Estimating nitrogen requirement of grain corn in Manitoba using optical spectral reflectance
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Optical sensors can measure optical (visible/near-infrared) reflectance and be used to assess crop canopy conditions. In this study, two hand-held active sensors (GreenSeeker® and CropCircleTM) and a passive aerial sensor (Red-Edge multi-spectral camera) were compared at three growth stages of grain corn (V4, V8 and V12) to predict in-season nitrogen (N) requirement. Active optical sensors have a light source. Passive sensors rely on sunlight; thus, their reflectance measurements are subject to changing sunlight conditions. Here a high reflectance area of canopy non-limited by N was used to standardize for variations in sunlight conditions between measurements days. The Normalized Difference Vegetation Index (NDVI) for all three sensors and the Normalized Difference Red-Edge index (NDRE) using the CropCircleTM and Red-Edge were also compared. Four site-years (2018-2021) of N response trials were combined to capture N response under different meteorological conditions and create a regional response model, adjusted for N fertilizer and corn grain prices to determine the optimum N rate to apply. Measured grain yield significantly increased (adjR2=0.40) with N supply (spring soil nitrate plus N rate). The maximum return to nitrogen (MRTN) using a current high price ratio ($N: $Corn) of 9.15:1 was 177 kg N/ha for 7,986 kg grain/ha. Two methods were used to make N addition recommendations. The first was using a quadratic response model for grain corn yield to N supply. The second was the optical sensor approach compared the difference between a non-limited area and the field estimate using canopy spectral reflectance. The optical sensor approach (187 kg/ha N) was the closest to the determined MRTN of 177 kg N/ha. Standardizing light conditions at V4 provided significant associations of NDVI and NDRE with yield regardless of the sensor. At V8, only Red-Edge NDVI and NDRE were improved by standardization. Standardization had no effect at V12. For determining in-season N addition to grain corn in Manitoba, it is best to determine NDRE using the CropCircle V12 (adjR2 = 0.62). However, it is recommended for Manitoba farmers to standardize reflectance values to an N non-limited crop area as they prefer earlier timing for top or side-dressing corn between the V4 to V8 developmental stages.