INTRODUCTION
What an exciting time this is to be working on the taxonomy and systematics of parasites! Assuming that for any monophyletic group of parasites, the primary goals of these fields are to collect and describe all species, to fully explore their definitive host associations (as that is required to ensure all species have been found), and to understand their phylogenetic relationships, the unprecedented array of methods and resources now available make it actually possible to achieve these goals. The methods and resources that have conspired to make this a unique and particularly productive moment in time are as follows: (1) Global interest in biodiversity has energized global surveys. (2) Globalization has made it possible to address questions and to interact with colleagues on a worldwide scale. (3) Enhanced digital technologies and microscopic capabilities have expedited the discovery and assimilation of characters in an evolutionary context. (4) The advent of web-based resources has greatly facilitated the sharing and dissemination of taxonomic and systematic information. (5) Molecular methods have become relatively rapid and easy to employ, at a comparatively low cost. (6) Methods for the analysis of phylogenetic relationships have reached a relatively sophisticated level.
What follows is an attempt to illustrate the above points using cestodes, and in particular, the cestodes of elasmobranchs (i.e. sharks and rays). This is a parasite/host system with which I have worked for over two decades and thus can attest to the fact that it has benefited greatly from the above advancements. Of course, at the outset, it is important to recognize that our ability to take advantage of recent technologies and opportunities relies heavily on the literally hundreds of years of work of the dedicated cestode taxonomists and systematists who preceded us (e.g. working in the 19th and early 20th centuries: Diesing, Dollfus, Fuhrmann, Guiart, Leuckart, Linstow, Lönnberg, Monticelli, Olsson, Perrenoud, Rudolphi, Van Beneden, Woodland, Yamaguti, Zschokke), and who have provided the foundation for the substantial advancements in this host/parasite system that have been made over the last three or so decades.
ADVANCES
Global interest in biodiversity has energized global surveys
Global interest in biodiversity has provided opportunities for the discovery of new taxa; over the last two decades alone hundreds of new species of elasmobranch tapeworms have been described. This interest has also allowed us to circumscribe the full extent of the diversity of particular groups of parasites and their potential hosts. With respect to an inventory of the cestodes of elasmobranchs, an understanding of the full scope of this endeavour is now well within our grasp and, happily, the total predicted diversity is not as overwhelming as one might initially have thought. As a community we are swiftly moving towards developing a robust estimate of the total number of elasmobranch species on the planet. In fact, efforts to describe new elasmobranch taxa have intensified over the last decade, with more than 140 species being described since Compagno's global checklist of elasmobranchs was published in 2005 (see Naylor et al. unpublished), many, not unexpectedly, coming from biodiversity hotspots (e.g. Last et al. Reference Last, White and Pogonoski2008; 2010a). Even if the current estimate of ∼1,200 known elasmobranch species (see Naylor et al. unpublished) were to double as a result of the discovery and description of additional species across the globe (an unlikely but workable assumption), the total number of potential host species for the cestodes that parasitize elasmobranchs would be relatively manageable at ∼2,400.
At present, seven of the 18 recognized orders of cestodes parasitize elasmobranchs. These are the Cathetocephalidea, Diphyllidea, Lecanicephalidea, Litobothriidea, Rhinebothriidea, Tetraphyllidea and Trypanorhyncha. Many of these cestode groups are notoriously host specific. With the exception of the tetraphyllidean genus Chimaerocestus, which parasitizes holocephalans (see Williams and Bray, Reference Williams and Bray1984), all seven of these orders consist of species that exclusively parasitize elasmobranchs as adults and, in many cases, individual species parasitize only a single species of elasmobranch (see Caira, Reference Caira1990; Caira and Jensen, Reference Caira and Jensen2001; Caira and Healy, Reference Caira, Healy, Carrier, Musack and Heithaus2004). Thus, given what we know of elasmobranch diversity, a realistic estimate of the number of species in each of these seven orders is also within our grasp, as is an estimate of the total number of species of tapeworms parasitizing all elasmobranchs. It is important to recognize that these seven orders vary widely in the breadth of their host associations, and thus also their potential diversity. For example, all known species of the order Litobothriidea parasitize a subset of lamniform sharks (i.e. thresher and goblin sharks and their kin). In total, 14 species of lamniform sharks are currently recognized (Compagno, Reference Compagno, Fowler, Cavanagh, Camhi, Burgess, Cailliet, Fordham, Simpfendorfer and Musick2005). Some of these species have never been examined for cestodes; I have examined several, such as the mako, great white and common thresher sharks on multiple occasions and found them to lack litobothriideans; four species are known to be parasitized by one or even two members of this order (Dailey, Reference Dailey1969; Reference Dailey1971; Kurochkin and Slankis, Reference Kurochkin and Slankis1973; Caira and Runkle, Reference Caira and Runkle1993; Olson and Caira, Reference Olson and Caira2001). Thus, it is fair to say that the total number of species in the order Litobothriidea is unlikely to exceed 28, and if our negative data are correct, the number is probably substantially lower. On the other hand, species of individual tetraphyllidean genera, such as Phoreiobothrium, parasitize sharks of the more speciose families Carcharhinidae and Sphyrnidae (see Caira and Jensen, Reference Caira and Jensen2001; Caira et al. Reference Caira, Richmond and Swanson2005a). Current estimates suggest that, in combination, these families include at least 63 species of sharks (Compagno, Reference Compagno, Fowler, Cavanagh, Camhi, Burgess, Cailliet, Fordham, Simpfendorfer and Musick2005). Once again, a number of these shark species have never been examined for cestodes; some have been examined and found to lack Phoreiobothrium (e.g. Curran and Caira, Reference Curran and Caira1995); a number have been examined and found to be parasitized by one or even two members of this genus (see Caira et al. Reference Caira, Richmond and Swanson2005a and Linton, Reference Linton1924, respectively). Thus, it is reasonable to predict that the total number of species of Phoreibothrium is unlikely to exceed 126 species and, if our negative data are correct, the number is again probably somewhat lower. Of course the accuracy of such estimates is predicated on the quality, quantity and taxon spread of the subset of elasmobranchs that have been examined for cestodes to date. Such estimates also assume that the cestode fauna of a species of elasmobranch does not vary substantially across its distribution. However, I believe the sample, in terms of both elasmobranch taxon spread and geographic coverage (see http://tapewormdb.uconn.edu/hosts.php), is sufficient to support such estimates. In fact, we are now at a point at which robust predictions about the additional hosts of certain taxa, in particular genera of elasmobranch cestodes, can be generated (e.g. see Caira et al. Reference Caira, Jensen and Rajan2007; Pickering and Caira, Reference Pickering and Caira2008; Caira and Jensen, Reference Caira and Jensen2009; Desjardins and Caira, Reference Desjardins and Caira2011).
While an understanding of the scope of the challenge does not make the challenge easier to meet, a robust estimation of the proportion of the fauna that remains to be discovered allows the resources required to complete the inventory to be realistically evaluated, which is a fine start.
Globalization
Globalization has vastly improved communication and has greatly facilitated travel to distant and remote places. Global communities of investigators who focus on specific groups of parasites are emerging; in collaboration, these teams are advancing the taxonomy and systematics of their respective parasite groups. For example, in 1993, the first International Workshop on Cestode Taxonomy and Systematics was held at the Museum d'Histoire Naturelle de Genève, Geneva Switzerland. The goals of this meeting were to bring together Cestodologists from multiple countries and with expertise in a wide range of cestode groups, to discuss a variety of issues involving both old and new problems in the taxonomy and systematics of cestodes. Among other issues, the need for unification of terminology and the paucity of details available for comparative work were addressed. The results were a call for an improvement in the level of detail presented in alpha taxonomic work and also, given the vast potential of molecular data as a source of characters, a call for the collection of material suitable for the application of a range of study methods (see Mariaux and Vaucher, Reference Mariaux and Vaucher1994).
Five additional international workshops have been held since (see Caira et al. Reference Caira, Scholz and Georgiev2006 for a summary of four of these). Each of these gatherings resulted in one or more tangible products that helped consolidate and advance the field. The second workshop, held at the University of Nebraska-Lincoln in 1996, led to the generation of the first formal morphology-based phylogenetic hypotheses for many of the cestode orders (e.g. Rego et al. Reference Rego, de Chambrier, Hanzelová, Hoberg, Scholz, Weekes and Zehnder1998 for Proteocephalidea; Hoberg, Reference Hoberg1999 for cestodes overall; Ivanov and Hoberg, Reference Ivanov and Hoberg1999 for the Diphyllidea; Hoberg et al. Reference Hoberg, Jones and Bray1999 for the Cyclophyllidea; Beveridge, Reference Beveridge1999 for the Trypanorhyncha; Bray et al. Reference Bray, Jones and Hoberg1999 for the Pseudophyllidea; Caira et al. Reference Caira, Jensen and Healy1999 for the Tetraphyllidea and to some extent the Lecanicephalidea), providing important frameworks for future phylogenetic work. The third workshop was held at the Central Laboratory of General Ecology, at the Bulgarian Academy of Sciences, in Sofia, Bulgaria in 1999. Efforts there focused on standardizing terminology to reflect homology across cestode groups, with special emphasis placed on larval terminology. At that meeting the community of international experts agreed to present the consensus of their discussions in a series of papers, each published under the pseudonym Lenta Chervy (i.e. Bulgarian for “tapeworms”). The first in this series of papers (Chervy, Reference Chervy2002) treated larval terminology. In 2002, the fourth workshop was held at the University of Connecticut in Storrs, Connecticut. The Global Cestode Database (http://tapewormdb.uconn.edu) was formally launched (see below) and discussions of terminology focused on embryonic envelopes and the unique surface elaborations of cestodes, called microtriches. The discussion of microthrix terminology continued at the fifth workshop, held at the Institute of Parasitology, Academy of Sciences of the Czech Republic, České Budějovice, Czech Republic in 2005, as did discussion of embryonic envelopes. Recommendations for standardized terminology for embryonic envelopes were subsequently published by Conn and Swiderski (Reference Conn and Swiderski2008). Consensus on microthrix terminology was finally reached at the sixth international workshop, held in 2008 in Smolenice, Slovakia. These discussions culminated in a second paper by Lenta Chervy, presenting a unified terminology for microtriches (Chervy, Reference Chervy2009).
The community also now has a good sense of its strengths and weaknesses with respect to global collecting efforts. Fortunately, there has been a strong tradition of working on the taxonomy and systematics of the cestodes of elasmobranchs in a geographic framework. As a consequence, at least some data are available for many regions of the globe. For example, late in the 19th century and throughout the 20th century, substantial progress was made as a result of work conducted by Linton (e.g. Reference Linton1889, Reference Linton1890, Reference Linton1908, Reference Linton1924) and Campbell (e.g. Reference Campbell1969, Reference Campbell1970) in New England, off the eastern Atlantic seaboard, Shipley and Hornell (e.g. Reference Shipley, Hornell and Herdman1905, Reference Shipley, Hornell and Herdman1906), and Southwell (e.g. Reference Southwell1911, Reference Southwell1925) in Sri Lanka, Yamaguti (e.g. Reference Yamaguti1934, Reference Yamaguti1952) in Japan, Euzet (e.g. Reference Euzet1959) and Neifar (e.g. Reference Neifar, Tyler and Euzet2001, Reference Neifar, Euzet and Ben Hassine2002) and their collaborators in the Mediterranean Sea off France and Tunisia, Subhapradha (Reference Subhapradha1955) in India, Riser (Reference Riser1955) off the coast of California, Robinson (Reference Robinson1959a,Reference Robinsonb) and Alexander (Reference Alexander1963) in New Zealand, Williams (e.g. Reference Williams1968a, Reference Williamsb) in Great Britain, Carvajal and collaborators (Reference Carvajal1974, Reference Carvajal and Dailey1975) in Chile, beginning in the 1970s, Brooks and colleagues (e.g. Reference Brooks1977; et al. Reference Brooks, Mayes and Thorson1981), Marques (Reference Marques, Brooks and Monks1995, Reference Marques, Brooks and Lasso2001) and Ivanov (e.g. Reference Ivanov and Campbell2002, Reference Ivanov2006) and collaborators in Costa Rica and throughout South America including freshwater habitats, Beveridge (e.g. Reference Beveridge1990a, Reference Beveridgeb) and Butler (Reference Butler1987) in Australia, Wojciechowska (e.g. Reference Wojciechowska1991a, Reference Wojciechowskab) in Antarctica.
Whereas the trend in the past has been for investigators to attempt to describe their local faunas on their own, a more collaborative approach has recently been taken by a number of investigators. This has greatly facilitated progress towards the description of a wide array of taxa from specific regions. For example, in 1985 Beveridge undertook a survey of the cestodes of elasmobranchs of Australia. While he treated many of the trypanorhynchs (e.g. Beveridge Reference Beveridge1990a, Reference Beveridgeb), he also engaged collaborators to conduct work on that order (e.g. Campbell and Beveridge Reference Campbell and Beveridge1993, Reference Campbell and Beveridge2002; Beveridge and Campbell, Reference Beveridge and Campbell2005), and made specimens of other orders readily available to other experts (e.g. Caira and Ruhnke, Reference Caira and Ruhnke1990; Nasin et al. Reference Nasin, Caira and Euzet1997). Similarly, Justine, working in New Caledonia, provided opportunities for expert collaborators to survey a diversity of cestodes from elasmobranchs of that region (e.g. Caira et al. Reference Caira, Tracy and Euzet2004; Beveridge and Justine, Reference Beveridge and Justine2006; Reference Beveridge and Justine2007). Two projects funded by the National Science Foundation's Biotic Surveys and Inventories programme, which were collaborative by design, resulted in the dissemination of specimens and ultimately the description of a wide diversity of cestodes from elasmobranchs of the Gulf of California (e.g. Ruhnke et al. Reference Ruhnke, Curran and Holbert2000, 2006b; Ghoshroy and Caira, Reference Ghoshroy and Caira2001; Jensen, Reference Jensen2001; Olson and Caira, Reference Olson and Caira2001; Tyler, Reference Tyler2001; Healy, Reference Healy2003; Campbell and Beveridge, Reference Campbell and Beveridge2006) and Borneo (e.g. Caira et al. Reference Caira, Mega and Ruhnke2005b; Healy, Reference Healy2006; Jensen, Reference Jensen2006; Reyda and Caira, Reference Reyda and Caira2006; Beveridge, Reference Beveridge2008).
In combination, work to date has revealed that, while some data on the tapeworms of elasmobranchs are available for many geographic areas, major regions of the globe remain poorly sampled. These regions include, but are not limited to, much of the continent of Africa and much of the coast of South America, particularly along its eastern seaboard, as well as the northwestern Pacific Ocean coast. Given the rather limited/restricted distributions of many elasmobranchs, it follows that major gaps/deficiencies also exist in the elasmobranch taxa surveyed. For example, as has been noted previously (Caira and Jensen, Reference Caira and Jensen2001), given their substantial diversity, catsharks, squaliform sharks, round stingrays, skates, electric rays and guitarfish remain very poorly sampled for cestodes worldwide. However, with support from the National Science Foundation's Planetary Biodiversity Inventory programme, we have recently been able to begin to address these identified deficiencies and have already expanded knowledge of the cestodes of catsharks (e.g. Caira et al., Reference Caira, Kuchta and Desjardins2010) squaliform sharks (e.g. Noever et al. Reference Noever, Caira, Kuchta and Desjardins2010), skates (e.g. Rodriguez et al. in press), electric rays (Jensen et al. in press), and guitarfish (e.g. Fyler and Caira, Reference Fyler and Caira2010; Ivanov and Caira, Reference Ivanov and Cairain press), from localities including southern Africa and the northwestern Pacific Ocean coast. The fact that the community of cestodologists was able to identify specific gaps in our global sampling for all major groups of vertebrates known to host cestodes is likely to have contributed to our ability to make a convincing case for the need for this “Survey of Cestodes from the Vertebrate Bowels of the Earth.”
Enhanced microscopical capabilities and digital technologies
Parasites, by definition, are smaller than their hosts; many are tiny indeed. As a consequence, their ‘simplicity’ is sometimes overestimated because many of their key features are ultramicroscopic. Over the past two decades, scanning electron microscopy (SEM) has been greatly simplified by the more automated designs of the newer instruments and it has become more widely available. It has been used to great effect in the taxonomy and systematics of elasmobranch cestodes, often providing insights into the recognition of species and even genera. For example, although it is clear that microthrix patterns are relatively invariable among species in some genera of elasmobranch cestodes (e.g. among species of Acanthobothrium see Fyler and Caira, Reference Fyler and Caira2006; Reyda and Caira, Reference Reyda and Caira2006), microthrix patterns have been used to great taxonomic effect in other groups. For example, the diphyllideans (see Tyler, Reference Tyler2006), lecanicephalideans (see Jensen, Reference Jensen2005), and an array of tetraphyllidean genera such as Paraorygmatobothrium (see Ruhnke et al. Reference Ruhnke, Healy and Shapero2006a; Ruhnke and Carpenter, Reference Ruhnke and Carpenter2008) and Orectolobicestus (see Ruhnke et al. Reference Ruhnke, Caira and Carpenter2006b), have all been shown to exhibit microthrix patterns that are interspecifically variable. In addition, the advent of new methods for processing specimens, such as air-drying in hexamethyldisilizane rather than critical-point drying, have greatly expedited the preparation of specimens for SEM.
The fact that images can now be captured and processed digitally has not only accelerated the rate at which morphological data can be generated, but it has significantly reduced the cost of generating images as dark-room methods and materials are no longer required. This is true for both light and electron microscopy. Furthermore, the emergence of digital technologies and enhanced storage capabilities has greatly facilitated the process of describing species. In working to identify new species, it is now feasible to generate images of every specimen under consideration for the determination of type series. For the preparation of descriptions of new species, measurements can now be generated using image analysis software. Images of multiple taxa can be examined simultaneously for the purposes of generating differential diagnoses. Furthermore, digital technologies have made it possible and practicable to thoroughly document host identities while in the field.
Advent of web-based resources has greatly facilitated the sharing and dissemination of taxonomic and systematic information
Web-based resources designed to facilitate the broad dissemination of taxonomic and systematic data are now available for a variety of taxa, and they are continuing to expand rapidly. This is key given that these data are essential for establishing the identity of species (and higher taxa) with certainty. Unfortunately, historically there has been a trend to apply the readily accessible taxon names of cestodes, collected from one host species and/or geographic region, to cestode specimens from all closely related host species in different regions. For example, for decades Euzet's (Reference Euzet1959) monograph on the Tetraphyllidea of the elasmobranchs from the Mediterranean Sea stood as the primary synthetic work on that order. As a consequence, species treated therein were often reported by other authors from congeneric host species in distant geographic localities. For example, in their study on the ecology of the cestode fauna of the shark Mustelus canis off the New England coast, Cislo and Caira (Reference Cislo and Caira1993) originally identified a small species included in that fauna as Calliobothrium lintoni, a species treated in detail by Euzet (Reference Euzet1959) from hosts identified as the dusky smooth-hound shark, M. canis, in the Mediterranean. It was not until several years later that it was recognized that, not only is M. canis absent from the Mediterranean Sea, but the small species of Calliobothrium it hosts off New England actually represented a species new to science (see Nasin et al. Reference Nasin, Caira and Euzet1997). Similarly, Alexander (Reference Alexander1963) working with a related shark, Mustelus lenticulatus, off New Zealand, and Yamaguchi et al. (Reference Yamaguchi, Yokoyama, Ogawa and Taniuchi2003) working with Mustelus manazo off Japan, both identified cestodes from their respective sharks as C. verticillatum of Euzet (Reference Euzet1959), despite the fact that Euzet's specimens were collected from sharks identified as M. canis in the Mediterranean Sea. As noted by Pickering and Caira (Reference Pickering and Caira2008), while the specimens from these three hosts (and localities) resemble one another in a number of respects, it is likely they are not, in fact, conspecific.
In an effort to avert such erroneous identifications and the underestimation of diversity and host specificity they foster, the Global Cestode Database (http://tapewormdb.uconn.edu) was developed in collaboration with expert cestodologists from around the world. Although the database began with treatments of the cestode groups that parasitize elasmobranchs, it has more recently expanded to encompass all cestode groups. Although not yet complete, the primary goal of the database is to provide access to taxonomic information presented in the original descriptions of all cestode species and genera. The assembling of this taxonomic information into a single resource that is widely available via Internet, has already had an impact on cestode taxonomy in countries around the world, such as Bulgaria (e.g. Vasileva et al. Reference Vasileva, Dimitrov and Georgiev2002) and Argentina (e.g. Ivanov, Reference Ivanov2009). This resource not only facilitates the articulation of morphological characters for examination in a phylogenetic context, but it also greatly simplifies the tracking of global cestode diversity.
Web-based access to host information has also greatly enhanced the accuracy of host identifications and has facilitated collaborative work on entire parasite faunas of individual hosts and/or species. For example, the elasmobranch host specimen database (http://tapewormdb.uconn.edu/hosts.php) associated with the Global Cestode Database provides detailed specimen and locality data and in many cases images of ∼6,000 elasmobanch specimens from which we have collected cestodes over the past two and a half decades. The information provided can be easily and rapidly updated as, for example, molecular data and expert advice on host identities become available (see below). In the cases of potentially undescribed host species, it also provides a stable location for the storage of information and images of potentially undescribed host taxa (e.g. Fyler et al. Reference Fyler, Caira and Jensen2009; Kuchta and Caira, Reference Kuchta and Caira2010; Ivanov and Caira, Reference Ivanov and Cairain press).
Molecular methods
Recent advances in molecular methods now allow for the generation of large amounts of sequence data which, in the context of morphological work, can inform the exploration of species boundaries, and can be applied to questions of phylogenetic relationships. Such data are also proving to be of enormous importance with respect to the verification and/or determination of accurate host identities. Examples of each are provided below.
Molecules and parasite species boundaries
Tapeworms of elasmobranchs, and in particular, tetraphyllidean, lecanicephalidean and rhinebothriidean species exhibit a notoriously high degree of host specificity (see Caira, Reference Caira1990; Caira and Jensen, Reference Caira and Jensen2001; Caira and Healy, Reference Caira, Healy, Carrier, Musack and Heithaus2004). For example, each species of the hooked tetraphyllidean genus Acanthobothrium parasitizes a single species of batoid (e.g. see Ghoshroy and Caira, Reference Ghoshroy and Caira2001; Fyler and Caira, Reference Fyler and Caira2006); species of the non-hooked tetraphyllidean genus Paraorygmatobothrium each parasitize a species of carcharhiniform shark (e.g. see Ruhnke and Thompson, Reference Ruhnke and Thompson2006; Ruhnke et al. Reference Ruhnke, Healy and Shapero2006a; Ruhnke and Carpenter, Reference Ruhnke and Carpenter2008). In each case, congeners differ subtly in morphology. Hundreds of years of work on various morphological aspects of cestodes have resulted in the identification of suites of characters that vary interspecifically and thus are considered to be important for the characterization of species. This has involved sorting such characters from those that are less informative because they are prone to fixation artifacts (given that worms are soft-bodied, pressure applied when mounting can result in wider and/or longer worms) and/or those that exhibit some degree of intraspecific variation (e.g. testis number). What if species are phenotypically plastic and the morphological differences seen among congeners parasitizing different host species are actually the result of differences in habitat offered by different host species? Or, to put it another way, what if this perceived host specificity is actually the result of host-induced variation? If this were shown to be the case, a complete re-assessment of the taxonomic framework of this system would be in order. Until recently, this question has gone largely unanswered.
Fortunately, the application of molecular methods has done much to put this concern to rest. Sequence data from a variety of elasmobranch cestode taxa have now been generated for several molecular markers. In particular, the nuclear genes 28S rRNA and 18S rRNA and the mitochondrial gene 16S rRNA have been explored in some detail in elasmobranch cestodes (e.g. Olson et al. Reference Olson, Ruhnke, Sanney and Hudson1999, Reference Olson, Littlewood, Bray and Mariaux2001, Reference Olson, Caira, Jensen, Overstreet, Palm and Beveridge2010; Waeschenbach et al. Reference Waeschenbach, Webster, Bray and Littlewood2007; Fyler, Reference Fyler2009; Healy et al. Reference Healy, Caira, Jensen, Webster and Littlewood2009; Palm et al. Reference Palm, Waeschenbach, Olson and Littlewood2009). Data generated have fully supported existing morphological species concepts. For example, Fyler et al. (Reference Fyler and Caira2009) and Fyler and Caira (Reference Fyler and Caira2010) generated partial 28S rRNA data for a total of 13 species of the tetraphyllidean genus Acanthobothrium and found that, whereas replicate samples of each “morphospecies” were essentially identical in sequence, all 13 species were molecularly distinct. None of the 31 species for which partial 18S rRNA and 28S rRNA sequence data were generated for two or more congeners (representing two genera of tetraphyllideans and three genera of rhinebothriideans) by Healy et al. (Reference Healy, Caira, Jensen, Webster and Littlewood2009) was identical in sequence. Similarly, in none of the cases of the 11 trypanorhynch genera for which partial 18S rRNA and 28S rRNA data were generated for two or more species by Olson et al. (Reference Olson, Caira, Jensen, Overstreet, Palm and Beveridge2010) were congeners found to be identical in sequence. Although this issue was not among the primary goals of any of these works, the results serve to support the current identities of elasmobranch tapeworm species, and thus the morphological species concept that has historically been applied in this system. Its precise correspondence with molecular variation suggests that the morphological variation seen among tapeworms inhabiting different hosts is not merely a phenotypic response to the different environments offered by different hosts.
Molecules and host identifications
Global work on elasmobranch diversity has raised questions about some of our more long-standing host identifications, possibly seriously jeopardizing our overall understanding of this parasite/host system. For example, as noted by Fyler and Caira (Reference Fyler and Caira2010), between the first (1994) and second (2009) editions of The Sharks and Rays of Australia by Last and Stevens, the number of elasmobranchs known from Australia rose from 190 to 307 species. Perhaps of even more concern is the fact that certain host species (e.g. Rhynchobatus djiddensis) originally reported from that country in 1994, and from which cestodes have been described (e.g. Acanthobothrium bartonae; see Campbell and Beveridge, Reference Campbell and Beveridge2002), by 2009 were no longer considered to occur in Australia. Thus, the great inroads that have been made in the taxonomy and description of new elasmobranch taxa over the past 10 years has exacerbated issues with existing host identifications in this system. In many cases, it is a matter of the distribution of elasmobranch species being shown to be more limited than previously thought and thus, species once considered to be relatively broadly distributed have been determined to actually represent complexes of species (e.g. Last and Stevens, Reference Last and Stevens2009; White et al. Reference White, Last, Naylor, Jensen, Caira, Last, White and Pogonoski2010a). In such instances, the identities of the hosts of the associated cestodes become suspect and often require revision. In other cases, for example in regions that have previously been poorly sampled, sharks and rays completely new to science are being discovered (e.g. Last et al. Reference Last, White and Pogonoski2010a); it is important that parasitologists become aware of such novelty.
During the period between the discovery and description of new host species, in an effort to stabilize the identities of hosts of cestodes described from such taxa, sequence data can be of great value. This was, in fact, the primary motivation for K. Jensen and me to establish a collaborative project with G. Naylor (Naylor et al. unpublished). This project aimed to generate ∼1,044 base pairs of the elasmobranch barcoding gene, NADH dehydrogenase subunit 2, for as wide a diversity of elasmobanchs as possible both to aid in the recognition of potential host novelty as well as to provide a means of grounding field identifications. Thus, for example, based on the results of Naylor et al. (unpublished), Fyler et al. (Reference Fyler, Caira and Jensen2009), Kuchta and Caira (Reference Kuchta and Caira2010) and Ivanov and Caira (unpublished) were able to establish the identities of the potentially novel species of hosts from which their newly described cestodes were collected.
An interesting byproduct of this parasitological endeavour was that the host data generated by Naylor et al. (in review) have been used to inform some of the taxonomic decisions of elasmobranch experts both in the description of novel elasmobranch taxa (e.g. Compagno et al. Reference Compagno, White, Cavanagh, Last, White and Pogonoski2010; Last et al. Reference Last, Fahmi, Naylor, Last, White and Pogonoski2010b), and also in the resurrection and/or revision of existing elasmobranch taxa (e.g. White et al. Reference White, Last, Naylor, Jensen, Caira, Last, White and Pogonoski2010a,Reference White, Last, Naylor, Last, White and Pogonoskib,Reference White, Last, Naylor, Harris, Last, White and Pogonoskic).
Molecules and phylogenetic relationships
Without question, our understanding of the phylogenetic relationships of the cestodes of elasmobranchs is much richer for the application of molecular methods. Such methods have, for example, supported the somewhat controversial erection of the orders Cathetocephalidea by Schmidt and Beveridge (Reference Schmidt and Beveridge1990) and Litobothriidea by Dailey (Reference Dailey1969), which were each originally based on the unusual morphologies of their included taxa (see Caira et al. Reference Caira, Mega and Ruhnke2005b). Similarly, the non-monophyly of the order Tetraphyllidea, a hypothesis that became plausible following formal phylogenetic analyses of morphological characters (e.g. Brooks et al. Reference Brooks, Hoberg and Weekes1991; Brooks and McLennan, Reference Brooks and McLennan1993; Hoberg et al. Reference Hoberg, Mariaux, Justine, Brooks and Weekes1997; Caira et al. Reference Caira, Jensen and Healy1999, Reference Caira, Jensen, Healy, Littlewood and Bray2001), has been supported by all molecular studies that have included multiple representatives of this order, regardless of the markers or methods of analysis employed (e.g. Mariaux, Reference Mariaux1998; Olson and Caira, Reference Olson and Caira1999; Olson et al. Reference Olson, Ruhnke, Sanney and Hudson1999, Reference Olson, Littlewood, Bray and Mariaux2001; Waeschenbach et al. Reference Waeschenbach, Webster, Bray and Littlewood2007). Analysis of molecular data, in combination with an examination of selected morphological features, resulted in the erection of the new order Rhinebothriidea by Healy et al. (Reference Healy, Caira, Jensen, Webster and Littlewood2009), an action that alleviated some of the issues with the non-monophyly of the Tetraphyllidea. More recently, investigation of a combination of molecular and morphological data resulted in the most comprehensive phylogenetic analyses of the order Trypanorhyncha conducted to date (see Palm et al. Reference Palm, Waeschenbach, Olson and Littlewood2009; Olson et al. Reference Olson, Caira, Jensen, Overstreet, Palm and Beveridge2010). Although the results of these works were generally consistent with one another, they differed in a number of respects from the morphological tree of Campbell and Beveridge (Reference Campbell, Beveridge, Khalil, Jones and Bray1994). However, they included a number of elements that were consistent with the morphology-based phylogenies for this order presented by Palm (Reference Palm2004). In addition, Olson et al. (Reference Olson, Caira, Jensen, Overstreet, Palm and Beveridge2010) proposed the formal bisection of the order Trypanorhyncha into the monophyletic suborders Trypanobatoida and Trypanoselachoida. Not only are these subgroups supported by molecular data, but, as their names imply, their members are generally associated with batoids (stingrays and their relatives) in the former case and with sharks in the latter.
At this point, the phylogenetic relationships of the cestodes of elasmobranchs are beginning to come into focus. It will be interesting to see how molecular data will serve to illuminate the major questions that persist in this system. The Tetraphyllidea remain non-monophyletic, despite removal of the rhinebothriideans, and the phylogenetic relationships of the lecanicephalideans have yet to be explored in the context of molecular and/or morphological data. Furthermore, the morphology-based hypothesis of the interrelationships among the diphyllideans of Ivanov and Hoberg (Reference Ivanov and Hoberg1999) and Tyler (Reference Tyler2006) remain to be evaluated in a molecular framework. Nonetheless, it seems clear that these and other relationships will be best understood using an iterative process of combining morphological and molecular data, such as has so effectively served to inform our understanding of cestode taxa to date.
Methods of phylogenetic analysis
Our ability to generate enormous amounts of data (molecular or otherwise) for phylogenetic analyses would be of limited value if methods for the analysis of such large amounts of data were not available. The last decade has seen unprecedented progress in the development of methods that not only make it feasible to analyze large quantities of data, but that also incorporate elements (e.g. models, assumptions) that make it possible to accommodate the various biases inherent in certain types of data. The spectrum of available resources include programs for conducting parsimony, maximum-likelihood and Bayesian analyses; many are available free of charge. Although these new resources are certainly not unique to host/parasite work, they have played a key role in advancing modern parasite taxonomy and systematics and, in fact, most of the phylogenetic studies cited above employed one or more of these methods. For a sampling of the wide array of available resources see http://evolution.genetics.washington.edu/phylip/software.html.
THREATS TO REALIZATION OF THE POTENTIAL OFFERED BY THESE ADVANCES
It is somewhat ironic that, as efforts by taxonomists to develop realistic estimates of the diversity of taxa on the planet have intensified, so too has the pressure to complete the task of describing all of these taxa. In fact, there is a growing sense of impatience with taxonomists. This has led to an appeal for the development of a scheme to expedite the rate at which species are discovered and described which, not unexpectedly, has prompted a move towards developing methods to molecularly characterize species. To some, the notion of ‘barcoding’ ideally fits this perceived need (Hebert et al. Reference Hebert, Cywinska, Ball and deWaard2003; Jansen, Reference Jansen2004). As has been noted by a number of authors (e.g. Moritz and Cicero, Reference Moritz and Cicero2004; DeSalle et al. Reference DeSalle, Egan and Siddall2005; Lefébure et al. Reference Lefébure, Douady, Gouy and Gibert2006) it is important to recognize the distinction between DNA barcoding and DNA taxonomy. In the former case, in which barcode data are generated for described species or even as part of the process of describing species, the method is employed as a diagnostic tool to aid in the identification of known species. When done in a collaborative manner, with appropriate vouchering, and involvement of taxonomic experts, such endeavours can serve the needs of taxonomy to great effect and take advantage of the centuries of effort that have been dedicated to developing a viable taxonomic system and body of biological knowledge for each target taxon. As noted by Ebach and Holdredge (2005), molecular data are meaningful only when they are placed in a rich biological context. As described above, great strides have been made in the taxonomy and systematics of elasmobranchs and their tapeworms based on collaborations that have fostered the integration of morphological and molecular data.
DNA taxonomy, however, is an entirely different undertaking for it attempts to discover and characterize species based on molecular data alone. Clearly, it has both substantial intellectual (e.g. Lipscomb et al. Reference Lipscomb, Platnick and Wheeler2003; Ebach and de Carvalho, Reference Ebach and deCarvalho2010) and methodological (e.g. Mortiz and Cicero, Reference Moritz and Cicero2004; Lohse, Reference Lohse2009; Goldstein and DeSalle, Reference Goldstein and DeSalle2011) limitations. While this strategy is, by definition, molecular, on its own it simply does not qualify as taxonomy. For, even if the issues surrounding the determination of species boundaries using molecular criteria alone are overlooked for a moment, while this strategy may contribute to the question “How many?”, except in an impracticably superficial way, it is ineffective at addressing the question “Which ones?”
As an example, it is interesting to contemplate exactly how our discovery efforts would proceed were the molecular characterization of new species to become the norm in our system. To set the stage, the delicate nature of the bodies of tapeworms, coupled with the notoriously poor condition of unmounted museum material, mean that the discovery of new taxa generally requires the collection of new material. Much of the novelty discovered to date has come from fieldwork conducted on trawling vessels. On such vessels trawling gear is set, towed for 30–60 min and then hauled on board the vessel where the bag containing the catch is emptied on the deck. This process is then repeated 50–150 times, depending on the number of days at sea. Each catch often consists of literally tons of teleosts and a by-catch consisting of hundreds of specimens of elasmobranchs. It is not uncommon when working in poorly sampled areas for these catches to yield tens to hundreds of novel species of tapeworms across the numerous host species examined, and also one or more novel species of elasmobranch. Our current protocol is as follows: prior to each trip we consult with elasmobranch taxonomists and taxonomic keys to elasmobranchs for the region (or a region nearby if one is not available for the target region) to gain a basic understanding of potential hosts in the area. While on the vessel we sort a subsample of elasmobranch specimens from each haul by eye with the goal of examining tapeworms from as many species of elasmobranchs as possible over the course of the trip. Each host specimen chosen for necropsy is assigned a unique specimen number and a small tissue sample and a series of photographs are taken; if the host specimen looks suspiciously unlike any known from the region (i.e. if novelty is suspected from visual inspection), it may be retained. The spiral intestine of each host is then opened and it and its contents are examined under a dissecting microscope. The tens, hundreds or thousands (in the case of stingrays which host a particularly complex cestode fauna) of tapeworms found are sorted to order or genus if time (or weather!) allows. Each set of specimens of the taxa found is subdivided into two lots, one that is preserved in buffered formalin for morphological work and one that is preserved in ethanol for molecular work. In some cases, suspected novel taxa are already apparent at this point and are set aside for priority examination. Regardless, upon returning to the lab, the specimens fixed in formalin are distributed to taxonomic experts working on each order and ultimately prepared for study with light and scanning electron microscopy. The specimens fixed for molecular work are stored to be used as required.
In contrast, were the discovery of novelty to be placed in the hands of the ‘molecular technicians’ envisioned by some, I can only imagine the following scenario. To be fair to those advancing/promoting the notion of molecular taxonomy, I will assume that the molecular technician is equipped with the infamous ‘tricorder’ (e.g. Savolainen et al. Reference Savolainen, Cowan, Vogler, Roderick and Lan2005) and thus has access to some type of information about the taxa involved. Thus, on a trawler, working in a novel region, the technician, tricorder in hand, would be faced with a catch consisting of literally tons of animals that includes hundreds of elasmobranchs. Unfamiliar with elasmobranch morphology (but hopefully able to distinguish an elasmobranch from a teleost) he/she would focus his/her tricorder-poking on the elasmobranch specimens. However, in the absence of morphological training he/she would be faced with ‘poking’ all of the elasmobranch specimens. Presumably the tricorder would come pre-loaded with sequence data characterizing all known elasmobranch species. When the result of a poke failed to match existing sequence data, the technician would assume he/she had discovered a new species of elasmobranch and presumably would assign it a new ‘name’. It is unclear how he/she would then proceed, because the issue of vouchering is curiously missing/absent from most discussions of molecular taxonomy. And now for the worms – there is no getting around the fact that the technician would actually need to dissect the elasmobranchs. In each case, once the technician had opened the spiral intestine, unfamiliar with tapeworm morphology (but hopefully able to distinguish a tapeworm from a nematode), he/she would focus his/her efforts on the tapeworms. Given that individual hosts may be infected with tens, hundreds or thousands of worms, this would involve rather a lot of poking, this time perhaps with his/her ‘micro-tricorder’ for many of the tapeworms of elasmobranchs are tiny. Again, assuming that the micro-tricorder had come pre-loaded with sequence data for all known species of elasmobranch tapeworms, specimens for which pokes failed to result in matches would be assumed to represent new species and new ‘names’ would be assigned by the technician. Again, I am unclear as to how/if vouchering of the ‘new’ tapeworms would be handled. In the end, what would the technician actually have achieved? Well, future tricorder-bearing technicians were they to poke any of these ‘new taxa’ again, would know its technician-assigned name.
Admittedly, the above scenario was developed in the absence of enlightenment from advocates of DNA taxonomy as to how they would envision the process of discovering novelty in this field situation might proceed without the involvement of taxonomic experts. It serves, however, to illustrate that considerations of the practical applications of DNA taxonomy should extend well beyond discussions of gene choice, etc. (e.g. Casiraghi et al. Reference Casiraghi, Labra, Ferri, Galimberti, De Mattia, Nimis and Vignes2010); it also illustrates how potentially unrealistic and inefficient the generation of molecular characterizations of species in the absence of morphological observations and outside of traditional taxonomic frameworks is likely to be. Morphology-based delimitations of taxa allow triage in the field. In this system, quick, informed visual inspection facilitates the rapid sorting of tons of hosts to allow detailed efforts to focus on potentially novel specimens; similarly it allows the rapid sorting of hundreds of tapeworm specimens (even in the short time between hauls). It is important to note that all examples of DNA taxonomy that have appeared in the literature have come from studies done in the context of decades or even centuries of morphological characterizations. For example, Hebert et al.'s (2004) infamous discovery of cryptic skipper butterfly species using barcoding data was conducted in a framework of 25 years of natural history observations. Were species to be characterized solely using molecular data, rapid visual triaging would no longer be possible. Instead, sorting would require poking (and sequencing) of all specimens. The greater the number of species characterized solely on molecular grounds, the greater the amount of poking that would be required.
TAXONOMY: IT IS NOT FOR THE FEINT OF HEART
In reality, whether it is acknowledged by practitioners or not, successful DNA barcoding and DNA taxonomic efforts (e.g. the Moorea barcode project http://www.mooreabiocode.org/about; Ferri et al. Reference Ferri, Barbuto, Bain, Galimbert, Uni, Guerrero, Ferté, Bandi, Martin and Casiraghi2009) ultimately intimately rely on traditionally trained taxonomists, and specifically those trained in the morphological recognition of taxa. This brings us back to the need to culture and support traditional taxonomy and the training of its practitioners. Unfortunately, the morphological work required to establish the unique identities of species is extremely challenging and time-consuming and is most certainly not for the feint of heart! As a consequence, fewer and fewer students are choosing to negotiate this formidable (but rewarding) path rather than the more immediately satisfying (albeit less informative) path involving purely molecular methods. But, as so effectively articulated by Goldstein and DeSalle (2011), if the full potential of the taxonomic opportunities provided by our current circumstances is to be realized, an integrated approach combining traditional taxonomic work and molecular data is required. The notion that peer pressure is driving the insistence of morphology (over molecules) as the primary data source for the description of new species (see Cook et al. Reference Cook, Edwards, Crisp and Hardy2010) is absolutely correct. For these ‘peers’ are the expert experienced taxonomists who are striving to prevent the destruction of the infrastructure that is actually required to conduct taxonomy in the real world.
Although barcoding endeavours do, in fact, rely heavily on traditional taxonomy, it is unfortunate that attempts to garner funding for barcoding have failed to embrace the need for an integrated approach. Despite statements to the contrary (e.g. Gregory, Reference Gregory2005) such efforts are, as noted by many authors previously (e.g. deCarvalho and Bockmann, Reference deCarvalho and Bockmann2008) either explicitly or implicitly, jeopardizing the development of the morphological framework required for expediting the discovery and characterization of our planet's biodiversity in a practicable fashion. Furthermore, should future discovery and characterization of species proceed in a morphological vacuum as it would in the hands of DNA taxonomists, and thus outside of the context of traditional taxonomy, many of the questions of interest to taxonomists, systematists and phylogeneticists, such as those exploring adaptations to parasitism, morphological character evolution, the intrigue of life-cycles, intermediate host associations and co-evolution (or lack thereof) will be difficult if impossible to address in taxa other than the very small proportion of the planet's diversity described to date. Taxa circumscribed purely on a molecular basis are so uni-dimensional as to be almost uninteresting in a biological context. They certainly fall short of the goal of conceptualizing a species as articulated, for example, over a half a century ago by Simpson (Reference Simpson1961). It is disturbing that there are those who would so readily yield quality for quantity, in advocating for DNA taxonomy over traditional taxonomic treatments.
Fortunately, the concept of DNA taxonomy has not been embraced by those working on the tapeworms of elasmobranchs (or their hosts). Instead, there is great enthusiasm for an integrated approach, and work using morphology and molecules has progressed in concert. As noted above, great inroads have been made with most orders of elasmobranch tapeworms using a combination of these methods in the context of the traditional taxonomic framework. Furthermore, as a result in part from funding from NSF's PEET (Partnerships for Enhancing Expertise in Taxonomy) programme, many young individuals with taxonomic expertise in elasmobranch tapeworms (each of whom has experience working both with morphology and molecules) are now actively working in this area (e.g. K. Jensen, C.J. Healy, F.B. Reyda, C.A. Fyler, M. Pickering).
EXTRAPOLATING TO OTHER HOST/PARASITE SYSTEMS
While it may appear that the situation with elasmobranch tapeworms is unique, this is most assuredly not the case. Clearly, co-operation on a global scale is key if the potential of synergistic activities is to be fully realized. Fortunately, global communities of investigators whose research focuses on other parasite taxa have begun to assemble. For example, the 6th International Symposium on Monogenea was held in Cape Town, South Africa in 2009; the 7th International workshop on the Acanthocephala was held in Maine, USA in 2010 (Richardson, 2010); digenean workers have begun to organize on an international scale and the first workshop on these parasites is likely to be held in the relatively near future. Workshops on various aspects of nematode identification have been the tradition among nematologists for over a decade. While taxonomy and systematics have not necessarily been the primary topics of discussion at most of these meetings, the tradition of such regular gatherings is an important first step. Co-operative networks help develop a united community voice and greatly enhance efficiencies making arguments for funding more persuasive. The engagement of students is critical for the future success of such networks. Synergism appreciably increases as such communities grow in membership. Funding agencies must recognize the value of such global efforts and endeavour to facilitate and support training and the development of the infrastructure required to maximize the outcomes of such collaborations. Although it is disappointing that federal granting agencies in other countries have not followed suit, the United States National Science Foundation's PEET programme is a good example of such a funding initiative and, as noted above, has played a major role in enhancing tapeworm taxonomy and systematics. The decision-making process in the allocation of resources by any funding agency should carefully consider the difference between the long-term utility of the products of integrative taxonomy and those of barcoding endeavours.
CONCLUSIONS
In many respects, this is a unique moment in time for taxonomy and systematics, and certainly for such studies focused on the tapeworms of elasmobranchs. A global perspective has catalyzed collaborations around the world; such collaborations have been reinforced by web-based taxonomic resources such as the Global Cestode Database. Estimations of global diversity for elasmobranch tapeworm orders are within our grasp. Enhanced microscopical capabilities and digital technologies have greatly facilitated the articulation and documentation of morphological characters and host identities. Advances in molecular methods have reduced the cost and effort required to generate enormous amounts of molecular data. It is now possible to place parasite and host data in a phylogenetic context as a result of advances in computer resources for analyzing data. Advances in parasite taxonomy and systematics are best made in the context of collaborations and the synergisms they foster. These collaborations should involve host and parasite taxonomic experts and data from morphology and molecules. Assuming that the enthusiasm for barcoding does not undermine the taxonomic progress made to date, the taxonomy and systematics of elasmobranch tapeworms have a very bright future.
ACKNOWLEDGEMENTS
I am grateful to Jane O'Donnell, Kirsten Jensen, Bob Capers and Charles Henry for their thoughtful comments throughout the development of this manuscript.
FINANCIAL SUPPORT
This work was supported in part by the National Science Foundation Planetary Biodiversity Inventory Program grant Nos. DEB 0818696 and DEB 0818823.
