Nitrate removal, denitrification and nitrous oxide production in the riparian zone of an ephemeral stream

Riparian zones are important features of the landscape that can buffer waterways from non-point sources of nitrogen pollution. Studies of perennial streams have identified denitrification as one of the dominant mechanisms by which this can occur. This study aimed to assess nitrate removal within the riparian zone of an ephemeral stream and characterise the processes responsible, particularly denitrification, using both in-situ and laboratory techniques. To quantify rates of groundwater nitrate removal and denitrification in-situ, nitrate was added to two separate injection–capture well networks in a perched riparian aquifer of a low order ephemeral stream in South East Queensland, Australia. Both networks also received bromide as a conservative tracer and one received acetylene to inhibit the last step of denitrification. An average of 77 ± 2% and 98 ± 1% of the added nitrate was removed within a distance of 40 cm from the injection wells (networks with acetylene and without, respectively). Based on rates of N2O production in the network with added acetylene, denitrification was not a major mechanism of nitrate loss, accounting for only 3% of removal. Reduction of nitrate to ammonium was also not a major pathway in either network, contributing <4%. Relatively high concentrations of oxygen in the aquifer following recent filling by stream water may have reduced the importance of these two anaerobic pathways. Alternatively, denitrification may have been underestimated using the in-situ acetylene block technique. In the laboratory, soils taken from two depths at each well network were incubated with four nitrate-N treatments (ranging from ambient concentration to an addition of 15 mg N l−1), with and without added acetylene. Potential rates of denitrification, N2O production and N2O:N2 ratios increased with nitrate additions, particularly in shallow soils. Potential rates of denitrification observed in the laboratory were equivalent in magnitude to nitrate removal measured in the field (mean 0.26 ± 0.12 mg N kg of dry soil−1 d−1), but were two orders of magnitude greater than denitrification measured in the field with added acetylene. The relative importance of assimilatory vs. dissimilatory processes of nitrate removal depends on environmental conditions in the aquifer, particularly hydrology and its effects on dissolved oxygen concentrations. Depending on seasonal conditions, aquifers of ephemeral streams like the study site are likely to fluctuate between oxic and anoxic conditions; nevertheless they may still function as effective buffers. While denitrification to N2 is a desirable outcome from a management perspective, assimilation into biomass can provide a rapid sink for nitrate, thus helping to reduce short-term delivery of nitrate downstream. Longer-term studies are needed to determine the overall effectiveness of riparian buffers associated with ephemeral streams in mitigating nitrate loads reaching downstream ecosystems.