Monitoring the sugar to nitrogen ratio in wheat Leaves with hyperspectral remote sensing

Chino. 碳氮代谢反映植株生理状况和生长活力，是小麦籽粒产量与品质形成的生理基础，因而叶片糖氮比的实时无损监测对小麦生长诊断和氮素管理具有重要意义。本研究的主要目的是通过分析小麦叶片糖氮比与冠层高光谱参数的定量关系，确立小麦叶片糖氮比的定量监测模型。采用不同蛋白质含量的小麦品种在不同施氮水平下进行了连续3年大田试验，于小麦不同生育期采集田间冠层高光谱数据并测定叶片糖氮比值，进而分析建立冠层高光谱参数与叶片糖氮比的回归模型。小麦叶片糖氮比随施氮水平的提高而下降，随生育进程呈“高-低-高”动态变化模式。利用高光谱对叶片糖氮比进行监测的适宜时期为拔节期至灌浆中期，其中开花期最好。水分特征参数FWBI和Area980与叶片糖氮比关系密切，指数方程拟合决定系数（R2）分别为0.762和0.768，估计标准误差（SE）分别为1.27和1.28。色素特征参数（R750-800/R695-740）-1和VOG2为变量，指数方程拟合决定系数（R2）分别为0.718和0.712，估计标准误差SE分别为1.87和1.95。经不同年际独立试验数据的检验表明，以参数FWBI、Area1190、（R750-800/R695-740）-1和VOG2参数为变量建立的叶片糖氮比监测模型表现很好，预测精度R2分别为0.627、0.618、0.691和0.795，预测相对误差RE分别为19.2%、18.7%、17.9%和18.3%。与色素指数和水分指数相关的特征光谱参数可以有效地评价小麦叶片糖氮比的变化状况，利用FWBI、Area1190、（R750-800/R695-740）-1和VOG2 4个参数可以对生长盛期的小麦叶片糖氮比进行可靠的监测。

Inglés. Carbon and nitrogen metabolism in crop plants reflects plant physiological status and growth activity, and provides physiological basis for yield and quality formation in wheat. Thus, non-destructive and quick assessment of soluble sugar to nitrogen ratio is necessary for growth diagnosis and nitrogen management in wheat production. The objectives of this study were to determine the relationships between leaf soluble sugar to nitrogen ratio and ground-based canopy hyper-spectral reflectance and spectral parameters, and to derive regression equations for monitoring leaf soluble sugar to nitrogen ratio in winter wheat (Triticum aestivum L.) with canopy hyper-spectral remote sensing. Three field experiments were conducted with different wheat varieties and nitrogen levels across three growing seasons, and time-course measurements were taken on canopy hyperspectral reflectance and leaf soluble sugar to nitrogen ratio during the period of experiment. Then, regression analyses were made on the relationships between leaf soluble sugar to nitrogen ratio and canopy hyper-spectral reflectance and spectral parameters. The results showed that the soluble sugar to nitrogen ratio in wheat leaves decreased with increasing nitrogen rates, taking on the trends of high-low-high pattern with growth stages. The proper time for monitoring leaf soluble sugar to nitrogen ratio should be from jointing to mid-filling, with best stage as anthesis. FWBI and Area980 of water-index were highly correlated with leaf soluble sugar to nitrogen ratio, with the determination of coefficients (R2) as 0.762 and 0.768 (exponential), respectively, and the standard errors (SE) as 1.27 and 1.28, respectively. (R750～800/R695～740)-1 and VOG2 of pigment-index were also significantly related to leaf soluble sugar to nitrogen ratio, with R2 as 0.718 and 0.712 (exponential), respectively, and SE as 1.87 and 1.95, respectively. Thus, the exponential equation with some sensitive water-index and pigment-index as variables could well describe the dynamic change patterns in estimating the leaf soluble sugar to nitrogen ratio in wheat, with better performance in water-index than in pigment-index. Testing of the monitoring models with independent dataset indicated that FWBI, Area1190, (R750-800/R695-740)-1 and VOG2 were the best indicators to estimate leaf soluble sugar to nitrogen ratio, with the predictive precision (R2) of 0.627, 0.618, 0.691 and 0.795, respectively, the relative error (RE) of 19.2%, 18.7%, 17.9% and 18.3%, respectively. Overall, the soluble sugar to nitrogen ratio in wheat leaves could be estimated by vegetation indices based on spectral characteristics related to water and pigment absorption, and FWBI, Area1190, (R750-800/R695-740)-1 and VOG2 could be used for reliably estimating the leaf soluble sugar to nitrogen ratio in wheat leaves.