Wednesday, April 20, 2011

You're Doing It Wrong: Birds With Teeth

Above: If only Charlie from Always Sunny could visit the Mesozoic.

My previous post on beaked theropods left one thing a little too ambiguous for my taste. We know that many non-avian theropods had beaks, but we also know that, famously, many of these also possessed teeth. An issue I've often come across in palaeontography is how exactly to restore this. Should the teeth erupt directly from the beak? Were they segregated to different portions of the jaw? Did the beak edge overlap an inset tooth row? This is an issue which I've rarely seen discussed online, let alone in the literature, so hopefully this post can be a starting point to suss things out. Because it's one of the most famous examples and one that's been frequently discussed in the lit, I'll focus specifically on Hesperonris regais here.

Look at almost any life restoration of Hesperornis, and it will show a keratinous beak covering the entire extent of the upper and lower jaws. I say "almost" only hypothetically, because I've literally never seen a hersperornithine drawn any other way (please link me one if you can). Here's a link to a google Image search for Hesperornis, and every reconstruction is the same. Some, like the beautiful painting above (by artist Larry Felder), clearly show teeth erupting directly from the tomia (edge) of a continuous keratin beak. As discussed last time, the continuous appearance of this beak is likely incorrect in itself, since non-avian birds probably all had 'compound rhamphotheca' made up of several distinct plates that are often visible in life.
Above is bit of anatomy lingo to make this discussion easier. The rhamphotheca ("beak") is usually divided (and literally divided, in the case of compound beaks) into several segments or regions. The figure is figure 5 from Heironymus & Witmer, 2010.

Now, for comparison, here are two views of the skull of Hesperornis regalis. On top is a ventral view of the skull showing the premaxilla and maxilla, from Elzanowski 1991. The bottom is a right lateral view of the skull taken from Heilman, 1926.
A few things to note here. The dentary teeth continue almost all the way to the tip of the jaw, though the very tip (and small predentary that was probably present) were toothless. In the ventral view of the upper jaw, you can see indentations where the lower teeth would have locked into the premaxilla. If there was a hard beak present, it would have been pitted to accommodate the dentary teeth (these are labelled dp, dental pits, in the top figure above). However, you can also see that the indentations are inset to the jaw a bit. The edges of the upper jaw slightly overhung the lower jaw, which would have allowed for the tomia, if it was there, to not come into contact with the lower teeth, causing wear any time the mouth closed.

The back of the jaws are a different story. As you can see in the lateral view, upper teeth are present only in the maxilla, not the premaxilla, and therefore restricted to the very back of the mouth. This can also be seen in the 'dental grooves' (dg) in the ventral view of the skull.

According to Heironymus & Witmer 2010, in both Ichthyornis and Hesperornis, the premaxillary nail and mandibular nail were the most heavily keratinized part of the beak. This is where the beak would have been most solid, like a normal bird bill. The same authors note that the simple presence of teeth in the maxilla and dentary of these species probably means that they lacked the latericorn and ramicorn parts of the beak entirely, and that the presence of cornfield rhamphotheca on the edges of the jaws may be unique to modern birds. However, as I noted above, the premaxilla in Hesperornis is also toothless and provides area for a tomia of some kind to be present. This would have been somewhat softer tissue, like the more pliable bills of ducks and geese. Further support for the presence of a beak on the premaxilla comes from the presence of a rhamphothecal groove on the dorsal part in front of the naris (visible but unlabeled in the figure above).

So how far did the beak extend? Heironymus & Witmer found that the latericorn almost always extends to the back of the subnarial bar in birds. This is a process of the premaxilla that extends back to separate the naris from the maxilla. Basically, this means that the beak will very rarely, if ever, extend onto the maxilla itself. As you can see in the lateral figure above, the maxilla in Hesperornis even compensates for this limitation by extending a bit forward underneath the subnarial bar to extend the tooth row anteriorly a bit past the possible full extent of the beak.

So, based on the evidence above, Hesperornis probably had a beak like the recon I whipped up below. The toothless, pointed tips would have been solid, normal beak, while the rest would have been more like stiffened skin grading into normal skin and feathers toward the back of the skull. At no point would the teeth have occupied the same physical space as the rhamphotheca. Basically, the rhamphotheca never seems to have housed tooth sockets. The beak and the teeth were segregated to different parts of the jaws. In short, no Mesozoic birds had "teeth in their beaks" as is often stated and depicted in art, but rather had both beaks and teeth, in different parts of the skull, and presumably serving different roles in food capture and processing.
-Elzanowski, A. (1991). "New observations on the skull of Hesperornis with reconstructions of the bony palate and otic region." Postilla, 207: 1-20.
-Heironymus, T.L. and Witmer, L.M. (2010). "Homology and evolution of avian compound rhamphothecae." The Auk, 127(3): 590-604.


  1. Cool.

    I started on a 3D Hesperornis a couple months ago (had to put it on the sideburner for now), and hit this question.

    Interesting I had no idea about the Heironymus and Witmer paper. It makes sense.

    I started to follow the "type specimen" art examples out there (though I do like Darren Naish refering to trends in Palaeo-art as Memes), as you noted they ALL have teeth in beaks. I was curious though if this made sense.

    Thank you for providing a technical backing up about my curiosity. I thought a beak AND teeth in the same spot would be both redundant and counter productive to the bird. Waste of energy to grow two fish catching mechanisms for the same space, and I couldn't see how the two together would help catch fish any better. Specializing one or the other makes way more sense.

    Great article interested to see what other people have to say.

    1. Hi, i want to know if you finishd the hesperornis in 3d
      sorry for my english

  2. Thanks for the discussion and the pdf link... Although Hesperornis and Ichthyornis were discovered here in Kansas, there's always been a question about how you can actually have "birds with teeth".... I'm pleased to see that someone took a serious look at it.

  3. It dawned on me after reading the first beak post that all the descriptions saying hesperornithines had "teeth in their beaks" were wrong. Good to see it covered in more detail! (I also went and corrected all my hesperornithine drawings, though I didn't add the compound rhamphotheca.)

  4. And you've changed the name of your website, I see. (Quick heads up, the thumbnails there don't show up, and nor do images directly linked from there.)

  5. Ouch, thanks for the heads up! I changed the folder system for my paleoart as well but obviously missed a few spots :) Was getting a bit tired of having to spell out the old name every time...

  6. "Teeth in their beaks" is a misnomer of a sort, but also apt, in another sort: Typically, "we" view the entire preorbital region of the skull as the "beak" or "bill" of the bird, rather than the rhamphothecal structure itself, or the toothless portion of a jaw.

    I am loath to respond to this in other ways, but this looks good anyways, matt!

  7. Interesting. Really drives in the point that you can't assume that what's true for the crown group is true for the stem.

  8. This comment has been removed by a blog administrator.

  9. It is good to see that we have finally agreed on this matter (the possibility of partially overlapping of these elements (tooth and beak, side by side)) as we discussed in the other forum.

  10. Nice one.....keep it up