|Edmontosaurus at the Denver Museum of Nature and Science wonders how you get by with just two sets of teeth.|
Tooth structure and dental batteriesYou may remember a few years back that Erickson et al. (2012) published on the structure of hadrosaurid teeth. To briefly recap, Erickson et al. (2012)'s interpretation was that hadrosaur teeth were made up of six different tissues: hard enamel and soft dentine (orthodentine; also known as dentin, just to keep us honest), the two basic reptilian tissues; equivalents of mammalian secondary dentine and coronal cementum; and two unique tissues, one composed of giant tubules found in filled pulp cavity branches, and a thick mantle dentine. These six tissues were found to be variably distributed in teeth, so that the teeth would work differently depending on how worn they were, and also were distributed differently between different hadrosaur lineages, giving different branches different capacities for attacking plants, such as teeth adapted for shearing or grinding. (My immediate question at the time was whether this could be used to distinguish different species; I don't think anyone's quite tried this, but on the other hand we've only got so many Trachodon teeth that maybe we don't want to subject them to destructive analysis.) Erickson and Zelenitsky (2014) followed up on this with detailed study of Hypacrosaurus stebingeri teeth, using specimens from the egg to the old-duckbills'-home. They found the teeth to change as the animal aged, pointing to changes in feeding and diet throughout life.
Recently, this model has been challenged. LeBlanc et al. (2016) reported that hadrosaur teeth did not include new unusual tissues. Instead, they found enamel, dentine, and plain old cementum, with some features taken as new tissues actually being modifications of other tissues, such as tubules being blood vessels trapped in rapidly growing dentine and cementum. The hadrosaur tooth at its most basic appears to be based around a shifting of tissues. Instead of enamel caps and cementum at the roots, enamel is shifted to one face and cementum to the opposite face. This pattern is similar to ever-growing incisors in certain mammals, such as lagomorophs and rodents (Bramble et al. 2017). These mammals may be better dental analogs for hadrosaurs than the usual large herbivorous mammals.
LeBlanc et al. (2016) and a follow-up, Bramble et al. (2017) interpret the structure of the teeth as part of the evolution of the dental battery, not (directly) dietary preference. For reference, the teeth of hadrosaurs are packed into dental batteries. A dental battery can have as many as 300 teeth, distributed in as many as 60 columns, with multiple teeth in use in any tooth position (LeBlanc et al. 2016). To have a functional hadrosaur dental battery, you need to prevent the simple "tooth in, tooth out" replacement cycle of most dinosaurs, you need to keep the teeth together, and you need to keep the teeth useful for their entire existence in the battery. According to LeBlanc et al. (2016), the first problem was solved by isolating the pulp cavity of each erupted tooth with a layer of dentine. The second problem was solved not by fusing the teeth together, as was previously thought, but by liberal application of periodontal ligaments, promoted by accelerated production of cementum. The ligamentous connections would have also allowed migration of tooth families within the battery in response to growth and uneven use (Bramble et al. 2017), as well as providing stress absorption (LeBlanc et al. 2016). Another consequence is that hadrosaurs would never have shed teeth, instead grinding them to nothing. This means that isolated fossil hadrosaur teeth would have come off of dental batteries post-mortem (LeBlanc et al. 2016; Bramble et al. 2017). The final problem was solved by filling the pulp cavity with more dentine. This killed the tooth but also made it possible to be ground down entirely, without the risk of pain or infection from an exposed pulp cavity (LeBlanc et al. 2016) (as well as giving hadrosaurs mouthfuls of zombie teeth).
|The quick tour of the business end of a duckbill, from LeBlanc et al. (2016) (Figure 1; full caption here). "a" shows a whole skull and jaws, "b" is a section through an upper dental battery, "c" is a lower dental battery as seen from the tongue's perspective, "d" is a section through a lower battery, "e" highlights the grinding surface of a lower dental battery, and "f" and "g" highlight the sediment between teeth, showing they are not cemented but were joined by soft tissue that is now lost.|
The downfall of pleurokinesisSpeculation, wild and otherwise, about the use of the hadrosaur dental battery has gone on more or less since the dental battery was recognized in the first place. If you're curious about the various historical proposals, I recommend the overviews in Weishampel (1984) and Nabavizadesh (2014). To make a long story short, after several fairly chaotic decades of proposals and counter-proposals, a consensus emerged in the 1980s around David Weishampel (1984)'s pleurokinesis model (kinesis in skulls being movement anywhere but the jaw joint). In this model, the skull of a hadrosaur is composed of several linked blocks, with varying degrees of motion available to them relative to the other blocks. When the hadrosaur closed its jaws, the maxillae would rotate outwards slightly against the lower jaws, among a group of other movements through the rest of the skull, and anything trapped between the upper and lower tooth rows would be ground by the opposing dental batteries.
|Just for reference. As before, the code is L=lacrimal; PMX=premaxilla; PO=postorbital; PRF=prefrontal; QJ=quadratojugal; SQ=squamosal; 1=naris; 3=orbital; 4=infratemporal fenestra|
Pleurokinesis fell out of favor in the first decade of the 2000s as it became clear that dinosaur skulls in general were not as kinetic as previously supposed (Holliday and Witmer 2008), and more sophisticated computer analysis showed that an unrealistic amount of movement would be needed across the entire skull to produce the expected movement of the maxillae, including separation of well over 1 cm between certain bones (Rybczynski et al. 2008). The rejection of pleurokinesis meant a new model was needed. One of the reasons it had become the consensus in the first place was that it satisfied a number of characteristics that could be seen in the skulls and teeth of hadrosaurs; in fact, as pleurokinesis was first foundering, a study of dental microwear was published that supported the model, or at least comparable motion (Williams et al. 2009). The answer appears to lie in the other part of the cranial equation: essentially, you can produce the same kind of wear if you make the lower jaws do the moving instead of the maxillae. The idea that the lower jaws could rotate along their long axis was first proposed early in the 20th century, but lost favor by the 1960s. It has made an impressive recovery, though; recent analyses of the problem of the less-than-flexible skull began circling this natural alternative almost immediately (Holliday and Witmer 2008; Bell et al. 2009). Cuthbertson et al. (2012) and Nabavizadeh (2014) have more fully articulated it.
With the lower jaw, there are really only three joints to worry about: the two jaw joints (left and right), which in hadrosaurs feature the quadrate of the upper jaw meeting the articular of the lower jaw; and the joint at the anterior end of the jaw where the two mandibles meet each other and the beak-forming predentary bone. It turns out that the hadrosaur jaw joint is nearly a ball-and-socket (quadrate ball into articular socket), which naturally permits mobility, and the anterior joint is pretty wishy-washy: the two mandibles are held in the embrace of the premaxilla, but nothing is fused or interdigitating or anything like that. Instead. the two mandibles sit on a broad surface of bone and were bound to each other and the predentary with soft tissues (Nabavizadeh 2014). Therefore, it certainly looks like we can get that great grinding action by keeping the upper dental batteries in place but rotating the mandibles. In practice, the full process would have incorporated palinal (fore and aft) movement as well as rotation (Nabavizadeh 2014).
Eating crustaceansMy graduate supervisor, Dr. Karen Chin, strikes again with another whodungit mystery. This one ties in with her previous work concerning woody material in Two Medicine Formation coprolites (Chin 2007). In this case (Chin et al. 2017), the fossil feces come from three horizons in the lower half of the middle Kaiparowits Formation of Utah, and feature abundant chopped-up bits of decayed conifer wood; as in the 2007 publication, the wood was pre-rotted by fungi, which makes the wood easier to digest. The Kaiparowits and Two Medicine coprolites also show evidence of the attentions of dung beetles and tiny snails. It is generally difficult to assign coprolites to any particular genus or species, but just as faunal and size considerations indicate that the Two Medicine coprolites were produced by Maiasaura, the Kaiparowits coprolites were probably produced by hadrosaurs, or perhaps large ceratopsians. A woody diet could have been in response to periods when other food wasn't available, or something that was sought for specific nutritional reasons...
The most noteworthy component of the coprolites was fragmented crustacean cuticle. The exact type of crustacean is not known, but they were big enough that a hadrosaur (or ceratopsian) would have noticed them. Given the presence of crustacean material in at least 10 coprolites at multiple stratigraphic levels, this was not a one-off activity driven by some wacky gryposaur with a taste for crayfish. Chin et al. noted that crustaceans and other invertebrates can be found today hiding out in rotted wood, so it could have been a two-for-one deal: crunch some wood, add a little protein. Whether the dinosaurs were specifically seeking out individual crustaceans, or just focusing their feeding in areas that could produce crustaceans, such as rotted wood, cannot be determined (and in truth there doesn't have to be an association between the rotted wood and the crustaceans; a dinosaur could have eaten wood in the morning and gone off after crustaceans in the afternoon, for example). Although hadrosaurs (and ceratopsians) are usually placed firmly among the herbivores, it is of course hardly unknown for herbivorous animals to add a little extra to their diets. Chin et al. postulated that crustaceans could have been sought as a dietary supplement by female dinosaurs during egg-laying, for additional calcium and protein. In this case, herbivorous mammals are poor analogues, given the scarcity of mammalian egg-layers, and we have very little to compare with among modern birds, which as a rather hard and fast rule do not get into the one-ton-or-greater range. We would be looking at a behavior which today only takes place at the smaller scale of modern herbivorous birds.
|The things you can find in dinosaur dung! This is Figure 2 from Chin et al. (2017) (go here for full caption). Parts "a" and "b" show woody material, while the rest show different aspects of crustacean remains.|
ReferencesBell, P. B., E. Snively, and L. Shychoski. 2009. A comparison of the jaw mechanics in hadrosaurid and ceratopsid dinosaurs using finite element analysis. The Anatomical Record 292(9):1338–1351.
Bramble, K., A. R. H. LeBlanc, D. O. Lamoureux, M. Wosik, and P. J. Currie. 2017. Histological evidence for a dynamic dental battery in hadrosaurid dinosaurs. Scientific Reports 7, article 15787. doi:10.1038/s41598-017-16056-3.
Chin, K. 2007. The paleobiological implications of herbivorous dinosaur coprolites from the Upper Cretaceous Two Medicine Formation of Montana: why eat wood? PALAIOS 22(5):554–556.
Chin, K., R. M. Feldmann, and J. N. Tashman. 2017. Consumption of crustaceans by megaherbivorous dinosaurs: dietary flexibility and dinosaur life history strategies. Scientific Reports 7, article 11163.
Cuthbertson, R. S., A. Tirabasso, N. Rybczynski, and R. B. Holmes. 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.
Erickson, G. M., and D. K. Zelenitsky. 2014. Osteohistology and occlusal morphology of Hypacrosaurus stebingeri teeth throughout ontogeny with comments on wear-induced form and function. Pages 422–432 in D. Evans and D. Eberth, editors. Hadrosaurs. Indiana University Press, Bloomington, Indiana.
Erickson, G. M., B. A. Krick, M. Hamilton, G. R. Bourne, M. A. Norell, E. Lilleodden, and W. G. Sawyer. 2012. Complex dental structure and wear biomechanics in hadrosaurid dinosaurs. Science 338(6103):98–101.
Holliday, C. M., and L. M. Witmer. 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.
LeBlanc, A. R. H., R. R. Reisz, D. C. Evans, and A. M. Bailleul. 2016. Ontogeny reveals function and evolution of the hadrosaurid dinosaur dental battery. BMC Evolutionary Biology 16:152. doi:10.1186/s12862-016-0721-1.
Nabavizadeh, A. 2014. Hadrosauroid jaw mechanics and the functional significance of the predentary bone. Pages 467–481 in D. Evans and D. Eberth, editors. Hadrosaurs. Indiana University Press, Bloomington, Indiana.
Rybczynski, N., A. Tirabasso, P. Bloskie, R. Cuthbertson, and C. Holliday. 2008. A three-dimensional animation model of Edmontosaurus (Hadrosauridae) for testing chewing hypotheses. Palaeontologica Electronica 11(9A).
Weishampel, D. 1984. Evolution of jaw mechanics in ornithopod dinosaurs. Advances in Anatomy Embryology and Cell Biology 87:1–109.
Williams, V. S., P. M. Barrett, and M. A. Purnell. 2009. Quantitative analysis of dental microwear in hadrosaurid dinosaurs, and the implications for hypotheses of jaw mechanics and feeding. Proceedings of the National Academy of Sciences 106(27):11194–11199.