Experiences Developing and Using a Regional Corn Nitrogen Response Trial Database.

Nitrogen rate recommendations for corn in the Midwestern United States have historically been based on N response data. Scientists across states took different approaches to making such recommendations, however, and in many cases neither the data nor the rationale for those recommendations can be reconstructed. Based on these and other issues, such as the need for updating response data, an effort was undertaken to develop a regional Midwest approach to corn N rate recommendations, with the approach to be based directly on results from recent N response field trials. This required sharing of response data from all participating states and development of an analytical and recommendation approach for rate calculation. Underlying principles include that datasets of response trials represent a population of potential corn yield responses to N, and analysis of that data for maximum economic net return to N (MRTN) directly provides rate guidelines. Key data accumulated were site metadata and regression model parameters fit to yields from each response trial. Datasets were developed and maintained in each participating state, and shared for accumulation in one database used for web-based calculation of N recommendations (Corn Nitrogen Rate Calculator). Metadata useful in differentiating N recommendations are state, crop rotation, within-state geographic region, and soil potential. Data accumulation and this regional approach are currently implemented in seven states. An important aspect of the approach is the ease and speed with which N response trial datasets can be updated, with the data incorporated in the on-line calculator where they have an immediate influence on N rate guidelines. Development of the database was not intended as a research tool or to make N response data public. Rather, the database and calculator were developed specifically to provide a consistent approach to corn N rate recommendations based on documented and current N response data.