A second line of evidence proposed in support of a tropical origin for many longdistance intercontinental migrants comes from their choice of habitat. The overall species numbers in the Palaearctic-Asian and Nearctic-Neotropical systems are similar (about 338 species in both), and much greater than in the Palaearctic-Afrotropical system (186 species, excluding seabirds). The main deficiency in the Palaearctic-Afrotropical system is in forest birds. In total, about 48 species of long-distance migrants breed in west Palaearctic forest, while about 112 such species breed in Nearctic forests and 107 in east Asian forests (Moreau 1972, Rappole 1995). However, while nearly all the Asian and Nearctic species that breed in forests also winter in some type of forest, only five Palaearctic migrants winter in African tropical forests, mainly round the edges, namely European Honey Buzzard Pernis apivorus, Eurasian Golden Oriole Oriolus oriolus, European Pied Flycatcher Ficedula hypoleuca, Collared Flycatcher F. albicollis and Wood Warbler Phylloscopus sibilatrix (Monkkonen et al. 1992). The remaining species (apart from aquatic and coastal birds) winter in savannah and scrub (for comparison between Palaearctic-Afrotropical and Nearctic-Neotropical systems, controlling for phyl-ogeny, see Bohning-Gaese & Oberrath 2003).
These differences between migration systems throw further light on their evolution. Throughout the Tertiary Period, including the glacial cycles, the east Palaearctic and Nearctic forest species have had continuity of forest from their breeding to their wintering areas, while the west Palaearctic forest species have been separated from the Afrotropical ones by a broad belt of arid scrub and desert. These intervening habitats could have acted in the past as a filter, screening out any species not adapted to those habitats and letting through those that were: that is, mainly forest species from the American and Asian tropics and scrub species from Africa. Any forest-dependent species from tropical Africa would throughout their history have confronted, in any attempted northward range expansion, thousands of kilometres of arid scrub or desert (during the glacial periods, the Sahara was much better vegetated than today, but apparently by scrub rather than high forest). The same filter could have also prevented European forest species from invading Africa, but the other facts on species numbers and relationships presented above, together with genetic evidence from some species, strongly suggest movement mainly from the tropics outwards (or at least from lower to higher latitudes). This would in any case be expected as the commonest pattern, because many more bird species occur at low than at high latitudes, and because the bout of re-colonisation following each glaciation would have been mainly from low to high latitudes.
Nevertheless, the poverty of the west Palaearctic forest avifauna cannot be attributed entirely to the effects of the Sahara as a barrier to spread. The forests in Europe were much more thoroughly obliterated by the glaciations than were the forests of North America and eastern Asia. This process is known from palaeontological evidence to have eliminated large numbers of the plants and animals found in west Palaearctic forests in pre-glacial times (many of which still survive today in Asian forests (at least at the generic level). Among birds, extinctions are likely to have included migratory and resident species. We thus have two explanations for the small numbers of forest species in Europe compared with other regions, but it is clear that not many forest species could have colonised Europe from Africa in recent post-glacial times (say the last 15 000 years). The opposite is true for scrub species, however, which may account for the predominance of scrub species among current Eurasian-Afrotropical migrants.
Was this article helpful?