Some problems of community ecology: processes, patterns and species persistence in ecosystems
2017-01-05T03:15:15Z (GMT) by
In this thesis a variety of approaches are examined and used to explore the dynamics, patterns and structure of ecological communities. I address the problem of "how it is possible for a number of species to coexist ... all competing for the same sorts of materials" (Hutchinson 1961). An ensemble model is used in an attempt to capture those factors that make for long-term community coexistence. The model is also used as a vehicle to explore - as well as to generate questions and hypotheses relating to - topics currently being examined by community ecologists. For example, the persistence and stability of ecological communities, or the true (sometimes hidden) nature of the interaction between a pair of species can be analysed with the aid of the model. I also study the problem of how, by directly analysing field-data, one might detect evidence of any community-wide processes that explain coexistence. The model makes use of the Generalized Lotka-Volterra equations, and is primarily based on the fundamental consumer-resource interaction, so that in the main, competition communities are investigated. The design of the model permits an analytical study of multi-species systems (say 5 to 100 species). This contrasts with analyses of models normally presented in the literature which usually describe communities of only two or three interacting species. One feature of the ensemble model is that it makes allowance for environmental variations (which cause structural and/or population disturbances) by simulating the totality of possible states to which an ecosystem can be disturbed. It was found that feasibility - the requirement that all equilibrium populations of a system are positive - is a key factor. In fact, virtually all of the model's feasible states were stable. Feasibility was thus found to be a more critical factor than stability - even though it is the latter property which is normally concentrated on, in studies of model-ecosystems. The model presents an interpretation of communities that spend most of their lifetime close to an equilibrium. This limited view was then naturally extended, and it became possible to analyse communities that experience a relatively high disturbance rate, and therefore spend only a minor part of their lifetime close to any equilibrium. It is shown that persistent communities can possess the important qualities of conservation and recovery, without necessarily appearing to possess a stable equilibrium. The model demonstrates that environmental variability may promote coexistence. An examination is made of how community coexistence depends on species' relative competitive abilities and upon their abilities to "spread risks". As well, the response of a community to species invasions is analysed, and a species extinction curve is derived that corresponds qualitatively to that obtained from field-data on the Hawaiian avifauna. The notion of a competition community is then discussed. Although a pair of species might appear to be competing when viewed in isolation, their interaction could well be facilitative if viewed within a community context. This phenomenon appears to be prevalent 1n nearly all of the observed competition communities I examined, and can be attributed to hidden "indirect effects" between species. The ensemble model provides an explanation as to why these facilitations occur so frequently. A detailed null test is performed in order to deduce whether bird distributions on some archipelagos are nothing more than random assemblages, as has been argued by Connor and Simberloff (1979). The design of the null test 1s unique and makes use of a specially formulated C-score statistic to determine the checkerboard patterns within biogeographic data. The test adheres faithfully to the constraints outlined by Connor and Simberloff, whereas other attempts reported in the literature have failed to do so. The data is shown to have signicantly large checkerboard distributions when compared to a null model. Even so, analysis of the New Hebrides bird data (when examined at the family level) indicates that it is the "coexistence principle" which shapes community organization, rather than the "competitive exclusion principle".