This Article Appeared in a Journal Published by Elsevier. the Attached Copy Is Furnished to the Author for Internal Non-Commerci

Home , Bluntnose stingray, Smalleye stingray, Trypanorhyncha

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy

International Journal for Parasitology 40 (2010) 223–242

Contents lists available at ScienceDirect

International Journal for Parasitology

journal homepage: www.elsevier.com/locate/ijpara

Evolution of the trypanorhynch tapeworms: Parasite phylogeny supports independent lineages of sharks and rays q

Peter D. Olson a,*, Janine N. Caira b, Kirsten Jensen c, Robin M. Overstreet d, Harry W. Palm e, Ian Beveridge f a Department of Zoology, The Natural History Museum, Cromwell Road, London SW7 5BD, UK b Department of Ecology & Evolutionary Biology, The University of Connecticut, 75 N. Eagleville Road, Storrs, CT 06269-3043, USA c Department of Ecology & Evolutionary Biology, University of Kansas, Haworth Hall, 1200 Sunnyside Ave., Lawrence, KS 66045-7534, USA d Gulf Coast Research Laboratory, The University of Southern Mississippi, 703 East Beach Drive, Ocean Springs, MS 39564, USA e Heinrich-Heine-University Düsseldorf, Institute for Zoomorphology, Cell Biology, and Parasitology, Universitätsstrasse 1, 40225 Düsseldorf, Germany f Department of Veterinary Science, University of Melbourne, Veterinary Clinical Centre, 250 Princes Highway, Werribee, Vic. 3030, Australia article info abstract

Article history: Trypanorhynch tapeworms (Platyhelminthes: Cestoda) are among the most diverse and abundant groups Received 22 May 2009 of metazoan parasites of elasmobranchs and are a ubiquitous part of the marine food webs that include Received in revised form 24 July 2009 these apex predators. Here we present a comprehensive analysis of their phylogeny, character evolution Accepted 27 July 2009 and host associations based on 10 years of sampling effort, including representatives of 12 of 15 and 44 of 66 currently recognized trypanorhynch families and genera, respectively. Using a combination of ssrDNA and lsrDNA (Domains 1–3) for 79 and 80 taxa, respectively, we maintain one-to-one correspondence Keywords: between molecules and morphology by scoring 45 characters from the same specimens used for sequenc- Trypanorhyncha ing, and provide museum vouchers for this material. Host associations are examined through likelihood- Cestoda Elasmobranchii based ancestral character state reconstructions (ACSRs) and by estimating dates of divergence using strict Neoselachii and relaxed molecular clock models in a Bayesian context. Maximum parsimony and Bayesian inference Ribosomal RNA analyses of rDNA produced well-resolved and strongly supported trees in which the trypanorhynchs Phylogeny formed two primary lineages and were monophyletic with respect to the diphyllidean outgroup taxa. Host-parasite coevolution These lineages showed marked differences in their rates of divergence which in turn resulted in differing Hypnosqualea support and stability characteristics within the lineages. Mapping of morphological characters onto the Ancestral character state reconstruction tree resulting from combined analysis of rDNA showed most traits to be highly plastic, including some previously considered of key taxonomic importance such as underlying symmetries in tentacular armature. The resulting tree was found to be congruent with the most recent morphologically based superfamily designations in the order, providing support for four proposed superfamilies, but not for the Tentacularioi- dea and Eutetrarhynchoidea. ACSRs based on the combined analysis of rDNA estimated the original hosts of the two primary parasite lineages to be alternatively rajiform batoids and carcharhiniform sharks. This fundamental split provides independent support for rejecting the notion that rays are derived sharks, and thus supports the most recent molecular phylogenies of the Neoselachii. Beyond the basal split between shark- and ray-inhabiting lineages, no pattern was found to suggest that the trypanorhynchs have closely tracked the evolutionary histories of these host lineages, but instead, it appears that host-switching has been common and that the subsequent evolution of the parasites has been ecologically driven primarily through overlap in the niches of their shark and ray hosts. Using a relaxed molecular clock model calibrated by means of host fossil data, the ray-inhabiting lineage is estimated to have diversified around the Jurassic–Cretaceous boundary, whereas the shark-inhabiting lineage is estimated to have diversified later, in the Middle Cretaceous. Although the large error associated with the estimated divergence dates prevents robust conclusions from being drawn, the dates are nevertheless found to be consistent in a relative sense with the origins of their major hosts groups. The erection and definition of the suborders Try- panobatoida and Trypanoselachoida, for the major clades of trypanorhynchs parasitizing primarily rays and sharks, respectively, is proposed for the two primary lineages recovered here. Crown Copyright Ó 2009 Published by Elsevier Ltd. All rights reserved.

q Note: Nucleotide sequence data reported in this paper are available in the GenBank, EMBL and DDBJ databases under the Accession Nos. DQ642743–DQ642804 and FJ788108–FJ788110 (28S lsrDNA) and DQ642903–DQ642966 and FJ788111–FJ788112 (18S ssrDNA). * Corresponding author. Address: Department of Zoology, The Natural History Museum, 709 Darwin Centre I, Cromwell Road, London SW7 5BD, UK. Tel.: +44 (0) 207 942 5568; fax: +44 (0) 207 942 5151. E-mail address: [email protected] (P.D. Olson). URL: http://www.olsonlab.com (P.D. Olson).

224 P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242

1. Introduction dea in terms of their species diversity in the marine realm (Caira and Reyda, 2005). The order Trypanorhyncha Diesing, 1863 (Platyhelminthes: Comprehensive taxonomic accounts of the group have been gi- Cestoda: Eucestoda) comprises one of the most ubiquitous and ven by Dollfus (1942), Campbell and Beveridge (1994) and Palm readily recognized groups of marine helminths. With four armed, (2004). Diagnosis and classification have been based chiefly on lar- retractable tentacles (Fig. 1), the often large-bodied larval stages val type and features of their rhyncheal apparatus (Fig. 2) that con- of these tapeworms are commonly observed in the flesh of marine sists of a system of hydrostatic eversion by bulbs within the scolex teleosts as well as a wide variety of invertebrate intermediate and retraction by muscles within the tentacles (cf. the simple ten- hosts, such as crustaceans which harbour the first larval stage of tacles of the diphyllobothriid tapeworm Haplobothrium; Kuchta most tapeworm groups with aquatic life cycles (Palm and Caira, et al., 2008), and may be characterized by what appears to be four 2008). They do not parasitize humans, except through accidental basic underlying symmetries in the arrangement of their armature infection (see Heinz, 1954; Kikuchi et al., 1981; Fripp and Mason, (Dollfus, 1942; Campbell and Beveridge, 1994). This unique appa- 1983), nor do they cause substantial economic loss to fisheries, ratus makes specific identification in the larval form possible. In although parasitized flesh can reduce commercial demand despite contrast, the larval forms of most other marine tapeworm groups being harmless (indeed, some indigenous peoples actually prize typically lack any species-specific characters and can only be iden- parasitized flesh for its enhanced flavour; see Overstreet, 2003). tified with certainty through genetic analysis. Similarly, whereas As adults, they are almost exclusively enteric parasites of elasmo- all other recognized tapeworm orders are defined by unique com- branchs and among tapeworms are second only to the Tetraphylli- binations of characters (see Khalil et al., 1994), the complex try-

Fig. 1. Scanning electron micrographs of scoleces of trypanorhynchs representing 11 of the 15 families. (A) Eutetrarhynchidae (Paroncomegas araya). (B) Mixodigmatidae (Halsiorhynchus macrocephalus). (C) Rhinoptericolidae (Rhinoptericola megacantha). (D) Aporhynchidae (Aporhynchus sp.). (E) Gilquinidae (Gilquinia squali). (F) Gymnorhyn- chidae (Gymnorhynchus isuri). (G) Lacistorhynchidae: Lacistorhynchiinae (Dasyrhynchus sp.). (H) Lacistorhynchidae: Grillotiinae (Hornelliella annandelei). (I) Otobothriidae (Otobothrium mugilis). (J) Pseudotobothriidae (Pseudotobothrium arii). (K) Sphyriocephalidae (Hepatoxylon sp.). (L) Tentaculariidae (Tentacularia sp.). Scale bars: A–E, I, J, 200 lm; F, G, L, 500 lm; H, 1 mm; K, 2 mm. Author's personal copy

P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 225

Fig. 2. Hook armature patterns of trypanorhynch cestodes. (A) Homeoacanthous armature in which the hooks are arranged in regular quincunxces, Nybelinia queenslandensis. (B) Typical heteroacanthous armature in which the hooks are arranged in ascending half circles on each side of the tentacle, diminishing in size towards the end of the hook row; the figure illustrates the termination of the hook rows in ascending arcs, Parachristianella baverstocki. (C) Atypical heteroacanthous armature in which the hooks are arranged in ascending half circles on each side of the tentacle, as in typical heteroacanths, but there is an interpolated row of accessory or intercalary hooks (in black) on the external surface of the tentacle where the hook rows terminate, Pseudolacistorhynchus heroniensis; (D) Poeciloacanthous armature: definitions of this type of armature vary, but in the case illustrated, the principal rows of hooks (open) terminate short of the midline of the external surface of the tentacle, while in the midline is a single file of hooks termed a chainette; in this example, there are multiple rows of intercalary hooks compared with the single row shown in C, Dasyrhynchus giganteus. panorhynch rhyncheal apparatus is considered a synapomorphy over, where monophyly has not been supported, the two primary that underpins the naturalness of the group. The fact that prelimin- lineages indicated by molecular analyses tend to be placed either ary molecular data have supported the existence of two indepen- in immediate succession to one another (e.g. Olson et al., 2001) dent trypanorhynch lineages (i.e. paraphyly; see Olson and or are grouped in a clade together with their most likely sister Tkach, 2005 for a review) is likely to be a result of the disparate group (see below), the Diphyllidea (e.g. Waeschenbach et al., rates of evolution between these lineages (i.e. among-lineage rate 2007; Olson et al., 2008). For these reasons, here we assume mono- variation) rather than the non-natural nature of the order. More- phyly of the trypanorhynchs, and focus instead on resolving its Author's personal copy

226 P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 internal structure and examining the historical associations formes) and sawsharks (Pristiophoriformes) leading in succession formed with their hosts. to true skates and rays (Rajiformes) (Fig. 3). Soon after, molecu- Morphologically-based views of trypanorhynch evolution con- lar-based phylogenies began to appear that supported the basal tinue to differ with respect to the traits considered and the relative split of the Neoselachii, once again reverting to the earlier hypoth- importance given to them. This has resulted in conflicting ideas on esis of sharks and rays as independent, reciprocally monophyletic the plasticity of various characters and host associations, as well as lineages, these lineages often referred to as the Selachimorpha competing classifications and ideas as to their evolution. The pre- and the Batoidea, respectively. Now supported by a variety of mito- liminary cladistic analysis by Beveridge et al. (1999), for example, chondrial and nuclear genes, molecular systematic studies (Fig. 3) resulted in clades that largely reflected the nature of their tentac- have been unanimous in their rejection of the Hypnosqualea, ular armature and differed significantly in its systematic implica- implying that dorsal–ventral flattening has occurred indepen- tions from the arguments of Palm (1997, 2004) who considered dently in both the shark and ray lineages (Douady et al., 2003; the armature of less importance for establishing higher taxa. Until Winchell et al., 2004; Naylor et al., 2005; Human et al., 2006). Here recently, molecular data from trypanorhynchs had been applied we show that a basal split in the parasite phylogeny similarly leads only to the two extremes of the taxonomic spectrum: to represent to independent lineages whose ancestral hosts were alternatively the group using a small number of exemplar taxa in higher (i.e. sharks or rays, providing to our knowledge the first independent ordinal) level analyses of the Cestoda (e.g. Olson and Caira, 1999; evidence for refuting the hypothesis that batoids are derived Olson et al., 2001) and to examine levels of divergence within a sharks. species (e.g. Palm et al., 2007). Here we provide a comprehensive analysis of trypanorhynch interrelationships based on a combination of ssrDNA and lsrDNA data which have proven to be informa- 2. Materials and methods tive for estimating interrelationships within other tapeworm groups, e.g. Proteocephalidea (de Chambrier et al., 2004) and Rhin- 2.1. Specimen collection, fixation and taxonomic representation ebothriidea (Healy et al., 2009); see also Olson and Tkach (2005). A majority of the sequence data collected for the present study Trypanorhynchs were obtained from hosts collected from vari- were made publicly available in 2007 and were used recently by ous parts of the globe over a period of more than 10 years using the Palm et al. (2009), together with additional taxa representing the following methods: off Heron Island, Queensland, Australia using Indo-Pacific region, in order to address specific family and genus- hand spears or hook and line; in the Gulf of Mexico using hook level questions. Their work focused on the implications of these and line; off Malaysian Borneo by artesanal or commercial trawling data with respect to the classification system and ideas of try- vessels, long-lines, or in the cases of smaller stingrays, hand spears; panorhynch evolution proposed by Palm (2004) and discussed off Victoria, Australia from commercial trawling vessels; off the their morphological evolution by way of annotating 13 putatively Wessel Islands, Australia using the commercial trawling vessel synapomorphic characters taken from the literature onto their FV Ocean Harvest; off coastal Argentina; off the Azores with com- phylogenetic estimate. Here we present a comprehensive suite of mercial trawling vessels in conjunction with the Department of analyses in which character evolution is inferred by way of explicit Fisheries, University of Azores, off the island of Faial; in the North coding of 45 traits observed, interpreted and scored from the actual Sea via trawling in conjunction with the Scottish Marine Labora- specimens sequenced, and employ parsimony to arbitrate charac- tory (University of Aberdeen, Scotland); off New Caledonia and ter state changes across our molecular-based topology. In this French Polynesia in conjunction with the Institut de Recherche way we maintained one-to-one correspondence between informa- pour le Développement, Nouméa using trawling vessels; off Mon- tion from molecules and morphology and thus avoid the need to tauk, Long Island (USA) by sports fishermen using hook and line; make assumptions or generalizations with regard to character in Florida Bay (USA) using hook and line; off coastal Senegal in con- state consistency within supra-specific taxa (i.e. genera, families, junction with artesanal fishermen using stationary nets. Full de- etc.). Full collection data for the 66 trypanorhynch species in com- tails of the trypanorhynch species, hosts and collection mon between the two studies are also presented here for the first information are provided in Table 1. time. We formally revise the classification of trypanorhynchs with To enable complete character coding we strove to collect spec- the aim of moving away from the retention of paraphyletic taxa imens as adults, rather than their larval stages: 84% of the species (e.g. Eutetrarhynchoidea of Palm et al., 2009) to a system in which are represented by adult worms collected from their elasmobranch all major groups are monophyletic. We then present an in-depth definitive hosts, with the remainder being plerocercus or plerocer- analysis of their host associations through statistically-based coid stages collected primarily from teleost intermediate hosts. ancestral host reconstructions and estimates of divergence times Specimens were fixed and stored in 95–100% ethanol at 4 °C; in with respect to the geologic time scale and discuss the implications the most cases, representative specimens were also fixed in 10% of these results in light of elasmobranch evolution (see below). In seawater buffered formalin to facilitate morphological analysis. addition, we take the opportunity to re-evaluate and contrast some These specimens were prepared as whole mounts according to of the conclusions drawn by Palm et al. (2009). conventional techniques. Given that 12 of the 15 additional species A long standing question in elasmobranch evolution has con- included by Palm et al. (2009) were represented by larval stages, cerned the monophyly of sharks relative to rays, with the early, and only one of the three additional species represented by adult pre-cladistic idea consisting of a basal split of the Neoselachii into forms was vouchered, no attempt was made to expand our sample sharks and rays (e.g. Bigelow and Schroeder, 1948, 1953; Compagno, to include species beyond those for which substantial new material 1973). This notion was later challenged by a cladistic-based was available. However, our study included three species and one hypothesis of Shirai (1992, 1996) who united the shark groups Chl- genus beyond those analyzed by Palm et al. (2009). amydoselachiformes, Hexanchiformes, Squaliformes, Squatinifor- Collections herein included representatives of all five currently mes and Pristiophoriformes, and the true rays, or ‘batoids’; the recognized superfamilies (Eutetrarhynchoidea, Gymnorhynchoi- latter four dorsal–ventrally flattened groups being collectively dea, Lacistorhynchoidea, Otobothrioidea, and Tentacularioidea), termed ‘hypnosqualeans’ (Shirai, 1996). Based on this hypothesis, 12 of 15 families and 40 of 66 genera (plus an additional five re- rays were considered to have resulted from an evolutionary trajec- cently described genera i.e. Fossobothrium, Iobothrium, Oncomego- tory towards increasingly dorso-ventrally flattened bauplans, with ides, Sagittirhynchus and Vittirhynchus (see Beveridge and hypothesized intermediate forms such as angel sharks (Squatini- Campbell, 2005; Beveridge and Justine, 2006)) not recorded in Author's personal copy

P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 227

mod. from Shirai, 1992 & 1996 (morphology) mod. from Douady et al., 2003 (12S + 16S rRNA) Hydrolagus Raja BATOIDS hexanchiforms Urobatis Heterodontus Ginglymostoma most squaliforms Alopias Squalus/Cirrhigaleus Carcharias Carcharodon Squatiniformes Lamna Isurus Pristiophoriformes Hypnosqualea Isurus BATOIDS Scyliorhinus Carcharhinus SHARKS Heterodontiformes Mustelus Orectolobiformes Hexanchus Galea Heptranchias Carcharhiniformes Squatina Pristiophorus Lamniformes Centrophorus Squalus

Winchell et al., 2004 (18S + 28S rRNA) mod. from Naylor et al., 2005 (RAG-1) mod. from Human et al., 2006 (Cyt-B)

Acipenser Chimaera Chimaera Gymnura Latimeria Raja Gymnura Dasyatis Hydrolagus Pristis Myliobatis BATOIDS Urobatis BATOIDS Callorhinchus Aetobatus Amblyraja Potamotrygon Pristis Dasyatis Rhinobatos Urobatis Chlamydoselachus Chlamydoselachus Pristiophorus Rhinobatos BATOIDS Pliotrema Heptranchias Squatina Raja Squalus Heterodontus Centroscymnus Oxynotus Isurus Isurus Centroscyllium Squalus Carcharodon Lamna Deania Squatina Lamna Cetorhinus Dalatias Pristiophorus Carcharias Alopias Squalus Heterodontus SHARKS Alopias Squatina Alopias Orectolobus Odontaspis Pristiophorus Megachasma Stegostoma Pseudocarcharias Chlamydoselachus Mitsukurina Carcharhinus Mustelus Heterodontus Mustelus SHARKS Apristurus Mustelus Triakis SHARKS Mustelus Mitsukurina Galeocerdo Galeocerdo Prionace Carcharodon Apristurus Odontaspis Sphyrna Sphyrna Scyliorhinus Sphyrna Negaprion Carcharias Haploblepharus Alopias Haploblepharus Haploblepharus Mitsukurina Halaelurus Holohalaelurus Hemiscyllium Apristurus Galeus Orectolobus Scyliorhinus Scyliorhinus Poroderma Poroderma

Fig. 3. The cladistic, morphologically-based hypothesis of Shirai (1996) contrasted with recent molecular-based estimates of elasmobranch phylogeny.

Palm (2004). Thus the representation is nearly comprehensive at lected from the same host individuals; ‘paragenophores’ sensu the level of family and includes over 60% of known genera, of Pleijel et al., 2008) were also deposited when available as members which 40% are represented herein by multiple species. In a few of a series. Accession numbers and museums in which vouchers cases, the same species collected from different hosts were in- have been deposited are given in Table 1. cluded to test the conspeciﬁcity of the parasites (i.e. Callitetrarhyn- chus gracilis, Oncomegoides celatus and Prochristianella sp.). 2.3. Molecular phylogenetic analyses

2.2. Specimen voucher deposition 2.3.1. Outgroup choice The precise phylogenetic position of the Trypanorhyncha within To anchor sequence data to actual specimens and enhance tax- the Eucestoda, and thus the identities of its nearest relatives, re- onomic collections, stained, whole-mounted vouchers were depos- main problematic. From a traditional viewpoint (see review in Ho- ited in well-curated, publicly available collections for the majority berg et al., 1997), the group most often linked to trypanorhynchs is of the newly collected material. Where possible, molecular vouch- the Diphyllidea, members of which are also restricted to elasmo- ers represent the actual specimens from which tissue was obtained branchs (primarily rays) and are similarly ‘bothriate’ (i.e. possess for the extraction of genomic DNA (gDNA) (i.e. ‘hologenophores’ a scolex consisting of weakly muscular bothria). In molecular phy- sensu Pleijel et al., 2008), and generally comprise the anterior logenetic analyses of the Cestoda, these two orders are also gener- scolex plus posterior portion of the strobila. Additional whole- ally found in close proximity, although poor nodal support and a mounted specimens representing the same collections (i.e. col- lack of resolution among the bothriate and monofossate tapeworm Author's personal copy

228 P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242

Table 1 Taxonomic and alphabetic listing of species.

Classificationa Taxon ex Host species (common nameb), Host accession numberc, collection locality [specimen voucher accession number(s)] 18S ssrDNA/28S lsrDNA sequence accession numbersd (unique specimen code) Platyhelminthes:Neodermata:Cestoda:Eucestoda DIPHYLLIDEA Van Beneden in Carus (1863) DITRACHYBOTHRIDIIDAE Schmidt (1970) Ditrachybothridium macrocephalume ex Apristurus laurussonii (Iceland catshark), Goban Spur off Ireland, Northeastern Atlantic [BMNH 2004.1.6.1–5] DQ642903/ AY584864 (Dibm2) ECHINOBOTHRIIDAE Perrier, 1897 Echinobothrium chisholmae ex Glaucostegus typus (Giant shovelnose ray), Heron Island, Queensland, Australia [BMNH 2000.8.3.4–7] AF286986/AF286922 (Eho) Echinobothrium harfordi ex Leucoraja naevus (Cuckoo ray), North Sea, UK [BMNH 2000.8.3.4–7] AF286986/AF286922 (Ehar) Echinobothrium sp. ex Raja sp. (skate), Northeastern Atlantic off Ireland [BMNH 2003.3.6.23–28] DQ642904/AY584862 (Echb) MACROBOTHRIDIIDAE Khalil and Abdul-Salam (1989) Macrobothridium rhynchobati ex Glaucostegus typus (Giant shovelnose ray), Shoal Bay, Northern Territory, Australia [BMNH 2004.3.18.101] AF124463/AY584861 (Mac) TRYPANORHYNCHA Diesing (1863) EUTETRARHYNCHOIDEA Guiart (1927) EUTETRARHYNCHIDAE Guiart (1927) Dollfusiella tenuispinis ex Dasyatis sabina (Atlantic stingray), Gulf of Mexico, Mississippi, USA [BMNH 2008.5.21.2] DQ642958/DQ642796 (DNA-00–88) Dollfusiella sp. ex Himantura fava (honeycomb whipray), BO-24, South China Sea off Sematan, Sarawak, Malaysia [LRP 4269] DQ642959/DQ642797 (Dolb) Dollfusiella sp. ex Himantura pastinacoides (Round whip ray), BO-76, Sulu Sea off Kampung Tetabuan, Sabah, Malaysia [LRP 4270] DQ642962/DQ642800 (Dolb 3) Dollfusiella geraschmidti ex Urolophus paucimaculatus (Sparsely-spotted stingaree), Queenscliffe, Victoria, Australia [BMNH 2001.1.25.6–7] DQ642955/DQ642793 (Dfsp) Dollfusiella martini ex Trygonorrhina fasciata (Southern fiddler), Queenscliffe, Victoria, Australia [BMNH 2001.1.25.2–4] DQ642964/DQ642802 (Dolm) Dollfusiella michiae ex Rhina ancylostoma (Bowmouth guitarfish), NT-103, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3683] DQ642966/ DQ642804 (Dmic1) Dollfusiella ocallaghani ex Himantura jenkinsii (Pointed-nose stingray), NT-33, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3661] DQ642961/DQ642799 (Doca1) Dollfusiella spinulifera ex Glaucostegus typus (Giant shovelnose ray), Coral Sea, Heron Island, Queensland, Australia [BMNH 1999.9.16.1–2; QM G217372] DQ642965/DQ642803 (D1-D6) (Doll) Mecistobothrium brevispine ex Rhinoptera bonasus (Cownose ray), MS05–49, off Ship Island, Gulf of Mexico, Mississippi, USA [LRP 4278] (no 18S)/FJ788110 (MS05–49-10) Mecistobothrium johnstonei ex Pastinachus cf. sephen (cowtail stingray), NT-44, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3667] DQ642936/DQ642774 (NT44E) Oncomegas australiensis ex Aetobatus cf. narinari (spotted eagle ray), NT-76, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3678] DQ642957/DQ642795 (Onco1) Oncomegas sp. ex Aetobatus cf. narinari (spotted eagle ray), NT-76, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3679] DQ642956/ DQ642794 (Omeg1) Oncomegoides celatus ex Dasyatis microps (Smalleye stingray), NT-108, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3698] DQ642934/ DQ642772 (NT108A) Oncomegoides celatus ex Himantura jenkinsii (Pointed-nose stingray), NT-33, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3662] DQ642935/DQ642773 (NT33D) Parachristianella baverstocki ex Pastinachus cf. sephen (cowtail stingray), NT-44, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3669] DQ642937/DQ642775 (Pbav1) Parachristianella indonesiensis ex Rhynchobatus cf. australiae (wedgefish), NT-49, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3673] DQ642939/DQ642777 (Pbav5) Parachristianella monomegacantha ex Himantura draco (Dragon stingray), NT-106, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3694] DQ642943/DQ642781 (Pmon2) Paroncomegas araya ex Potamotrygon motoro (Ocellate river stingray), Rio coastline, Santa Fe, Argentina [BMNH 2003.3.31.23; 2004.3.18.100] DQ642963/ DQ642801 (Pmeg) Prochristianella clarkeae ex Rhynchobatus cf. australiae (wedgefish), NT-7, Gove Harbor off Nhulunbuy-Gove, Gulf of Carpentaria, Northern Territory, Australia [BMNH 2003.3.31.23] DQ642947/DQ642785 (Pcla1) Prochristianella hispida ex Dasyatis say (Bluntnose stingray), Gulf of Mexico, Mississippi, USA DQ642946/DQ642784 (DNA-01–97) Prochristianella macracantha ex Taeniura lymma (Bluespotted ribbontail ray), BO-80, Celebes Sea off Kunak, Sabah, Malaysia [LRP 4271] DQ642932/DQ642770 (Proch2) Prochristianella sp. ex Dasyatis sabina (Atlantic stingray), Gulf of Mexico, Mississippi, USA [BMNH 2008.5.21.4] DQ642945/ DQ642783 (DNA-00–90) Prochristianella sp. ex Himantura pastinacoides (Round whip ray), BO-76, Sulu Sea off Kampung Tetabuan, Sabah, Malaysia [LPR 4272] DQ642933/DQ642771 (Proch) Prochristianella sp. ex Sphyrna tiburo (Bonnethead), M-00–2502, Gulf of Mexico, Mississippi, USA DQ642931/ DQ642769 (DNA-00–16) Prochristianella sp. ex Dasyatis sabina (Atlantic stingray), Gulf of Mexico, Mississippi, USA DQ642930/DQ642768 (DNA-01–122A) Prochristianella sp. ex Dasyatis sabina (Atlantic stingray), Gulf of Mexico, Mississippi, USA DQ642944/ DQ642782 (OVR-36b) Tetrarhynchobothrium sp. ex Carcharhinus melanopterus (Blacktip reef shark), Coral Sea, Heron Island, Queensland, Australia [BMNH 2001.1.26.1] AF287002/ AF286965 (Ttm) Tetrarhynchobothrium sp. ex Himantura cf. gerrardi (sharpnose stingray), BO-23, South China Sea off Sematan, Sarawak, Malaysia [LRP 4273] DQ642960/ DQ642798 (Trhy) Trimacracanthus aetobatidis ex Trygonorrhina fasciata (Southern fiddler), Queenscliffe, Victoria, Australia [BMNH 2001.1.25.1] DQ642942/DQ642780 (Tria) MIXODIGMATIDAE Dailey & Vogelbein, 1982 Halysiorhynchus macrocephalus ex Pastinachus cf. sephen (cowtail stingray), NT-44, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3664] DQ642940/DQ642778 (Haly2) Trygonicola macropora ex Himantura cf. gerrardi (sharpnose stingray), BO-23, South China Sea off Sematan, Sarawak, East Malaysia [LRP 4274] DQ642941/ DQ642779 (Tryg) RHINOPTERICOLIDAE Carvajal & Campbell, 1975 Rhinoptericola megacantha ex Rhinoptera bonasus (Cownose ray), east of Ship Island, Gulf of Mexico, Mississippi, USA [BMNH 2008.5.21.1] DQ642954/DQ642792 (DNA-01–100L) Shirleyrhynchus aetobatidis ex Himantura cf. uarnak (honeycomb stingray), BO-82, Celebes Sea off Semporna, Sabah, Malaysia [LRP 4275] DQ642938/DQ642776 (Shir) Author's personal copy

P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 229

GYMNORHYNCHOIDEA Dollfus, 1935 APORHYNCHIDAE Poche, 1926 Aporhynchus sp. ex Etmopterus spinax (Velvet belly lantern shark), AZ-8, Horta, Faial, Azores, Portugal [LRP 4279] FJ572911/FJ572947 (Apor/TE-157) GILQUINIIDAE Dollfus (1942) Gilquinia squali ex Squalus acanthias (Piked dogfish), North Sea, UK AJ287516/AF286966 (Gsq) Sagittirhynchus aculeatus ex Centrophorus sp. (gulper shark), L’ile des Pins, New Caledonia [MNHN JN70A] DQ642907/DQ642745 (Gng1) Vittirhynchus squali ex Squalus melanurus (Blacktailed spurdog), Coetlogon, New Caledonia [MNHN JN73A] DQ642905/DQ642743 (Gng2) GYMNORHYNCHIDAE Dollfus (1935) Chimaerarhynchus rougetae ex Squalus cf. megalops (spurdog), Baie de Santal, Lifou, New Caledonia [MNHN JN4A] DQ642906/DQ642744 (Chir) Gymnorhynchus isuri ex Isurus oxyrinchus (Shortfin mako), Montauk, Long Island, New York, USA [LRP 3711] DQ642909/DQ642747 (Gymn) Molicola uncinatuse ex Thyrsites atun (Snoek), Apollo Bay, Victoria, Australia [BMNH 2004.3.18.102] DQ642908/DQ642746 (Moli) LACISTORHYNCHOIDEA Guiart (1927) LACISTORHYNCHIDAE Guiart (1927) GRILLOTIINAE Dollfus (1942) Dasyrhynchus giganteuse ex Caranx hippos (Crevalle jack), 7 nautical miles south of East Ship Island, Gulf of Mexico, Mississippi, USA [LRP 4283] FJ788112/ FJ788109 (MS05–48-4) Grillotia erinaceus ex Amblyraja radiata (Throny skate), North Sea, south of Fair Isle, Scotland AJ228781/AF286967 (Geri) Grillotia pristiophori ex Pristiophorus nudipinnis (Shortnose sawshark), San Remo, Victoria, Australia [SAMA 28386] DQ642925/DQ642763 (GripA) Grillotia roweie ex Coryphaenoides armatus (Abyssal grenadier), Porcupine Sea Bight, Northern Atlantic Ocean [BMNH 2003.3.7.1–15] DQ642927/DQ642765 (1393536) Paragrillotia similis ex Ginglymostoma cirratum (Nurse shark), GC-1, Florida Bay, Florida, USA [LRP4280] FJ572918/FJ572954 (TE-51) Pseudogilquinia microbothria ex Sphyrna mokarran (Great hammerhead), NT-112, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3706] DQ642928/DQ642766 (Dmag1) Pseudogilquinia pillersie ex Lethrinus atkinsoni (Pacific yellowtail emperor), Heron Island, Queensland, Australia [BMNH 2004.3.18.98–99] AJ287496/AF286964 (Das) Protogrillotia sp. ex Carcharhinus amboinensis (Pigeye shark), NT-113, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [SAMA 28646] DQ642929/ DQ642767 (NT113A) LACISTORHYNCHIINAE Callitetrarhynchus gracilis ex Carcharhinus amboinensis (Pigeye shark), NT-113, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3709] DQ642920/DQ642758 (Calg1) Callitetrarhynchus gracilis ex Carcharhinus melanopterus (Blacktip reef shark), Heron Island, Queensland, Australia [QM G 217612] AJ287487/AF286970 (Ctg) Callitetrarhynchus gracilise ex Lethrinus nebulosus (Spangled emperor), Ningaloo, Western Australia, Australia [BMNH 2008.5.21.3] DQ642921/DQ642759 (Calls) Diesingium lomentaceum ex Mustelus mustelus (Smooth-hound), SE-99, Eastern Atlantic Ocean off Soumbédioune, Senegal [LRP 3713] DQ642922/DQ642760 (Dies) Floriceps minacanthuse ex Euthynnus affinis (Kawakawa) Heron Island, Queensland, Australia [QM G 217612] AF287003/AF286971 (Flo) Floriceps saccatuse ex Diodon hystrix (Spot-fin porcupinefish), Pointe Hauru, Moorea, French Polynesia [BMNH-2001.1.31.1–2] DQ642919/DQ642757 (Flos) Hornelliella annandalei ex Stegastoma fasciatum (Zebra shark), NT-109, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3703] DQ642924/ DQ642762 (Horn1) Lacistorhynchus dollfusi ex Mustelus antarcticus (Gummy shark), Apollo Bay, Victoria, Australia [BMNH-2001.1.23.8–13] DQ642923/DQ642761 (Ldol) Pseudolacistorhynchus heroniensise ex Plectropomus leopardus (Leopard coralgrouper), Heron Island, Queensland, Australia [QM G217515–7] AJ287519/AF286968 (Ghe) PTEROBOTHRIIDAE Diesing, 1850 Pterobothrium lintonie ex Choerodon venustus (Venus tuskfish), Heron Island, Queensland, Australia [SAMA V4080] AF287004/AF286973 (Plin) Pterobothrium platycephalum ex Himantura sp. (whipray), NT-96, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3680] DQ642926/ DQ642764 (Ptpl1) OTOBOTHRIOIDEA Dollfus (1942) OTOBOTHRIIDAE Dollfus (1942) Fossobothrium perplexum ex Anoxypristis cuspidata (Knifetooth sawfish), NT-65, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3714] DQ642914/DQ642752 (Onsp1A) Iobothrium elegans ex Himantura jenkinsii (Pointed-nose stingray), NT-33, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [SAMA 28634] DQ642916/DQ642754 (NT33A) Otobothrium mugilis ex Sphyrna mokarran (Great hammerhead), NT-99, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 4276] DQ642912/ DQ642750 (NT99A) Otobothrium propecysticum ex Sphyrna mokarran (Great hammerhead), NT-112, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [SAM 29101] DQ642913/DQ642751 (Otsp) Otobothrium carcharidis ex Carcharhinus dussumieri (Whitecheek shark), NT-53, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3675] DQ642911/DQ642749 (Ocar1) Poecilancistrium caryophyllume ex Bairdiella chrysoura (Silver perch), 8 miles south of north end of Chandeleur Light, Gulf of Mexico, Mississippi, USA [LRP 4284] FJ788111/FJ788108 (MS05–71-1) Proemotobothrium linstowi ex Rhynchobatus cf. australiae (wedgefish), NT-49, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 3672] DQ642917/DQ642755 (Prol) Proemotobothrium sp. ex Himantura jenkinsii (Pointed-nose stingray), NT-33, east of Wessel Islands, Arafura Sea, Northern Territory, Australia [LRP 4282] DQ642915/DQ642753 (Proe) PSEUDOTOBOTHRIIDAE Palm, 1995 Parotobothrium ballie ex Chaetodon lunula (Raccoon butterflyfish), Ningaloo, Western Australia, Australia [BMNH 2008.5.21.5] DQ642918/DQ642756 (Paran) Pseudotobothrium arii ex Lamiopsis temminckii (Broadfin shark), BO-74, South China Sea off Mukah, Sarawak, Malaysia [LRP 4277] DQ642910/DQ642748 (Poto) Pseudotobothrium dipsacume ex Choerodon venustus (Venus tuskfish), Heron Island, Queensland, Australia AJ287552/AF286972 (Otb3) TENTACULARIOIDEA Poche (1926) SPHYRIOCEPHALIDAE Pintner (1913) Hepatoxylon trichiurie ex Prionace glauca (Blue shark), Montauk, New York, USA AF124462/AF286969 (Hep) Sphyriocephalus viridis ex Dalatias licha (Kitefin shark), Goban Spur, off southwest Ireland [BMNH-2000.1.18.4] AJ287576/AF286974 (Sph) TENTACULARIIDAE Poche, 1926 Heteronybelinia estigmena ex Carcharhinus limbatus (Blacktip shark), BO-58, South China Sea off Mukah, Sarawak, Malaysia DQ642951/DQ642789 (Nyb3) Kotorella pronosoma ex Dasyatis say (Bluntnose stingray), Gulf of Mexico, Mississippi, USA DQ642950/DQ642788 (OVR-08) Kotorella sp. ex Taeniura lymma (Bluespotted ribbontail ray), BO-87, Celebes Sea off Semporna, Sabah, Malaysia DQ642949/DQ642787 (Nyb8) Nybelinia aequidentata ex Rhinoptera cf. neglecta (cownose ray), BO-83, Celebes Sea off Semporna, Sabah, Malaysia DQ642952/DQ642790 (Nyb7)

(continued on next page) Author's personal copy

230 P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242

Nybelinia africana ex Lamiopsis temminckii (Broadﬁn shark), BO-74, South China Sea off Mukah, Sarawak, Malaysia DQ642948/DQ642786 (Nyb5) Nybelinia sphyrnae ex Sphyrna lewini (Scalloped hammerhead), BO-69, South China Sea off Mukah, Sarawak, Malaysia DQ642953/DQ642791 (Nyb4) Nybelinia queenslandensis ex Carcharhinus melanopterus (Blacktip reef shark), Heron Island, Queensland, Australia [QM G217521–31, G218030; BMNH 2004.3.18.103–104] AF287005/AF286975 (Nyq) Tentacularia coryphaenaee ex Prionace glauca (Blue shark), Montauk, New York, USA AF124461/AF286976 (Ten)

Abbreviations (museums in bold): BMNH, The Natural History Museum, London, England; GCRL, Gulf Coast Research Laboratory, Ocean Springs, Mississippi, USA; LRP, Lawrence R. Penner Parasitology Collection, University of Connecticut, Connecticut, USA; MNHN, Collection du Muséum National d’Histoire Naturelle, Paris, France; QM, Queensland Museum, Brisbane, Queensland, Australia; SAMA, South Australian Museum Helminthological Collection, Adelaide, South Australia, Australia. a Diphyllidea classification follows Khalil et al. (1994); Trypanorhyncha classification follows Palm (2004). b Common names of hosts follow FishBase (Froese, R., Pauly, D., 2004. FishBase: world wide web electronic publication. www.fishbase.org, version Dec, 2008). c Where listed, additional host and collection information can be found via the world wide web through the Host Specimen Database link at www.tapeworms.org. d Accession numbers beginning ‘DQ’ were made publicly available by us in 2007. e Larval specimen. orders has left their relative positions open to question (for a re- alignment of the two genes was constructed initially using MUS- view see Olson and Tkach, 2005). Although it might appear desir- CLE (Edgar, 2004) with default parameters and subsequently able to include a large number of candidate sister lineages of the examined by eye using MacClade (Maddison, D.R., Maddison, Trypanorhyncha as outgroup taxa, sequence divergence among W.P., 2005. MacClade 4: Analysis of phylogeny and character evo- such groups would require the removal of considerably more posi- lution, version 4.08. Sinauer, Sunderland, MA, USA). Regions of tions due to the lack of recognizable homology (see Appendix 1 in ambiguous alignment (e.g. regions with considerable length varia- Olson and Caira, 1999). Moreover, previous rDNA-based studies tion among the sequences) were defined in a NEXUS character have already demonstrated rDNA data to be insufficient in strongly exclusion set. Regions containing gaps in a majority of taxa were arbitrating on this question, whilst nevertheless supporting a close also excluded from analyses even if such regions were alignable association (often as sister group) between the Diphyllidea and the among a minority of taxa possessing insertions. The resulting 18S Trypanorhyncha (e.g. Waeschenbach et al., 2007; Olson et al., partition comprised 2186 positions of which 400 positions were 2008). For these reasons, only members of the Diphyllidea were in- excluded (18%) and 338 (19%) were parsimony informative. The cluded as outgroups. 28S partition comprised 1651 positions of which 509 were excluded (31%) and 559 (49%) were parsimony informative. The con- 2.3.2. DNA amplification and sequencing catenated alignment including the character exclusion set is The complete 18S ssrRNA (minus the distal 50 and 30 priming re- available for download in the form of an executable NEXUS file gions) and partial (Domains 1–3) 28S lsrRNA genes were character- from www.olsonlab.com. ized for all trypanorhynch taxa and for diphyllidean outgroup taxa where such sequences were not already available from GenBank 2.3.4. Phylogenetic analyses based on the works of Olson and Caira (1999), Olson et al. (2001) The 18S, 28S and combined data partitions were analyzed by and Bray and Olson (2004). In total, 63 new 18S and 65 new 28S se- the methods of maximum parsimony and Bayesian inference. Max- quences were characterized. Tissue samples from ethanol-pre- imum parsimony analyses were conducted with PAUP* (Swofford, served specimens were soaked overnight in Tris–EDTA buffer and D.L., 2003. PAUP*, Phylogenetic Analysis Using Parsimony (*and Ò TM the total gDNA extracted using the Qiagen DNeasy tissue kit fol- Other Methods). Version 4.0b10. Sinauer, Sunderland, MA, USA) lowing manufacturer-recommended protocols, with the exceptions using a heuristic search strategy with 1000 search replicates, ran- that the incubation period with proteinase-K was extended to over- dom-addition taxon sampling and tree-bisection reconnection night in a rotating incubator and the final elution volume was re- branch-swapping, with all characters run unordered with equal duced to 200 ll. One to three microliters gDNA were used as a weights and gaps treated as missing data. Nodal support was as- TM template in 25 ll reactions using Ready-To-Go (Amersham Phar- sessed by bootstrap analysis based on 1000 replicates with three macia Biotech) PCR beads and the following thermocycling profile: full heuristic searches/replicate. Bayesian inference was conducted 3 min denaturation at 94 °C; 40 cycles of 30 s at 94 °C, 30 s at 52 °C, using MrBayes (Ronquist and Huelsenbeck, 2003; ver. 3.1.2). Mod- 2 min at 72 °C; and 7 min extension at 72 °C. The complete 18S gene els of nucleotide substitution were evaluated for both data parti- was amplified using primers Worm-A and Worm-B (see Littlewood tions independently using MrModelTest (Nylander, J.A.A., 2004. and Olson, 2001 for a complete list of 18S primer definitions) and MrModelTest, program distributed by the author. Evolutionary the D1–D3 domains of the 28S amplified using primers LSU-5 and Biology, Upsala University, Sweden, version 2), and for both the 1200R (see Olson et al., 2003 for 28S primer definitions). PCR ampli- most parameter-rich model (i.e. general-time-reversible including cons were either gel-excised or purified directly using Qiagen Qia- estimates of invariant sites and gamma distributed among-site rate TM quick columns, cycle-sequenced from both strands using ABI variation) was found to provide the best fit to the data based on the TM BigDye chemistry, alcohol-precipitated, and run on an ABI Prism Akaike information criterion. The following model parameters TM 377 automated sequencer. The products were sequenced bidirec- were thus used in MrBayes: nst = 6, rates = invgamma, ngamma- tionally using the PCR primers and a variety of internal primers cat = 4. The metropolis-coupled Markov chain Monte Carlo param- listed in the publications of Littlewood and Olson (2001) and Olson eters used were ngen = 2,000,000, nchains = 4, and samplefreq = 100 et al. (2003). Contiguous sequences were assembled and edited with default prior probabilities. Two separate runs were performed TM using Sequencher (GeneCodes Corp., ver. 3.1.1), screened for pos- (default behaviour in MrBayes 3.1.2) and convergence between sible contamination using BLASTn (McGinnis and Madden, 2004) them examined using the sump command. Consensus trees with and submitted to GenBank under Accession Nos. DQ642743– mean branch lengths were constructed using the sumt command DQ642804 and FJ788108–FJ788110 (28S) and DQ642903–DQ642966 with the contype = allcompat option and ignoring the topologies and FJ788111–FJ788112 (18S). saved during the initial n-generations before log-likelihood values and substitution parameters reached stationarity (i.e. ‘burn-in’). 2.3.3. Sequence alignment and data partitions Burn-in was estimated by plotting the log probabilities against The molecular data sets consisted of 79 trypanorhynch (80 for generation and visualizing the plateau in parameter values. Burn- 28S which included a second species of Mecistobothrium; i.e. Mec- in values were 2000 for 28S and 800 for both the 18S and com- istobothrium brevispine) and five diphyllidean taxa. A concatenated bined data sets (representing the first 200,000 and 80,000 of a total Author's personal copy

P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 231 of 2,000,000 generations in each analysis). Nodal support was esti- taxon sampling used here. A listing of the characters and their mated as the mean posterior probabilities (Huelsenbeck et al., states is provided in Table 2. A matrix of morphological characters 2001) as calculated by the sumt command. (Supplementary Data S1) was constructed based, with a few exceptions, on observed features of the specimens used for sequencing 2.3.5. Stability analyses (i.e. holo- or paragenophores). In cases in which specimens were In order to identify relatively unstable taxa whose inclusion can damaged, incomplete or unavailable, codings were based on recent diminish support across the tree (see Wilkinson, 2003), we ranked descriptions of the species. taxa by their leaf stabilities (Thorley and Wilkinson, 1999) given the set of topologies resulting from the combined Bayesian analy- 2.4.2. Character analysis sis: a total of 19,200 tree topologies not including those estimated The morphological character state changes of a subset of the 45 during burn-in, using a Python program developed by T. Hill (Nat- characters in Table 2 were mapped onto the topology of the Bayes- ural History Museum, London). Leaf stability of a taxon is based on ian inference tree resulting from combined analysis of 18S and 28S the average of the support (here maximum posterior probabilities) rDNA data using MacClade with accelerated transformation. Char- for each subtree containing that taxon and any other pair of taxa. acter state changes shown in bold in Table 2 were those of key importance to previous trypanorhynch classification schemes 2.4. Coding and analysis of morphology and/or based on our results, showed systematic utility (i.e. help to define natural groups). 2.4.1. Coding The morphological characters on which we have focused are 2.5. Analysis of host associations and phylogenetic dating based on those of Beveridge et al. (1999). However, a number of the characters associated with tentacle armature have been refined 2.5.1. Host association matrix construction into more discrete and thus unambiguously codeable units. Four A matrix of host associations coded at the levels of host family, additional characters were added to accommodate the expanded host order and a division between batoids and sharks was

Table 2 Character descriptions.

1. Uterus: 0 = uterus pre-formed in midline of segment; 1 = uterus growing from anlage at end of uterine duct to form inverted U-shaped gravid uterus 2. Larval type: 0 = plerocercoid; 1 = plerocercus or merocercoida 3. Sensory fossettes: 0 = absent; 1 = present 4. Bothria: 0 = four; 1 = two 5. Genitalia: 0 = single; 1 = paired 6. Bothria: 0 = sessile; 1 = pedicellate 7. Prebulbar organ: 0 = absent; 1 = present; 9 = inapplicable 8. Retractor muscle origin: 0 = at base of bulb; 1 = middle of bulb; 2 = in tentacle sheath; 9 = inapplicable 9. Relative length of pars bothrialis to pars vaginalis: 0 = pars bothrialis equal to or greater than pars vaginalis; 1 = pars bothrialis much shorter than pars vaginalis; 9 = inapplicable 10. Genital pore position: 0 = ventrosubmarginal; 1 = marginal 11. Internal seminal vesicle: 0 = absent; 1 = present 12. External seminal vesicle: 0 = absent; 1 = present 13. Accessory seminal vesicle: 0 = absent; 1 = present 14. Hermaphroditic duct: 0 = absent; 1 = present 15. Scolex posterior margin: 0 = acraspedote; 1 = craspedote 16. Uterine pore: 0 = present; 1 = absent 17. Uterus configuration: 0 = central, symmetrical; 1 = deviating anteriorly towards genital pore 18. Posterior margin of uterus: 0 = simple; 1 = with two posterior diverticula 19. Tentacular bulb length: 0 = short (length to width ratio less than 5:1); 1 = long (length to width ratio greater than 5:1); 9 = inapplicable 20. Gland cells within tentacular bulbs: 0 = absent; 1 = present; 9 = inapplicable 21. Pars post bulbosa size: 0 = small to absent; 1 = prominent; 9 = inapplicable 22. Bothrial margins: 0 = relatively thin; 1 = prominently thickened; 2 = elevated throughout dorsal and ventral aspects of scolex 23. Testes distribution: 0 = exclusively preovarian; 1 = some testes postovarian 24. Pintner’s cells in scolex: 0 = absent; 1 = present 25. Tentacle length: 0 = short (less than 20–25 principal rows of hooks); 1 = long (more than 20–25 principal rows of hooks); 9 = inapplicable 26. Basal swelling on tentacle: 0 = absent; 1 = present; 9 = inapplicable 27. Distinctive basal armature: 0 = absent; 1 = present; 9 = inapplicable 28. Hook type: 0 = solid; 1 = hollow; 9 = inapplicable 29. Metabasal armature type: 0 = homeomorphous; 1 = heteromorphous; 9 = inapplicable 30. Hook pattern: 0 = homeoacanthous (quincunxes); 1 = heteroacanthous (half spirals); 9 = inapplicable 31. Arrangement of hook files 1(10) on internal surface: 0 = convergent; 1 = divergent; 9 = inapplicable 32. Hook symmetry: 0 = rotational symmetry; 1 = glide reflection internal/external; 2 = glide reflection bothrial/abothrial; 9 = inapplicable 33. Intercalary hooks: 0 = absent (typical); 1 = 1 band (atypical); 2 = multiple bands (multi-atypical); 9 = inapplicable 34. Chainette: 0 = absent; 1 = present; 9 = inapplicable 35. Number of chainette bands: 0 = single; 1 = double; 2 = multiple; 9 = inapplicable 36. Chainette form: 0 = simple; 1 = with wings; 9 = inapplicable 37. Satellite hooks: 0 = absent; 1 = present; 9 = inapplicable 38. Macrohooks on base of tentacle: 0 = absent; 1 = single or series; 9 = inapplicable 39. Falcate macrohooks on base of tentacle: 0 = absent; 1 = present; 9 = inapplicable 40. Bill-hooks on base of tentacle: 0 = absent; 1 = present; 9 = inapplicable 41. External tentacular sheath: 0 = basal region of tentacle without unarmed sheath; 1 = basal region of tentacle with unarmed sheath; 9 = inapplicable 42. Hermaphroditic vesicle: 0 = absent; 1 = present 43. Principal hook rows: 0 = continuous; 1 = interrupted; 9 = inapplicable 44. Posterior extremities of tentacular bulbs: 0 = directed posteriorly; 1 = directed laterally; 9 = inapplicable 45. Tentacular apparatus: 0 = present; 1 = absent

a Characters indicated in bold are mapped onto the tree in Fig. 6. Author's personal copy

232 P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 constructed using MacClade to examine the distribution of host Tracer, and consensus topologies and 95% confidence intervals associations with respect to the phylogeny of trypanorhynchs. Host for the nodal ages calculated using TreeAnnotator (both programs associations for each species consisted of the actual hosts, in the part of the BEAST software package). ESSs of some parameters cases of specimens collected from definitive hosts, in combination were found to be insufficient after 10,000,000 generations for the with previously reported definitive hosts as compiled in Palm relaxed clock analysis and thus this analysis was rerun with the (2004). Those trypanorhynch species infecting multiple host taxa chain increased to 15,000,000. The first 1000 and 1500 saved trees, were coded as multistate for any and all relevant taxonomic levels. for the strict and relaxed clock analyses respectively, were re- For two taxa collected as larvae, Pseudogilquinia pillersi and Pseud- moved as burn-in prior to consensus analysis using TreeAnnotator. otobothrium dipsacum, definitive hosts have yet to be identified The resulting linearized consensus trees were drawn using FigTree (Palm, 2004) and thus character states for these taxa were coded ver. 1.0 (Rambaut, A., Edinburg, UK, tree.bio.ed.ac.uk/software/); as question marks. The matrix is shown in Supplementary Data S1. and further annotated using Adobe Illustrator.

2.5.2. Ancestral character state reconstructions (ACSRs) 3. Results The absence of a comprehensive molecular data set that includes the same species of elasmobranchs that host the parasite 3.1. Analysis of 18S and 28S rRNA species analyzed herein prevented any type of congruence analysis of independently-derived host and parasite phylogenies using Both individual (Fig. 4) and combined (Fig. 5) analyses resulted methods such as TreeMap (Page, 1994) or Jungles (Charleston, in well-resolved and strongly supported topologies characterized 1998). Instead, the distribution of known, contemporary host asso- by a pattern consisting of two major lineages, one (upper clade ciations across the parasite phylogeny was used to identify the in Fig. 5) consisting of the Eutetrarhynchidae and one family of most likely ancestral host ‘states’ for nodes subtending major the former Tentacularioidea (i.e. the Tentaculariidae), and the clades. ACSRs were estimated for the three hierarchical levels at other (lower clade in Fig. 5) consisting of the remaining try- which host associations were coded in the matrix in Supplemen- panorhynch taxa. These two major lineages showed marked differ- tary Data S1 (i.e. family, order and shark versus batoid) using ences in divergence rates that can be seen in their relative branch BayesTraits (Pagel et al., 2004; ver. 1.0, available from www.evolu- lengths: the upper clade (Fig. 5) is characterized by relatively long tion.rdg.ac.uk) with the multistate model of evolution. By this internal and terminal branches, subtended by comparatively short method, the degree of uncertainty in the hypothesized ancestral basal internal branches, resulting in well-supported clades, albeit host states could be estimated from the data. with less support for the interrelationships of the clades themselves. In contrast, the lower clade (Fig. 5) is characterized by con- 2.5.3. Phylogenetic dating siderably shorter but nevertheless more equitable internal and Divergence dates under strict and relaxed clock models were terminal branches, providing less support for the monophyly of estimated by Bayesian analysis using BEAST (Drummond and individual clades but greater support for their interrelationships. Rambaut, 2007; ver. 1.4, available from www.bio.ed.ac.uk). In the These differences are reflected in their stability profiles, such that absence of any direct evidence of tapeworm ancestry (such as fos- the more equitable branch lengths that characterized members silized specimens), calibration was achieved using proxy dates of the lower lineage also led to greater stability in the positions based on the fossil record of the host groups. Although we are of individual taxa and thus higher leaf stability values (Fig. 5). Both aware that the estimated dates of first appearance of elasmobranch lineages, however, showed high clade stability (as indicated by the groups differ among authors (e.g. see Maisey et al., 2004), the rel- thick lines in Fig. 5) suggesting that membership of the individual atively broad taxon coverage provided in a single work by Cappetta clades in each lineage is stable. Thus extinct lineages notwith- et al. (1993) made comparisons across families feasible. Thus the standing, molecular data strongly suggest that the trypanorhynchs age at the root of the ingroup was estimated as the age of the ear- split into two separate major lineages early in their evolution, after liest known (i.e. minimum age) representative of the extant host which their diversification followed disparate evolutionary groups, as reported by Cappetta et al. (1993). This was 183 million tempos. years ago (Myr) based on the oldest known fossil of the Hexanchi- dae. Thus the ingroup clade was defined as a taxon set and the prior distribution for the age of its root was set to an exponential 3.2. Analysis of morphology curve with a zero offset of 183 Myr and an exponential mean of 10 Myr. Such a distribution is considered to be most appropriate Seventeen of the 45 morphological characters in Table 2 have or for minimum age dates as it avoids sampling more recent dates now appear to have potential systematic utility. These are shown whilst accommodating for uncertainty in the fossil record (see in Fig. 6 mapped onto the Bayesian inference tree resulting from Ho, 2007). The speciation tree prior was set to a Yule birth process combined analysis of 18S and 28S rDNA data. Mappings of only and other priors set as defaults. The substitution model was the three of these (Characters 3, 8 and 45, the latter being the apomor- GTR + I + G with four rate categories and empirical base frequen- phic loss of the rhyncheal apparatus) were found to be non- cies, as employed above using MrBayes, and the length of the chain homoplasious. (i.e. No. generations) was set initially to default values of 10,000,000 with samples saved every 1000 generations. 3.3. Analysis of host associations and estimated divergence dates The SD of the relaxed uncorrelated lognormal clock model (i.e. ‘ucld.stdev’ parameter) provides an indication of how clock-like ACSRs suggest that the two major lineages of trypanorhynchs divergence rates are, with values of zero indicating no variation diverged together with the two major lineages of neoselchians, in rates, and values over one (i.e. SD > mean rate) indicating high the Batoidea (rays) and Selachimorpha (sharks) (Fig. 7), and thus degrees of rate heterogeneity (Drummond et al., 2006). For our support recent molecular estimates of elasmobranch phylogeny combined data set that value was 0.618, suggesting that our data that hypothesize a basal divergence of shark and ray lineages are neither highly clock-like nor highly heterogeneous, and thus (see Fig. 3). Host switching from ray to shark hosts and from shark we chose to conduct and evaluate separate analyses using the strict to ray hosts is seen in both the ‘ray-inhabiting clade’ (RIC) (upper as well as relaxed clock models. Resulting prior and posterior dis- clade) and ‘shark-inhabiting clade’ (SIC) (lower clade), respectively, tributions and effective sample sizes (ESSs) were examined using but estimates show such events to be restricted largely to the re- Author's personal copy

P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 233

Nybelinia queenslandensis Shirleyrhynchus aetobatidis Nybelinia aequidentata Parachristianella indonesiensis Heteronybelinia estigmena Parachristianella baverstocki Kotorella sp. Halysiorhynchus macrocephalus Nybelinia africana Trygonicola macropora Kotorella pronosoma Trimacracanthus aetobatidis Nybelinia sphyrnae Parachristianella monomegacantha Tentacularia coryphaenae Prochristianella sp. (Ovr-36b) Rhinoptericola megacantha Prochristianella sp. (DNA-00-90) Dollfusiella geraschmidti Prochristianella hispida Oncomegas sp. Prochristianella clarkeae Oncomegas australiensis Nybelinia africana Dollfusiella sp. (Dolb) Kotorella pronosoma Tetrarhynchobothrium sp. (Trhy) Kotorella sp. Dollfusiella ocallaghani Nybelinia queenslandensis Dollfusiella sp. (Dolb3) Heteronybelinia estigmena Dollfusiella tenuispinis Nybelinia aequidentata Dollfusiella spinulifera Nybelinia sphyrnae Dollfusiella michiae Tentacularia coryphaenae Paroncomegas araya Rhinoptericola megacantha Dollfusiella martini Oncomegas sp. Tetrarhynchobothrium sp. (Ttm) Oncomegas australiensis Parachristianella baverstocki Dollfusiella geraschmidti Shirleyrhynchus aetobatidis Oncomegoides celatus (NT108A) Parachristianella indonesiensis Oncomegoides celatus (NT33D) Trimacracanthus aetobatidis Mecistobothrium johnstonei Parachristianella monomegacantha Prochristianella sp. (DNA-01-122A) Halysiorhynchus macrocephalus Prochristianella macracantha Trygonicola macropora Prochristianella sp. (Proch) Prochristianella sp. (DNA-00-90) Prochristianella sp. (DNA-00-16) Prochristianella sp. (Ovr-36b) Tetrarhynchobothrium sp. (Ttm) Prochristianella hispida Tetrarhynchobothrium sp. (Trhy) Prochristianella clarkeae Dollfusiella ocallaghani Prochristianella sp. (DNA-01-122A) Dollfusiella sp. (Dolb3) Prochristianella sp. (Proch) Dollfusiella sp. (Dolb) Prochristianella sp. (DNA-00-16) Dollfusiella tenuispinis Prochristianella macracantha Dollfusiella spinulifera Oncomegoides celatus (NT108A) Paroncomegas araya Oncomegoides celatus (NT33D) Dollfusiella michiae Mecistobothrium johnstonei Dollfusiella martini Mecistobothrium b revispine Grillotia pristiophori Iobothrium elegans Grillotia erinaceus Proemotobothrium linstowi Grillotia rowei Proemotobothrium sp. Pterobothrium lintoni Fossobothrium perplexum Pterobothrium platycephalum Poecilancistrum caryophyllum Pseudolacistorhynchus heronensis Otobothrium mugilis Paragrillotia similis Otobothrium p ropecysticum Posterior probabilities: Hornelliella annandalei Otobothrium ca rcharidis 28S rDNA 18S rDNA Callitetrarhynchus gracilis (Calg1) Pseudotobothrium dipsacum 90-100% Callitetrarhynchus gracilis (Ctg) Pseudotobothrium arii Floriceps minacanthus Parotobothrium balli < 90% Callitetrarhynchus gracilis (Calls) Callitetrarhynchus gracilis (Calg1) Floriceps saccatus Callitetrarhynchus gracilis (Ctg) Diesingium lomentaceum Floriceps minacanthus Bootstrap percentages: Lacistorhynchus dollfusi Callitetrarhynchus gracilis (Calls) 90-100% Dasyrhynchus giganteus Floriceps saccatus 0.01 changes 0.01 changes Protogrillotia sp. Diesingium lomentaceum 80-89% Pseudogilquinia pillersi Lacistorhynchus dollfusi Pseudogilquinia microbothria Pseudolacistorhynchus heronensis Proemotobothrium sp. Paragrillotia similis Iobothrium elegans Hornelliella annandalei Proemotobothrium linstowi Pterobothrium lintoni Poecilancistrum caryophyllum Pterobothrium platycephalum Fossobothrium perplexum Grillotia pristiophori Otobothrium mugilis Grillotia erinaceus Otobothrium p ropecysticum Grillotia rowei Otobothrium ca rcharidis Dasyrhynchus giganteus Pseudotobothrium dipsacum Protogrillotia sp. Pseudotobothrium arii Pseudogilquinia pillersi Parotobothrium balli Pseudogilquinia microbothria Gilquinia squali Gilquinia squali Vittirhynchus squali Vittirhynchus squali Chimaerarhynchus rougetae Chimaerarhynchus rougetae Sagittirhynchus aculeatus Sagittirhynchus aculeatus Aporhynchus sp. Aporhynchus sp. Hepatoxylon trichiuri Hepatoxylon trichiuri Sphyriocephalus viridis Sphyriocephalus viridis Molicola uncinatus Molicola uncinatus Gymnorhynchus isuri Gymnorhynchus isuri Ditrachybothridium mac rocephalum Ditrachybothridium mac rocephalum Echinobothrium sp. Echinobothrium sp. Macrobothridium rhynchobati Macrobothridium rhynchobati Echinobothrium chisholmae Echinobothrium chisholmae Echinobothrium harfordi Echinobothrium harfordi

Fig. 4. Results of independent analyses of 28S and 18S rDNA by Bayesian inference showing nodal support as posterior probabilities and bootstrap percentages. cent past. At least one exception is seen in the clade encompassing clades around the Triassic–Jurassic boundary (210 Myr), or slightly the majority of tentacularioid species (i.e. Nybelinia and its rela- older than the calibration date of 183 Myr. Diversification of the tives) which is estimated to have evolved from a host switch from RIC is estimated to have occurred near the start of the Cretaceous rays to sharks that occurred 40 Myr, with multiple, subsequent (140 Myr), whereas diversification of the SIC is estimated to have independent host switches back to rays (Fig. 7). ACSRs of the host occurred later, in the latter half of the Cretaceous (100 Myr), orders estimated the ancestral host of the RIC to be a member of based on a relaxed clock model. The strict clock model estimated the Rajiformes, whereas the likelihood of it being a member of an- even less equitable divergence dates, pushing the RIC back into other batoid group (i.e. Pristiformes or Torpediniformes) was neg- the Upper Jurassic and the SIC forward into the Upper Cretaceous ligible (i.e. <1%; see pie-charts in Fig. 7). In the SIC, the most likely (data not shown). However, the greatest discrepancy between ancestral host was estimated to be a member of the Carcharhini- the two models was found in the size of the error estimated, with formes, although the case is less clear cut than that above, as the the relaxed clock producing far larger error bars than the strict squaliform and lamniform sharks showed notable likelihood val- model. For example, whereas the strict clock estimated an error ues. At the level of host family, the likelihoods become highly dis- for the time of the diversification of the SIC that spans roughly persed and there is no evidence that either the finer scale the latter half of the Cretaceous, the same error bar based on the diversification of the host orders or families are reflected in the relaxed model spans from the Upper Jurassic to the end of the Cre- parasite phylogeny (but see Section 4). taceous, thus overlapping with the error associated with the origin Both the relaxed (Fig. 7) and strict (results not shown) clock of the RIC (Fig. 7). The relaxed model thus incorporates far greater analyses estimate the split between the two major trypanorhynch uncertainty in the divergence time estimates. Author's personal copy

234 P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242

HOST ASSOCIATIONS LEAF STABILITY VALUES Nybelinia queenslandensis BATOIDS SHARKS 61 Nybelinia aequidentata RAJIFORMES CARCHARHINIFORMES Heteronybelinia estigmena PRISTIFORMES SQUALIFORMES TORPEDINIFORMES LAMNIFORMES TEN 0.837 97 Kotorella sp. PRISTIOPHORIFORMES Nybelinia africana 0.907 96 78 ORECTOLOBIFORMES Kotorella pronosoma HEXANCHIFORMES 0.942 100 Nybelinia sphyrnae 0.961 100 Tentacularia coryphaenae Rhinoptericola megacantha Dollfusiella geraschmidti 0.05 Oncomegas australiensis 100 Posterior probabilities: Oncomegas sp. Dollfusiella sp. (Dolb) 62 90-100% Tetrarhynchobothrium sp. (Trhy) < 90% 0.885 100 Dollfusiella ocallaghani Dollfusiella sp. (Dolb3) Bootstrap percentages: 0.975 Dollfusiella tenuispinis No. 90-100% Dollfusiella spinulifera No. 50-89% 52 Dollfusiella michiae 0.979 100 Paroncomegas araya Dollfusiella martini Tetrarhynchobothrium sp. (Ttm) 61 Parachristianella baverstocki TBA TEN 97 Shirleyrhynchus aetobatidis 0.953 93 Parachristianella indonesiensis 0.730 99 Halysiorhynchus macrocephalus 78 69 Trygonicola macropora 0.946 100 Trimacracanthus aetobatidis Parachristianella monomegacantha 0.987 100 Prochristianella sp. (Ovr-36b) 100 58 Prochristianella sp. (DNA-00-90) 0.915 100 Prochristianella hispida Prochristianella clarkeae Oncomegoides celatus (NT108A) 100 73 Oncomegoides celatus (NT33D) Mecistobothrium johnstonei Prochristianella sp. (DNA-01-122A) 0.970 57 Prochristianella macracantha 100 Prochristianella sp. (Proch) Prochristianella sp. (DNA-00-16) 94 Grillotia pristiophori 85 Grillotia erinaceus 53 Grillotia rowei Pterobothrium lintoni 100 Pterobothrium platycephalum Pseudolacistorhynchus heroniensis 99 Paragrillotia similis 1.000 Hornelliella annandalei 96 Dasyrhynchus giganteus Protogrillotia sp. LAC 98 ? Pseudogilquinia pillersi 99 63 Pseudogilquinia microbothria 96 Callitetrarhynchus gracilis (Calg1) 98 Callitetrarhynchus gracilis (Ctg) 56 Floriceps minacanthus 100 ? Callitetrarhynchus gracilis (Calls) 0.997 93 Floriceps saccatus Diesingium lomentaceum 62 Lacistorhynchus dollfusi 0.968 66 63 Iobobothrium elegans 99 Proemotobothrium linstowi Proemotobothrium sp. 0.681 99 Poecilancistrum caryophyllum Fossobothrium perplexum

TSE 82 Otobothrium mugilis Otobothrium propecysticum 0.878 95 Otobothrium carcharidis 0.986 97 73 ? Pseudotobothrium dipsacum OTO 99 Pseudotobothrium arii Parotobothrium balli Gilquinia squali 99 Vittirhynchus squali 91 Chimaerarhynchus rougetae GYM 100 0.999 100 Sagittirhynchus aculeatus TEN Aporhynchus sp. 90 0.999 Sphyriocephalus viridis 98 Hepatoxylon trichiuri 1.000 79 Gymnorhynchus isuri 100 Molicola uncinatus Ditrachybothridium macrocephalum Echinobothrium sp. Macrobothridium rhynchobati Echinobothrium chisholmae Echinobothrium harfordi

0.870 0.896 0.922 0.948 0.974 1.0

Fig. 5. Tree resulting from combined analysis of 18S and 28S rDNA by Bayesian inference (BI) showing nodal support as posterior probabilities and bootstrap percentages. Thick branches indicate clades found in 100% of the 19,200 topologies estimated by BI; values shown on (or to the left of) the thick branches show the average stability value for the members of the clade (i.e. the internal stability of the clade). Histogram on the right shows the relative Leaf Stability Value of each taxon. Host orders are indicated with symbols adjacent to terminal taxon names; question marks indicate species for which the elasmobranch (deﬁnitive) host is unknown. EUT, Eutetrarhynchoidea; GYM, Gymnorhynchoidea; LAC, Lacistorhynchoidea; OTO, Otobothrioidea; TBA, Trypanobatoida; TEN, Tentacularioidea; TSE, Trypanoselachoida. Author's personal copy

P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 235

Nybelinia queenslandensis Nybelinia aequidentata 7(0) 29(0) Heteronybelinia estigmena 30(0) 29(1) Kotorella sp. 31(9) 32(1) 33(9) Nybelinia africana 34(9) Kotorella pronosoma 4(0) 20(0) Nybelinia sphyrnae Tentacularia coryphaenae 2(0) 32(0) 31(0) 32(1) 28(0) 28(1) Rhinoptericola megacantha Dollfusiella geraschmidti 30(0) Oncomegas australiensis 31(9) 33(9) Oncomegas sp. 34(9) 29(1) Dollfusiella sp. (Dolb) 29(0) Tetrarhynchobothrium sp. (Trhy) 31(1) Dollfusiella ocallaghani Dollfusiella sp. (Dolb3) 32(2) Dollfusiella tenuispinis† Dollfusiella spinulifera

28(0) Dollfusiella michiae 32(0) Paroncomegas araya 30(0) 31(9) 32(2) 33(9) 34(9) Dollfusiella martini Tetrarhynchobothrium sp. (Ttm) 4(0) Parachristianella baverstocki 4(0) 20(0) 33(2) Shirleyrhynchus aetobatidis 34(1) Parachristianella indonesiensis 7(1) 16(1) 20(1) Halysiorhynchus macrocephalus Trygonicola macropora Trimacracanthus aetobatidis 28(0) Parachristianella monomegacantha Prochristianella sp. (Ovr-36b)† Prochristianella sp. (DNA-00-90)†

32(2) Prochristianella hispida Prochristianella clarkeae 33(1) Oncomegoides celatus (NT108A)

32(1) Oncomegoides celatus (NT33D) Mecistobothrium johnstonei 32(2) Prochristianella sp. (DNA-01-122A)† Prochristianella macracantha Prochristianella sp. (Proch)

16(1) 33(1) Prochristianella sp. (DNA-00-16)† Grillotia pristiophori 33(2) 14(1) 14(0) Grillotia erinaceus 34(0) Grillotia rowei 4(0) Pterobothrium lintoni 16(1) 34(0) Pterobothrium platycephalum 33(2) Pseudolacistorhynchus heroniensis Paragrillotia similis Hornelliella annandalei 34(0) Dasyrhynchus giganteus 33(2) Protogrillotia sp. 33(1) Pseudogilquinia pillersi 34(1) Pseudogilquinia microbothria Callitetrarhynchus gracilis (Calg1) 16(1) Callitetrarhynchus gracilis (Ctg) Floriceps minacanthus Callitetrarhynchus gracilis (Calls) Floriceps saccatus

16(1) Diesingium lomentaceum 8(1)*14(1) Lacistorhynchus dollfusi 34(1) Iobobothrium elegans Proemotobothrium linstowi

33(2) 14(0) Proemotobothrium sp. 32(2)Poecilancistrum caryophyllum Fossobothrium perplexum 33(1) 32(2) 7(0) 16(0) Otobothrium mugilis 20(0) 3(1)*16(1) Otobothrium p ropecysticum† 7(9) Otobothrium ca rcharidis 8(9)* 14(0) Pseudotobothrium dipsacum 20(9) 28(0) 32(2) 28(9) 30(0) Pseudotobothrium arii 29(9) 31(9) Parotobothrium balli 30(9) 33(9) 16(1) 31(9) 34(9) Gilquinia squali 32(9) 34(1) 33(9) Vittirhynchus squali 34(1) 34(9) 13(1) Chimaerarhynchus rougetae 2(0) 45(1)* 4(1) Sagittirhynchus aculeatus 8(2)* 4(0) 29(0) Aporhynchus sp. 30(0) Sphyriocephalus viridis 31(9) 32(0) 33(2) Hepatoxylon trichiuri 33(9) 34(1) Gymnorhynchus isuri 34(9) 13(1) 16(1) Molicola uncinatus Ditrachybothridium mac rocephalum † Echinobothrium sp.† Macrobothridium rhynchobati † Echinobothrium chisholmae† Echinobothrium harfordi †

Fig. 6. Character state distributions of selected morphological characters (with states in parentheses) mapped using accelerated transformation onto the Bayesian inference topology resulting from combined analysis of 18S and 28S rDNA data, as shown in Fig. 5. indicates species for which morphological data were not coded. For character descriptions see Table 2. Asterisks indicate characters that are non-homoplasious. Character ambiguities are mapped on the tree at their basal-most potential position. Author's personal copy

236 P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242

Kotorella sp. GTR+I+G / RELAXED CLOCK Heteronybelinia estigmena Nybelinia queenslandensis Nybelinia aequidentata Nybelinia africana HOST ASSOCIATIONS Kotorella pronosoma Nybelinia sphyrnae Tentacularia coryphaenae BATOIDS SHARKS Rhinoptericola megacantha Oncomegas australiensis RAJIFORMES CARCHARHINIFORMES Oncomegas sp. PRISTIFORMES SQUALIFORMES Dollfusiella geraschmidti Dollfusiella sp. (Dolb) TORPEDINIFORMES LAMNIFORMES Tetrarhynchobothrium sp. (Trhy) PRISTIOPHORIFORMES Dollfusiella ocallaghani Dollfusiella sp. (Dolb3) ORECTOLOBIFORMES Dollfusiella tenuispinis 93% Rajidae Dollfusiella spinulifera HEXANCHIFORMES Dollfusiella michiae 3.5% Carcharhinidae Paroncomegas araya 1.2% Myliobatidae Dollfusiella martini 2.3% other families combined Tetrarhynchobothrium sp. (Ttm) Parachristianella baverstocki Shirleyrhynchus aetobatidis TRYPANOBATOIDA Parachristianella indonesiensis Halysiorhynchus macrocephalus Trygonicola macropora Trimacracanthus aetobatidis Parachristianella monomegacantha Prochristianella sp. (Ovr-36b) RIC Prochristianella sp. (DNA-00-90) Prochristianella hispida Prochristianella clarkeae Oncomegoides celatus (NT108A) Oncomegoides celatus (NT33D) Mecistobothrium johnstonei Prochristianella sp. (DNA-01-122A) Prochristianella macracantha Prochristianella sp. (Proch) Prochristianella sp. (DNA-00-16) Grillotia pristiophori Pseudolacistorhynchus heronensis Grillotia rowei Pterobothrium lintoni Pterobothrium platycephalum TRYPANORHYNCHA Grillotia erinaceus Paragrillotia similis Hornelliella annandalei Dasyrhynchus giganteus Protogrillotia sp. ? Pseudogilquinia pillersi Pseudogilquinia microbothria Callitetrarhynchus gracilis (Calg1) Callitetrarhynchus gracilis (Ctg) Floriceps saccatus Floriceps minacanthus Callitetrarhynchus gracilis (Calls) Diesingium lomentaceum Lacistorhynchus dollfusi Iobobothrium elegans Proemotobothrium linstowi Proemotobothrium sp. Poecilancistrum caryophyllum SIC Fossobothrium perplexum Otobothrium mugilis Otobothrium propecysticum Otobothrium carcharidis ? Pseudotobothrium dipsacum TRYPANOSELACHOIDA Pseudotobothrium arii Parotobothrium balli 57% Carcharhinidae Gilquinia squali Vittirhynchus squali 16.2% Alopiidae Chimaerarhynchus rougetae 8.2% Lamnidae Sagittirhynchus aculeatus 3.6% Etmopteridae Aporhynchus sp. Sphyriocephalus viridis 1.5% Centrophoridae Hepatoxylon trichiuri 1.5% Squalidae Gymnorhynchus isuri 12% other families combined Molicola uncinatus Macrobothridium rhynchobati DIPHYLLIDEA (OUTGROUP) Echinobothrium chisholmae Echinobothrium harfordi Echinobothrium sp. Ditrachybothridium macrocephalum

PERMIAN TRIASSIC JURASSIC CRETACEOUS TERTIARY

290 245 210 140 65 1.6 0 mya Urolophidae Hexanchidae Orectolobidae Somniosidae Pristiophoridae Potamotrygonidae Scyliorhinidae Etmopteridae Rhinobatidae Squalidae Ginglymostomatidae Myliobatidae Rhynchobatidae

Alopiidae Hemiscylliidae Parascylliidae Rajidae Pristidae Dasyatidae Gymnuridae Rhinopteridae

Triakidae Torpedinidae Sphyrnidae

Carcharhinidae Lamnidae Dalatiidae Stegastomatidae Rhincodontidae

Fig. 7. Trypanorhynch chronogram based on the combined data set using a relaxed, uncorrelated clock model. Confidence intervals (95%) for nodal dates shown in light grey, except for the calibration node shown in black. Dashed drop-lines indicate earliest known fossil dates from Cappetta et al. (1993), where known, for extant elasmobranch families relevant to the current analysis. Host associations at the level of order (sensu Shirai, 1996) are indicated with solid symbols representing batoids and open symbols representing sharks; question marks indicate species for which the elasmobranch (definitive) host is unknown. Likelihood-based ancestral character state reconstructions (ACSRs) of the hosts are shown as pie charts for major divergences in the trypanorhynch tree, with the small charts showing the likelihood of shark versus ray ancestral hosts, and the larger charts the likelihoods of specific host super-orders. ACSRs at the level of host family are shown as percentages for the two basal most nodes of the ingroup. RIC, ray-inhabiting clade; SIC, shark-inhabiting clade.

4. Discussion has not been widely endorsed. In fact, in four of the most comprehensive treatments of the order (Dollfus, 1942; Campbell and 4.1. Implications of the molecular data for trypanorhynch classification Beveridge, 1994; Beveridge et al., 1999; Palm, 2004), suborders were not recognized because the system of Guiart was thought The trypanorhynchs have been divided historically into two to be non-natural. Following Pintner (1913, 1931), Dollfus (1942) suborders based primarily on whether they possess a plerocercoid focused his efforts to provide a classification of the trypanorhynchs or plerocercus larva. Referred to as the Acystidea and Cystidea, on features associated with the size, shape and arrangement of the respectively, by Guiart (1927), he later replaced these names with tentacle hooks, coining the formal terminology for hook features Atheca and Thecaphora to avoid redundancy with existing echino- that remains in use today (e.g. Campbell and Beveridge, 1994). derm names (Guiart, 1931). This system, but with the former ter- The result of Dollfus’ efforts was the recognition of four major minology, was followed in several important subsequent works groups, which he referred to as the Homéacanthes, Heteracantha (e.g. Fuhrmann, 1931; Schmidt, 1986). However, recognition of typical, Heteracantha atypical and Pécilacanthes. These groups these two suborders, or of suborders in trypanorhynchs in general, were slightly redefined by Campbell and Beveridge (1994) to re- Author's personal copy

P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 237

flect hook pattern symmetry, and were recognized formally as the to be para- or in a few cases polyphyletic, both here and by Palm superfamilies Homeacanthoidea, Heteracanthoidea, Otobothrioi- et al. (2009): Nybelinia, Kotorella, Dollfusiella, Tetrarhynchobothrium, dea and Poecilacanthoidea. Palm (2004), believing that tentacle Parachristianella, Prochristianella, Proemotobothrium, Callitetrarhyn- armature should not be used to define higher trypanorhynch taxa chus and Floriceps. This is despite the fact that the latter two genera owing to the potential homoplasy associated with similar modes of were said to be monophyletic by Palm et al. (2009). In the case of attachment, took a somewhat different approach. His classification Callitetrarhynchus gracilis, whereas the two specimens collected emphasized other features of the scolex such as bothrial pits, from sharks were found to group together on our tree, the third prebulbar organs, complete or partly reduced rows of hooks and specimen, collected as a larva from a bonyfish, grouped well away larval type to define five superfamilies: Tentacularioidea, Gymno- from the adult specimens of their conspecific. In an attempt to re- rhynchoidea, Lacistorhynchoidea, Otobothrioidea and Eutetrarhyn- solve this apparent anomaly, the relevant voucher specimens were choidea, the memberships of which had little correspondence to compared with all specimens of C. gracilis available in museum col- the superfamilies of Campbell and Beveridge (1994). Palm et al. lections (British Museum of Natural History, UK: Muséum National (2009) persisted in recognizing these taxa despite the fact that d’Histoire Naturelle, France; United States National Parasite Collec- the topology of the trees resulting from their molecular work show tion, USA; South Australia Museum, Australia; Natural History Mu- three of the five families to be non-natural as circumscribed by seum Vienna, Austria; Muséum d’Histoire Naturelle Genève, Palm (2004). Switzerland) but no discrete morphological differences were de- The results of the molecular analyses conducted here (Fig. 5) are tected among specimens. However, the formal investigation of inconsistent with both the suborders Atheca and Thecaphora, and the monophyly of these and other trypanorhynch genera was be- deviate substantially from the superfamilial classification scheme yond the scope of this study and will require more focused sam- of Campbell and Beveridge (1994). Our results are most consistent pling efforts. with the scheme proposed by Palm (2004) and subsequently endorsed by Palm et al. (2009), but even here several major differ- 4.2. Nomenclatural recognition of two primary lineages ences exist. These are due in part to differences in tree topology, but also result from whether monophyly of higher taxa is required. The most conspicuous result of the molecular analyses con- For example, Palm et al. (2009) accepted the Otobothrioidea as a ducted here is the suggestion that trypanorhynchs in fact comprise superfamily, despite its derived placement among the Lacistorhyn- two independent lineages. One of these consists of the choidea. In our analyses, the Lacistorhynchoidea grouped as sister Eutetrarhynchoidea and one family of tentacularioid, both of which to the Otobothrioidea and thus we recognize both taxa as mono- should now be considered to belong to a single superfamily. The phyletic superfamilies. Palm et al. (2009) persisted in the accep- other consists of the Lacistorhynchoidea, Otobothrioidea, Gymno- tance of a paraphyletic Eutetrarhynchoidea with their recognition rhynchoidea as well as the sphyriocephalid genera Sphyriocephalus of the Tentacularioidea as an independent superfamily, despite and Hepatoxylon, which were considered by Palm (2004) to belong its derived position among the eutetrarhynchoids on their trees. to the Tentacularioidea but should now be considered with the In an effort to circumscribe monophyletic superfamilies, our Gymnorhynchoidea. Although supported by unique suites of char- scheme would include the taxa previously assigned to the acters discussed below, host associations of these two lineages are Eutetrarhynchoidea and the Tentacularioidea in the same super- perhaps their most consistent distinguishing features (see Fig. 7). family. However, given that Tentaculariidae is the older of the We thus propose that they be formally recognized as suborders, two names and thus has priority, the group, if recognized as a sin- and in recognition of the ray associations of the first lineage pro- gle superfamily, should be referred to as the Tentacularioidea. pose the name Trypanobatoida, and in recognition of the shark Based on our results, we concur with Palm et al. (2009) that the associations of the second lineage, propose the name Trypanosela- Gymnorhynchoidea is monophyletic only if it is considered to in- choida. These names reflect not only the most common host asso- clude the Sphyriocephalidae, a family formerly placed in the Tenta- ciations within each lineage, but are hypothesized via ACSRs to be cularioidea (e.g. see Campbell and Beveridge, 1994; Palm, 2004). their ancestral host groups (see below). Our results provide little support for the family-level classification of Palm (2004) and amplified in Palm et al. (2009). Neither the 4.3. Implications of molecular data for interpretations of character Lacistorhynchidae nor either of its subfamilies (Lacistorhynchinae evolution and Grillotiinae) are monophyletic. Similarly, neither of two families of Eutetrarhynchoidea included here (Eutetrarhynchidae and Characters that have been employed historically in the circum- Rhinoptericolidae) is monophyletic. In fact, these results suggest scription of subgroups (i.e. suborders, superfamilies, families) have that the decision of Palm (2004) to synonymize Shirleyrhynchidae included the number of bothria (e.g. Blainville, 1828; Southwell, Campbell and Beveridge, 1994 with Rhinoptericolidae Carvajal and 1929; Pintner, 1931; Dollfus, 1942; Campbell and Beveridge, Campbell, 1975 was ill advised. With the exception of the mono- 1994; Palm, 2004), larval type (e.g. Diesing, 1850; Braun, 1894; generic Aporhynchidae, neither of the other families of Gymno- Guiart, 1927; Schmidt, 1986; Palm, 2004), presence or absence of rhynchoidea (Gymnorhynchidae or Gilquiniidae) as defined by ciliated bothrial pits (= sensory fossettes) (e.g. Linton, 1889; Palm (2004) and Beveridge and Justine (2006) is monophyletic. Southwell, 1929; Campbell and Beveridge, 1994; Palm, 2004), Whereas our results suggest that the Tentaculariidae is monophy- presence or absence of pre-formed uterine pores (e.g. Pintner, letic as currently circumscribed, recognition of that family raises 1913; Campbell and Beveridge, 1994), presence or absence of serious issues with the familial level classification of species in prebulbar organs (e.g. Palm, 2004), and an extensive suite of char- some of its immediate sister genera such as Rhinoptericola and acters associated with tentacular armature patterns (e.g. Pintner, Oncomegas, which do not collectively form a monophyletic assem- 1913, 1931; Dollfus, 1942; Campbell and Beveridge, 1994). It is blage. In addition, and contrary to Palm et al. (2009), our results on these kinds of characters that we therefore focus our morpho- suggest that both the Otobothriidae and Pseudotobothriidae are logical discussions. Such characters of interest are shown in monophyletic. Fig. 6 mapped onto the topology of the combined molecular tree Our work also calls into question the monophyly of many of the topology (Fig. 5). genera included in this study suggesting that the generic-level Among the principal insights provided by these analyses was classification of trypanorhynchs in most superfamilies is in need the fact that features employed previously to recognize suborders of revision. Phylogenetic analyses revealed the following genera and/or superfamilies are unreliable indicators of genealogy. For Author's personal copy

238 P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 example, larval type (Character 2) fails to circumscribe the two also lacking from Rhinoptericola megacantha and Shirleyrhynchus major clades and although the derived state of this character (i.e. aetobatidis. Origin of the tentacle retractor muscles within the mid- lacking a blastocyst) more or less defined the Tentaculariidae, this dle of the bulb (Character 8) unambiguously unites the Otoboth- condition also occurred in two genera that should be considered rioidea and the Lacistorhynchoidea. An accessory seminal vesicle members of the superfamily Gymnorhynchoidea (i.e. Hepatoxylon (Character 13) is restricted to the Gymnorhynchoidea. While sev- and Sphyriocephalus, both in the Sphyriocephalidae). This explains eral species within this superfamily were coded as lacking this fea- the lack of support for the family-level classification of Palm (2004) ture, the unusual genital structures seen in some of these species, in which the Sphyriocephalidae was placed in Gymnorhynchoidea such as Gymnorhynchus isuri, bear further consideration in light of and not the Tentacularioidea. Similarly, the major hook features the tree presented here. The presence of an hermaphroditic duct including armature type (Character 29), hook pattern (Character (Character 14) is restricted to the Lacistorhynchoidea and Otoboth- 30), hook symmetry (Character 32) and chainette type (Character rioidea and, in fact, with a few exceptions appears to characterize 34) were found to be homoplasious and fail to define major groups. these two superfamilies. The terminal genitalia of the exceptions, For example, homoplasy in hook pattern reveals the Homéacanthes which apparently lack a hermaphroditic duct, should be more clo- of Dollfus (1942), Homeacanthoidea of Campbell and Beveridge sely scrutinized given that congeners of some of these species (e.g. (1994), Tentacularioidea of Palm (2004), as well as the Heteracan- Grillotia australis, not included here) have been found to exhibit a tha of Dollfus (1942) and Heteracanthoidea of Campbell and vagina that runs within the cirrus sac. The relatively derived posi- Beveridge (1994) to be polyphyletic. Some of this homoplasy, how- tion of Aporhynchus sp. among the Gymnorhynchoidea serves to ever, is likely to result from difficulties in interpreting hook pat- confirm the suggestion by Beveridge (1990) that the absence of terns and this has been mooted previously. For example, all elements of the rhyncheal apparatus (Character 45) seen in Bussieras (1970) and Beveridge and Campbell (1996) noted that the Aporhynchidae represents a loss rather than a plesiomorphic whereas the metabasal armature of the Sphyriocephalidae and absence of this suite of traits. Hepatoxylidae was homeoacanthous, their basal armature was heteroacanthous. Homoplasy in intercalary hooks (Character 33) 4.4. Morphological characters underpinning recognition of the reveals the Heteracantha of Dollfus (1942) and the Otobothrioidea Trypanobatoida and Trypanoselachoida as circumscribed by Campbell and Beveridge (1994) also to be polyphyletic. Homoplasy in the presence of a chainette (Character Beyond the robust molecular support for the proposed suborders 34) suggests that the Pécilacanthes of Dollfus (1942) and the Poe- Trypanobatoida and Trypanoselachoida, these lineages can be cir- cilacanthoidea of Campbell and Beveridge (1994) are polyphyletic. cumscribed to some extent by unique suites of morphological char- The notion that there might exist homplasy in chainettes was first acters. For example, all Trypanobatoida except the tentacularoids, advanced by Beveridge and Campbell (1989) who recognized that possess prebulbar organs and gland cells within the tentacular there is considerable diversity in hook types and arrangements bulbs and all Trypanobatoida exhibit retractor muscles that origi- currently considered to comprise a ‘‘chainette”. We suspect that nate at the bases of the tentacular bulbs. In contrast, the Trypanos- a more careful examination of the criteria employed to recognize elachoida lack prebulbar organs and gland cells within the this feature would do much to resolve at least some of the apparent tentacular bulbs and, with the exception of the Gymnorhynchoidea, homoplasy. exhibit retractor muscles that originate somewhere near the middle There are a number of additional apparently homoplasious of the tentacular bulbs. Furthermore, it could be argued that while characters which, although not incorporated into formal classifica- the Trypanobatoida exhibit simple terminal genital ducts, the tion schemes, have been considered of importance to try- Trypanoselachoida exhibit more complex terminal genital ducts panorhynch classification. For example, solid hooks (Character that include hermaphroditic ducts (Lacistorhynchoidea and 28) appear multiple times within the Trypanobatoida and only Otobothrioidea), accessory seminal vesicles (Gilquiniidae and once within the Trypanoselachoida. The evolution of four bothria Gymnorhynchidae) and extensions of the internal seminal vesicle (Character 4), an important character for distinguishing families beyond the muscular wall of the cirrus sac into which the retracted in Palm (2004), occurs multiple times within both suborders. The cirrus extends (Grillotia, Gymnorhynchus), none of which have been derived state occurs independently at multiple points within the observed in the former suborder. In the Trypanobatoida, modifica- Eutetrarhynchoidea, once within the Lacistorhynchoidea and oc- tions of the terminal genital ducts are restricted to the presence of curs in most of the Gymnorhychoidea, with the exception of Hep- paired internal seminal vesicles (e.g. Tetrarhynchobothrium) and atoxylon and Sphyriocephalus. Once again, the potential for vaginal diverticula (Zygorhynchus) neither of which have been con- homoplasy in this character has been mooted previously (e.g. sidered in the present analysis. Although the morphological under- Beveridge et al., 1999). pinnings of the proposed suborders require additional investigation, In contrast, a number of features were found to have systematic strong molecular support for these primary lineages suggests that utility. Among these is the presence of bothrial fossettes (Character reciprocal illumination may reveal additional characters. 3) which provides support for Palm’s superfamily Otobothrioidea uniting taxa that exhibit fossettes on the margins of their bothria. 4.5. Implications for previous interpretations of trypanorhynch However, it is notable that some otobothrioid species not included evolution in our analyses lack (i.e. Otobothrium curtum) or have partially reduced (i.e. Symbothriorhynchus spp.) fossettes. The convergent Existing explicit morphology-based phylogenetic hypotheses arrangement of hook files (Character 31) is restricted to one of for the trypanorhynchs (e.g. Campbell and Beveridge, 1994; Bever- the two clades of eutetrarhynchoids, but is somewhat homoplas- idge et al., 1999; Palm, 2004) are highly incongruent. It is likely ious. Pre-formed uterine pores (Character 16) are restricted to that differences in the methods employed in the generation of the Trypanoselachoida, but within this clade are conspicuously ab- these trees account for much of this incongruence. For example, sent from the Otobothrioidea, from various species in the Laci- the tree presented by Campbell and Beveridge (1994) was meant storhynchoidea and possibly also the Gymnorhynchoidea. The merely to summarize their thoughts on trypanorhynch phyloge- presence of prebulbar organs (Character 7) and gland cells within netic relationships; on this tree they mapped 13 characters they the tentacular bulbs (Character 20) are both restricted to the Try- considered to be of use in supporting these relationships. In con- panobatoida, however, in both cases within this suborder, these trast, the trees of Beveridge et al. (1999) and Palm (2004) were features are lacking from the Tentaculariidae. The latter feature is the result of matrix-based phylogenetic analyses. However, the Author's personal copy

P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 239 characters on which these trees were based differed. Whereas the contemporary hosts in this clade (see Fig. 7). This is in contrast to former was based on 44 characters, 17 of which concerned tentacle our understanding of elasmobranch phylogeny in which the lamn- armature, the latter expanded the characters to a total of 60, 24 of iform sharks are considered to represent the sister taxon of the car- which concerned tentacle armature. In addition, these authors charhiniform sharks (see Fig. 3), whereas the squaliform sharks are handled the challenges imposed by the unique nature of many of thought to share a more recent common ancestor with the angel the features of trypanorhynchs differently. Whereas Beveridge sharks (Squatiniformes), saw sharks (Pristiophoriformes) and re- et al. (1999) assumed that the simplest condition of each of their lated groups (see Fig. 3). Thus even such major diversification characters was plesiomorphic, Palm (2004) assumed (with a few events in the host phylogeny do not appear to be mirrored by exceptions) the most common state of each character to be the parasite phylogeny. However, these data remain to be evalu- plesiomorphic. ated in the context of a comprehensive, molecular-based phylog- Beyond the level of major subgroups, all of these morphological eny of the elasmobranchs, such that topological congruency hypotheses also differ substantially from the trees resulting from analyses can be conducted without the confounding problem of the molecular work conducted here. Certainly our formal use of missing data. outgroup comparisons is likely to contribute to some of these dif- The level of fidelity shown to the final, elasmobranch hosts has ferences. However, the taxonomic category of interest among these been shown to be highly variable among trypanorhynchs (Palm studies also differs. Whereas our analyses were specimen (and thus and Caira, 2008), and thus one would predict different evolutionary species) based, Campbell and Beveridge (1994) presented their re- influences (i.e. phylogenetic versus ecological) depending on sults in the context of families, and both Beveridge et al. (1999) and whether the specific trypanorhynch group examined showed high Palm (2004) focused on genus-level taxa, thereby using combined or low levels of host specificity. The Trypanoselachoida and Trypa- codings for their terminal taxa. The latter approaches are clearly nobatoida each contain examples of both extremes of host-speci- invalid in cases where families and genera have been shown to ficity, and include species (although rare) that parasitize both be non-natural. sharks and rays (e.g. species of Dollfusiella, Parachristianella and It is similarly difficult to compare the results of the character Lacistorhynchus). In most cases, host switching inferred from the analysis employed here, based on formal coding and parsimony parasite phylogeny can be readily explained by ecological, rather mapping of 45 characters for each of our 79 ingroup taxa, with than phylogenetic, influences. For example, the clade within the the discussion of the 13 morphological characters assumed to rep- Trypanoselachoida including species of Grillotia and Pterobothrium resent important morphological synapomorphies in Palm et al. contains species parasitising rays that are closely related to species (2009). This is not only because the method employed to obtain parasitising pristiophoriform and orectolobiform sharks, both host the character mapping presented by the latter authors is not artic- groups including many bottom-feeding taxa whose ecological ulated, but also because the mappings presented do not appear to niche overlaps extensively with that of the batoids (Wetherbee be consistent. For example, the character ‘‘solid hooks” is indicated and Cortés, 2004). The opportunity for host-switching in such as a synapomorphy for their clades Eutetrarhynchoidea and Tenta- groups is thus great, and evidence of ecologically driven host- cularioidea, but no indication is given that one of the relatively ma- switching combined with the existence of generalist species that jor clades of this group (including a diversity of Dolfusiella and can infect a broad spectrum of elasmobranchs suggest that any Tetrarhynchobothrium and Paraoncomegas spp.) is comprised of physiological differences between sharks and rays have not pro- species that exhibit hollow hooks. Similarly, the symbolic coding vided a significant barrier to colonization. Such differences are of the armature pattern ‘‘homeoacanthous” is inconsistent with therefore unlikely to explain a dichotomous phylogenetic pattern the actual mapping of homeoacanthous armature presented on separating the Trypanoselachoida from the Trypanobatoida, and the tree itself. Nevertheless, some similarities were found. For thus this early diversification is best explained through coevolu- example, number of bothria, which was shown to be remarkably tion, whereas subsequent diversification events are more likely homoplasious by Palm et al. (2009), was similarly homoplasious to be explained by ecological factors. in our analysis, and features such as prebulbar organs, found to As with the morphological analyses, direct comparisons with represent a synapomorphy for the batoid-infecting clade was sim- the treatment of Palm et al. (2009) are difficult to make due to ilarly synapomorphous in our study. the differences in taxa considered and methodologies employed. For example, the families of definitive hosts considered by Palm 4.6. Host-parasite coevolution and the Hypnosqualea hypothesis et al. (2009) represent only a subset of the families of elasmobranchs known to host species in the trypanorhynch genera under That the phylogeny of the trypanorhynchs is characterized by a consideration in our analyses (see Palm, 2004); omitted were the basal split leading to two major lineages, the ancestral hosts of Etmopteridae, Hemiscylliidae, Heterodontidae (and its order Het- which were alternatively a carcharhiniform shark or a rajiform ba- erodontiformes), Hexanchidae, Parascylliidae, Scyliorhynidae, toid, we find to be compelling evidence in support of rejection of Somniosidae and Urotrygonidae. Moreover, in a number of in- the Hypnosqualea hypotheses, or the notion that rays represent a stances different suites of host families are presented for parasite lineage of derived sharks. This finding is congruent with the most genera replicated across the tree topology. Of perhaps greater sig- recent molecular phylogenetic studies of elasmobranchs as illus- nificance, is the fact that no attempt was made to determine ances- trated in Fig. 3 (i.e. Douady et al., 2003; Winchell et al., 2004; tral host states or ages of divergence, leaving conclusions to be Naylor et al., 2005; Human et al., 2006), and provides a source of made on the basis of the assumption that common equals primi- evidence derived independently from the hosts themselves. Be- tive, which as demonstrated above, can be misleading (e.g. ances- yond this basal diversification, however, the phylogeny of the par- tral hosts of the Gymnorhynchoidea). asites does not appear to be strongly influenced by the pattern of their host’s evolution (i.e. to the extent the latter is known). In 4.7. Divergence time estimates the diversification of the Trypanoselachoida, for example, it could be argued that a cladogenic event in the Middle Cretaceous led Without a fossil record, calibrating the phylogenetic tree of a to the evolution of a squaliform-hosted lineage, the Gymnorhyn- parasite group can only be done by assuming that they are at least choidea. However, ACSRs of the host groups predict the ancestral as old as their hosts, which not only leaves considerable room for host of the Gymnorhynchoidea to be a member of the Lamnifor- error but cannot be tested directly. In fact, simply estimating a date mes, despite the fact that squaliform sharks are the more common of divergence between sharks and rays is highly tenuous despite Author's personal copy

240 P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 their relatively extensive fossil record. The resulting uncertainty ten than not to show significant levels of congruence. This presum- associated with the divergence estimates makes support for most ably accounts for the fact that the field of host-parasite coevolution scenarios permissible and thus we simply cannot draw robust con- has generated greater advances in methodology than in the gener- clusions. From a relative standpoint, the divergence dates of the ation of empirical data: very few actual data sets have been found two major lineages of trypanorhynchs is at least consistent with that show clear coevolutionary patterns (see Page, 2003). Even the the fossil record of the hosts in that, according to Cappetta et al. ubiquitously used example, pocket gophers and their chewing lice (1993), the earliest evidence of the Rajiformes comes from the (Hafner and Page, 1995), has been shown to exhibit an unusually Toarcian (Lower Jurassic; 183 Myr) with the appearance of the high level of coevolution when chewing lice and their mammalian Rhinobatidae, whereas the Carcharhiniformes first occur slightly hosts are compared more broadly (Taylor and Purvis, 2003). later in the Bathonian (Upper Jurassic; 167 Myr) with the Although few people would argue that the process of coevolution appearance of the Scyliorhinidae, and not until the Turonian has been a non-existent or inconsequential part of parasite evolu- (Middle Cretaceous; 93.5 Myr) do we first see the appearance tion (especially among parasites having little independent means of carcharhiniform sharks that host contemporary trypanorhynchs of dispersal, such as chewing lice), finding empirical evidence has (i.e. Triakidae) (Cappetta et al., 1993). Diversification of the Tryp- proven exceedingly difficult in most groups. For these reasons, anoselachoida thus correlates well with the first appearance of the simple lack of congruence itself, as noted by Page (2003), does the Triakidae in the Middle Cretaceous (see Fig. 7), whereas diver- not necessarily signal lack of cospeciation. However, it obviously sification of the Trypanobatoida is estimated to have occurred provides no evidence for it. 40 Myr earlier around the Jurassic–Cretaceous boundary. The lat- Advancing the work presented here would benefit from addi- ter date falls significantly later than that the first appearance of the tional sampling of the trypanorhynch fauna, particularly where contemporary ray groups (even given a 60 Myr error bar), but nev- that would introduce new genera or host associations not repre- ertheless predates the diversification of the Trypanoselachoida, sented herein, but given our broad sampling of their diversity, will which is consistent with the host record in a relative sense. depend chiefly on obtaining a more comprehensive understanding Whereas the ancestral host group of the Trypanobatoida is of elasmobranch evolution. Ideally what is required is a compre- estimated to be a member of the Rajiformes with effectively no hensive and well calibrated tree incorporating multiple molecular support for alternative batoid orders (i.e. Pristiformes or Torpedin- divergence rate estimates as well as multiple calibration points iformes), that of the Trypanoselachoida is more equivocal, with the based on fossil evidence. Most such data are available today, except Carcharhiniformes having a likelihood of only 52%, the Lamnifor- with regard to the molecular characterization of elasmobranchs, in mes receiving 34%, the Squaliformes 10%, and the remaining which different genes are available for different subsets of taxa. A shark orders 1% each. With regard to the fossil record, this node concerted effort is thus needed to produce a more comprehensive corresponds equally well with first appearances of both the car- analysis of their evolution, which in turn will have utility for charhiniform (i.e. Triakidae) and lamniform (i.e. Alopiidae) shark understanding the evolution of the flatworm, arthropod, annelid, hosts of contemporary trypanorhynchs, albeit error associated nematode, ‘protozoan’ and other parasitic animals for which the with this estimate is great enough to encompass the first appear- great diversity of elasmobranchs play host (Caira, 1990). ance of the squaliform sharks (i.e. Squalidae; Upper Jurassic) as well. Thus while the Squaliformes is calculated to have a very Acknowledgements low likelihood of being the ancestral host group of the Trypanos- elachoida, and while its date of first appearance shows less We are grateful to the following individuals for assisting us correspondence with the estimated divergence of the Trypanosela- with the collection of hosts from which the trypanorhynchs exam- choida than with either the Carcharhiniformes or Lamniformes, the ined here came: Gui Mesendez of the University of Azores for spec- level of uncertainty associated with the analysis does not allow us imens from Horta, Annie Lim for specimens from Sarawak, to dismiss it. Malaysia, Mabel Manjaji-Matsumoto for specimens from Sabah, An obvious question concerning divergence rates is whether or Malaysia, Warren and Mildred Servatt for specimens from Florida not the disparity seen in the parasite phylogeny, in which the Try- Bay, Cheikh Tidiane Ba for assistance with collections off Senegal, panobatoida shows a markedly greater average divergence than Jean-Lou Justine for collections from New Caledonia, Rod Bray for that of the Trypanoselachoida (Figs. 4–7), is reflected in differing specimens from the North Sea, and Veronica Ivanov for specimens rates between sharks and rays. The question is difficult to address from Argentina. In addition, we thank Stephen Curran, Ash Bullard given our limited understanding of the host evolution. The most and Ronnie Palmer for their assistance in the collection and pro- comparable data set is that of Winchell et al. (2004) who also used cessing of specimens at the Gulf Coast Research Laboratory, Missis- a combination of lsrDNA and ssrDNA and thus the comparison is sippi; Bill and Ray Passey for allowing JNC and KJ to work on the FV not confounded by choice of genes. In their work the batoids do ap- Ocean Harvest off the Wessel Islands; and Nancy Kohler and Lisa pear to show a greater divergence than the majority of the shark Natanson of the National Marine Fisheries Service for facilitating lineages (their Fig. 2). A major exception, however, is the carchar- our collections from Montauk, Long Island. Special thanks to Tobias hiniform sharks that show even greater divergence than the rays, Hill for performing stability analyses for us using his unpublished whereas this disparity is not reflected in the parasite phylogeny. Python script; to Lukas Rüber for discussions concerning probabi- Moreover, studies based on mitochondrial genes (i.e. Douady listic methods of character state reconstruction and to Andrew et al., 2003; Human et al., 2006) do not show this disparity either Mead for fielding questions concerning the use of BayesTraits; to between sharks and rays or between the carcharhiniform and James Cotton and Julia Day for advice regarding phylogenetic dat- other sharks. Thus the lack of correspondence in the disparity of ing using BEAST; and to Peter Forey for assistance with the litera- rates in the parasite and host phylogenies remains an outstanding ture concerning the elasmobranch fossil record. JNC and KJ were question. supported in part by National Science Foundation (NSF) award Nos. DEB 0542846, DEB 0542941, DEB 0818696, DEB 0818823 4.8. Host-parasite coevolution and an NSF EPSCoR First Award to KJ. RMO was supported in part by the NSF (DEB 0529684) and through the National Oceano- Because cospeciation events become masked through time via graphic and Atmospheric Administration (NA08NOS4730322 and host switching, extinction and lineage sorting, comparisons of NA17FU2841). PDO was supported in part by a Marshall-Sherfield independently-derived host and parasite phylogenies fail more of- Fellowship from the Marshall-Aid Commemoration Commission Author's personal copy

P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242 241

(UK) and through a fellowship held by DTJ Littlewood from the Guiart, J., 1931. Considérations historiques sur la nomenclature et sur la Wellcome Trust (043965). classification des Tétrarhynques. Bulletin de l’Institut Océanographique de Monaco 575, 1–27. Hafner, M.S., Page, R.D.M., 1995. Molecular phylogenies and host-parasite cospeciation: gophers and pocket lice as a model system. Philos. Trans. R. Soc. Appendix A. Supplementary data Lond. 349 (B), 77–83. Healy, C.J., Caira, J.N., Jensen, K., Webster, B.L., Littlewood, D.T.J., 2009. Proposal for a new tapeworm order, Rhinebothriidea. Int. J. Parasitol. 39, 497–511. Supplementary data associated with this article can be found, in Heinz, H.J., 1954. A case of tetrarhynchid (cestode) infection in man. Rev. Ecuat. the online version, at doi:10.1016/j.ijpara.2009.07.012. Entomol. Parasitol. 2, 227–230. Ho, S., 2007. Calibrating molecular estimates of substitution rates and divergence times in birds. J. Avian Biol. 38, 409–414. References Hoberg, E., Mariaux, J., Justine, J.-L., Brooks, D.R., Weekes, P.J., 1997. Phylogeny of the orders of the Eucestoda (Cercomeromorphae) based on comparative morphology: historical perspectives and a new working hypothesis. J. Beveridge, I., 1990. Revision of the family Gilquiniidae Dollfus (Cestoda: Parasitol. 83, 1128–1147. Trypanorhyncha) from elasmobranch fishes. Aust. J. Zool. 37, 481–520. Huelsenbeck, J.P., Ronquist, F., Nielsen, R., Bollback, J.P., 2001. Bayesian inference of Beveridge, I., Campbell, R.A., 1989. Chimaerarhynchus n. g. and Patellobothrium n. g., phylogeny and its impact on evolutionary biology. Science 294, 2310–2314. two new genera of tryparnohynch cestodes with unique poeciloacanthous Human, B.A., Owen, E.P., Compagno, L.J.V., Harley, E.H., 2006. Testing armatures, and a reorganization of the poeciloacanthous trypanorhynch morphologically based phylogenetic theories within the cartilaginous fishes families. Syst. Parasitol. 14, 209–225. with molecular data, with special reference to the catshark family Beveridge, I., Campbell, R.A., 1996. New records and descriptions of trypanorhynch (Chondrichthyes: Scyliorhinidae) and the interrelationships within them. Mol. cestodes from Australian fishes. Rec. South Aust. Mus. 29, 1–22. Phylogenet. Evol. 39, 384–391. Beveridge, I., Campbell, R.A., 2005. Three new genera of trypanorhynch cestodes Khalil, L.F., Jones, A., Bray, R.A. (Eds.), 1994. Keys to the Cestode Parasites of from Australian elasmobranch fishes. Syst. Parasitol. 60, 211–224. Vertebrates. CAB International, Wallingford. Beveridge, I., Justine, J.L., 2006. Gilquiniid cestodes (Trypanorhyncha) from Kikuchi, Y., Takenouchi, T., Kamiya, M., Ozaki, H., 1981. Trypanorhynchiid cestode elasmobranch fishes off New Caledonia with descriptions of two new genera larva found on the human palatine tonsil. Jpn. J. Parasitol. 30, 497–499. and a new species. Syst. Parasitol. 65, 235–249. Kuchta, R., Scholz, T., Brabec, J., Bray, R.A., 2008. Suppression of the tapeworm order Beveridge, I., Campbell, R.A., Palm, H.W., 1999. Preliminary cladistic analysis of Pseudophyllidea (Platyhelminthes: Eucestoda) and the proposal of two new genera of the cestode order Trypanorhyncha Diesing, 1863. Syst. Parasitol. 42, orders, Bothriocephalidea and Diphyllobothriidea. Int. J. Parasitol. 38, 49–55. 29–49. Linton, E. 1889. Notes on Entozoa of marine fishes of New England, with Bigelow, H.W., Schroeder, W.C., 1948. Sharks. In: Tee-Van, J., Breder, C.M., descriptions of several new species. Ann. Rep. US Comm. Fish Fisheries for Hildebrand, S.F., Parr, A.E., Schroeder, W.C. (Eds.), Fishes of Western North 1886: 453–511. Pl I–VI. Atlantic. Part 1. Yale University, New Haven, pp. 59–576. Littlewood, D.T.J., Olson, P.D., 2001. Small subunit rDNA and the Platyhelminthes: Bigelow, H.W., Schroeder, W.C., 1953. Sawfishes, guitarfishes, skates and rays. In: signal, noise, conflict and compromise. In: Littlewood, D.T.J., Bray, R.A. (Eds.), Tee-Van, J., Breder, C.M., Hildebrand, S.F., Parr, A.E., Schroeder, W.C. (Eds.), Interrelationships of the Platyhelminthes. Taylor and Francis, London and New Fishes of Western North Atlantic. Part 1. Yale University, New Haven, pp. 1–514. York, pp. 262–278. Blainville, H.M.de., 1828. Article Vers. Dictionnaire des Sciences Naturelle. Tome LII, Maisey, J.G., Naylor, G.J.P., Ward, D.J., 2004. Mesozoic elasmobranchs, neoselachian Paris. pp. 365–625. phylogenetics and the rise of modern elasmobranch diversity. In: Arratia, G., Braun, M., 1894. Cestodes. In: Bronn, H.P. (Ed.), Klassen und Ordnungen des Tintori, A. (Eds.), Mesozoic Fishes 3—Systematics, Paleoenvironments and Thierreichs. Band 4: Vermes. Abteilung I. C. F. Winter’sche Verlagshandlung, Biodiversity. Verlag Dr. Friedrish Pfeil, München, pp. 17–56. Leipzig, pp. 927–1731. McGinnis, S., Madden, T.L., 2004. BLAST: at the core of a powerful and diverse set of Bray, R.A., Olson, P.D., 2004. The plerocercus of Ditrachybothridium macrocephalum sequence analysis tools. Nucleic Acids Res. 32, W20–W25. Rees, 1959 from two deep-sea elasmobranchs, with a molecular analysis of its Naylor, G.J.P., Ryburn, J.A., Fedrigo, O., López, A., 2005. Phylogenetic relationships position with the order Diphyllidea and a checklist of the hosts of larval among the major lineages of modern elasmobranchs. In: Hamlett, W.C. (Ed.), diphyllideans. Syst. Parasitol. 59, 159–167. Reproductive Biology and Phylogeny of Chondrichthyes: Sharks, Batoids and Bussieras, J., 1970. Nouvelles observations sur les Cestodes tétrarhynques de la Chimaeras. Science Publishers, Enfield, pp. 1–25. collection du Musée Océanographique de Monaco. Ann. Parasitol. Hum. Comp. Olson, P.D., Caira, J.N., 1999. Evolution of the major lineages of tapeworms 45, 5–12. (Platyhelminthes: Cestoidea) inferred from 18S ribosomal DNA and elongation Caira, J.N., 1990. Metazoan parasites as indicators of elasmobranch biology. In: factor-1a. J. Parasitol. 85, 1134–1159. Pratt, H.L., Gruber, S.H., Taniuchi, T. (Eds.), Elasmobranchs as living resources: Olson, P.D., Tkach, V.V., 2005. Advances and trends in the molecular systematics of advances in the biology, ecology, systematics, and the status of the fisheries. the parasitic Platyhelminthes. Adv. Parasitol. 60, 165–243. NOAA Tech. Rep. NMFS, Seattle, pp. 71–96. Olson, P.D., Littlewood, D.T.J., Bray, R.A., Mariaux, J., 2001. Interrelationships and Caira, J.N., Reyda, F., 2005. Marine eucestodes. In: Rohde, K. (Ed.), Marine evolution of the tapeworms (Platyhelminthes: Cestoda). Mol. Phylogenet. Evol. Parasitology. CSIRO Press, Canberra, pp. 92–104. 19, 443–467. Campbell, R.A., Beveridge, I., 1994. Order Trypanorhyncha Diesing, 1863. In: Khalil, Olson, P.D., Cribb, T.H., Tkach, V.V., Bray, R.A., Littlewood, D.T.J., 2003. Phylogeny L.F., Jones, A., Bray, R.A. (Eds.), Keys to the Cestode Parasites of Vertebrates. CAB and classification of the Digenea (Platyhelminthes: Trematoda). Int. J. Parasitol. International, Wallingford, pp. 51–148. 33, 733–755. Cappetta, H., Duffin, C., Zidek, J., 1993. Chondrichthyes. In: Benton, M.J. (Ed.), The Olson, P.D., Poddubnaya, L.G., Littlewood, D.T.J., Scholz, T., 2008. On the position of Fossil Record 2. Chapman and Hall, London, pp. 593–609. Archigetes and its bearing on the early evolution of the tapeworms. J. Parasitol. Charleston, M., 1998. Jungles: a new solution to the host/parasite phylogeny 94, 898–904. reconciliation problem. Math. Biosci. 149, 191–223. Overstreet, R.M., 2003. Presidential address: flavor buds and other delights. J. Compagno, L.J.V., 1973. Interrelationships of living elasmobranchs. Zool. J. Linn. Soc. Parasitol. 89, 1093–1107. 53, 15–61. Page, R.D.M., 1994. Parallel phylogenies: reconstructing the history of host-parasite de Chambrier, A., Zender, M., Vaucher, C., Mariaux, J., 2004. The evolution of the assemblages. Cladistics 10, 155–173. Proteocephalidea (Platyhelminthes, Eucestoda) based on an enlarged molecular Page, R.D.M. (Ed.), 2003. Tangled Trees: Phylogeny, Cospeciation and Coevolution. phylogeny, with comments on their uterine development. Syst. Parasitol. 57, The University of Chicago Press, Chicago. 159–171. Pagel, M., Meade, A., Barker, D., 2004. Bayesian estimation of ancestral character Diesing, K.M., 1850. Systema Helminthum, vol. I. Vindobonae, xvi. states on phylogenies. Syst. Biol. 53, 673–684. Dollfus, R.P., 1942. Études critiques sur les Tétrarhynques du Museum de Paris. Palm, H.W., 1997. An alternative classification of trypanorhynch cestodes Arch. Mus. Natl. Hist. Nat. Paris 19, 1–466. considering the tentacular armature as being of limited importance. Syst. Douady, C.J., Dosay, M., Shivji, M.S., Stanhope, M.J., 2003. Molecular phylogenetic Parasitol. 37, 81–92. evidence refuting the hypothesis of Batoidea (rays and skates) as derived Palm, H.W., 2004. The Trypanorhyncha Diesing, 1863. PKSPL-IPB Press, Bogor. sharks. Mol. Phylogenet. Evol. 26, 215–221. Palm, H.W., Caira, J.N., 2008. Host specificity of adult versus larval cestodes of the Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by elasmobranch tapeworm order Trypanorhyncha. Int. J. Parasitol. 38, 381– sampling trees. BMC Evol. Biol. 7, 214. 388. Drummond, A.J., Ho, S.Y.W., Phillips, M.J., Rambaut, A., 2006. Relaxed phylogenetics Palm, H.W., Waeschenbach, A., Littlewood, D.T.J., 2007. Genetic diversity in the and dating with confidence. PLoS Biol. 4, e88. trypanorhynch cestode Tentacularia coryphaenae Bosc, 1797: evidence for a Edgar, R.C., 2004. MUSCLE: multiple sequence alignment with high accuracy and cosmopolitan distribution and low host specificity in the teleost intermediate high throughput. Nucleic Acids Res. 32, 1792–1797. host. Parasitol. Res. 101, 153–159. Fripp, P.J., Mason, P.R., 1983. Spurious human infection with a trypanorhynchiid Palm, H.W., Waeschenbach, A., Olson, P.D., Littlewood, D.T.J., 2009. Molecular tapeworm. S. African J. Sci. 79, 473. phylogeny and evolution of the Trypanorhyncha Diesing, 1863 Fuhrmann, O., 1931. Dritte klasse des cladus Plathelminthes. Cestoidea. In: (Platyhelminthes: Cestoda). Mol. Phylogenet. Evol. 52, 351–367. Kükenthal, W., Krumbach, T. (Eds.), Handbuch der Zoologie. Walter de Pintner, T., 1913. Vorarbeiten zu einer monographie der tetrarhynchoideen. Gruyter and Co., Berlin, pp. 141–416. Sitzungs. Kaiserl. Akad. Wissensch. Wien. Math.-Naturwissen. Klasse 122, Guiart, J., 1927. Classification des Tétrarhynques. Assoc. Franc. Pour L’avanc. Sci. 50 171–253. ème session, Lyon, 1926, 397–401. Author's personal copy

242 P.D. Olson et al. / International Journal for Parasitology 40 (2010) 223–242

Pintner, T., 1931. Über fortgesetzte tetrarhynchenuntersuchungen. II. Anz. Taylor, J., Purvis, A., 2003. Have mammals and their chewing lice diversiﬁed in Akademie Wissensch. Wien. Math.-Naturwissen. Klasse 68, 141–142. parallel? In: Page, R.D.M. (Ed.), Tangled Trees: Phylogeny, Cospeciation and Pleijel, F., Jondelius, U., Norlinder, E., Nygren, A., Oxelman, B., Schander, C., Coevolution. The University of Chicago Press, Chicago, pp. 240–261. Sundberg, P., Thollesson, M., 2008. Phylogenies without roots? A plea for the Thorley, J.L., Wilkinson, M., 1999. Testing the phylogenetic stability of early use of vouchers in molecular phylogenetic studies. Mol. Phylogenet. Evol. 48, tetrapods. J. Theor. Biol. 200, 343–344. 369–371. Waeschenbach, A., Webster, B.L., Bray, R.A., Littlewood, D.T.J., 2007. Added Ronquist, F., Huelsenbeck, J., 2003. MRBAYES 3: Bayesian phylogenetic inference resolution among ordinal level relationships of tapeworms (Platyhelminthes: under mixed models. Bioinformatics 19, 1572–1574. Cestoda) with complete small and large subunit nuclear ribosomal RNA genes. Schmidt, G.D., 1986. Handbook of Tapeworm Identiﬁcation. CRC Press Inc., Boca Mol. Phylogenet. Evol. 45, 311–325. Raton. Wetherbee, B.M., Cortés, E., 2004. Food consumption and feeding habits. In: Carrier, Shirai, S., 1992. Squalean Phylogeny: A New Framework of ‘Squaloid’ Sharks and J.C., Musick, J.A., Heithaus, M.R. (Eds.), Biology of Sharks and their Relatives. CRC Related Taxa. Hokkaido University Press, Sapporo. Press, New York, pp. 225–246. Shirai, S., 1996. Phylogenetic interrelationships of Neoselachians (Chondrichthyes: Wilkinson, M., 2003. Missing data and multiple trees: stability and support. J. Vert. Euselachii). In: Stianssy, M.L.J., Parenti, L.R., Johnson, G.D. (Eds.), Inter- Paleo. 23, 311–323. relationships of Fishes. Academic Press, San Diego, pp. 9–34. Winchell, C.J., Martin, A.P., Mallatt, J., 2004. Phylogeny of elasmobranchs based on Southwell, T., 1929. A monograph on cestodes of the order Trypanorhyncha from LSU and SSU ribosomal RNA genes. Mol. Phylogenet. Evol. 31, 214–224. Ceylon and India. Part 1. Ceylon J. Sci. Sect. B. 15, 169–317.