The science of paleoecology suffers from a lack of conceptual frameworks. Paleoecologists have been overconcerned with the inadequacies of the fossil record: as a result, community palmecology has historically developed very slowly. At the community ecosystem level, the need for a theoretical framework is so great that paleoecology must ‘borrow’ the hypotheses of modem ecology. Consideration of the stability-time hypothesis of Sanders in conjunction with the physical setting of transgression and regression has permitted the structuring of three community types and the interpretation of their behavior under variations in the physical environment. These community types (opportunistic. stable mature, relict mature) are recognizable in the fossil record and examples are given from the Upper Pennsylvanian of the Appalachian Basin.