Effects of Nitrogen Nutrition on the Growth, Yield, and Reflectance Characteristics of Corn Canopies

Remote sensing offers the potential to determine rapidly the physiological condition of crops over large areas. To realize this potential, information is needed on the seasonal reflectance patterns of healthy and stressed crop canopies. A field experiment was conducted in which spectral and agronomic measurements were collected from corn (Zea mays L.) canopies grown under four N treatment levels (0, 67, 134, and 202 kg/ha) on a Raub silt loam (Aquic Argiudoll) soil. Data were collected at the Purdue Agronomy Farm, West Lafayette, Ind., on 11 dates during 1978 and 12 dates during 1979. Reflected radiation of 0.4 to 2.4 µm wavelength was measured with an Exotech 20C spectroradiometer and used to compute a reflectance factor. Agronomic data collected coincidently included biomass, leaf area index, plant height, crop development stage, percent soil cover, and surface soil moisture percent. Data were analyzed to determine: 1) the relationship between the spectral and agronomic characteristics of the canopies and 2) the spectral differences and separability of the four treatments. Nitrogen treatment effects were seen across the entire wavelength interval measured. Red reflectance was increased and near infrared reflectance was decreased from N-deprived canopies. Spectral differences between treatments were seen throughout each growing season. The near infraredred reflectance ratio (0.76 to 0.90 µm)/(0.63 to 0.69 pm) differed more between treatments than single band reflectance measures. Variations in canopy reflectance due to wetting and drying of soils were reduced by the use of the near infraredhed reflectance ratio. The spectra of the four N treatments were significantly different (Duncan's multiple range test, P = 0.05) on August dates. Early and late in the season only two spectral classes were resolved. Of the spectral variables examined, the near infraredhed reflectance ratio most effectively separated the treatments. Discrimination results for both years were similar in spite of larger treatment differences in 1979 than in 1978. Differences in spectral response between treatments were attributable to varying soil cover, leaf area index, and leaf pigmentation values, all of which changed with N treatment. The results further confirm the potential of remote sensing for monitoring the growth and development of crops.