中国农业气象

• 论文 • 上一篇    下一篇

 稻纵卷叶螟为害水稻的冠层光谱特征及叶绿素含量估算

 周晓,包云轩,王琳,杜正朕,唐倩,陈粲   

  1.  1.南京信息工程大学气象灾害预报和评估协同创新中心,南京 210044;2.南京信息工程大学江苏省农业气象重点实验室,南京 210044;3.南京信息工程大学气象灾害教育部重点实验室/气候与环境变化国际合作联合实验室,南京 210044
  • 收稿日期:2019-09-21 出版日期:2020-03-20 发布日期:2020-03-20
  • 作者简介:周晓,E-mail: 18751903716@163.com
  • 基金资助:
     国家自然科学基金项目(31601221;41975144);江苏省重点研发计划(现代农业)(BE2019387);中国博士后科学基金(2016M1884)

 Hyperspectral Features of Rice Canopy and SPAD Values Estimation under the Stress of Rice Leaf Folder

 ZHOU Xiao, BAO Yun-xuan, WANG Lin, DU Zheng-zhen, TANG Qian, CHEN Can   

  1.  1. Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science & Technology, Nanjing 210044, China; 2. Jiangsu Key Laboratory of Agricultural Meteorology, Nanjing University of Information Science & Technology, Nanjing 210044; 3. Key Laboratory of Meteorological Disaster, Ministry of Education(KLME)/Joint International Research Laboratory of Climate and Environment Change(ILCEC), Nanjing University of Information Science & Technology, Nanjing 210044
  • Received:2019-09-21 Online:2020-03-20 Published:2020-03-20
  • Supported by:
     

摘要:  2017-2018年以南京农业大学农业气象与昆虫生态实验室培养的二龄稻纵卷叶螟幼虫为试验材料,在南京信息工程大学农业气象试验站水稻大田内,设置网罩控制下不同投虫量处理,以无虫量为对照,利用便携式波谱仪和SPAD-502叶绿素仪测定水稻全生育期冠层光谱反射率以及倒一叶叶绿素相对含量(Soil and Plant Analyzer Development,SPAD),分析不同初始虫量条件下、不同生育期水稻冠层的原始光谱、三边参数和SPAD值的变化规律。在此基础上,利用观测光谱中与SPAD值相关性较强的波段,计算基于光谱数据的植被指数和三边参数,并以这两类指数为自变量,建立水稻叶绿素相对含量的回归估算模型。结果表明:(1)稻纵卷叶螟为害下的水稻冠层光谱反射率与对照差异显著(P<0.05),其中可见光波段的反射率在分蘖-孕穗期低于对照,在扬花-成熟期高于对照;(2)在近红外波段,表现为虫害处理的反射率低于对照,且随着投虫量的增加反射率呈下降趋势,其中以300头·百株-1(4级)处理近红外波段的反射率最低;(3)随着投虫量的增加,红边位置发生明显“蓝移”。(4)SPAD值随投虫量的增加逐渐降低,4级处理最低;(5)归一化植被指数(Normalized Difference Vegetation Index,NDVI)、红边归一化植被指数(Red-Edge Normalized Difference Vegetation Index,NDVI705)以及红边位置与SPAD值的相关性较好;(6)利用观测光谱数据构建了SPAD单因子估算模型和多元逐步回归模型,其中以NDVI指数模型的估算效果最优,R2达到0.72,且高于其它估算模型,说明利用水稻冠层光谱参数,建立全生育期稻纵卷叶螟为害下水稻叶片叶绿素相对含量估算模型的方法是可行的,且具有快速、无损的特点,利于实现持续、动态和长期定位观测。

关键词:  稻纵卷叶螟, 水稻冠层光谱, 植被指数, 叶绿素相对含量估算模型

Abstract:  The second instar larvae of Cnaphalocrocis medinalis Guenee (Rice Leaf Folder, RLF) cultured in the Laboratory of Agrometeorology and Insect Ecology of Nanjing Agricultural University were used as the experimental objects in 2017-2018. In the rice field of the agricultural meteorological experimental station of Nanjing University of Information Science & Technology, six levels treatments, including 50, 100, 200, 300, 400 RLFs·100plants-1and Control Treatment (CK) were set, in order to test the relationship between spectral characteristics and SPAD values under such different initial pest population condition. During the experiments, the spectral reflectance of rice canopy and the SPAD values of the flag leaves were measured by ASD Field Spec3 (Analytical Spectral Devices Inc) and SPAD-502 (Soil and Plant Analyzer Development), respectively. The variation patterns of original spectrum, trilateral parameters and SPAD values in different growth stages were analyzed. Consequently, the vegetation indices and trilateral parameters based on the observed spectral data were calculated using the reflectance values in the bands with strong correlations with SPAD. And the regression estimation models of the SPAD valuesof rice were established.The results showed that: (1) the spectral reflectance values of rice canopy damaged by RLFs were significantly different from those of the control (P<0.05). In the visible bands, the reflectance values of the former were lower than those of the latter during tillering-booting stages, and was higher than the latter during the flowering-maturity stages.(2) In the near-infrared band, the reflectance values of the injured treatments were lower than those of the control. And they decreased with the increase of the pest’s population, among which the reflectance of 300 RLFs·100plants-1(Level 4) was the lowest.(3) With the increase of the pest’s population, the red-edge position values appeared the ‘blue-shift’ trend. (4) The SPAD values decreased gradually with the increase of the pest’s population, and the Level 4 treatment had the lowest values. (5) The Normalized Difference Vegetation Index (NDVI), Red-Edge Normalized Difference Vegetation Index (NDVI705) and the position of red-edge had a close correlation with SPAD values.(6)The single factor estimation model and multiple stepwise regression model of SPAD were constructed using the hyperspectral features of rice. The exponential model with NDVI had the best estimation effect, with the highest coefficient of determination (0.72) among all the estimation models. The results indicate that it is feasible to use the spectral parameters of rice canopy to establish a model for estimating the SPAD values of rice leaves under the stress of rice leaf folder in the whole growth stages. This method is fast and nondestructive, and is convenient for continuous, dynamic and long-term located observation.

Key words:  Cnaphalocrocis medinalis Guenee, Canopy spectrum, Vegetation indices, Estimation models of SPAD

中图分类号: