Introducing 'trident': a graphical interface for discriminating groups using dental microwear texture analysis
Résumé
This manuscript introduces trident, an R package for performing dental microwear texture analysis and subsequently classifying variables based on their ability to separate discrete categories. Dental microwear textures reflect the physical properties of the food, the feeding ecology of a given species, and niche partitioning when considering multi-specific communities. The trident package comes with independent functions and a user interface, trident, enabling easy and fast proficiency. It can import .SUR files, then remove aberrant peaks and possibly polynomial surfaces. Next, it can measure up to 24 texture parameters and their statistics of heterogeneity, generating 384 variables. It also ranks any number of variables using five different methods, displays the results in multivariate analyses, and exports the results into R, providing access to its large asset of libraries.
We then present these features in three case studies, showing how trident helps answer questions commonly investigated by paleontologists and archaeologists. In the first case study, we separate four groups of domestic pigs based on their dietary composition. In the second case study, we identify microwear texture patterns in a large database of 15 primate species and relate these patterns to biomechanical and ecological factors. The third case study investigates the dental microwear textures of four extant ruminants to infer the diet of an extinct antelope from the Pleistocene of Greece. These case studies show how trident can leverage dental microwear texture analysis results.
Fichier principal
Thiery_etal_2024_final.pdf (2.7 Mo)
Télécharger le fichier
Suppl material.zip (8.99 Mo)
Télécharger le fichier