BoneProfileR: The next step to quantify, model, and statistically compare bone section compactness profiles
Jordan Gônet, Michel Laurin, Marc Girondot
Abstract
Bone sections have been widely used for over a century to assess life history traits of extant and extinct vertebrates, and they remain extremely useful. The characterization of their geometric and microanatomical properties forms the core of many descriptive and inferential studies. Bone compactness in particular has been associated with various attributes in individuals, such as habitat use or specific morphofunctional traits. A method implemented in software was published in 2003 to reduce the complexity of bone sections to facilitate quantitative analyses. We improve this method using better statistical tools (maximum likelihood and Bayesian analysis), a new function to model compactness, and we propose new enhanced software, BoneProfileR, as an R package, along with an online version for users unfamiliar with this language. Application of the new method to the femur of an extant hedgehog (Erinaceus europaeus) and a Permian temnospondyl (Eryops megacephalus) allows us to quantitatively and objectively describe the cross-sectional microanatomy of these two taxa.