Recognizing both denitrification and nitrogen consumption improves performance of stream diel N2 flux models

There is keen interest to enhance denitrification within intervening aquatic habitats between agricultural areas and downstream aquatic ecosystems to reduce nitrogen (N) loading impacts to receiving ecosystems. We conducted a series of measurements to examine whole system in situ diel denitrification estimates in experimental ditch and stream environments using a Bayesian one‐station diel N2 flux model. Model estimates revealed complex patterns that indicate fluxes may be controlled by the balance of both N2 production via denitrification and consumption driven by physical or biological processes associated with strong diel patterns in environmental conditions. We investigated potential improvements in model fits to observed data associated with the addition of a N2 consumption term to represent biological (N2 fixation) or physical (bubble formation and N2 scavenging) mechanisms associated with daytime photosynthesis. We also expanded the current one‐station diel flux model to a two‐station model to estimate denitrification in discrete reaches. Our modified diel N2 flux models improved model fit significantly across three metrics (Nash–Sutcliffe efficiency, root mean square ratio, and percent bias) increasing their utility in shallow, open canopy, lotic systems. While more studies are needed to understand specific mechanisms associated with N2 consumption processes in small agricultural drainages as well as environmental conditions affecting their relative importance, these results improve estimates of N2 flux where dynamic conditions and heterogeneity of habitats create severe diel patterns in factors controlling dissolved gas concentrations and prohibit accurate estimates of N2 flux using existing models.