Pathobiology and first report of larval nematodes (Ascaridomorpha sp.) infecting freshwater mussels (Villosa nebulosa, Unionidae), including an inventory of nematode infections in freshwater and marine bivalves

Little information is available on host-parasite relationships between bivalves and larval nematodes. Herein, we describe nematode larvae (likely stage 2) in the infraorder Ascaridomorpha infecting the foot, intestine, and mantle of a freshwater mussel (Alabama rainbow, Villosa nebulosa [Conrad, 1834]) and detail histopathological changes to infected tissues. A total of 43 live mussels from the South Fork of Terrapin Creek, Alabama, were collected between 2010 and 2014, with 14 sectioned for histopathology and 29 dissected. Of the 14 sectioned mussels, 5 appeared to be uninfected, and 7, 1, and 1 had histozoic infections observed in the foot and intestine, intestine only, and mantle edge and foot, respectively. Twenty-three of 29 (79%) of the mussels dissected were infected by live nematodes, and mean nematode abundance was 8.3 (CL = 5.23–13), with 2 mussels infected with >100 nematodes each. Thus, with a total of 32 of the 43 collected mussels observed with nematodes, overall infection prevalence was 74.4% (CL = 0.594–0.855). The 18S rDNA of this nematode was 99% similar to that of several ascaridids (species of Kathlaniidae Lane, 1914 and Quimperiidae Baylis, 1930) that mature in aquatic/semi-aquatic vertebrates; the recovered 18S phylogenetic tree indicated this nematode from V. nebulosa shares a recent common ancestor with Ichthyobronema hamulatum (Ascaridomorpha: Quimperiidae; GenBank Accession Number KY476351). Pathological changes to tissue associated with these infections comprised focal tissue damage, but a cellular response was not evident. The Alabama rainbow possibly represents an intermediate or paratenic host. Given these results, the nematode is likely not pathogenic under normal stream conditions; however, high intensity infections in the foot could inhibit pedal extension and retraction; which would have demonstrable health consequences to a freshwater mussel. Based on our review of the bivalve mollusc parasite literature, a collective biodiversity of 61 nematodes reportedly exhibit some degree of symbiosis (from commensal to parasitic) with 21 bivalves (28 nematode spp. from 17 marine bivalve spp.; 33 nematode spp. from 4 freshwater bivalve spp.); only four records exist of putatively parasitic nematodes from Unionida. The present study represents the first description of a nematode species that invades the tissues of a Unionidae species.

1. Introduction "Freshwater mussels" are unique bivalve molluscs (Mollusca, Bivalvia, Unionida) because they are parasites of fishes during their larval period and because they use their gills for brooding glochidia, respiration, and filter feeding (Barnhart et al., 2008). North America is historically known for its high species richness of mussels comprising approximately 298 species (Margaritiferidae: 5, Unionidae: 293) (Williams et al., 2017). However, much of this fauna has declined, and as much as 71% of the mussel species across the continental U.S. may be imperiled (Williams and Neves, 1995). The dwindling of mussel populations is largely thought to stem from habitat degradation, toxic https://doi.org/10.1016/j.ijppaw.2019.05.006 Presumed to be P. pectinis (Cobb, 1930) Atlantic Ocean and Mytilus californianus and M. edulis were sampled were sampled, but the authors did not disclose whether nematodes occurred in one or both Mytilus spp.
l Crassostrea virginica and C. rhizophorae were sampled, but the authors did not disclose whether nematodes occurred in one or both Crassostrea spp. m Did not specify if worms were attached to the surface of the mantle or if they were embedded in tissue.    The authors observed nematodes in histological sections, but did not report the infection site.
contaminants or a synergism of these problems (Hughes and Parmalee, 1999;Grabarkiewicz and Davis, 2008). Pathogens and parasitic infections could be contributing factors, but the biodiversity of metazoan parasites and other etiological agents of Unionida is understudied relative to marine bivalves (Lauckner, 1990;Grizzle and Brunner, 2009) and a direct cause-effect relationship between the presence of a given parasite in a freshwater mussel and demonstrable physiological dysfunction is typically lacking in published reports. Based on our review of the bivalve mollusc parasite literature, conservatively 61 nematode species have been reported from 21 bivalve species (28 nematode spp. from 17 marine bivalve spp. [ Table 1], 33 nematode spp. from 4 freshwater bivalve spp. [ Table 2]) totaling 58 sources of literature. However, 33 articles only reported nematodes at the genus level or higher, including 6 sources in which the listed nematode species, genus, family or order was presumed or in which the authors stated that the nematode resembled a named species. Also, 11 articles did not specify a host species or listed the host at the genus level or higher (Tables 1 and 2). Also, four articles represent studies in which marine or freshwater bivalves were challenged with Angiostrongylus cantonensis (Cheng and Burton, 1965;Cheng, 1966;Knapp and Alicata, 1967 [ Knapp and Alicata (1967) additionally examined Crassostrea virginica, Ruditapes philippinarum, and Mactra thaanumi for natural A. cantonensis infections, but did not observe infection. Many reports of nematodes from freshwater bivalves were observations of putatively commensal species from the shell surface or mantle cavity, and we know of at least two studies that have reported free-living nematodes from marine bivalves (Korringa, 1954;Anderson and Bourne, 1960). To the best of our knowledge, there are only four records of putatively parasitic nematodes from Unionida. Clark and Wilson (1912) and Coker et al. (1921) reported Ascaris-like worms infecting the alimentary canal of unspecified North American freshwater mussels from the Maumee River Basin and it is unclear whether these reports are from Unionidae and/or Margaritiferidae and from what localities. Wilson and Clark (1912) reported Ascaris sp. from the stomach of Pyganodon grandis in Indiana. More recently, Lopes et al. (2011) described Rhaphidascaris sp. (as Hysterothylacium sp.) from the pericardial cavity of Rhipidodonta suavidicus (as Diplodon suavidicus [Hyriidae]) from the Aripuana River, Brazil. Histozoic nematodes have principally been reported from marine bivalves and from a variety of tissues. Although some molluscs may serve as intermediate, definitive, or paratenic hosts for nematodes (Grewal et al., 2003;Morley, 2010), literature regarding histozoic roundworms from bivalves largely represents observations of larvae in marine bivalves and there is typically little or no information concerning gross and/or histopathology that would enable us to better understand these host-parasite relationships (e.g., Cobb, 1930;Cheng, 1975a;Sprent, 1977;Vázquez et al., 2006;Lopes et al., 2011).
Much of the literature about parasites in freshwater mussels concerns members of Unionidae, consisting of biodiversity surveys using a gross inspection of tissues or studies having a taxonomic focus (Grizzle and Brunner, 2009). Few investigations have used histology to characterize host-parasite relationships at the cellular level (Antipa and Small, 1971;Huehner and Etges, 1981;Müller et al., 2015;McElwain et al., 2016). These gaps in our knowledge represent a barrier to our understanding of mussel health. Given the above, the lack of histopathological studies on parasites of mussels in freshwater habitats is a bottleneck to our understanding of species declines.
While describing the tissues of Villosa nebulosa towards producing the first unionid histological atlas (McElwain and Bullard, 2014), small nematodes were observed in the foot and other tissues. Herein we describe histopathological changes to the foot, and intestine of V. nebulosa from Alabamaan investigation that represents the first description of a nematode species that invades the tissues of a Unionidae species.

Histological processing
The mussels sampled for histology consisted of 14 individuals. Regarding histological methodology, the valves of each mussel were propped open with wooden dowels to facilitate proper fixation. Mussels were immersed in 10% neutral buffered formalin for 48 h, rinsed in tap water to remove buffer salts, and dehydrated in a graded ethanol series. Formalin fixed mussels were removed from their shells by excising soft tissues from the nacre using a scalpel and divided into pieces by cutting through the visceral mass with a grossing knife. Each sample was processed for routine paraffin embedding using the Tissue-Tek ® Mega-Cassette ® System and a Tissue-Tek VIP E300 (Sakura ® Finetek, Inc., Tokyo, Japan) automated tissue processor. Following tissue processing, pieces of visceral mass were embedded using a Tissue-Tek Thermal Console 4585/7 (Sakura Finetechnical Co., LTD, Tokyo, Japan).
Before sectioning, paraffin blocks were immersed for 1 min in an ice-water mixture immediately before sectioning. Paraffin blocks were sectioned at 4 μm thickness using a Reichert-Jung Biocut 2030 (Wetzlar, Germany), immediately thereafter moved to a Boekel Scientific 145701 lighted tissue floatation bath water (Feasterville, Pennsylvania) at 43°C and pre-mixed with histology adhesive and lifted with forceps. Slides with paraffin sections were placed into a stainless steel 50 slide staining rack and heated to 63°C for 45 min to remove excess paraffin and stained in a Sakura Finetek automated slide stainer with fume hood (Tiyoda MFG, USA, Torrance, California) using Harris's hematoxylin and eosin as per Luna (1968). Stained slides were photographed using a digital single lens reflex camera mounted on a Leica DM 2500 compound microscope (Wetzlar, Germany).

Mussel dissection
A total of 29 individuals were necropsied to obtain nematodes for pathology and for taxonomic diagnosis based on morphological and phylogenetic molecular analyses. Approximately 5 mm 3 pieces of tissue were excised from the foot of live mussels using straight dissecting scissors. Each piece of tissue was placed into a Petri dish filled with deionized water and subdivided into smaller pieces using a scalpel. Small samples of pedal tissue were then wet mounted and gently compressed between two 10 × 8 x ¼ inch plates of glass. Compressed tissues were carefully inspected for roundworms with a Meiji Techno RZDT stereomicroscope (Meiji Techno Co., Ltd., San Jose, California) at a high magnification under bright field and dark field illumination. Infected pieces were removed from the plates and reserved in a small dish filled with deionized water while uninfected pieces were discarded. Each infected piece of tissue was gently teased apart using finetipped forceps. Fibrous tissue was carefully removed from the vicinity of each worm, using 0.20 mm diameter BioQuip Minuten pins, each mounted in a BioQuip pin vise. Individual worms were transferred to a separate dish of deionized water and allowed to crawl freely to remove any attached debris.

Nematode processing and taxonomic identification
For a morphological diagnosis, worms were fixed in a small dish A. McElwain, et al. IJP: Parasites and Wildlife 10 (2019) 41-58 containing glacial acetic acid until they became straightened, then transferred to a vial containing 70% ethanol. Nematodes intended as whole-mounts were photographed with the aid of a stereo-dissecting microscope and fiber optic light source, rinsed with distilled water, immersed in 95% glacial acetic acid for 5-10 min, fixed and cleared in 5 parts glycerin plus 95 parts 70% ethanol (EtOH) ("70 + 5"), mounted on glass slides using glycerin jelly, and studied with a Leica DM 2500 microscope with differential interference contrast (DIC) optical components. For gene sequencing, worms were fixed in 2.0 ml cryo-storage vials containing 95% ethanol or RNAlater™ and stored at −20°C. Using the pooled (4) EtOH-preserved and microscopically-identified nematodes from Alabama rainbow, total genomic DNA (gDNA) was extracted using DNeasyTM Blood and Tissue Kit (Qiagen, Valencia, California, USA) as per the manufacturer's protocol with one exception: the proteinase-K incubation period was extended overnight and the final elution step used 100 μl of elution buffer to increase the final DNA concentration. Inhibitors were removed from extracted DNA using OneStepTM PCR Inhibitor Removal Kit (Zymo Research, Irvine, California, USA). Amplification and sequencing of the small subunit ribosomal DNA (18S) used the set of primers described in Floyd et al. (2005). PCR amplifications were performed using a total volume of 25 μl with 2 μl of DNA template, 0.2 μM of each primer along with 1 × buffer, 3 mM MgCl 2 , 0.2 mM dNTP mixture, and 0.15 μl Taq polymerase (5 U/μl) (Promega, Madison, Wisconsin, USA). The thermocycling profile comprised 5 min at 94°C for denaturation, 35 repeating cycles at 94°C for 30 s for denaturation, 54°C for 30 s for annealing, and 72°C for 1 min for extension followed by a final 10 min at 72°C for extension. All PCR reactions were carried out in a MJ Research PTC-200 (BioRad, Hercules, California, USA). PCR products (10 μl) were verified on a 1% agarose gel and stained with ethidium bromide. PCR products were purified by microcentrifugation with the QIAquick PCR Purification Kit (Qiagen, Valencia, California, USA) according to manufacturer's protocols except that the last elution step was performed with autoclaved nanopure H 2 O rather than with the provided buffer. DNA sequencing was performed by ACGT, Incorporated (Wheeling, Illinois, USA). Reactions were sequenced using BigDye terminator version 3.1, cleaned with magnetic beads (CleanSeq dye terminator removal kit), and analyzed using an ABI 3730 XL or 3730 Genetic Analyzer. Sequence assembly and analysis of the chromatogram was performed with Geneious version 11.0.5 (http://www.geneious.com). All nucleotide sequence data were deposited in GenBank.
A preliminary NCBI BLAST (https://blast.ncbi.nlm.nih.gov) search showed high genetic similarity (> 97%) between the new sequence and those of Cosmocercoidea and Seuratoidea; therefore representatives of these superfamilies were used in the phylogenetic analysis. The dataset for phylogeny consisted of 23 taxa belonging to Atractidae, Cosmocercidae and Kathlaniidae (Cosmocercoidea), and Cucullanidae, Quimperiidae and Seuratidae (Seuratoidea) plus Zeldia punctata (Thorne, 1925) (Cephaloboidea: Cephalobidae) as outgroup, chosen according to previous studies (Choudhury and Nadler, 2016;Pereira and Luque, 2017;Sokolov and Malysheva, 2017). Sequences with less than 800bp and without genetic overlapping were excluded. Except for Cucullanidae, which is considered monophyletic (Choudhury and Nadler, 2016), all available taxa assigned to Cosmocercoidea and Seuratoidea were included in the analysis. Sequences were aligned using T-Coffee (Notredame et al., 2000), then evaluated by the transitive consistency score, to verify the reliability of aligned positions and, based on score values, ambiguous aligned positions were trimmed (Chang et al., 2014). The phylogenetic tree was generated in MRBAYES (Huelsenbeck and Ronquist, 2001), using Bayesian inference and nodal supports estimated by Bayesian posterior probability after running the Markov chain Monte Carlo (2 runs 4 chains) for 4 × 10 6 generations, with sampling frequency every 4 × 10 3 generations and discarding the initial ¼ of sampled trees (1 × 10 6 ) as burn-in. The model of evolution (TIM3+I + G) and its fixed parameters for phylogenetic reconstruction were chosen and estimated under the Akaike informative criterion with jModelTest 2 (Guindon and Gascuel, 2003;Darriba et al., 2012).

Statistical analyses
Prevalence and mean abundance values defined according to Bush et al. (1997), were calculated in the Quantitative Parasitology Program (QPweb Version 1.0.14, Reiczigel et al., 2019). The 95% confidence limits (CL) of prevalence were calculated using the Sterne's exact method, and those for mean abundance were calculated using bootstrapping with 2000 replications.

Prevalence, abundance, and site of infections
From the total sample of mussels, the prevalence of nematodes was 74.4% (CL = 0.594-0.855, n = 43). Of the 14 individuals examined with histology, 9 were infected. Of these 9 individuals, 7 mussels displayed nematodes in the foot and intestine, one individual presented nematodes only in the intestine, and one individual had nematodes in the mantle edge and foot. From the sample of 29 individuals necropsied, 23 were infected. Based on dissections, intensity ranged from 0 to 39. Mean abundance was 8.3 (CL = 5.23-13, n = 26). However, two of the 29 individuals were infected with > 100 nematodes, and one additional infected individual was excluded from this analysis because its tissues were placed directly into a 2.0 ml cryo-storage vial containing RNAlater™. Nematodes were more abundant in the foot than in other tissues, and therefore we focused our attention on extricating nematodes from foot tissues during dissections.

Histopathology
The foot of V. nebulosa is mainly composed of bundles of somatic muscle that become branched near the ventral margin. Hematoxylin and eosin stained sections of foot revealed the presence of two distinct groups of cells having a granular cytoplasm. In the medial portion of the foot, there are pale, basophilic granulocytes while darker, violet, basophilic granulocytes are located laterally and ventrally (Fig. 1). Nematodes were principally located medially in the ventral region of the foot (Fig. 2). Pedal musculature of V. nebulosa is well organized with large fascicles medially and the myofibers overlap as they near the ciliated epithelium (Fig. 3). Infected tissue typically displayed an irregular, medial-ventral gap in the somatic musculature containing roundworms (Fig. 4). Nematodes were typically concentrated in this area, but were occasionally isolated in more dorsal and lateral regions of the foot. At a higher magnification, myofibrils appear to be densely packed and intricately interlaced (Fig. 5). Worms were typically arranged in different orientations. Surrounding the focus of worms, the myofibers and granulocytes were intact and histologically indistinguishable from uninfected tissue (Fig. 6). At high magnification, roundworms were closely positioned to myofibrils and the tissue locally conformed to the curvature of the worms. The medial aspect of the infected area contained a small amount of fibrous debris (Fig. 6). A cellular response to the nematodes was not observed.
Intestinal nematodes were located within the epithelium of the fourth limb of the intestine, which is characterized by a major typhlosole in the dorsal aspect of the visceral mass (McElwain and Bullard, 2014). Uninfected intestinal tissue was characterized, as typically, by a simple, ciliated columnar epithelium. Some parts of this epithelium may be pleated. The columnar epithelium also contains teardropshaped columnar cells that contain eosinophilic granules. The epithelium is surrounded by a lamina propria and loose connective tissue resembling adipocytes (Fig. 7). Nematodes always appeared to be threaded through the columnar epithelium and infected cells were      intact with little cellular changes apparent except for a small, irregular gap surrounding the wormspotentially an artifact of histological processing (Fig. 8).
The literature contains limited information about tissue damage or potential host responses associated with nematodes infecting bivalves. Pathological changes to infected tissues have mainly been reported as localized discolorations or cysts in marine bivalves (Table 1). Shipley and Hornell (1904) reported specimens of Ascaris meleagrinae encysted in gonad, mantle, stomach and mouth and Echinocephalus uncinatus encysted in the adductor of Pinctada imbricata. Shipley and Hornell (1904) also stated that E. uncinatus may occasionally become embedded in the nacre. Encysted E. uncinatus have also been reported from the adductor of Placuna placenta (Willey, 1907) and from unspecified tissues of Pinna sp. (Baylis and Oxon, 1920). In one case, an unidentified nematode appeared to have been encased within a pearl in P. placenta (Herdman and Hornell, 1906). Tissue discolorations associated with nematodes have been documented by Shipley and Hornell (1904), Ko et al. (1975), Lester et al. (1980), Lichtenfels et al. (1980), McLean (1983, Deardorff (1989). Pink cysts in the adductor of Pinctada margaritifera contained larvae of E. uncinatus (Shipley and Hornell, 1904). Ko et al. (1975) observed green spots around E. sinensis infecting the digestive diverticula, stomach, intestine, and mantle of Magallana gigas. Lester et al. (1980) reported a brown discoloration to the adductor of Spisula solidissima infected with Sulcascaris sulcata. S. sulcata was also associated with an unspecified color change to the gonad of A. gibbus (Lichtenfels et al., 1980). The gonad of A. gibbus typically becomes bright orange when they are ready to spawn (Miller et al., 1979). Yellow-brown spots occurred in the adductor of A. ventricosus infected with E. pseudouncinatus (McLean, 1983). S. sulcata was also associated with a yellow-brown discoloration of the adductor of A. gibbus (Deardorff, 1989). Brownish spots indicated the presence of encapsulated nematodes in Tagelus plebeius (Vázquez et al., 2006). It remains uncertain whether tissue discolorations are the result of cellular damage, immunological responses to parasites or secretions from parasites. In some cases, nematodes may be pigmented, but it is unclear if this coloration contributes to the coloration of infected tissue (Shipley and Hornell, 1904;Cummins, 1971;Perkins et al., 1975;Lichtenfels et al., 1978;Cannon, 1978). Other observations of gross pathological changes to tissue have included caseous tissue near the adductor of A. balloti associated with larval ascaridoid worms (Cannon, 1978). Also, Payne et al. (1980) reported a slight thickening of the infection site associated with nematodes presumed to be a Sulcascaris sp. in S. solidissima, and some worms occurred in watery cysts. In the present investigation, we did not observe any obvious gross pathological changes to the tissues of Villosa nebulosa infected with nematodes.
Parasitological investigations, in which histology was a focus, have provided little insight into the cellular changes to host tissues that occur during a roundworm infection. Burton (1963) observed small nematodes measuring 75 μm in diameter within the digestive diverticulum. Burton (1963) also reported dense concentrations of leukocytes, encapsulation responses, and gastric and intestinal ulcers, but it was unclear whether such observations were associated with nematode infections or not. Cheng (1966) experimentally infected Crassostrea virginica with Angiostrongylus cantonensis. A preponderance of hemocytes was observed within and around hemolymph vessels of infected oysters as compared to uninfected oysters. Leukocytes surrounded nematodes between 10 and 14 days post infection. Cheng (1967) reported unidentified nematodes coiled near digestive diverticula of C. virginica, but no other details were provided. Cheng (1975a) reported E. crassostreai from Magallana gigas. E. crassostreai did not cause appreciable histopathological changes to the gonoduct lining, but a 0.15 mm tunic of connective tissue fibers, hemocytes, myofibers and Leydig cells were observed around the gonoducts. Nematodes infecting the ovaries were associated with displaced, shrunken, ruptured or compressed ova. In a follow-up study, Cheng (1975b) reported brown cells in the reaction complex, but the function of these cells remains unclear. Ko et al. (1975) reported E. sinensis from M. gigas. Worms were located in Leydig tissue and gonoducts of male and female oysters. Intensity of tissue reactions varied from no apparent host response to a conspicuous response. Host responses included infiltration of amoebocytes around the worms and extensive fibroplasia. Infected oysters also displayed enlarged gonoduct lumen, desquamation, erosion of ciliated epithelium and metaplasia of pseudostratified columnar epithelium into cuboidal or squamous epithelium. Cannon (1978) observed an encapsulation response to larval ascaridoid worms infecting the adductor of A. balloti. An encapsulation response occurred in histological sections of the adductor of A. gibbus infected with S. sulcata (Deardorff, 1989). Murchelano and MacLean (1990) provided an update on the histopathological changes associated with nematodes in bivalves reported by Burton (1963), Cheng (1966), Lichtenfels et al. (1978), and Perkins et al. (1975) with higher resolution images of S. sulcata infecting S. solidissima and unidentified nematodes infecting C. virginica. Kim and Powell (2004) observed unidentified nematodes infecting the digestive gland, visceral mass between body wall and underlying muscle tissue, foot, muscle tissue, gill of S. solidissima. Nematodes were frequently associated with hemocytic infiltration (Kim and Powell, 2004). Vázquez et al. (2006) reported nematodes infecting T. plebeius were either free or encapsulated in tissues. The capsule consisted of a dense aggregation of hemocytes and sometimes bundles of muscle fibers appeared to comprise the outer capsule wall.
Irrespective of whether V. nebulosa individuals exhibited a low or high infection intensity, there did not appear to be a cellular response to  nematode presence. Metazoan parasites infecting bivalves are sometimes associated with an encapsulation response characterized by hemocytes surrounding the parasite and, in some cases, fibrosis (Pauley and Becker, 1968;Cheng and Rifkin, 1970;Huehner and Etges, 1981;Feng, 1988). The lack of a host response associated with nematodes in V. nebulosa may be indicative of some form of immunological suppression (Sorci et al., 2013).
The minute size of and lack of discernible genitalia and demonstrable cuticular features in our specimens indicated that they were larval specimens (Moravec, 1998). The presence of a genital anlagen (primordium) comprising a single small cell strongly suggested that they represented a second stage larva, L2 because the genital anlagen develops and differentiates late in the L3 or during the L4. Further, given that they comprised unencysted, unencapsulated larvae infecting a mollusc (rather than a vertebrate), we suspect that they may comprise an L2. As such, it was not possible to diagnose them to a family; however, noteworthy is that they demonstrated morphological features consistent with larval specimens of species of both Seuratoidea and Cosmocercoidea (Anderson, 2000; i.e., Cosmocercoidea: having esophagus with three discernible sections, including an anteriorly distinct pharynx and slender tail; Seuratoidea: esophagus lacking bulb at its base). However, confident diagnosis based upon morphology alone is tenuous because the larval types for species of these superfamilies have not been morphologically diagnosed. These larval nematodes may represent an innominate taxon.
Regarding the taxonomic identification and phylogenetic placement   of the nematode larvae herein, they claded within a well-supported assemblage of quimperiid and kathlaniid species (all in the Ascaridomorpha) and had morphological features of both groups. As such, combined with the available morphological and life history evidence, we conservatively identified our nematode larvae as a species within the nematode infraorder Ascaridomorpha (i.e., Ascaridomorpha sp.). The biodiversity of nematodes associated with bivalve molluscs is poorly understood. Most records are from commercially important marine species. Many records of nematodes associated with freshwater bivalves are from Dreissena polymorpha, an invasive species that has become established in many water bodies throughout Europe and North America (Karatayev et al., 2000(Karatayev et al., , 2003Quinn et al., 2014). The lack of a specific identity for nematodes associated from bivalves reported in the literature may be an indication that nematodes are difficult to identify as larvae because species-specific morphological characters are underdeveloped. Additionally, it may not be feasible to identify nematodes from histological sections of intermediate hosts potentially because their small size makes them difficult discern during routine microtomy and because worms may be coiled or arranged in a sinuous manner making it laborious to characterize their entire anatomy (e.g., Burton, 1963;Deardorff, 1989;Murchelano and MacLean, 1990;Vázquez et al., 2006).
Our observations represent the first description of a nematode species that invades the tissues of a Unionidae species. Clark and Wilson (1912), Wilson and Clark (1912), and Coker et al. (1921) reported "Ascaris sp." and "Ascaris-like" nematodes infecting North American unionids and or margaritiferids; however, these authors did not morphologically diagnose the nematodes and, to the best of our knowledge, no specimen was deposited in a curated museum. Although we attempted to morphologically diagnose our nematode specimens infecting V. nebulosa to family, the morphological features of these larval specimens were inadequate to do so. Moreover, the molecular phylogenetic results were equivocal, to some extent, based on the small fragment (835 bp) and the self-evident systematic revision needed among the Quimperiidae, Kathlaniidae and other families of Cosmocercoidea and Seuratoidea, or whether these superfamilies are valid (all of which were recovered as paraphyletic or polyphyletic in the 18S phylogeny). While the prevalence among the sampled mussels was high, we do not know if these nematodes are locally abundant in the South Fork of Terrapin Creek or how common these nematodes are throughout the range of V. nebulosa.

Conclusions
Since this study was based on field collections, there are several gaps in our understanding of the host-parasite relationship. The route of infection is presently indeterminate; we are uncertain if eggs or larvae are ingested by mussels or if larvae penetrate the integument from the surrounding sediment and migrate to other tissues. Secondly, we do not know whether V. nebulosa represents a paratenic or intermediate host.
A wide range of predators, some of which are potential definitive hosts, may feed on freshwater mussels. Therefore, mussels may be indicators of ecosystem health (Haag, 2012). Additionally, it is unclear to what extent nematodes may impair the function of host tissues. For example, in mussels in which we observed a large number of nematodes, we are uncertain if pedal extension and or retraction was impaired. Unionids use their foot to burrow into specific sediments and the impairment of pedal extension and retraction could make them vulnerable to becoming dislodged.  Table 3 Species whose sequences of the 18S rDNA were retrieved from GenBank and used for phylogenetic analysis.