Predicting presence of proazulenes in theAchillea millefolium group

Phytosociological classification of vegetation is often justified on the grounds it provides a means of predicting properties other than floristic composition. It is difficult to find examples inliterature where this possibility of prediction is actually used. In this paper, we examine the qualitative variation in proazulenes of theAchillea millefolium group as a function of cytological, abiotic and floristic factors. Sixty-six sites in natural populations were examined. Twenty to forty individuals per site were tested in the field for the presence of proazulenes. Ploidy level of collected specimens was examined to differentiate between taxa within the aggregate. Soil analyses were also performed and species composition of the sites was recorded. The floristic data was classified and attributed to phytosociological units. Canonical correspondence analysis was used to investigate the relationships between environment and floristic data, t-tests were performed and correlation coefficients were calculated to distinguish the mutual relationships between environmental factors, populations and relative frequency of proazulenes-containing individuals. In the area under study, hexaploid populations of yarrows were characterized by a low frequency of proazulenes. They were confined to dry and nutrient-poor habitats of fallows, slightly moist grasslands and fringes. Tetraploid populations consisted of different types. Both high and low frequencies of proazulenes-containing individuals occurred. A high frequency of such plants per site was positively correlated with phosphate, magnesium and manganese and negatively with carbonate and hydrogen ion concentration in soil. The results show that effective prediction of proazulenes-containing individuals is possible from the species composition of the sites, which is a comparatively simple procedure avoiding more expensive cytological and chemical assessment.