Hierarchical approaches to the study of ecological process and pattern

An approach based on hierarchy theory is applied to three specific problems involving the relationships between levels of analysis in ecology. The first problem involves “emergence” --the inability to predict the behavior of the whole from its parts. “Relative” emergence is defined to arise from a lack of sufficient information, as opposed to “absolute” emergence, which refers to inherent unpredictability. The observation of relative emergence is nontrivial because it indicates a deficiency of lower level (or reductionist) analyses in certain situations and thus the appropriateness of higher level analyses designed to overcome these deficiencies. To analyze relative emergence, a hierarchical conceptual model is developed, which is then used to clarify several controversies involving emergence and to help assess the benefits and pitfalls of higher level analyses. If a taxonomic hierarchy can be assumed to reflect a hierarchy of ecological similarity of species, then higher taxonomic levels may be useful in detecting larger scale ecological patterns. Patterns shown by specific and generic level binary similarity coefficients are compared for chironomid species data from a polluted Ohio stream. The ecological similarity of congenerics is assessed using the trend in the species:genus ratio along the pollution gradient. No trend toward ecological similarity was found and it appears that the generic level may best be viewed as resulting in a random loss of species information. A hierarchical approach assumes that a hierarchy of pattern reflects a hierarchy of process. This means that small-scale “stochastic” factors, such as small differences in larval recruitment, should be manifested as small-scale “noise” added on to the dynamics, while large-scale phenomena, such as large changes in the average predation rate, should affect the pattern on a large scale. This chapter uses simulation models of New England and Australian rocky intertidal communities to show how this assumption may be violated. The model results indicate that three characteristics of the Australian rocky intertidal community may allow small-scale. “stochastic” processes to have larger scale effects than in the New England community.