| 不同施氮水平下棉花叶片最大羧化速率的高光谱估测 |
| |
| 引用本文： | 张鑫磊,刘连涛,孙红春,张科,白志英,董合忠,李存东,张永江.不同施氮水平下棉花叶片最大羧化速率的高光谱估测[J].农业工程学报,2020,36(11):166-173. |
| |
| 作者姓名： | 张鑫磊 刘连涛 孙红春 张科 白志英 董合忠 李存东 张永江 |
| |
| 作者单位： | 河北农业大学农学院/省部共建华北作物改良与调控国家重点实验室/河北省作物生长调控重点实验室,保定 071001;山东棉花研究中心,济南 250100 |
| |
| 基金项目： | 河北省自然科学基金，国家自然科学基金（31871569），河北省现代农业产业技术体系（HBCT2018040201） |
| |
| 摘 要： | 植物叶片最大羧化速率(Maximum Carboxylation Rate,Vcmax)是表征植被光合能力的重要参数之一,研究叶片Vcmax对氮素的响应对预测植被光合作用和未来植被生产力具有重要意义。该研究以国欣棉9号为试验材料,设置4个施氮水平,测定棉花不同生育时期叶片生理参数和叶片反射光谱,探究Vcmax与生理参数及光谱指数的关系,建立叶片Vcmax最优反演模型。结果显示,棉花叶片Vcmax在苗期至蕾期不受施氮水平的影响,随着生育进程的推进,氮素逐渐成为影响叶片Vcmax高低的主要因素之一。此外,叶片Vcmax与含氮量的相关程度最高,远高于叶片Vcmax与叶绿素、比叶重的关系。对叶片Vcmax与光谱指数进行回归分析发现,由蓝光和红边波段组合的光谱指数能较好的预测叶片Vcmax,其中表现最好的为归一化差值植被指数(Normalized Difference Vegetation Index,NDVI697,445)和比值植被指数(Ratio Vegetation Index,RVI445,694),决定系数(R^2)均大于0.75。最后,以生理参数和光谱指数为基础,采用一般线性回归和多元逐步回归构建棉花叶片Vcmax估算模型,结果表明,由RVI445,694、光化学植被指数(Photochemical Vegetation Index,PRI)、修正型归一化差值植被指数(Modified Normalized Difference Vegetation Index,mND705)所构建的多元逐步回归模型精度最高(R^2=0.809,RMSE=16.93μmol/(m^2·s)),叶片含氮量和叶绿素所构建的多元逐步回归模型次之(R2=0.801,RMSE=17.01μmol/(m^2·s)),高于其他单变量线性回归模型。研究表明,利用高光谱指数可以有效地估测棉花叶片最大羧化速率,结果可为叶片Vcmax准确反演和评估光合能力提供支撑。
|
| 关 键 词： | 棉花 模型 最大羧化速率 高光谱 植被指数 |
| 收稿时间： | 2020/2/3 0:00:00 |
| 修稿时间： | 2020/5/11 0:00:00 |
Hyperspectral estimation of the maximum carboxylation rate of cotton leaves under different nitrogen levels |
| |
| Zhang Xinlei,Liu Liantao,Sun Hongchun,Zhang ke,Bai Zhiying,Dong Hezhong,Li Cundong,Zhang Yongjiang.Hyperspectral estimation of the maximum carboxylation rate of cotton leaves under different nitrogen levels[J].Transactions of the Chinese Society of Agricultural Engineering,2020,36(11):166-173. |
| |
| Authors: | Zhang Xinlei Liu Liantao Sun Hongchun Zhang ke Bai Zhiying Dong Hezhong Li Cundong Zhang Yongjiang |
| |
| Institution: | 1.College of Agronomy, Hebei Agricultural University / State Key Laboratory of North China Crop Improvement and Regulation/ Key Laboratory of Crop Growth Regulation of Hebei Province, Baoding, 071001 China;;2.Cotton Research Center, Shandong Academy of Agricultural Sciences, Ji''nan, 250100, China |
| |
| Abstract: | As the core of photosynthesis,the accurate prediction of maximum carboxylation rate(Vcmax)is crucial to photosynthetic rate and vegetation productivity.At present,there are many studies which have predicted Vcmax by physiological parameters and spectral data.However,the studies mainly focus on forest,and study on leaf Vcmax of cotton has not been reported.In this experiment,“Guoxinmian 9”was selected as the experimental material,and four nitrogen application levels were set to determine the physiological parameters and leaf reflection spectra of cotton leaves in different growth periods.The relationships were explored between leaf Vcmax and physiological parameters,spectral indexes,and the hyperspectral indexes that could accurately retrieve the leaf Vcmax of cotton were screened out.Then leaf Vcmax estimation models were constructed.The results showed that the leaf Vcmax of cotton was not affected by the level of nitrogen application from the seedling stage to the bud stage due to the nitrogen in the soil without nitrogen fertilizer is enough to ensure that the leaf Vcmax of cotton maintain a high level from the seedling stage to the bud stage.As the growth period progressed,nitrogen became one of the main factors affecting the Vcmax of the leaves,and the correlation between leaf Vcmax and nitrogen content(R^2=0.717)was higher than that of chlorophyll and leaf mass area.The Vcmax of the leaves had strong correlations with the combination of blue light and red-edge wave band through regression analysis of spectral indexes.The Normalized Difference Vegetation Index(NDVI)using reflectance at 697 and 445 nm,as well as Ratio Vegetation Index(RVI)using reflectance at 445 and 694 nm had the best fitting effects,and their values of R2 all exceed 0.75.In addition,from the 27 predecessor vegetation indexes,three vegetation indexes with a higher degree of correlation with leaf Vcmax were obtained:Photochemical Reflectance Index(PRI),Modified Chlorophyll Absorption Ratio Index(MCARI),Modified Normalized Difference Vegetation Index(mND705),and the absolute value of the correlation coefficient of them were greater than 0.6.Finally,based on physiological parameters and spectral index,the estimation models of leaf Vcmax were established by general linear regression and multiple stepwise regression.The accuracy of the multiple stepwise regression model established by RVI445,694,PRI and mND705 was highest(R^2=0.809,RMSE=16.93μmol/(m^2·s)),followed by the multiple stepwise regression model established by nitrogen and chlorophyll content in leaves(R2=0.801,RMSE=17.01μmol/(m^2·s)).In summary,leaf Vcmax of cotton is more sensitive to leaf nitrogen compared other physiological parameters.The accuracy of the leaf Vcmax estimation model based on leaf spectrum established in this study is higher than that using leaf nitrogen content,chlorophyll and leaf mass area as independent variables,indicating that it is feasible to invert leaf Vcmax through leaf spectrum.It is shown that hyperspectral index can effectively estimate the Vcmax of cotton leaves,and the results can provide support for accurate inversion of leaf Vcmax and assessment of photosynthetic capacity. |
| |
| Keywords: | cotton models leaf maximum carboxylation rate (Vcmax) hyperspectral vegetation index |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《农业工程学报》浏览原始摘要信息 |
| 点击此处可从《农业工程学报》下载免费的PDF全文 |
|
