- Urban ecosystems are characterized by human disturbance and changes in the amount, types, and spatial arrangement of wildlife habitat. The relative importance of habitat and human associated variables to urban birds is unknown. In 1999, I surveyed spring bird and plant communities along 54 perennial streams
in the Portland, Oregon metropolitan region. My objectives were to determine (1) what habitat and human-associated variables were related to avian community measures, (2) how far from riparian areas did relationships between avifauna and canopy cover and development extend, and (3) habitat characteristics associated with individual bird species. I used Principal Components Analysis to cluster and reduce the number of explanatory variables, and regressed bird community variables against the resulting principal components. Total and nonnative bird abundance were negatively related to the first principal component, PC 1 (high-scoring PC 1 sites had wide, structurally complex native forests) (r2 = 0.38, P < 0.0001 and r2 = 0.63, P < 0.0001, respectively). Several weak (r2 = 0.17-0.31, P < 0.003) curvilinear relationships emerged between overall and native species richness and diversity versus PC 1. Neotropical migratory bird (NMB) species richness and diversity were negatively related to human development variables, and positively related to native shrubs (adjusted r2 = 0.21 and 0.18, respectively; P < 0.003). I regressed bird community variables against canopy cover and street density for each of ten 50-m buffers around sites, then plotted the model fit (r)
against buffer width. Native and nonnative abundance measures were negatively related to canopy cover. Total abundance was best predicted by canopy cover 300-350 m from the stream whereas normative abundance was predicted nearly equally by canopy cover in each buffer. The relationships between total species richness, native species richness, total diversity and native diversity were
curvilinearly related to canopy cover, with canopy in the first 50 m appearing most
important. Neotropical migratory bird species richness, diversity and abundance were negatively related to street density within 150 m of the stream. Brown-headed Cowbird (Molothrus ater) abundance was increasingly, negatively related to canopy cover as spatial scale increased. I used logistic regression to identify habitat and/or human-associated variables that predicted the occurrence of
32 species. The occurrence of four species (American Crow [Corvus brachyrhynchos], House Finch [Carpodacus mexicanus], Red-breasted Nuthatch [Sitta canadensis], and Vaux's Swift [Chaetura vauxi]) was positively, and five
species (Black-headed Grosbeak [Pheucticus melanocephalus], Common Yellowthroat [Geothlypis trichas], Rufous Hummingbird [Selasphorus rufus], Steller's Jay [Cyanocitta setlleri], and Swainson's Thrush [Catharus ustulatus]) negatively related to human-associated variables. My results suggest that increasing urban canopy cover is the most valuable land management action for
many breeding bird species. Preserving large undeveloped habitats and controlling urban sprawl may help prevent regional declines of NMB species.