Almost 50 years ago, Simmons26 called gyrocotylideans a “century-old enigma” and this status still persists despite the advent of more advanced identification methods3. The poor understanding of the group (e.g., the complete life cycle of none of the species is known) is linked with the scarcity of available data and the biological peculiarities of these tapeworms and their holocephalan hosts. In particular, most of the host species are rarely available deep-sea dwellers, which often could not be examined fresh or were frozen with their parasites prior to examination. If isolated alive, gyrocotylideans exhibit an unusual morphological variability due to the contraction of their large bodies and as a result of different fixative procedures which were tested to ensure their relaxation (e.g.27). Despite these issues, several comprehensive studies have been conducted, e.g.15,16,21,28, which provided deep insight into the biology, ecology and taxonomy of these enigmatic tapeworms. Nevertheless, the poor quality of the specimens studied and the use of different, not always appropriate, methods of parasite fixation, unintentionally affected the quality of morphological descriptions of most gyrocotylidean species, which prevented the establishing of clear morphological borders to delimit individual species. As a result, the informative value of morphological traits used for species delimitation should be re-assessed, based on the simultaneous use of molecular data, i.e., the use of hologenophores to match morphology and molecular data. Existing problems with species delimitation and morphological variability even led to complete omission of morphological characterisation of two new species described just recently6.
Herein, the genotyping of the Gyrocotyle spp. specimens acquired in Taiwan revealed four distinct genotypes, each one more related to the North Atlantic isolates identified as “Gyrocotyle urna” off Ireland (the isolate is genetically diverse from G. urna off Norway), “G. rugosa” off Alaska (probably misidentified, see below), G. discoveryi off Ireland and G. confusa off Norway, respectively, than to each other.
In addition to casting doubts on the restriction of gyrocotylideans to individual oceans, our data also question the proclaimed strict host specificity3,7, because specimens of Gyrocotyle sp. genotype 3 were found in two hosts species, which are not the closest relatives to one another—C. phantasma and C. cf. argiloba (Fig. 4). Broader host specificity was also reported for G. fimbriata, which was found in Hydrolagus colliei and Chimaera phantasma, and for G. rugosa, recorded in Callorhinchus callorynchus and C. milii14,15,24,29. Gyrocotyle urna was also found in several holocephalans, including Chimaera monstrosa, Callorhinchus callorynchus, Hydrolagus ogilbyi Waite and H. colliei24,29,30. In contrast, Bandoni & Brooks16 revised the host spectrum of this parasite, considering C. monstrosa as the only host of G. urna.
The suitability of the molecular markers employed for this group also requires attention, because a considerable amount of phylogenetic information was also lost in the un-rooted dataset due to treatment of the numerous gaps in the 28S rRNA alignment. The involvement of partial COI gene sequences seemed to be informative for estimating gyrocotylidean phylogeny, because we obtained a no-gap COI alignment and improved support for some nodes in the three-gene network. The suitability of this marker requires assessment employing further taxa, because except for our isolates off Taiwan and Argentina, only a single sequence of the COI gene (i.e., that of G. urna off Norway; GenBank acc. no. JQ268546) is currently available.
A single specimen of Gyrocotyle sp. genotype 4 was conspicuously different morphologically from the remaining ones by having few folds on the lateral margins, many acetabular spines, a narrow funnel and a small rosette. However, its formal description as a new species would be premature, because only a single specimen was found. Morphological differences among the specimens of the other genotypes were not so obvious, even though a careful examination of the hologenophores allowed us to find several morphological traits that were characteristic for particular genotypes (see “Results” section). Among them, the number of acetabular spines and the distribution of the body spines and their size may be potentially useful for species differentiation, especially because the body contraction can hardly affect them. Since body contraction cannot be absolutely excluded even when live specimens are properly fixed, its effect could be overcome to some degree by an evaluation of ratios related to the main body dimensions (e.g., length of uterine sac/total body length) rather than comparison of total measurements of internal structures.
The specimens off Taiwan most probably represent several new species, but we decided not to describe them formally as new taxa, mainly because of the shortage of comparative data. In addition to these specimens, two hologenophores of Gyrocotyle rugosa off Argentina were examined, which made it possible to characterise the type species of the genus. The host of G. rugosa described by Diesing10 was questionable until Callorhynchus antarcticus (= C. callorynchus—see31) off New Zealand was finally established as its currently accepted type host3,32. Gyrocotyle rugosa was found in coastal waters of South America, South Africa and New Zealand as a parasite of C. callorynchus and C. milii, suggesting its broader host specificity16,24. Our specimens from C. callorynchus off Argentina were identified as G. rugosa based on crenulated (i.e., without any folds) lateral margins, a tiny uterine sac, a branched uterus and embryonated eggs in the uterine sac; the latter two traits are unique to this species21. Genetically, it clustered with an unspecified isolate of Gyrocotyle from C. milii off Australia, and these specimens seem to be conspecific.
In contrast, an isolate from Hydrolagus colliei off Alaska identified as G. rugosa (GenBank acc. nos. AF286925 and AF124455) was apparently misidentified, because (i) it was found in an unrelated definitive host (H. colliei belongs to the family Chimaeridae, whereas the type host to the family Callorhinchidae), (ii) its distant geographic origin (the type locality of G. rugosa is unclear, but it is definitely in the Southern hemisphere), and (iii) its genetic divergence from our isolate of G. rugosa from the type host off Argentina. The isolate from H. colliei may represent Gyrocotyle fimbriata or G. parvispinosa, which have been reported from this host off the Pacific coast of North America, but its identification was not possible because morphological vouchers were not available to the present authors.
Gyrocotylideans were generally considered to be oioxenous, i.e. strictly specific parasites sensu Euzet and Combes33, with each gyrocotylidean species parasitising a single holocephalan species. Although several species were reported from two or more hosts species16,24, these findings are usually considered as misidentifications due to the unclear taxonomy of the order. Moreover, some holocephalans, such as Ch. monstrosa, H. colliei, H. affinis, and Ca. callorynchus, were often found to harbour two or more gyrocotylidean species, one common and the other rare9,10,21,22,23. Our findings of Gyrocotyle sp. genotypes 1 and 3 in Ch. phantasma and Gyrocotyle sp. genotypes 2, 3 and 4 in Ch. cf. argiloba suggested stenoxenous host specificity (i.e., the occurrence in a few closely related hosts) of gyrocotylideans, because the specimens of genotype 3 were found in both species of Chimaera. The obvious genetic similarity of our G. rugosa specimen from Ca. callorynchus and the isolate of Gyrocotyle sp. from Ca. milii also questions the strict specificity of this group, but morphological vouchers of the latter, which are necessary for the confirmation of their conspecificity, are not available.
Our genetic analyses provided insight into the interrelationships among the gyrocotylideans, even though the absence of a suitable outgroup did not enable us to broadly assess the possible evolutionary scenario of this earliest evolving group of tapeworms. Moreover, genetic data on only half of the nominal species of Gyrocotyle are available, not considering the possibility of misidentifications of previously sequenced specimens, for which hologenophores are not available. However, some clues of host-parasite coevolution can be inferred from the network. The mutual genetic distance of species/genotypes from the same host species suggests multiple colonisation events rather than co-speciation with their hosts within the order. It seems that G. phantasma might have been colonised by Gyrocotyle sp. genotype 1 or genotype 3, because these two genotypes are not the closest relatives in our analyses. The same pattern is obvious for C. cf. argiloba parasitised by Gyrocotyle sp. genotype 2, 3 and 4, and also for C. monstrosa, which harbours G. urna, G. confusa and G. nybelini. Indeed, Colin et al.27 considered these species from C. monstrosa to be conspecific, but our genetic data support the validity of three separate and genetically distant species. Moreover, G. nybelini formed by far the most distant lineage among all isolates, which may suggest the validity of the genus Gyrocotyloides Furhmann, 1931.
Genetic divergence of congeneric tapeworms from the same host species was also observed in several elasmobranch/teleost-cestode assemblages, e.g., Acanthobothrium spp. (Onchoproteocephalidea) and the mumburarr whipray Urogymnus acanthobothrium Last, White & Kyne; Echeneibothrium spp. (Rhinebothriidea) and the yellownose skate Dipturus chilensis (Guichenot); and Pseudoendorchis spp. (Onchoproteocephalidea) and the catfish Pimelodus maculatus Lacepède34,35,36.
The aim of this paper was to provide new insight into the phylogenetic relationships within the enigmatic order Gyrocotylidea, but, in particular, to demonstrate the lack of geographical patterns in the distribution of most its species and the limited suitability of current morphological characteristics for species circumscription. Herein, we have outlined a methodology (fixation of live specimens with hot fixative and the exclusive use of hologenophores) that should be used in future taxonomic, ecological and biogeographical studies of gyrocotylideans in order to reliably circumscribe their actual species diversity and to unravel associations with their hosts, a relict group of marine vertebrates. Gyrocotylideans represent one of the key groups of parasitic flatworms (Neodermata) in terms of a better understanding of their evolutionary history and the switch of free-living flatworms to parasitism.
Source: Ecology - nature.com
