The macroevolutionary processes underlying large-scale biotic replacements are still poorly understood. Opinion remains divided regarding the roles of mass extinction, biotic interaction, and environmental perturbations in these replacement events. Previous attempts to test replacement hypotheses have largely focused on taxonomic diversity patterns. Taxonomic data alone, however, provide little insight about ecological interactions and hence other approaches are needed to understand mechanics of biotic replacements. Here I propose a conceptual model of replacement based on predation-mediated biotic interactions, and attempt a test using analysis of the Cenozoic replacement of the gastropod family Aporrhaidae by a closely related group, the Strombidae.
Taxonomic, morphologic, and geographic data analyzed in this study all suggest a replacement of aporrhaids by strombids following the end-Cretaceous mass extinction. While most of the taxonomic replacement was associated with a mass extinction, some replacement also occurred during background times and was mediated by higher origination rates in strombids rather than by higher extinction rates in aporrhaids. Morphologically, the replacement was largely confined to the portion of the morphospace unaffected by the end-Cretaceous extinction. At a global scale, the geographic overlap between the two groups declined through the Cenozoic, reflecting increasing restriction of aporrhaids to colder, temperate waters while strombids flourished in the tropics. However, at a finer geographic scale a more mosaic pattern of replacement is evident and coincides with Eocene and Oligocene climatic fluctuations.
The results of this study suggest that mass extinction, long-term biotic interaction, and environmental change can all play significant roles in biotic replacements. Since the relative importance of each factor would vary from one event to another, an understanding of the general nature of large-scale biotic replacements requires a knowledge of the relative intensities of each of these processes.