Disentangling effects of forest harvest on long-term hydrologic and sediment dynamics, western Cascades, Oregon.

• Increased sediment transport after forest harvest is not attributable to one process. • Both changes in sediment supply and streamflow may contribute to increased transport. • Long-term paired catchment data suggest increased sediment supply predominates. • Suspended sediment and bedload transport recover at different rates after harvest. • Full recovery to pre-harvest sediment transport conditions may require many decades. The magnitude of sediment yield following forest timber harvest is controlled by increases in both sediment supply and streamflow. Since the relation between sediment transport and streamflow typically follows a power law, small increases in streamflow may translate into large increases in sediment transport. Interpreting the geomorphic effects of streamflow increases is confounded by the fact that timber harvest influences both the hydrologic regime and sediment supply of a watershed simultaneously, making it difficult to isolate the streamflow effect alone. Here we report on a novel approach to this problem using long-term data from two paired catchments located in the H.J. Andrews Experimental Forest, Oregon, USA. We use observed streamflow from the treated (clearcut) and control watersheds to reconstruct a natural streamflow time series for the treated watershed, one that represents streamflow response in conditions prior to harvest. We combine this reconstructed natural streamflow time series with observed relations between streamflow and sediment transport to quantify the background sediment yield and disentangle the relative effects of changes in hydrology and sediment supply. Results indicate that while increases in streamflow can account for modest increases in sediment transport, this is dwarfed by the increased supply of sediment that accompanies most timber harvest. These results have broad relevance to forest timber harvest and fuel management practices worldwide and can be used to constrain or set bounds on likely effects of more modest (i.e., thinning) techniques. [ABSTRACT FROM AUTHOR]