Predicting Wheat Yield by Principal Component and Regression Techniques based on Morpho-Physiological and Quality Traits

The study reviewed morpho-physiological and quality traits of 194 wheat genotypes to identify the traits determinant to wheat yield exploring principal component and regression techniques. The experiment was performed at Chaudhary Charan Singh Haryana Agricultural University, Hisar (Haryana) during 2023-24 crop season. Principal component analysis identified eight components that cumulatively explained 66.30 % of the total variation of 25 morpho-physiological and quality traits. The first principal component was associated with grain yield, grain filling duration, harvest index and 1000-grain weight and also showed strong association with quality traits viz., crude protein, gluten content and sedimentation value. The coefficient of multiple regression (R2) explained 86.74% of the variability and showed positive correlation between the dependent and independent variables. The stepwise regression analysis identified four traits - harvest index, biological yield per plot, chlorophyll content, and days to 50% heading - as significant contributors to grain yield in wheat. These variables were sequentially retained in the model based on their statistical significance indicating substantial proportion of the variation in yield. Hence, these traits emerged as the most reliable predictors of grain yield and may serve as important selection criteria in wheat improvement programmes.