What, if anything, is a hagfish?

by Greg Mayer

Linnaeus thought hagfish were worms, not fish, but there has been considerable controversy about which fish they are closest to. Are hagfish the earliest diverging of all extant vertebrates, or are they closer to lampreys?  The latter hypothesis, which we might call the cyclostome hypothesis (because hagfish and lampreys have been grouped in the taxon Cyclostomata), was favored for many decades. But in the late 20th century, people began to argue that lampreys were closer to jawed fish (gnathostomes), making cyclostomes paraphyletic (i.e. ancestral rather than sister to gnathostomes), which would mean that hagfish were their own group, an early and primitive branch.

In my post, I said recent molecular data had moved us back to the cyclostome hypothesis. Philippe Janvier, one of the most prominent proponents of the paraphyly hypothesis, has come round back to the cyclostome hypothesis, and has an excellent, brief, discussion of the history of the issue and the evidence.  So, hagfish, it seems, are cyclostomes. Money quote:

The results of Heimberg et al. (11) are certainly certainly the most convincing contribution ever published in support of cyclostomes monophyly…, Although I was among the early supporters of vertebrate paraphyly (6, 7), I am impressed by the evidence provided by Heimberg et al. (11) and prepared to admit that cyclostomes are, in fact, monophyletic [i.e. holophyletic].

This is, by the way, an excellent example of how a good scientist accepts new evidence, and alters his views accordingly.


Janvier, P. 2010. microRNAs revive old views about jawless vertebrate divergence and evolution. Proceedings of the National Academy of Sciences (USA) 107:19137-19138. pdf (may not be open access)


  1. Digitus Impudicus
    Posted November 22, 2011 at 12:56 pm | Permalink

    I think they are sloser to lampreys myself.

  2. Nicholas
    Posted November 22, 2011 at 1:03 pm | Permalink

    Cyclostomes or not, they continue their prominent role in my nightmares.

  3. starskeptic
    Posted November 22, 2011 at 1:10 pm | Permalink

    “What, if anything, is a hagfish?”

    -why is there a hagfish rather than nothing?..

  4. Posted November 22, 2011 at 1:23 pm | Permalink

    “…making cyclostomes paraphyletic (i.e. ancestral rather than sister to gnathostomes)”

    How does paraphyly of cyclostomes translate into hagfishes being ancestral to gnathostomes? Early branching does not equal ancestral. It would simply mean that (modern!) hagfishes are sister to other vertebrates including lampreys rather than a lampreys+hagfishes clade being sister to other vertebrates.

    • Posted November 22, 2011 at 1:52 pm | Permalink

      That confused me, too. But he’s not saying hagfishes are ancestral to gnathostomes; he’s saying cyclostomes are ancestral to gnathostomes. Which I think makes even less sense. What am I missing?

    • whyevolutionistrue
      Posted November 22, 2011 at 1:59 pm | Permalink

      It doesn’t make hagfishes ancestral to gnathostomes; it makes cyclostomes ancestral to gnathostomes. As Janvier makes explicit, part of the appeal of the paraphyly hypothesis was that under it hagfish and lampreys become stepwise approaches to the gnathostome condition:

      This phylogenetic pattern implied that cyclostomes could throw light on the early steps of the assembly of the vertebrate body plan and that hagfishes could document the most generalized condition for a number of vertebrate characters.

      Alas, it appears not to be so. As he makes even clearer in a later paper, he now regards cyclostomes as quite divergent from the cyclostome/gnathostome common ancestor, and of not much use in inferring the most general conditions of the Vertebrata. Rudimentary vertebrae, by the way, have now been reported in hagfish.


      Gai, Z., P.C.J. Donoghue, M. Zhu, P. Janvier & Marco Stampanoni. 2011. Fossil jawless fish from China foreshadows early jawed vertebrate anatomy. Nature 476-324-327.

      Ota, K.G., S. Fujimoto, Y. Oisi, & Shigeru Kuratani. 2011. Identification of vertebra-like elements and their possible differentiation from sclerotomes in the hagfish. Nature Communications 2:373, 6 pp.

      • randyextry
        Posted November 22, 2011 at 5:38 pm | Permalink

        I think the terminology is confusing because if cyclostomes are paraphyletic, you could say lampreys are sister to gnathostomes, but when grouped with hagfish (in the group cyclostomes) they are ancestral to gnathostomes.

        Also, saying “hagfish and lampreys become stepwise approaches to the gnathostome condition” seems misleading to me in a “if people evolved from monkeys, why are there still monkeys” kind of way.

  5. Torbjorn Larsson, OM
    Posted November 22, 2011 at 2:22 pm | Permalink

    Thanks! I was just wondering how the new results were seen by the consensus.

    Speaking of the consensus, as a layman I was not aware that tunicates could be descendant of lancelets. So I googled, and stumbled on the facts that:

    a) tunicates makes cellulose


    b) we have small fibers of cellulose in our dermis, so have spider webs, and so has silk. (At least according to a 60’ish paper that is web accessible).

    Suddenly I feel halfway to a plant. Even more so, since slime molds express it in their tissue analog in fruiting bodies.

    That polarized cell layer bound to cellulose and proteins, with cell-cell adhesion mediated by ancestral catenins, is termed “epithelium” by Dickinson et al.

    “Taken together with earlier results (14), our work shows that the non-metazoan Dictyostelium discoideum has a bona fide polarized epithelium consisting of a single layer of structurally and functionally polarized cells that secrete proteins into a luminal space (Fig.

    So if catenins are ancestral to Amoebozoa, and perhaps (the ability to form) epithelium tissues, why not cellulose?

    What, if anything, were the Amoebozoa/Ophistokonta ancestor?

  6. Dominic
    Posted November 23, 2011 at 3:36 am | Permalink

    This is really interesting – I need to see diagrammatic representation though as my feeble brain finds it harder to grasp in written form!

  7. TJR
    Posted November 23, 2011 at 8:10 am | Permalink

    (Apologies in advance if anyone finds this offensive or just not funny).

    Do you get gay Hagfish?

    If so, are they called Faghagfish?


  8. Jim Thomerson
    Posted November 25, 2011 at 1:50 pm | Permalink

    The ancestral argument is bogus. If hagfishes and lamperys are not monophyletic, then hagfishes could be the plesiomorphous sister group of lamperys + gnathosomes. The group cyclostomata would have to be defined to reflect this; excluding the hagfishes, I suppose.

    If hagfishes and lamperies are monophyletic, then the cyclostomes are the plesiomorphous sister group of the gnathostomes. I see no need for the term Holophyletic = Monophyletic.

    • whyevolutionistrue
      Posted November 25, 2011 at 2:08 pm | Permalink

      The term is needed because monophyletic, as initially defined, include both paraphyletic and holophyletic. If you’re going to use the term monophyletic to mean holophyletic, you should always say something like “monophyletic sensu Hennig”. (Willi Hennig being the German entomologist who redefined it to mean what holophyletic means. Since you used the word “plesimorphous”, I’m sure you know who Hennig is, but other readers probably don’t.)


  9. Jim Thomerson
    Posted November 25, 2011 at 7:59 pm | Permalink

    OK, I see the utility of Holophyletic. I suppose I assumed that anyone discussing cladistic relationships would be familiar with Henning, although they might not agree with his thinking. I have always thought of Monophyletic and Paraphyletic as conflicting concepts.

One Trackback/Pingback

  1. […] monophyly”. It turns out that we were probably right.  The last major opponent has graciously conceded defeat, as a result of higher-level DNA function & structure linking lampreys & hagfishes […]

%d bloggers like this: