Using ground-based spectral reflectance sensors and photography to estimate shoot N concentration and dry matter of potato

Two years experiments were set up to evaluate the performance of different vegetation indices (VI) to estimate shoot N concentration (Nc) and shoot dry matter (DM) for a potato crop grown under different nitrogen (N) treatments. Possibilities to improve the performance of VI using normalization by leaf area index (LAI) or camera-derived ground cover fraction (GC) were also investigated. Results indicated that Nc was significantly correlated to RRE (Near-infrared divided by red edge reflectance) and RRE/GC with a coefficient of determination (R2) of 0.62 and 0.78, respectively, indicating that inclusion of auxiliary parameter GC together with RRE substantially improved the correlation as compared to using only RRE. However, no significant correlation between Nc and RVI (Ratio Vegetation Index, near-infrared divided by red reflectance) or NDVI (Normalized Difference Vegetation Index) was found. However, DM was highly correlated to RVI and NDVI. Moreover, DM showed significant relationship (R2 = 0.86) with GC, highlighting its versatile usefulness in estimating agronomic variables DM and Nc, which are the core variables to assess N status of crops for a better N application.