| Abstract: | Estuaries have long been thought to be effective traps for river-borne contaminants; however, accurately predicting removal efficiencies remains problematic. In the Quinnipiac River, CT, marsh system, patterns of metal accumulation (Ag, Cd, Cu, Pb) in sediments were used to estimate heavy metal removal efficiencies. Linear and multivariate regression models were fit to sediment metal concentrations to estimate removal curves and surfaces, which avoided the large temporal variability commonly encountered when using direct water column measurements. Second order regressions normalized to the shortest distance to the river channel were found to have the best fit (r>0.67) with the lowest standard error (<32%). The heavy metal removal efficiencies estimated by comparing total metal accumulation in marsh sediments with riverine flux were Ag=21%, Cd=6%, Cu=14%, and Pb=17%, which are comparable to the maximum removal efficiency of riverine suspended particulate matter (28%). |
