Reliability of the Breeding Bird Survey: Effects of restricting surveys to roads

Breeding Bird Surveys (BBS), which are widely used to monitor trends in avian populations (e.g. Robbins et al. 1989, Sauer and Droege 1993), are conducted along roads but are used to infer changes in regionwide populations. Such inferences may be inaccurate if trends in habitat along roads differ from regionwide trends. For example, if forest cover regionwide remained constant but forest cover along roads declined (due for example to development), then BBS data for species found primarily in the forest might show declines despite regional populations being stable. We investigated this issue by measuring change in forest cover in western (i.e. unglaciated) Ohio (Fig. 1). Change in forest cover between 1963 and 1988 was determined for: (a) the complete study area; (b) areas 0 to 140 m from a road (inner roadside strip); and (c) areas 141 to 280 m from a road (outer roadside strip).