Evaluation of Maize Yield Variability Based on Soil Properties and Principal Component Analysis

Successful implementation of soil and crop management program requires quantitative knowledge of site characteristics and interactions that affect crop yield. Soil properties are one of the most important site variables affecting within- field yield variability. The objective of this research was to identify intercorrelations among soil properties (chemical, physical and biological) using principal component analysis (PCA) and their relationships with maize yield variability in field. Site variables (18) and maize yield were measured in selected parts of Astan Quds agricultural fields in Mashhad city. The principal component analysis was used to reduce the site variables numbers and remove multicollinearity among variables. The first four PCs with eigenvalues>1 accounted for > 67% of variability in measured soil properties. Soil properties were grouped in four PCs as: (PC1) Soil highly descriptive fertility potential, (PC2) Soil moderately descriptive fertility potential, (PC3) Soil permeability potential, (PC4) Soil aggregation potential. The results showed that the factor of soil highly descriptive fertility potential explained 43% of variance and accounted for 77% of the yield variability in the field. Principal component analysis allows explaining a major part of crop yield variability by removing the multicollinearity.