The Impact of Fossils from the Northern Dominican Republic on Origination Estimates for Miocene and Pliocene Caribbean Reef Corals

Documenting the pattern of biodiversity change on global, regional, or local scales is an important use of paleontological data. Long-term trends of taxonomic richness and the rate of first and last occurrences of taxa can be used to study effects of environmental change on ecosystems and is of significant relevance during this time of increased concern over the effects of anthropogenic environmental change. Compilations of taxon occurrences provide the basic data used to describe the distributions of taxa in time and space. Paleontologists have been producing hard-copy and digital compilations for many decades (for example, Valentine, 1969; Raup, 1972; Sepkoski, 1982, 2002), and a large suite of sophisticated methodologies have been developed for the statistical analysis of these data (for example, Gilinsky and Signor, 1991). Most of the existing compilations are limited to particular taxonomic groups within restricted geographic and stratigraphic scope, but some intend to be global in scope (for example, Alroy, 2007). However, no compilation can ever be comprehensive or truly authoritative because ongoing research based on new field sampling as well as study of existing museum collections will continue to generate new information requiring the enlargement or modification of data compilations. Indeed, one of the most difficult tasks for compilation managers is determining how best to increase the utility of synthetic compilations. Is it better to prioritize the addition of new data over increasing the accuracy of existing information? Similar compromises must be met to determine which data to include and which to exclude in the absence of unlimited resources. For any given question, it is likely that some subsets of data will be more influential than other subsets, and a rational choice would be to prioritise increasing the accuracy of the more influential data rather than the existing practice of concentrating effort based on particular research interests of data contributors, the ready availability of existing data sets, or the search for ‘even’ coverage. The aim of this chapter is to explore a basic methodological approach to estimate the relative influence of subsets of data. We will apply these new techniques to ongoing study of long-term patterns of taxonomic turnover dynamics in Cenozoic Caribbean Scleractinian reef-coral occurrences to determine the importance of collections from the Dominican Republic in our understanding of patterns of species origination and extinction in response to regional environmental change.