Oxygen supply and demand as controls of denitrification at the microscale in repacked soil

<p>The controlling factors of biotic denitrification in soil as a source of the greenhouse gas nitrous oxide (N<sub>2</sub>O) and of dinitrogen (N<sub>2</sub>) are still not fully understood due to the challenges in observing processes that co-occur in soil at microscopic scales and the difficulty to measure N<sub>2</sub> fluxes. N<sub>2</sub>O production and reduction depend on the extent of anoxic conditions in soil, which in turn are a function of O<sub>2</sub> supply through diffusion and O<sub>2</sub> demand by soil respiration in the presence of an alternative electron acceptor (e.g. nitrate).</p><p>This study aimed to explore microscopic drivers that control total denitrification, i.e. N<sub>2</sub>O and (N<sub>2</sub>O+N<sub>2</sub>) fluxes. To provoke different levels of oxygen supply and demand, repacked soils from two locations in Germany were incubated in a full factorial design with soil organic matter (1.2 and 4.5 %), aggregate size (2-4 and 4-8mm) and water saturation (70%, 83% and 95% WHC) as factors. The sieved soils were repacked and incubated at constant temperature and moisture and gas emissions (CO<sub>2</sub> and N<sub>2</sub>O) were monitored with gas chromatography. The <sup>15</sup>N tracer application was used to estimate the N<sub>2</sub>O reduction to N<sub>2</sub>. The internal soil structure and air distribution was measured with X-ray computed tomography (X-ray CT).</p><p>The interplay of anaerobic soil volume fraction (ansvf) as an abiotic proxy of oxygen supply and CO<sub>2</sub> emission as a biotic proxy of oxygen demand resulted in 81% and 84% explained variability in N<sub>2</sub>O and (N<sub>2</sub>O+N<sub>2</sub>) emissions, respectively. These high values dropped to 5-30% when only ansvf or CO<sub>2</sub> was considered indicating strong interaction effects. The extent of N<sub>2</sub>O reduction in combination with ansvf and CO<sub>2</sub> even increased the explained variability for N<sub>2</sub>O fluxes to 83%. Average O<sub>2</sub> concentration measured by microsensors was a very poor predictor due to the extreme variability in O<sub>2</sub> at short scales in combination with the small footprint of the micro sensors probing only 0.2% of the entire soil volume. The substitution of predictors by independent, readily available proxies for O<sub>2</sub> supply (diffusivity based on air content) and O<sub>2</sub> demand (SOM) leads to a reduction in predictive power.</p><p>To our knowledge this is the first study analyzing total denitrification in combination with X-ray CT image analysis, which opens up new perspectives to estimate denitrification in soil and also contribute to improving models of N<sub>2</sub>O fluxes and fertiliser loss at all scales and can help to develop mitigation strategies for N<sub>2</sub>O fluxes and improve N use efficiency.</p>