ICHTHYOLITHS: AN OPEN-SOURCE R PACKAGE FOR EXPLORING MORPHOLOGICAL DISPARITY IN MICROFOSSIL FISH TEETH AND ELASMOBRANCH DENTICLES AND ITS APPLICABILITY IN MACHINE LEARNING-BASED MICROFOSSIL IDENTIFICATION

Ichthyoliths – microfossil fish teeth and shark scales – are preserved in nearly all aquatic sediments. They are composed of dissolution-resistant calcium phosphate, and are the most numerically abundant fossil record of marine vertebrates. The ichthyolith microfossil record thus has the potential to provide valuable insights into how fish and shark communities evolve throughout Earth’s history and in response to environmental changes. However, their small size and relative rarity has left this rich fossil record of ichthyoliths overlooked for paleontological studies. Further, while ichthyoliths are numerically abundant, they are found in isolation – preserved separately from any body fossils or more complete specimens of the fish and sharks that produced them, presenting a challenge to interpret the microfossils in a traditional taxonomic framework. Here we present a pair taxon-free morphological frameworks for studying fish and shark evolution using ichthyoliths – one for teeth and one for denticles. These morphological frameworks are underpinned by a new R package ichthyoliths (github.com/esibert/ichthyoliths) which facilitates morphological disparity analyses across studies, and provides the tools to explore denticle and tooth morphology with fossil and modern taxonomic and ecological frameworks. Here we present a series of vignettes of both denticle and tooth morphological studies, demonstrating the power of this new tool, and discuss the potential to use this in machine learning-based analyses which will further increase efficiency and repeatability of morphological studies on these abundant yet under-studied microfossils.