Abstract
Aim Conservation planning for migratory species requires knowing how populations are connected throughout the annual cycle. We measured migratory connectivity for five migratory songbird species by combining stable‐hydrogen isotopes in feathers (δ 2 H F ) with citizen science data on breeding range abundance. Location Caribbean Basin and North America. Taxa Neotropical migratory birds: Mniotilta varia , Setophaga ruticilla , Seiurus aurocapilla , Setophaga americana , and Setophaga discolor . Methods We analyzed δ 2 H F in feathers grown on North American breeding grounds and sampled at 20 Caribbean Basin sites across most of the non‐breeding range of each species. We made Bayesian assignments to breeding origin by combining δ 2 H F values and North American Breeding Bird Survey abundance data. For each species, we used cluster analysis to group Caribbean sites into non‐breeding regions that shared birds with similar breeding ground assignment probabilities. We then mapped assignment probabilities for the non‐breeding regions of each species onto breeding ground δ 2 H F surfaces and estimated the strength of migratory connectivity. Results Apart from some general similarities, the five species demonstrated different patterns of migratory connectivity. Species with larger non‐breeding ranges exhibited stronger migratory connectivity compared to those with smaller non‐breeding ranges. Despite geographical overlap, birds from each non‐breeding region had mostly unique breeding ground origins. Main conclusions The contrasting patterns of migratory connectivity shown by the five species suggest that geographic distribution during the non‐breeding period is a poor indicator of breeding ground origin. Our results indicate that full annual cycle conservation planning, which requires migratory connectivity information, should proceed on a species‐specific basis.