Indicator Species Analysis, A Divisive Polythetic Method of Classification, and its Application to a Survey of Native Pinewoods in Scotland

Abstract

Indicator species analysis is a divisive polythetic method of numerical classification applicable to large sets of qualitative or quantitative data. It incorporates a key which enables new data to be assigned to the classificatory framework. Each dichotomy is established in several steps. First, a one-dimensional reciprocal averaging ordination is computed. The stands are then divided into two groups according to whether they fall on one side or the other of the centre of gravity of the ordination. Five indicator species are then identified which discriminate as well as possible between the two groups of stands. These are then used to construct a secondary ordination. The balance between the indicator species in the secondary ordination provides an objective and easily applied criterion for identifying the two groups of stands which can be used conveniently in the form of a key. The method has been applied to a survey of the native pinewoods of Scotland. The main floristic variation in these woods is related to two gradients: wetness and pH. The indicator species analysis helped to highlight the gradients, and provided a means of pigeon-holing the stands into acceptably homogeneous groups as a basis for further, more detailed, studies