On the Causes of Temporal Change in Communities: Modification of the Biotic Environment

Similarities among communities occurring in a variety of habitats are adduced in support of a new mechanism for sequential changes in species composition, modification of the biotic environment, consisting of interactions among residents and between residents and immigrants, which complements the conventional mechanism, modification of the abiotic environment. Novel interpretations of successional phenomena among pelagic organisms of the open ocean and sessile ones of the nearshore benthos and old fields of the coastal plain are provided; other explanations have been suggested in a few cases, but most of these observations are new. Patterns common to all three confirm the importance of interactions, particularly predatory ones (Glasser 1979), in shaping and maintaining the structure of communities, although life histories and modification of the abiotic environment are important too. Beyond the constraints of a theory that was too narrow to include animals as well as plants, students of determinants of ultimate structure in natural communities seem to have been misled by boundaries, both real and imaginary. Ignorance of the influence of size and isolation on processes affecting the patchiness of a mosaic is responsible for the argument concerning whether community development is deterministic or stochastic, and the distinction between large sessile occupants of primary space (i.e., plants and plantlike animals) and other organisms is responsible for the notion that succession proceeds toward monocultures in the absence of disturbance. Community development is a composite of processes that may be divided into those that determine the arrival of propagules at sites (i.e., colonization) and those that determine whether species colonize successfully, subsequently undergoing population growth, increased niche breadth, overlap, competition for resources, and so on (i.e., succession). The limits of this partition become apparent with the production and dispersal of new propagules, but it is useful nonetheless. Current conceptions of colonization are more stochastic than those about succession, but that may reflect the inadequacy of our present understanding more than anything else. Nonetheless, the vagaries of colonization do seem to dominate community development on small proximate sites as a consequence. If predators, in the most general sense of that word (Lubchenco 1979), are included among the members of communities, then succession does not proceed toward uneven distributions of individuals, biomass, or space among one or a few species, but rather toward more even distributions of whatever among many. Communities are less vulnerable to disturbance in this structural configuration than others, supporting the putative relationship between diversity and stability (Odum 1969). Although it has become fashionable to attack such empirical generalizations (e.g., Connell and Slatyer 1977, pp. 1136-1137), this analysis suggests that at least one of those about succession is well founded. The contrary notion results from considering the effects of extraordinary perturbations or misinterpreting the results of simple models (e.g., Goodman 1975). Other apparent exceptions ought to be scrutinized before they are accorded stature as observations for which viable community theories must account. Those observations described here are consistent with one such theory (Glasser 1979), although it awaits more critical evaluation. Our understanding of colonization seems most likely to increase through expansion of conceptual models that, unlike those of Horn (1975) and Simberloff (1978), are based upon life histories. These determine the probabilities of individuals arriving at sites and colonizing successfully, the former as functions of their ages and the latter of characteristics of the present residents. Such models will permit us to frame biological hypotheses about the differences in those probabilities from site to site and species to species. These are the patterns we should seek to explain (MacArthur 1967), for through this process we may eventually understand community development.