Spectral Response And Agronomic Parameter Model Of Peanut Under Different Fertilization Treatments

In order to understand the relationship between spectral reflectance of peanut leaves and chlorophyll content,yield,yield components and quality under different fertilization conditions,it provides scientific basis for growth monitoring and yield estimation.Of this study is to design the 7 kinds of fertilization treatment,repeat 3 times of field experiment,the main growth period determination of peanut leaf spectral reflectance and functional leaf chlorophyll relative content of peanut after harvest yield and quality,to explore different fertilization conditions the different development stages of the peanut chlorophyll content（SPAD）model,the peanut yield and yield components models for predicting the factor spectrum estimation model and the spectral quality of peanuts.The main results are as follows:（1）The SPAD values of the leaves under different fertilization treatments showed a similar trend of overall increase and then decrease with the advance of different fertilization treatments within 70 days after peanut flowering.（2）There is a certain quantitative relationship between the spectra of peanut and chlorophyll in different growth periods,and the spectral parameters of peanut in pod stage have the highest correlation with chlorophyll,so the estimation model of SPAD of peanut under different fertilization treatments in pod stage is established.The power function model y=0.0235x^1.3791 with spectral parameter PSSRb was the best.The quadratic model y=0.0005x²+0.1209 x-2.4545 with spectral parameter m SR705 was the best for high fertilizer treatment.The power function y=0.0045x^1.2923model with spectral parameter m ND₇₀₅ was used to estimate the best effect.The best effect was estimated by using the spectral parameter m SR705 to establish the quadratic curve y=-0.008349 x²+1.098943x-28.021739.The quadratic curve y=0.3883 x²-34.887 x+787.04,which was established by the spectral parameter m SR705 under the treatment of"low fertilizer+calcium fertilizer+rhizobia",had the best estimation effect.The single linear model y=0.1756 x-3.3486 with spectral parameter PSSRb was used to estimate the best effect of"low fertilizer+zinc fertilizer+rhizobia"treatment.The y=0.0216 x²-1.7216 x+38.101 estimation model with spectral parameter PSSRb under the treatment of"low fertilizer+calcium fertilizer+zinc fertilizer+rhizobia"had the best effect.（3）The spectral reflectance and spectral parameters of peanut at pod stage and full fruit maturity stage were the best for predicting flower yield.In the peanut pod setting stage,the spectral reflectance model at the original spectrum R719 was selected,and the quadratic multivariate model y=-15428x²+9747.4 x-1502.1 was established to estimate the best effect.In the mature stage of peanut satiation,the spectral parameter SDr was selected to model the independent variable,and the established conic y=-9398.1x²+8457.7 x-1865.5had the best estimation effect.（4）The number of full peanut fruit was the best predictor in the ripening stage.The original spectral reflectance at the characteristic band of 400nm was selected for prediction,so the power function model y=10785x^2.0621 was established.（5）The prediction effect of peanut fruit mass was the best in peanut pod stage.The power function model y=2.5931x-1.949 was established by using the spectral reflectance of R719.（6）The best effect was to predict the protein content of peanut in the pod stage,among which the first-order differential spectrum at R671 was selected as the independent variable,and the best effect was obtained from the conic model y=-8E+08x²+424275x-31.318.（7）The prediction of oil content was the best at the ripening stage of full fruit.The quadratic curve model y=253.22 x²-400.85 x+216.2 was established by using the spectral parameters（sdr-sdb）/（SDr+SDb）.