Support for a Lipless, Cheekless Dinosaur World

First, let me set up this discussion with an image:

Snout cross-sections and tissue analogues and arrangement.

Sure you’ve seen this one before. A cross-section through the skull of Velociraptor mongoliensis, with some extreme simplification. Red indicates gum tissue, peach indicates skin (with volume from fats), blue indicates the oral ligaments, and yellow the labial scales; in the bird, light orange indicates rhamphotheca. Two models are drawn from paleoart, where “Dinosaur” indicates the “standard” model where the upper teeth overlay the dentition of the lower jaw, and were slightly covered in “lips,” but covered the “lips” of the lower jaw; the second in a reconstruction from Greg Paul’s “Raptors,” in which the labial scales are removed from the “Lizard” model, owing as it does to Paul’s general reluctance to illustrate facial squamation; this bears the caveat that Paul exposed the teeth of the jaw at the rostral end. I argued in that post that I felt the extant phylogenetic bracket favored the “Lizard” model for dinosaurs. Here, I attempt to affirm that argument.

[Note: I heed Ashley Morhardt’s caution in the first “lip” post that overuse of terms like “lip” and “cheek” for pretty much anything around the mouths of fossil animals is something to be avoided. These terms typically apply to mammals, and no other animal, especially reptiles — of which dinosaurs (including birds) and crocs make a part of  — has these precise tissues; when analogues exist, these should be readily differentiated, and when speculation is used, this should be pointed out. I have adjusted my terminology to follow this caution, and so should you.]

Extraoral tissues in extinct animals are difficult to assume beyond direct observation. To project what might be possible, we have to use the extant phylogenetic bracket (Witmer, 1995), and thus use the relation of similar taxa, or the placement of these taxa among close relatives. But we can also use direct tissue analogues, features of the bones or preserved soft tissue, and infer the structure of tissues involved.

This includes: Musculature, ligaments, presence of rhamphothecae or cornified skin. Some types of nonbony cranial tissues, especially integument like rhamphothecae and cornified skin, and bosses such as the nasal horns of rhinoceratids and the frontal/parietal horns of various ruminants, are directly associated with bone and thus leave markers on it (Hieronymous et al., 2009):

From Hieronymus & Witmer, 2007.

Integument is relatively easy to assess for extant, recently extinct, or securely bracketed taxa (such as fossil mammals), but it is less secure when the subjects are very, very distinct. Missing from the data above, but related to cornified sheaths, is the analogue on bone for avian rhamphotheca: Hieronymus & Witmer (2010) identified numerous analogues among avian rhamphotheca, but as of yet have not integrated the two papers. Yet, it seems relatively simple: rhamphothecae are often quite closely associated with the underlying bone, but often becomes very thin and extends over relatively featureless bone at its caudal limits. Where rhamphothecae are closely associated, though, dense and anastomosing nutrient grooves abound, the bone surface is smooth rather than pitted or pectinate, and foramina are numerous and oblique, as they are concentrated running through the cortical and upper trabecular bone.

Bird skulls lack features of integument they are not expected to have, which helps underscore the predictive power of Hieronymous and Witmer, while they do possess found tissues: rhamphotheca, smooth skin, rugose bosses produced from thick, cornified skin. What can be inferred from fossil crurotarsans/pseudosuchians and dinosaurs, bracketed as they are for known crocodilians and birds, is a trickier issue: No living archosaur preserves “lips,” or in fact any form of free-moving ligamentous structures, and this is the easiest form of “lips” we can project onto these taxa.

Hypothetical arrangements of extraoral tissues on a standardized archosaur-line diapsid, which is not meant to depict any particular group. For the first image, muscles in red represent the circum-oral musculature of mammals, while the large solid red represents the masseter of mammals and psuedomasseter of parrots; pink represents the supraoral muscles which “lift” the muscle bands. For the fourth image, “rhamphotheca & teeth” only mesial loss of dentition of both jaws is shown, however the complex of possible arrangements of teeth and rhamphotheca are more variable than would be useful to show in this figure — it would dominate the described possibilities.

I’ve been spending some time looking at what fossils can infer for their tissues, to answer the question of whether dinosaurs had lips, or cheeks, or similar structures.

In 1997 and 1998, Michael Papp and Larry Witmer presented at the Society of Vertebrate Paleontology on the possibility of a lack of any sort of buccal soft-tissue in Leptoceratops gracilis, replaced instead by a possible rhamphothecal-like “beak” that would progress on the former “cheeky” jaw margins.

Illustration of Leptoceratops gracilis in the bone, and in the flesh, from Bill Parsons.

Kelli Whitlock talked to Larry Witmer after his presentation when the information was publishedpublicized. Said Witmer: “One of the things I like to take pride in is that this theory has the possibility of being proved wrong[….] I like to think it can survive testing, but at least it can be tested.” I am inclined to think that Witmer was, in fact, correct, although this reconstruction doesn’t work — in this form — for all ornithischians.

There is a strong effort to reconcile this research with the unusual jaws of ankylosaurian and even stegosaurian ornithischians, in which the dental row is somewhat inset (but not by much). However, while I would be cautious on broad extrapolation, I would like to note that for those who find the idea of “cheeks” useful for ornithischians, direct evidence should come first. Direct tissue homologies, extrapolation of the extant phylogenetic bracket, and functional considerations of the jaws should be first, foremost, and only in the gaps this information doesn’t cover should speculation enter the framework. This is the case for Panoplosaurus mirus, in which not only do the dental rows seem strongly inset (they are not) but lateral expansion of the nasopharyngeal sinuses (forming extreme loops; Witmer & Ridgely, 2008) causes the lateral margins of the skull to expand beyond the lateral extend of the mandible at the same part. This has led to a subjective impression that there was a distinctive, broad cheeky overhand.

Skull of Panoplosaurus mirus scanned and with internal sinues, nasopharyngeal passages and endocranium filled. From Witmer & Ridgely, 2008.

The main culprit is the presence of thin, vertically oriented bones in the “buccal cavity” of the skull, in this species and in others.

Skull of CMN 2759, holotype of Panoplosaurus mirus Lambe, 1919. Left lateral view, unlike the right view, preserves large plates in the buccal region. From Wikimedia Commons.

If there was no “cheek,” what is this thing doing in the skull? What is it attached to? And precisely, what is its relation to extraoral tissues surrounding the jaw? While it has largely been assumed that these ossicles supported a full buccal tissue that expanded from the skull to the mandible, no such tissue homologue exists except in mammals and a few (very few) birds, namely parrots (which have a unique muscle, m. pseudomasseter), as well as this odd bird:

Gymnogyps californianus, the California condor, sporting a sleek new pair of “cheeks”. From and copyright of Sonoma State University, California.

But this last feature doesn’t involve muscle, and there is no mechanical function known for such a tissue, making me suspect it serves a visual feature, either for species or sexual recognition. New world vultures, generally, have small rhamphothecae, which means that much of the oral margins are covered instead by skin and/or feathers.

I am thinking then that the “cheeks” of vultures and parrots are red herrings; they do not help us figure out what tissues were present, merely they they could be, if we were to be very, very loose on what our analogues were. To propose there is a tissue there would require more direct evidence of its presence, muscle or otherwise. This data does not exist, despite the bone, the inset dentition, and the “appearance” there must be a cheek.

What about “lips”? Well, this is where things become grounded, but it is sure to disappoint: Modern lizards lack mobile lip-like structures (remember, we’re using “lips” to refer only to the mammalian sort, not the muscle-less, sessile sort in reptiles), instead preserving a sessile, mainly ligamentous tissue (see the images up top) which connective tissue constrains the shape to the margins of the fleshy jaw. This is normally fine, except when two things become involved: loss of functional dentition, and mobility of bones. In the latter, snakes lack extensive muscles of the oral margin, and in fact lack them around the jaw, and mostly preserve thin muscles connecting the major mobile bones. The ability of snakes to distend and expand the features of their “lips” are largely powered by pressure from behind by the movement of the maxilla and palatal bones; during ingestion of large prey, the jaws largely expand outside of the maximum expansion of the overlying ligaments surrounding the jaw margins, and as such, are not involved in the mobility of said “snake lips.” This leaves loss of functional dentition:

Skull and jaw of UCA.5, Uromastyx hardwickii. Modified from and copyright of DigiMorph.

The spinty-tailed lizards, or mastigures, or Uromastyx sp., are a group of Agama-like iguanian lizards adapted to dry habitats. Like many other Agama-like lizards, including chameleons, the dentition (which is acrodont) fuses early on to the jaw, forming an apparent saw-like edge of the underlying bone. However, dental tissue remains distinct. Rostrally, the premaxilla bears an apparent bony “beak;” early in ontogeny, this region of the jaw is occupied by small teeth (Cooper & Poole, 2009), which are successively lost, while development of the premaxillary bone becomes “beak like.” However, no rhamphotheca is formed, and the hard beak-like structure remains encased in soft-tissue. In Sphenodon punctatus, the tuatara, the rostral teeth of the upper jaw are retained in ontogeny (e.g., see Jones, 2008), and progressively enlarge and become fused to the upper jaw, forming a “psuedobeak” that has been positively compared to rhynchosaurs. Instead, a lip surrounds the upper jaw of both groups of lacertilian.

The tuatara (Sphenodon punctatus), courtesy SidPix from Flickr, used under a CC BY license, and Uromastyx maliensis, from kingsnake.com’s Uromastyx forum. Extraoral tissue does not compensate for the rostral toothless “beaks” in both groups.

Interestingly, Uromastyx lizards are almost exclusively herbivorous, which trend toward omnivory; tuataras, in contrast, are almost exclusively carnivores, and it doesn’t help that tuataras also exhibit a large rostral mandibular crown, rather caniniform in aspect.

Beak-like structures also exist in fish, especially parrotfishes (Scaridae) and puffer and porcupine fishes (Tetraodontiformes), which display some amazing similarities not shared with their closest relatives. Of course, the functional similarities are convergent, as the “beaked” taxa are not, in fact, very closely related. Regardless, the large distinction here with these “beaked” species mentioned, these fish lack an actual beak; instead, their “beaks” are formed from their teeth (Fraser et al., 2012). In some taxa, this forms a broad plate crossing the entire functional mandible or upper jaw, split in the middle, thus giving the appearance of only four functional teeth, total. The margins between tooth and bone can be so slim that it appears as though the teeth and jaws have become one.

Putting Everything Together

This won’t be the last word on the subject, and I have more to say. It is my understanding, from these lines of evidence, that soft tissue correlates prefer a lacertilian form of integument, even in ornithischians. It is so important to emphasize the need to look into tissue morphology and relationship before one speculates that certain structures are “true” that I’ve been illustrating Mesozoic reptiles “in the flesh” for the last 12 years now with the warning — in HUGE FLASHING RED LIGHTS  — that these reconstructions are hypothetical. But, moreover, that research here, unpublished, and in process, does not support the presence of buccal tissues crossing the maxilla to the mandible. I’ve been drawing ornithischians like this —

Goyocephale lattimorei, a basal pachycephalosaur for which integument is unknown, but which can be inferred from Tianyulong confuciusi and Psittacosaurus sp., and is here favorable to the latter.

— with the understanding that I am entirely wrong! Functional considerations of the lack of buccal tissues in ornithischians is one issue involved in this, and is what I will be getting to with the next few posts on this subject. This will include “chewing” in hadrosaurs, and the functional considerations of a cheekless ankylosaur. I remain fascinated in extending this to taxa such as rhynchosaurs, dicynodonts, and so forth, especially when whopping huge teeth in taxa like Epidexipteryx hui (featured here) and Suminia getmanovi seem like they should have use swaths of extraoral tissue. Stay tuned.

Bellwood, D. R. 1994. A phylogenetic study of the parrotfishes family Scaridae (Pisces: Labroidei), with a revision of genera. Records of the Australian Museum, Supplement 20:1-86.
Cooper, D. F. G. & Poole, J. S. 2009. The dentition and dental tissues of the agamid lizard, Uromastyx. Journal of Zoology 169(1):85-100.
Fraser, G. J., Britz, R., Hall, A., Johanson, Z. & Smith, M. M. 2012. Replacing the first-generation dentition in pufferfish with a unique beak. Proceedings of the National Academy of Sciences, Philadelphia 109(21):8179-8184.
Hieronymus, T. L. & Witmer, L. M. 2010. Homology and evolution of avian compound rhamphothecae. Auk 127:590–604. [PDF link]
Hieronymus, T. L., Witmer, L. M., Tanke, D. H. & Currie, P. J. 2009. The facial integument of centrosaurine ceratopsids: morphological and histological correlates of novel skin structures. The Anatomical Record 292:1370–1396. [PDF link]
Jones, M. E. H. 2008. Skull shape and feeding strategy in Sphenodon and other Rhynchocephalia (Diapsida: Lepidosauria). Journal of Morphology 269:945–966.
Witmer, L. M. & Ridgely, R. C. 2008. The paranasal air sinuses of predatory and armored dinosaurs (Archosauria: Theropoda and Arnkylosauria) and their contribution to cephalic architecture. The Anatomical Record 291:1362-1388. [PDF link]