Response Of Soil Carbon Nitrogen And Microbial Characteristics To Water Regulation And Hyperspectral Monitoring

Drought and lack of rain is one of the important factors limiting agricultural production in northern China.Reasonable water and fertilizer applications are the basis of promoting the healthy growth of crops,however,the transformation of material and energy in soil are highly related to the function of soil microorganism.Therefore,the effects of water regulation on soil carbon nitrogen content and microbial characteristics were studied in combination with the water requirement of crops at different growth stages,it is of great significance to promote the development of water-saving agriculture in China.Winter wheat and soybean are the main crops in China,and multiple cropping of wheat and soybean is an important way to improve the utilized ratio of farmland.This study was carried out with different water regulation treatments on the multiple cropping of wheat and soybean to observe the changes of soil organic carbon,total nitrogen,microbial biomass carbon,nitrogen content,sucrose and urease activities.Moreover,the relationship between soil carbon nitrogen content and microbial characteristics and soil moisture content has been revealed as well.On this basis,the method of principal component analysis(PCA)was used to construct a comprehensive index(CI)of soil carbon nitrogen and microbial characteristics in study,and the origin spectral reflectance were transformed by reciprocal logarithm(lg(1/R),Lg),first derivative(FD),multiple scattering correction(MSC),standardized normal variable(SNV)and continuous removal(CR),and the PLS was finally used to build the hyperspectral monitoring model and realize the rapid monitor on soil carbon nitrogen content and microbial characteristics under the mode of multiple cropping of wheat and soybean.The main conclusions were as follows:(1)There were large differences in wheat and soybean fields soil organic carbon,total nitrogen,microbial carbon,nitrogen content,sucrase and urease activities,which were affected by the weather,fertilization and other factors.However,the response of each index data on water regulation in the same growth stage was basically same.As the drought stress intensifies,it shows the trend of increasing,decreasing or increasing first and then decreasing,in which the first increasing then decreasing is the most,and the water treatment corresponding to the highest point is different.(2)Various indexes have good correlation among them.The constructed index of CI based on PCA can represent 95.173% of the original information,and all indexes and CI were normal distributed.(3)Crop type and water regulation had no effect on the spectral reflectance of soil with different particle sizes.While the water regulation had effect on the spectral reflectance of soil with the same size,and the spectral reflectance of soil with different particle sizes varied at the same wavelength.It shows that the spectral reflectance of soil with the same particle size at the same wave length increased with the increase of drought stress,while that of soil with the same water treatment decreased with the increase of soil particle size.(4)The spectral transformation of Lg,FD,MSC,SNV and CR have different effects on the change of soil spectral reflectance,however,lead to obvious difference on the correlation between the soil spectra and the soil organic carbon,total nitrogen,microbial biomass carbon,nitrogen content,sucrase,urease activity and CI,on which,the FD,MSC,SNV,CR obtained better effect,while LG produced negative effect.(5)The predictive performance of all models of soil microbial biomass carbon,nitrogen content and sucrase activity was not good,while models of soil total nitrogen content and urease activity performed good with R~2> 0.6 and RPD>1.4.Moreover,the monitor model of CI achieved good prediction based on the FD transformation which commonly produced good effect on improving the model accuracy.