Impact of nitrogen and environmental conditions on corn as detected by hyperspectral reflectance

Abstract

Indices derived from hyperspectral reflectance spectra have the potential to be used as indicators of environmental stress in crops. This study uses canopy-scale, ground-based measurements of hyperspectral reflectance to demonstrate the temporal patterns in corn development under imposed fertility (N rate) and environmental (water availability) stresses. In 1998, two large areas in a 30-ha corn ( Zea mays, L.) field near Ottawa, Canada (45°18′N, 75°44′W) were supplied with 99 and 17 kg N ha −1, while the balance of the field received the recommended rate of 155 kg N ha −1. Reflectance measurements were taken nine times using a portable spectroradiometer at georeferenced locations within these areas. Individual reflectance-based indices demonstrated the relative differences between application rates and identified both nitrogen and water stresses at various times in the growing season. No single index was able to describe the status of the corn crop throughout the season. Canonical discriminant analysis provided accurate classification of samples by N rate during early, mid, and late season conditions with overall success rates of 70%, 88%, and 93%, respectively. A shift in importance from green-based derivatives to red-based derivatives was noted from mid to late season and attributed to the natural reduction in green pigments as the crop entered senescence. Canopy-scale photochemical reflectance index (PRI) was shown to be correlated with canopy radiation use efficiency (RUE). Mid-season water stress affected the relationship. Multiple years of data are required to demonstrate robust relationships between hyperspectral indices and corn ecophysiological status because of the interaction between environmental and nutrient stresses. Identifying areas of fields sensitive to weather-induced stresses will allow better management of N application. Timing the collection of hyperspectral image data at early and late vegetative phase could enhance precision agriculture by allowing supplemental nutrient application, identifying stress patterns and aid in yield forecasting.