Consistent range-wide pattern in fall migration strategy of Purple Martin (<i>Progne subis</i>), despite different migration routes at the Gulf of Mexico

The migration rate of Nearctic—Neotropic songbirds is expected to be influenced by whether the route is around or across migration barriers such as the Gulf of Mexico. To examine factors that influence fall migration strategies, we used light-level geolocators to track the journeys of 91 Purple Martins (Progne subis) originating from breeding colonies across the eastern range of the species. We expected individuals that crossed the Gulf of Mexico to have slower migration rates, and more stopover days in Central America to refuel after the crossing, than birds that took routes around the gulf. Owing to expected variability in conditions experienced by individuals en route, we anticipated that departure date would be a poor predictor of arrival date within and among populations. Despite widely separated breeding origins, one-way journeys of >7,000 km, and high variability in departure dates and routes, individuals showed a strikingly similar fall migration strategy. Fall migration featured a rapid (450 km day-1) initial migration covering ≥2,000 km, followed by prolonged stopovers and a slower rate of travel before the birds continued to South America. Contrary to predictions, route explained little of the variation in the overall migration rate or the rate to Central America. Stopover duration in Central America was unrelated to whether birds crossed or circumnavigated the Gulf of Mexico. As expected, breeding location (primarily longitude) was the strongest predictor of the routes that birds took at this barrier. Within-breeding-region departure date alone predicted much of the variation in arrival date at the first winter roost, but route was not a significant factor. Our results reveal a consistent range-wide pattern in fall migration strategy, with route and migration timing predicting little of the variation in rate or stopover duration.