What is the most relevant soil structure parameter to describe field-measured N2O emissions?

• Field N 2 O flux can be both explained by various bulk soil and structure properties. • Thresholds for structure properties are proposed beyond which N 2 O flux can be high. • Gas diffusivity is a relevant soil structure parameter to describe field N 2 O flux. Nitrous oxide (N 2 O) emissions from soil are partly controlled by aeration and gas transfer in soil, and thus by soil structure. The intensity of N 2 O emissions is usually expressed according to the water filled pore space (WFPS), calculated using the soil bulk density. These factors, even if they describe the soil structure and the water proportion in the porous network, do not inform about porous network characteristics among scales and their connectivity. The aim of this work was therefore to determine (1) to what extent the soil structure of an agricultural soil controlled N 2 O emissions during a snap-shot campaign and (2) which metric of gas transfer or soil structure was the most appropriate to describe the N 2 O emission variability at field scale. N 2 O emissions were measured with a mobile chamber on a maize crop after fertilization with several soil management practices resulting in four soil states (strip-till versus tillage, compacted soil versus uncompacted) with contrasting soil structure. Soil cylinders and bulk soil were sampled from 24 plots exhibiting a strong gradient in N 2 O emissions. Classical soil physical and chemical properties were measured, including soil bulk density and water filled pore space. Soil structure also was characterized quantitatively by X-ray tomography at meso and macro scales, and indirectly by gas transfer parameters. Clear differences were observed between low and high emission plots in terms of soil structure, soil temperature and nitrate concentration. However, soil structure appeared more strongly connected to N 2 O emissions, and some thresholds on soil structural indicators were relevant to disentangle high and low N 2 O fluxes. Some structural indicators at both scales (e.g. porosity, surface density) and gas transfer parameters (relative gas diffusivity, air permeability) were good descriptors of the observed N 2 O fluxes. Nevertheless, the gas transfer parameters can be easily measured over a short period of time, whereas the soil structure indicators determined from 3D images require an acquisition and a processing phase that can be time consuming. A good compromise to evaluate the field N 2 O flux potential from an easy measure would be to evaluate the relative gas diffusivity, which directly controls the diffusion of oxygen in soil and thereby the microbial processes of N 2 O production.