摘要
【目的】快速、准确地获得渍害胁迫下小麦SPAD状况。【方法】基于田间试验,分析了渍害胁迫8 d对8个小麦品种SPAD的影响及其与17个高光谱特征参数的相关关系,并建立小麦SPAD值的灾损估算模型。【结果】与正常小麦相比,受渍8 d时的SPAD值差异并不明显;而停止渍水7 d后,受渍小麦SPAD值较正常小麦显著降低;小麦SPAD值与17个高光谱特征参数中AR指数的相关性最高(r=-0.84),且其减少量随受渍时间的变化与品种的耐渍能力基本吻合;与基于高光谱特征参数的小麦SPAD灾损估算模型相比,基于偏最小二乘回归方法建立的模型估算精度较高,模型决定系数R^2达到0.984。【结论】AR指数的减少量可以用来反映不同品种小麦受渍情况,且基于PLSR建模方法可以有效进行渍害胁迫下小麦SPAD的灾损估算。
【Background】The waterlogging disaster is one of the serious disasters in wheat production of China,especially in the middle-lower reaches of the Yangtze River,where waterlogging disaster caused by excessive rainfall in the middle and late period of wheat growth was the major limiting factor for high and stable yield of wheat.Hyperspectral technique has been an effective method to monitor the crop growth as a rapid and non-destructive approach.And it is of great significance to clarify SPAD changes under waterlogging stress and their correlation with hyperspectral characteristic parameters for rapid identification of wheat disaster situation,timely understanding of crop damage,guidance of agricultural production management and yield estimation.【Objective】The purpose of this study aimed to present a non-destructive and rapid monitoring method for wheat SPAD monitoring under waterlogging stress.【Method】In order to reveal the effect of 8 d waterlogging stress to SPAD of wheat,a plot experiment was conducted after booting stage based on the waterlogging characteristics of precipitation in the growth period of wheat in this area.This paper analysed the effects of waterlogging stress on SPAD of 8 wheat varieties,as well as the correlation with 17 hyperspectral parameters,and then constructing predictive models for wheat SPAD waterlogging based on hyperspectral parameters.【Result】Compared with normal wheat,there was no significant difference for SPAD under 8 d waterlogging,while the SPAD was significantly decrease after stopping waterlogging for 7 d.The influence of waterlogging on wheat SPAD was significantly different with varieties of wheat,and the decrease of SPAD was significantly with weak resistance to waterlogging.Among the 17 hyperspectral characteristic parameters,the correlation between AR index and SPAD was the highest with the r of-0.84,and the change of its reduction amount with the time of waterlogging was consistent with the variety’s ability to resist waterlogging disaster.Moreover,as compared with the linear model built with hyperspectral characteristic parameters,the accuracy of partial least squares regression(PLSR)model based on the hyperspectral reflectance for SPAD reduction estimation was significantly improved,with the R^2 of 0.984.【Conclusion】These results indicated that the AR could be used for reflecting the change of SPAD of different varieties wheat,and the PLSR modeling method could be applied as an effective method for the damage estimation of wheat SPAD under waterlogging stress.Thus,hyperspectral technique could be used as a more effective method for obtaining wheat growth parameters as comparing with traditional destructive sampling methods,and provides a theoretical foundation and technical support for area monitoring and optimal management of wheat waterlogging.
作者
李燕丽
熊勤学
卢碧林
李继福
李磊
田广丽
LI Yanli;XIONG Qinxue;LU Bilin;LI Jifu;LI Lei;TIAN Guangli(Farmland Irrigation Research Institute,CAAS/National Agro-ecological System Observation and Research Station of Shangqiu,Shangqiu 476001;Agriculture College of Yangtze University,Jingzhou 434025,China;Jingzhou Agricultural Technology Extension Center,Jingzhou 434025,China)
出处
《灌溉排水学报》
CSCD
北大核心
2020年第12期41-48,共8页
Journal of Irrigation and Drainage
基金
农业部商丘农业环境科学观测实验站开放基金项目(FIRI2018-07-02)
国家重点研发计划项目(2018YFD0301301)
国家自然科学基金项目(31871516)。
关键词
渍水
小麦
高光谱
SPAD
监测
waterlogging
wheat
hyperspectral
SPAD
monitoring
