SYSTEMS-LEVEL RESPONSE OF CROP NITROGEN REMOVAL TO NITROGEN INPUTS IN U.S. AGRICULTURE

The response of nitrogen (N) removal by crops to an increase in fertilization strongly determines the profitability and sustainability of agricultural systems and informs nutrient management decisions at the producer level. These response functions are analyzed for agronomic and economic optima achieved at field scales for evaluating production, economic, and environmental goals. However, such assessments are lacking for entire regional agroecosystems to allow understanding of the response of N removal collectively across all crops grown during the year (N_removal) to total N fertilization (i.e., all manageable N sources, N_in) historically. Here, we address this knowledge gap by leveraging a large-scale N budget and statistical techniques to characterize space-time variability and trends in historical (1987–2016) county-level N_removal, N_in, and nitrogen use efficiency (NUE) across the conterminous U.S. (CONUS). We intend to evaluate crop belt-specific characteristics of diminished returns in N_removal to N_in response and change over time. N_in, N_removal, and NUE were subject to drastic spatial variation in long-term mean values, interannual variability, and long-term change, which were quantified and mapped to understand their spatiotemporal distributions. Pooled across all counties and years, N_removal shows diminished returns when N_in reached 163 kg ha^-1 (N_{in, bp}). Upon quantifying and analyzing year-specific diminished returns, we found that N_{in, bp} has increased during 1987–2016, and so has the NUE achieved prior to attaining diminished returns. The proportion of counties (6%–22%) where N_in exceeds N_{in, bp} also increased, and counties that repeatedly demonstrated such exceedance during 1987–2016 were identified. Values of N_in,bp are specific to crop belts within the U.S., the majority of which also show increased N_{in, bp} over time. Specifically, barley, beans, and sugarbeets (198 kg ha^-1), and alfalfa and barley (190 kg ha^-1) belts showed notably greater N_in,bp relative to the national mean (163 kg ha^-1), while N_in,bp for corn grain and soy belts was similar to the national mean. Overall, these findings represent a comprehensive assessment of how systems-level N_removal across U.S. agriculture has historically responded to change in N_in, a prerequisite for guiding mitigation and adaptation policy and efforts.