Precise nitrogen management is crucial for ensuring the yield and quality of brewing sorghum in Southwestern China, yet robust region-specific nitrogen nutrition diagnostic tools remain lacking. In this study, field experiments were conducted at three sites during 2023–2024, including four major cultivars and six nitrogen application rates. Classical nonlinear regression and hierarchical Bayesian hierarchical approaches were employed and compared to construct critical nitrogen dilution curves (CNDCs). Bayesian hierarchical analysis was further used to assess the uncertainty and quantify the sources of variation attributable to cultivars and environment. The results showed that parameter estimates differed between the two methods. Parameter a varied both among cultivars (95% credible interval width: 0.75–1.00) and experimental sites (0.73–1.00). In contrast, parameter b varied substantially across cultivars (0.088–0.108) but ranged from 0.096 to 0.099 across different sites. Soil organic matter, soil total nitrogen concentration, growing degree days, and rainfall showed significant correlations with parameter a, while parameter b was associated with thermal time during the vegetative stage. The preliminary regionally universal CNDCs (classical: critical nitrogen concentration (Nc) = 2.15 W⁻ 0.175 ; Bayesian: Nc = 2.42 W⁻ 0.205 ) showed good performance across cultivars and sites, with nitrogen nutrition index (NNI) fitting RMSE ≤ 0.14 and highly consistent outputs between the two modeling approaches (R 2 = 0.98). This study establishes the first robust region-specific CNDCs and provides a reliable tool for precision nitrogen management in brewing sorghum production in Southwestern China.
Yang et al. (Wed,) studied this question.