Roadside versus remote: Is the habitat sampled by the Breeding Bird Survey in Maryland representative of statewide conditions?

Abstract

The North American Breeding Bird Survey (BBS) has received criticism that the bird habitat sampled along the 24.5 mile long roadside transects may not be proportional to regional totals. If true, trends in bird populations recorded by the BBS may not be sensitive predictors of regional or continental change in songbird abundance. To test whether the approximately 60 BBS routes in Maryland representatively sample the state's habitat, a geographic information system (GIS) database was compiled of significant bird habitat identified from remotely sensed landcover and land-use information (e.g., Multi-Resolution Land Characteristics Consortiumclassified Landsat Thematic Mapper imagery, etc.). These GIS data layers were analyzed to determine the statewide acreage of identified habitats as well as the acreage in each of the major physiographic regions of Maryland. Regional and statewide totals were also extracted for the subsample of habitat within 30 m of the BBS transects. The results of the comparison of regional and statewide habitat totals with the BBS sample showed very low proportional difference for nearly all of the identified habitat parameters. For Maryland and perhaps other urbanizing states, the BBS provides an accurate sample of available songbird habitats.