Comparison of Ground Based Canopy Sensing Technologies for Corn Nitrogen Management in Minnesota.

Various crop canopy sensing tools are being used to manage nitrogen, but their utility to predict N needs in Minnesota remains unclear. The objectives of this study are to compare the effectiveness of different canopy sensing technologies at predicting corn (Zea mays L.) yield at different development stages, and compare their capability to determine in season N deficiency. Six to seven N rates at 35 to 45 kg N ha^-1 increments were pre-plant applied in six fields throughout Minnesota and a rate of 100 to 125 kg N ha^-1 split-applied 1/3 pre-plant and 2/3 sidedress at V2, V4, V6, V8, and V12 development stages.  Canopy sensing measurements with SPAD, Crop Circle (NDRE) and GreenSeeker (NDVI), were obtained at V4, V8, V12 and R1. Grain yield was determined at harvest. Sampling earlier than V8, regardless of the tool, resulted in poor yield predictive power and N deficiency detection, likely because of mineralization of soil organic N exceeds plant requirements at early stages. GreenSeeker was well correlated to yield (R²=0.87) at V8 but index saturation past V8 weakened the correlations. Crop Circle had increasing predictive power until V12 (R²=0.95); by R1 tassels likely interfered with the measurements. Correlation with SPAD was 0.83 (V8), 0.73 (V12), and 0.94 (R1). Nitrogen deficiency was not detected by GreenSeeker at any stage. At V8, Crop Circle and SPAD underestimated- and at V12 and R1 overestimated-N needed to maximize yield. Sensor relative critical values ranged from 0.98 to 1.00. Using a typical 0.95 critical value at V8 the agronomic optimum rate was underestimated by 99 and 43 Kg N ha^-1 by SPAD and Crop Circle, respectively. Overall, the later the sensing the higher the yield predictive power, however the potential to improve yields by later N application diminishes, unless irrigation ensures N incorporation into the root zone.