Four basins were chosen from a population of beaver-occupied basins in the lower Alsea drainage area of the Oregon Coast Range for proximity to one another and a high proportion of reaches flanked by mature stands. Within basins, three each of impoundments caused by beaver, impoundments caused by other factors (debris jams) and unimpounded stream reaches were randomly chosen for further study. Dams were further visually classified as new dams (no more than one year old in places that appeared to have no impoundment history), recolonization dams (new dams in areas impounded in the recent past) and old dams (dams that appeared to be many years old; frequently had willows growing from the cuttings anchoring the dams in place). In order to place unimpounded and jam sites in the same geomorphic context as the impoundment sites, the site selection (for all site classes) within a basin was limited to the set of sites typically occupied by beaver as described by the Habitat Suitability Index (HIS) for beaver in the Coast Range (Suzuki and McComb 1998). Suzuki and McComb (1998) concluded that beaver preferentially located their dams in streams with <3% gradient, a stream width of 3-4 m, and valley floors of 25-30m. We limited the search for dams to areas in third order basins to streams with a <5% gradient and a stream width of <5m. Reaches sampled were surrounded by mature forest stands.
To sample the riparian forest communities, one 10 x 30m transect extending from the edge of the water into the forest was established on each side perpendicular to the stream at each site, and subdivided into six 5 x 10 m plots. Thus, the forested transect included portions of the emergent and littoral zones. We measured cover of shrub species and number of individual trees measuring >15 cm dbh in each plot, as well as distance of the plot from center of impoundment or stream and estimated height of the plot above water level (at plot center).