THE ROLE OF SEDIMENT SUPPLY IN CHANNEL INSTABILITY AND STREAM RESTORATION

Today, stream channel erosion and instability are widespread in much of the U.S.; this instability is manifested in bed and bank erosion. Using examples from two distinctly different watersheds in southwest Ohio we describe the interaction between sediment supply and channel instability and examine the implications of system-wide instability for stream restoration. The instability observed in both watersheds is a result of excess sediment transport capacity relative to the amount of sediment available for the stream to transport. Channels have narrowed by a factor of 3 to 4 between 1930s and 2005 and channel beds have incised into glacial sediments. Lateral migration of these channels decreased after 1960, signaling the onset of widespread channel incision. This recent channel instability is largely a consequence of reduced sediment supply over several past decades. Soil loss rates in 2005 were 4 to 20 times lower than in the 1920s as a result of elimination of cropping on marginal lands and improved soil conservation methods. The construction of impoundments has further reduced sediment delivery downstream, and there is evidence that urbanization and climate change have exacerbated this imbalance by increasing sediment transport capacity including peak flows. Thus the decrease in sediment supply occurred during a time when peak stream flows have likely been increasing, resulting in an imbalance between sediment supply and sediment transporting power in the stream systems. Because sediment supply and transport are governed by watershed-scale conditions, channel instability will persist at these watersheds for some time to come. Reach-scale restoration will neither improve water quality nor prevent continued system-wide channel instability. Stream restoration efforts should recognize these systemic factors. Achieving a dynamically stable channel with no net erosion under conditions of declining sediment supply is likely difficult to impossible. We advocate an approach that uses setbacks/buffers that allow streams to develop new, dynamic floodplains with which to reestablish connections while selectively protecting significant sites with stabilization measures.