I. Summary Despite the fact that a substantial research effort has been directed at the analysis of species abundance patterns over the years, there are many important issues which are not sufficiently explored and understood. This review examines the conceptual and methodological frameworks of analysis in species abundance patterns, with particular emphasis on interpreting patterns in the context of community structure and organization. Historically, this discipline underwent a largely monotonous pattern of development (Section III) in that successive researchers proposed different models which were often claimed to be superior to those previously proposed, followed by seemingly piecemeal analyses of empirical data with reference to a particular model. This has resulted in a conspicuous lack of broad perspective in the study of species abundance patterns, despite a number of reviews produced. In the first place, therefore, it is important to clarify the conceptual basis of analysing species abundance patterns, including the definition of community and sample to which different models inevitably refer (Section IV). Basic characteristics of species abundance models proposed to date are briefly reviewed and recommendations for standardizing the presentation of rank-abundance graphs are given. In the context of unravelling community structure, the concept of niche is highly relevant in the analysis of species abundance patterns (Section V). Thus particular attention is given to the development of niche-oriented models with a consideration of the relationship between resource and species abundance. Furthermore, with respect to the question of how total energy or resource is divided among species in a community, the current debated issue of density-body size allometry is reviewed. Interpretation of models and patterns is a major problem in this discipline (Section VI). It is recognized that models, particularly niche-oriented ones, are useful in suggesting possibilities underlying community organization. Thus they should be regarded as an aid for interpreting patterns rather than as a precise and rigid description of mechanisms involved. Consideration is given to some important dichotomies including global versus community-specific patterns, contemporary versus evolutionary processes, species- versus process-oriented interpretation and equilibrium versus non-equilibrium communities. Apart from problems associated with interpretation, there are practical problems of testing models (Section VII). Despite empirical studies in the past, this is again a neglected area, and particular difficulties arise with stochastic species abundance models. As a way forward, a simulation method of testing stochastic-type models is introduced and the importance of making replicated observations is emphasized. Section VIII deals with topics which are closely related to the analysis of species abundance patterns, namely diversity indices, species-area relationships and application to environmental assessment. Section IX identifies aspects to be considered in future investigations and emphasizes the importance of spatial/temporal variability and of integrating the analysis of species abundance patterns and other analyses of community structure, whether observational, experimental or analytical, in order to achieve a better understanding of ecological communities. © 1993, Academic Press Limited.
