B41J-0183:
Identification and characterization of anthropogenic nitrogen fluxes using stable isotopes and reactive hydrologic modeling

Thursday, 18 December 2014
Matthew T O'connell, University of Virginia Main Campus, Charlottesville, VA, United States, Stephen A Macko, Univ Virginia, Charlottesville, VA, United States and Youtong Fu, Triad Engineering Application, Inc., Ashburn, VA, United States
Abstract:
The Najinhe watershed is a topographically diverse, heavily agricultural watershed in northeastern China that provides opportunities for identification of the impact of land use on nitrogen cycling. In addition to agricultural soil amendments, seasonal variation in atmospheric flow introduces a signal of dry and wet deposition from urban and desert atmospheric N sources. Both agricultural amendments and atmospheric sources are significant sources of reactive N, at estimated annual rates of 450kg/hectare and 30kg/hectare respectively in the nearby North China Plain.

Land use, both historic and current, influences the biological processing of nitrogen in a particular area. Soil conditions, including moisture, texture, and organic content, control the capacity of a parcel for processing reactive nitrogen. Compounds derived from natural and anthropogenic sources exhibit characteristic ratios of stable isotopes of nitrogen and oxygen that serve as tracers of origin as well as integrators of biological processes. Analysis of bulk soils (including both organic and inorganic N) in the system shows δ15N ranging from 1.3 – 8.6 ‰ suggesting varying influence of anthropogenic fertilizers, soil organic nitrogen, and atmospheric sources based on land use.

A distributed hydrologic model coupled with one focusing on reactive transport is able to help determine locations with the highest impact on the dissolved N in this system. Spatial statistical methods are employed to determine the biogeochemical influence of model locations whereas δ18O and δ15N measurements from NO3- and NH4+ in surface water and soil extracts are used to calibrate and validate model predictions based on measured precipitation and streamflow values. Sources are integrated using a Bayesian mixing model to determine likely fate and transport parameters for various N inputs to the watershed.

The application of the coupled hydrologic and transport models to a village scale catchment suggests integration and expansion to larger watersheds on the basin scale. Identification of sensitive parcels on multiple spatial scales can direct targeted land management efforts to mitigate ecological and health effects of reactive N in surface waters.