Abstract
BACKGROUND: High-latitude bird migration has evolved after the last glaciation, in less than 10,000-15,000 years. Migrating songbirds rely on an endogenous migratory program, encoding timing, fueling, and routes, but it is still unknown which compass mechanism they use on migration. We used geolocators to track the migration of willow warblers (Phylloscopus trochilus yakutensis) from their eastern part of the range in Russia to wintering areas in sub-Saharan Africa. Our aim was to investigate if the autumn migration route can be explained by a simple compass mechanism, based on celestial or geomagnetic information, or whether migration is undertaken as a sequence of differential migratory paths possibly involving a map sense. We compared the recorded migratory routes for our tracked birds with simulated routes obtained from different compass mechanisms. RESULTS: The three tracked males were very similar in the routes they took to their final wintering sites in southern Tanzania or northern Mozambique, in their use of stopover sites and in the overall timing of migration. None of the tested compass mechanisms could explain the birds' routes to the first stopover area in southwest Asia or to the destination in Southeast Africa without modifications. Our compass mechanism simulations suggest that the simplest scenarios congruent with the observed routes are based on either an inclination or a sun compass, assuming two sequential steps. CONCLUSIONS: The birds may follow a magnetoclinic route coinciding closely with the tracks by first moving west, i.e. closer to the goal, and thereafter follow a constant apparent angle of inclination to the stopover site. An alternative would be to use the sun compass, but with time-adjustments along the initial part of the migration to the first stopover, and thereafter depart along a new course to the winter destination. A combination of the two mechanisms cannot be ruled out, but needs to be confirmed in future studies.