ABSTRACT
The thesis developed in this paper is that the phenomenon of bird migration is founded on (1) pre-adaptations for flight that reduce the cost of long-distance directional movement, and (2) intensive reproductive effort that makes the use of the northern short season of high food abundance very beneficial for breeding. It is proposed that the first step in the evolution of migration is long-distance pre-breeding dispersal of a fraction of the juveniles of southern species to northern latitudes, where they breed sind then disperse south. Further adaptations for timing and navigation mechanisms turn the dispersal into directional seasonal migration. Finally, reproductive isolation, drift, and selection result in speciation of the northern-breeding migratory fraction from the nonmigratory southern-breeding ancestor species. The evidence supporting this hypothesis is scanty and circumstantial, and includes the high incidence of migration among northern species, several examples of species that represent steps in speciation of Palearctic migrants from resident African relatives, and the high incidence in many avifaunas of “vagrants” and “occasional breeders” that may represent the initial stages of an ongoing evolution towards migration. Research attention should be paid to the phenomenon of post-fledging juvenile dispersal as a potential precursor of the evolution of migration.