A multivariate approach to study the maintenance of soil fertility and soil health in relation to yield and nutrition of rice (Oryza sativa)

Abstract

Understanding the relationship between soil properties and grain yield and yield related traits is an important objective in crop production programme. The purpose of this study was to evaluate maintenance of soil health in terms of soil chemical properties in relation to rice yield using multivariate analytical methods. Canonical correlation analysis is one of the most popular multivariate analysis techniques to estimate dependent relationship between soil properties with grain yield and yield attributing characters .In this study, five canonical correlation coefficients ( CCCs) were estimated and the first three of them were significant( R1: 1, R2: 0.99 and R3: 0.84, p<0.001).The findings obtained from the CCA indicate that available N, available Cu and Zn in soil and N content in grain had the largest contribution for the explanatory capacity of canonical variables estimated from grain yield, straw yield, number of panicles/m2 and filled grains / panicle.