FORUM

The Bobbit worm dilemma: a case for DNA (Reply to Salazar-Vallejo et al. 2011. Giant Eunicid (Annelida) in shallow tropical and temperate seas. Rev. Biol. Trop. 59-4: 1463-1474)

Anja Schulze Department of Marine Biology, Texas A&M University at Galveston, P.O. Box 1675, Galveston, TX 77553; [email protected]

Received 04-I-2011. Corrected 20-iv-2011. Accepted 31-v-2011.

Whoever came up with the name “Bobbit Not only is the name misleading with worm”? It must be a fairly recent idea, given regard to the nature of the worms, it also leads that the “regretful incident” in the Bobbit to taxonomic confusion. Some sources speci- family only happened in 1993. Unfortunately, fically apply it to what they believe is it is not clear who coined the name or when aphroditois, others use it for any large eunicid. it was first used. The name does not do the Given that not even eunicid taxonomists agree worms justice and is misleading. For example, on the identity of E. aphroditois (well, taxono- one of the explanations for the name that comes mists rarely agree with each other, but that’s up when googling “Bobbit worm” is that the a different story), it is not too surprising that female bites off the male’s penis after copula- untrained aquarists or recreational divers lump tion (seriously? Since when do they have peni- all the species together under one easily memo- ses?!). Neither are there any reports of eunicids rable common name. attacking body parts of unsuspecting humans Does size matter? Salazar-Vallejo et al. in their vicinity. The jaws resemble scissors, cite reports of eunicid worms of over 3 m leng- which is another possible explanation for the th. The Australian museum even holds a speci- name, although actually, the “Bobbit incident” men that reportedly was nearly 6 m long when was committed with a carving knife. collected (Fauchald 1992 and pers. comm.). Eunice species can indeed be voracious (As a side note: six meters is long, but it’s still predators with impressive jaws. They are often far from the longest worm ever reported. That accidentally introduced into reef tanks, hidden honor falls to the nemertean Lineus longissimus in “live rock”, and wreak havoc with other which can grow to 30 m in length and probably tank inhabitants. This behavior has given them at least twice as long when fully stretched out a bad reputation in the aquarist’s community [McIntosh 1873-1874]). The main problem and beyond. One children’s book even inclu- with using size in taxonomic studies is that des them in a list of the most “terrifying and even big worms start out as tiny worms and ugly sea creatures” (Christiansen 2008). While probably take many years to reach their full ugliness is a matter of taste, I would agree that, length and segment number. In most cases we in their own little world, they are probably do not know which characters are size-depen- quite terrifying. But that’s still a far stretch dent. In traditional, type-based , this from the Bobbits. can pose serious problems.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 59 (4): 1475-1477, December 2011 1475 Color is another issue. One of the rea- Molecular data may not be necessary for each sons Salazar-Vallejo et al. were able to shed and every species description, but for compli- some light on the taxonomic distinctness of cated taxonomic problems, such as that of the different “Bobbit worms” is that they could large eunicids, DNA would add an indepen- identify different color patterns from photo- dent, objective line of evidence. graphs taken in different geographic regions. An impediment for molecular work on Most traditional taxonomists working with is that museum and ecological collec- museum collections do not have that option, tions are usually formalin fixed. Formalin is because the color is lost in the preserved spe- still the best solution for preserving cimens. Some early taxonomists such as de morphology, but it crosslinks with DNA and Quatrefages had the talent, patience and funds proteins, making the extraction of quality DNA to illustrate their collections in color, based on difficult if not impossible. Protocols for DNA live specimens, but many others did not. Even extraction from formalinized tissue or for enzy- today, few taxonomic descriptions of anne- matically repairing DNA do exist (Schander lids include color illustrations or photographs. & Kenneth 2003, Skage & Schander 2007), Color patterns may be very distinctive in many but they are cumbersome and not reliably cases, but color can also be a confounding successful. New specimens would have to be factor. To what degree, for example, does color collected from the type locations and tissue vary with diet? How much color variation is samples preserved in an appropriate manner based on phenotypic plasticity and how much while ensuring that voucher specimens are has a genetic basis? If we find different color available for morphological studies. In many morphs in otherwise morphologically similar cases, especially for large worms, a small piece individuals, how do we decide whether we are of tissue, such as a parapodium, can be preser- dealing with a single, polymorphic species or ved for molecular work, whereas the remainder with separate species? One of the few studies of the specimen is fixed in formalin. That way, that specifically addresses the latter issue is that the same specimen can be studied morphologi- by Nygren et al. (2011) of Harmathoe imbrica- cally and genetically. ta. This species has 10 different color morphs A DNA barcoding approach would be the along the Scandinavian coasts but there is no simplest solution. A few individuals of confir- genetic evidence suggesting they represent med from the type locality different species. and, if possible, each of the potential other Salazar-Vallejo et al. conducted a lot of Eunice species proposed in Salazar-Vallejo et detective work to reconstruct the taxonomic al.’s paper could be sequenced for one or two history of Eunice aphroditois and other large genes from the mitochondrial genome, such Eunice species and make a convincing case as the common barcoding gene cytochrome for splitting E. aphroditois into several species c oxidase subunit I (COI) or 16S ribosomal reflecting different morphologies. Their main RNA, and the sequences of the proposed spe- conclusion is that more type material, including cies could be compared with each other. The neotypes and topotypes, needs to be examined approach is not without flaws and often leads to tease apart species and delineate their diffe- to new questions. For example, how different rences and intraspecific variation. do two sequences, or sets of sequences, have to While I believe that the suggested work is be before they are considered separate species? crucial for resolving this and similar taxono- The substitution rate in the mitochondrial geno- mic dilemmas, I would also argue that mole- me and consequently the amount of variation cular data should be added to this equation. among species and individuals within species The decreasing costs for DNA sequencing can vary from taxon to taxon (Galtier et al. allow more and more researchers access to the 2009). Usually, if the interspecific variation for technology or opportunities for collaboration. COI is greater than the intraspecific variation, a

1476 Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 59 (4): 1475-1477, December 2011 more convincing case can be made for species replace morphology in taxonomic studies, but boundaries. If not, the researchers should take will add a new dimension and perhaps clarity a second look at the morphology and maybe in delineating species. Rather than using one re-evaluate some of the characters used in or the other data source, I believe that “Bobbit their identifications. Sometimes, genetic data worms” are a good case in point for a combi- show great differences among species that are nation of morphological and molecular data morphologically identical or almost identical. to provide a more complete picture of the An example for such a cryptic species com- evolutionary history of these terrifying and plex in eunicids is the genus Palola (Schulze ugly creatures. 2006). On the other side of the spectrum, the mitochondrial sequence data can help clarify ACKNOWLEDGMENTS whether the observed variation among sam- ples, such as the color morphs in Harmathoe Kristian Fauchald and Jon Norenburg hel- imbricata mentioned above, reflects species ped retrieve some of the information on large differences or intraspecific variation. eunicids and nemerteans. Elizabeth Borda pro- A more thorough approach would be to vided comments that improved the manuscript. generate a multi-gene phylogeny for the entire The work was supported by NSF AToL grant genus Eunice in which all of the species in DEB-1036186. question are included. A great start was the paper by Zanol et al. (2010) who included 25 species of Eunice in their phylogenetic analysis REFERENCES of . Of the larger species mentioned Christiansen, P. 2008. Terrifying and ugly sea creatures. in Salazar-Vallejo et al.’s paper only E. rous- Gareth Stevens, New York, USA. saei and E. cf. violaceomaculata were included and appeared as sister species. Note the “cf.” Galtier, N., B. Nabholz, S. Glémin & G.D.D. Hurst. 2009. Mitochondrial DNA as a marker of molecular diversi- in E. violaceomaculata: the fact that not even ty: a reappraisal. Mol. Ecol. 18: 4541-4550. experienced eunicid taxonomists could confi- dently identify the species is a further reminder Nygren, A., E. Norlinder, M. Panova & F. Pleijel. 2011. that we need help from sources other than mor- Colour polymorphism in the Harmothoe imbricata (Linnaeus, 1767). Mar. Biol. Res. 7: 54-62. phology. It would be interesting to know where the other large Eunice species fall in the tree. Schander, C. & H.M. Kenneth. 2003. DNA, PCR and One of the most important finding by Zanol et formalinized tissue - a short review and pro- al. (2010) was that Eunice is not monophyletic, tocols. Organisms Diversity & Evolution 3: 195-205. warranting a revision of the entire family. As Schulze, A. 2006. Phylogeny and genetic diversity of such, it is possible that some of the species palolo worms (Palola, Eunicidae, Polychaeta) from proposed by Salazar-Vallejo et al. belong to the tropical North Pacific and the Caribbean. Biol. different genera. Bull. 210: 25-37. In conclusion, Salazar Vallejo et al.’s Skage, M. & C. Schander. 2007. DNA from formalin-fixed approach of examining color images of “Bobbit tissue: extraction or repair? That is the question. Mar. worms” from around the world highlights some Biol. Res. 3: 289-295. of the taxonomic problems encountered in this group. I fully support the call for examining Zanol, J., K.M. Halanych, T.H. Struck & K. Fauchald. 2010. Phylogeny of the bristle worm family Eunici- more topotypes and neotypes of the species in dae (, Annelida) and the phylogenetic utility question, but strongly suggest incorporation of noncongruent 16S, COI and 18S in combined of molecular data as well. Genetics will never analyses. Mol. Phyl. Evol. 55: 660-676.

Rev. Biol. Trop. (Int. J. Trop. Biol. ISSN-0034-7744) Vol. 59 (4): 1475-1477, December 2011 1477