STATISTICAL TESTING FOR PALEOCOMMUNITY RECURRENCE - ARE SIMILAR FOSSIL ASSEMBLAGES EVER THE SAME

Observations of the recurrence of similar fossil assemblages through l ong intervals of geologic time punctuated by rapid changes in both the composition and structure of fossil assemblages has recently resulted in the concept of ''coordinated stasis'' to describe this pattern in the behavior of paleocommunities through time. Coordinated stasis may imply the existence of ecological mechanisms that actively maintain pa rticular community structures and thus have a significant impact on th e process of macroevolution. However, before such mechanisms can be in voked, it must be shown that ecological stasis as observed in the foss il record is more significant than the persistence of similar communit y types due to the repeated reinvasion of recurring habitats from a pe rsistent species pool (the ''null hypothesis'' for paleocommunity recu rrence). Analysis of the relationship between communities and paleocom munities and an expansion of the ecological hierarchy at the community /paleocommunity level allows the creation of a rigorous definition of the entities composing the local paleocommunity (the samples collected within a stratigraphic horizon at a single locality), paleocommunity (groups of samples shown to be statistically identical), and paleocomm unity type (groups of samples that are similar but can be shown to be statistically different). These definitions permit the development of the null model for paleocommunity recurrence and establish a base-leve l of variability within a local paleocommunity that permits rigorous s tatistical comparisons to be made between paleoecological samples at l arger temporal and geographic scales. Paleoecological samples from fou r marine tongues in the Middle Pennsylvanian Breathitt Formation are u sed to analyze species abundance variability at several spatial and te mporal scales and to lest for paleocommunity recurrence. Although recu rrence of statistically identical local paleocommunities occurs at sin gle localities and between localities within individual marine units, it is usually not detectable between marine tongues, suggesting that p aleocommunity recurrence in the marine strata of the Breathitt Formati on is the recurrence of paleocommunity types and does not falsify the null hypothesis.