stricto and U. nanus into viable juvenile mussels. Observations on equal groups of fish kept under identical conditions revealed a higher efficiency of larval transformation in P. phoxinus compared to C. microstomus. The number of successfully transformed juveniles per specimen in C. microstomus ranged from 44 to 134, mean = 82.2, SD = 27.27), while in P. phoxinus it ranged from 26 to 211 (mean = 114.1, SD = 63.67). In total, C. microstomus produced 904 juvenile mussels, while P. phoxinus produced 1255 juvenile mussels. The results of constructed GLZ (log-poisson) model showed that the difference in number of juveniles collected during the whole excystation period between species was not statistically significant (F(1, 119) = 0.46, p = 0.501) and also, that it was not significantly influenced by the fish weight (F(1, 119) = 0.99, p = 0.323). The number of juvenile mussels (pooled across both host species) differed significantly between most of the collections dates (F = 246.2, df1 = 5, df2 = 119, p < 0.001) indicating, that juvenile mussels release showed dynamic changes over the collection dates; there was no statistically significant difference in the number of juveniles collected on the 22nd and 33rd and on the 37th and 40th days post-infestation (see Tab. S1 in the Supplementary Materials).

The model also indicated a statistically significant interaction between host species and collection date (F = 37.9, df1 = 5, df2 = 119, p < 0.001), showing that the number of successfully metamorphosed juvenile mussels varied between species in a collection-specific manner (Fig. 2).

A random effect was also found to be statistically significant, although it explained less than 23% of the variation in the data (η² = 0.223, SE = 0.088, p = 0.012). To further illustrate the data, a figure displaying the mean percentage of juveniles collected at each time point has been included, which visually depicts the differences in excystation patterns for both host species in a cumulative diagram (Fig. 3).

Both host fish species exhibited similar maximum time frames for larval transformation. The first transformed, viable juveniles of U. crassus were collected from both species on the 22nd day post-infestation, and the last juvenile mussels detached from the fish by the 40th day after infestation, with a pronounced peak in juvenile collection on the 26th day (Fig. 2). No juveniles were collected during the control checks conducted on the 19th and 44th days post-infestation. The potential excystation period lasted from 18 to 21 days (mean = 20.45, SD = 1.21) in C. microstomus and from 11 to 21 days (mean = 18.36, SD = 3.88) in P. phoxinus. The potential parasitism period lasted from 37 to 40 days (mean = 39.45, SD = 1.21) in C. microstomus and from 33 to 40 days (mean = 37.91, SD = 2.77) in P. phoxinus. Mean duration of successful parasitism was 26.93 days (SD = 3.9 days) in C. microstomus and 28.39 days (SD = 3.5 days) in P. phoxinus.

The size of 11 specimens of P. phoxinus ranged between 48 mm and 90 mm (median: 72, SD = 11.51), while the size of 11 specimens of C. microstomus ranged between 63 mm and 127 mm (median: 82, SD = 17.56). The calculated weigth of 11 specimens of P. phoxinus ranged between 0.95 g and 6.86 g (median: 3.41, SD = 1.68), while for 11 specimens of C. microstomus it ranged between 1.94 g and 17.66 g (median: 4.45, SD = 4.36).

Fig. 2 The differences in number of juvenile mussels in consecutive counts between studied species: C. microstomus (empty circles) and P. phoxinus (solid squares); points: predicted average, whiskers: 95% confidence interval, DPI (Days post-infestation): days from exposition of fish to U. crassus glochidia; gray markers indicate last and first control in which no juveniles were detected (estimated start and finish point).

Fig. 3 Cumulative percentage of juvenile mussels collected from fish in consecutive steps: juveniles collected from C. microstomus (empty circles) and P. phoxinus (solid squares); points: predicted average, whiskers: 95% confidence interval, DPI (Days post-infestation): days from exposition of fish to U. crassus glochidia; grey markers indicate last and first control in which no juveniles were detected (estimated start and finish point).

3.2 Mitochondrial divergence in the Warkocz River population of the U. crassus species complex

The uncorrected p-distances within the U. crassus species complex revealed substantial genetic differentiation between the two haplotypes obtained from the Warkocz River population (KJ525912 and KJ525917, Kilikowska et al., 2020), which differed by 2.8%. The first haplotype (KJ525912) showed the lowest divergence from U. nanus (p = 0.010), followed by U. vicarius (0.023) and U. bruguierianus, U. carneus, and U. crassus sensu stricto (each 0.028). In contrast, the second haplotype (KJ525917) was nearly identical to U. crassus sensu stricto (p = 0.003) and showed low divergence from U. gontierii (0.010), while distances to other taxa were markedly higher (≥0.023). Both Warkocz haplotypes exhibited substantially greater divergence from more distantly related taxa such as U. sesirmensis (0.057 - 0.062) and U. tumidiformis (0.078 - 0.083). Overall, the observed pattern indicates pronounced mitochondrial heterogeneity within the Warkocz River population, with one haplotype closely affiliated with U. crassus sensu stricto and the other showing the strongest affinity to U. nanus, indicating the presence of two coexisting species from the U. crassus species complex in the Warkocz River: U. crassus sensu stricto and U. nanus.

4 Discussion

Unionid mussels (Unionida) represent one of the classic examples of tight coevolution, in which one group of organisms has effectively made its reproductive success dependent on another group (host fish). Consequently, the prospects for their conservation are closely linked to the availability of suitable host fish. In some studies on host suitability, binary tests (presence/absence) are performed to determine whether a given fish species is capable of transforming glochidia or not (reviewed in Lopes-Lima et al., 2017). However, as recently demonstrated by Dołęga et al. (2025), the influence of fish species that reject larvae may also be indirect and depend on the probability of larval interception by dead-end hosts, as well as on the abundance of both fish and mussel species. In turn, Douda et al. (2014) identified differences in the ability of U. crassus to infest particular host fish species between nearby and recently isolated mussel populations, indicating a significant effect of even slight genetic differences among populations.

In both groups of animals, substantial progress has recently been made in phylogenetic research. Assuming after Douda et al. (2014) that phylogenetic differences, often not detectable without detailed genetic analyses, play an important role in shaping interactions between mussels and fish, studies incorporating recent taxonomic advances gain particular importance. Such an approach may explain many of the observed differences in infestation success and transformation efficiency among mussel and fish populations.

In the case of mussels, species identification can be challenging due to the high phenotypic plasticity of shell morphology (e.g., Zając et al., 2018, Egg et al., 2025) and the lack of clear diagnostic morphological traits distinguishing recently described species (Lopes-Lima et al., 2024). Furthermore, the possibility of hybridization demonstrated in our study further complicates species identification based solely on morphology. In fish, ongoing advances in phylogenetic analyses are leading to taxonomic revisions, revealing previously unrecognized diversity that may translate into differences in host suitability and transformation efficiency. There is therefore a need to effectively reinitiate research programmes aimed at assessing the susceptibility of different fish lineages to glochidial infestation and transformation, incorporating recent advances in genetics.

In the present study, we tested a mussel population whose precise species identity was uncertain, as the taxonomic revision by Lopes-Lima et al. (2024) had not yet been available at the time of sampling. Retrospectively, we can assign these individuals to two lineages (U. crassus and U. nanus) dominating in this part of Europe, which, according to our data, can hybridize. At the same time, following the recognition of a new potential fish host species, C. microstomus, we demonstrated successful transformation of larvae attributed to U. crassus in the traditional sense. This finding is of considerable importance, as all three taxa involved (U. crassus, U. nanus, and C. microstomus) are widespread in this part of Europe, and conservation strategies for the U. crassus complex must necessarily account for their interactions. Although our experiment does not allow us to determine precisely which mussel species is responsible for successful transformation, we can state with confidence that conservation of C. microstomus may contribute to the persistence of the co-occurring members of the U. crassus complex: U. crassus and U. nanus.

The results of the statistical model indicate, that overall transformation of U. crassus/nanus larvae on C. microstomus did not differ significantly from that observed on the well-established host species P. phoxinus. However, the rapid development of larvae on C. microstomus, with a statistically significant advancement in juvenile excystation (juvenile collection on 22nd day post infestation, Fig. 2), suggests that the ontogenetic transition is not hindered in this host, even in comparison to P. phoxinus. Although the overall time frames of juvenile mussel excystation and parasitism duration are similar in both hosts, a more detailed view on excystment timing—at the level of host individuals and even individual glochidia—reveals differences in the temporal patterns of transformation between the two host species. This is consistent with observations reported by other authors (Tab. S4, Supplementary Materials), where both transformation success and temporal parameters are usually lower for Cottus sp. than for P. phoxinus. The generally shorter duration of parasitism is suggested to play a minimal role for U. crassus complex larvae, as according to Taeubert et al. (2012b), glochidia of U. crassus doesn’t grow significantly during encystation, confining benefits of parasitic stage of mussels life to dispersal rather than individual condition.

Given the diversity of methodologies and study objectives in the literature, direct comparisons remain difficult, as already noted by Douda (2013). In study by Douda et al. (2012) C. gobio and P. phoxinus were identified as primary hosts, each exhibiting high transformation rates (>50% of attached larvae). C. gobio hosted on average 365 glochidia per fish, with 57.3% successfully transforming c. 209 juveniles per fish, whereas in our study the mean number of juveniles detached from C. microstomus was 82.2 (SD = 27.27), however, initial infestation levels were not quantified in our study. In contrast, Taeubert et al. (2012a) reported that C. gobio lost approximately 90% of glochidia, classifying it as a relatively unsuitable host. They also noted variation in host suitability among populations of C. gobio, possibly reflecting obscured lineage-specific responses within the C. gobio–U. crassus interaction.

According to in situ observations by Schneider et al. (2018), during the reproductive period of U. crassus, C. gobio carried fewer glochidia per individual than P. phoxinus, despite both species being simultaneously active and exposed. This may suggest that P. phoxinus is a more effective host in situ possibly due to microhabitat preferences and behavioural traits that increase susceptibility to infestation (Ćmiel et al., 2018). A more detailed comparison of literature data with the results of the present study is provided in Table S4 of the supplementary materials.

In the Warkocz River population, Mioduchowska et al. (2016) identified individuals exhibiting discordant mitochondrial lineages: Fcox1 haplotypes corresponded to U. crassus, whereas Mcox1 haplotypes were assigned to U. nanus. These lineages correspond to the northern and southern clades described in Poland and were later recognized as distinct species within the U. crassus complex by Lopes-Lima et al. (2024). This pattern provides clear evidence of interspecific gene flow and hybridization within the complex. Such hybrids may result from secondary contact following historical divergence, but they may also indicate that speciation has proceeded under conditions of incomplete reproductive isolation, allowing ongoing genetic exchange. The coexistence of mitochondrial genomes from two nominal species within single individuals raises important questions regarding species boundaries and the applicability of traditional species concepts in freshwater mussels.

In this study, we demonstrated that C. microstomus is an effective host species for one or both species of the U. crassus complex occurring in Poland (U. crassus and/or U. nanus). Despite taxonomic differences and geographical separation from other Cottus species used in previous studies, C. microstomus supports successful infestation and transformation of glochidia of one or both of these two representatives of the U. crassus complex. Both C. microstomus and the U. crassus complex have wide distributions across diverse river habitats, yet both are highly susceptible to environmental changes, making them prone to co-extinction - a process documented for associated fish–mussel systems (Modesto et al., 2018). Our findings demonstrate that uncovering cryptic host–parasite relationships is not only of academic interest but constitutes a prerequisite for effective conservation planning.

Acknowledgments

The authors would like to thank Mr. Michał Bień, the Director of Polish Angling Association in Kielce, for help in obtaining specimens of C. microstomus from the river Mierzawa. We are also very grateful to dr Michał Nowak, for turning our focus on taxonomy of European Cottus species. We would also like to sincerely thank the anonymous reviewers for their valuable comments, which contributed to the improvement of the manuscript. The study was conducted partially within the framework of LIFE+ project “Life4Delta” (LIFE17-NAT_PL_000018) and in part by statutory funds of the Institute of Nature Conservation, Polish Academy of Sciences. All procedures leading to creation of present manuscript were carried out in accordance with applicable ethics and current legal requirements in force in Republic of Poland.

Supplementary Material

Tab. S1. p-values of pairwise contrasts post-hoc test for performed repeated Measures generalized Linear Model (log, Poisson) of the total number of juveniles collected in each control (aggregated for both species); DPI – days post infestation of fish.

Tab. S2. Results of pairwise contrasts post-hoc test for performed repeated Measures generalized Linear Model (log, Poisson) of the differences in number of juveniles collected in given control between both host species (P. phoxinus and C. microstomus); DPI – days post infestation of fish.

Tab. S3. Estimates of evolutionary divergence between COI sequences of U. crassus complex; references: A - Kilikowska et al. 2020, B - Lopes-Lima et al. 2021, C - Lopes-Lima et al. 2024; the referenced citations are included in the main text of the manuscript, in the “References” section.

Tab. S4. Reference to cited literature on host suitability for larvae of U. crassus complex (comparisons between C. gobio and P. phoxinus); A – results obtained from artificial infestation, N – results obtained from natural infestation; DPI – days post infestation; *- only data from a trial performed concurrently with C. gobio were considered in the table; where possible, data from the present study were presented in the same format as in the cited studies for comparison; the referenced citations are included in the main text of the manuscript, in the “Literature” section.

References

All Figures

Fig. 1 A map of geographical ranges of discussed species and the study location; A) range of occurrence of C. microstomus: a polygon with hatched fill -spatial data was provided by IUCN (MfN Berlin, 2021b) and updated by authors to include drainages of Neman and Venta, as no available spatial data for Krasnaya river drainage was found; and C. gobio: a polygon with dotted fill -spatial data provided by IUCN: (MfN Berlin, 2021a); points indicate exact sites of collection of specimens used in the study by Lopes-Lima et al. (2024) to assess U. crassus complex genetic divergence (limited to the map range): black points - U. crassus, white points – U. nanus, grey points – sites of sympatric occurrence of both species; the red diamond indicates the area of this study; B) the sites of collection of animals in this study: 1- P. Phoxinus specimens used during artificial infestation, U. crassus complex specimens used in artificial infestation, U. crassus complex individuals used for genetic reassessment of the population inhabiting the river Warkocz; 2- C. microstomus specimens used during artificial infestation.

In the text

Fig. 2 The differences in number of juvenile mussels in consecutive counts between studied species: C. microstomus (empty circles) and P. phoxinus (solid squares); points: predicted average, whiskers: 95% confidence interval, DPI (Days post-infestation): days from exposition of fish to U. crassus glochidia; gray markers indicate last and first control in which no juveniles were detected (estimated start and finish point).

In the text

Fig. 3 Cumulative percentage of juvenile mussels collected from fish in consecutive steps: juveniles collected from C. microstomus (empty circles) and P. phoxinus (solid squares); points: predicted average, whiskers: 95% confidence interval, DPI (Days post-infestation): days from exposition of fish to U. crassus glochidia; grey markers indicate last and first control in which no juveniles were detected (estimated start and finish point).

In the text