Date of Award
1985
Degree Type
Dissertation
Degree Name
Doctor of Philosophy
Abstract
The reader of this thesis is presented with a view of Vegetation Science as a dynamic, developing field. Changes are seen as producing increased complexity in the methodology and increased emphasis on prediction. As pointed out in the text, early work has mainly been descriptive and the methods of analysis used were not so much prediction oriented as they were a vehicle for presenting current descriptions of vegetation structure and function. Method development is traced from the early descriptive schemes through more evolved ones which have reflected different and often opposing concepts of what vegetation is. The most noted are the ideas of discrete types versus the idea of continuous variation which negates the existence of types.;The methodology of Vegetation Science continues to see changes. It is proposed that at this point in the evolution of the field, it will be more productive, in a statistical sense, if methods are applied in concert rather than individually. With this in mind three method types are considered. The three are complementary both in their aims and through the revealing of information which enables the user to more successfully apply the others. The concept at the base of these methods presented in the thesis is that of continuity of vegetation types. The first method type is that of simulation modelling. Past efforts have been effective mainly in simulating the dynamics of small systems. A model has been developed to effectively simulate a broad system, an entire ecoregion in the Yukon Territory. The second method type, time static modelling, a novel method in that it incorporates formally the idea of continuity of types, is concerned with the prediction of eventual states of the vegetation following perturbation, without concern for time. The final method type is a fairly recent method, called non-linear predictive ordination, which by imbedding specific resemblance measures, helps to reveal the complex underlying structure of vegetation data. The three types of methodologies are inevitable when the aim is to lay a complex system open to interpretation.
Recommended Citation
Fewster, Peter Humphreys, "Methodologies Of Predicting Vegetation Response To The Environment And To Perturbation On A Regional Level: A Thesis In Three Parts Treating Temporal And Time Static Models" (1985). Digitized Theses. 1410.
https://ir.lib.uwo.ca/digitizedtheses/1410