Determination of nitrogen dilution curves of corn, canola, and spring wheat in Canada using classical and Bayesian approaches

Determining the dynamics of crop nitrogen (N) uptake is essential to optimize N management and increase the sustainability of crop production. The concept of critical N concentration helps to achieve this goal by defining the minimum shoot N concentration required to produce the maximum aerial biomass at a given time during the crop cycle. The objective of this study was to determine new critical N dilution curves of the main annual field crops used in Canada (spring wheat, canola and corn) using a classical approach well established and a new Bayesian approach and compare these curves with the curves previously published. A total of 557 data for the three crops were used, covering a wide range of pedoclimatic conditions, crop management and cultivars across Canada. These data were processed following a classical statistical approach involving two steps to determine critical points used to calculate the parameters of the critical N dilution curve and then validate these curves against N limiting and non-limiting points. Data were also processed with a new one step Bayesian approach which allows estimating the 95% credibility intervals of the calculated critical curves. For spring wheat, our analysis confirmed the results of a previous study showing that the critical N curve established for winter wheat was not adequate for spring wheat. The critical N dilution curve determined with the classical approach (Nc = 4.14 W−0.51) performed well in separating the N limiting to the N non-limiting data of the validation dataset. The use of a larger dataset than in the previous study conducted for spring wheat in eastern Canada allowed us to demonstrate that new critical, minimum and maximum N dilution curves for spring wheat has a larger domain of validity than suggested by results of previous studies. Our study shows the strong interaction between plant N and water status, and that N dilution concept allowed us to reveal the occurrence of interaction between water and N stresses for crops in Western Canada, explaining the lower efficiency of N fertilizer in western than in eastern regions. For corn, the refined critical N dilution curve was very close to the critical N dilution curve determined in France and in eastern Canada, but the Bayesian approach allowed to propose a specific critical N curve for Canada (Nc = 3.07 W−0.40). In addition, minimum and maximum N dilution curves for corn were determined in Canada's pedoclimatic conditions. For canola, the new critical (Nc = 4.26 W−0.26), minimum and maximum curves determined in this study were close to the curves found in previous studies during the vegetative stages. In addition, a new critical N curve valid after the flowering stage was determined.