Determining the Influence of Water Deficiency on NPK Stress Discrimination in Maize using Spectral and Spatial Information

Abstract

A necessary pre-requisite for variable rate application of plant nutrients to manage inputs in a site-specific manner is a technique to detect nutrient stresses in plants in real-time. Spectral reflectance of plant canopy may provide a non-destructive and rapid technique to detect nutrient and water stress in plants. In this study, we investigated the effect of water on nutrient stress discrimination (nitrogen, N; phosphorus, P; and potassium, K) based on the visible and near infrared reflectance of maize leaves grown under controlled greenhouse conditions. The analysis of the spectral data indicated that it was challenging to detect the nutrient deficiencies if the plant water status was unknown. However, if a priori knowledge of water stress existed, such as a well-watered situation, the ability to discriminate nutrient stress improved significantly. The analysis of whole plant behaviour (i.e., mean reflectance data from all leaves within a plant) versus leaves located at a specific location within a plant indicated that knowledge of spatial location of leaves within a plant helped to identify N and P stress and NPK as a multi-stress condition more clearly.