Nitrogen determination in irrigated rice using spectral reflectance

Abstract

Precision N management using optical radiation sensors is a promising management strategy. Using a combination of three spectral reflectance bands, 22 vegetation indices (VIs) were calculated and evaluated for their efficiency in estimating N status in irrigated rice (Oryza sativa L.) during growth stages. The results obtained here have showed a promising strategy to include crop stages as covariate in generalist models to predict the N status parameters for irrigated rice. This approach is interesting because it reduces the need for specific models, with different structures, for each crop stage. In addition, including crop stages as a covariate in the prediction models allows knowing the rice N status according to the crop stage, which is essential for efficient N management in commercial crops. The results obtained here show the beginning of vegetative stage (V1–V9) significantly affects the prediction of all N status parameters. The dry leaf biomass (DLB), leaf area index (LAI), leaf nitrogen uptake (LNU), and nitrogen nutrition index (NNI) can be adequately predicted with combinations of just two VIs. These results show the importance of using active sensors with more than two fixed bands, preferably including a red-edge band, for effective crop N status estimation.