Nitrous oxide (N₂O) transfer velocity and the effect of Ammonium on N2O fluxes from an agricultural drain

Abstract

Indirect nitrous oxide (N₂O) emissions from rivers and drains are poorly quantified and the uncertainity surrounding the emission factor for dissolved N₂O in rivers (EF5-r) is high. Scaling the exchange of N₂O across the water – air interface is important in order to estimate the indirect N₂O emissions from rivers and drains.Therefore, this study was designed to measure the drain water N₂O emission transfer velocity (kN₂O) and the effect of ammonium (NH₄⁺) on drain water N₂O yield. The field experiment released ¹⁵N-NH₄⁺, with Bromide (Br⁻) as a conservative tracer, RhodamineWT as a visual tracer, and propane (C₃H₈) as a conservative tracer gas, into a drain. Visual tracer and conservative tracers allowed the drain water velocity and dilution factor to be determined, respectively. The C₃H₈ and N₂O gases were collected using headspace equilibrium and floating chamber methods. The rate of C₃H₈ escape from water to air was used to measure the N₂O transfer velocity (kN₂O). The gas transfer velocity model which as explained by O’Connor and Dobbins (1958), equation 2.9 as interpreted by Wilcock (1982) was also used to measure kN₂O. The ¹⁵N₂O fluxes allowed the NH₄⁺ contribution to the N₂O flux to be determined. The ¹⁵N enrichment in biofilms and aquatic plants also allowed nitrogen assimilation to be evaluated.

The measured kN₂O using headspace equilibrium and the kN₂O estimated from Wilcock (1982) were 7.49 ± 0.72 m day-1 and 8.65 ± 1.23 m day-1, respectively. To measure the hydro physical variations in kN₂O, the current study data and the data from other New Zealand studies were evaluated. The results showed that for shallow drains, water depth < 1 m, the value of kN₂O increased with decreasing water depth.

There was an inverse relationship between ¹⁵N-NH₄⁺-N enrichment and the ¹⁵N enrichment of the N₂O evolved. However, there was no relationship between ¹⁵N-NO₃⁻ enrichment with ¹⁵N-N₂O enrichment indicating N₂O was produced due to nitrification of added NH₄⁺-N. Following ¹⁵N tracer addition, biotic components of the drain ecosystem were highly enriched but returned to near background levels after 10 days, demonstrating that NH₄⁺-N assimilation and mineralization occurred.

The calculated EF5-r values were significantly lower than the IPCC default EF5-r value; 0.0025 kg (2.5 g) N₂O-N per kg NO₃⁻-N. The measured EF5-r values, derived using the surface C₃H₈ flux measurements and the headspace C₃H₈ equlibrium concentrations, were 7.66E-6 (7.66 mg) and 8.20E-6 (8.20 mg) kg N₂O-N per kg NO₃⁻-N, respectively. However, the EF5-r values derived using floating chambers were significantly lower (3.92E-11 kg; (0.0004 mg) N₂O-N per kg NO₃⁻-N) and showed chamber methodology as an inefficient technique for gas flux measurements on flowing water surfaces. In contrast, the kN₂O measured using the Wilcock (1982) model, that used drain water speed and depth produced higher EF5-r; 9.48 E-6 kg (9.48 mg) N₂O-N per kg NO₃⁻-N. Thus, this study demonstrated; (1) the tracer gas addition technique is a relaible and accurate method for N₂O flux determinations from drains, (2) in shallow drains ( water depth < 1 m) the kN₂O value was inversely correlated with the water depths, (3) the added ¹⁵N tracer permitted the role of NH₄⁺ in N₂O production to be assessed and (4) the IPCC EF5-r default value overestimated N₂O emissions from agricultural drains.