Fertility Potential Classification of Rice Growing Soils Using Principal Component Analysis

Fifteen rice growing soils of different types were evaluated for their fertility potential considering thirty characters and using principal component analysis. The principal component analysis synthesized five components explaining 77 per cent of total variability. Component scores were calculated on the basis of which the soils were graded.

In case of the first component, some of the problematic soils gave good ranks which might have resulted from the inclusion of undesired soil characters affecting its fertility. Whereas grading with respect to the second, third and fourth components appeared to be more logical. It is suggested that the undesired characters which give unnecessary high scores to the problem soils while applying principal component analysis for fertility potential evaluation should be eliminated.