Multivariate Approaches to a Limestone Grassland Community: III. A Comparative Study of Ordination and Association-Analysis

Abstract

A study based on quantitative and qualitative forms of the same data was made employing separate techniques of analysis, each appropriate to data of one of these kinds. The methods used exemplify ordination and association-analysis and the objective of the study was to evaluate the techniques employed. Both methods are derivations of factor analysis and were applied in forms corresponding to the R-and Q- techniques of factor analysis. It was concluded that the ordination technique employed results in a more informative and ecologically more satisfying model of vegetation structure than the form of association-analysis used. On the other hand, ordination required more elabotate data and the system of inter-relationships provided were less convenient for some purposes than classification. The advantages of association-analysis were considered to be the ease of collection of suitable data and the practical convenience of the classification produced; its principal limitations were the uncertain ecological status of the categories defined and failure to examine interrelationships between them.