Determining the most important soil fertility properties affecting rice yield in paddy fields using principal component analysis

Multi-variate statistical methods such as principal component analysis (PCA) and regressions could be used to facilitate the interpretation of complex relationships. The objective of this study was to determine the most important soil fertility properties affecting rice yield in the paddy fields. For this purpose, soil samples were taken from the plow layers of 119 points with suitable distribution in the paddy fileds located in Shaft and Fouman cities of Guilan province. Then after, physical and chemical properties of the soil fertility were measured and analysed using descriptive statistics, principal component analysis and regression methods. Results showed that three PCs with eigen values greater than one named as “k and it’s preservation factors”, ”Total N and it’s provider factors” and ”P and Thickness of plow layer” are respectively explained more than 67.4% of the variability in the soil physical and chemical properties and 55% of the yield variability. In addition, the corresponded properties to the PCs explained 80% of the yield variability. Consequently, in order to increase the yield, management practices such as proper fertilizer applications of nitrogen, potassium and phosphorous and proper tillage for creating suitable plow layer are recommended.