Active‐Optical Reflectance Sensing Corn Algorithms Evaluated over the United States Midwest Corn Belt

CORE IDEAS: Active‐optical reflectance sensor algorithms perform poorly outside the area for which they were originally developed. The red edge waveband is more sensitive to N stress than the red waveband. Some active‐optical reflectance algorithms are dependent on the sensor for which they were developed. Uncertainty exists with corn (Zea mays L.) N management due to year‐to‐year variation in crop N need, soil N supply, and N loss from leaching, volatilization, and denitrification. Active‐optical reflectance sensing (AORS) has proven effective in some fields for generating N fertilizer recommendations that improve N use efficiency, but locally derived (e.g., within a US state) AORS algorithms have not been tested simultaneously across a broad region. The objective of this research was to evaluate locally developed AORS algorithms across the US Midwest Corn Belt region for making in‐season corn N recommendations. Forty‐nine N response trials were conducted across eight states and three growing seasons. Reflectance measurements were collected and sidedress N rates (45–270 kg N ha⁻¹ on 45 kg ha⁻¹ increments) applied at approximately V9 corn development stage. Nitrogen recommendation rates from AORS algorithms were compared with the end‐of‐season calculated economic optimal N rate (EONR). No algorithm was within 34 kg N ha⁻¹ of EONR > 50% of the time. Average recommendations differed from EONR 81 to 147 kg N ha⁻¹ with no N applied at planting and 74 to 118 kg N ha⁻¹ with 45 kg of N ha⁻¹ at planting, indicating algorithms performed worse with no N applied at planting. With some algorithms, utilizing red edge instead of the red reflectance improved N recommendations. Results demonstrate AORS algorithms developed under a particular set of conditions may not, at least without modification, perform very well in regions outside those within which they were developed.