How Does Sensor-Based Variable Rate N Application Differ from Uniform N Rate?: Lessons Learned from Corn on-Farm Studies in Alabama.

Adoption of sensor-based variable rate nitrogen (N) by corn producers in the Southeast USA is still in progress. Questions related to the accuracy of the sensors assessing N status, significant yield differences with uniform N application, and N savings are often raised among producers.  Corn On-farm studies were conducted over two years in Alabama to evaluate differences of sensor-based variable rate and uniform rate N application, and to evaluate the on-the-go performance of the rate controller by comparing differences between the prescribed N rate and the as-applied N rate. Three different treatments were evaluated: two variable rate N treatments using the Great Plains Oklahoma algorithm vr. 1.3, consisting of N applied at V6 (T1) and V8 (T2) growth stages, and an uniform rate chosen by the producer (T4). Every treatment consisted of six rows spanning the length of the field replicated six or seven times. A GreenSeeker RT200® system mounted on a Hagie Sprayer was used to collect the canopy reflectance data with the RT200 providing the target N rate to a Raven VIPER Pro®. Data from 2013 showed that at field 1, there were no statistical yield differences between T1, and T4. The producer’s treatment (T4) did out-yield T1 by 0.44 Mg ha^-1 but an additional 25 Kg N ha^-1 was applied for T4. Significant yield differences among treatments were also observed at field 2. The lowest yield and amount of N applied resulted from T2. Although there were no significant differences between T1 and T4 (producer’s rate), T4 did out-yield T1 by 0.44 Mg ha^-1 with less than 9 Kg N ha^-1 applied for T4.
Evaluation of the rate controller performance, comparison of the prescribed versus as-applied N rates, involved grouping the data into rate quartiles (<25^th, 25^th - 50^th, 50^th - 75^th, >75^th percentile). Overall for both fields, data indicated that at low rates (rate values in the lower quartile, 25^th percentile), the as-applied rate was higher than the prescribed rate. In contrast, at high rates (rate values in the upper quartile, 75^th percentile), the as-applied rate was lower than the prescribed rate. Small differences existed between the rates located in the 50^th percentile of the distribution. Even though these differences existed, data showed that the as-applied rate did follow the general prescribed rate trend. Although this study only involved two site-years, results appeared promising not only for improving nitrogen use efficiency, but also using crop sensors for assessment of corn N need differences early in the growing season.