variation, linear regression models were calculated using soil physicochemical properties as regressors.
Principal component analysis (PCA) was applied to substantially reduce the multivariate data set into
few uncorrelated master variables called principal components (PCs). For their intuitive interpretation
in terms of SQs, the magnitude and the sign of the loading associated with each initial variable were
used. For SQ variation across soil types, the score values of the PCs were calculated and the descriptive
statistics presented in the form of box plots. For the assessment of their effects on maize growth, correlation
analyses were carried out. All statistical analyses were performed using SAS statistical package
(SAS Institute, 1999).