Natural abundance nitrate (NO3 (-)) isotopes represent a powerful tool for assessing denitrification, yet the scale and context dependence of relationships between isotopes and denitrification have received little attention, especially in surface soils. We measured the NO3 (-) isotope compositions in soil extractions and lysimeter water from a semi-arid meadow and lawn during snowmelt, along with the denitrification potential, bulk O2, and a proxy for anaerobic microsites. Denitrification potential varied by three orders of magnitude and the slope of δ(18)O/δ(15)N in soil-extracted NO3 (-) from all samples measured 1.04 ± 0.12 (R (2) = 0.64, p < 0.0001), consistent with fractionation from denitrification. However, δ(15)N of extracted NO3 (-) was often lower than bulk soil δ(15)N (by up to 24 ‰), indicative of fractionation during nitrification that was partially overprinted by denitrification. Mean NO3 (-) isotopes in lysimeter water differed from soil extractions by up to 19 ‰ in δ(18)O and 12 ‰ in δ(15)N, indicating distinct biogeochemical processing in relatively mobile water versus soil microsites. This implies that NO3 (-) isotopes in streams, which are predominantly fed by mobile water, do not fully reflect terrestrial soil N cycling. Relationships between potential denitrification and δ(15)N of extracted NO3 (-) showed a strong threshold effect culminating in a null relationship at high denitrification rates. Our observations of (1) competing fractionation from nitrification and denitrification in redox-heterogeneous surface soils, (2) large NO3 (-) isotopic differences between relatively immobile and mobile water pools, (3) and the spatial dependence of δ(18)O/δ(15)N relationships suggest caution in using NO3 (-) isotopes to infer site or watershed-scale patterns in denitrification.
Keywords: Isotope mass balance model; Mobile water; Nitrification; Redox; Snowmelt.