Résumé
Analysis of data on diversity consists mostly of analysis of crossing tables of taxonomic units-species, populations, cultivars, etc.-and of variables that characterize the diversity. These variables are quantitative agronomic characters, morphological descriptors, mostly qualitative, biochemical markers (isozymes), or molecular markers (RFLP, RAPD), often coded as presence or absence. The taxonomic units are generally observed individually, but they can also correspond to populations; the observations are thus of means of characters, of allele frequencies, and so on. In any case, the analysis of these data aims to discover a possible structure of taxonomic units by analysis of resemblances or dissimilarities between these units.