Detecting effects of nitrogen rate and weather on corn growth using micrometeorological and hyperspectral reflectance measurements

Improved nitrogen management is one of the main challenges of precision agriculture. We need to apply nitrogen rates customized to fit local crop needs. The response of corn (Zen mays L.) to two rates of nitrogen, representative of the range used in variable nitrogen application, was assessed over a growing season by means of micrometeorological and hyperspectral reflectance techniques. The experimental field, which was located near Ottawa, Ont., Canada, was ploughed in early May and fertilized in mid-May. Most of the field area (23.7 ha) received the recommended nitrogen rate of 155 kg N ha(-1). A sub-optimal rate of 99 kg N ha(-1) was applied on 3.7 ha and a low rate of 17 kg N ha(-1) was applied on 2.6 ha. The field was tile-drained and was not irrigated. The vertical flux densities of water vapour and CO2 were measured in the 99N and 155N areas throughout the growing season using two eddy-covariance measurement systems. Hyperspectral reflectance was measured in the three areas, and the temporal variability in the water band index (WBI) is reported. The 1998 growing season was characterized by precipitation that was above normal in June but with rainfall deficits in July and August. The eddy-covariance measurements showed that maximum evapotranspiration occurred earlier than maximum net CO2 uptake in the corn canopy. For leaf area index (LAI) greater than I, corn in the 155N area had a water use efficiency (WUE), defined as the ratio of daily net crop CO2 uptake to evapotranspiration, of 18.9 g CO2 kg(-1) H2O compared to 20.4 g CO2 kg(-1) H2O for the 99N area. Greater evapotranspiration, related to faster leaf expansion in the 155N area, was mainly responsible for the lower efficiency. Total biomass accumulation, calculated using eddy-covariance and soil chamber CO2, fluxes, was generally similar for the 99N and 155N application rates, with values of 2211 and 2183 g dry matter m(-2), respectively. As well, harvest yields were similar for the 155N and 99N areas wuth dry grain yields of 0.85 and 0.83 kg m(-2), respectively. Yield of the 17N area was 0.80 kg m(-2). Spatial yield variability decreased with higher N application rates. Corn in the 155N area was more affected by water deficit during the active growing period than corn in the 99N area. This was also supported by the WBI results. The recommended rate of nitrogen application did not induce an economically significant increase in grain yield, because of the rainfall deficit in July and August. This experiment shows that tools for assisting producers in making management decisions regarding nitrogen application rates need to include a weather component. Such tools will improve the sustainability of nitrogen management. (C) 2001 Published by Elsevier Science B.V.