Automatic Classification of Phytosociological Data on the Basis of Species Groups

Abstract

In phytosociological tables species and relevés are ordered in such a way that the species with a more or less similar distribution pattern over the relevés are grouped together and, similarly, relevés with a more or less similar species content are placed next to one another. As stated previously (Werger 1974a), membership of a species to a species-group is not necessarily determined by association with any individual species in that group, but can also be based on the restriction to the general distribution range in the table of a group of associated species. The same applies mutatis mutandis to the relevés belonging to a relevé group (cf. Hull 1964–65). In this way clusters of relevés characterized by groups of differential species are easily spotted in the table (cf. Van der Maarel et al. 1978).

Nomenclature of species follows usage in Werger (1973).

Contribution from the International Working Group on Data Processing m Phytosociology. Part of this paper it based on an unpublished lecture presented at the Prague meeting of the Working Group (Stockinger & Holzner 1973b). Copies of that manuscript and of a description of the mathematical background of the method (Stockinger & Holzner 1973a) are available from the first author The computer program can be studied at Vienna, Nijmegen, and at the C.S.I.R.O., St. Lucia Qld., Australia (Dr. M B. Dale).

Supported by a grant from the ‘Fonds zur Förderung der wissenschaftlichen Forschung’, Austria.