Evaluation of Vegetation Indices (VI) for Their Sensitivity to Corn Biomass and Chlorophyll Content Changes Associated with Nitrogen Fertilization.

Application of canopy sensors for nitrogen (N) management in corn production in the Southeast US requires first the identification of the relationship between field-measured crop status  and sensor-measured canopy spectral reflectance, therefore reflectance values can be used to estimate in-season yield potential.  A nitrogen test was conducted between 2010 and 2012 at three locations in Alabama to evaluate different vegetation indices (VIs) for biomass and chlorophyll (Chl) content assessment at various corn growth stages. The experiment comprised of five different N rates; 0, 56, 112, 168, 224, 280 kg ha^-1 applied at planting. At each location, data of SPAD, LAI, and canopy spectral reflectance were collected at V6, V8, V10 growth stages. Spectral reflectance data was collected using two commercial sensors brands; GreenSeeker (GS) and Cropcircle (CC). GreenSeeker data was used to calculate the normalized difference vegetation index (NDVI) and CC sensor was calibrated to collect reflectance at 670nm (Red), 730nm (Red edge), 760nm (NIR) wavelengths. Ten VIs were calculated using CC data including NDVI, red-edge NDVI (NDRE), simple ratio (S ratio), simple ratio red-edge (S ratio-RE), inverse simple ratio red-edge  (Inv. S ratio-NDRE), Carter and Miller index (CSM), Carter and Miller red-edge index (CSM-RE), chlorophyll index red-edge (Chl In RE), and Modified Datt index (Modif Datt). A canonical correlation analysis (CCA) was conducted to identify the VIs best correlating with field-measured crop status variables and the type of relations existing between both groups of variables. Canonical correlations were calculated per growth stage, location, and year. Preliminary results showed that VIs including the red-edge wavelength had the highest correlation with corn biomass and Chl content. Higher correlation of NDRE compared to NDVI was found in seven of nine growth stage-location-year combinations. Results show that VIs including the red-edge wavelength seems to be more sensitive to Chl content and biomass changes at multiple vegetative stages.