Nitrogen Status Diagnosis And Precision Nutrient Management Of Winter Wheat Using Unmanned Aerial Vehicle-based Remote Sensing In The North China Plain

Precision agriculture aims to manage spatial and temporal variability using geographic information technology?GIS?,variable rate technologies?VRT?,yield monitors,global positioning systems?GPS?,and remote sensing technology?RS?,associated with key aspects of crop production for optimum profitability,sustainability,and protection of the environment.However,there is still a lack of precision management of winter wheat based on GIS and RS technology small-scale farming systems at the village level in the North China Plain.This study evaluated different approaches to non-destructive winter wheat N status diagnosis and developed precision nitrogen?N?management?PNM?strategies using eight-rotor UAV with multi-spectral camera Mini-MCA and e Bee fixed-wing UAV with multi-spectral camera Parrot Sequoia+at plot and village scales involving different N rate treatments,management systems,sites and years.At the same time,villahe scale precision management strategies based RS and GIS technology and geostatistical methods were also developed,and the potential of fertilizer-saving and nutrient use efficiency were evaluated.The main results were concluded as follows:?1?The potential of real-time N status diagnosis of winter wheat at plot scale was estimated by using eight-rotor UAV remote sensing.The top vegetation index(VI_top)and the red edge normalized vegetation index?NDRE?had a good performance to estimate aboveground biomass?AGB?and plant nitrogen uptake?PNU?,with no significant difference.Compared with two mechanistic and a semi-empirical strategies,the most practical and effective method for N status diagnosis is to use NDRE for estimating N nutrition index?NNI?non-destructively,with diagnosis accuracy rate of73?86%at the key growth stages.The potential of real-time N status diagnosis of winter wheat at village scale was estimated by using fixed-wing UAV remote sensing.The red normalized vegetation index?NDVI?and NDRE had a similar good performance as VI_topto estimate AGB and PNU,respectively.The most practical and stable method for N status diagnosis was proposed,which was to calculate N sufciency index?NSI?with vegetation indices and then to estimate NNI non-destructively,with diagnosis accuracy rate of 57-59%at the key growth stages.?2?Eight-rotor UAV remote sensing had a good the potential of yield estimation,which explained 89?93%of the yield variation.Scenario analysis using Fertilizer Optimization Algorithm?NFOA?was performed to evaluate PNM at plot scale.Compared with farmer management?FM?and regional optimum N management?RONM?,the PNM system decreased N application rate by 21?40%and 17?37%,and increased partial factor productivity of applied N?PFP_N?by 27?66%and 32?59%,respectively.Fixed-wing UAV remote sensing had a good potential for estimation of winter wheat yield?R²=0.85?.The green window strategy combined with N diagnosis results could be used in PNM at village scale.Based on scenario analysis,the difference between the green normalized vegetation index?GNDVI?and VI_topfor recommended N application rate was not significant,which was similar as economically optimal N rates.?3?Using geostatistical methods at village scale,precision management?PM?based on RS and GIS technology was also developed to manage the spatial variation of soil and plant nutrients.Compared with FM under the scenario analysis,the PM system decreased fertilizer rate for N,P₂O₅and K₂O by 44?68%,62%,respectively.The corresponding reductions relative to regional optimum management?ROM?were24?56%,48%and 93%,respectively.Compared with FM,PM increased profit by 1387and 1424?ha^-1in 2017 and 2018,respectively.Compared with ROM,PM increased profit by 834 and 871?ha^-1in 2017 and 2018,respectively.