Developments in data science solutions for carnivore tooth pit classification

Abstract

Competition for resources is a key question in the study of our early human evolution. From the first hominin groups, carnivores have played a fundamental role in the ecosystem. From this perspective, understanding the trophic pressure between hominins and carnivores can provide valuable insights into the context in which humans survived, interacted with their surroundings, and consequently evolved. While numerous techniques already exist for the detection of carnivore activity in archaeological and palaeontological sites, many of these techniques present important limitations. The present study builds on a number of advanced data science techniques to confront these issues, defining methods for the identification of the precise agents involved in carcass consumption and manipulation. For the purpose of this study, a large sample of 620 carnivore tooth pits is presented, including samples from bears, hyenas, jaguars, leopards, lions, wolves, foxes and African wild dogs. Using 3D modelling, geometric morphometrics, robust data modelling, and artificial intelligence algorithms, the present study obtains between 88 and 98% accuracy, with balanced overall evaluation metrics across all datasets. From this perspective, and when combined with other sources of taphonomic evidence, these results show that advanced data science techniques can be considered a valuable addition to the taphonomist’s toolkit for the identification of precise carnivore agents via tooth pit morphology.

3D Insights into the Effects of Captivity on Wolf Mastication and Their Tooth Marks; Implications in Ecological Studies of Both the Past and Present

“…Samples were collected during the winter months of 2010 and 2011. The first of these samples have been previously included in publications including [ 32 , 41 , 44 , 45 , 46 , 50 , 53 , 57 ], while the latter by [ 45 , 53 , 58 ]. Both samples were produced by groups of adult individuals, seven individuals in the case of Cabárceno and five in the case of Hosquillo.…”

“…This process, frequently referred to as Generalized Procrustes Analysis (GPA), consists of multiple superimposition procedures (translation, rotation, and scaling) that aid in the quantification and visualization of minute displacements of individual landmarks in space. Nevertheless, considering observations made by Courtenay et al [ 44 , 45 , 46 , 53 ], which reveal tooth mark size to be an important conditioning factor in morphological variance, GPA was then subjected to multiple allometric analyses to test for shape-size relationships [ 96 ]. For allometric analyses, multiple regressions were performed using the logarithm of centroid size to estimate shape, testing for the goodness-of-fit to conclude whether (1) shape is notably affected by size but also (2) to see if differences in shape-size relationships for different samples are present.…”

“…This can be calculated using a combination of FPR and an inverse function (Equation (S5.4)). Considering how Courtenay et al [ 45 ] found the point of maximum curvature of calibration curves to be at p = 0.3681, p(H 0 ) was defined here using this value as the optimal limit in Equation (S5.4).…”

“…For more details on the present use of BFB and FPR values, consult [ 103 ], ref [ 104 ], or the summary reported in Supplementary Appendix 3 of Courtenay et al [ 45 ]. For more details on p(H 0 ) consult Courtenay et al [ 106 ].…”

“…Likewise, while the inclusion of advanced data science techniques such as Computational Learning strategies was able to present a more efficient means of processing this data [ 44 ], these studies are still limited in terms of sample size and can only be considered the first step in a promising direction. From this point, more consequent efforts have been made to increase sample sizes with equally promising results [ 32 , 45 ] while also improving the means of extracting this information [ 46 ].…”