Management Zone Delineation for Variable Rate Seeding in Soybeans.

It is estimated that half of corn planters have the capacity for variable rate seeding (VRS), which may improve annual profits by allocating seeds more effectively within a field. Farmers frequently plant soybeans with a planter that can also be used for corn, and they have been adapting VRS techniques for soybean planting. Creating a seeding rate prescription has two steps: delineating management zones and assigning a seeding rate to each zone. Effective prescriptions will create zones that have different ideal seeding rates and have low within-zone variability. The objectives of this research are to compare different prescription methods in: 1) the relative importance they grant different soil characteristics, and 2) how homogenous the zones they create are. On-farm cooperators in Preble, Darke, and Morrow County, Ohio provided VRS prescriptions, each made with different methods. Half acre grid soil samples were taken before planting and were analyzed for soil physical and chemical properties. The soil properties were considered input variables in a decision tree formula (using the R package rpart) to determine which variables best predict the prescription’s management zones. In Darke County, pH was identified as the most predictive soil property, followed by potassium. CEC was the variable best at predicting zone in Preble and Morrow County. In Preble, calcium, magnesium, potassium, and texture were also used in zone delineation. Potassium and phosphorous were used in Morrow County, following CEC. The decision tree error is an indicator of how heterogeneous the soil is within a management zone, and the Darke County prescription had the least heterogeneous zones. Prescriptions based off of high resolution grid soil sampling were more successful at creating homogenous zones. Future studies will evaluate which soil characteristics most impact idea seeding rate, which will help further evaluate the effectiveness of a prescription methods.