Why the Long Face? The Mechanics of Mandibular Symphysis Proportions in Crocodiles

Abstract

BackgroundCrocodilians exhibit a spectrum of rostral shape from long snouted (longirostrine), through to short snouted (brevirostrine) morphologies. The proportional length of the mandibular symphysis correlates consistently with rostral shape, forming as much as 50% of the mandible’s length in longirostrine forms, but 10% in brevirostrine crocodilians. Here we analyse the structural consequences of an elongate mandibular symphysis in relation to feeding behaviours.Methods/Principal FindingsSimple beam and high resolution Finite Element (FE) models of seven species of crocodile were analysed under loads simulating biting, shaking and twisting. Using beam theory, we statistically compared multiple hypotheses of which morphological variables should control the biomechanical response. Brevi- and mesorostrine morphologies were found to consistently outperform longirostrine types when subject to equivalent biting, shaking and twisting loads. The best predictors of performance for biting and twisting loads in FE models were overall length and symphyseal length respectively; for shaking loads symphyseal length and a multivariate measurement of shape (PC1– which is strongly but not exclusively correlated with symphyseal length) were equally good predictors. Linear measurements were better predictors than multivariate measurements of shape in biting and twisting loads. For both biting and shaking loads but not for twisting, simple beam models agree with best performance predictors in FE models.Conclusions/SignificanceCombining beam and FE modelling allows a priori hypotheses about the importance of morphological traits on biomechanics to be statistically tested. Short mandibular symphyses perform well under loads used for feeding upon large prey, but elongate symphyses incur high strains under equivalent loads, underlining the structural constraints to prey size in the longirostrine morphotype. The biomechanics of the crocodilian mandible are largely consistent with beam theory and can be predicted from simple morphological measurements, suggesting that crocodilians are a useful model for investigating the palaeobiomechanics of other aquatic tetrapods.

A Giant Pliosaurid Skull from the Late Jurassic of England

“…Liopleurodon [16] and Pliosaurus [34]; and also in Simolestes [16]) suggests a functional correlation between macropredatory lifestyles and a short mandibular symphysis. Finite element modelling and beam theory approaches to understanding the mechanical implications of symphysis length in crocodiles indicates that short symphyses perform better under loads used for feeding on large prey [97]. This suggests that symphysis shortening in pliosaurids was an adaptation for macropredation, and is consistent with the appearance of other correlates of macropredation such as trihedral teeth [20] (in Pliosaurus ) and a strongly anisodont dentition in these taxa.…”

New dyrosaurid remains (Crocodyliformes, Mesoeucrocodylia) from the Paleocene of the Paraíba Basin, NE Brazil

“…However, this type of rostrum shape is fragile, which restricts potential prey size (SchwarzWings, 2014). Long snouts are an advantage to many crocodiles due to the reduction in hydrodynamic drag for swimming and prey capture (Walmsley et al, 2013). The homodont dentition with acute apices also represents an advantage for catching small prey (Schwarz-Wings, 2014).…”

Accounting for differences in element size and homogeneity when comparing Finite Element models: Armadillos as a case study

“…Therefore, these extreme values cannot be stabilized in a final value by changing the mesh. On the other hand, 90% of the values of the stress distribution can be stabilized in its final value despite the variation of the size of the elements (80% inside the whiskers and the lower 10%) following the idea proposed in Walmsley et al (2013), where the maximum value is not analysed for the presence of artefact values. For our example, Chlamyphorus truncatus, the stabilization of the values happened at around 25,000 nodes (Mesh S07 of the Figure 5.1).…”

“…Another important point that should be analysed is the influence of artificial noise produced in the results of FEA (such as Dumont et al, 2005, or Walmsley et al, 2013. The artificial noise is a numerical singularity due to the presence of artificially high stress values at points where displacement boundary conditions are applied (Marcé-Nogué et al, 2011).…”