A multivariate approach to a specific problem of grouping maize cultivars

The use of multivariate techniques in the analysis of multivariate problems is illustrated by comparing the results of univariate and multivariate techniques applied to the problem of establishing the nutritional requirements of, and the acid tolerance differences between maize cultivars. Forty-eight maize cultivars were statistically separated into three groups, tolerant, intermediate and intolerant, using a univariate approach. A principal components analysis was then carried out to study the grouping at a multivariate level. The variates included were grain yield, plant height and ten leaf chemical analyses: Al, Mg, P, Ca, K, Mn, Zn, Fe, N and Cu. A non-hierarchical classification was applied to classify cultivars into the three tolerance classes. The univariate method resulted in different groupings for each variate under study, while the multivariate approach ensured one single classification of all cultivars into the three groups.