Although all main aspects of migratory behaviour have been shown to have heritable components, mainly through artificial selection and cross-breeding in captivity (Chapter 20), genetic change is not easy to demonstrate in wild birds. The most convincing way is to test wild birds in standard conditions in captivity, but this requires compliant species and suitable facilities. The assumption is that, if individuals taken from the wild in different years or from different regions express behavioural differences when held under identical controlled conditions, these differences are likely to have a genetic basis. This conclusion is strengthened if the trend is maintained in captive-bred offspring from these birds, unaffected by parental effects or experience in the wild, because only genetic effects are maintained through the generations. Such a test has been made on samples of Blackcaps Sylvia atricapilla randomly collected as nestlings from south Germany and hand-raised each year over a 13-year period (Berthold 1998, Pulido & Berthold 2004). In successive samples of birds, the amount of autumn migratory activity was found to decline towards a later onset and reduced intensity (less activity per night). This was precisely the result expected if the population had responded genetically to ameliorating environmental conditions, so at least in this species later departure and shorter migration may partly represent a genetic response resulting from natural selection.
In any population the rate of evolutionary change is limited by: (1) the amount of genetic variation within the population at the time; (2) the strength and consistency of the selection pressure; and (3) the extent to which selection on one trait causes parallel changes in others, which could be beneficial or detrimental. Genetic variance is often reduced in populations that have suffered recent numerical declines in which much of the variance was lost (genetic bottlenecks). Such variance can be increased again by immigration and gene flow from another population, or in the longer term by mutation and other means. Immigration and gene flow can also have deleterious effects if they break up locally adapted gene complexes, and render the local population less well adapted to local conditions.
Single selection events, such as spring storms, can cause rapid genetic change in the arrival dates of populations, as explained in Chapter 20, but counter-selection pressures could rapidly reverse the situation, and change arrival dates back to their original state. Selection pressures must act consistently in the same direction over several generations if they are to have any more than temporary effects on the genetic composition of a population. Most selection probably acts to stabilise the gene pools of populations rather than to change them. Moreover, some migratory traits (notably incidence, intensity and timing) are part of a syndrome of co-adapted traits (Pulido & Berthold 2003), so selection on one trait is likely to have strong simultaneous effects on the others. If this is disadvantageous in the new conditions, it may take many generations of selection to dissociate the beneficial traits from the detrimental ones before evolutionary change can occur. Evolutionary change may thus be rapid or slow, depending on the circumstances.
An important aspect of global warming is that temperatures have increased more in some regions than others, and more at some times of year than others. While the timing of spring migration could be influenced by weather conditions along the whole migration route, the timing of egg-laying depends on conditions in the breeding area. Any discrepancy between conditions en route and in the breeding area can worsen any mismatch between breeding and food supply. Moreover, in the breeding areas themselves, birds may respond more or less rapidly than their food organisms to climatic changes, so that birds cease to arrive and breed at the optimal time. An apparent example is provided by Pied Flycatchers Ficedula hypoleuca nesting in the Netherlands, where climate change has advanced the food supply on which breeding depends, but spring migration has not advanced sufficiently to allow the birds to make best use of this food supply, as they did in the past (Both & Visser 2001). The birds thereby suffered reduced breeding success and, in areas with the biggest mismatch, population levels declined by about 90% over a 20-year period (Both et al. 2006). Such mismatches can only be rectified in the longer term by changes in the genetic controlling mechanism, so as to trigger spring migration at an earlier date with respect to conditions in winter quarters. The longer the migratory journey, the less likely is weather in the breeding and wintering areas to be correlated. Long-distance migrants could have no indication from their wintering areas of how spring is developing in their breeding areas. Their departure dates from wintering areas are triggered by a photoperiodically timed endogenous rhythm, evolved through natural selection, which ensures that they arrive in breeding areas at an appropriate date (with minor variation according to prevailing conditions) (Chapter 12). Only by further evolution acting on this endogenous control mechanism is the trigger date for departure likely to be changed. In this situation, the selection pressure to migrate earlier is applied in the breeding area, but the action to accomplish an earlier arrival occurs weeks earlier in the wintering area, hundreds or thousands of kilometres away (Visser et al. 2004). Changing this control mechanism may be a relatively slow process, perhaps explaining why the arrival dates of long-distance migrants are less well correlated with temperatures in breeding areas than are the arrival dates of short-distance migrants, wintering nearer to breeding areas. Another mismatch was found in the American Robins Turdus migratorius that breed at high elevations in the Rocky Mountains of Colorado and whose spring arrival dates advanced by two weeks over a 20-year period. At the same time, winter snowfall increased and took longer to melt, giving a mismatch between arrival dates and the exposure of bare ground-feeding areas (Inouye et al. 2000).
These examples raise a general point that the photoperiodic responses of many birds, through which their annual cycles are timed, may become less reliable predictors of seasonal change in food supplies as climate change alters the phenology of their food supplies. This is not a new problem, as it is faced by all birds which expand their breeding ranges into different regions, but it may take time for them to adjust genetically to new situations, during which time they could perform less well than usual (though not necessarily with effects on population levels).
Was this article helpful?