GENETIC AND TRAIT VARIABILITY OF GYRINICOLA REVEALS THE EXISTENCE OF AT LEAST FOUR SPECIES WITHIN THE UNITED STATES

Tadpole-infecting pinworms of the genus Gyrinicola Yamaguti, 1938 occur in larval anurans through a quasi-global distribution. These monoxenous pinworms are a rarity, because they are among the few oxyurids to complete their life cycles in an aquatic environment (Anderson, 1988). Six species are currently recognized as valid, including Gyrinicola batrachiensis (Walton, 1929) in North America, Gyrinicola tba (Dinnik, 1930) and Gyrinicola chabadamsoni Planade and Bain, 2008 in Europe, Gyrinicola japonica Yamaguti, 1938 in Japan, Gyrinicola chabaudi Araujo & Artigas, 1983 in Argentina and Brazil, and Gyrinicola dehradunensis Maity, Rizvi, Bursey & Chandra, 2019 in India. This group of pinworms was originally recognized as a monotypic genus and the sole member of the family Gyrinicolidae within Oxyuroidea (Yamaguti, 1938). Since then, the taxon has experienced several hierarchical shifts (Yamaguti, 1938, 1961; Skrjabin et al., 1960; Chabaud, 1965; Petter and Quentin, 1976; Adamson, 1981a; Hasegawa and Asakawa, 2004; Sata and Nakano, 2020; Walker et al., 2023). The taxonomic volatility presented above resulted from a misinterpretation of the structure of the spicule and the number of ovaries in the type species G. japonica. In their reevaluation of the morphology of the Japanese tadpole pinworm, Sata and Nakano (2020) redescribed the species as possessing a singular spicule with a sclerotized V-shaped gubernaculum, a characteristic typical of Pharyngodonidae. Further, they also noted the presence of 2 ovaries. The phylogenetic reconstruction associated with the morphological reevaluation (Sata and Nakano, 2020) placed the taxon within the Oxyuroidea and based on the topology of the single terminal taxon these authors suggested the resurrection of the Gyrinicolidae.

Species diversity of tadpole-dwelling pinworms has been documented across North America previously. Walton (1929) described Pharyngodon batrachiensis Walton, 1929 from tadpoles of northern leopard frogs (Lithobates pipiens) collected in Michigan. Walton (1933) later described Pharyngodon armatus Walton, 1933 from tadpoles of northern leopard frogs (L. pipiens) and green frogs (Lithobates clamitans) from Virginia, but Walton did not describe males in either study. The male of P. armatus was described by Holomon (1969) based on material collected from green frogs (L. clamitans) from Ohio. However, while studying tadpoles from eastern and central Canada, Adamson (1981a) assigned his specimens to Gyrinicola and considered them identical to P. batrachiensis and P. armatus, which became synonyms of G. batrachiensis (Walton, 1929). Recent studies have documented considerable morphological variability in specimens assigned as G. batrachiensis across the central plains of the United States (Rhoden and Bolek, 2011; Pierce et al., 2017). An assessment of the genetic variability of specimens identified as G. batrachiensis across the eastern portion of the United States allowed Walker et al. (2023) to identify greater-than-expected genetic divergence, suggesting the existence of cryptic species.

Herein we conduct an integrative investigation of Gyrinicola diversity, which utilizes molecular and morphological information collected from several locations and multiple host species to identify possibly cryptic species and reveal overlooked diagnostic characters (Nadler and Pérez-Ponce de León, 2011). The integrative approach is especially important when studying nematodes, and other micrometazoans, that feature a limited set of organs and are relatively small, a combination that causes difficulty in species determinations reliant on physical characteristics alone (Kirillova and Kirillov, 2020; Alnaqeb et al., 2022a, 2022b). A robust phylogeny can, therefore, serve as an excellent framework for character interpretation and assist in determining species synergistically with physical characteristics (Dayrat, 2005; Will et al., 2005; De Sousa et al., 2019). In utilizing these 2 lines of evidence an integrative approach unifies criteria as suggested by De Queiroz (2007) in supporting species determination and delimitation.

Tadpole and nematode collection followed the methods detailed in Walker et al. (2023), per Southern Illinois University, Carbondale Institutional Animal Care and Use Committee Protocol 21-015. Sampling sites were highly variable ranging from small urban water features such as fountains and ditches to larger more rural water bodies such as ponds and lakes. In larger waterbodies, which featured more habitat heterogeneity, areas with dense algae and aquatic vegetation tended to yield more tadpoles, as these areas have ample resources for feeding. Nematodes selected for DNA extraction were imaged before dissection on a Leica DFC295 (Leica Microsystems, Buffalo Grove, Illinois) connected to a Leica DM25000 (Leica Microsystems) through Leica camera capture software, with measurements given in micrometers. Whole nematodes were washed in 1X PBS buffer. The washed nematodes were then sectioned into 3 longitudinal sections; the anterior and posterior ends of the nematodes, which feature most of the diagnostic characters, were preserved as hologenophores, and the internal sections were used for DNA extractions. The lots selected for this study included infrapopulations with abundances >5; this allowed the morphometric characterization of several organisms from a single host and attempted extractions in at least 2 specimens per host. Names of the localities and their abbreviations are provided in Table I. Specimens were deposited in the Harold W. Manter Laboratory of Parasitology (HWML) in Lincoln, Nebraska, and the U.S. National Parasite Collection, National Museum of Natural History, Smithsonian Institution, Washington, D.C. (USNM). Collection numbers generated for specimens deposited in scientific collections are provided in Table II. All descriptions follow the terminology of Chitwood and Chitwood (1974) and Petter and Quentin (1976), especially for the structures of the cephalic cuticle and spicule.

We followed protocols detailed in Walker et al. (2023), except for those worms to be sequenced in an Illumina platform. In those cases, the elution buffer used to store DNA was replaced with 10 mM Tris-Cl (Fisher BioReagents, Fair Lawn, New Jersey), to prevent inhibition of reactions induced by EDTA at concentrations >0.1 mM.

Methods for the amplification of the ribosomal genes 18S and 28S also follow Walker et al. (2023). For ITS1, 5.8S, and ITS2, hereafter referred to as ITS, amplification was performed using the forward primer ITS-18S 5′-TTGATTACGTCCCTGCCCTTT-3′ (Vrain et al., 1992), and the reverse primer NC2 5′-TTAGTTTCTTTTCCTCCGCT-3′ (Gasser et al., 1993) with the following cycling conditions: 94 C for 5 min, 35 cycles of (94 C for 30 sec, 53 C for 30 sec, and 72 C for 60 sec), and 72 C for 7 min. Amplicons were sequenced using commercial providers (Eurofins Genomics, Louisville, Kentucky, and MCLAB, South San Francisco, California), using the amplifying primers above and the internal primer NC5 5′-ACGTCTGGTTCAGGGTTGTT-3′ (Gasser et al., 1993).

Extractions for Next-Generation sequencing were processed at the Roy J. Carver Biotechnology Center at the University of Illinois at Urbana-Champaign (Urbana, Illinois) for library preparation and shotgun sequencing. Library preparation consisted of size selection of 250-nt insert lengths. The average insert size following assembly was 189 nt. Sequencing was performed on an Illumina NovaSeq 6000 (Illumina, San Diego, California), which generated an average of 5,000,000 (250 paired-end reads) per sample for a total of 11 sequence libraries for Gyrinicola specimens, and 1 sequence library for a specimen, cf. Spauligodon giganticus was used as an outgroup due to its placement in previous studies (Sata and Nakano, 2020; Walker et al., 2023).

Raw PCR sequences were assembled in Sequencher v.3.5 (Gene Codes Corporation, Ann Arbor, Michigan). Illumina data were assembled using Geneious Prime v.2020.1.2 (Biomatters Inc., Newark, New Jersey). To retrieve a nearly complete nrDNA cistron from Illumina data, a reference assembly of 1 sample (RC01MO05 referenced to 28S Sanger-sequence data of the same specimen) was performed using the medium–low sensitivity setting with 25 iterations; mapped reads were saved and then assembled to their consensus with high sensitivity settings to create a final reference sequence used for all other samples. We mapped the remaining libraries to the nrDNA reference sequence using 10 iterations of medium–low sensitivity settings, followed by 5 iterations of used reads mapped to their own generated consensus sequences with high sensitivity settings. All cistrons were then evaluated using a Geneious alignment with free end gaps and 65% similarity against PCR data from worms in their infrapopulation. The regions that aligned with the 18S, ITS, and 28S were then extracted and used in future tree building.

To retrieve mtDNA from the Illumina data, a reference assembly of 1 sample—RC01MO05 referenced to the oxyurid Syphacia obvelata, GenBank accession number KT900946 (Wang et al., 2016)—was performed using the medium–low sensitivity setting with 25 iterations, which ultimately produced 3 consensus sequences of varying lengths. All raw reads were then mapped against the 3 consensus sequences to extend the ends. The largest consensus structure was retained for future use. To verify the consensus structure, the raw data for RC01MO05 were then reference assembled against the largest consensus sequence using the high sensitivity setting and 10 iterations. After visual evaluation, a 50% majority-rule consensus sequence was generated; here this sequence is referred to as the mtDNA reference sequence. The remaining nematodes were then mapped to the mtDNA reference sequence using 10 iterations of medium–low sensitivity settings, followed by 10 iterations of used reads mapped to their own generated consensus sequences, this time using high sensitivity settings. The retrieved mtDNA sequences were then annotated against a user reference library containing the published mitochondria of Enterobius vermicularis (EU281143), Aspiculuris tetraptera (KT764937), Wellcomia compar (MW059037), and Wellcomia siamensis (GQ332427; Kang et al., 2009; Park et al., 2011; Wang et al., 2016). Five gene regions annotated in all specimens as cox1, cox3, nad3, nad4, and nad5 were then extracted. The extracted mtDNA genes were then evaluated via alignment to one another and translation to amino acids to ensure no errant stop codons were present. All taxa including outgroups are members of the clade III nematodes as defined by Blaxter et al. (1998). As the placement of the Gyrinicola was previously investigated by Sata and Nakano (2020), and Walker et al. (2023), the number of species in this analysis was limited to Gyrinicola and a select few oxyurids. The limitation was done to reduce the amount of noise introduced by large data sets and focus on the clades within the North American Gyrinicola.

Sequences for 28S, ITS, and 18S were concatenated in Mesquite and aligned on the CIPRES Scientific gateway (Miller et al., 2010) using MAFFT on XSEDE through a Q-INS-i alignment (Katoh and Standley, 2013; Maddison and Maddison, 2020). The Q-INSI-i alignment utilizes a CONTRafold method to predict secondary structure via the use of conditional log-linear models to predict parameters of folded structure (Do et al., 2006). The concatenated nrDNA data matrix features 31 operational taxonomic units (OTUs) and 7,097 characters with 974 parsimony informative characters.

The genes cox1, cox3, nad3, nad4, and nad5 were concatenated in Mesquite and aligned using the MAFFT version 7 online sever using default settings and auto selection of alignment method (Kuraku et al., 2013; Katoh et al., 2019). The alignment method G-INS-I, which is a highly accurate progressive alignment type, was employed to generate the alignment. This alignment method offers an additional iteration for refinement (Katoh et al., 2013). The concatenated mtDNA matrix features 16 OTUs and 5,621 characters with 2,533 parsimony informative characters.

A concatenation of the nrDNA matrix and mtDNA matrix was also performed on the previously aligned matrices in Mesquite. The concatenated matrix features 14 OTUs and is 12,718 characters with 2,843 parsimony informative characters.

Model testing was performed using jModelTest2 on XSEDE on the CIPRES scientific gateway (Miller et al., 2010; Darriba et al., 2012). The Akaike information criterion (AIC) was used as the selection method for finding the best model. The model set was restricted to 3 to allow for analysis using MrBayes (see the following). All matrices returned GTR+I+G as the best mode of evolution. Maximum-likelihood inference and bootstrap support for the branches were calculated using 1,000 replicates constructed on the CIPRES scientific gateway using RAxML-HPC2 on XSEDE (Stamatakis, 2005). Estimation of the posterior clade probability was performed using MrBayes on XSEDE via the CIPRES scientific gateway (Ronquist et al., 2012). Four chains were set to run for 30,000,000 generations with resampling every 1,000 generations and a burn-in of 25%. Convergence was assessed by examining the potential scale reduction factor and visualization of the generated TRACE plot. The resulting RAxML and Bayesian trees were visualized using FigTree v.1.4.4 (Rambaut, 2012) and Seaview v.5.05 (Gouy et al., 2010). Final trees showing both posterior probability and bootstrap values, as well as clade labeling, were constructed in the GNU Image Manipulation Program (GIMP) v.2.1.0 (The GIMP Development Team, 2019). To evaluate distances uncorrected P-distance matrixes were also generated using Phylogenetic Analysis Using Parsimony (PAUP*) v.4.0a (Swofford, 2002).

The resulting phylogeny of all concatenated genes was used as an input for a Bayesian implementation of the Poisson tree processes (bPTP). This method was selected because bPTP species delimitation models speciation by using branching patterns and the number of substitutions between and within branches to predict species. Given that we had a large tree with several gene regions for evaluation, this method seemed efficient. The species delimitation analysis was conducted on the online server described by Zhang et al. (2013). The input tree was unrooted, and distant outgroups were automatically removed by the program. The number of MCMC generations was set to 500,000, thinning was set to 100, burn-in was set to 0.25, and a seed of 123 was used. The convergence of chains was evaluated via a likelihood trace plot provided by the server.

Whole nematodes were cleared on slides with either lactophenol or glycerin, depending on the number of nematodes recovered from a sampling locality. Lactophenol tended to clear nematodes more rapidly and with less risk of crumpling the cuticle; however, nematodes placed in lactophenol could not be used for DNA analysis; therefore, when few nematodes were recovered from a locality glycerin was used, which allowed mounted nematodes to function as possible backups for extraction if any of the other extractions from the same locality failed to produce useable DNA. Following clearing nematodes were imaged and measured with specialized software (Leica Microsystems, Buffalo Grove, Illinois); all measurements are in micrometers. Measurements for morphological analysis were standardized by the total length of the specimen to create a proportion data set, which controlled for size variation in specimens. Specimens were also assigned a Gyrinicola species number (1–4) based on their grouping in the bPTP analysis of the concatenated data set.

Because of the presence of multiple groupings (potential species of Gyrinicola), and multiple dependent variables (morphological characteristics), multiple analysis of variance (MANOVA) paired with a canonical variate analysis was chosen as the means of data testing and interpretation. MANOVA accounts for multiple dependent variable comparisons of variance without increasing the global Type I error rate (Wilks, 1932). A canonical variate analysis was paired with the MANOVA to illustrate differences between groups. This combination reduces variables into 2 canonical dimensions, which explain the variance in the data; the variance can then be visualized in centroids of the putative species groupings. Tests were performed using SAS® ONDEMAND FOR ACADEMICS using proc glm (general linear models) features, and the macros “Equate” and “Canplot,” which produce plots for canonical analysis (Friendly, 2001; SAS Institute Inc., 2014). Separate analyses of proportion data were performed on female and male data; programs are provided in Suppl. Materials S1. An analysis of raw data was also performed and is provided in Materials S1. A post hoc power analysis was calculated using the program G*Power (version 3.1.9.6), to determine the achieved power of results utilizing the sample size, group number, variable number, and Pillai Trace generated by the MANOVA (Faul et al., 2009). A post hoc Tukey test was also performed on the data sets using the SAS software above to aid in the interpretation of characteristics; a Bonferroni correction was used for this Tukey test, resulting in a highly conservative alpha value of 0.007 rather than 0.05; the code for this implementation is given in Materials S1 (Dunn, 1961).

Analysis of the nrDNA data matrix shows maximum support for the monophyly of North American Gyrinicola and places the group with maximum support as a sister to a newly sequenced cf. Spauligodon giganticus (Fig. 1). The North American Gyrinicola form 3 major clades, with clades 1 and 2 being reciprocally monophyletic (0.99/82) and clade 3 acting as sister to the other 2 clades (1.00/84.0). Clade 1 contains members from Michigan, the type locality of G. batrachiensis, and New York. Clade 2 contains members from Florida, Missouri, Iowa, Illinois, Kentucky, and Virginia, the type locality for P. armatus. Clade 3 contains nematodes from Illinois, Oklahoma, Iowa, and Washington. There is no resolution within clades and all clades include pinworms that infect several frog species.

The mtDNA also yields maximum support for the monophyly of North American Gyrinicola and places the group sister to the cf. Spauligodon giganticus (1.00/99.6; Fig. 2). In this topology the North American Gyrinicola is also split into 3 clades and they are congruent with the topology shown in Figure 1. The support within the clades is strong for most internal clades. All 3 clades have maximum support, though in this mtDNA analysis, a nematode from Florida is placed sister to the rest of the worms in clade 2 (1.00/100) and a nematode from Urbana (Illinois) is placed sister to the rest of the worms in clade 3 (1.00/100).

The analysis of the concatenated nrDNA and mtDNA data matrix also shows maximum support for the monophyly of North American Gyrinicola and features the same resolution into 3 clades present in the independent analysis of nrDNA and mtDNA. However, the resolution for the splitting of clades 1 and 2 is slightly weaker (0.94/50.7; Fig. 3).

Uncorrected P distance matrix of nrDNA showed intraclade variation among Gyrinicola clade 1 of 0–0.3%. The variation among Gyrinicola clade 2 is between 0 and 0.8%. Finally, variation among Gyrinicola clade 3 ranges from 0 to 2.4%. The average genetic distance between clades 1 and 2 is 2.3%; the average genetic distance between clades 1 and 3 is 3.4%; finally, the average genetic distance between clades 2 and 3 is 2.7% (Table III). Full results are hosted on OpenSIUC (https://opensiuc.lib.siu.edu/zool_data/21/).

Uncorrected P distance matrix of mtDNA showed intraclade variation among clade 1 Gyrinicola of 3%. Variation between 0 and 1.8% was seen among clade 2 Gyrinicola. Variation between 0.1 and 8.1% was seen among clade 3 Gyrinicola. The average genetic distance between clades 1 and 2 is 19.6%; the genetic distance between clades 1 and 3 is 19.2% (Table IV). Full results are hosted on OpenSIUC (https://opensiuc.lib.siu.edu/zool_data/21/).

Species delimitation based on the topology resulting from the concatenated genes suggested a range of 2–11 species present within the analysis with the highest posterior probability tree containing 4 species. Clade 1 was suggested as a putative species with low Bayesian support (0.52). Clade 2 was also suggested as a species with higher Bayesian support (0.82). Clade 3 was suggested to contain 2 possible species; 1 possible species included a specimen from Urbana that was placed at the root of the group with high Bayesian support (0.91); the second possible species included the remaining specimens with moderate Bayesian support (0.77; Fig. 4).

At least 1 of the meristic traits of females varied significantly using the Gyrinicola species identification guided by the bPTP analysis (F_21,101.5 = 8.65, Pr > F < 0.0001; Table V). Canonical variate analysis yielded 2 major canonical vectors of near equal weights (Table V). These 2 canonical vectors explained a vast majority of the variance in the data (88.1%). Can1 (48% of variance) was primarily a contrast of vulva placement (pervu = −1.30), against placement of the excretory pore (perep = 0.77) and maximum width relative to body length (permw = 0.63). Can2 (40% of variance) was primarily a contrast of the width of the esophageal bulb relative to the body length (perbw = 2.83), against the length of the esophagus relative to body length (perel = −1.15; Fig. 5). Achieved power of 1.00 calculated (groups = 4, response variables = 7, effect size = 1.51, alpha = 0.05, F_21,111 = 1.65). All results are provided in Materials S1. Scatterplots of significant measurements noted above are given in Figure 6.

At least 1 of the meristic traits of males varied significantly with the Gyrinicola species identification assigned in the bPTP analysis (F_21,101.5 = 8.65, Pr > F < 0.0001; Table VI). Canonical variate analysis yielded 2 major canonical vectors of differing weights (Table VI). These 2 canonical vectors explained a vast majority of the variance in the data (96%). Can1 (62% of variance) was primarily a contrast of the length of the esophagus relative to body length (perel = −4.08), against maximum width relative to body length (permw = 4.42). Can2 (33% of variance) was primarily a contrast of the width of the esophageal bulb relative to body length (perbw = −1.63) and max width relative to body length (permw = −1.16), against spicule length relative to body length (perspi = 1.27; Fig. 7). Achieved power of 0.9997 calculated (groups = 4, response variables = 7, effect size = 1.44, alpha = 0.05, F_21,36 = 1.86). All results are provided in Materials S1. Scatterplots of significant measurements noted above are given in Figure 8.

Post hoc Tukey tests revealed significant variation in all morphological characteristics, except for excretory pore, placement in female specimens (Fig. 9). Post hoc Tukey tests suggested no significant variation in any morphological characteristics in male specimens at the selected alpha value of 0.007 (Fig. 9).

Discriminate function analysis of female morphological traits showed clear distinction among the 4 assigned species of Gyrinicola. Of the 4 clades assigned, 3 were capable of identification with 100% frequency. Clade 3, however, had a single member misattributed as a member of clade 1 and another member misattributed as a member of clade 4, leaving it with a proper identification frequency of 83%. Among the total sampling of 45 individuals, the error rate was relatively low, with only 4% of members misattributed (Table VII).

The discriminate function analysis of male morphological traits showed a clear distinction among the assigned Gyrinicola species numbers. Of the 4 clades assigned, 4 were capable of identification with 100% frequency. Among the total sampling of 20 individuals, the error rate was 0% (Table VIII).

Gyrinicola batrachiensis (Walton,1929)
(Synonyms: Pharyngodon batrachiensisWalton, 1929)
(Figs. 10, 11; Tables IX, X)

Cephalic extremity with 3 indistinct lips: 1 dorsal lip, 2 latero-ventrals. Outer cephalon features 4 large and well-developed compound papillae, each including 2 sensory pores. Two prominent amphids located between dorsal lip and latero-ventral lips. Six delicate projections, termed labial lobes, overhang the stoma; 2 are lateral, 2 are subdorsal, and 2 are subventral (Figs. 10F, 11A). The buccal cavity is shallow and wide and does not form a chitinized capsule. A delicate cuticular membrane projects into the stoma. Cuticular tooth, onchium, present in buccal region, visibly more prominent in relaxed specimens (Fig. 10A, F). Esophagus clearly divided into 3 portions: corpus, isthmus, and a globular posterior bulb. Cuticle thick and distinctly annulated, transversely, beginning short distance from the anterior end. Annulations become indistinct near the caudal end, near the anal region.

Worms with a relatively cylindrical large body, which tapers abruptly into a postanal filament. Body widest near the vulva. Filiform tail ranges from 11 to 25% of the organism’s total length (Figs. 6, 9, 10D; Table IX). Lateral alae absent. Entire esophagus generally 22% of total length, in some individuals it represents 25% of total length (Figs. 6, 9; Table IX). Esophageal bulb is wider than it is long, bears a terminal apparatus. Width of the bulb is typically 9% of the total length of worm, and about half the maximum width (Figs. 6, 9; Table IX). Nerve ring on the anterior third of corpus, within the first 4–6% of the total organism. Excretory pore opens posteriad to the junction of esophageal bulb and intestine. Vulva opens in anterior half of the body, roughly 45% of the way down from the mouth, slightly salient. Vagina muscular (Figs. 6, 9; Table IX). Uterus didelphic. One uterine branch coils dorsally; second uterine branch coils ventrally. Dorsal branch is longer and contains thick-shelled eggs; ventral branch is shorter and contains thin-shelled eggs. Thick-shelled eggs contain 1–8-cell stage embryos, generally around the 2–4-cell stage nearest to the vulva. Thick-shelled eggs featuring lateral, triradiate shell crests, triangular in cross section. Length of thick-shelled egg more than twice as long as width (Fig. 10G; Table IX). Thin-shelled eggs contained fully developed larva (Fig. 10E; Table IX). Posterior anal lip salient.

Smaller than females (Fig. 10C); prominent lateral alae extending from near excretory pore to near level of cloaca. Entire esophagus typically 22% of total length, up to 24% in some individuals (Fig. 8; Table IX). Esophageal bulb slightly wider than it is long. Width of the bulb is typically around 5% of total body length, and about half the maximum width (Fig. 8; Table IX). Nerve ring typically located at equator of corpus, distance from anterior end represents 10% of total length, proportionally further from anterior end than in female. Excretory pore between 27 and 37% of the way down the body (Table IX). Testis single, flexed in anterior half of the body. Opening of cloaca on prominent genital cone. Spicule 4% of body length, simple somewhat sigmoid; manubrium, calomus, and lamina not well differentiated, tapers to a sharp filiform terminal point (Figs. 8, 10B). Gubernaculum absent. Four pairs of caudal papillae present. First pair placed ventrally immediately precloacal. Second pair fused, placed ventrally, near the tip of the genital cone. Third pair placed laterally on genital cone, at level of cloaca. Fourth pair placed ventrally on the filiform tail, a short distance from the genital cone (Figs. 10B, 11B). Post cloacal filiform with a conical base, tapers immediately at level of third pair of papillae tail tapering to a fine terminus. Post cloaca to posterior distance around 25% of total length but may range to just under 30% of the total length in some individuals (Table IX).

Rana pipiens

Rana catesbiana, Rana clamitans.

Douglas Lake, Michigan.

Springfield, New York.

9 September 2021; 18 November 2021.

100%; 86%.

Large intestine.

From Michigan females: USNM 1606990–1606997, HWML 118592–118599, 118605–118606; Michigan males: USNM 1606998–1606999, HWML 118600–118602. New York Females: HWML 118603–118604.

USNM 1283147.

The 5 currently described species of Gyrinicola are as follows: G. tba (Dinnik, 1930), G. chabadamsoni Planade and Bain, 2008, G. chabaudi Araujo and Artigas, 1983, G. japonica Yamaguti, 1938, G. dehradunensis Maity, Rizvi, Bursey and Chandra, 2019. Of these species G. batrachiensis is differentiated by the number of labial lobes surrounding the mouth since G. chabadamsoni features 12 and G. chabuadi and G. tba feature 3, whereas G. batrachiensis features 6 (Table X). In the case of G. japonica, G. batrachiensis can be differentiated by the lack of a uterine pouch that contains subadult males and developing embryos; such a pouch is present in G. japonica. The North American G. batrachiensis most closely resembles the Indian G. dehradunensis in having the same number of labial lobes, and the lack of a uterine pouch. These 2 worms can be distinguished by the presence of larvated thin-shelled eggs in G. batrachiensis and by the atrophy of 1 of the uterine tracts evident in G. batrachiensis when they infect toads.

Gyrinicola armatus (Walton,1933)
(Synonyms: Pharyngodon armatusWalton, 1933)
(Figs. 11, 12; Tables X, XI)

Cephalic extremity with 3 indistinct lips: 1 dorsal lip, 2 latero-ventral. Four large and well-developed compound papillae present on the outer cephalon, each papilla includes 2 sensory pores. Two prominent amphids located between dorsal lip and latero-ventral lips. Six labial lobes overhang the stoma, 2 lateral, 2 subdorsal, and 2 subventral (Figs. 11C, G, 12E). Buccal cavity is shallow and wide and does not form a chitinized capsule. Delicate cuticular membrane projects into stoma. Cuticular tooth, onchium, present in buccal region, prominent in relaxed specimens (Fig. 12B, E). Esophagus clearly divided into 3 portions: corpus, isthmus and globular posterior bulb. Cuticle thick and distinctly annulated, transversely, beginning short distance from the anterior end. Annulations become indistinct near the caudal end, near the anal region.

Robust worm with a relatively cylindrical body, which tapers abruptly into a post anal filament. Body widest near the vulva. Filiform tail ranges from 14 to 34% of total length, generally 30% (Figs. 6, 9, 12A; Table XI). Lateral alae absent. Esophagus generally 21% of the organism’s total length, 25% in some individuals (Figs. 6, 9; Table XI). Esophageal bulb bears an apparatus and is wider than it is long. Width of the bulb is typically less than a 5% of the body length, and slightly less than half the maximum width (Figs. 6, 9; Table XI). Distance from anterior end to nerve ring represents 5–7% of body length. Excretory pore opens slightly posteriad to the esophageal–intestine junction. Vulva salient, situated near anterior third to anterior half of the body, representing 36–49% of body length (Figs. 6, 9; Table XI). Vagina muscular. Uterus didelphic. One uterine branch coils dorsally. Second uterine branch coils ventrally. Dorsal branch is longer and contains thick-shelled eggs. Ventral branch is shorter and contains thin-shelled eggs. Thick-shelled eggs present in 1–8 cell-stage but were generally around the 2–4-cell stage nearest to the vulva. Thick-shelled eggs featuring lateral, triradiate shell crests, triangular in cross section. Thick-shelled eggs 2 times longer than wide (Fig. 12F; Table XI). Thin-shelled eggs contain fully developed larva (Fig. 12G; Table XI). Posterior anal lip salient.

Smaller than females, somewhat fusiform bodies (Fig. 12D). Prominent lateral alae present extending from near excretory pore to near level of cloaca. Entire esophagus represents 20% of body length, 25% in some individuals (Fig. 8; Table XI). Esophageal bulb bears an apparatus and is only slightly wider than it is long. Width of the bulb represents 5% of body length, about half the maximum width (Fig. 8; Table XI). Nerve ring at the equator of esophagus; distance from anterior end represents 10% of body length. Distance from excretory pore to anterior end 28% of body length, ranging from 26 to 31% (Table XI). Testis single, flexed in anterior half of the body. Opening of cloaca on prominent genital cone. Single spicule is 4% of body length, simple and sigmoid, calomus narrower than manubrium and lamina (Figs. 8, 12C; Table XI). Gubernaculum absent. Four pairs of caudal papillae present. First pair placed ventrally immediately precloacal. Second pair fused, placed ventrally, near the tip of the genital cone. Third pair placed laterally on genital cone, at level of cloaca. Fourth pair placed ventrally on the tail filament, a short distance from genital cone (Figs. 11D, 12C). Postcloacal filiform tail features wide base; filament tapers to a fine pointed terminus. Distance from cloaca to posterior terminus represents 20% of body length, 24% in some individuals (Table XI).

Rana pipiens and Rana clamitans.

Rana catesbiana, Acris crepitans, Rana gyrlio.

Virginia.

Charleston, Illinois; Redford, Missouri; Cadiz, Kentucky; Berea, Virginia.

25 September 2020; 6 August 2021; 8 August 2021; 28 May 2022.

100%; 80%; 60%; 100%

Large intestine.

From Charleston females: USNM 1607000–1607003, HWML 118607–118609, 118621–118623; Charleston males: USNM 1607004–1607006, HWML 118610–118611. Virginia females: HWML 118612–118614. Missouri females: USNM 1607007–1607008, HWML 118615, 118620. Kentucky females: USNM 1607009–1607011, HWML 118616–118618. Florida females: HWML 118619.

USNM 1329276; USNM 1366835; USNM 1366836; USNM 1366837.

Gyrinicola armatus most closely resembles G. batrachiensis, in having the same number of projections of the labial lobes around the oral cavity, the lack of a uterine pouch, and the presence of larvated thin-shelled eggs. Specimens of G. armatus can be differentiated from G. batrachiensis in 7 meristic characters (Table X). The vulva of G. armatus, while variable, tends to be more anterior, the length of the tail tends to contribute more to total length and thick-shelled eggs tend to be wider. Male specimens are somewhat difficult to distinguish, but in males of G. armatus, the nerve ring and excretory pore tend to be placed slightly more posteriorly and specimens are stouter; finally, the esophagus and pharyngeal bulb are generally smaller. In his original description, Walton noted a sclerotized region present near the end of the buccal cavity for which he named the nematode (Walton, 1933). This sclerotized region is noted in specimens from Charleston, Virginia, Missouri, and Kentucky, but is also present in all specimens of Gyrinicola examined in the present study, as well as the recently described G. dehradunensis. The relative prominence of the sclerotized structure is associated with the rigidity, or relaxation, of the esophagus at the time of preservation; it is less prominent in specimens with tensed esophageal muscles. Males were not found among the specimens that were used for the original description; further, the cephalic papillae of females were not described. The description of those characters was provided in Holomon (1969). The measurements of the presently redescribed specimens are consistent with those documented by these authors (Table XI).

Gyrinicola gulabrevioris n. sp.
(Figs. 11, 13; Tables X, XII)

Cephalic extremity with 3 indistinct lips: 1 dorsal lip, 2 latero-ventral. Four large and well-developed compound papillae present on the outer cephalon, each includes 2 sensory pores. Two prominent amphids located between dorsal lip and ventral lips. Six delicate projections, termed labial lobes overhang stoma, 2 lateral, 2 subdorsal, and 2 subventral (Figs. 11E, 13E). Buccal cavity is shallow and wide and does not form a chitinized capsule. A delicate cuticular membrane projects into stoma. Cuticular tooth, onchium, present in buccal region, visibly more prominent in relaxed specimens (Fig. 13B, E). Esophagus clearly divided into 3 portions: corpus, isthmus and a globular posterior bulb. Thick cuticle transversally annulated, annulations begin a short distance from the anterior end. Annulations become indistinct near the anal region.

Large worm with a relatively cylindrical body, which tapers abruptly into a postanal filament. Body widest near the vulva. Filiform tail represents 10–24% of body length (Figs. 6, 13A; Table XII). Lateral alae absent in females. Entire esophagus represents 17% of body length, up to 21% in some individuals (Figs. 6, 9; Table XII). Bulb bears an apparatus and is wider than it is long. Width of the bulb is 4% of body length, slightly less than half the maximum width (Figs. 6, 9; Table XII). Nerve ring in anterior third of corpus representing 3–6% of body length. Excretory pore opening slightly posteriad to the esophageal–intestine junction (Table XII). Vulva opens near midpoint of body, distance from anterior end represents 44–54% body length (Fig. 6; Table XII). Vagina muscular. Uterus didelphic. One uterine branch coils dorsally. Second uterine branch coils ventrally. Dorsal branch is longer and contains thick-shelled eggs. Ventral branch is shorter and contains thin-shelled eggs. Thick-shelled eggs present in 1–8-cell stage but were generally around the 2–4-cell stage nearest to the vulva. Thick-shelled eggs featuring lateral, triradiate shell crests, triangular in cross section. Thick-shelled eggs length slightly more than 1.5 times as long as they are wide (Fig. 13F; Table XII). Thin-shelled eggs contain fully developed larva; some were noted to have hatched in utero (Fig. 13G; Table XII). Posterior anal lip salient.

Cylindrical body, smaller than female (Fig. 13D). Prominent lateral alae present extends from near excretory pore to near level of cloaca. Entire esophagus is 20% of body length, up to 24% in some individuals (Fig. 8; Table XII). Esophageal bulb bears an apparatus and is generally as wide as it is long. Width of the bulb is typically 5% of the body length, and around 70% of the maximum width (Fig. 8; Table XII). Nerve ring is in posterior third of corpus, distance from anterior end represents 12% body length. Distance of anterior end to excretory pore 26–31% of body length, generally closer to 28% (Table XII). Testis single, flexed in anterior half of the body. Opening of cloaca on prominent genital cone. Single spicule present, simple, around 5% of the organism’s length; manubrium ventrally bent, calomus narrower than manubrium, lamina not well differentiated (Figs. 8, 13). Gubernaculum absent. Four pairs of caudal papillae present. First pair placed ventrally immediately precloacal. Second pair fused, placed ventrally, near the tip of the genital cone. Third pair placed laterally on genital cone, at level of cloaca. Fourth pair placed ventrally on the tail, a short distance from the genital cone (Figs. 11F, 13C). Postcloacal filiform tail features a narrow base, which at level of fourth pair of papillae tapers to a fine point. Filiform tail represents 23% of body length, up to 30% in some individuals (Table XII).

USNM 1606972.

USNM 1606973.

HWML 118558–118559.

Rana catesbiana.

Pseudacris regilla; Rana clamitans.

Carbondale, Illinois.

Teal Ridge, Oklahoma; Port Orchard, Washington; Mount Pleasant, Iowa; Renton, Washington.

23 June 2021.

4 September 2020; 22 July 2021; 8 July 202; 21 July 2021.

N/A; 100%; 100%; 100%; 75%.

Large intestine.

Carbondale females: USNM 1606974–1606977, HWML 118560–118563. Carbondale males: USNM 1606978–1606980, HWML 118564–118565. Washington (Port Orchard) females: HWML 118566–118568, 118775–118777. Washington (Renton) females: HWML 118569–118570, 118778, 118779. Iowa females: HWML 118571, 18780. Oklahoma female: HWML 118781.

urn:lsid:zoobank.org:act:E64C432D-B296-4E9E-B1B5-3F0F3FD80E30

The species name is a combination of the following Latin words: gula, meaning throat, and brevis meaning short. The combination of the 2 is used to draw attention to the short esophagus in comparison to the total length of the organism which is present in the more commonly found females.

The newly described G. gulabrevioris is characterized by the presence of a relatively short esophagus as identified by the esophagus total length ratio (Table X). The pinworm G. gulabrevioris most closely resembles the previously described G. batrachiensis, in having the same number of projections of the labial lobes around the oral cavity, similar ratios of the tail to body lengths, and placement of the vulva. In female specimens, of G. gulabrevioris, the esophagus tends to be shorter in comparison to the total length, the nerve ring tends to be more anterior, and the thick-shelled eggs tend to be shorter. Male specimens are difficult to distinguish, but in males of G. gulabrevioris, the nerve ring and excretory tend to be placed slightly more posteriorly, specimens are narrower, the spicule to body length is shorter, the esophagus is slightly larger, and pharyngeal bulb is slightly wider.

Gyrinicola moohsia n. sp.
(Figs. 11, 14; Tables X, XIII)

Cephalic extremity with 3 indistinct lips: 1 dorsal lip, 2 ventral. Four large and well-developed compound papillae present on the outer cephalon, each includes 2 sensory pores. Two prominent amphids located between dorsal lip and ventral lips. Six delicate projections, termed labial lobes overhang the stoma, 2 lateral, 2 subdorsal, and 2 subventral (Fig. 14C). The buccal cavity is shallow and wide. A delicate cuticular membrane projects into the stoma. Cuticular tooth, onchium, present in buccal region, visibly more prominent in relaxed specimens (Fig. 14B). Esophagus clearly divided into 3 portions: corpus, isthmus, and a globular posterior bulb. Cuticle thick and transversally annulated, annulations begin a short distance from the anterior end. Annulations become indistinct near the anal or cloacal regions.

Large worm with a relatively cylindrical body, which tapers abruptly into a post anal filament. Maximum body width at level of vulva. Filiform tail represents 25–33% of body length (Figs. 6, 14A; Table XIII). Lateral alae absent. Entire esophagus represents 25% of body length, up to 32% (Figs. 6, 9; Table XIII). Bulb bears an apparatus and is wider than it is long. Width of the bulb is typically 6% body length, slightly more than half of the maximum width (Figs. 6, 9; Table XIII). Nerve ring in the anterior third of corpus, distance to anterior end represents 6% of body length. Excretory pore opens at level of esophagus/intestine junction. Vulva situated near anterior third to anterior half of the body, distance to anterior end represents 33–49% of body length (Figs. 6, 9; Table XIII). Vagina muscular. Uterus didelphic. One uterine branch coils dorsally. Second uterine branch coils ventrally. Dorsal branch is longer and contains thick-shelled eggs. Ventral branch is shorter and contains thin-shelled eggs. Thick-shelled eggs present in 1–8-cell stage but were generally around the 2–4-cell stage nearest to the vulva. Thick-shelled eggs featuring lateral, triradiate shell crests, triangular in cross section. Length of thick-shelled eggs less than 2 times its width (Fig. 14D; Table XIII). Thin-shelled eggs contain fully developed larva (Fig. 14F; Table XIII). Posterior anal lip salient.

Worm fusiform, compact (Fig. 14G). Prominent lateral alae present extending from near excretory pore to near level of cloaca. Esophagus represents 24% of body length (Fig. 8; Table XIII). Esophageal bulb bears an apparatus and is generally as wide as it is long. Width of the bulb is typically 6% of body length, and around a half of the maximum width (Fig. 8; Table XIII). Nerve ring at equator of corpus, distance to anterior end represents 11% of body length. Distance from anterior end to excretory pore represents 29% of body length, ranging 27–32% (Table XIII). Testis single, flexed in anterior half of the body. Opening of cloaca on prominent genital cone. Single spicule present, typically 4% of body length, manubrium ventrally bent, calomus narrower than manubrium, lamina not well differentiated (Figs. 8, 14E). Gubernaculum absent. Four pairs of caudal papillae present. First pair placed ventrally immediately precloacal. Second pair fused, placed ventrally, near the tip of the genital cone. Third pair placed laterally on genital cone, at level of cloaca. Fourth pair placed ventrally on base of filiform tail (Figs. 11H, 14E). Postcloacal filiform tail features a very short basis, wider than it is long. From its base it tapers gradually to a fine-pointed terminus. Filiform tail represents 25% of body length, 33% in some individuals (Table XIII).

USNM 1606981.

USNM 1606982–1606984, HWML 118578–118579.

HWML 118580.

Rana clamitans.

Urbana, Illinois.

27 September 2020.

68%.

Large intestine.

Urbana females: USNM 1606985–1606988, HWML 118581–118584. Urbana males: USNM 1606989, HWML 118585–118586, 118589–118591.

urn:lsid:zoobank.org:act:0D69F306-B5DC-45AE-B51D-11B896519752

In the Myaamia language traditionally spoken by the indigenous peoples of Illinois the noun moohsia designates a worm/bug. The noun is used in apposition of the genus name.

The newly described G. moohsia can be distinguished from the other globally distributed Gyrinicola in the same manner as the other North American Gyrinicola (Table X). The pinworm G. moohsia is characterized by having a wider esophageal bulb as identified by the esophageal bulb length ratio (Table X). The pinworm G. moohsia most closely resembles the previously described G. armatus, in having the same number of projections of the labial lobes around the oral cavity, similar tail lengths, and maximum widths. In female specimens, of G. moohsia, the esophageal bulb tends to be much wider in comparison to the total length, and the thick-shelled eggs tend to be shorter. Male specimens are difficult to distinguish, but in males of G. moohsia, the nerve ring, and excretory pore tend to be placed slightly more anteriorly, the esophagus is slightly larger, and the pharyngeal bulb is much wider.

In the comparative remarks included in his original description of the nematode P. armatus Walton wrote, “the variations seem to be of specific importance” in reference to the characteristics he used to distinguish the nematode from the previously described P. batrachiensis (Walton, 1933). The present study represents a continued investigation of these “variations,” which are present in the North American Gyrinicola, from an integrated perspective. Some of these morphological differences have been observed in organisms infecting different species of tadpoles occurring in sympatry (Rhoden and Bolek, 2011; Pierce et al., 2017). The present investigation includes an increased representation of nuclear ribosomal genes, an analysis of morphological characteristics, and for the first-time amplification of mitochondrial genes for the Gyrinicola. This integrative approach is key when studying Oxyurid nematodes; the variation of oxyurids can be difficult to evaluate on the morphological basis upon which the tradition of nematode identification sits, however, it can also be difficult to distinguish using only DNA. It is for this reason that several researchers advocate for a more integrative approach to taxonomy (Dayrat, 2005; Will et al., 2005; Nadler and Pérez-Ponce de León, 2011; Sloan et al., 2017; Pereira et al., 2018; De Sousa et al., 2019).

Morphological data and phylogenetic data are not always in congruence with one another. Two recent studies of nematode parasites of amphibians and squamates represent the strong dichotomy of results returned when morphological hypotheses are reevaluated with new tools. Screening of specimens of Oswaldocruzia has shown that what was considered a complex of species present across Europe, may be a single species with host-induced morphological variability (Kirillova and Kirillov, 2020). Phylogenetic work with a variety of species in Parapharyngodon and Thelandros showed instances of genetic discordance and revealed that some species were wrongly assigned to a genus, likely because identification relied heavily on host taxonomic identity and a few potentially homoplastic characters (De Sousa et al., 2019). The dichotomy of results in the previously mentioned studies underscores the need to identify and characterize the species within a broader context of comparative biology by utilizing multiple lines of evidence in support of a species.

Analysis of nrDNA was consistent with the previous analysis of the 18S and 28S genes by previous researchers, though, in this nrDNA analysis, there was greater support for internal clades being present in the Gyrinicola (Sata and Nakano, 2020; Walker et al., 2023). The increased support for the internal clades was likely because of increased sampling of the nrDNA through the inclusion of the more variable ITS regions. The phylogeny based on mtDNA was congruent in topology with previous, and current, nrDNA analysis (Fig. 1). The North American Gyrinicola split into the same 3 clades present within the nrDNA analysis. The branch lengths of the mtDNA tree were long; however, the outgroups with all available genes were somewhat distant as shown in the nrDNA analysis (Figs. 1, 2). Given the consistency in the trees, it appears both sources of genetic information reflect the same distinction among the North American Gyrinicola groupings.

The bPTP species analysis performed on the concatenated genes suggested 4 species were present in the analysis. The first 2 clades were identified as separate species with low support (G. batrachiensis = 0.52) and moderate support (G. armatus = 0.82) respectively, while the third clade was suggested to consist of 2 species with moderate support (G. gulabrevioris = 0.77) and high support (G. moohsia = 0.91) respectively, with the nematode from Urbana, Illinois being separate from other clade-3 nematodes (Fig. 4).

Genetic distances reflect the genetic divergence among groups, and although they alone cannot suggest speciation, they can be compared to values among closely related groups. When the distance comparisons are made to groups that are already recognized as valid species, the valid species can serve as a benchmark for whether or not further investigation should take place in the group of interest. The genetic distances calculated for specimens included in each of the clades are consistent with the identification of 4 species in the bPTP species-delimitation analysis. Intraspecific genetic distances among members of clade 3, namely G. moohsia and G. gulabrevioris, were consistently greater in both analyses. In the nrDNA analysis, genetic distances within clade 3 were higher than the other Gyrinicola clades with differences of 0.3–4% compared with differences of 0–0.5% and 0–2.4% observed in clade 1 and clade 2, respectively (Table III). The distances seen within clade 3 in the nrDNA analysis (0.3–4%) were also similar to the differences seen between different species of Spauligodon (0–3%) in the previous analysis (Walker et al., 2023). In the mtDNA analysis, the maximum genetic distances within clade 3 were much like the difference between the 2 species of Wellcomia (8.1% vs. 8.2%). Between G. batrachiensis and G. armatus there was an average interspecific distance of 19.6%, whereas the interspecific distance between G. gulabrevioris and G. moohsia was 7.4%, on average (Table IV). One possible explanation for this genetic difference is that there is, perhaps, even greater taxonomical diversity than described herein, which has yet to be discovered and properly named. Clues to this diversity would include biological processes and modes of reproduction that may be inherent to each lineage or species, which could only properly be documented with intensive in vivo laboratory study. The foundation for this hypothesis is the variations in chromosome numbers encountered by Adamson (1981c). Evaluation of chromosome counts with extremely fresh specimens may assist future researchers in making these species determinations, but was beyond the scope of the current study. Given the congruence of the genetic distances, and bPTP species analysis, 4 clades representing species were retained for the morphological analysis, regardless of the cause of the noted variation.

In the morphological analysis, the MANOVA was found to be instrumental in clarifying the distinction of the putative species, as it was able to test for variations within and between the putative species groups. Results for female and male tests were highly significant, which suggested that at least 1 of the physical characteristics measured varied between the 4 proposed species suggested in the bPTP analysis. The 3 most important areas of morphological variation in females, as indicated by the standardized canonical variables, appear to be vulva placement, relative esophagus length, and relative pharyngeal bulb width (Fig. 5; Table V). In males, the most important areas of morphological variation appear to be the relative esophagus length, relative maximum width, as well as relative pharyngeal bulb length, and relative spicule length (Fig. 7; Table VI). The canonical variate analysis, furthermore, showed clear differentiation in centroids of the 4 proposed species in both females and males (Figs. 5 and 7, respectively). In female specimens, Gyrinicola batrachiensis and G. gulabrevioris appeared to be the most like one another, with some overlap of the centroids being present, whereas G. armatus and G. moohsia were morphologically distinct. In male specimens, G. batrachiensis and G. armatus were the most like one another, but no overlap was present in any of the 4 centroids, suggesting clear morphological distinction amongst males.

It has been shown that Gyrinicola possesses wide phenotypic variation which varies with the host (Adamson, 1981a; Rhoden and Bolek, 2011; Peirce et al., 2017). While bufonids and hylids were sampled for this study none of the bufonids sampled yielded any nematodes, and the hylids sampled did not yield enough nematodes for significant morphological interpretations. These groups will require further study to determine trends in their variation, in particular, sequencing of toad infecting Gyrinicola will help greatly help clarify their placement and add support to the reproductive plasticity of the group. Among ranid tadpoles Rhoden and Bolek (2011) noted several differences in nematodes infecting various species; in our morphological analysis, we were able to note several other factors with large vector values not noted as significant between previously compared ranid tadpoles. These morphological factors included the following: vulva placement in females, esophagus length in both sexes, max width in both sexes, excretory pore placement in females, spicule length, and tail length in males. Our analysis also showed significance without consideration of host identity; this suggests that although these nematodes feature host-induced morphological differences, it is likely some of the morphological variations are constrained in a phylogenetic context, as the signal is still seen when only putative species identity is considered. The morphological variability of Gyrinicola can make the interpretation of species identity difficult without a taxonomic framework for interpretation. It is, therefore, unsurprising that the species of North American Gyrinicola were considered to be a single species before our investigation. The integrative analysis herein yields results that are consistent in the detection of 3 clades within the putative G. batrachiensis. This signal resulted in the detection of 4 species using topology-based methods to detect species limits.

Although our study is integrative, it is by no means all-encompassing; the North American Gyrinicola will require further study. To clarify the morphological variation of the putative North American species, as well as document the role ploidy as a reproductive strategy or barrier, the phylogeny we present can be used to guide experimental infections in vivo as presented in Adamson (1981a, 1981c). This in vivo work could easily be paired with analysis of any possible differences in the development times, transmission, and intensity of infection; given the starting infection number was controlled, this work could mirror that of Childress et al. (2017) and Rhoden and Bolek (2011) utilizing various natural hosts. The life history factors discussed above may be strongly tied to host type and/or geographic range; thus they would require detailed study to elucidate. Host effects of the different North American species could also be investigated to see if all species were consistent with the mutualist effects proposed by previous studies (Pryor and Bjorndal, 2005). The life cycle and development of Gyrinicola batrachiensis was also studied in depth by Adamson, and similar studies of the remaining tadpole pinworms may help elucidate the purposes of the widely varying genital tracts found in the genus and lead to an investigation of how the various reproductive tracts evolved (Adamson, 1981b; Brigitte et al., 2008). Geological history could also play a major role in how this group of worms has diversified, though this will be difficult to access until the sampling effort is increased greatly and distributions have been clarified for the various species of Gyrinicola. Given our current discovery of new species, it is likely differences exist in 1 of the foregoing factors; as such the documentation of the global diversity of Gyrinicola and the characterization of their biological features represents a promising area for future study.

Finally, many of the original descriptions of tadpole pinworms are lacking in detail and vary widely in the use of terminology for various features; the understanding of the group would, therefore, be greatly improved by a holistic investigation that clarified and defined such features. It is clear that all species of Gyrinicola present in North America feature 6 labial lobes and possess a ventral uterine branch holding thin-shelled eggs (Table XIII). Given the rarity of males, we recommend basing species identifications on characters present in females; in particular, we suggest the use of the vulva placement, relative esophagus length, and relative pharyngeal bulb width. The species diversity for this genus may be greater than what we are currently reporting, as there are vast regions of the continent that have yet to be surveyed and the role of ploidy as a reproductive barrier has yet to be tested. As such, the North American Gyrinicola, and the genus as a whole, represent a fascinating and wide area for future study.

This work was funded by the National Science Foundation (NSF) through Division of Undergraduate Education award number 1564969. Research methods involving anuran hosts collected for this study were approved under Southern Illinois University, Carbondale Institutional Animal Care and Use Committee Protocol 21-015. Wild specimens were collected under state-associated scientific collection permits issued to F.A.J., M.A.W., E.A.Z., G.J.L., and F.B.R., as stated in Walker et al. (2023).