Spatiotemporal dynamics and controls of greenhouse gas emissions in agricultural ditches

Agricultural ditches have been identified as emission hotspots of the three main atmospheric greenhouse gases (GHGs) i.e., carbon dioxide (CO₂), methane (CH₄) and nitrous oxide (N₂O). However, due to lack of representative GHG emission data from these small waterbodies, their contribution to the total GHG emissions from the agricultural sector are reported as highly uncertain. In this study the aim was to 1) quantify the magnitude of GHG emissions and understand their spatiotemporal dynamics in agricultural ditches, and 2) identify gas specific controls that regulate those dynamics.

In our two-year field study, dissolved concentrations and fluxes of CO₂, CH₄ and N₂O were measured in ditches, as well as concentrations in groundwater, of a clay soil dominated agricultural catchment located in central Sweden. Sampling was carried out biweekly during the growing season (April–November). Additional sampling for water chemistry and runoff was made to aid assessment of GHG controls. Local floating chamber-based GHG fluxes were further scaled for the entire ditch network for representative catchment scale emission estimates.

The results showed that both dissolved GHG concentrations and their respective fluxes were highly variable in space and time. In general, higher GHG concentrations and fluxes were observed during the summer and after rain events, but patterns were gas and site-specific. Across all sites, N₂O concentrations were positively related to dissolved inorganic nitrogen (DIN), and CO₂ to total organic carbon (TOC), whereas patterns for CH₄ were more unpredictable. Groundwater data also revealed highly variable gas-specific patterns over time. While groundwater N₂O concentrations was on average close to concentrations observed in the ditch, CO₂ was 5-fold and CH₄ 1000-fold higher in groundwater indicating a gas-variable terrestrial source contribution to observed ditch emissions. Our study identified ditches as significant but also highly variable sources of CO₂, CH₄ and N₂O emissions to the atmosphere. This further highlights the need for appropriate sampling and scaling designs that can capture these high spatiotemporal dynamics to provide representative catchment-based emission estimates.