How do varying nitrogen fertilization rates affect crop yields and riverine N2O emissions? A hybrid modeling study

Headwater streams in agricultural areas constitute significant sources of nitrous oxide (N2O) due to nutrient enrichment; however, their emissions are often overlooked in current environmental impact assessments. This scarcity highlights the importance of developing advanced decision tools to evaluate these contributions and create effective mitigation strategies. Our study establishes the first integrated modeling framework that combines a process-based model SWAT+ with a linear mixed model (LMM) to predict N2O emissions from a headwater agricultural river system in Belgium under diverse climate change and fertilization scenarios. In particular, the calibrated and validated SWAT+ model was used to simulate streamflow, nutrient transport, and crop yields under these scenarios, from which, together with biochemical data collected from sampling campaigns, riverine N2O emissions were predicted via LMM. Our results revealed hydrologically driven patterns in riverine N2O emissions, with peak emissions in winter and spring, driven by precipitations enhancing shallow subsurface flows, carrying leached nutrients from fields to the river, and fueling N2O emissions. These phenomena were intensified under climate change scenarios, especially during combined wetter and hotter winters and springs, which elevated headwater N2O emissions by 40 %. Moreover, when coupling these conditions with a 20 % increase in fertilizer rates, riverine N2O emissions would be boosted by 83 %. These findings underscore the importance of integrating land-surface and river processes, to effectively quantify the feedback loop between river nutrient enrichment and climate change under the influence of agricultural practices, and to support comprehensive mitigation strategies under the warming climate.