You are not currently logged in.
Access JSTOR through your library or other institution:
Breeding Bird Abundance in an Urbanizing Landscape in Coastal Southern California
Douglas T. Bolger, Thomas A. Scott and John T. Rotenberry
Vol. 11, No. 2 (Apr., 1997), pp. 406-421
Stable URL: http://www.jstor.org/stable/2387615
Page Count: 16
You can always find the topics here!Topics: Species, Habitat conservation, Urban habitats, Birds, Habitat fragmentation, Conservation biology, Shrubs, Breeding, Wildlife habitats, Modeling
Were these topics helpful?See something inaccurate? Let us know!
Select the topics that are inaccurate.
Preview not available
The relative influence of local versus landscape heterogeneity on species abundance is an important consideration in the design and management of nature reserves in heterogeneous landscapes. We described the spatial patterns of resident breeding bird abundance at a landscape scale in an urbanizing landscape gradient in coastal San Diego County, California (USA). The gradient spanned approximately 260 km2 and included a large, mostly contiguous block of undeveloped shrub habitat mosaic which graded into an area of internally fragmented habitat and finally into an area characterized by discrete patches of habitat in an urban matrix. We quantified the distribution of resident bird species in this landscape with point counts and used stepwise logistic regression and canonical correspondence analysis to explore the relative strengths of association between bird abundance and local habitat variation versus landscape variation for the 20 most common bird species. We measured local habitat conditions in a 100-m radius around each point count station. Landscape variables were extracted from a GIS database of this region and included habitat patch size, distance to developed edge, and two indices of urban exposure. The addition of landscape variables to stepwise logistic regression analyses substantially improved the predictive power of the resulting models in 14 of 20 species. Model concordance (the ability of the model to correctly classify all pairs of divergent sites) increased from a mean of 56% without landscape variables to 76% when one or more landscape variables were selected for inclusion in the final model. Response of the 20 focal species to the landscape gradient differed markedly among species. Canonical correspondence analysis was used to rank species along a landscape response gradient from edge/fragmentation reduced species to edge/fragmentation enhanced species. The four species exhibiting the most edge/fragmentation reduced pattern were sparrows: Sage (Amphispiza belli), Rufous-crowned (Aimophila ruficeps), Lark (Chondestes grammacus), and Black-chinned sparrows (Spizella atrogularis). Costa's Hummingbird (Calypte costae) also showed a pattern of edge/fragmentation reduction. The opposite end of the gradient was occupied by species whose densities are elevated near edges and in fragmented areas. The middle of the gradient was occupied by species that were abundant, widely distributed, and showed no sensitivity to landscape position or shrub habitat type. The abundance of edge/fragmentation reduced species appears to be depressed within 200 to 500 m of an edge and the abundance of the edge/fragmentation enhanced species is elevated up to 1000 m from an edge, depending on the species. Conservation analyses, including GAP analysis, often assume that the relationship between species abundance and habitat is invariant with landscape position. In haterogeneous landscapes these analyses may underestimate the conservation risk to species in the edge/fragmentation sensitive category because their abundance varies strongly with landscape location.
Conservation Biology © 1997 Wiley