Prediction of grain yield and nitrogen uptake by basmati rice through in-season proximal sensing with a canopy reflectance sensor

The present study was conducted to establish prediction models for grain yield and nitrogen (N) uptake using normalized difference vegetation index (NDVI) measurements with the GreenSeeker optical sensor for different cultivar groups of basmati rice (Oryza sativa L.) and to define the optimum sensing timing. Sensor readings were collected at 21, 28, 35, 42, and 49 days after transplanting (DAT) from multi-cultivar and multi-rate N fertilization experiments conducted in 2016 and 2017. Prediction model established by regressing NDVI day(-1) as the determinant of plant biomass with grain yield and N uptake at maturity following exponential functions revealed that sensing the crop before or after 35 DAT (panicle initiation stage) was not accurate and did not predict satisfactorily the yield or N uptake potential. Regression analysis generated two potential and viable yield or N uptake prediction models: one for the basmati rice cultivar CSR30 (tall cultivar), and the other for a PB-PUSA (group of semi-dwarf cultivars). Validation of the prediction models using an independent experiment conducted in 2018 revealed that sensing the crop at the panicle initiation stage provide grain yield and N uptake predictions close to the observed grain yield (R-2 = 0.86, RMSE = 6.1%) and N uptake (R-2 = 0.75, RMSE = 8.5%). This study showed that yield and N uptake potential in basmati rice can be predicted using in-season NDVI data measured with the GreenSeeker optical sensor.

Automated summary

The study established prediction models for grain yield and nitrogen uptake using NDVI measurements with the GreenSeeker sensor in different cultivar groups of basmati rice. Sensing the crop at the panicle initiation stage provided accurate predictions for grain yield and N uptake potential. This demonstrates that in-season NDVI data can be used to predict yield and N uptake potential in basmati rice.