EFFECTS OF VARYING METHODS OF COMPOSITE TIMESCALE FORMATION ON BIODIVERSITY ESTIMATES

Several approaches to incorporating fossil recovery information in to the estimation of extinction and origination rates have been proposed. Successful implementation of these methods requires a robust and reliable method of creating composite sections, and of mapping presence/absence data from horizons within individual sections into the composite. Composite sections formed by graphic correlation and by automated methods such as constrained optimization by simulated annealing (CONOP) or ranking-scaling (RASC/CASC) may produce slightly different composite section solutions, and have varying levels of reliability. We present here several approaches to the comparison of composite section solutions, and to estimating the robustness of these solutions using the extensively studied graptolite fauna of the late Ordovician, including collections from the Yangtze region of China and Vinnini Creek, Nevada, as a case study.