Remote Sensing Of Wheat Powdery Mildew And Stripe Rust And Spore Concentration Dynamics Of Blumeria Graminis F.sp.triciti In The Field

In this study, wheat canopy hyperspectral reflectance of wheat powdery mildew and wheat stripe rust and both mixedly occurring zone was detected at different wheat growing stage in 2015, and digital images of 50m、150m and 250 m height was acquired using unmanned aerial vehicle at late grain filling stage. The relation between different disease index and severity and hyperspectral reflectance was analyzed, and characteristic wavelength and color parameters screened and disease index(DI) of wheat powdery mildew and disease severity of wheat stripe rust estimating model based on parameters established. The relation between color features extracted from digital images and disease index, average severity, yield, 1000-kernel weight was analyzed, of which estimating model based on color features were built. Conidia of Blumeria graminis f. sp. triciti(Bgt) in the air of Jingshuang16 and Zhongmai2 planting field was trapped using volumetric spore samplers, and meteorological factors monitored by meteorological station. The correlation between accumulation periods of spore concentration and DI of Jingshuang16 and Zhongmai2 planting field were analyzed and DI of wheat powdery mildew estimating model based on accumulation of spore concentration was established. The main results are as follows:There exists negative correlation of significant level between DI of wheat powdery mildew and canopy hyperspectral reflectance at near-infrared 780~880nm band, while the correlation at visible 380~780nm band is promiscuous and unstable. From anthesis to late grain filling stage, there is significant correlation between DI of wheat powdery mildew and DVI and SAVI and SDr. The correlation coefficient between disease severity of wheat stripe rust and canopy hyperspectral reflectance at 580~680nm band is over 0.8 at at anthesis, early grain filling and late grain filling stage, reaching significant level. At grain filling stage, the correlation between disease severity of wheat stripe rust and RVI, NDVI, SAVI, λred, dλred, SDr reaches significant level, of which there always exists negative correlation of significant level of RVI and NDVI from anthesis stage. And estimating model of disease severity of wheat stripe rust based on R689, R701 and R692 reaches significant level. When wheat powdery mildew and stripe rust mixedly occurring, the correlation between DI of wheat powdery mildew and canopy reflectance at near-infrared 780~880nm band will be decreased by stripe rust.In wheat powdery mildew occurring zone, the correlation between R、G、S of digital image and disease index, yield, 1000-kernel weight is the most significant, of which the correlation of S is the most stable, the exceeding significance of which existing continuously at 50m、150m and 250 m height. This study demonstrate that R captured from digital image of 50 m and 150 m height in wheat powdery mildew and wheat stripe rust mixedly occurring zone is significantly and negatively correlated to wheat yield and 1000-kernel.The significance and efficiency of the estimating model of DI of wheat powdery mildew in Zhongmai2 planting field based on the accumulation of spore density before the current week of disease measuring or on the accumulation of spore density before disease measuring is the best, while the most superior independent factor for Jingshuang16 is accumulation of spore density before disease measuring. Spore concentration in the air of Zhongmai2 planting field in 2014 and 2015 and that in the air of Jingshuang16 planting field is significantly and positively correlated to air temperature.