Much ado about something

by Greg Mayer

In a paper in press in Nature, Min Zhu and colleagues describe a new species of placoderm from the Silurian period of China. Placoderms are an extinct group of (usually) heavily armored jawed fishes that lived in the Silurian and Devonian. The new species is based on a beautifully preserved 3-D specimen, and is interesting, but it is being widely misreported in the press.

The  holotype of the new species, a three-dimensionally preserved specimen with head and trunk armour in anterolateral (a), lateral (b), anteroventral (c) and dorsal (d) views. A small part of the left trunk armour was accidentally sawed off as extraneous material and repositioned in b. Scale bars, 1 cm. e, Life restoration.

The holotype of the new species, a three-dimensionally preserved specimen with head and trunk armour in anterolateral (a), lateral (b), anteroventral (c) and dorsal (d) views. A small part of the left trunk armour was accidentally sawed off as extraneous material and repositioned in b. Scale bars, 1 cm. e, Life restoration.

To understand why this new species is interesting requires some background information. First, we need to know that while most modern vertebrates (backboned animals such as ourselves) have jaws, and are called gnathostomes (“jaw mouths”), not having jaws is the primitive condition (jawless vertebrates, represented today only by hagfish and lampreys, are called agnathans). The origin of jaws is thus a key episode in the vertebrate story.

Second, we need to know that there are four great groups pf gnathostomes, the placoderms, the acanthodians (another extinct group, often called ‘spiny fish’), chondrichthyans (cartilaginous fish: sharks, rays, and their relatives), and osteichthyans (bony fish: tuna, gars, goldfish, etc.; the tetrapods are descended from osteichthyans, and for our purposes can be included with them).

And finally, we need to know that the vertebrate skull is a composite of bones from three different sources: the chondrocranium, bones preformed in cartilage that surround the brain, the splanchocranium, bones preformed in cartilage that support the gill arches, and the dermatocranium, bones that ossify directly and cover most of the outside of the skull. Gnathostome jaws are formed by the anteriormost bones of the splanchocranium (the palatoquadrate in the upper jaw, and Meckel’s cartilage in the lower jaw), which are often covered over or replaced by dermal bones in development. (Chondrichthyans, lacking bone, have only the first two components in their skulls.)

Chondrocranium, blue, splanchocranium green, dermatocranium labeled 'dc'. From

Chondrocranium, blue, splanchocranium green, dermatocranium labeled ‘dc’. From

Okay, so what’s interesting? Placoderms have jaws, including the palatoquadrate and Meckel’s cartilage, which are accompanied by dermal bones that have usually been thought not to correspond very precisely to the dermal bones of osteichthyans. In the new fossil, Zhu and colleagues identify some dermal bones as being the same as in osteichthyans, most prominently the maxilla in the upper jaw and the dentary in the lower jaw (see first figure above). This is what’s interesting, because if true, it would mean that the osteichthyan condition is more widespread than previously known, and thus perhaps change some of our ideas on the relationships of the various gnathostome groups.

Another thing Zhu and colleagues do is a phylogenetic analysis of 75 taxa with 253 characters, but unfortunately for them the results are quite muddled, with no clear evidence that the ‘maxilla’ or other dermal jaw bones of the new placoderm are homologous to those of osteichthyans. These large data set analyses rarely produce convincing results, because it is the interpretation and analysis of the individual characters that most strongly influence the results, and these individual analyses are usually de-emphasized (or as in this case, hidden in the online supplement).

So where has the press gone wrong? First, some commentary, especially by scientists, has been directed toward the differing trees gotten by Zhu and colleagues, versus the one obtained by Davis et al. (2012) last year using essentially the same data set. This is inside baseball– the relationships among the four great gnathostome groups is quite interesting, but this paper does not resolve the question.

Popular media have been implying that jaws were not previously known in placoderms or fish in general, or that we would not expect jaws in fish this primitive or early. None of this is right. Placoderms are jawed fish, they are not the oldest known jawed fish, and the bones in this specimen do not apparently show a new or previously unknown condition (rather, the claim is that the condition in this fish resembles an already known condition). There is definitely something of interest here, but it’s not quite all that the media are portraying it as.


Davis, S. P., Finarelli, J. A. & Coates, M. I. 2012. Acanthodes and shark-like conditions in the last common ancestor of modern gnathostomes. Nature 486: 247–250.

Zhu, M., X. Yu, P.E. Alberg, B. Choo, J. Lu, Q. Qu, W. Zhao, L. Jia, H. Blom and Y. Zhu. 2013. A Silurian placoderm with osteichthyan-like marginal jaw bones. Nature in press.


  1. JBlilie
    Posted September 27, 2013 at 1:47 pm | Permalink

    Thanks very much! I was confused/misled by the press on this fossil.

    • Jesper Both Pedersen
      Posted September 27, 2013 at 2:05 pm | Permalink

      Ditto’ed. Great post.

    • Posted September 27, 2013 at 6:18 pm | Permalink

      Same here. I appreciate the clarification and distinctions made in this post very much.

    • Torbjörn Larsson, OM
      Posted September 28, 2013 at 7:18 am | Permalink


  2. Posted September 27, 2013 at 2:48 pm | Permalink

    Me too. I will give this to my students to read, Greg!

  3. D. Taylor
    Posted September 27, 2013 at 6:33 pm | Permalink

    Great post, Greg. Thank you so much. The background information was greatly appreciated and so clearly written. And isn’t that holotype a marvel of visualization!

  4. Diana MacPherson
    Posted September 27, 2013 at 6:35 pm | Permalink


  5. Steven Obrebski
    Posted September 27, 2013 at 7:04 pm | Permalink

    Some years ago I heard the story that Erik Stensiö, the famous Swedish fish paleontologist and his coworkers studied the detailed anatomy of the bony heads of Ostracoderms by sanding the fish fossil from the tip backward about a millimeter taking a photo, sanding some more, taking a photo etc. to obtain a series of serial photographic sections which then were enlarged. The apertures made by nerves and vessels could be seen in the enlargements, and cut out. The whole sequence could be hung up serially and wires inserted to trace the path of nerves and vessels, producing a neat physical “cat scan” for studying the anatomy in detail.

    Is this story at least approximately true?

    Today, of course appropriate synchrotron X Ray technology accomplishes the same thing. Gone are the days when paleontologists spent weeks carefully trying to avoid damaging the specimens while drilling away the rock around it.

    • Posted September 27, 2013 at 8:25 pm | Permalink

      Well, they still do a lot of that. But now they also have the option you mention, which is essential for delicate specimens.
      I remember something of the serial sectioning studies in Ostracoderms. Vaguely recall they were able to map the cranial nerves, showing that these features of nervous organization were remarkably ancient.

    • Eddie Janssen
      Posted September 28, 2013 at 2:21 am | Permalink

      In “The Rise of Fishes” by John A. Long, pages 57-58, this method is mentioned (plus a nice picture from Stensiö and his team). Long tells that it is called Sollas’ grinding technique after William Sollas. Each ‘slide’ was a tenth of a millimeter.

    • John Scanlon, FCD
      Posted October 1, 2013 at 8:52 pm | Permalink

      This grinding or shaving technique was also used for the ‘Visible Human’ project, which at the time (early 90s) seemed very next-century. The 1-mm slices now look like meatloaf, beside various scanning technologies that provide much higher resolution, but at least they’re in full-colour visible light.

      “Gone are the days when paleontologists spent weeks carefully trying to avoid damaging the specimens while drilling away the rock around it.” – I wouldn’t say they’re gone, the traditional craft of fossil preparation remains essential to create permanent display and study specimens, which are things of beauty as well as the objective source of current and future digital datasets.

      Synchrotron scans are rather limited in the dimensions of the specimen, generally at the millimetre level. None of your T-rexes.

  6. Steven Obrebski
    Posted September 27, 2013 at 7:08 pm | Permalink

    Forgot to mention above that the Ostracoderm cranium was entirely encased in bone.

  7. Steven Obrebski
    Posted September 27, 2013 at 7:10 pm | Permalink

    Oops, I did mention it!
    That dinner wine sure was good!

  8. gravelinspector-Aidan
    Posted September 28, 2013 at 4:22 am | Permalink

    I hadn’t previously ‘got’ why Meckles’ Cartilege is considered important in studies of cranial anatomy, but that’s cleared it up nicely. As I recall, one of it’s (M’s C) evolutionarily later peregrinations has been the subject of discussion about the origins of dinosaurs and mammals and mammal-like reptiles in the late Permian to early Triassic.
    “Serial sectioning” as mentioned above is a well-known technique. It’s use in the past has probably been restricted by the difficulty of representing and manipulating the data. Many geologists are fairly good at “3d thinking” (probably we’re culled by the early joys of trying to understand geological maps ; some people just can’t do it, and drop out), so can handle assembling different views of different fossils crushed and distorted in different ways. So presenting a page of photographs or camera lucida drawings is sufficient. But the pretty pictures that you can generate from serial sectioning are … well prettier. And probably easier to work with.

    • Steven Obrebski
      Posted September 28, 2013 at 8:40 am | Permalink

      Your 3D thinking comment brings to mid that at the Walker Museum at the University of Chicago, E. C. Olson, an expert on ancient amphibians taught a course in higher vertebrate cranial anatomy. At the end of the course you were supposed to be able to place yourself in the hypophyseal fossa, where the pituitary gland resides, point in different directions, and indicate the bones you would encounter. An “insiders view” of cranial anatomy.

      This nice post by Greg Myer has got me perusing my copy of Robert Carroll’s Vertebrate Paleontology and Evolution and resurrecting fond memories of taking a course in fish paleontology from Bobb Schaeffer at the American Museum of Natural History in NY in the late 50’s, which solidified my interest in evolution, and was certainly a romantic excursion into the subject for an undergraduate.

  9. Reginald Selkirk
    Posted September 28, 2013 at 5:58 am | Permalink

    Deep sea critters

  10. marksolock
    Posted September 28, 2013 at 11:03 am | Permalink

    Reblogged this on Mark Solock Blog.

  11. Christine Janis
    Posted September 29, 2013 at 7:57 pm | Permalink

    While this is a great find, as far as I can see it only reinforces what I’ve been teaching (and writing about) for years: at some level, the dermal head shield (dermatocranium) seen in ostracoderms must be homologous with that in gnathostomes (placoderms and osteichthyians), and its absence in chondrichthyans must represent a secondary loss. The other possibility, that the dermatocranium of osteichthyes is a neomorph without history in the similar structure in earlier vertebrates, seems highly unlikely. The apparent presence of osteichthyan-like marginal mouth bones in a placoderm is intriguing, but (at least for me) hardly a huge shock —- more like confirmation of the bleedin’ obvious. But perhaps that’s just me.

One Trackback/Pingback

  1. […] reading at the SOURCE: Greg […]

%d bloggers like this: