Information conveyed by management zones and optimal soybean seeding rate

Abstract

The application of variable rate seeding (VRS) enables soybean [Glycine max (L.) Merr.] growers to allocate seed more efficiently on heterogeneous farmland and thus potentially increases profit by reducing seed cost and/or increasing yield. Successful implementation of VRS depends on the appropriate division of management zones where producers apply different seeding rates. We construct an economic conceptual model for evaluating the information conveyed by management zones. The model implies that distinct management zones should differ by marginal seed productivity when evaluated at the same seeding rate. Based on this model and a dataset collected from 10 on-farm trials in Ohio and Michigan in 2017 and 2018, we implement a quadratic model regression analysis describing yield response to seeding rate. Results show that management zones in four out of 10 trials do not display statistically significant differences in marginal seed productivity. The small differences across management zones could result from the soybean plant's ability to compensate for yield due to fewer plants by increasing branches per plant and/or from inappropriately identified management zones. The estimated profit gains from applying VRS across the 10 trials vary from $0.20 to $30.60 ha⁻¹. We also find that increasing seeding rate and mean yield decreases yield variability and yield skewness. Overall, the difference in marginal seed productivity is crucial for the appropriate determination of VRS management zones.