Estimation of understorey grass production in Acacia tortilis-Cenchrus ciliaris silvopastoral system through Principal Component Regression analysis

摘要

This study deals with the establishment of relationships of understorey grass production in a 4x4 m spaced trees of Acacia tortilis-Cenchrus ciliaris based silvopastoral system. Data recorded on understorey grass production, aboveground (Photosynthetically active radiation) and belowground (soil temperature, moisture and tree roots at three depths, 0-15, 15-30 and 30-45 cm depths) factors were used in the study. Principal component regression analysis was employed for establishing relationships of understorey grass production with respect to the formulated principal components as predictors. While formulating the prediction models in estimation of understorey grass production, the predictor variables particularly the below ground variables (soil moisture, soil temperature and tree roots), were found to be highly correlated. The principal components removed the multicollinearity and served as predictors to meet the statistical criterion of being independents and provide best fitted relationships.