These procedures generally work on the same principle (illustrated in Fig. 1.8).
Abundances (biomasses or equivalent measures) are measured in a suite of taxa
(Fig. 1.8a) in each of a set of samples, each with a number of replicates. A similarity
(or dissimilarity) matrix is then calculated for all pairs of replicates across all of
the samples in the matrix (Fig. 1.8b). A number of measures of similarity can be
used, but the Bray–Curtis coefficient of similarity,