The tapeworms of the Bothriocephalidea, which are found in North American freshwater fish (in total 15 species - Figure 2), do not form a monophyletic group. Instead, they belong to five separate, not closely related lineages of all three families. Relatively little progress has been made on bothriocephalidean tapeworms in North America since 2000, with the exception of studies on species of Bothriocephalus.

The taxonomic history of this species-rich genus is confused due to the generally uniform morphology of its species, the inadequate descriptions of many taxa, and the poor quality of museum specimens (15). One of the most common species, Bothriocephalus cuspidatus, has been reported from 33 North American fish species (1, 16). A closer examination of ‘B. cuspidatus’ from different hosts based on specimens collected using uniform methods of heat kill and fixation, provided evidence that several separate species have been grouped under the name B. cuspidatus in the past (15).

The combination of morphological and molecular data provided evidence for divergence and host specialization patterns in these tapeworms. As a result, B. kupermani has been described from sunfishes (Lepomis spp.) and three other putative new taxa from Ambloplites rupestris, Micropterus dolomieu and M. salmoides are to be formally described (15, 17). These data, which demonstrate narrow host specificity of individual species, also have implications for faunistic surveys, biodiversity assessments and ecological studies.

A phylogenetic study by Brabec et al. (18) showed that the genus Bothriocephalus is an artificial assemblage of several not closely related lineages. Therefore, Scholz et al. (19) proposed the new genus Bothriocestus to include freshwater species previously placed in Bothriocephalus, including Holarctic B. claviceps (type species) and Nearctic B. cuspidatus and B. kupermani. The establishment of Bothriocestus was important to resolve long-standing phylogenetic disagreements, which were further complicated by the fact that the species of the new genus and those of Bothriocephalus do not differ by distinct morphological characters (19).

This order includes monozoic tapeworms, i.e., they have a single set of genital organs and no proglottisation (20, 21). Most of the North American species occur in suckers (Catostomidae), where the group has spread relatively recently. These fishes are abundant and widespread in North America. The high level of research activity over the last two decades has enabled the revision of four genera, namely Archigetes, Biacetabulum, Glaridacris and Promonobothrium [see (22–27)].

In addition, 17 new species have been described (most of them from the southern USA) in the following genera: Archigetes (3 spp.), Biacetabulum (5), Dieffluvium (1) Isoglaridacris (6) and Promonobothrium (2). Three new genera have been proposed, namely Homeomorpha [recently synonymized with Isoglaridacris by (28)], Megancestus and Pseudoglaridacris (23, 29, 30). These genera represent important phylogenetic lineages and were proposed to reconcile the genus classification with the molecular phylogenetic data.

Currently, 73 species of Caryophyllidea are recognized as valid in North America (Figure 2), which corresponds to an increase of 30% compared to the end of the 20th century. The distribution of all taxa shows that species diversity is increasing from north to south, possibly due to the greater diversity of catostomids in the southern part of the USA. Molecular data suggest that there may be undescribed species. There is a large gap in our knowledge of tapeworms of endemic catostomids in the western part of North America, as shown by the recent description of two new Isoglaridacris species from these fishes in northwestern Mexico (28).

The taxonomy of this group is probably the most difficult among all freshwater fish tapeworms, especially with regard to the polyphyletic genus Proteocephalus. In addition to freshwater fish, species of this group also infect amphibians, reptiles and even a mammal (13). In total, 15 species (instead of 23) of the Proteocephalus species-aggregate (= ‘true’ Proteocephalus or Proteocephalus sensu stricto) were recognized as valid (11). The Proteocephalidae are the second most species-rich group, but the number of recently described species is low (Figure 2; Supplementary Table S1).

The main reasons for this are the poor quality of historical material and the lack of molecular data. In addition, most research activities have focused on the revision of individual genera, and some of newly collected material, which most likely contains new species, is still being analyzed (unpublished data).

Significant achievements were the correction of misidentifications (e.g., a new species from sticklebacks), the clarification of host specificity (a new species of a newly proposed genus specific to gars) and the reorganization of the phylogenetic tree with regard to the genus and subfamily affiliation of tapeworms from ictalurids and gars (11, 12, 31–34). It is likely that future studies on new, properly fixed material, mainly from salmonids, will reveal the existence of further species (35, 36).

In contrast, little progress has been made in clarifying the inadequately clarified taxonomic situation of the Proteocephalus species-aggregate of salmonids and minnows. Hoffman (1) reported 11 species from salmonids, but most taxa were synonymised with P. longicollis on the basis of morphological studies, but without molecular data (13, 37, 38).

Recently, however, it was found that some of the synonymised species are indeed valid, such as P. exiguus, a specific parasite of North American whitefish (Coregonus spp.) (35, 39). Scholz et al. (36) commented on the taxonomic status and possible validity of Proteocephalus species described from North American whitefish (Coregonus spp.), but their taxonomic status cannot be clarified without a detailed morphological and especially molecular assessment of new, properly processed specimens.

In addition to these revisions, new genera for tapeworms specific to ictalurids and bowfin that are endemic to North America were established to reflect phylogenetic findings: (1) Essexiella was proposed to include Corallobothrium fimbriatum from the channel catfish Ictalurus punctatus; and (2) Laruella was proposed for Proteocephalus perplexus from bowfins, as it is phylogenetically distinct from the ‘true’ Proteocephalus species-aggregate (Figure 2) (12, 33).

Of three small orders reported from North American freshwater fishes, namely Amphilinidea (1 species, Amphilina japonica, occurs in sturgeon), Haplobothriidea (2 endemic species of Haplobothrium occur in bowfin) and Spathebothriidea (3 species occur in a wide range of freshwater and Diplocotyle olrikii also in marine fishes), no new species were described (Supplementary Table S1).

Although these orders are represented by only a small number of species in North America, they are important from a phylogenetic perspective (as evolutionary relicts), because of their host specificity (in evolutionarily ancient fish hosts such as sturgeons and bowfin), and because of the overall rarity of most species, with the exception of Bothrimonus sturionis and Cyathocephalus truncatus, which has been reported from a number of non-closely related fish hosts (1, 3). Moreover, B. sturionis was the first fish tapeworm reported in North America (40), but its validity remains unclear (41).

One challenge is the need to review two Haplobothrium species that have been reported in the past from the ruddy bowfin (Amia calva), following the recent recognition of a second Amia species in North America (42). The eyetail (or emerald) bowfin, Amia ocellicauda, has a large range and is found in the Great Lakes and Mississippi Basin, while A. calva is only distributed in the southeastern United States, particularly in Florida and adjacent states with short rivers leading to the Gulf of Mexico (42). Since Haplobothrium bistrobilae occurs exclusively in Florida, it is necessary to genetically compare the Haplobothrium tapeworms of both fish species. With strict host specificity (occurrence of H. globuliforme in A. ocellicauda and H. bistrobilae in A. calva), the distribution of the tapeworms would help to confirm host speciation.

Brabec et al. (18) assessed the relationships of the order, parasitising marine and freshwater bony fishes, using molecular phylogenetic analyses with multiple genes covering about 70% of the currently recognized genera. The new phylogenetic data challenged the morphological classification. In contrast, some patterns in host utilization and environments (freshwater versus marine species) were revealed, but biogeographic patterns are not evident [Figure 1; (18)]. Some of the most basal taxa including North American endemic genus Marsipometra from paddlefish Polyodon spathula are parasite of freshwater fish, but the origin of the order cannot be resolved based on the available molecular data [Figure 3; (18)].

The order appears to be monophyletic, but consists of non-monophyletic families. The family Bothriocephalidae forms the most derived lineage of the order, with a single freshwater clade I that includes North American taxa, currently classified in Bothriocestus, as well as the invasive Asian fish tapeworm, Schyzocotyle acheilognathi (syn. Bothriocephalus acheilognathi), which is of major veterinary and ecological importance (18, 48). Other North American genera (Eubothrium and Triaenophorus) belong to the most basal, but non-monophyletic family Triaenophoridae with Holarctic distribution of many freshwater taxa, with the exception of the North American endemic genus Marsipometra, Eubothrium tulipai from the northern pikeminnow Ptychocheilus oregonensis and Triaenophorus stizostedionis from the walleye Sander vitreus.

Molecular phylogenetic analyses of the Caryophyllidea (and also the Proteocephalidae – see below) do not support traditional taxonomic significance (14, 20). It was based on the position of the reproductive organs (testes, ovary, vitelline follicles and uterus) in relation to the internal longitudinal musculature, which appears to be homoplastic in both groups (13, 21).

As a result, Scholz et al. (49) reorganized the classification of the Caryophyllidea at the family level. Although all existing families remained valid, their composition was significantly changed to reflect the assumed phylogenetic relationships. Most endemic Nearctic taxa are currently placed in the monophyletic family Capingentidae, which represents a relatively recent radiation in North America, including the formation of numerous scolex types and the presence of an external seminal vesicle in most taxa (which is never present in other families). The remaining species belong to a largely Palaearctic group forming the family Caryophyllaeidae. With one exception (Caryophyllaeides fennica), all species of this family occurring in North America are invasive and were introduced to America with the common carp (Cyprinus carpio), which is not native to North America.

The North American taxa (family Capingentidae) are related to those of the Caryophyllaeidae, which occur in the cypriniforms (Cyprinoidei and Cobitoidei) and almost exclusively in the Palaearctic (Figure 4) (49). Interestingly, two species from the native Nearctic leuciscids, Edlintonia ptychocheila from the Northern pikeminnow and Pliovitellaria wisconsinensis from the golden shiner Notemigonus crysoleucas, are the most basal taxa of the North American species of the Capingentidae, which may indicate a host switch in their ancestors from cyprinoids to suckers (Catostomidae) and a subsequent radiation in these hosts (49).

Similar to the Caryophyllidea, the previous classification was largely artificial and did not correspond to the actual relationships of the individual groups (13). Molecular phylogenetic studies have not supported the existing subfamilies, most of which are not monophyletic (50, 51). The most species-rich genera, including Proteocephalus with species that parasitise freshwater fishes in North America, are polyphyletic, and form several distantly related lineages (13, 51).

The North American fish proteocephalids have been divided into two subfamilies, the Proteocephalinae (species of ‘Proteocephalus’), which are characterized by a simple scolex, and the Corallobothriinae, which possess a metascolex, including tapeworms of Nearctic catfishes (Siluriformes: Ictaluridae) (13). The metascolex is the posterior part of a divided scolex consisting of folds of tissue that generally enclose or conceal the suckers [see (10)]. Although the presence of a metascolex has long been used for classification, recent molecular studies show that it is a homoplastic feature that has evolved multiple times independently (13, 50, 51). Therefore, the North American ‘Corallobothriinae’ were transferred to a new subfamily, Essexiellinae, by Scholz et al. (31).

The North American members of the family Proteocephalidae currently belong to at least five different lineages [Figure 5; see also Figures 4–6 in (52)]:

The phylogenetic position of Testudotaenia testudo, a parasite of turtles and possibly accidentally found in bowfin, is not clear (54).

Interestingly, the oldest, relict fishes such as gars (Lepisosteiformes) and bowfin (Amiiformes), are definitive hosts of relatively recently evolved tapeworms, while the most recently evolved teleosts such as centrarchids and percids harbor tapeworms belonging to relatively basal Proteocephalidae (32, 51). This indicates several host switches of the Nearctic proteocephalids and their ancestors.

Recent molecular phylogenetic studies confirmed the monophyly and validity of all three small cestode orders (Amphilinidea, Haplobothriidea and Spathebothriidea) found in North American freshwater fishes, as circumscribed by Khalil et al. (55). Phylogenetic studies confirmed the basal position of the monozoic, non-strobilised Amphilinidea, which is one of the two earliest diverging orders of tapeworms (45). Overall, these three small orders – possibly due to their low veterinary importance and small number of species – remain poorly studied and require future investigation. No further changes have been made to the classification, with the exception of new molecular data and revision of the Spathebothriidea (41).

This order includes 15 species with a variable range of host specificity: 7 strict specialists (e.g., the bothriocephalideans Eubothrium tulipai and Triaenophorus stizostedionis), 4 congeneric specialists, 3 suprageneric specialist and 1 high generalist (invasive, non-native Schyzocotyle acheilognathi) (Figure 6; see also Supplementary Table S1).

Bothriocestus cuspidatus has been reported from 33 species of freshwater fishes from 13 families in 12 orders (1, 16). However, Choudhury and Scholz (15) showed that B. cuspidatus does not exhibit such broad host specificity. Instead, several species specialized on their own definitive hosts were grouped together as B. cuspidatus; they are congeneric or suprageneric specialists (15, 18). Several hosts have been reported based on immature or juvenile tapeworms that cannot be reliably identified by morphology alone.

Schyzocotyle acheilognathi, commonly known as Asian fish tapeworm, is a highly invasive cestode that infects a wide range of freshwater fish (> 300 species), especially cyprinoids (minnows in North America) (56). The parasite poses a significant threat to aquaculture and the biodiversity of native fish worldwide, including North America (57). It is most pathogenic in newly acquired host species, which appears to be the case in Australia, Europe and North America (1, 58, 59). The species was transferred to genus Schyzocotyle, based on a molecular phylogenetic study by Brabec et al. (18). The two species of this resurrected genus are characterized by a heart-shaped scolex with narrow, deep bothria (57).

The North American Eubothrium species have very different host patterns. The two species most commonly parasitising salmonids, Eubothrium crassum and E. salvelini, were reported by Hoffman (1) from at least 16 fish species from different families and orders. However, E. crassum mainly infects trout and salmon (Salmo spp.), while E. salvelini is typically associated with charr (Salvelinus spp.) and whitefish (Coregonus spp.). However, both species have been detected in rainbow trout (Oncorhynchus mykiss) in Europe (60), and this may also be the case in other Pacific salmon of the genus Oncorhynchus, although this remains to be confirmed. In contrast, the two remaining Eubothrium species in North American freshwater fishes, E. rugosum from the burbot, Lota lota, and E. tulipai, from the northern pikeminnow, are strict specialists [(61); Supplementary Table S1].

There are numerous records of bothriocephalideans in atypical or unusual fish hosts [see (1)], but these should be examined critically as they are likely accidental infections or postcyclic parasitism.

The Caryophyllidea comprise 73 species, the vast majority of which exhibit narrow host specificity: 29 strict specialists (e.g., six Isoglaridacris species, five Biacetabulum species and all three Penarchigetes species), 18 congeneric specialists, 18 suprageneric specialists, 7 low generalists and 1 high generalist (Archigetes sieboldi in the Cypriniformes and Atheriniformes) (Supplementary Table S1). The first two categories with the narrowest host specificity (tapeworms living in a single host species or genus) account for almost two thirds (64%) of all Caryophyllidea in North American freshwater fish (Figure 6).

In contrast, some species of Promonobothrium and Pseudoglaridacris have a broader range of definitive hosts and are found in fish of different genera or tribes, but rarely in fish of different subfamilies (Catostominae versus Ictiobinae) [(6, 22, 23); Supplementary Table S1].

Their host specificity is not sufficiently known, mainly because of the persisting taxonomic problems in this group and the existence of numerous unconfirmed reports of atypical or dubious hosts. The present analysis shows that of the 36 species of Proteocephalidae, 20 are strict specialists, 13 are congeneric specialists and 3 are suprageneric specialists. No species was categorized as low or high generalist (see Supplementary Table S1). Regarding the proportion of species in the category with the highest host specificity (strict specialist), more than half (56%) of the Proteocephalidae belong to this category (compared to 50% in the Bothriocephalidea, albeit with a low total number of species, and 40% in the Caryophyllidea) (Figure 6).

Although recent studies have provided new data on the actual host specificity of many species, there are still large gaps (36). An example of this is the so-called bass tapeworm, Proteocephalus ambloplitis, adults of which have been reported from a number of phylogenetically distant fish groups, notably bass, bowfin and gars. However, it appears that at least two species, including P. australis from gars, have been confused with P. ambloplitis (32). The same is true for metacestodes with a large apical glandular organ, which have been consistently (and automatically) identified as P. ambloplitis and are found in a wide range of numerous fish groups (1, 3). However, they actually belong to two or more different species (32, 62).

Nevertheless, most species of the Proteocephalidae appear to have a narrow host specificity as tapeworms of other orders, including several strict specialists such as species of the Proteocephalus species-aggregate of sticklebacks (Gasterosteidae), American yellow perch (Perca flavescens) and cavefish (Amblyopsidae) (11, 34, 53).

Laruella perplexa is a specific parasite of bowfin and has been morphologically and genetically characterized, albeit with a relatively complex genetic structure of isolates from different parts of North America (33). Therefore, the presence of L. perplexa (or another closely related and morphologically similar species) in the two recently distinguished bowfin species [see (42)] should be confirmed.

A recent study of tapeworms of gars (Lepisosteidae), ancient and relict group of fish, confirms a narrow host specificity and restricted distribution of these proteocephalids, with closely related species occurring in different zoogeographical regions, i.e., Nearctic and Neotropical (32).

Amphilina japonica (syn. A. bipunctata) (Amphilinidea), the only Amphilinidea from North America, is a specialist and occurs in the body cavity of the white sturgeon, Acipenser transmontanus in North America and of Sinosturio mikadoi in Japan. However, molecular data are still lacking to confirm the validity of this species, as it closely resembles A. foliacea from Eurasian sturgeons (63).

Haplobothrium globuliforme (Haplobothriidea) is a typical parasite of bowfin but has also been reported from the American eel, Anguilla rostrata. The latter fish was most likely only an accidental or postcyclic host. Similarly, a very rare H. bistrobilae, the validity of which remains to be confirmed, has been found in A. calva and in brown bullhead, Ameiurus nebulosus, although the latter host is questionable.

Two species of the relict order Spathebothriidea [see (41)] have a relatively broad spectrum of definitive hosts. Bothrimonus sturionis, a parasite of sturgeons (Acipenseridae), has been reported from 13 other fish hosts, including species from three salmonid genera, in Canada (1, 3). However, these reports are most likely a confusion with the morphologically similar Diplocotyle olrikii from marine and anadromous fish, as members of the genus Bothrimonus appear to be specific parasites of sturgeons (41, 64). The Holarctic Cyathocephalus truncatus was found in 16 species from six orders. With the exception of B. sturionis, the North American freshwater species of the order Spathebothriidea are high generalists.

The tapeworm fauna of the freshwater fishes of North America is unique and peculiar, with a high degree of endemism (4/5 of the species – Figure 2). Nearctic elements dominate (96 species, i.e., 79%), with a small proportion of Holarctic taxa (15, i.e., 12%) and a negligible number of Neotropical taxa (5 in Mexico, i.e., 4%) and species introduced to North America with their non-native fish hosts, in particular the common carp (6, i.e., 5% – Figure 3), which harbors five tapeworm species in North America (1 of the Bothriocephalidea, 4 of the Caryophyllidea).

The only species of Caryophyllidea with a circumboreal distribution, Caryophyllaeides fennica, was recently found by Oros et al. (65) in the chiselmouth Acrocheilus alutaceus, an endemic cyprinid in the northwestern Nearctic (Oregon). A very low genetic divergence between C. fennica representatives from the Palaearctic and the Nearctic indicates a relatively recent colonization of the Nearctic by this cestode across the Beringian land bridge (65).

There are at least 14 species (only 11% of all species) with a Holarctic distribution, namely Amphilina japonica (Amphilinidea), Bothriocestus claviceps, Eubothrium crassum, E. rugosum, E. salvelini, Triaenophorus crassus, T. nodulosus (Bothriocephalidea), Proteocephalus ambiguus, P. filicollis, P. macrocephalus, P. tetrastomus, P. torulosus (Proteocephalidae), Cyathocephalus truncatus and Diplocotyle olrikii (Spathebothriidea) (11, 19, 66, 67).

The vast majority of species occur in the USA (121 species), while less than half of the species diversity was found in Canada (58 species) and Mexico (14 species), including nine species in the Nearctic part of the country and only five Neotropical taxa (Supplementary Table S1).

As far as the main hydrological regions of North America are concerned, it is not possible to provide detailed and reliable data, as there are numerous misidentifications of individual records. It can roughly be summarized that the Bothriocephalidea and the Proteocephalidae have a similar species diversity in Canada and the USA, with the number of species increasing slightly from Canada toward the south of the USA, while very few species of these orders have been reported from Mexico. In the Caryophyllidea, there is a clear increase in species diversity from Canada (20 species) toward the south and south-east of the USA with the highest number of species. However, the freshwater fish tapeworms in Canada, the southern USA and Mexico are still little studied despite the high diversity of fish.