Characterization of the Solute Dynamics Signature of an Agricultural Coastal Plain Stream, North Carolina, US

The context of this research is the quantification of water quality benefits of stream restoration. This common mitigation practice has not received widespread acceptance, in part because of the lack of concordant evidence of its benefits, and magnitude thereof, on stream water quality. The reasons might be that the benefits are too small to be measurable, and/or that the uncertainties on the quantification methods are just too large to detect benefits. The only variable on which we have control are the quantification methods and their uncertainties. To document the water quality benefits of stream restoration, North Carolina Department of Transportation, North Carolina Forest Service and North Carolina State University have collaborated to conduct a pilot study where new monitoring techniques have been applied to minimize quantification uncertainties. We chose to take nutrients loads as the indicator of the bulk effect of restoration on water quality. In this work, it is necessary to report the characterization of the solute dynamics and calculate indicators of water quality benefits. For this, flow and water quality were monitored on a 'continuous' basis during the 16-month pre-restoration period in the Claridge Canal, a coastal plain stream to be restored in Goldsboro, North Carolina.;The results indicated that it was possible to calculate the concentrations of a suite of water quality parameters every 15-min combining absorbance spectra from in situ ultravioletvisual (UV-Vis) spectrophotometers, with water quality rating curves developed from Partial Least Squares Regressions (PLSR). We were able to construct robust water quality rating curves to measure nitrate (NO3-N), Total Dissolved Nitrogen (TDN), Total Kjeldahl Nitrogen (TKN), Dissolved Organic Carbon (DOC), and Total Phosphorus (TP) concentrations, using the absorbance data as index data. The method did not work well for ammonium (NH4-N) and phosphate (PO4-P), although this might have come from less than ideal calibration point pools. It was found that turbidity-based rating curves were better suited to calculate Total Suspended Solid (TSS) concentrations and loads.;Then, this research quantified the uncertainties on annual cumulative loads associated with the data generated by UV-Vis spectrophotometers. For this, we used bootstrap, random and stratified sample subsets from the original dataset and applied PLSR on these subsets to estimate the uncertainties on annual loads. The bootstrap resampling results demonstrated that uncertainties on annual loads associated with the application of UV-Vis spectrophotometers were much lower than those typically obtained from infrequent sampling, and could be within +/-5% measuring error for nitrate, TDN, TKN, DOC and TSS; and approximately +/-10% for TP. It was found that to obtain robust rating curves that would lead to precise and unbiased cumulative annual loads, it is desirable to stratify calibration concentrations.;At last, the unique flow and concentration datasets were used to characterize nutrients and material dynamics in the events and seasonal scales in this coastal plain watershed. In this research, a unique pattern of concentration troughs and peaks was found during flow peaks. Unlike most upland watersheds, the summer dynamics have highlighted the unique role of the near-stream zone, which in the flat coastal plain watersheds tend to be drier than the rest of the soil profile. We were able to associate events scales with seasonal variations. As a result, the role of stream denitrification and the release of DOC due to diagenetic processes were shown during warmer months.