A comparison of different approaches for measuring denitrification rates in a nitrate removing bioreactor

Denitrifying woodchip bioreactors (denitrification beds) are increasingly used to remove excess nitrate (NO(3)(-)) from point-sources such as wastewater effluent or subsurface drains from agricultural fields. NO(3)(-) removal in these beds is assumed to be due to microbial denitrification but direct measurements of denitrification are lacking. Our objective was to test four different approaches for measuring denitrification rates in a denitrification bed that treated effluent discharged from a glasshouse. We compared these denitrification rates with the rate of NO(3)(-) removal along the length of the bed. The NO(3)(-) removal rate was 8.73 +/- 1.45 g m(-3) d(-1). In vitro acetylene inhibition assays resulted in highly variable denitrification rates (DR(A1)) along the length of the bed and generally 5 times greater than the measured (NO(3)(-)-N removal rate. An in situ push-pull test, where enriched 15N-NO(3)(-) was injected into 2 locations along the bed, resulted in rates of 23.2 +/- 1.43 g N m(-3) d(-1) and 8.06 +/- 1.64 g N m(-3) d(-1). The denitrification rate calculated from the increase in dissolved N(2) and N(2)O concentrations (DR(N2)) along the length of the denitrification bed was 6.7 +/- 1.61 g N m(-3) d(-1). Lastly, denitrification rates calculated from changes in natural abundance measurements of delta(15)N-N(2) and delta(15)N-NO(3)(-) along the length of the bed yielded a denitrification rate (DR(NA)) of 6.39 +/- 2.07 g m(-3) d(-1). Based on our experience, DA(N2) measurements were the easiest and most efficient approach for determining the denitrification rate and N(2)O production of a denitrification bed. However, the other approaches were useful for testing other hypotheses such as factors limiting denitrification or may be applied to determine denitrification rates in environmental systems different to our study site. DR(N2) does require very careful sampling to avoid atmospheric N(2) contamination but could be used to rapidly determine denitrification rates in a variety of aquatic systems with high N(2) production and even water flows. These measurements demonstrated that the majority of NO(3)(-) removal was due to heterotrophic denitrification. (C) 2011 Elsevier Ltd. All rights reserved.