Friday, July 2, 2010

Three Ways to Split a Fish

Here's another squirt from the Third International Palaeontological Congress in London, where yesterday I attended a workshop on early vertebrates - the early evolution of backboned animals such as ourselves. Being as I am, particularly at the moment, a martyr to bipedality, this is something with which I can sympathise.

But I digress.

At the proper scale, all vertebrates are fishes, and what a varied bunch they are. There are the jawless fishes or Agnetha Agnatha - lampreys and hagfishes, the last remnants of a very varied and ancient assemblage of armoured fishes or ostracoderms, all of which are extinct. Then there are the Chondrichthyes - sharks and rays. And finally there are the Osteichthyes - the bony fishes, by far the most successful group of vertebrates, which happen to include as an offshoot all the land vertebrates, a small and bizarre group of fishes specialized for living in water of negative depth.
Cute Bunneh.
A bony fish specialized for living in water of negative depth. Recently.

There are, however, two other major groups of fishes that are now completely extinct, but in whose understanding might allow the unlocking of many enigmas - many, indeed, conundramatic riddles. These are the Acanthodians, or 'spiny sharks', mainly small, scaly, spiny fishes whose internal skeletons are rarely preserved, but are known mainly from smears of scales and spiny bits that look like the parts of a fish that the cat would have thrown up as indigestible. Acanthodians are known from the Silurian right up to the end of the Permian. Acanthodes, one of the best known acanthodians, is a Permian form, one of the last of its kind, and not necessarily representative of the group's much earlier heyday. The wider relationships of acanthodians has been a matter of controversy, with opinion shifting between allegiances with sharks or bony fishes.

Then there are the Placoderms, another large and very diverse group that emerged in the Silurian but disappeared at the end of the Devonian. These were often heavily armoured, with robust head- and trunk armour, and look superficially like some of the various jawless ostracoderms, for all that they have jaws. But the pattern and styling of the armour of placoderms is so unique that it has been hard to relate them to either bony fishes or sharks. Some placoderms were tiddlers. Others were, as we scientists say, mean meat-eatin' muthas.

Underlying all this is the vast morphological gulf that separates jawless vertebrates from the various kinds with jaws, or Gnathostomes. It's not just a question of the evolution of jaws (not a trivial restructuring in any case). The evolution of jaws was accompanied by wholesale changes in many other parts of the anatomy, from the wiring of the brain to the appearance of paired limbs. What one should like to do is get some idea of how the various features that make up a gnathostome were acquired, and which order. Without that, we are left with a vast evolutionary jump, from Agnetha to Frida Agnathans to Gnathostomes, without any further clue as to evolution's likely course. Did paired fins evolve before jaws? Did jaws evolve all at once, or by a series of various modifications? In order to help us answer such questions it would help if we had transitional forms, way stations between the jawless and jawed state. It turns out, however, that these forms have been with us all the
time, only we'd failed to recognise them. They are -- ba boom, tish -- the acanthodians and the placoderms.

Most people have become accustomed to thinking of placoderms and acanthodians as 'natural' groups. This is why there has been so much debate about whether acanthodians are closer to sharks or bony fishes. The question is unanswerable, unless one makes an intellectual leap and challenges the assumption that acanthodians are all one, natural group, like - say - sharks, or bananas. However, there is an alternative, in which acanthodians are less a natural group than a grade of evolution in which some acanthodians are closer to sharks, others are closer to bony fishes, and still others are offshoots from a time when these two lineages were as yet indistinct. This revelation has come about thanks to recent work on new and extremely rare material of acanthodian braincases. This insight is crucial as for the first time we can get an idea of what the latest common ancestor of sharks and bony fishes looked like - and it would have looked very much like an acanthodian.

But what of placoderms? Once one breaks this group up into its many constituent parts, the group starts to arrange itself as a series of grades between agnathans and gnathostomes. Of the two major placoderm groups, for example, the antiarchs look much more primitive and ostracoderm-like, whereas the arthrodires seem much more gnathostome-like.

This is, however, only the beginning of what promises to be an exciting if somewhat tricky path of research. For once one breaks up the monophyly (natural-group-ness) of acanthodians and placoderms, many things become possible - and many new problems surface. Crucially, placoderms of any kind look very different from acanthodians, so which group of placoderm is most closely related to which group of acanthodian? Did placoderms evolve into acanthodians? At present this seems as likely as girrafes evolving from unicycles. This is a circle whose squaring will come in the next few years of research into an old problem dramatically renewed, simply by breaking down a couple of cherished assumptions.


  1. I always liked the placoderms is a very interesting fish here is a picture of the great dino-artist, "The True" Z. Burian:

  2. ah yes, Burian. One of the all-time great palaeo artists. I grew up with his book Life Before Man.

  3. If I were not on vacation, I might be able to formulate an intelligent reply to this well thought-out post, but the field is too unfamiliar to me.

    It's just that when I think "jaws" I think about the neural crest cells which were co-opted to make them, the same that allowed even the most primitive of fish to make bones around their pharynx in the gill slits. It puts me in mind of New Head hypothesis by Gans and Northcutt way back when. I'm a little sorry not to see these evoked in some way or another, but it would probably be too speculative. The internal carotids by higher vertebrates are encased by neural crest-origin soft tissues, and the sphenoid is a marriage between an older post-sphenoid region of mesodermal origin and the newer pre-sphenoid surrounding the antehypophysis, that comes from neural crest. I'm sure there is something nifty to tease out here about placodal contributions to head development, and I'd look to a couple of US and UK labs I have in mind, to help find them.

  4. Darn, and I just broadcast my ignorance with the juxtaposition of "bones" and "pharynx". I'd have meant to type "cartilage"... and I seem to have lost a sentence in there in which I had tried to make reference to this article. Sorry about that. I also don't know how to drop my ancient handle and just get back to my real name. Well, you know me anyhow.

  5. Sorry ! is the specie of genus "Dinichthys":



  6. They must have been terrible massacres in the Devonian between, Dinichthys and Cladoselaches.

    Without the Chondrichthyes have survived until today, once again no rule of "law stronger", probably the law of the most adapted. Or the chance, maybe.

  7. SQUEEEEEEE!! Kyoot bunny!

    Ahem. Sorry.

    Neural crest is the best thing since sliced cephalochordate. But I'm probably biased. ;-)

  8. Ah yes, neural crest, that magical fairy dust that turns a blob of goo into a vertebrate, with eyes, and ears, and a face, and skin, and everything. The break-up of acanthodians and placoderms into a series of grade groups more or less related to crown gnathostomes should allow, in theory, for a more detailed dissection of the role of neural crest in the evolution of jaws and other structures than has been possible. That, and new methods of cell lineage tracing that can map the contributions of neural crest and endochondral tissue to structures such as the shoulder girdle, whose results can be tracked back by homology and mapped onto extinct groups such as placoderms. There's a great paper in Nature in 2005 from Matsuoka et al that goes into this.

  9. You have a thing for ABBA?

    Oh, and you missed writing about the most fascinating and exiting group of them all, the LUNGFISHES!

    But here I am slightly biased...

    IOWADIPTERUS for president! Errr... or not ;p

  10. Hi Babs. I'm afraid that after Uranolophus I kinda lost count. ;)

  11. Babs: you should try to find a record called 'Gold: A Tribute to ABBA' by a bunch of Swedish session men called 'The Black'. It's a collection of ABBA hits performed as heavy metal. 'Take a Chance On Me' done in the style of Metallica is quite ... Er ... Unforgettable.

  12. Haha, ok I have heard of that record before... Will try and find it!

    I'll do my best to make the Devonian lungfishes more memorable in the future... ;)

  13. All this amusing conversation reminds me of the song _Dunkleosteus_:


    (This reminds me of the comments of someone who is Scottish and disappeared a certain Asuracenturix the bard).

  14. I think that all fish must have an origin of Chondrichtyes, because they are the best-adapted, but have not found the oldest dating back to the beginning of the Cambrian but cartilage was not possible to find traces. Therefore the oldest Ghanostomata must be a kind of Chondrihctyes (cartilaginous fish, important and interesting to the cartilage but not its fossilization is possible), I hope you find it soon, to strengthen my hypothesis.

  15. Obviously, supporting my hypothesis occurred in certain investigations: Phylogenetically, the jaw appears
    bony fishes and with characteristics similar to those of
    mammals is evolutionarily older, and is present in fish such as shark cartilage. Since the beginning of development
    mandibular cartilage is related to the mandibular
    The jaw is formed by ossification
    intramembranous, but develops secondary cartilages differ from primary cartilage in embryological origin, organization and mode histological
    of growth regulation. Thus, initial membranous ossification, but later in development are added endochondral ossification centers.
    Both processes involve an initial condensation
    mesenchymal tissue and bone formation
    calcified. Membranous ossification is made directly, whereas endochondral ossification includes an intermediate step in which the cartilage regulates
    pattern of growth and bone development.

    Cromer 2 - Nature Network 0

  16. I apologize for the disorder of my comment is that I have the stomach too weak. Sorry!!

  17. @Alejandro: I hope you get well soon! Your hypothesis sounds interesting, but it's more complicated than that. Some early chondrichthyans had quite a lot of bony tissue, and many osteichthyans (such as coelacanths) are largely cartilaginous. Placoderms and acanthodians have bony dermal armour, though much of the endoskeleton is cartilaginous. Soe we can't always equate 'chondrichthyan' with 'cartilaginous'.

    Cromer 1 - Nature Network 0

    Yes, I no longer blog on Nature Network, for all sorts of reasons I won't go into here. But you have reminded me that I really ought to put up a report, if belated, on Cromer Is So Bracing 2010.

  18. _have bony dermal_

    Henry : ¿will not be small "bunny" skin ?

  19. That Matsuoka paper looks really interesting. In neuroanatomy, there's this controversy over where to place the accessory nucleus: special visceral efferent, or general somatic efferent? From a brief glance at the paper, it looks as if SVE might be the winner, since trapezius is described as a branchial muscle. Of course the classification might also depend on the patterns of gene expression in the developing motor neurons themselves.

  20. _Soe we can't always equate 'chondrichthyan' with 'cartilaginous'_

    Okay Henry, but also think that all fish with jaws "Ghanostomata", derived of the "Chondrichthyans" is the first, independent of whether or not posess cartilage.

    I am sure they are the most primitives, just missing getting at discovery this missing link.

  21. I am even more certain that the common ancestor between "Placoderms" and "Acanthodians" is a specie of class "Chondrichthyan".