Database-Driven N Decision Support: II. Using Rainfall and Feedback Information about Corn N Status and Yield Response.

Feedback provided to corn (Zea Mays L.) growers from information collected about the N status of many fields across numerous years can be used to refine nitrogen (N) management. We propose a three-element decision support system to make adjustments in future N management using databases developed in Iowa, Ohio, Illinois, and Michigan. The data collected included post-season corn N status (corn stalk nitrate test guided by aerial imagery), farmer management, observations of yield response to N, and site-specific amounts of rainfall. The Iowa database includes more than 3500 corn fields and 300 on-farm replicated strip trials conducted from 2006 through 2013, and the Ohio, Illinois, and Michigan database includes 617 corn fields from 2008 to 2013 and 144 on-farm replicated strip trials conducted from 2010 through 2013. The decision support system is based on (1) benchmarking the N status from individual farmers’ fields to the aggregate average N status of many fields across a state, watershed or grower group, (2) estimating the “Agronomic Risk” of “deficient” and the “Environmental Risk” of “excessive” corn N status for management practices with different forms and timing of N application, N rate applied, and early season rainfall; and, (3) estimating posterior predictive probabilities of an economic yield loss from reduced N applications or an economic yield response from increased N applications for different management categories of previous crop, forms and timing of N application, soil organic matter content, or early season rainfall. The benchmarking references were the 25, 50 and 75^th percentiles for distributions of N rates required to reach optimal corn N status for specific areas or watersheds. The proposed feedback-based database-driven decision support system can be also used in nutrient management planning and environmental conservation.