Geographic position and landscape composition explain regional patterns of migrating landbird distributions during spring stopover along the northern coast of the Gulf of Mexico

Annual migration of landbirds across the Gulf of Mexico (GOM) presents a unique opportunity to examine extrinsic processes operating at various spatial scales in determining animal distributions. Our objectives were to comprehensively quantify bird stopover densities across the northern GOM coast and model broad-scale factors explaining distributional patterns. We used weather surveillance radars to measure reflectivity of birds aloft at onset of nocturnal migratory flights and estimate bird stopover densities during four springs (2009-2012) for 6.7 million ha along the GOM. We aggregated bird densities to one longitudinal degree and 3 km of proximity to coast. Boosted Regression Tree models revealed that stopover density was related to year, longitude, proximity to coast, and amount of hardwood forest cover in the landscape. Average longitudinal patterns supported previous studies of broad-scale trans-Gulf migrant arrivals with highest density in Louisiana (92-93A degrees W) and lowest in Alabama (88-89A degrees W). Florida (83-84A degrees W) supported a second peak in migrant density, suggesting an eastern trans-Gulf route or contribution from trans-Caribbean migrants. Longitudinal patterns in migrant distributions varied strongly between years and appear generally related to variability in GOM wind patterns. Densities increased with proximity to coast, highlighting constraints on migrants to travel inland, especially in Florida's panhandle. Despite this, density was positively related to amount of forest cover more steeply along the immediate coast. Broad-scale stopover distributions of migrating landbirds along the GOM coast are heavily influenced by geographic constraints in the context of the GOM acting as a barrier to landbird migration.