Numerical phenetic taxonomy and its heuristic aspects

Abstract

The principal multivariate methods used in numerical taxonomy and their graphic representations are outlined. Advantages of, and problems with, numerical taxonomy and the definition of phenetic taxonomic characters, recent use of ultrastructural, biochemical and genetic tools and their application to taxonomy, and elucidation of non-cosmopolitanism and speciation is discussed. The application of a combination of molecular and numerical methodologies is recommended. The capability of numerical taxonomy to reveal new knowledge is explored. It could lead to recognition of multiple correlations hidden in the data, to ordination of populations, and the use of phenetic descriptors in defining multidimensional niches.