Diversity of trypanorhynch metacestodes in teleost fishes from coral reefs off eastern Australia and New Caledonia

Trypanorhynch metacestodes were examined from teleosts from coral reefs in eastern Australia and from New Caledonia. From over 12,000 fishes examined, 33 named species of trypanorhynchs were recovered as well as three species of tentacularioids which are described but not named. Host-parasite and parasite-host lists are provided, including more than 100 new host records. Lacistorhynchoid and tentacularioid taxa predominated with fewer otobothrioid and gymnorhynchoids. Five species, Callitetrarhynchus gracilis, Floriceps minacanthus, Pseudotobothrium dipsacum, Pseudolacistorhynchus heroniensis and Ps. shipleyi, were particularly common and exhibited low host specificity. Limited data suggested a higher diversity of larval trypanorhynchs in larger piscivorous fish families. Several fish families surveyed extensively (Blenniidae, Chaetodontidae, Gobiidae, Kyphosidae and Scaridae) yielded no trypanorhynch larvae. The overall similarity between the fauna of the Great Barrier Reef and New Caledonia was 45%. Where available, information on the adult stages in elasmobranchs has been included.


Introduction
The identification of significant threats to the coral reefs of the world [9,17] has been partly responsible for focussing attention on the full diversity of reefs rather than simply on the diversity of fish and corals, the most obvious examples of reef diversity. The contributions of other groups of invertebrates to diversity on reefs have been largely overlooked in the past [7,32]. Part of this ''hidden'' invertebrate diversity includes the endoparasites of vertebrates.
In recent years, teleost fish occurring on coral reefs have been recognised as harbouring a particularly diverse array of parasites [20]. Studies to date have focussed either on specific parasite groups such as the Monogenea (e.g. [33]) or Digenea (e.g. [13]), or more recently have examined the diversity of all helminth parasites found in or on specific families of fish such as the Lethrinidae or Serranidae [21][22][23].
Teleosts found on coral reefs are commonly infected with the larval stages (plerocerci, merocerci or plerocercoids -for terminology see Chervy, 2002 [12]) of cestodes of the order Trypanorhyncha, the adults of which are found in the stomach or spiral valves of elasmobranchs. Larval stages occur most commonly in the body cavity but may also be found in the musculature or other sites such as the gill arches [27]. They constitute a significant component of parasite diversity but have frequently been overlooked because of taxonomic difficulties in identification [27]. However, unlike other orders of cestodes found in marine fish, the larval stages have scolex features, including the distinctive tentacular armature, which are identical to those found in the adult and which allow specific morphological identification. Although taxonomic studies of this group of parasites are frequent, ecological studies are few, and while systematic collecting has been undertaken in several parts of the world (Gulf of Mexico, Gulf of California, Java, Borneo, Australia and Hawaii), there are few published descriptions of the faunas encountered in these areas (see Jensen, 2009 [19] for Gulf of Mexico and Palm and Bray, 2014 [29] for Hawaii). Some species of trypanorhynchs (e.g. Grillotia (Christianella) minuta van Beneden, 1858; Gilquinia squali Fabricius, 1794) have also been used as biological tags in teleosts [25] because the larval stages are readily identifiable and because they are long-lived in the intermediate host.
However, such ecological studies of these species are limited.
In this study, we examined the larval trypanorhynch cestode parasites of teleosts, and where applicable the corresponding adults in elasmobranchs, from the Great Barrier Reef (GBR) and compared them with those from similar reef environments in New Caledonia (NC). New Caledonia is separated from the GBR by about 1200 km of deep oceanic waters.

Materials and methods Great Barrier Reef (GBR)
Teleosts and elasmobranchs were collected opportunistically between 1986 and 2010. The two main collecting sites were Heron Island in the southern Great Barrier Reef and Lizard Island in the Northern Barrier Reef. Small numbers of parasites were collected on reefs between these two sites (Mossman, Townsville) and in these instances, the nearest geographical feature on the coast was recorded rather than the specific reef near which the collection was made (Fig. 1).
Metacestodes were collected mainly from body cavities of teleosts, although in some instances they were sought in regions of the body such as the gill arches and musculature. Metacestodes were removed from surrounding cysts (in the case of plerocerci) and the eversion of tentacles was achieved either by shaking vigorously or by applying pressure under a coverslip. Cestodes were fixed in 70% ethanol or 10% formalin and were stained with Celestine blue or carmine (Palm, 2004) [27], dehydrated in ethanol, cleared in methyl salicylate and mounted in Canada balsam. All specimens were identified by IB and have been deposited in either the British Museum (Natural History) (BMNH), the Queensland Museum, Brisbane (QM) or the South Australian Museum, Adelaide (SAM). Some of the records used in this compilation have been published previously in , 2001 [1,3], Beveridge et al., 2000, Campbell & Beveridge, 1996 [8], Palm, 2004 [27], Palm & Beveridge, 2002 [28] and Sakanari, 1989 [34].
Records of adults from elasmobranchs are included only for species in which larval stages have been identified in teleosts; these are based on both published data and specimens held in museum collections. Additional species of trypanorhynch cestodes from elasmobranchs have been found and their larval stages may be found in the future, but for the present study, these records have not been added.

New Caledonia (NC)
Fish were collected opportunistically between 2003 and 2009 generally by line fishing, occasionally by spear fishing and on occasions supplemented by fish obtained from a market. Collections were mainly off Nouméa (Fig. 1). All fish were measured, weighed and photographed. Methods for collection from several host families have been explained elsewhere [21][22][23]. Trypanorhynch plerocerci were opened and compressed between two slides or immersed in hot saline to evert tentacles. Plerocercoids found in the body cavity were also fixed under pressure to evert tentacles. Metacestodes were fixed in 70% ethanol or 10% formalin and were stained with Celestine blue or carmine [27], dehydrated in ethanol, cleared and mounted in Canada balsam. All specimens were identified by IB and have been deposited in the Muséum national  In the parasite-host list ( Table 1), authorities of cestodes are included and host species are listed in alphabetical order without authorities. In instances where both generic and specific names of cestodes have changed, synonyms have been included. In the host-parasite list ( Table 2), fish hosts are arranged in orders, families and genera, but within each group, the order is alphabetical. Authorities of fish are indicated and the parasites are arranged in alphabetical order without authorities.
Authorities of hosts or parasites which are indicated in the lists are not repeated in the text. The systematic arrangement of trypanorhynch taxa follows Palm (2004) [27]. All host names were verified in FishBase [15].

Species found and other data
Larval trypanorhynchs were recovered primarily from the body cavities of the teleosts examined (Figs. 2-7). Plerocerci were usually encountered attached to the mesentery enclosed within white envelopes (Fig. 2), although in some hosts melanisation of the cyst wall had occurred rendering the cysts brown (Fig. 3). Some brown or even black envelopes contained only remnants of plerocerci (Fig. 4). Plerocercoids of tentaculariids were found either in the body cavity or in the gastrointestinal lumen; the latter were not contained within a ''cyst''. Occasionally, plerocerci were found in the musculature and in the gill arches ( Fig. 7), although there was no systematic search of such sites for plerocerci. Merocerci of Molicola horridus occurred in the livers of a limited number of species of teleosts, but the intensity of infection was high and the infections were readily observable at autopsy (Fig. 6).
Species of larval trypanorhynch cestodes found in both teleost (as larvae) and elasmobranch (as adult) hosts at sites along the GBR and off NC are shown in Tables 1 and 2.
From the GBR, the specimens examined were obtained from the dissection of more than 9000 fish, although not all were specifically examined for trypanorhynch cestodes. Likewise, from NC, approximately 3800 fish were examined but the body cavity was not examined in every fish, as explained by Justine et al. [21][22][23]. Consequently, prevalence data were available for some species only and abundance data were not available; for most species only presence-absence data were available (with one exception from Lizard Island).
Members of the Tentacularioidea differ from other trypanorhynch metacestodes as they are present as plerocercoids (= post-larvae) rather than plerocerci [14] and may be found in intestinal contents as well as in the viscera. In New Caledonia, tentacularioids were frequently found in smaller schooling fishes, often being the only trypanorhynchs encountered in these fishes.
In total, 33 named species were found (Tables 1 and 3) as well as three species of tentaculariid cestodes to which no current name could be applied. Lacistorhynchoid and tentacularioid trypanorhynchs dominated the fauna in terms of numbers of species recovered (Table 3), with the otobothrioid and gymnorhynchoid trypanorhynchs being less numerous.
Prevalence data were obtained from 182 fish from various families collected during a single collecting trip to Lizard Island. The prevalence of trypanorhynch larvae was: 4/6 (77%) in scombrids, 5/7 (71%) in lethrinids, 2/13 (15%) in lutjanids, 8/9 (89%) in serranids and 1/109 (0.9%) in apogonids. Other fish families were represented by smaller numbers and were excluded. Remarks This species is similar to N. queenslandensis, but all measurements including those of the hooks are substantially smaller. In addition, the shape of the hooks differs (Fig. 8).
The hook shape aligns the species with N. lingualis (Cuvier, 1817), N. bisulcata (Linton, 1889), N. anthicosum Heinz & Dailey, 1974 and N. hemipristis Palm & Beveridge, 2002, but N. lingualis and N. bisulcata differ in having much larger scoleces (2025-2700 and 2500, respectively) and bulbs (365-425 and 450-505, respectively) while the latter two species have much larger hooks (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40). Consequently, these           Remarks This specimen clearly belongs to Heteronybelinia as the hooks differ markedly in shape on the bothrial versus the antibothrial surfaces of the tentacle. Hook sizes are closest to H. eureia (Dollfus, 1960), but the specimen differs from this species in the number of hooks per half spiral and by the fact that in this specimen the bulbs are entirely posterior to the pars bothrialis while in H. eureia, they do not extend beyond the pars bothrialis. Therefore, this specimen cannot be accommodated within any known species of Heteronybelinia. Remarks Two specimens have been identified as belonging to this species with its characteristic armature. In spite of the fact that the armature of both specimens is identical, scolex measurements differed substantially and for this reason, the measurements of both specimens are presented. The specimen from P. multifasciatus although quite flattened, corresponds more closely with the original description of the species, also from P. multifasciatus from Hawaii [10]. In the specimen from N. sammara, all measurements are shorter but the tentacular armature is identical.

General comments
Although the records of trypanorhynch infections listed here are based on the dissection of thousands of fish from both the GBR and NC, the data collected are based on opportunistic collecting and must be viewed in this light. Few prevalence or intensity data were collected and the data are based largely on the presence of trypanorhynch metacestodes. Fish examined that did not harbour metacestodes were not included in the data presented in the tables but representative examples have been indicated in the results.
In spite of these limitations, the large numbers of metacestodes collected from both regions provide a significant basis for comparing trypanorhynch metacestodes of teleosts inhabiting coral reefs.
Several features are evident from the data presented. In spite of potential differences in the fish faunas between the two regions examined and possible biases in sampling approaches, an extremely large number of fish specimens (thousands) was examined at each locality and even though the methods of examination varied to some degree, the study encompassed a wide range of fish families at both sites. Overall, 45% of the trypanorhynch species recorded here occurred in both regions. In addition, the trypanorhynch species most commonly encountered were similar in both locations. Records of adults from elasmobranchs from both of these regions provided additional information on potential life cycles and the collection included numerous new host and geographical records.

Host specificity
Notwithstanding the opportunistic nature of the collecting, several aspects of host specificity are detectable within the data set and are worthy of discussion particularly since Palm & Caira, 2008 [30] have shown that specificity of the larval stages of trypanorhynchs is generally lower than that of the adults. First, it is evident that several fish taxa were rarely infected with trypanorhynchs. Thus, despite examination of substantial numbers of Blenniidae, Chaetodontidae, Gobiidae, Kyphosidae and Scaridae, no trypanorhynchs were found in these taxa. Other taxa strikingly underrepresented, though heavily sampled, were the Acanthuridae, Pomacentridae and Echeneidae. We do not suggest that these taxa have been exhaustively examined, but certainly they are depauperate relative to families such as the Balistidae, Lethrinidae, Scombridae and Serranidae.
Among the teleost fishes that were infected, there was evidence of both stenoxenicity (parasitism of closely related species) and euryxenicity (parasitism of distantly or ecologically related species). In the stenoxenous category, Molicola horridus was seen in two species of Diodontidae, Pterobothrium australiense has been seen only in labrids (one record), Pseudogilquinia microbothria was found only in lethrinids (both in NC and the GBR) and Dasyrhynchus basipunctatus occurred overwhelmingly in tetraodontiforms (five species) although also once in a fistulariid. The apparently restricted distributions of such species are doubtless subject to refinement with further collecting but it seems highly unlikely that they will prove to be euryxenous in the same way as are some other species.
We detected some evidence of the absence of trypanorhynch species in particular fish groups. The best evidence comes from the family Serranidae which is probably the most thoroughly characterised for its trypanorhynch fauna. The serranid fishes collected tend to be large and easily examined for trypanorhynchs with which they are often heavily infected. Our results incorporate reports from 25 serranid species and of the 181 host/parasite combinations detected, 55 were from serranids; the next highest number of combinations came from the Lethrinidae with 14. The extent to which the characterisation of this family is comprehensive is demonstrated by the fact that six of the ten trypanorhynch species recorded in this family have been reported from more than one serranid species; three species were found in ten or more serranid species although four species were found in only one. We infer that the true trypanorhynch richness is thus not likely to be very much greater than the 10 species reported so far in this region. Thus, we predict that species that have been reported relatively frequently in other fishes are genuinely absent, rather than have simply not yet been collected. Most striking in this respect are the species of the Tentacularioidea. Twelve species of this superfamily are reported here in 34 host/parasite combinations, but none in serranids. The apparent absence of a range of species from the Serranidae thus appears consistent with the high host specificity seen for the species described above.
Several species showed remarkably low specificity. Thus, Callitetrarhynchus gracilis was reported here from five fish orders and 18 families, Floriceps minacanthus from two orders and six families, Pseudotobothrium dipsacum from three orders and six families, Pseudolacistorhynchus heroniensis from two orders and four families and Pseudolacistorhynchus shipleyi from three orders and five families. The absence of any detectable specificity in these species leads to the prediction that further sampling will lead to even larger host ranges for these species.
Callitetrarhynchus gracilis exhibited the widest host range and has a cosmopolitan distribution [27] with carcharhinid sharks as its primary definitive hosts in the Australian region [1]. Currently recorded in the intermediate stage from approximately 130 species of teleosts [16,27,29], 23 new host records have been added in the present study.
Floriceps minacanthus appears to be limited to the Indo-Pacific region, and again, its known definitive hosts are carcharhinid sharks [26], with adults having been reported from four species of Carcharhinus. However, the present record in Triaenodon obesus is the first from a shark not belonging to this genus. Plerocerci have been reported from 13 species of teleosts [27,29] from the Red Sea, Australia and off Indonesia and Hawaii while 14 new species of teleosts are reported here as hosts.
Pseudotobothrium dipsacum was also found in a wide variety of teleosts. It has previously been reported from numerous species of teleosts ranging from the west coast of Africa to Australia [4,27]. Eight new hosts, all from New Caledonia, have been added in the present study. In spite of its wide host range and distribution, its definitive hosts remain unknown.
Pseudolacistorhynchus heroniensis is known only from the GBR and from NC but is found in a wide range of teleosts, with 12 new teleost hosts being added in the current study. The only record of the adult parasite is a single collection from Stegostoma fasciatum from New Caledonia [6]. The specimens collected were either immature or hyperapolytic such that some doubt exists as to whether this is the usual definitive host species.
Pseudolacistorhynchus shipleyi occurs widely in the Indo-West Pacific, with the adults being found in Nebrius ferrugineus off Sri Lanka [2]. In the current study, eleven new intermediate host records are reported.
The above five species occurred in a wide variety of teleost hosts with serranids (25 species), carangids (5), balistids (5), scombrids (5) and sphyraenids (5) being most frequently encountered. The same five species of trypanorhynch were the most commonly encountered species both on the Great Barrier Reef and off New Caledonia in spite of obvious differences in the species of fish infected at the two localities. There was no intentional bias in collecting activities, but it may have been that more of these larger fishes were collected than other smaller taxa.
Other species of trypanorhynch had a more restricted host distribution. Limited data on prevalence based on a single series of collections from Lizard Island suggested that trypanorhynch larvae were prevalent in larger fishes (serranids, sphyraenids, scombrids, lutjanids) but in small fish (a single family, Apogonidae) they occurred at a very low prevalence. However, these data were based on a very small sample of fish and need to be interpreted with caution.
Overall, the patterns of host specificity seen here, a mixture of stenoxenicity and euryxenicity, resemble that reported by Chambers et al., 2000 [11] for tetraphyllidean (sensu lato) metacestodes of GBR fishes. In that study, metacestode Type 4 was found in two orders and 12 families, whereas Types 9 and 10 were found only in labrids. However, in the study of tetraphyllidean metacestodes it is often not possible to be confident that a single morphotype represents only one species whereas the complex morphology of trypanorhynch scoleces makes identification to species quite reliable.

Biogeography
Of the 33 trypanorhynch species reported here, 15 (45%) were found both in NC and on the GBR. Almost certainly this number underestimates the level of sharing between the two areas. Noticeably, the nine species reported in the largest number of host/parasite combinations were all found at both sites. Of the 21 species found in only one or two host/parasite combinations, only one (Molicola horridus) was found both in NC and on the GBR. It seems likely, or at least possible, that some species are restricted to one or other of the two sites but at present the evidence is generally marginal in this respect. The only robust parasitological study of which we are aware that has previously compared parasites of NC and the GBR is that of McNamara et al., 2012 [26] who analysed monorchiid trematodes of chaetodontids from NC and the GBR (as well as other sites in the Tropical Indo-West Pacific [TIWP]). Thirteen species of Hurleytrematoides Yamaguti, 1953 were found in total for the two sites of which just six were found at both sites for a similarity of 46%; four species were found only from the GBR and three only from NC. In every case, hosts suitable for the species not found in each area had been examined in numbers sufficient to suggest that they would have been found if present. The proportion of monorchiid species shared (46%) is thus remarkably similar to that for the trypanorhynchs. Given the much stricter specificity of monorchiids of chaetodontids (none known convincingly other than from chaetodontids) than of trypanorhynch metacestodes in general, we predict that further sampling for trypanorhynchs will see the levels of sharing increase.
Of the species found, eight (C. lintoni) with few, highly disjunct records are difficult to categorise. Nevertheless, with many of the trypanorhynch species encountered having extremely wide geographical distributions [31], it was not surprising that the species found on the GBR and from NC were broadly similar.

Localisation in host
Apart from potential differences in the species of fish present at the two sites studied, or their abundance and hence ease of obtaining a particular species, other factors may be involved such as the location of trypanorhynch metacestodes in the body of the teleost. Most are found in the body cavity and are easily recognised. However, the metacestodes of Gr. exile occur only in the gill arches of Sc. commerson [35] and this site is not always examined for the presence of metacestodes. Similarly, the metacestodes of Psg. microbothria cluster around the oesophagus of L. nebulosus (unpublished) while those of Pt. lintoni are found in the musculature (unpublished). Failure to examine sites other than the body cavities may lead to differences in the species recovered.

Life cycles
Combining the data obtained here with that available for adult trypanorhynchs in elasmobranchs in the same region has provided some insights into life cycles such as finding the adult of Pt. acanthotruncatum for the first time in Pristis zijsron. In addition, the definitive host range of F. minacanthus is expanded to include the shark Triaenodon obesus. Many life cycles remain to be identified, but broad scale collecting, such as that undertaken in this study, can be useful in identifying both potential intermediate and definitive hosts.
Species of Diodon warrant a particular mention as they are parasitised by several well-recognised trypanorhynch species including Floriceps saccatus and Molicola horridus. Infections with the latter species are particularly striking as much of the hepatic parenchyma may be replaced by metacestodes (Fig. 5). Species of Diodon are not only highly toxic [36], but can also inflate their bodies when threatened. As adults of these cestodes are found in large sharks such as Prionace glauca (Linnaeus, 1758) (see Dollfus, 1942) [14], it is tempting to assume that only large sharks are able to consume species of Diodon. Alternatively, it may be that the life cycles of these cestodes are completed using alternative intermediate hosts and their presence in species of Diodon indicates an occurrence in ''dead-end'' hosts. By comparison, in a study of the larval anisakid nematodes of teleosts off Lizard Island, Jabbar et al., 2012 [18] found no larval anisakids in their sample of tetraodontiform fishes, which would potentially be ''deadend'' hosts for these nematodes.

Conclusion
This is the first study to attempt to examine the trypanorhynch larval cestode fauna of coral reef teleosts in the west Pacific, examining reefs on the GBR and NC. The trypanorhynch fauna was dominated numerically by a small number of species at both sites with considerable similarity between the two localities examined. Although large numbers of teleosts were examined at both sites, it is most unlikely that the trypanorhynch fauna has been exhaustively surveyed and more detailed comparisons must await much more extensive sampling. Nevertheless, apart from characterising the general features of the fauna, this study has provided additional insights into host specificity and life cycles of these cestode parasites.

Conflict of Interest
The Editor-in-Chief of Parasite is one of the authors of this manuscript. COPE (Committee on Publication Ethics, http:// publicationethics.org/), to which Parasite adheres, advises special treatment in these cases. COPE wrote: ''Editors should not be denied the ability to publish in their own journal, but they must not exploit their position. The journal must have a procedure for handling submissions from the editor or members of the editorial board that ensures that peer review is handled independently of the author/editor. This process should be detailed once the paper is published.'' In this case the peerreview process was handled by Invited Editor Dominique Vuitton.