Can a river heal itself? Natural attenuation of metal contamination in river sediment

Sediment sampling of bed sediment from a large river contaminated by mining and smelting was used to determine rates of natural attenuation of metal concentrations. A "natural decay model" was developed from high-resolution temporal data and used to predict when restoration guidelines would be met without restoration and with various degrees of restoration success. The natural decay model estimates that in the most contaminated reaches it will take about 90 years for average concentrations of As, Cd, Cu, Pb, and Zn to fall below "probable effects concentrations" (PEC), i.e. levels above which we expect to see adverse environmental effects. At sites farther downstream, all metals will fall below PEC in <35 ± 8 years. It will take longer to reach "threshold effects concentrations" (TEC), i.e. concentrations at which no effects are expected. But, even in the most contaminated reaches, Cd, Pb, and Zn will reach TEC in <80 ± 57 years, while Cu and As will take ∼200 years. Model simulations with different levels of remediation success show that recovery is highly dependent on source reduction and how far the goal is from the basin background concentration. Furthermore, beneficial effects of restoration may be unexpectedly small: for example a likely decrease of ∼20% in the source concentration would shorten the time to reach the Cu PEC by only 13 years. We argue that conducting analyses like these can provide insight into remediation approaches and ultimately decrease the cost of restoration by identifying the role of natural attenuation in restoration design and implementation.