How complex must shape data be to model in vivo forces? Intraspecific level validation of in silico jaw strength estimates in a lizard.

A major problem in current biomechanical literature is the extent to which in silico data can be validated by in vivo data across taxonomic scales. Despite frequent incongruence between in silico and in vivo data gained from precisely the same individual, biologists and palaeontologists continue to publish in silico data of single bones intended to represent entire species. Here, we aim to bridge this gap by investigating whether jaw morphology alone can be used to validate biomechanical models on the intraspecific level in a phenotypically diverse lizard, Podarcis pityusensis. We tested this by investigating how effectively in vivo bite force measurements from eight populations of this species are predicted by biomechanical models. We used alcohol-preserved specimens from each location to generate population-average and male-average morphologies of mandibles and dentaries, from which we calculated mechanical advantage as well as strength estimates from finite element analysis. Overall, we found a general lack of population-level correlation between in vivo and in silico data; however, strength estimates from finite element analysis did follow the same bite∼size relationship as in vivo bite, suggesting that biomechanical analysis of even a single bone can produce useful bite force estimates. We encourage researchers to create in silico models with maximally complex shape data and caution that intraspecific variation is a crucial aspect of in vivo and in silico biomechanics.