江西鹰潭地区早稻氮素营养光谱诊断模型的构建与应用

 【目的】建立基于水稻冠层归一化植被指数NDVI的早稻氮素诊断模型,并利用模型指导水稻生产实践,实现水稻氮素追肥的精确定量。【方法】基于不同氮肥处理的田间试验数据,建立了水稻冠层NDVI与植株吸氮量之间的定量关系,并利用独立试验数据进行了检验,筛选出最佳的氮素光谱诊断模型。在此基础上,结合水稻高产栽培经验,对早稻的追肥用量进行了实时推荐,并和当地农户施肥模式及氮肥-产量曲线计算出的最佳氮肥用量和产量进行对比。【结果】水稻关键生育时期(苗期、分蘖期、拔节期、抽穗期和灌浆期)的冠层NDVI与水稻植株吸氮量都呈显著正相关,相关性高于整个生育期,指数函数拟合效果要优于线性函数。独立试验检验结果表明,单生育时期的预测效果高于整个生育期,指数模型优于线性模型;其中,分蘖期和拔节期的光谱诊断模型表现最佳,预测精度分别为0.907和0.941。推荐施肥应用表明,与常规农户施肥模式相比,光谱诊断施肥模式在产量不减少的情况下降低了氮肥用量,提高了氮素利用率,而产量与常规施肥没有显著差异。在低密度和高密度处理下,光谱推荐施肥模式的施氮量与最佳施氮量的误差分别为2.12%、-2.22%;而产量与最高产量的误差分别为-0.75%、-5.04%。【结论】光谱推荐施肥模式可在保证产量的情况下,降低氮肥用量,提高氮肥农学利用率,在水稻生产中有较好的推广利用价值。

Nitrogen Nutrition Diagnosis of Early Rice with NDVI and Its Application for Nitrogen Topdressing Recommendation at Yingtan,Jiangxi Province

【Objective】Establishment of the nitrogen uptake diagnosis model for early rice based on NDVI, and recommend- ation of the N topdressing rate. 【Method】 The quantitative relationship between canopy NDVI and plant nitrogen uptake (PNU) was studied using the data of field experiment with different N treatments, and the models were tested using independent experiment data. The optimum diagnosis model was established to recommend the N topdressing rate with canopy NDVI. Then the recommended N rate, grain yield and nitrogen use efficiency (NUE) were compared with those of farmer’s N management treatment.【Result】 Rice canopy NDVI was well positive correlated with PNU at key development stage, the correlation for single stage was better than whole stage, and exponential curve better fitted the relation than linear function, the same for the test results with independent experiment data. The prediction precision of PNU diagnosis model at tillering and jointing stages was 0.907 and 0.941, respectively. Compared with the farmer’s N management, equivalent grain yield was obtained with reduced N rate while NUE was improved. The error between the N rate of spectral diagnosis fertilization and the optimum N rate calculated with the N-yield curve was 2.12% at high transplanting density and -2.22% at low transplanting density, respectively, while the error for grain yield was -0.75% and -5.04%, respectively. 【Conclusion】 Spectral diagnosis fertilization mode can reduce nitrogen application rate and raise nitrogen use efficiency on premise of grain yield, and has a potential to be widely applied in rice N management.