A QUANTITATIVE NULL MODEL OF EXPECTED CHANGES IN ECOLOGICAL COMMUNITY STRUCTURE FOLLOWING AN EXTINCTION

Most previous work on biological recovery from extinction has focused on taxonomic categorizations of the surviving biota, often restricting discussion to a single higher taxonomic group. Alternatively, recoveries can be addressed using a more integrative, paleoecological perspective that incorporates analysis of paleocommunity structure through quantitative factors such as taxonomic richness, abundance, and dominance patterns. We present a quantitative null model of expected changes in Shannon-Wiener diversity (H), evenness (E), and taxonomic richness (S) as a function of percent extinction. The model has three major components: (1) initial values of H and E are determined from lognormal taxonomic abundance distributions for 10, 100, and 1000 taxa respectively; (2) the distributions are subjected to varying percentages of extinction at 5% intervals from 5 to 95%; and, (3) H and E are recalculated for post-extinction conditions. Two conditions were tested with respect to extinction—that extinction is either random with respect to taxonomic abundances, or that extinction preferentially targets rare taxa.

When extinction is random with respect to abundance, H decreases systematically with increasing levels of extinction, and E remains nearly constant except at the highest levels of extinction. The largest changes in H and E are observed after 60% extinction for 10 taxa, 70% for 100 taxa, and 80% for 1000 taxa. When pre-extinction evenness is low, the highest increases in E are observed.

When rare taxa preferentially go extinct, H decreases systematically as in the previous model. However, post-extinction values of E are greater than in the previous model and values of E increase consistently with extinction rates, even at the lowest percentages.

This null model will be useful in testing whether extinction is random with respect to taxonomic abundances or whether an extinction event has preferentially targeted either rare or abundant taxa. The null model will also be useful in directly comparing changes in community structure from the pre- to post-extinction intervals by determining the null starting conditions during initial periods of survival and recovery from extinction.