In contemplating the development of migration, two scenarios have been proposed: (1) that the breeding range is the original year-round home of a population, and the wintering range is the secondary home, visited to enhance survival through the most difficult season (e.g. Lack 1954, Cohen 1967); or (2) that the present wintering range is the original year-round home, and the breeding range is the secondary home visited to enhance reproductive prospects (Cox 1968, 1985, Levey & Stiles 1992, Rappole & Tipton 1995, Safriel 1995, Poulin & Lefebvre 1996, Bell 2000, Bohning-Gaese & Oberrath 2002, Rappole & Jones 2002). In theory, current migration patterns could have evolved in some species from one of these starting points, and in other species from the opposite starting point. However, as most of the migration that we now observe developed as birds spread from low to high latitudes at the end of the last glaciation, range expansion and migration almost certainly developed hand in hand. This not a question of the origin of migration, but of how past migratory patterns were progressively modified to produce those of today.
The process was evident recently in the northward expansion of the European Serin Serinus serinus (Chapter 21). This species is resident in southern Europe, but as it has extended its breeding range northwards, it has become increasingly migratory in the newly colonised areas. The same holds for other species elsewhere, including the Cattle Egret Bubulcus ibis in North America. Such recent colonists reveal that the occupation of higher latitudes can occur over a matter of decades, rather than millennia, if suitable habitat already exists. The colonisation of new breeding areas is achieved by normal dispersal processes, and migration develops secondarily in response to conditions in the newly occupied breeding areas, wherever individuals survive better, on average, by moving out for the winter than by staying in their breeding areas year-round. According to Berthold (1999), most bird populations are likely to contain a mixture of migratory and non-migratory individuals, or else a range of individuals with different thresholds for the initiation of migratory behaviour. This inherent variation within a population provides the raw material on which natural selection can act to produce a predominantly migratory population from a predominantly resident one, or vice versa (Chapter 20).
We can suppose that the post-glacial colonisation of northern regions involved a general northward movement of bird breeding ranges, associated with the northward advance of suitable vegetation, following ice melt. Evidence of this process derives from the post-glacial range expansions of bird-dispersed plants following ice melt, both in Europe and North America. Bird species that were non-migratory could have remained so, or become migratory in the more northern of the areas colonised. Species that were migratory during the glaciation (wintering at lower latitudes than their breeding refuges) could have moved both their breeding and wintering areas northward, or they could have retained the same wintering areas, and simply lengthened their migrations as their breeding range moved northward. The latter scenario could have held in species that, throughout the glaciation until now, have continued to winter in tropical or opposite-hemisphere regions.
Along similar lines, Rappole & Jones (2002) proposed the following steps in the evolution of tropical-temperate migration, although they were concerned with range spreads over several million years, and not merely those that followed the last glaciation. On their view:
1. Long-distance migrants from tropical to temperate regions originated as predominantly sedentary, tropical-breeding residents.
2. These birds tended to expand their breeding ranges into all adjacent and suitable habitat by normal dispersal processes.
3. Increased food availability, along with reduced competition and predation, in seasonal temperate habitats during summer allowed higher reproductive rates in temperate breeding populations relative to tropical populations of the same species (a difference still evident today).
4. These fitness benefits favoured rapid (in an evolutionary sense) invasion of temperate habitats by those tropical species able to exploit them.
5. Spreading out from the tropics into the more seasonal environments of higher latitudes, migratory genotypes, which already existed within the population, were favoured over resident ones, and under natural selection the expanding populations changed from predominantly resident to predominantly migratory, returning to lower latitudes for the winter. The latitude at which migration was favoured over residency is likely to have varied between species, depending on their particular food needs, but giving in each species a geographical gradient from wholly resident at the lowest latitudes, through partially migratory to wholly migratory at the highest latitudes.
Some authors have proposed the development of small-scale seasonal dispersal movements (tracking locally available resources) as a precursor to the spread to higher latitudes (e.g. Levey & Stiles 1992, Rappole 1995, Safriel 1995). However, individuals of every bird species disperse to some extent, and there is no reason why range expansion should not also have occurred in the most sedentary of resident species that developed migration secondarily as they spread into more seasonal environments (Bell 2000).
Was this article helpful?