'Unspecialised' dinosaur herbivores: not so boring after all

One of the central tenets of palaeobiology is that similar looking skeletal structures in different taxa convey similar functions in life. Hence, the presence of serrated teeth, like those of extant carnivorous varanid lizards, imply carnivory in theropods, and the convergent acquisition of long, graceful lower legs in gazelles and ornithopods suggests cursoriality in the latter. While some of these form/function relationships have proved relatively robust to quantitative, experimental testing, the generality of several classical form/function comparisons has been questioned by recent work. For example, experimental studies on living teleost fish have shown that skeletal morphology alone does not predict jaw movements: predictions made on bones alone fail and real jaw movements could only be deduced when soft tissues and nervous control mechanisms were factored in (e.g. Lauder 1995). These, and other similar studies, have shown that we should no longer rely uncritically on simple form/function correlations, but should test these assumptions through experiment or modelling. This will allow us to avoid erroneous functional predications that would otherwise resonate through ecological reconstructions and discussions of homology, as well as influencing other functional work.

Thanks to the development of new and refined experimental methods, as well as sophisticated computer modelling techniques, we are now in a position where we can test at least some of the mechanical properties of fossil skeletons (and of living tissues) in ways far more rigorous than the early comparative anatomists could have imagined. With this in mind, my colleagues Stephan Lautenschlager, Charlotte Brassey, David Button and I decided to look at the skull function of three different herbivorous dinosaurs to investigate some aspects of the form/function question.

We selected skulls of the Late Cretaceous therizinosaurian theropod Erlikosaurus (the subject of Stephan’s PhD), the Late Triassic sauropodomorph Plateosaurus (from David’s PhD thesis), and the Late Jurassic thyreophoran Stegosaurus (based on the complete, but disarticulated skull of ‘Sophie’ the NHM’s new specimen, which was CT scanned and reconstructed virtually by Charlotte). Although these taxa are widely separated in time and space, and are phylogenetically distant from each other, we chose them as their skulls are superficially similar in several respects, due to many of the features classically associated with a herbivorous diet. Many of these features were acquired convergently, though some are due to their shared deep phylogenetic heritage. All three taxa have skulls that are relatively elongate and narrow, with low snouts, and the snouts are relatively long in comparison to overall skull length. The external openings are large, the mandibles are slender with slightly depressed jaw joints, there is no evidence for substantial kinesis within the skull, and the teeth are coarsely denticulate, relatively small, numerous and did not occlude. Traditionally these features have been associated with ‘weak’, fast bites, a lack of sophisticated chewing mechanisms, or indeed of any real specialisation (e.g. Norman & Weishampel 1991). As a result, it’s generally been thought that these skulls would have functioned similarly in life, with corresponding ideas about probable food plants and ecological roles (e.g. reliance on ‘soft’ vegetation, lack of oral processing).

From left to right, skulls of Erlikosaurus, Stegosaurus and Plateosaurus (Image courtesy of Stephan Lautenschlager/University of Bristol)

            

However, when we subjected models of these skulls to multibody dynamic and finite element analyses, what we found surprised us (Lautenschlager et al. 2016). Instead of behaving similarly, each of the skulls has its own unique function. Stegosaurus had a higher than expected bite force, in the range of 166–321 N, which overlaps with that of some living mammalian herbivores. By contrast, those of Erlikosaurus and Plateosaurus were much lower and similar to each other (50–121 N and 46­–123 N, respectively). These differences in bite force were accompanied by differences in stress patterns within the skulls. Plateosaurus seems have experienced the lowest and most evenly distributed stress patterns (implying a skull adapted to deal with a variety of different forces), whereas overall peak stresses were much higher in Erlikosaurus and Stegosaurus. In Stegosaurus, stresses were concentrated in the snout, whereas in Erlikosaurus they seem to have been highest in the posterior part of the skull. In addition, the skull of Erlikosaurus experienced the greatest amount of deformation during biting, but those of both Stegosaurus and Plateosaurus experienced very little shape change. 

Finite element models of 'Sophie' the NHM Stegosaurus, the image at the rear grossly exaggerated to look at possible deformation patterns (image courtesy of Stephan Lautenschlager/University of Bristol)

These results imply that each taxon had quite different feeding strategies, a conclusion that differs from previous ideas about these ‘unspecialised’ herbivores. For example, the differences in maximum bite force suggest that these taxa might have been feeding on diverse sorts of vegetation, with the higher bite force of Stegosaurus implying that it was able to feed on a broader, or tougher, range of plant parts/types than either the ‘prosauropod’ or therizinosaur. This higher bite force was enabled by a larger jaw muscle mass in Stegosaurus and/or an arrangement of the jaw muscles that allowed more efficient conversion of muscle force into bite force. The lower bite forces of Plateosaurus in combination with its high cranial robustness are consistent with low fibre herbivory, dealing with soft vegetation that required little chewing, and/or omnivory (the skull could have withstood dealing with struggling small prey, for example). Erlikosaurus appears to have been specialised to use the tip of its snout in plucking vegetation, as the skull performs exceptionally badly when biting food at the back of the mouth. Nipping soft vegetation with the tips of the jaw is also consistent with its low bite forces.

            Previously, these three taxa were all thought to be relatively ‘boring’ herbivores that simply nipped and swallowed soft plants. It now seems that one was eating much tougher vegetation, another was a generalist that could exploit different food sources, and the third was a specialist with a rather delicate way of feeding itself. This work shows that first appearances based on simple application of the form/function paradigm can be misleading. Novel functions have now been revealed that would have gone unnoticed if it were not for detailed biomechanical modelling of each skull. This leads me to wonder what other functional surprises might be lurking in dinosaur skulls, especially as so few have been really thoroughly studied in this way.

References

Lauder, G.V. 1995. On the inference of function from structure. In Functional Morphology in Vertebrate Paleontology (ed. J.J. Thomason), pp. 1–9. Cambridge: Cambridge University Press.

Lautenschlager, S., Brassey, C., Button, D. J. & Barrett, P.M. 2016. Decoupled form and function in disparate herbivorous dinosaur clades. Scientific Reports 6: 26495. doi:10.1038/srep26495

Norman, D.B. & Weishampel, D.B. 1991. Feeding mechanisms in some small herbivorous dinosaurs: processes and patterns. In Biomechanics in Evolution (eds J.M.V. Rayner & R.J. Wootton, pp. 161–81. Cambridge: Cambridge University Press.