Potential linkage between sediment oxygen demand and pore water chemistry in weir-impounded rivers

Highlights: • Seasonal differences were found in SOD with the higher values in summer vs. autumn. • Linkage between SOD and pore water chemistry was more pronounced in summer. • Organic matter decomposition and nitrification contributed to SOD in each season. • Ammonia and fluorescent DOM in pore water explained ~ 59% of SOD variations. • Applying EEM-PARAFAC improved the prediction of SOD using pore water chemistry. Due to recent weir construction on four major rivers in South Korea, sediment has accumulated in the river bottom near the weirs, which has in turn raised concerns over the quality of overlying water. In this study, the seasonal and spatial variations of sediment oxygen demand (SOD) and the influencing factors were explored using pore water chemistry for the weir-impounded rivers. Muddy and sandy sediment samples were taken from 24 different sites along the four major rivers in summer and autumn, 2016. The SOD was measured in a laboratory based on 10-hour incubation at in situ temperature. The measured pore water chemistry included the concentrations of dissolved organic carbon (DOC), total dissolved nitrogen (TDN), inorganic nitrogen (NH₃-N, NO₃-N, NO₂-N), and phosphate phosphorous (PO₄-P), and the optical properties from UV absorption spectra and fluorescence excitation-emission matrixes coupled with parallel factor analysis (EEM-PARAFAC). Significant differences in SOD values between muddy and sandy sediments were found only in summer (p = 0.047). The higher SOD in summer versus autumn (p = 0.015) was attributed to seasonal temperature differences. The higher NH₃-N and the lower NO₃-N of the pore water samples in summer versus autumn suggested that organic nitrogen decomposition via an ammonification and nitrification process could operate as an important factor for the SOD variations in summer and autumn, respectively. Principal component analysis revealed the mutual contributions of nitrogen-associated processes and the organic composition in pore water to increasing SOD levels. NH₃-N in sediment pore water alone could be a good predictor for SOD. However, multiple regression analysis using NH₃-N, fluorescence index and terrestrial humic-like components improved the estimation capability for SOD variations.