It is easy to envisage how migration might evolve in species that have suitable areas for breeding and wintering adjacent to one another, giving a continuum of habitat between the two. This would then enable a gradual incremental lengthening of the journey. But it is less easy to understand the origin of those migrations in which the breeding and wintering areas are separated by hundreds or thousands of kilometres.
How can birds from one part of the world come to winter in another part so far away when the intervening area is unsuitable?
One possible way is through the loss of intermediate habitat or populations. In many species whose breeding and wintering areas are now contiguous, the birds from the most northerly areas often migrate furthest, over-flying the birds in intervening areas (leapfrog migration, see later). But if the intermediate populations for some reason died out, this would leave the outer areas, one for breeding and the other for wintering, with a gap between the two. The intervening population may die out through the gradual loss of intervening habitat, so that birds initially migrating in continuous habitat found themselves crossing first short then increasingly long stretches of unsuitable terrain.
The main difficulty is in explaining how landbirds could evolve in one step the ability to cross large stretches of sea or desert, which for most species offer no opportunities for refuelling. If a species were to evolve such a migration today, it would entail not a gradual change in migration physiology but an abrupt change, with markedly greater fat deposition and flight lengths. One obvious possibility is that barriers which are now wide, such as the Sahara Desert, were once narrow, so that a progressively longer migration could have evolved gradually as the barrier widened. Migrants would then have been able to lengthen their flights bit by bit, to match the need. Climate changes, with their effects on vegetation, could thus lead to the splitting of a once-contiguous vegetation type, in the way that the Sahara Desert now separates scrub to the north and south, or the boreal forest now separates open tundra from open steppe. This would enable birds to evolve gradually longer migrations as their breeding and wintering areas became progressively more separated. Alternatively, although a barrier may always have been wide, it may once have held more potential stopping places than are present now, perhaps more oases in the desert or more islands in the sea. If such stopping places disappeared gradually, this would again have allowed the progressive evolution of longer flights. This is the second way in which long flights over hostile areas might have evolved. It is happening today as stopover sites are lost, either through climatic change (as in the expansion of the Sahara Desert), or by more direct human action in destroying already fragmented habitats.
The difficulty of evolving a migration over the Sahara Desert (as it now is) may explain why most species known to have spread westward in Eurasia in historical times still return to traditional winter quarters in Southeast Asia, even though apparently suitable habitats exist in Africa to the south. It may also explain why so few forest birds (as opposed to scrub birds) cross the Sahara, because throughout the Tertiary Period, the west Eurasian forests were separated from the African forests by an even wider belt of arid land (now mainly the Sahara Desert).
The evolution of some migration routes, notably onto remote islands in the Pacific, is particularly hard to understand. Yet the entire population of the Bristle-thighed Curlew Numenius tahitiensis, which nests in western Alaska, now winters on islands in the central Pacific, along with large numbers of Pacific Golden Plovers Pluvialis fulva, Wandering Tattlers Tringa incana, and other species. These islands are volcanic, having arisen directly from the sea-bed, so were never connected to a continental land mass. Some are no more than a few million years old. Still other species, such as the Bar-tailed Godwit Limosa lapponica, fly from Alaska almost straight south across the Pacific to winter quarters in New Zealand
(Chapter 6). Such migration routes offer only a few scattered volcanic islands or coral atolls where birds could rest and feed. While in glacial times, when sea levels were lower, the numbers and sizes of islands were larger than today, the hazards of the long oversea migration must always have been great, posing problems of how it could have evolved.
One obvious mechanism involves progressive corner cutting. Imagine a shore-bird that once migrated from Alaska via the Aleutian Islands in the north to winter on the coasts of Southeast Asia and Australasia. By cutting across the Pacific initially over short distances in the north, and then over progressively longer distances further south, it could gradually shorten its total journey, but by making progressively longer overseas flights. The gradual elongation of overwater flights could also explain how migration to Pacific Islands evolved, if the islands were once stepping stones on a longer journey to Southeast Asia and Australasia, which was then foreshortened. These different hypothetical stages are shown by different species today, which make overwater flights of different lengths between their breeding and wintering areas. The hypothetical starting situation is still shown by stints and others that migrate from Alaska first west, then southward down the coasts of eastern Asia. Representing an intermediate stage, other species cross the northern Pacific to join the Asian coastline at some point, before proceeding southward along that coastline. At the extreme are the Bar-tailed Godwits Limosa lapponica, mentioned above, which now migrate, apparently mostly non-stop in autumn, from Alaska across the Pacific to New Zealand (Chapter 6).
Migration over other sea areas, such as the Gulf of Mexico, could have evolved in the same way, from an initial movement around the western Gulf coast to a gradual lengthening of the overwater flight, but a consequent reduction in the total journey length. So too could the long overwater route between northeastern North America and South America now taken in autumn by certain shorebirds and songbirds. This pattern could have started as a mainly coast-hugging eastern North American flight, crossing the Caribbean Islands, and then shifted progressively eastward over the Atlantic, to give a progressively more direct (and hence shorter) overall route, but again requiring a progressively longer, non-stop overwater flight (Figure 22.3). The advantage of a direct overwater flight is not only that it cuts the distance, but also that it reduces the risks of daytime predation from falcons and other predatory birds. This and other long overwater flights are undertaken mainly in autumn when winds are favourable, but in spring when winds are unfavourable the birds take a longer and safer overland route.
Corner-cutting might also explain the current migration pattern of Icelandic birds. This island has been colonised during the last 10 000 years by landbirds entirely from Europe. The predominant migration direction in these species within Europe is southwest, yet in departing from Iceland the same species fly southeast. Only in this way could they reach the nearest suitable wintering area in the British Isles. If these species migrated back along their most likely colonisation route, they would migrate first east and then southwest, but a progressive change from east to southeast would bring them more swiftly to their current wintering areas.
All these various mechanisms for the evolution of long-distance barrier crossing can be inferred from existing patterns of variation. They would allow gradual development of a long-distance migration system over seas or deserts, without the need for a sudden step-change in one or more aspects of migration behaviour.
This makes it easier to understand how such long and difficult migrations might have evolved. It does not of course prove that they did evolve in this way.
Was this article helpful?