Preserving the Past: Evaluating Conservation and Accuracy in Paleontological Microwear Analysis

Abstract

This research evaluates the reliability of dental microwear analysis through repeated molding and casting procedures on perissodactyl fossils. Statistically significant differences in scratch measurements were found across Brontotheriidae (UWBM #75645) casts (F (9, 203.79) = 2.01, p = 0.039), particularly between Generation 2 and Generation 8 (mean difference = 0.0334, 95% CI [0.0014, 0.0654], p_adj = 0.0330). Non- parametric analysis further revealed significant distributional differences between Generation 1 and Generations 2 (p = 0.0018), 5 (p = 0.0045), 9 (p = 0.0129), and 10 (p = 0.0129). In contrast, no significant differences were found among Subhyracodon (UWBM #35783) casts for either scratch (F (9, 88.40) = 1.17, p = 0.327) or pit measurements (F (11, 54.10) = 1.740, p = 0.089). These findings suggest that the molding and casting process can introduce both mean-level changes and distributional shifts in microwear features after multiple replication cycles, potentially affecting dietary interpretations. The observed differences between Generation 2 and 8 indicate that surface fidelity remains relatively stable for several replications before detectable changes emerge. Given the importance of microwear in reconstructing short-term feeding behavior and ecological change, even subtl distortions may compromise analytical accuracy. This research emphasizes the need for standardized molding and casting protocols and recommends best practices to reduce error and improve consistency in paleodietary studies.