The evolution of bird migration--a synthesis.

We approach the problem of the evolution of bird migration by asking whether migration evolves towards new breeding areas or towards survival areas in the non-breeding season. Thus, we avoid the ambiguity of the usually discussed "southern-home-theory" or "northern-home-theory". We argue that migration evolved in birds that spread to seasonal habitats through gradual dispersal to enhance survival during the non-breeding season; this in contrast to the alternative idea suggesting that migration evolved towards new breeding areas to increase reproductive success. Our synthesis is based on the threshold model explaining how migratory traits can change rapidly through microevolutionary processes. Our model brings former theories together and explains how bird migration, with the appropriate direction and time program, evolves through selection after genetically non-directed events such as dispersal and colonization. The model does not need the former untested assumptions such as competition as a reason for migration and for the disappearance of sedentary populations or higher reproductive success in temperate breeding areas. Our theory offers answers to questions such as how birds with a southern origin may gradually reach northern latitudes, why migration routes may follow historical expansion routes and why birds leave an area for the non-breeding season and move back instead of breeding on their wintering grounds. The theory proposes gradual change through selection and not sudden changes such as long distance dispersal or mutations and can be applied to migration at all latitudes and in all directions. The scenario provides a reasonable concept to understand most of the existing migratory phenomena on the basis of the ecology and genetics of migratory behaviour.