科研进展

【论文题目】NTRI: A novel spectral index for developing a precise nitrogen diagnosis model across pre- and post-anthesis stages of maize plants（NTRI：一种用于玉米开花前后精确氮诊断模型的新型光谱指数）

【作者】中英文

Yuzhe Tang（唐彧哲），Fei Li（李斐），Yuncai Hu（胡云才），Kang Yu（于康）

【摘要】

Context: Accurate and real-time diagnosis of crop nitrogen (N) status is essential for effective precision N management. Integrating the N nutrition index (NNI) with spectral non-destructive rapid monitoring technologies offers a promising approach to precision N management for field crops. However, applying spectral sensing technologies for providing fertilizer recommendations based on real-time plant N nutrition diagnosis for drip-irrigated maize in arid regions remains challenging.

Objective: Our study set out to leverage spectroscopic techniques to accurately diagnose maize N status at pre- and post-anthesis. Our goal was to develop a novel spectral index that could guide site-specific fertigation strategies in arid environments.

Methods: The comprehensive field experiments with three maize varieties and different N levels were conducted from 2021 to 2023 in Inner Mongolia, China. Spectral reflectance, biomass, and leaf N concentrations were determined at various layers of maize plants across five growth stages. A Bayesian model to estimate leaf-based NNI was employed to develop leaf-based critical N concentration dilution curves for different ecological sites. The nitrogen nutrient triangle ratio index (NTRI), a key outcome of our research, was constructed using first-

order derivative spectral reflectance between 680 and 750 nm. We then compared the NNI prediction accuracies of NTRI with 29 published spectral indices, ensuring the robustness of our findings.

Results: Compared to NNI prediction models based on twenty-nine published spectral indices, our newly developed NTRI demonstrated a superior correlation to NNI (R² = 0.83). Independent validation confirmed NTRI’s robustness, yielding an RMSE of 0.11 % and RE of 9.6 %, surpassing existing indices.

Conclusions: Pre-anthesis N diagnosis was most sensitive to spectral diagnosis from the latest fully expanded leaf, while post-anthesis N diagnosis relied on ear leaves. NTRI’s accuracy and resistance to varietal and interannual variability highlight its potential application for real-time N monitoring.

Significance: Our innovative spectral index NTRI significantly advances spectral N nutrition diagnostics, enabling leaf-layer sensing and smart fertigation systems. This breakthrough paves the way for sustainable, high-yield maize production in arid regions.

【关键词】Remote sensing; Nitrogen management; Nitrogen nutrition index; Spectral index