Restoring In-Stream Legacies of Log Driving and Splash Damming: a Crucial Step Forward

Log driving and splash damming were among the earliest reported management disturbances in rivers of the Pacific Northwest. A splash dam would span the width of a river, creating an upstream reservoir in which logs were collected until the spillway was opened to release a large flood. While effective at moving logs, splash damming dramatically altered streams, with historical photographs of splashed streams showing long stretches scoured to bedrock and little habitat complexity. Log drives and splash dams were routinely used from the 1880s through the 1950s in Oregon to transport logs to downstream mills before extensive road networks were constructed. Recent literature implicates log driving and splash damming as key culprits in declines of salmon population abundances and suggests that habitats in affected streams have not recovered. Despite this, few formal studies have quantified whether log drives and splash dams left an environmental legacy and none of these studies examined effects over a large area. Building on previous efforts in the Pacific Northwest, we mapped and created a database of 213 log drives and 232 splash dams in western Oregon, primarily by searching archived historical documents and maps. We also conducted field searches that located remnants of splash dams, suggesting that their effects on streams may persist. Using data available from regional monitoring programs, we compared habitat conditions between splashed and unsplashed areas to see if an environmental legacy of splash damming could still be detected. We found that splashed reaches generally had more exposed bedrock, fewer deep pools, and fewer pieces of key large wood than unsplashed reaches. Our findings indicate the potential for widespread, persistent habitat degradation from splash damming and so have relevance for efforts to monitor, research, and restore the effects on streams of more current forestry practices.  For example, intensive active in-channel and riparian restoration may be essential for splashed streams because structures needed to capture and shape gravel and wood into functioning habitat are lacking and so greatly limit the likelihood of natural recovery. The return for such large investments in restoration can be maximized by targeting splashed streams that have the intrinsic potential and forest-cover conditions needed to sustain recovery of high-quality habitat.