Ophiotaenia currani, Chambrier & Kudlai & McAllister & Toma & Scholz, 2023

Chambrier, Alain de, Kudlai, Olena, McAllister, Chris T., Toma & Scholz, s, 2023, Discovering high species diversity of Ophiotaenia tapeworms (Cestoda: Proteocephalidae) of watersnakes (Colubridae) in North America, International Journal for Parasitology: Parasites and Wildlife 22, pp. 255-275 : 265-273

Ophiotaenia currani n. sp. Figs. 3I–K , 6A–C .

urn:lsid:zoobank.org:act:3F881E29-D443-4878-9CD6-903493D14814

Material examined: Four mature and two juvenile whole-mounted specimens from Nerodia fasciata confluens (Blanchard) , host USA 22, Ocean Springs, Mississippi, USA, V. V. Tkach, March 2009 (MHNG- PLAT-0063341) .

Type- and only host: Broad-banded watersnake, N. fasciata confluens ( Colubridae : Natricinae).

Type- and only locality: Ocean Springs, Jackson County, Mississippi, USA (latitude 30.4113100 ◦, longitude – 88.8278100 ◦).

Type-material: Holotype – complete specimen ( MHNG-PLAT-0063341 ); 1 paratype – mature specimen ( USNM 1690929 ).

Representative DNA sequences: New sequences of 28S rDNA ( OR 663966) and cox 1 ( OR 660032) (isolate from host USA 22, Mississippi – Figs. 9 and 10 ).

Etymology: The new species is named in honour of Stephen S. Curran, currently at Auburn University, Alabama, for his generous and continuous assistance to the present authors in capturing hosts of proteocephalid and other tapeworms in Mississippi from 2009 to 2019.

Description (based on four specimens; for other measurements – see Table 1): Proteocephalidae . Medium-sized worms, about 9 cm long and up to 0.7 mm wide, flattened dorsoventrally, with proglottids very elongated, up to 3.2 mm long. Strobila acraspedote, anapolytic. Immature proglottids wider than long to longer than wide (length: width ratio 0.46–4.76); mature proglottids longer than wide (length: width ratio 3.83–4.74); pregravid proglottids much longer than wide (length: width ratio 5.87–8.44); gravid proglottids conspicuously longer than wide (length: width ratio 8.07–9.04). Tegument thin, about 5 in thickness.

Scolex spherical, aspinose ( Fig. 3I–K ), wider than neck, with four uniloculate, spherical suckers ( Fig. 3I–K ), 65–80 in diameter; sucker diameter/scolex width ratio 35–40%. Apical organ absent, but cells of granular content concentrated in scolex apex ( Fig. 3I–K ). Neck up to 160 wide; unsegmented zone posterior to scolex to first recognisable proglottids long (up to 16–24 mm).

Inner longitudinal musculature developed, consisting of numerous small, isolated muscle fibres. Osmoregulatory canals slightly sinuous, divide testicular fields, situated 15–25% of proglottid width from lateral margins ( Fig. 6A and B ). Ventral canals thin-walled, 15–20 in diameter. Dorsal canals thick-walled, about 4–5 in diameter, never reaching to level of vitelline follicles ( Fig. 6A and B ). Terminal genitalia between osmoregulatory canals.

Testes spherical, in one layer and two lateral fields on both sides of uterine stem ( Fig. 6A and B ). Testes not reaching ovary posteriorly and anterior margin of proglottids ( Fig. 6A and B ). Vas deferens strongly coiled, reaching to mid-line of proglottid, sometimes slightly overlapping it, occupying small area directed posteriorly ( Fig. 6A and B ). Cirrus sac oval to pyriform, thin-walled, 190–270 long by 100–120 wide, length: width ratio 1.75–3.44 ( Fig. 6C ). Cirrus occupies up to 90% of length of cirrus sac ( Fig. 6C ). Genital atrium deep and narrow; genital pores irregularly alternating, markedly pre-equatorial ( Fig. 6A and B ).

Ovary bilobed, 300–490 wide, with long and narrow lateral lobes ( Fig. 6A and B ). Vagina anterior (80%) or posterior (20%, n = 61) to cirrus sac, without vaginal sphincter ( Fig. 6C ); terminal part of vagina near genital atrium lined with thick layer of intensely stained cells ( Fig. 6C ). Mehlis’ gland 55–70 in diameter.

Vitelline follicles spherical to elongate, arranged in two lateral, longitudinal columns on lateral side of proglottid ( Fig. 6A and B ), interrupted at level of terminal genitalia (cirrus sac and vagina; Fig. 6C ). Follicles approaching, but not reaching, anterior or posterior margin of proglottids; posteriorly, vitelline follicles overpassing mid-length of ovary ( Fig. 6A and B ).

Primordium of uterine stem ventral. Uterus reaching anterior margin of proglottid but not overpassing ovarian isthmus posteriorly ( Fig. 6B ). Uterine development of type 2 of de Chambrier et al. (2004), with uterine diverticula formed before appearance of first eggs in uterus, occupying about 14–19% of proglottid width in mature proglottids. In pregravid proglottids, diverticula occupy up to 33% of proglottid width. In gravid proglottids, diverticula occupy up to 63% of proglottid width. Eggs (measured in whole mounts), with embryophore 25–32 in diameter and oncospheres 15–18 in diameter.

Differential diagnosis

The new species differs from congeneric tapeworms found in watersnakes occurring in North America, including O. perspicua redescribed here and the other new species described in this paper, by having (i) few testes (89–118 in the new species vs> 150 in the other species); (ii) a long cirrus sac, the length of which is 35–50% of the proglottid width in O. currani n. sp. (<30% in other species, usually about 20–25%;

Table 1); (iii) elongate proglottids that are much longer (4 × in mature proglottids and up to 9 × in gravid proglottids) than wide in the new species (in contrast to much shorter in other species).

Remarks

Ophiotaenia currani n. sp. was found only once and may be a relatively rare parasite of N. f. confluens. Six specimens were found in a single N. f. confluens examined in March 2009.

Ophiotaenia tkachi n. sp. Figs. 3L , 6D and 7 , 8A, B, E–H .

urn:lsid:zoobank.org:act:945662BA-27D6-4488-8759-57A218830661

Material examined: Four specimens from Nerodia fasciata confluens , host USA 21, Hammond, Louisiana, USA, V.V. Tkach, 16 July 2002 ( MHNG-PLAT- 0063340; IPCAS C-939/1); one mature spec. without scolex (hologenophore) ( MHNG-PLAT- 0130134) and fragments of immature specimen from N. f. confluens, host US 952, 4.8 km north of Idabel at Turner site, Oklahoma, USA, CTM, 10 June 2019; one whole-mounted spec. without scolex (hologenophore) and cross sections from N. erythrogaster , host US 1118, Hochatown, McCurtain Co., Oklahoma, CTM, 26 August 2021 ( MHNG-PLAT- 0063340); cross sections of one spec. (hologenophore) from N. sipedon , host US 1128, Bear Creek at Bear, Garland Co., Arkansas, CTM, 27 May 2021 ( MHNG-PLAT-0150187); one complete spec. (hologenophore) from A. piscivorus US 953, Hochatown, McCurtain County, Oklahoma, USA, CTM, 23 June 2019 ( IPCAS C-939/2; MHNG-PLAT- 0130136) (this specimen was morphologically characterised by Scholz et al., 2023a).

Type-host: Broad-banded watersnake, Nerodia fasciata confluens ( Colubridae : Natricinae).

Additional hosts (not confirmed molecularly): Nerodia erythrogaster ; N. sipedon ; Northern cottonmouth, A. piscivorus (postcyclic host – Scholz et al., 2023a).

Type-locality: Hammond, Tangipahoa Parish, Louisiana, USA (latitude 30.504358 ◦; longitude – 90.461197 ◦).

Distribution: USA (Arkansas, Louisiana, Oklahoma).

Type material: Holotype – complete specimen ( MHNG-PLAT-0063340 ); 2 paratypes – specimens without scolex ( IPCAS C-939/1; USNM 1690928 ).

Representative DNA sequences: New sequences of 28S rDNA (OR663962–OR663965) and cox 1 (OR660028–OR660031) (isolates from hosts US 952, 953, 1118 and 1128 – Figs. 9 and 10 ).

Etymology: The new species is named in honour of Vasyl V. Tkach, University of North Dakota, eminent helminthologist, who kindly provided the present authors with extensive material of proteocephalid tapeworms for taxonomic and phylogenetic studies, including the specimens evaluated in the present work.

Description (based on two complete specimens from N. f. confluens

265

( USA 21), Louisiana; measurements of the specimen without scolex from the same host species ( US 952) from Oklahoma in brackets; for other measurements – see Table 1). Proteocephalidae . Large worms, up to 9.7 mm [12.8 mm] long and 1.1 mm [1.3 mm] wide, flattened dorsoventrally, with proglottids elongated, up to 1.7 mm [2.2 mm] long. Strobila acraspedote, anapolytic, with about 164 immature proglottids (up to appearance of spermatozoa in vas deferens), 4 mature proglottids (up to appearance of eggs in uterus), 6 pregravid proglottids (up to appearance of hooks in oncospheres), about 2 gravid proglottids; in total 176 proglottids in holotype. Immature proglottids wider than long to longer than wide (length: width ratio 0.28–1.18) [0.22–1.97]; mature proglottids slightly longer than wide (length: width ratio 1.20–1.46)

266

[2.12–2.31]; pregravid proglottids longer than wide (length: width ratio 1.30–1.51) [2.11–2.23]; gravid proglottids longer than wide (length: width ratio 3.42–4.28) [1.21–3.87]. Tegument thick, 10–12 [5–8] in thickness.

Scolex spherical, aspinose ( Fig. 3L ), wider than neck, with four uniloculate, elongated suckers ( Fig. 3L ); sucker diameter/scolex width ratio 37–43%. Apical organ absent, but cells of granular content concentrated in scolex apex ( Fig. 3L ). Neck up to 175 wide; unsegmented zone posterior to scolex to first recognisable proglottids long (up to 6–7 mm).

Inner longitudinal musculature well-developed, consisting of numerous small isolated muscle fibres ( Fig. 8A and B ). Osmoregulatory canals slightly sinuous, divide testicular fields, situated 21–35% [22–33%] of proglottid width from lateral margins ( Fig. 7A and B ). Ventral canals thin-walled, 15–20 [15–20] in diameter. Dorsal canals thick-walled, about 4–5 [8–14] in diameter, never reaching to vitelline follicles. Terminal genitalia between osmoregulatory canals.

Testes spherical to oval, in one or two layers and two dense lateral fields on both sides of uterine stem ( Figs. 7A, B , 8A ). Testes reach ovary posteriorly, sometimes slightly overlapping it, anteriorly reaching anterior margin of proglottids ( Fig. 7A and B ). Vas deferens strongly coiled, reaching mid-line of proglottid, sometimes slightly overlapping it, occupying elongated area ( Fig. 7A–C ). Cirrus sac oval to pyriform, thin-walled, 150–280 [220–295] long by 70–120 [120–145] wide ( Figs. 6D and 7C ); length: width ratio 1.48–2.00 [1.59–1.95]. Cirrus occupies up to 80% [75%] of length of cirrus sac. Genital atrium deep, narrow ( Figs. 6D and 7C ), genital pores irregularly alternating, markedly pre-equatorial ( Fig. 7A and B ).

Ovary bilobed, 695–810 [720–920] wide, with long and narrow lateral lobes ( Fig. 7A and B ). Vagina anterior (44%) or posterior (56%, n = 80) [mainly posterior – 92%; n = 36] to cirrus sac, without vaginal sphincter ( Fig. 7C ) or with a few muscle fibres surrounded vaginal canal ( Fig. 6D ). Terminal part of vaginal canal forming several plies, making canal flower-shaped in cross-section ( Figs. 6C and 7C, D ). Distal part of vaginal canal lined with thick layer of intensely stained cells ( Figs. 6C and 7C, D ). Mehlis’ gland 60–90 [65–100] in diameter.

Vitelline follicles spherical to elongate, arranged in two lateral, longitudinal columns on lateral side of proglottid ( Fig. 7A and B ), interrupted at level of terminal genitalia on ventral side ( Fig. 7A–C ), but uninterrupted dorsally ( Fig. 6D ). Follicles approaching, but not reaching, anterior or posterior margin of proglottids ( Fig. 7A and B ). Vitelline follicles overlap lateral margins of ovary ( Fig. 7A and B ).

Primordium of uterine stem ventral. Uterus reaching anterior margin of proglottid but not overpassing ovarian isthmus posteriorly. Uterine development of type 2 of de Chambrier et al. (2004), with uterine diverticula formed at appearance of first eggs in uterus, occupying about 25% [17%] of proglottid width in mature proglottids. In pregravid proglottids, diverticula occupy up to 43% [26%] of proglottid width ( Fig. 7A and B ). In gravid proglottids, diverticula occupy up to 62% [60%] of proglottid width.

Eggs (measured in whole mounts) spherical, with embryophore 18–22 in diameter and oncospheres 12–15 in diameter with 3 pairs of embryonic hooks. [Eggs unripe (without fully formed oncospheres), spherical, with hyaline outer envelope 50–70 in diameter connected with membraneous plies to embryophore 28–33 in diameter ( Fig. 8E–H ).]

Differential diagnosis

The new species can be distinguished from all but one ( O. laruei n. sp.) of the Ophiotaenia species that parasitize watersnakes in North America by the combination of the following characteristics: (i) the terminal (distal) portion of the vaginal canal is not tubular, but forms multiple folds (in other species the vaginal canal is tubular); (ii) numerous, densely packed testes (> 180 per proglottid vs. <200); (iii) robust strobila formed by mature proglottids that are only slightly longer than wide (length to width ratio 1.20–1.46), whereas proglottids are more elongate in other taxa.

Ophiotaenia tkachi n. sp. differs from O. laruei n. sp. in the smaller size of the scolex (width <230 μm vs.> 280 μm in O. laruei n. sp.), shape and

267

size of the suckers (spherical vs triangular; <90 μm in diameter in O. tkachi vs.> 160 μm in diameter), a relatively longer cirrus sac (its length corresponds to 19–26% of the proglottid width vs only 13–15% in O. laruei ), and a slightly smaller number of testes (<270 vs.> 270).

Remarks

The Arkansas and Oklahoma specimens resemble those from Louisiana in most morphological characters, including the peculiar structure of the distal part of the vaginal canal, the shape of the proglottids, the relative size of the cirrus sac, the position of the genital pore, the number of uterine diverticula, the relative size of the Mehlis’ gland, etc. ( Table 1). They differ slightly from those from Louisiana in the number of testes, position of the vagina, and a relatively longer ovary in the worm from N. f. confluens from Oklahoma ( Table 1), but these differences are considered to be intraspecific variability among conspecific specimens from different definitive hosts.

Molecular data are not available for type specimens because attempts to obtain their DNA sequences have failed, probably due to the poor quality of DNA in samples collected more than 20 years ago. However, sequences of two genes for tapeworms from the type host from Oklahoma, two congeneric species of watersnakes from Arkansas and Oklahoma, and the northern cottonmouth from Oklahoma are available. These tapeworms have identical partial sequences of 28S rDNA (D1–D3 region) and differ only slightly in the sequences of cox 1 ( Fig. 10 and Supplementary Tables 1 and 2); therefore, they are considered conspecific. Based on their morphological similarity (see above), these tapeworms are considered to belong to O. tkachi .

Ophiotaenia faranciae ( MacCallum, 1921) Hilmy, 1936 .

Syns Taenia faranciae MacCallum, 1921 ; Proteocephalus faranciae ( MacCallum, 1921) Harwood, 1932 .

Material examined: Fragments of one specimen (without scolex) from Farancia abacura abacura (Holbrook) , Pennsylvania, USA, collected by N. Rothman on 5 March 1958 ( USNM 1352975 ) .

Type- and only host (see Remarks regarding report from Nerodia cyclopion [Dum´eril, Bibron et Dum´eril): Red-bellied mudsnake, Farancia abacura (Holbrook) ( Colubridae : Dipsadinae).

Type-locality: Zoological Park, New York, USA.

Distribution and records: USA (Arkansas, Florida, Texas) ( MacCallum, 1921; Harwood, 1932; Brooks, 1978; McAllister et al., 2008).

Type-material: Not available.

Representative DNA sequences: No molecular data available.

Remarks

268

The original description was very brief and included only some measurements of specimens found in a mudsnake from the Zoological Park in New York, USA ( MacCallum, 1921). This specimen was severely shrunken and had a partially retracted scolex. Despite the incompleteness of the original description and the lack of further data on its morphology, the species appears to be well separated from other taxa of Ophiotaenia from the colubrid snakes of the Nearctic by a large scolex (width 600 μm) with large suckers (diameter 200 μm) and numerous (390–420) testes ( MacCallum, 1921). Harwood (1932) redescribed this species based on three specimens he found in a road-killed snake in Houston, Texas. He gave some additional measurements (see Table 1). McAllister et al. (2008) found O. faranciae in western mudsnake,

269

Farancia abacura reinwardtii Schlegel from Arkansas. Ophiotaenia faranciae was also reported from Nerodia cyclopion cyclopion by Brooks (1978), but this report is doubtful given the strict host specificity of most Ophiotaenia species (see de Chambrier et al., 2021; Diard et al., 2022).

Species inquirendae

Ophiotaenia lactea Leidy, 1855 – species inquirenda.

Material examined: None.

Type- and only host: Tropidonotus sipedon (= Nerodia sipedon ).

Type locality: Not indicated, possibly eastern USA (Pennsylvania?) .

Type-material: Not deposited.

Remarks

This species was very briefly described by Leidy (1855) on the basis of immature specimens and no illustrations were included. The original description does not include taxonomically important characters, except for the shape of the scolex, which is probably as broad as the neck (“Head small, continuous with the neck”), and there is no type material. Therefore, O. lactea was considered a species inquirenda by La Rue (1911, 1914) and de Chambrier et al. (2017).

Anderson (1935) discussed the validity of O. lactea . Although he mentioned some metrical differences between this species and O. perspicua , he concluded that the former taxon was a senior (!) synonym of O. perspicua “in recognition of complete, clear description [= of O. perspicua ], rather than priority.” However, the available data do not support this possible synonymy, as species of Ophiotaenia have quite strict host specificity of species of Ophiotaenia and another species

( O. tkachi n. sp.) occurs in N. sipedon , not O. perspicua ( Fig. 8 ).

Ophiotaenia variabilis ( Brooks, 1978) Ammann et de Chambrier, 2008 – species inquirenda Fig. 3M–P , 6E .

Syn.: Proteocephalus variabilis Brooks, 1978 .

Material examined: Holotype from Nerodia rhombifer rhombifer, Ouachita Parish, Monroe , Louisiana ( USNM 1369891); two paratypes from N. cyclopion, Rockefeller Wildlife Refuge, Cameron Parish , Louisiana ( USNM 1369892), all collected by Daniel R. Brooks; one voucher, N. cyclopion , Head of Island, Livingston Parish, Louisiana, Lance W. Fontenot ( USNM 1380490).

Type-host: Northern diamond-backed watersnake, Nerodia rhombifer rhombifer ( Colubridae : Natricinae).

Additional hosts: Mississippi green watersnake, Nerodia cyclopion ( Colubridae : Natricinae); Mexican garter snake, Thamnophis eques (Reuss) ; blackbelly garter snake, Thamnophis melanogaster (Peters) .

Type-locality: Louisiana, USA.

Type-material: Holotype – fragments of at least two (!) specimens on four slides ( USNM 1369891 – Fig. 3M and N ); 2 paratypes – two markedly different specimens on two slides ( USNM 1369892 – see Remarks and Fig. 3O and P ).

Distribution and records: USA – Louisiana; Mexico – Estado de Mexico, Jalisco and Michoacan ´( Brooks, 1978; Fontenot and Font, 1996; P´erez-Ponce de Le´on et al., 2001).

Representative DNA sequences: Not available.

Remarks

This species was described as Proteocephalus variabilis by Brooks (1978) based on specimens from N. rhombifer (type host) and N. cyclopion , all from Louisiana. Although the specific name refers to the variability of meristic characters in this species, the data given in the original description are extraordinarily variable compared to those of other species of Ophiotaenia . For example, the number of testes (77–253), the number of lateral uterine diverticula (49–90), and the position of the genital pore (15–48%) substantially exceed the variation reported for these characters in other congeneric species (see Table 2 in de Chambrier et al., 2021).

Moreover, different values are given for the latter feature in the original description: “Genital pores alternating regularly in anterior 15–35% (30%) of proglottid” and later in the text “Genital pores in anterior 24–48% (32%) of proglottid.” The diameter of suckers (up to

160 μm) accounts for up to 80% [sic!] of the scolex width in O. variabilis (probably the length of the suckers was confused with the diameter because the suckers are elongate and not spherical – Fig. 3M ).

An examination of the holotype and two paratypes from the USNM revealed that at least two different species were used for the species description. In addition, the quality of the type specimens is poor. In particular, the scolices of the two specimens designated as holotype are collapsed and severely deformed – they are unnaturally narrow ( Fig. 3M and N ). The scolex of the probable ‘true’ holotype (neither of the two specimens on the slide is labelled as holotype) is 193 μm wide and the suckers are 118–120 μm long ( Fig. 1M ); their shape and position differ markedly from what Brooks (1978) depicted in his Fig. 15 .

In the mature proglottids of the holotype, 92–123 testes (x = 112, n = 7) were counted by the present authors, which falls within a very wide, quite unusual range given for the species in its original description (77–253). In contrast, the vaginal sphincter, indicated as absent in the original description, is very well developed ( Fig. 6E ) and is present in the holotype and both paratypes. The gonopores of the holotype are slightly pre-equatorial (at 37–43% of proglottid length), whereas they are much more anterior (at 21–30% of proglottid length) in one of the two paratypes.

The two paratypes from N. cyclopion from Louisiana differ strikingly in size: one paratype (no. 1232-23) is minute and has a small scolex 225 μm wide with spherical suckers 75–80 μm in diameter ( Fig. 3O ), while the other paratype (no. 1232-24) is much larger, with the broken scolex at the margin of the coverslip; its width is about 430 μm and the deformed suckers measure up to 230 μm in length ( Fig. 3P ), i.e., they are three times larger than those of the first paratype.

Thus, it is evident that the original description of O. variabilis was based on a mixture of at least two morphologically distinct species from two different watersnake species. Unfortunately, the type material of O. variabilis is so poor that it is not possible to clearly identify the species identity of individual fragments and to adequately define the species. For this reason, O. variabilis is considered a species inquirenda.

Phylogenetic relationships of Ophiotaenia spp. from watersnakes.

A phylogram resulting from the maximum likelihood analysis based on the 28S rDNA sequences of proteocephalids included in Alignment 1 is presented in Fig. S1 . The maximum likelihood analyses based on the 28S rDNA sequences in Alignment 2 recognised four species of Ophiotaenia in our material obtained from watersnakes ( Fig. 9 ). Seven novel sequences of Ophiotaenia spp. obtained from the watersnakes clustered in a well-supported clade. Within this clade the sequences of O. currani n.

sp. (one isolate), O. europaea (two isolates), and O. tkachi n. sp. (four isolates) formed a subclade. The sequence divergence in this subclade ranged between 0 and 0.4% (0–4 nt) with sequences of O. europaea and O. tkachi n. sp. being identical and differed from a sequence of O. currani

n. sp. by 0.42% (4 nt). Sequences of O. perspicua and O. grandis formed a well-supported subclade (supported in the Bayesian inference analysis). Sequences of two isolates of O. perspicua were identical and differed from a sequence of O. grandis by 1.1% (15 nt) (see Supplementary Table 1 for detail).

The maximum likelihood analyses based on the cox 1 alignments (Alignments 3 and 4) produced phylograms with different topology ( Figs. 10–11 ). In the phylogram resulting from analysis of Alignment 3 ( Fig. 10 ) the sequences of all isolates of Ophiotaenia spp. from watersnakes formed a well-supported clade. The sequence divergence between species in this clade ranged between 7.6 and 13.4% (122–216 nt) with O. currani n. sp. and O. tkachi n. sp. demonstrating the lowest interspecific divergence, and O. grandis and O. perspicua demonstrating the highest interspecific divergence. The intraspecific divergence of O.

tkachi n. sp. was 0–1.7% (0–27 nt) (see Supplementary Table 2 for detail).

In the phylogram resulting from analysis of Alignment 4 ( Fig. 11 ) the sequences of O. europaea , O. currani n. sp. and O. tkachi n. sp. formed a well-supported clade. In contrast to the results of the 28S rDNA analysis,

270

271 272

the cox 1 phylogram demonstrated the distinct status of three species. The sequence divergence between species in this clade ranged between 3.4 and 8.2% (19–45 nt) with O. europaea n. sp. and O. currani n. sp. demonstrating the lowest interspecific divergence, and O. tkachi n. sp. and O. currani n. sp. demonstrating the highest interspecific divergence.

The intraspecific divergence of O. tkachi n. sp. was lower 0–1.6% (0–9

nt) compared to that one in Alignment 2 with longer cox 1 sequences. Interestingly, sequences of O. grandis and O. perspicua branched apart from the clade with O. europaea , O. currani n. sp. and O. tkachi n. sp., although this position was not supported. The sequence divergence between these species was 13.3% (73 nt) (see Supplementary Table 3 for detail).

Figures

Fig. 3. Scolices of Ophiotaenia spp. from watersnakes (Colubridae) in North America. A, B – holotype and voucher of O. perspicua La Rue, 1911, Illinois, USA (USNM 1348631); C, D – O. perspicua from host US 951, Oklahoma, USA (MHNG-PLAT-0130131); E – O. perspicua from host USA 13, Tennessee (MHNG-PLAT-0035370); F, G – anterior parts and scolices of O. perspicua from the same host individual as holotype, Illinois, USA (USNM 1351794); H – holotype of Ophiotaenia laruei n. sp., Illinois, USA (USNM 1351794); note large, almost triangular suckers; all specimens from Nerodia rhombifer rhombifer; I–K – Ophiotaenia currani n. sp. from host USA 22, Mississippi, USA (MHNG-PLAT-0063341); L – holotype of Ophiotaenia tkachi n. sp. from host USA 21, Louisiana, USA; all specimens from Nerodia fasciata confluens; M, N – holotype (?) and voucher of Ophiotaenia variabilis (Brooks, 1978) species inquirenda from N. rhombifer, Louisiana, USA (USNM 1369891); O, P – paratypes of O. variabilis species inquirenda from Nerodia cyclopion, Louisiana, USA (USNM 1369892).

Fig. 6. Ophiotaenia spp. from watersnakes (Colubridae) in North America. A–C – early pregravid proglottid, gravid proglottid and terminal genitalia of Ophiotaenia currani n. sp. from Nerodia fasciata confluens (USA 22), Mississippi, USA (MHNG-PLAT-0063341), ventral view; D, E – terminal genitalia of O. tkachi n. sp. from N. fasciata confluens (US 952), Oklahoma, USA, and holotype of O. variabilis (Brooks, 1978) species inquirenda from N. rhombifer, Louisiana, USA (USNM 1369891), dorsal view.

Fig. 9. Maximum likelihood phylogram based on 28S rDNA data for selected members of the Proteocephalidae. Bootstrap support from maximum likelihood (ML) and Bayesian inference (BI) nodal support are indicated as ML/BI; values <0.90 (BI) and <70 (ML) are not shown. The scale bar indicates the expected number of substitutions per site. The newly generated sequences are highlighted in bold. The hosts are indicated by silhouettes. The sequence of O. europaea obtained from a fish intermediate host is indicated by an asterisk. Abbreviations of the U.S. states where the samples were collected: AR – Arkansas; MS – Mississippi; OK – Oklahoma. Abbreviations of biogeographical regions: NEA – Nearctic; NEO – Neotropical; PAL – Palaearctic.

Fig. 10. Maximum likelihood phylograms of cox1 data (Alignment 3; 1608 nt) for selected members of the Proteocephalidae. Bootstrap support from maximum likelihood (ML) and Bayesian inference (BI) nodal support are indicated as ML/BI; values <0.90 (BI) and <70 (ML) are not shown. The scale bar indicates the expected number of substitutions per site. The newly generated sequences are highlighted in bold. Abbreviations of the U.S. states where the samples were collected: AR – Arkansas; MS – Mississippi; OK – Oklahoma. Abbreviations of biogeographical regions: NEA – Nearctic; NEO – Neotropical.

Fig. 7. Ophiotaenia tkachi n. sp. from Nerodia fasciata confluens, host USA 21, Louisiana, USA, holotype (MHNG-PLAT-0063340). A, B – pregravid proglottids, ventral view; C, D – vaginal canal of unusual structure, ventral view (C) and en face view (D).

Fig. 8. Ophiotaenia tkachi n. sp. from Nerodia fasciata confluens, host US 952 (MHNG-PLAT-0130134) (A, B, E, F) and N. erythrogaster, host US 1118 (G, H), and O. perspicua La Rue (1911) (C, D) from Nerodia rhombifer rhombifer (US 951), all samples from Oklahoma, USA. A, B – cross sections at the testicular and ovarian level, respectively; C, D – eggs; E–H – unripe eggs (without fully formed oncospheres), drawn in distilled water.

Fig. 1. Ophiotaenia perspicua La Rue, 1911 (USNM 1351794) (A) and O. laruei n. sp. (B) from Nerodia rhombifer, Illinois, USA (USNM 1696448), Ophiotaenia sp. (C) from Thamnophis sirtalis, Michigan, USA (USNM 1351911), and O. perspicua (D) from N. rhombifer and N. fasciata, Illinois, Alabama, Texas and Louisiana, USA. A–C – unmounted specimens from vials; D – mounted specimens; note erroneous labelling of the holotype of O. perspicua (USNM 1348631; dotted) as Proteocephalus variabilis Brooks, 1978.

Fig. 11. Maximum likelihood phylograms of cox1 data (Alignment 4; 549 nt) for selected members of the Proteocephalidae. Bootstrap support from maximum likelihood (ML) and Bayesian inference (BI) nodal support are indicated as ML/BI; values <0.90 (BI) and <70 (ML) are not shown. The scale bar indicates the expected number of substitutions per site. The newly generated sequences are highlighted in bold. Abbreviations of the U.S. states where the samples were collected: AR – Arkansas; MS – Mississippi; OK – Oklahoma. Abbreviations of biogeographical regions: NEA – Nearctic; NEO – Neotropical.