An examination of the floristic zone concept with special reference to the northern limit of the Carolinian zone in southern Ontario

Publisher Website
Google Scholar
Abstract
Several features of the concept of floristic zonation, as a special case of biogeographical zonation in general, are examined. A distinction between description and hypothesis testing provides a preliminary systematization of methods used to detect the presence of zonation by analysis of distribution records. It is argued that the use of range limits largely fails to take the statistical nature of distribution records into account. Accordingly, three ordinational methods which do not depend upon range limits are applied to presence–absence data for 100 species of vascular plants occurring in southern Ontario. Two of these methods, canonical correlation and principal-components analysis, explore the idea that linear combinations of differential weightings upon the members of a set of taxa may demonstrate clearer geographical trends than merely treating all taxa as equally important trend-setters. The results of the study appear to bear this out. No sharply demarcated zonal boundaries emerged from the analyses, but two of the methods indicated the presence of weakly developed tension zones which may be taken as evidence of floristic zonation in southern Ontario.

This publication has 0 references indexed in Scilit: