Using Multiple Years of Historical Yield Data to Create Spatial Management Zones.

Variable rate seeding prescriptions for maize are often based on management zones derived from historical spatial yield data for individual fields. The objectives of our study were to spatially compare management zones created from historical spatial corn and soybean yield data using management zone creation methods and to evaluate maize yield response to plant population within management zones created with these same methods. This study was conducted at 6 sites with 5 or more years of prior yield data. Crop rotations were primarily corn-soybean for 2 locations except 1 year at 1 location where corn followed corn. At the third location we had 2 years of popcorn data to go along with corn and soybean data. The historical yield data files were processed to ensure positional accuracy and cleaned to remove anomalies in the spatial data prior to implementing the zone creation methods. Zones were developed using a frequency analysis tool in the Markov Chain Extension for ArcView 3.2, and ArcMap 10.2 as well as the FarmRite tool in SST Summit. Zones were developed using either a) corn and soybean yield data or b) only corn yield data. Plant population trials were established in 3 of the fields in 2015 and the other 3 fields in 2016. The size of the trials ranged from 12 to 61 ha. Individual plots ranged from 9 to 12 m wide by 400 to 1400 mg long. Seeding rate treatments ranged from 66,667 to 103,704 seeds ha^-1 in increments of 7,407 seeds ha^-1. Grain yield and moisture were estimated with calibrated GPS-enabled yield monitors. Grain yield response to plant population was evaluated on a whole field basis and within management zones created by the different methods to determine a) whether spatial differences existed for yield response and b) whether zone creation methods differed in their functionality for this specific variable rate decision objective.