NEW PERSPECTIVES FROM AGE-OLD CONCEPTS: SAMENESS AND SIMILARITY IN PALEONTOLOGY

Similarity and dissimilarity are foundational concepts in paleontology, underlying popular methods such as comparative morphometrics, taphonomy/fidelity studies, and quantitative ecospace analyses. Here we discuss popular similarity/dissimilarity indices as they relate to paleoecological studies and demonstrate the existence of possible degenerate cases which lead to inaccurate characterizations of the relationship between two or more samples. We propose a hierarchical approach to similarity/dissimilarity and describe a new index capable of recognizing degrees of community-level functional similarity in terms of functional traits, entities, and groups. Specifically, we propose a generalized hierarchical version of the Tversky index which uses fuzzy set theory as its basis, allowing users to account for subtle differences between two samples, while avoiding the computational complexity and information loss associated with ordination-based and classical set theoretic techniques. The ability to decompose the proposed index into asymmetric components allows users to recognize the sources of variation driving the resulting similarity value, and can be used as a means of describing specific relationships between samples (e.g. the extent to which an older ecosystem structure still existed at a later time period). Our results suggest that the proposed index is sensitive to a number of biologically and/or macroevolutionarily relevant patterns. We demonstrate the utility of this new index by exploring rates of functional change as they relate to key macroevolutionary events, including mass extinctions. Our results shed new light on the relationship between extinction selectivity and rates of functional change in shallow marine ecosystems.