Catchment controls of denitrification and nitrous oxide production rates in headwater remediated agricultural streams.

Heavily modified headwater streams and open ditches carry high nitrogen loads from agricultural soils that sustain eutrophication and poor water quality in downstream aquatic ecosystems. To remediate agricultural streams and reduce the export of nitrate (NO), phosphorus and suspended sediments, two-stage ditches with constructed floodplains can be implemented as countermeasures. By extending hydrological connectivity between the stream channel and riparian corridor within constructed floodplains, these remediated ditches enhance the removal of NO via the microbial denitrification process. Ten remediated ditches were paired with upstream trapezoidal ditches in Sweden across different soils and land uses to measure the capacity for denitrification and nitrous oxide (NO) production and yields under denitrifying conditions in stream and floodplain sediments. To examine the controls for denitrification, water quality was monitored monthly and flow discharge continuously along reaches. Floodplain sediments accounted for 33 % of total denitrification capacity of remediated ditches, primarily controlled by inundation and stream NO concentrations. Despite reductions in flow-weighted NO concentrations along reaches, NO removal in remediated ditches via denitrification can be masked by inputs of NO-rich groundwaters, typical of intensively managed agricultural landscapes. Although NO production rates were 50 % lower in floodplains compared to the stream, remediated ditches emitted more NO than conventional trapezoidal ditches. Higher denitrification rates and reductions of NO proportions were predicted by catchments with loamy soils, higher proportions of agricultural land use and lower floodplain elevations. For realizing enhanced NO removal from floodplains and avoiding increased NO emissions, soil type, land use and the design of floodplains need to be considered when implementing remediated streams. Further, we stress the need for assessing the impact of stream remediation in the context of broader catchment processes, to determine the overall potential for improving water quality.