Abstract
For decades, researchers have successfully used ground-based surveys to understand localized spatial and temporal patterns in stopover habitat use by migratory birds. Recent technological advances with WSR-88D radar now allow such investigations on much broader spatial scales. Both methods are assumed to accurately quantify patterns in migrant bird communities, yet information is lacking regarding relationships between radar estimates of migration and different ground-based monitoring methods. From 2005 to 2007, we monitored migrant communities on or near two Department of Defense installations in the spring (Ft. Polk Military Complex, LA; U.S. Army Test and Evaluation Command, Yuma Proving Ground, AZ) and on two installations in the fall (Ft. Polk Military Complex, LA; Eglin Air Force Base, FL) using both ground-based transect surveys and radar imagery of birds aloft. We modeled daily changes in migrant abundance and positive and negative species turnover measured on the ground as a function of radar estimates of migrant exodus and input densities. Radar data were not significant predictors of any response variable in any season either in the southeastern or southwestern United States, indicating a disparity between the results obtained using different methods. Multiple unique sources of error associated with each technique likely contributed to the conflicting outcomes, and researchers should take great care when selecting monitoring methods appropriate to address research questions, effects of management practices, or when comparing the results of migration studies using different survey techniques.