Assessing nitrous oxide and nitrate leaching mitigation potential in US corn crop systems using the DNDC model

Abstract

Nitrous oxide (N2O, a potent greenhouse gas; GHG) is emitted at relatively high rates from corn-based agricultural systems. N2O mitigation strategies, in addition to reducing GHG emissions, ideally would not increase other harmful pollutants (such as leached nitrate, NO3, or volatilized ammonia, NH3) or decrease crop yield. We used the Denitrification-Decomposition model (DNDC) to simulate an array of single and combined interventions to corn management across broad range of physical conditions (climate and soil) in the Midwest US. We assumed a typical crop management baseline of continuous corn or corn-soy with conventional tillage and urea-ammonium nitrate fertilizer broadcast to the soil surface. Interventions included fertilizer nitrogen (N) form, controlled-release N, addition of nitrification inhibitors, N rate, split sidedress N applications, sub-soil N placement, and reduced tillage. Single-factor interventions which reduced N2O in all locales included a change to urea fertilizer, nitrification inhibitors, reductions to N rate, and use of N injection: urea fertilizer and nitrification inhibitors both reduced N2O emissions on average (30 and 9%, respectively) while simultaneously reducing NO3 leaching and mostly neutral effects to yield; N rate reductions reduced N2O (11%) but had modest negative effects to yield. Other single-factor interventions increase N2O emissions or N leaching on average but could have beneficial effects under some conditions. Combined interventions frequently include (>50% of the time) urea, nitrification inhibitors, and reduced N. Interventions that, even when combined with other interventions, do not reduce N2O emissions and N leaching include anhydrous ammonia fertilizer, controlled release, and injection. Controlled-release fertilizer results were contrary to those reported in most field studies indicating that DNDC's simulation of linear N release over time may be too simplistic to replicate field conditions.