Australasian Science: Australia's authority on science since 1938

Why the Long Face?

Photo: Guy Nolch

Head shape is a good indicator of the predator’s tendency to feed on either small or large prey. Photo: Guy Nolch

By Christopher Walmsley & Colin McHenry

The jaw strength of crocodiles can be predicted by simple linear measurements that could provide new insights into the diets of extinct marine reptiles.

Crocodiles and their relatives have been on the planet for around 200 million years, and today are the largest of all living reptiles. They are found in many countries all over the globe, preferring to inhabit warm waters in regions like Africa, South-East Asia and Australia. This iconic predator has stood the evolutionary test of time, managing to live and even thrive in environments where many others failed.

Crocodiles have adapted to their environment in many ways, including pressure sensors in the skin that help to detect prey in the water and an internal navigation system that allows them to find their way home even after being relocated thousands of kilometres away. The way these predators have managed to adapt so successfully to their environments is fascinating, and it’s easy to see why many scientists choose to study them.

Among the 23 living species of crocodile there is an amazing range of diversity in their head shape. This spectrum ranges from animals with very long, narrow, pincer-like jaws, such as the gharial in India and the false gharia in Sumatra and Borneo, through to the short, broad jaws of the spectacled caiman in Central and South America and the dwarf crocodile in West Africa.

Fig 1

This diversity in head shape is not unique to crocodiles. It’s also seen in other aquatic predators such as toothed whales, dolphins and a variety of fossil marine reptiles, all of which present the same spectrum of head shape seen in crocodiles (Fig. 1).

For both dolphins and crocodiles, head shape is a good indicator of the predator’s tendency to feed on either small or large prey. At the short, broad end of the spectrum animals such as the Nile crocodile are capable of taking prey much larger than themselves. At the other end of the scale, species that eat the smallest, most agile prey – mainly fish – have extremely long, narrow jaws; examples include river dolphins and the gharial.

In these long, slender-snouted forms, the mandibular symphysis – where the two halves of the lower jaw fuse together – is very long. For example, the symphysis in gharials is more than 50% of the length of the skull.

Long symphyses have been described in some fossil marine reptiles and mammals, and palaeontologists, noting the link between a long symphysis and a fish diet, have used that feature to reconstruct the feeding behaviour of these extinct forms. However, the reasons why a long symphysis should be associated with eating fish have not been studied – until now.

The flip side of fish-eating crocodiles and dolphins with long symphyses is that species that feed on very large prey have short symphyses. Saltwater crocodiles, alligators and killer whales can all feed on large prey, including mammals. In all of these species the symphysis is a small proportion of the length of the jaw.

Following our previous work on the biomechanics of crocodile skulls, we hypothesised that the short symphysis in these species is dictated by the bio­mechanics of their feeding behaviour. For both crocodiles and alligators, prey that is too large to be swallowed whole is broken up into small bite-sized chunks, either by vigorous shaking or even whole body twisting (the infamous “death roll”) by the predator. We hypothesised that the forces resulting from these shaking and twisting behaviours requires a broader jaw with a short symphysis for strength.

But how can we test this? Going out into the field and physically testing the jaw biomechanics of large crocodiles presents some obvious challenges, so we turned to computer modelling. We used a computational engineering approach called finite element analysis (FEA), which is widely used by engineers to design planes, cars, boats, buildings, bridges, and many other structures. FEA uses a 3D computer model of a structure, applies loads to the model and then calculates the resulting strain on each component.

Strain is a measure of how much a structure has deformed from its original geometry. The higher the strain, the more likely a structure is to permanently deform or even completely break.

With the development of high-powered computers, this technique has become very popular with biologists, biomedical scientists and palaeontologists, who have used FEA to model the biomechanics of everything from hip replacements to the biting strength of Tyrannosaurus rex.

Fig 2

For our study we used a comparative approach and selected seven different species of crocodile, and used FEA to simulate the sorts of biting, shaking and twisting loads crocodiles use when feeding on large prey. The seven species between them encompassed the entire spectrum of skull shape, from very short symphysis to very long symphysis forms (Fig. 2).

In order to use FEA to perform our analysis we had to generate 3D models of each crocodile skull. This process is very time-consuming, often taking months just to get the geometry of a single specimen into the FEA software.

This geometry was then brought into the FEA software, and the model was constrained to mimic the loads resulting from biting, shaking and twisting during feeding. When the models were properly loaded and restrained, the simulation was run and results extracted.

Fig 3

After performing these simulations the clear result was that long symphyses were not as strong for all types of loads, making them more likely to break when feeding on large prey. Furthermore, the amount of strain a jaw was under was directly proportional to the length of the symphysis when shaking or twisting, and proportional to jaw length when biting (Fig. 3).

This leads to the conclusion that the animal’s biomechanical response to a load could be predicted, with a fair amount of accuracy, by knowing nothing more than a few simple measurements. This sort of result is fantastic if you’re interested in crocodiles since it quantifies what aspects of skull shape are most important when feeding on large prey.

But is this result only specific to crocodiles, or does it tell us something about jaw mechanics in a wide range of aquatic predators?

Unfortunately, on its own this result isn’t enough to make any grand statements about skull shape and feeding behaviours in other aquatic predators, despite the obvious parallels.

So we delved a little deeper, and to evaluate that question – which is of particular interest to palaeontologists – we compared the results of our complex FEA models with the predictions of basic engineering beam theory. Similar to the FEA models before, we simulated biting, shaking and twisting load cases on simple representations of the lower jaw for each specimen used in the study. These representations only captured the main geometric features like length, symphysis length and width.

In most cases, the results of the complex models agreed very closely with these predictions, suggesting that the results are highly relevant to a broad range of aquatic predators, including dolphins and fossil marine reptiles. The one exception encountered was twist feeding, so future work will focus on what makes this behaviour biomechanically special for crocodiles (and perhaps other species).

This work reveals why crocodiles that feed on large prey have short symphyses. By extension, it also shows why crocodiles that have long symphyses do not feed on large prey – they simply aren’t biomechanically suited to that behaviour.

But it does not yet reveal why crocodiles that feed on small prey have such long symphyses. While there is conjecture about that question, there is as yet no data. We suspect that the answer lies in the hydrodynamic efficiency of the elongate jaws, and that maybe a long symphysis is better streamlined and this gives animals an advantage when capturing very agile prey like fish.

We plan to explore this further using another computational engineering technique called computational fluid dynamics. This technique is used to analyse how fluid flows over or past an object, allowing engineers to make cars, planes and even ship hulls more streamlined.

This interface between biology and engineering is one of the things that make biomechanics research such an exciting and dynamic science.

Christopher Walmsley is a PhD candidate and Colin McHenry is a lecturer in the Department of Anatomy and Developmental Biology at Monash University.