Laserfiche WebLink
The Science of Stream Restoration - NYTimes.com <br />"Most agencies want to spend the money making things happen and not spend the money <br />finding out if they work," Dr. Dietrich said. <br />David R. Montgomery, a geomorphologist at the University of Washin on, agreed. Monitoring <br />"involves a lot of people and thought and expertise," he said. "And you don't have any new <br />projects to show for it." <br />As a result, the academic and government scientists said in their report, "Many opportunities <br />to learn from successes and failures, and thus to improve future practice, are being lost." <br />Nowadays, Dr. Montgomery said, most people agree that the best approach is to create <br />landforms and water flows that streams can maintain naturally. "But how you translate that <br />into action and at this stream rather than that stream really requires a lot of work to figure <br />out," he said. With an ailing waterway, he said, "sometimes there's a clear line between the <br />symptoms and the cause, and sometimes there's not." <br />Project failure comes in many forms. Often, Dr. Dietrich said, people design projects in hopes <br />of creating "a meandering channel with relatively low banks that look nice." Then, he said, "a <br />large storm can come through and completely wipe it out," leaving shallow channels traveling <br />around sandbars in multiple threads, what geologists call a braided channel. <br />"In most of those cases," he added, "the restorer has taken a system that is naturally braided <br />and forced it into a form. The channel simple defeated it by being its natural dynamic self." <br />At other failed sites, restorers install boulders or other stabilizing armor only to see the water <br />shift around it, leaving piles of rubble midstream. In the Pacific Northwest, people tried to <br />improve stream flow by removing bank side logs and branches only to realize that the debris <br />http://www.nytimes.com/2008/06/24/science/24str...i?_r=1&oref=slogin&ref=science&pagewanted=print (3 of 7) [6/24/2008 12:37:57 PM]