The above discussion argues in favour of current migration systems developing as species spread from lower to higher latitudes, becoming more migratory as they reached increasingly seasonal environments. This is the situation that would have affected all those species that colonised high latitude areas after the last glaciation, and is evident today in species that are still spreading north. However, we can imagine that the same process in reverse occurred repeatedly through the Pleistocene, as glacial conditions advanced from high to lower latitudes. In those conditions, previous resident populations would need to become migratory in order to survive the colder winters, before being finally obliterated from their higher latitude breeding areas entirely by the advancing cold, and surviving only at the lowest latitudes. As we are now in an interglacial, the existing migration system is likely to have been established the other way round, as explained above, but there are also recent examples of resident species introduced to higher latitudes by human action then developing migration to lower latitudes secondarily, and thus acquiring new wintering areas. The most recent involves the House Finch Carpodacus mexicanus, introduced from California (where it is sedentary) to the New York area. As the population expanded, the migratory habit developed, and now a large proportion of the once sedentary population migrates to winter further south (Able & Belthoff 1998). The same happened earlier in the Common Starling Sturnus vulgaris, after birds supposedly from the resident British stock were introduced to the New York area. Now that the species has spread over much of North America, the most northerly breeding populations are migratory or partially migratory, and only the most southern ones are sedentary. Providing that 'genes' for migration persist in populations through both resident and migratory phases in their history, the one type can develop from the other whenever distribution or conditions change. There is probably no time when conditions are not changing somewhere in the range.
Was this article helpful?