Multiscale relationships of wintering birds with riparian and wetland habitat in the Willamette Valley, Oregon

Relatively little is known about wintering bird communities in lowlands of the Pacific Northwest, yet, the condition of winter habitats potentially can pose a major constraint on bird populations. I initiated a study to determine: (1) Which avian species use lowland riparian and wetland habitats in the Willamette Valley during winter, and with what intraseasonal regularity? (2) To what extent are these species associated—individually and collectively—with landscape and site-scale habitat variables? (3) To what extent are these species associated—individually and collectively—with particular landscape vs. site-scale variables? (4) How are the apparent associations affected by survey procedures and data analysis methods?; Two survey procedures (area counts and point counts) in 1995 and 1996 found 104 species during 10 visits to 142 riparian and wetland sites. Area counts mainly surveyed waterbirds and raptors; point counts surveyed passerines. From area counts, mean richness ranged 0 to 11 species among sites and 1 to 5 species among weeks at an average site; mean number of individuals ranged 0 to 1016 among sites and from 6 to 199 among weeks at an average site. From point counts, richness ranged 1 to 6 species both among sites and among weeks at an average site. Mean number of individuals ranged 0 to 1016 per site and 6 to 199 among weeks at an average site.; I individually examined 64 species to identify relative degree of association with site-scale and landscape variables. Winter Wren, Bewick's Wren, Hutton's Vireo, and Varied Thrush were significantly associated with woody cover at one or more spatial scales. Passerine abundance was greatest at sites with slow-flowing water. Richness was associated with soil inundation category, peaking at intermediate frequencies and interacting with canopy cover and wooded patch extent. Richness of non-passerines was greater where standing water occupied >6% of the landscape nearest the site. Conclusions for many species and community metrics were sensitive to choice of survey methods, variables, measurement procedures, and statistical analysis tests. Ratings generated by OFWAM, a general tool for categorizing overall wetland habitat quality, did not significantly parallel measured richness and abundance of wintering birds.