A conceptual framework for judging the precision agriculture hypothesis with regard to site-specific nitrogen application

In this study we derive a test that aims to show that the economic optimum nitrogen applications in site-specific management zones are different. In addition, a confidence interval is derived for the difference between optima, and here we show that separate confidence intervals for either optimum are not as good as the former for judging the precision agriculture hypothesis. Both the test and the confidence interval for the difference are based on modeling the returns above N fertilizer cost by quadratic functions of the nitrogen rate. The data are from an N-rate trial at a heterogeneous field in Thalhausen, Germany. This trial could barely support the economic benefit of precision agriculture. A statistical proof that high- and low-yielding zones should be discriminated against just failed to meet the standard 5% level. We also discuss what design of nitrogen fertilizer levels should be used for the test and the confidence intervals.