Journal of Ornithology

Much of our knowledge about migratory behaviour comes from single-population tracking studies. Some such studies had a considerable impact on migration research at large. However, it is often unclear how representative such studies are for other populations, even of the same species. We compared migration corridors and schedules of GPS-tracked adult Western Marsh Harriers (Circus aeruginosus) from two nearby breeding areas within the West African–Eurasian flyway: the Low Countries (LC, N = 6) and southern Sweden (SW, N = 4). Assuming the migration patterns revealed by decade-old satellite-tracking of SW harriers are representative for the species, we expected LC harriers to make a narrow loop migration involving a western spring detour out of Africa, but according to an earlier schedule tuned to the earlier onset of spring in the Low Countries. In autumn, LC harriers migrated significantly further west than SW harriers all the way to their common non-breeding range. In spring, both groups detoured westward across the Sahara, as expected, and diverged towards their breeding areas after reaching mainland Europe. LC harriers migrated slightly earlier than SW harriers in autumn. However, LC harriers unexpectedly left their non-breeding sites up to a month before SW harriers, after which they made long stop-overs in northwest Africa. Late-departing SW harriers forewent these stop-overs, and thus caught up with LC harriers in northwest Africa, so that both groups reached their breeding areas simultaneously. While we anticipated strong overlap between LC and SW migration corridors in spring, we failed to anticipate the earlier and more variable spring departures of LC harriers. Early spring departures did not result in earlier arrivals by LC harriers. Instead, we suspect they departed early to escape faster deteriorating foraging conditions at their non-breeding sites. Such environmental modulation of migratory behaviour may complicate generalization of migration patterns in other birds.