Revisiting the critical nitrogen dilution curve for rice

The critical nitrogen (Nc) dilution curve is commonly applied in crop nitrogen (N) management, growth analysis, crop modelling, and remote sensing. More than 10 rice Nc dilution curves have been published, showing large variation across varying genotype × environment × management (G × E × M) conditions. This study aims to investigate the uncertainties in rice Nc dilution curves and their parameters, validate the variability and reliability of published Nc curves, with an attempt to provide universal Nc dilution curves that can be applied across varying G × E × M conditions. Nc dilution curves from rice field experiments with different N application rates, covering 61 G × E × M combinations, were evaluated by using Bayesian theory. The correlation between the parameters of the derived dilution curves and genotypic, environmental, and management drivers were analyzed. The results indicate that Nc dilution curve parameters A 1 and A 2 are significantly associated and affected by the changes in G × E × M conditions. The Nc curves obtained in this study provide different Nc estimations compared to the published Nc curves. Variations in genotype, planting conditions, and management practices have significant impact on curve parameters. Our newly derived Nc dilution curves based on rice genotypic differences and growing regions provide more reliable estimations of Nc. This finding provides new insights into how Nc changes with genotypic and environmental factors, contributing to enhanced N management strategies and improved modelling of rice N-biomass relationships across environments. • We revisited the critical nitrogen concentration (Nc) dilution curve for rice with data covering 61 G × E × M combinations. • The Nc dilution curve parameters A 1 and A 2 change with genotype and environments. • Most published Nc dilution curves of rice are less reliable in estimating Nc. • We derived Nc dilution curves based on rice genotypes and planting environments, which enable more reliable Nc estimation. • Our results help improve rice N diagnosis and modelling of rice-N relationships.