With a mean annual discharge exceeding 17,000 m3 sec-1, and nitrate (NO3-N) concentrations up to 3 grams m-3, the Lower Mississippi River (LMR) is a superhighway for nitrate transport to the sea, and a driver of coastal eutrophication. This problem could be lessened by reducing nitrate inputs to the river, and/or enhancing nitrate sinks within the river system. Because of its restricted light environment, there is little opportunity for nitrate removal by primary production in the main channel. Conversely, in the photic zone of connected backwater lakes, algal production, supported by assimilatory nitrate reduction, can be very high; while below the photic zone, conditions favor anaerobic nitrate respiration, i.e. denitrification. We applied a multi-pronged mass-balance approach to assess the potential of LMR backwater systems to modify nitrate fluxes. First, we measured temporal patterns of algal biomass and production, and nitrate fluxes, in the inflow, water column, and outflow of several oxbow lakes and a side channel. Second, in one lake we estimated denitrification rates directly by accumulation of hypolimnetic molecular nitrogen, using membrane inlet mass spectrometry. From these data, we developed a simulation model for nitrate fluxes between the river and backwaters as functions of discharge, nitrate inputs, and hydraulic residence time.
Results/Conclusions
Nitrate inflowing from the river was sequestered or removed rapidly from backwaters by uptake into the food web, and more so by hypolimnetic denitrification (N2-N accumulation = 2.4 µmol l-1 d-1). Model results indicate nitrate removal capacity varies on an annual basis as a function of river discharge. Most often, backwaters are a nitrate sink, with a removal rate of up to 22 tonnes ha-1 y-1. However, in years of very high discharge the same systems may be sources of nitrate to the river channel, a likely consequence of terrestrial inputs to backwaters from the adjacent agricultural landscape. Hence, the role of backwater habitats as sources or sinks of nitrate to the main channel appears to vary with the timing and magnitude of the flood regime, and the intersection of hydrologic connection and land use within the levees.