Co-occurrence of in-stream nitrogen fixation and denitrification across a nitrogen gradient in a western U.S. watershed

It is frequently assumed that nitrogen (N2) fixation and denitrification do not co-occur in streams because each process should be favored under different concentrations of dissolved inorganic nitrogen (DIN), and therefore these processes are rarely quantified together. We asked if these processes could co-exist by conducting a spatial survey of N2 fixation using acetylene reduction and denitrification using acetylene block [with and without amendments of carbon (C) as glucose and nitrogen (N) as nitrate]. Rates were measured on rocks and sediment in 8 southeastern Idaho streams encompassing a DIN gradient of 26–615 µg L−1. Sampling at each site was repeated in summer 2015 and 2016. We found that both denitrification and N2 fixation occurred across the gradient of DIN concentrations, with N2 fixation occurring primarily on rocks and denitrification occurring in sediment. N2 fixation rates on rocks significantly decreased 100× across the DIN gradient in 1 year of the study, and amended (with N and C) denitrification rates increased 10× across the DIN gradient in both years. Multiple linear regression and partial least squares models with environmental characteristics measured at the scale of entire stream reaches showed that C and phosphorus were positive predictors of amended and unamended denitrification rates, but no significant model could explain N2 fixation rates across all streams and years. This, coupled with the observation that detectable rates of N2 fixation occurred primarily on rocks and denitrification occurred primarily on sediment, suggests that microhabitat scale factors may better predict the co-occurrence of these processes within stream reaches. Overlooking the potential co-occurrence of N2 fixation and denitrification in stream ecosystems will impede understanding by oversimplifying the contribution of each process to the N cycle.