Analysis of within-Field Yield Variability in Cotton on Large-Scale Field Tests in North Louisiana.

Nowadays, enhanced crop productivity can be partly attributed to wide-scale adoption of site specific management of agricultural inputs through precision farming technology. A demonstration trial adopting a variable rate technology/optical sensing system has been established in six fields in Tensas Parish, LA: A, B, C in 2012 and D, E, and F in 2013. Large strip plots consisted of 12 rows x ~ 300 m-long were established then divided into 10 x 30-m long subplots. The amount of total N applied to each subplot along with lint yield were determined. Deep core samples were collected from five geo-referenced sampling points in each strip plot before planting and after harvest, then divided into 0-15, 15-30, 30-45, 45-60 and >60 cm sections, processed and analyzed for inorganic N content. Using a combine cotton picker and a yield monitor, seed cotton yield for each of the subplots was determined. At planting, the widest range of soil inorganic N within the 0-15 cm depth was recorded in Site B with values from 8 to 62 kg N ha^-1 while the lowest range was from 9 to 28 kg N ha^-1 for Site F. These observations were in agreement with the variability on the amount of N fertilizer applied to each subplot with standard deviation values of 44 and 6 kg N ha^-1 for Site B and Site F, respectively. With lint yield as variable, the highest variability was recorded in Site A (131 kg ha^-1) and F (215 kg ha-¹) while the lowest variability was observed in Site C (50 kg ha^-1) and in Site E (163 kg N ha^-1) in 2012 and 2013, respectively. The application of variable N rate based on sensor reading was most profitable in Site A with $32.65 ha^-1 followed by Site C with $28.01 ha^-1 return.