The effect of parameter uncertainty on whole-field nitrogen recommendations from nitrogen-rich strips and ramped strips in winter wheat

Abstract

This paper estimates the relative profitability of using optical reflectance-based measures to predict crop needs for topdress nitrogen application to hard red winter wheat ( Triticum aestivum). The data are from nitrogen yield response experiments where midseason optical reflectance data were recorded. Both nitrogen-rich strips and ramped strips are considered. Unlike past research, optimal nitrogen recommendations are calculated with and without accounting for parameter uncertainty. The expected profit-maximizing strategy is to follow the historical extension advice of applying 90 kg ha −1 preplant nitrogen using anhydrous ammonia. This strategy is more profitable than the best optical reflectance-based prediction system by $18.74 ha −1. When anhydrous ammonia is unavailable and preplant nitrogen must be applied as dry urea, the extension recommendation and the optical reflectance-based predictors are not significantly different. Ramped strips are no better than nitrogen-rich strips. Accounting for estimation uncertainty in the parameters increases expected profit by about $10 ha −1.