Adaptations For Migration

Assuming that migration evolved as an adaptation to life in seasonal environments, it may have been an important aspect of bird behaviour for a long time; it may even have occurred in the earliest species. Nevertheless, the ice ages evidently played a major role in its recent development in high-latitude regions. At the height of the last glaciation, much of the northern hemisphere above about 50°N was covered with ice, and generally devoid of terrestrial life. Colonisation of these areas by plants and animals followed the retreat of the ice, which began about 10 000-14 000 years ago and continues to this day. This means that some of the longest and most impressive of current bird migrations must have developed within this period, as birds spread gradually from lower to higher latitudes to occupy the newly available habitats. At the start of this process, when birds were confined to lower vegetated latitudes, they may have been resident or shorter distance migrants. Moreover, the last 2.5 million years have seen more than 20 successive glaciations of varying severity, so over this whole period, bird migration systems must have been in continual flux, and many populations are likely to have passed through alternating sedentary and migratory phases, each lasting for many thousands of years. The ability to migrate was perhaps ever-present in these populations, never disappearing but becoming suppressed or re-activated according to prevailing conditions and the changing pressures of natural selection (Berthold 1999).

It would be misleading to divide bird species neatly into migratory and sedentary. Both types of behaviour can be found in a single species or even in a single population. In the northern hemisphere, many widely distributed bird species are completely migratory in the north of their breeding range, completely resident in the south, and partially migratory in between, with the proportion of birds leaving any particular locality corresponding to the degree of seasonal reduction in food supplies (Chapter 13). Where populations are entirely migratory or entirely resident, their behaviour is usually considered as obligatory, and under firm genetic control. In contrast, partial migration can apparently arise in two different ways. In one way 'obligatory migratory' and 'obligatory resident' individuals occur intermixed in the same area. In the other way, the behaviour of each individual is optional (facultative), enabling it to behave differently in different circumstances, staying in the breeding area in years when conditions are favourable there, and leaving in other years (Chapter 12). This latter system can result in big year-to-year changes in the ratio of migrating to resident individuals, in particular breeding populations, with the same individuals behaving differently in different years, depending on conditions.

Although migration requires adaptations in morphology, physiology and behaviour, it is not hard to see how it might have evolved, because transitional or intermediate stages still occur. For instance, some bird species do not migrate at all, others travel only short distances, and yet others long distances. The full range of variation can be found among different populations of the same species, or even among individuals in the same population, as mentioned above. Further, the main adaptations needed for long-distance migration, such as seasonal fat reserves, timing mechanisms and orientation skills, are all found in less developed form in non-migratory birds, as well as in other animals. These features are all necessary for effective migration to develop, but may have arisen independently of migration, in different contexts.

The main features that set migration apart from other long-distance movements are that it involves a two-way journey in more or less fixed directions. An ability to perform a return movement is necessary before any migration pattern (as defined above) can evolve. Again, however, such an ability is already present in non-migratory birds, as well as in other animals, but operates over smaller distances. All birds are able to return repeatedly to their nests, as well as to particular feeding or roosting sites, and the same is true for other animals. In addition, many non-migratory birds move locally away from their nesting areas after each breeding season and return for the next. Even juveniles, after wandering widely in the non-breeding season, normally return to settle near their natal sites to breed, and other non-migratory birds revisit the same wintering sites in successive years (Chapter 17). In resident populations, such individual movements are short distance and localised, but they provide a basis from which the longer return movements of migration might evolve, step by incremental step, each extension being beneficial in its own right.

The ability of even resident birds to return to their home areas from at least a few kilometres away has been shown repeatedly in displacement experiments in which individuals were trapped, transported to a different location and released (Chapter 9; Matthews 1968, Wiltschko & Wiltschko 1999). Many of the displaced individuals quickly re-appeared at their capture sites. There is, however, a difference between the needs of short-distance and long-distance moves. In theory, short-distance moves could be performed entirely on the basis of learned landscape features, as a bird gets to know its local area. But to return from long distances requires something more, an ability to orientate by different globally available signposts, as provided by celestial and magnetic cues. Such an ability is almost certainly inherent to some degree in all birds, as it is present in lower animals, including the reptiles from which birds evolved. If birds differ in any way from other animals, it is in their use of stars to orientate, a mechanism documented in birds (Chapter 9), but not so far in other animals.

Similarly, it is not hard to imagine how, given a sense of location, directional preferences could evolve from random dispersal movements. Any birds living in seasonal environments that move long distances after breeding are more likely to meet favourable conditions in some directions than in others. Moreover, at high latitudes, individuals with an inherent tendency to move towards lower latitudes after breeding are likely to survive better than any that move in the opposite direction, so that over generations directional preferences could become fixed by natural selection. It is not just the directions, but also the distances, and hence the specific wintering areas, that could be fixed in this way, the birds from each population wintering wherever they can reach and survive best, taking account of the mortality costs of getting there and back. Suppose that the birds from a certain breeding area have heritable tendencies to fly particular directions and distances at migration time and back again in spring, but that these directions and distances differ from bird to bird. Some birds will then reach suitable areas and survive to breed again, others will reach less suitable areas and survive in small numbers, and yet others will reach unsuitable areas and die. Thus those individuals with the most beneficial migratory behaviour will perpetuate themselves, and in this way the migratory habits of a population could become fixed. Only in populations (like some nomadic ones), which on balance are as likely to find food in one direction as in any other, is no directional preference likely to become fixed by natural selection.

All birds have timing mechanisms which ensure that breeding and moult occur at appropriate and consistent times of year and, in migrants, such mechanisms also promote movements at appropriate dates. They ensure that birds arrive in their breeding areas each spring in time to take advantage of the favourable season, and leave after breeding in late summer or autumn before deteriorating conditions reduce their survival chances. These timing mechanisms are basically endogenous (within the bird) but are entrained by seasonal changes in daylength and modified by other environmental conditions (Chapter 11). Migrants also need behavioural adaptations for responding appropriately to weather conditions, both before take-off and during the journey.

Many species travelling long distances over unfavourable terrain lay down substantial body reserves for the flight. Again the accumulation of body fat is common in all kinds of animals in preparation for periods of privation, when food is scarce or unavailable, and many birds lay down body reserves in preparation for breeding or in winter in association with severe weather and long nights. In accumulating migratory fat, therefore, birds are merely modifying a pre-existing facility for a different purpose, rather than developing a completely new adaptation. It goes with long periods of fasting and endurance performance.

Because of the time it takes, long-distance migration often involves modification of other parts of the annual cycle, particularly the timing of moult (Chapter 11). Most small passerines, resident and migrant, moult in summer after breeding, but some long-distance migrants set off soon after breeding, and postpone their moult until after they have reached their tropical wintering areas. Among the Eurasian warblers, as indicated by phylogenetic analyses, summer moult is the ancestral pattern, while winter moult has evolved independently 7-10 times within this group (Svensson & Hedenstrom 1999). As these birds colonised northern breeding areas in post-glacial times, thereby lengthening their migrations, summer moult gave way to winter moult. Other patterns, such as split moults and twice-yearly moults, also seem to have evolved from the ancestral state of summer moult, in adaptation to various migration patterns. Many of the evolutionary transitions from resident to migratory or vice versa, as well as changes in the extent and pattern of migration, apparently occur without phylogenetic constraint. Many bird genera bear striking witness of this, as they include a wide spectrum of residents, short-distance migrants and long-distance migrants among closely related species. And in some species, as mentioned already, the same range of variation occurs between populations occupying different parts of the breeding range.

Migration also involves morphological changes, as birds become more adapted for long-distance flight. Compared with residents, closely related migrants usually have longer and more pointed wings, and somewhat shorter tails and smaller bodies (Figure 20.1; Leisler & Winkler 2003). Pointed wings and short tail are most efficient during level flight because they reduce drag, whereas more rounded wings and longer tail give greater manoeuvrability and more rapid lift at take-off (Kerlinger 1989, Rayner 1990). Wing shape is thus a compromise between conflicting selection pressures, the balance being drawn differently in resident and migratory populations. Change is achieved mainly by altering the relative lengths of different feathers, but in some types of birds bone lengths (femur, ulna and carpo-metatarsus) also correlate with migratory distance, as does the size of the sternum and coracoid bones, giving greater surfaces for flight muscle attachment (Calmaestra & Moreno 2000).

In several groups of birds of widely different body shapes, correlations have thus emerged within groups between morphology and migration distance

Figure 20.1 The wing formulae of two Old World wetland warblers clearly fall into two groups - rounded wings of the comparatively sedentary (or totally resident) Cetti's Warbler Cettia cetti and pointed wings of the highly migratory Sedge Warbler Acrocephalus schoenobaenus. Note the relative lengths of the outer flight feathers, which give greater 'aspect ratios' in the migrants. From Svensson (1975), in Mead (1983). See also Winkler & Leisler (1992).

Figure 20.1 The wing formulae of two Old World wetland warblers clearly fall into two groups - rounded wings of the comparatively sedentary (or totally resident) Cetti's Warbler Cettia cetti and pointed wings of the highly migratory Sedge Warbler Acrocephalus schoenobaenus. Note the relative lengths of the outer flight feathers, which give greater 'aspect ratios' in the migrants. From Svensson (1975), in Mead (1983). See also Winkler & Leisler (1992).

Sedge Warbler

Cetti's Warbler

Sedge Warbler

Cetti's Warbler

(Winkler & Leisler 1992, Marchette et al. 1995, Monkkonnen 1995, Lockwood et al. 1998, Leisler & Winkler 2003), and in Calidris sandpipers also between wing morphology and relative fuel load (itself correlated with migration distance) (Burns 2003). Yet again, such differences occur even between sedentary and migratory populations of the same species (Alerstam 1990a, Fiedler 2005). For example, among different populations of Blackcaps Sylvia atricapilla: (1) wing length, aspect ratio and wing pointedness increase; (2) wind-load decreases; (3) slots on the wing-tips become relatively shorter; (4) the alula becomes shorter in relation to wing length; and (5) the tail becomes shorter in relation to wing length, with increasing migratory distance. These changes are significantly greater than expected from the simple trend of increasing body mass from southern to northern populations (Fiedler 2005). Moreover, it is not only external features which are modified in migratory birds, but also internal ones, including the brain. Migratory birds have a more highly developed hippocampus, and a more effective spatial memory than non-migratory ones, another difference evident in comparisons between resident and migratory populations of the same species (Cristol et al. 2003).

For the most part, then, migratory birds do not possess any fundamentally different adaptations from residents, whether orientation mechanisms, physiological, morphological or other features. Migration simply involves the further development or modification of features already present in non-migratory populations. These features are necessary before effective migration can evolve, but the key novel adaptation is the evolution of narrow directional preferences to unknown areas from the situation of no such preferences shown at the population level in dispersal and nomadism. Directional preference away from the breeding area is a heritable migratory trait in its own right, distinct from other traits (Bell 2000). Like other inherited traits, each is amenable to the action of natural selection, either independently or in association with other traits. Moreover, all can be modified in an incremental manner, in which each small appropriate change brings fitness benefits. This is a firm basis from which the different movement patterns of birds, and their associated adaptations, can be moulded.

If existing bird movements were arranged in order of increasing specialisation, the sequence might run: multi-directional one-way dispersal, nomadism, multidirectional return dispersal (including altitudinal movements), facultative return migration over restricted directions, and obligate return migration over restricted directions. Dispersal movements are performed by all bird species, usually by the young after they become independent of parental care, and at the population level they can occur in any direction from the starting point (Chapter 17). Nomadism can be regarded as a form of dispersal performed repeatedly through life and usually over longer distances, as birds apparently move in any direction from one area of temporary suitability to another (Chapter 16). Dispersive migration is also multi-directional, but involves a return journey (Chapter 17). True return migrations across latitudes involve movements in restricted directions, but whereas facultative movements are stimulated largely by conditions at the time, and hence vary from year to year, obligate movements occur consistently every year, usually between fixed breeding and wintering areas, and seem more firmly under endogenous control. It is possible that this is the sequence in which the most impressive and fixed bird migrations have evolved (Terrill 1990).

Listing the different types of bird movements in this way reinforces the point that migration itself (or any other type of movement) is not a trait in its own right, but an attribute made up of several components, each of which can be independently modified by selection to give the variety of movement patterns we see today. These variable traits are all pre-requisites for the evolution of effective migration, but their origins are independent of migration. One important goal for comparative studies of migration should therefore be to identify the genes responsible for each component of migration, and establish their distribution across different bird families.

Although migration is probably an ancient phenomenon, developed in many kinds of animals long before birds arose, this need not mean that all kinds of animals use the same control systems. Similarly, orientation and navigation mechanisms are probably also ancient, but not all animals necessarily rely on the same environmental cues. Ideas on the development of migration within birds have been put forward by various authors (e.g. Lack 1954, Merkel 1966, Baker 1978, Gauthreaux 1982a, Terrill 1990, Bell 2000, Rappole & Jones 2002), and most contain elements that can be incorporated into a coherent modern theory, consistent with recent research findings. Some aspects of the origin and inheritance of migration remain speculative, but others that were uncertain only a few decades ago have since found support in controlled breeding experiments (see later).

0 0

Post a comment