| To deal with the key issues of quick access to crop nutrient information for precision agriculture technology system, by means of different monitoring tools, we carried out the winter wheat variable rate fertilization experiment on the National Precision Agricultural Research Basement of Xiao Tangshang, Beijing from 2013 to 2014. In this experiment, we used a set of Nitrogen-rich Strip (N-rich Strip) as a reference to the nitrogen nutrition management. N-rich Strip treatment supplied with 150%of the recommended N value at the over-wintering and green returned stage season. And a conventional uniform nitrogen recommended fertilization treatment, a no nitrogen treatment were setted as the control trial in this experiment. Variable-rate nitrogen treatments arranged based on parameters acquired by ASD, Multiplex 3, SPAD and Dualex in wheat jointing stage.In this study, crop canopy growth monitoring indicators (OSAVI, REPLI, NBI-R) and individual crop nutrient monitoring indexes (SPAD, NBI) which correlated good with leaf N contentin jointing stage, was compared with N-rich Strip corresponding parameter to determine nitrogen status. Then formulate fertilizer algorithm for each plot was established by the difference between parametes of variable-rate plot and the N-rich strip plot. In this study, wheat growing dynamic change among different growth period, yield and grain quality were compared in order to identify which parameters was sensitive to crop nutrient diagnosis. Furthermore, the spatial structure for all parameters were analyzed based on the theory of geostatistics, then Kringing interpolation was applied to the parameters.The spatial varation for all parameters among different growth stages were analyzed and compared. The results are summarized as below:a) The N-rich Strip indicators such as biomass, LAI and leaf nitrogen content were significantly higher than other treatments, and it can be used as an effective reference to the nutritional status evalution in field to guide variable rate fertilization application. Compared to the conventional uniform fertilizer treatments on the same fertilization level, plot of variable rate fertilization based on N-Rich Strip control, with the premise of high yield and good quality, have obvious advantages in reduced the spatial variation of growth parameters, yield and grain quality. The study indicated that variable fertilization based on N-Rich Strip reference can balance the difference of soil fertility and crop condition foundation to obtain relatively uniform growth and output, so as to achieve the purpose of variable rate fertilization.b) Yield variance for othervariable fertilization treatment and uniform treatment had no significant difference. But yield for no nitrogen treatment and N-rich Strip treatment was lower than other treatment. There are large differences in fertilizeramount and coefficient of variation for wheat growth parameters between the different variable rate fertilization treatments. Compared to variable rate fertilization based on crop individual indicators (SPAD, NBI), the effect of variable rate fertilization for the variable rate fertilization treatments based on crop canopy indicators (REPLI, NBI-R), was more obviously. Variable rate fertilization treatments based on REPLI, on the basis of biggest spatial variability on growth condition and fertilizer rate, obtained the highest yield level and relatively minimal difference degree, and its effect of variable rate fertilization was the most obvious under the condition of this experiment.c) The comparation of the Kriging interpolation figure of each parameters and leaf N at different stages, as well as the spacecorrelation analysis for each figure showed that, indicators based on the crop individual parameters (SPAD, NBI) is superior to the crop canopyindicators (OSAVI, NBI-R and REPLI). The spatial correlation between leaf nitrogen and yield, grain quality reach the best in jointing and heading stage Meanwhile, NBI and REPLI at tilling or heading stage, had the best spatial correlation with yield and grain quality. They are reliable to predict yield, quality and their spatial variation.d) Under this experimental condition, the crop individual growth indicators (SPAD, NBI) and crop canopy indicators’(OSAVI, NBI-R and REPLI) performance obtained different results in guiding variable rate fertilization. They also showed different spatial variation. With the development of agricultural mechanization industrialization, for the application prospects in the future, compared to the other field condition monitoring means in this study, crop nutrient management based on high spectral information will get more application space. How to extract and mining the abundant information contained in hyperspectral efficiently, the accurate quantitative inversion of crop growth information and quantify its real differences in space, is the key issue to hyperspectral appliction in crop nutrient management. |
PDF Full Text Request