Summary of Findings and Conclusions Prepared by Alaska Forest Association After the passage of amendments to Alaska’s Forest Resources and Practices Act in 1990, Sealaska and the Alaska Forest Association wanted to ensure that effectiveness monitoring, required under the Act, would occur. They therefore initiated a project, in cooperation with state and federal resource management and protection agencies, to evaluate the effectiveness of buffer zones as well as other elements of the Act and its accompanying regulations. For this study, "The Effectiveness of Riparian Buffer Zones for Protection of Salmonid Habitat in Alaska Coastal Streams," a combination of repeat in-channel surveys and low-elevation aerial photographs were used to analyze potential large woody debris. The data were also used to determine effects of buffer zones on short- and long-term supply of large woody debris. The study reviewed streams in unlogged areas as well as streams protected by buffer zones. The study looked at data taken over four years, 1994-97. It monitored changes in how large woody debris fell into streams. The study examined the interaction between large woody debris, the structure of channels, and fish habitat. Findings The study affirms previous research and adds new knowledge concerning the sensitivity of fish habitat to changes in large woody debris. Large woody debris is the most important factor for forming pools. Large woody debris also contributes to retaining gravel that forms spawning habitat. Increases in large woody debris loading can increase the frequency of small to medium-size pools. The effectiveness of large woody debris to form habitat depends on the type of channel it falls in and the amount of large woody debris already in the stream. Large woody debris is most effective in alluvial channels (deformable stream beds with clay, silt, sand or gravel or similar material deposited by running water.) Large woody debris is far less effective in confined channels that have boulder or bedrock substrate. tIn alluvial channels, maximum pool development is associated with a large woody debris load of 400 pieces per kilometer. A "piece" is defined as any wood that was a minimum of 10 centimeters in diameter and 2 meters long. At higher loadings, pool development is relatively insensitive to additions. The study analyzed 38 buffer zones and more than 11,000 trees. Nearly all of trees that tall enough to possibly become large woody debris-- 94 percent-- grew within 20 meters of the stream. That affirmed the effectiveness of the 20- meter zone. Most large woody debris--94 percent at unlogged areas and 75 percent at standard buffer zones--was derived from the inner portion of the zone, within 10 meters of the stream. Trees beyond 20 meters contributed no more than 4 percent of large woody debris. Therefore, tree removal from more than 20 meters away from the stream will not affect the potential supply of large woody debris. The study looked at selective harvest of timber in standard buffer zones. A harvest of 1-12 percent of the original stand, as well as the windthrow that followed logging, did not significantly diminish the potential supply of large woody debris in the inner part of the buffer zone, zero to 10 meters from the stream. Conclusions A 20-meter buffer zone is more effective for providing large woody debris than a wider buffer zone or an unlogged area. The outer portion, 10-20 meters, of a buffer zone will provide more trees to streams several years after logging as a result of elevated windthrow activity than unlogged areas. Since maximum pool development in alluvial channels occurs with a large woody debris loading of 400 pieces per kilometer, and large woody debris load is related to recruitable tree density in the 20-meter buffer zone, a recruitable tree density of 275 trees per kilometer could be retained in a 20- meter buffer to maintain an optimal supply of large woody debris over the long term. The calculation notes increased windthrow shortly after logging. Where recruitable tree densities exceed 275 trees per kilometer, some trees could be harvested as long as approximately 65 percent of the recruitable trees are retained within the first 10 meters and the size composition of those trees is similar to tree size in the pre-harvest stand. The capacity of the buffer zones to maintain a future supply of large woody debris was directly related to the pre-harvest recruitable tree density. Where natural stocking levels were high, recruitable tree densities remained above the optimal level following tree loss from selective timber harvest and post- harvest windthrow. Where pre-harvest stocking levels were near or below the optimal level, avoiding selective timber harvest at some sites could have minimized further reductions in the density of recruitable trees. Therefore, the standard buffer is effective for maintaining the optimum density of recruitable trees at most sites where natural stocking is sufficient. Implementation of current buffer zone regulations should include evaluation of stand composition. If the recruitable tree density and height distribution, or size composition, of trees were known, then the number and size of trees to be harvested could be determined for each buffer zone. The density of recruitable trees needed to supply large woody debris in bedrock channels was not clearly defined. Windthrow may reduce the potential long-term supply of large woody debris in a small percentage of buffer zones. A stand's orientation to prevailing winds and soil wetness could result in high windthrow after logging. To minimize those incidents, vulnerability to high windthrow should be considered during buffer design. Some streams have a naturally low supply of large woody debris. Designing buffers to take advantage of natural disturbances may provide long-term benefits to fish habitat. * * * * *