Stacking of Canopy Spectral Reflectance from Multiple Growth Stages Improves Grain Yield Prediction under Full and Limited Irrigation in Wheat

Grain yield (GY) prediction for wheat based on canopy spectral reflectance can improve selection efficiency in breeding programs. Time-series information from different growth stages such as flowering to maturity is considered have high accuracy predicting GY and combining this multiple could effectively accuracy. For this, 207 cultivars lines were grown full limited irrigation treatments, their was measured at the flowering, early, middle, late grain fill stages. The potential of temporal evaluated by a new method stacking data. Twenty VIs derived used input feature support vector regression (SVR) predict each stage. predicted values trained linear (MLR) establish second-level model. Results suggested that (R2) data single ranged 0.60 0.66 0.35 0.42 respectively. increased an average 0.06, 0.07, 0.07 after two, three, four stages, respectively, under irrigation. Similarly, irrigation, 0.03, 0.04, 0.04 Stacking important increase application stable model usefulness obtained phenotyping platforms.