In my last post on the issue of recovering the probable structure of extra-oral tissues in dinosaurs, bracketed as they are between crocs and birds, but with lacertilians and other reptiles outside of that grouping, I suggested that one important direction this research must take is that it must account for ornithischians. Specifically, I said that Larry Witmer was correct in his assumption that Leptoceratops gracilis likely had nothing “cheek-like” around its jaws. I attempt here to explain this in more detail.
Bust of the Morrison “polacanthine” ankylosaurian Gargoyleosaurus parkpinorum.
First, tissues leave marks on bone, especially when they attach to said bone. Keratinous sheets lying close leave dense networks of nutrient grooves on bone, which is normally smooth; the tendons that link muscle to bone leave spiky bone surfaces in small, regular regions and generally in relief of the surrounding cortical surface, due to Sharpey’s fibers; bosses and cornified skin leave rough surfaces with dense foramina penetrating the bone; ligaments leave striae and raised lips with minute rough surfaces; and dental endothelium (or “gum”) leaves smooth bone without irregularities.
When it comes to discussing the option of whether any ornithischian dinosaur had a muscular or nonmuscular flap between upper and lower jaw, it must first be known what tissues were present in the jaw region. I will not speak with authority on this point, as I’ve not undertaken a direct examination of jaws of any variety of ornithischians by which I could say they LACK such markers on them, yet only that so far, no markers have been reported by systematists and paleontologists describing the skulls of dozens upon dozens of diverse hadrosaur, ceratopsian and ankylosaur or stegosaur jaw.
A histological or surficial map has yet to be made of any ornithischian jaw, or even to compare described regions among taxa, instead focusing research on ultrastructure and minute details of particular regions. What I’m covering next is instead partly mechanical and partly based on personal observation, based on what I’ve had the opportunity, with what my mere eyeball has had the opportunity to examine a variety of ornithischian skulls in person or in paper. I am leaving unpublished research out of this.
So, tissue distribution:
Cross-sections through the rostral maxilla and dentary in three “cheeked” ornithischian clades – Hadrosauroidea, Edmontosaurus annectens; Ankylosauridae, Euoplocephalus tutus; and Ceratopsidae, Centrosaurus apertus. Letters refer to the three possible models, with all taxa showing sessile, lacertilian “lips” (A), and a muscle-less, fleshy “cheek” (B), with the additional model of an ankylosaur (C, based on Panoplosaurus mirus) with a buccal plate (B and C for Euoplocephalus tutus only) butno “cheek”. Euoplocephalus tutus follows Coombs, 1978 and inferences from Witmer & Ridgely, 2008, while Edmontosaurus annectens and Centrosaurus apertus follow inferences from Witmer & Ridgely, 2008, Bell et al., 2009 and Cuthbertson et al., 2012. C follows a suggestion from Tracy Ford. Black, bone; peach: flesh and fats; red, gum; orange, cornified skin/pad or rhamphotheca; blue oral ligaments. Grey with highlights represents the nasopharyngeal passage and sinuses.
It was interesting to me to look for examples where a clear and distinct lamina or ridge exists that separates the lateral surface of the dentary into two faces: a ventrolateral face, and a buccal face. The surficial structure of the buccal face tends to be smooth, but bears large foramina; conversely, the ventrolateral face bears numerous nutrient grooves and appears with a rougher texture. Is this the boundary of a buccal cavity? I argue here that regardless of the presence of sessile, lacertilian-style oral tissues (model A) or a flap of skin (model B), the lateral margin of the tooth row and an extended region of the lateral surface further ventral would be covered in gum tissue, and thus this surface alone might not allow a covering “cheek” analogue. Problematically, not all ornithischians possess a distinct margin between these two faces, so that it may be inferred easily.
In Centrosaurus apertus (following Bell et al., 2009), the margins are distinct, and the faces demarcated by a clear ridge that crosses from the coronoid caudally to essentially merge with the lateral ridge of the predentary. This has other implications, but we won’t get into that here. In Edmontosaurus annectens, rather, the lateral margin of the dentary is more rounded, and the “faces” are far less distinct from one another: the jaw seems smooth all the way down (using Bell et al., 2009, who in turn used an indeterminate lambeosaurine mandible from the Dinosaur Park Formation, meaning it could be all sorts of things).
One thing I noticed (based now largely on discussions with others and thus not my own conclusions, but for which I don’t have a citation handy — if you have the cite, let me know!) is that the presence of a distinct separation of faces may be apparent only as an artifact of rotation of the mandibular ramus about its long axis:
Ovals representing a coronal section through a mandibular ramus, here the dentary (with a triangular tooth on top for orientation). A, with the adjusted long and short main axes of the section, and grading through 20 degrees of rotation, then another 20 degrees. The tooth position maintains position on the dorsal margin of the jaw, which adjusts labially or lingually depending on ramus rotation. B, ramus section with labial thickening (representing the ridge, indicate by small red arrows). Notice the greater demarcation of “faces” when the mandible is merely rotated. C, paired mandibular sections with apical teeth not oriented while mandibular rami are, maintaining lingual distance during rotation and thus forming a serially proportionately narrower oral cavity and larger buccal cavity without producing a distinct ridge. White sections represent Meckel’s canal.
In many ways, the depth of the buccal “cavity” is a product of the rotation of the mandible, and the ventral width relative to the dorsal. And in these taxa, the ventral margin may be thicker than the dorsal, causing rotation to “form” a “buccal ridge.” This ridge then becomes a feature that appears firstly out of rotation, secondly out of morphology (i.e., a lamina develops from an extension of a coronoid process, or caudally from the predentary region). Hadrosaurs, perhaps as a product of the extreme roundness of the lateral surface — nonetheless, bulging outwards — have the most vertical of mandibular rami shown here, followed by ceratopsians which have rotated the ramus lingually by about 15°; ankylosaurians have mandibular rami which are nearly 40° from the vertical, and consequently appear to have the deepest buccal cavity, but these appear to be a factor related not to actual osteological features but to the bone’s orientation. Perhaps a greater clue to the presence of buccal “depth” would be from the lateral margins of the jugal and maxilla, instead: in hadrosaurs, however, this creature is still rounded, and shallow, while in ceratopsians and ankylosaurs a distinct, sharp margin is produced and in the latter forms an overhanging shelf. Does this necessarily force a presence of “cheeks”? No. For certainly, these features also appear in lacertilians.
UCA.5, skull of Uromastyx hardwickii, in lateral view; coronal sections taken at the red lines. Blue arrows indicate “buccal cavities.” Uromastyx lizards do not have “cheeks”, but rather sessile “lips”. Modified from and copyright to Digimorph.
Jaw mechanics of ankylosaurids are limited. The skulls are akinetic, but apparently the mandibles are not (Vickaryous & Rybczynski, 2001), permitting flexibility for each mandibular ramus relative to the other, due to the open joint between each dentary and colateral cotylar sockets on the posterior predentary. This affords the ability of the jaws to engage in a complex array, despite restrictions due to the lingual curvature of the tooth row in all taxa.
One of the trickier issues, as I mentioned in the previous post on this topic, was that some ankylosaurians possess a bony plate lateral to the buccal cavity, a buccal plate, and that this has naturally caused some to suspect this plate affirms a fleshy “cheek” of some sort extended across the plate’s lingual surface. I question this for two reasons, barring any new data that will reveal that it can be accomodated:
1. The presence of an elastic structure of any sort or flexible region of skin lingual to the plate would only be effective around the margins of the plate, and cannot necessarily incorporate it. If it did, the plate would need to be weakly attached to the skin in this area to prevent distortion of the tissue around it. Plates in this region are also unintuitive to the function of the “cheek”: opening and closing would provide the flesh required flexibility, and it can be assumed that dermal plates should form in a region where the skin becomes stiff and requires either reinforcement, or are migrated to protect a point of weakness, but develop elsewhere.
2. “Cheeks” — as they were — are not needed for the ostensible role they are implied to be required for: retaining food within the oral cavity.
A model for extra oral tissues use in an ankylosaurid skull. A, jaws at rest; B, jaws in abduction, with mandibular rami rotated labially (outward); C, jaws in adduction, with mandibular rami rotated lingually (inward). 2 in B and C represent the inclusion of a bolus (deep orange) which here is sized to roughly be the diameter of the rostral dental row, which the caveat that the bolus can be large (as wide as the caudal tooth row, or rostral beak.
Depending on the elasticity of the mandibular tissues, the bolus in the jaw is largely contained. This model does not ensure that no loss of food during oral processing occurs, but this model does consider that virtually all volume of the bolus remains within the jaws. Moreover, the bulk of the material may dump into the floor of the mouth between mandibles. You will notice however that this is based on the absence of a tongue; I’ve not forgotten it, it is merely absent based on lack of data suggesting its size, shape, or possible extent. It should also be notable lizards can process food and often incur slippage out of the sides of the mouth, having nothing in the way of a “cheeky” tissue, yet manage to get along fine, so whether the actual — and likely low — volume of lost food is relevant remains to be seen. How efficient was the food stuffs of particular taxa for their environment, and how long they must feed, and how effective they are in processing those foods? These questions are generally pointed at sauropods; ornithischians less so.
Problematic with this model is in fact the issue of bolus retention, however, depending on the tensility of the extra-oral tissues, and their extent, this can be somewhat mitigated without having to involve novel tissues for which no direct evidence exists. The plate on the side of the jaw, and its suggestion of a full “cheek” is useful in helping resolve this, but not only has it not been preserved in many ankylosaurians, it has been preserved only in some nodosaurids, while not even the fantastic bony preservation of the Gobi Desert has permitted preservation in various Mongolian ankylosaurids and “shamosaurs”. Lack of preservation of a thing, of course, does not mean preservation of that thing’s lack.
Tracy Ford suggested an odd compromise:
Two schemes: A fleshy, true-lip-less model in Ankylosaurus magniventris, and a true-cheek-less Panoplosaurus mirus with buccal plate intact as a unit with the mandible. From Tracy Ford, used with permission. Clicking on this image brings you to his website.
Are both of these mechanisms possible? Yes, and certainly I think they are more likely than one in which a full fleshy “cheek” is present. There are, of course, possibilities (I say, waffling) that some ornithischians had developed a fleshy “cheek,” almost certainly one not comprised of muscle, while others did not. Ornithischia is a complex clade, and its backbone was a series of lineages all based on small, bipedal taxa with jaws full of teeth and no real distinct inset of the jaws that might even suggest a “cheek,” yet has produced three clades all convergent on the lingually-rotated mandibular ramus: thyreophorans, in which ankylosaurs (and especially ankylosaurids) have gone leagues beyond the condition seen in Scelidosaurus harrisonii or stegosaurians; ceratopsians; and iguanodontians. Yet desite this convergence, the relative buccal cavity to the oral cavity width doesn’t seem to have increased by much in a given lineage. While Coombs (1978) shows how Sauropelta edwardsorum had remarkably inset maxillary tooth rows, this inset variation hasn’t been strongly tested, and we are left wondering if the inset really did remain stable across Ornithischia when it becomes this inset. No matter; another thing for researchers to look into.
However, I would like to indicate that, as with a variety of herbivorous lizards, including those with relatively complex jaw movements but somewhat akinetic skulls (as with Uromastyx hardwickii; Throckmorton, 1978, Schwenk, 2000), many herbivorous ornithischian dinosaurs also appear to have akinetic or largely akinetic skulls and complexly mobile mandibles (Holliday & Witmer, 2008, but notably – hadrosaurids: Cuthbertson et al., 2012; ankylosaurids: Vickaryous & Rybczynski, 2001; and ceratopsians: Haas, 1955). If the lack of a “cheek” is no limit to the diverse but herbivorous diet of uromastyx lizards, it should be no issue for larger herbivores.
Here, I present four variable reconstructions of the head of Panoplosaurus mirus as it might appear in life, with sessile, lacertilian-style “lips;” a fleshy, boneless “cheek”; a fleshy “cheek” in which the bone plate has been placed and anchored to the mandible; and finally, following Tracy Ford’s idea, one in which the bony plate operates as an extension of the mandibular extra-oral tissues and not as part of a “cheek.” I leave it to the reader to consider which option is best, fitting sense or the data.
Bell, P. R., Snively, E. & Shychoski, L. 2009. A comparison of the jaw mechanics in hadrosaurid and ceratopsid dinosaurs using Finite Element Analysis. The Anatomical Record 292(9):1338-1351.
Coombs, W. P., Jr. 1978. The families of the ornithischian dinosaur order Ankylosauria. Palaeontology 21:143-170.
Cuthbertson, R. S., Tirabasso, A., Rybczynski, N. & Holmes, R. B. 2012. Kinetic limitations of intracranial joints in Brachylophosaurus canadensis and Edmontosaurus regalis (Dinosauria: Hadrosauridae), and their implications for the chewing mechanics of hadrosaurids. The Anatomical Record 295(6):968-979.
Haas, G. 1955. The jaw musculature in Protoceratops and in other ceratopsians. American Museum Novitates 1729:1-24. [PDF]
Holliday, C. M. & Witmer, L. M. 2008. Cranial kinesis in dinosaurs: Intracranial joints, protractor muscles, and their significance for cranial evolution and function in diapsids. Journal of Vertebrate Paleontology 28(4):1073-1088. [PDF]
Rybczynski, N. & Vickaryous, M. K. 2001. Evidence of complex jaw movement in the Late Cretaceous ankylosaurid Euoplocephalus tutus (Dinosauria: Thyreophora). pp.299-317 in Carpenter (ed.) The Armored Dinosaurs. (Indiana University Press, Bloomington & Indiana.)
Schwenk, K. 2000. Feeding in lepidosaurs. pp.175–291 in Schwenk (ed.) Feeding: Form, Function, and Evolution in Vertebrates. (Academic Press, New York.)
Throckmorton, G. S. 1976. Oral food processing in two herbivorous lizards, Iguana iguana (Iguanidae) and Uromastyx aegyptius (Agamidae). Journal of Morphology 148:363-390.
Witmer, L. M. & Ridgely, R. c. 2008. The paranasal air sinuses of predatory and armored dinosaurs (Archosauria: Theropoda and Ankylosauria) and their contribution to cephalic structure. The Anatomical Record 291(11):1362-1388. [PDF]
Jaime A. Headden
The Bite Stuff 
Are you suggesting that the “cheek” osteoderm in Panoplosaurus may have been displaced from elsewhere on the body?
Not quite. I suggest they may come from the same place, but may also actually be further lingual than they should be. I associated them in the reconstruction from further ventral and largely adjacent to the lateral mandibular osteoderms. Thus, they are adjacent ossifications of the dermis, and not broadly separated as has been somewhat suggested. Yes, I show a “cheek”less nodosaur as Panoplosaurus mirus without the plates, but this may also be the case in OTHER ankylosaurs, of which only some preserve plates. I make no claim that it is the most accurate one, but should be the null premise, which we made add a buccal plate (fourth image at the bottom) rather than a “cheek”.
Surely if the mastication process itself was intended to keep the bolus within the mouth, the animal would exhibit an underbite, no? The overbite arrangement seems suited for shearing stuff off and letting gravity do its work, dropping the smaller pieces or sorting them into the “processed” inbox created by the cheek/equivalent structure. Unless that beautiful battery of teeth found in Hadrosaurs was solely for the purpose of ensuring they didn’t swallow whole anything wider than their gullet, I don’t see how that would work at all.
Forgive my ignorance of Uromastyx lizards but how do they combat the problem of extra-oral droppage? Do they let the excess drop or do they employ a similar technique to my cats: the head-tilted free-falling chew?
Let me start by noting that cats, like most carnivorans, have by extension of being carnivorans, have a reduced masseter in comparison to, say, a ruminant. Or rodent. Definitely a rodent. Carnassial-bearing animals use precise occlusion and force meat-processing to one side of the jaw, and thus typically tilt their heads to “chew”; this is the same behavior a cat takes when gnawing on almost anything, or like a dog on your shoe (I don’t have a dog; I don’t know). These teeth do work with the upper set distinctly wider than the lower set of teeth, but with angles of shearing edges and premolar occlusion offset, making the jaws require slight “disarticulation” of one side to occlude the other.
Ruminants have bilateral occlusion, in which both sides of the jaws are in perfect occlusion at the same time. This also results in some of the more grazer-oriented taxa nearly flat nolars with transversal ridges, and oral processing occurs through bilateral translation rather than unilateral orthal “sawing”, as in cats. They “grind” or “chew” their food from side to side, which is a good reason to have substantive cheeks (or rather, buccal cavities bounced by cheeks) to retain food. The same is also true of rodents.
Carnivorans are inefficient oral processors: their jaw anatomy is suited to merely cutting up food into bits small enough to swallow — the gut does the rest of the work. This is true generally of browsing herbivores as well, which utilize their teeth to pulverize or render food into small parts, but not process it in lieu of the gut … not so much as grazers do. So browsers can afford less efficient teeth, grazers not, and the more complex the tooth or row, the less efficient the oral processing, such as development of unilateral precise occlusion.
It would be keen of me to describe at this point the morphology of the tooth rows in hadrosaurs, ceratopsians, or ankylosaurs, but references I’ve linked to (including some with pdfs) actually do this for most of these taxa. I will abstract that by saying that these taxa have lower teeth that are generally lingual to the upper, and occlude inside the upper; moreover, all groups are capable and in some cases specialized for unilateral precise occlusion, rather than bilaterally. This tells me that they will favor jaw mechanics that control the bolus’ position in the jaw better than those whose function is to re-render continuously a small portion of the food in the oral cavity. Thus, there is less need to retain the bolus in the mouth. It is about rendering food as quickly as possible. This may also be why the oral cavity in hadrosaurs and ceratopsians tend to be so narrow at the front, or completely: food is not meant to be retained. These animals should thus not require substantive time spent “keeping food in”: if it can be chopped small enough (and chop indeed is the operative mechanic) then swallowing is the next course.
As for Uromastyx lizards, this appears to be the same as with most agamids: acquire, chop up, and swallow … let the gut do the processing. I will have to check my references, but my understanding is that Uromastyx have a substantially larger body relative to head size than most agamids, relating to the massive gut due to its extremely herbivorous diet.
You know, a thought occurs. The hadrosaur Edmontosaurus is known from like four or five different mummified individuals. Wouldn’t it be easy enough to take a look at the preserved skin around those specimens to see whether or not cheeks are present?
This is one area where I would love to be corrected on, this example of disconfirmation. However, as far as I know, of the known hadrosaur mummies and even the few ankylosaur “pseudomummies” (including specimens of Panoplosaurus mirus, like the one mentioned this post and in this one), none of them preserve the buccal region. Part of this is may be due to preparation, as some of the more famous hadrosaur mummies were collected for the AMNH, and in that time exposure of the dentition presumed important beyond preservation of what may never have been identified as “cheek” material. Since then, mummies like “Leonardo” have been found, but sadly lacking material on the cheeks, including whether they had “lips” — assumed to be a preservational thing, and not disconfirmation by saying they had NO integument around the buccal region. So, unfortunately, mummies tell us little on this at the moment.
A second line of evidence can also come from lagerstätten, in which carbonized traces or even impressions of the integument can occur which can reveal whether they had “lips” or “cheek” tissue. I have to survey the fossils of the various preserving facies in which ornithischians, lizards, and such occur to see whether they preserve the relevant regions to consider the question further, however. This did not occur before this more theoretical set of observations. Testing them is the primary goal, but first I must make the proposition.
Very interesting about Uromastyx. Seems “deeply inset toothrows indicate cheeks” is yet another functional myth perpetrated by those unfamiliar with living taxa.
Pingback: The Enfluffening | The Bite Stuff
Pingback: All Your Guessing Games | The Bite Stuff
Pingback: Facial Expressions | The Bite Stuff
Which ankylosaur species have been found with these buccal plates? Would Pinacosaurus have had them?
Only some nodosaurid specimens have been recovered with them. Edmontonia spp. and Panoplosaurus spp. as I recall. I do not have a definitive list.
Okay, thanks for the info!