Sometimes large numbers of predators accumulate at stopover sites, relative to the numbers of potential prey. This can make individual birds more vulnerable than usual. In addition, when accumulating body reserves, migrants often feed intensively, reducing vigilance and spending more time in places where the danger is greater. Accipiters and falcons are often seen catching birds at stopover sites (Rudebeck 1950, Kerlinger 1989, Lindstrom 1989, Moore et al. 1990). On the basis of studies at Falsterbo in south Sweden, Lindström (1989) estimated that raptors (mainly Sparrowhawks Accipiter nisus) removed 10% of all Chaffinches Fringilla coelebs and Bramblings F. montifringilla during the six-week period of autumn migration. This rate of predation was much greater than expected from the annual mortality rate in these species, if losses had been evenly spread through the year.
Although prey-birds can respond to the presence of predators through greater vigilance and selection of safer habitats, both these measures may reduce feeding rates (Metcalfe & Furness 1984, Lindström 1990). This is because continual scanning for predators takes time which could otherwise be used for feeding, while secondary habitats usually offer less food. Moreover, the weight increase associated with fuel deposition reduces lift-off speed and agility, supposedly making prey easier for a predator to catch (Chapter 5). Little wonder that some birds increase their vigilance and feed in safer places as they become heavier (Burns & Ydenberg 2002).
Disturbance at stopover sites, caused by natural predators or people, can have marked effects on the rates and extent of weight gain by migrants, and hence on subsequent survival or breeding success. Pink-footed Geese Anser brachyrhychus that migrate from the Netherlands to breed in Spitzbergen make a major feeding stop at Vesterälen in north Norway. For many years, this population had been increasing. From about 1993 on, however, local farmers began systematically to disturb geese from their grasslands. This continual harassment prevented most of the geese from accumulating adequate body reserves, as was apparent from their body shapes ('abdominal profiles'). In contrast, geese using small areas of undisturbed habitat still achieved large reserves. Progressively, the geese began to abandon this site in favour of a less-good site further south, their spring-summer mortality increased, and the mean production of young per pair declined. The population stopped increasing (Madsen 1995, Drent et al. 2003).
A second example of the influence of disturbance came from an experiment in Canada, designed to curb the population growth of the Greater Snow Goose Chen caerulescens atlantica. On the main spring staging areas beside the St Lawrence River, the geese had been protected since 1917, and their numbers had increased greatly, but in the 1990s a spring hunt was reinstated. The main effects of this hunting came not from the numbers killed, but from the effects of disturbance on the accumulation of body reserves and subsequent nest success. The migration and breeding behaviour of large samples of radio-tagged geese were compared between two non-hunting years and the first two years with spring hunting (Mainguy et al. 2002). In the non-hunting years, 85% of the 80 radio-tagged females identified on the spring staging areas in the St Lawrence River valley were subsequently found in breeding areas on Bylot Island, where 56% were known to have nested. By contrast, in hunting years, only 28% of 80 radio-tagged females identified in the St Lawrence valley reached the nesting areas on Bylot Island, and a mere 9% nested. The differences between years in these proportions were statistically significant (proportion present, X2 = 57.6, P < 0.001; proportion nesting, x2 = 16.9, P < 0.001). They were not due to loss of birds to spring hunting, because most of the missing radio-tagged birds turned up again in the autumn. Rather, many radio-marked geese had shortened their migration and stopped on Baffin Island, where they did not breed. Moreover, females shot near the nest in hunting years were lighter in weight (F = 12.7, df = 37, P < 0.001), and contained less breast muscle (F = 12.3, df = 36, P <0.001)
and abdominal fat (F = 6.1, df = 34, P < 0.02) than some shot near the nest in earlier years with no spring hunt. Egg-laying in the colony was delayed about a week in both hunting years, and clutches were significantly smaller, compared with four previous years (Bety et al. 2003). The implication was that spring disturbance at the St Lawrence staging sites reduced the feeding rates and body condition of geese which survived the hunt, adversely affecting their subsequent migration and reproduction.
In another study, the disturbance effects of autumn shooting of waterfowl were tested by setting up experimental reserves in two Danish coastal wetlands (Madsen 1995). Over a five-year period, these undisturbed reserves became important staging areas for waterfowl, increasing the national totals of several species. Hunted species increased the most in these reserves, some four- to 20-fold, while non-hunted species increased two- to five-fold. Furthermore, most quarry species stayed in the area for up to several months longer each winter than in earlier years. No declines in bird use were noted at other Danish sites still open to hunting, so the accumulation of birds in the reserves was attributed to the short-stopping of birds that would otherwise have migrated further south. In this and other studies, hunting disturbance emerged as a major factor influencing the migration and winter distribution of waterfowl, on both local and regional scales. But it is not just waterfowl that are susceptible to human disturbance. The mean body mass of Sandhill Cranes Grus canadensis dropped by more than 7% at a staging area in North Dakota after the hunting season was brought forward from November (when most of the cranes had left) to September (Krapu & Johnson 1990).
Other changes in the behaviour of migrant birds towards shorter stopovers at certain sites have been attributed to enhanced predation risk associated with the increasing numbers of Peregrine Falcons Falco peregrinus and other raptors as they recovered from the organochlorine pesticide impacts of earlier years. For example, the Strait of Georgia in British Columbia is a major autumn staging site for Western Sandpipers Calidris mauri on southward migration. Birds trapped on the extensive mudflats of the Fraser estuary were significantly heaver (by 10%) than others caught on the small mudflat of Sidney Island (Ydenberg et al. 2002). The weight difference could not be attributed to seasonal timing, age or sex effects, but was linked with vulnerability to predation. The open expanse of the Fraser estuary offered safety from avian predators, but a lower fattening rate, while the small Sidney Island was more dangerous, but offered a higher fattening rate. The inference was that sandpipers arriving in the Strait with little fat (and hence more rapid escape responses) chose to take advantage of the high feeding rate at small dangerous sites like Sidney Island, whereas individuals encumbered by higher fat reserves elected to feed in larger but safer sites such as the Fraser estuary. Large, open sites are safer because they make it difficult for raptors to approach undetected, giving the shorebird prey earlier warning and longer escape times. From 1985, as Peregrine numbers increased, average migratory body mass and stopover durations of Western Sandpipers Calidris mauri at Sidney Island fell steadily (Ydenberg 2004). An accompanying steep decline in sandpiper numbers at Sidney Island was accounted for by shortening stopovers (mean 8.4 days, falling to 2.7 days), rather than by fewer individuals using the site. Under greater danger from predation, these birds seem to have switched from a long stay/high fuelling strategy at this site to a short stay/low fuelling strategy, using only safer sites for further weight gain. The authors suggested that such behavioural adjustments could be widespread among shorebird species, and that predation could be a major factor shaping the migratory routes, timing and behaviour of shore-birds (Lank et al. 2003a).
Predation at stopover sites is not the only form of predation endured by migrants. Because many falcons hunt high above the ground, beyond the limits of human vision, predation on high-flying diurnal migrants could easily be underestimated. In many regions, Peregrines Falco peregrinus and other bird-eating falcons are so widespread that day-flying migrants could be under continual risk of attack. In addition, falcons and accipiters are frequently seen hunting from oil rigs and ships at sea, taking passing migrants (e.g. Ellis et al. 1994), and large gulls are often seen chasing and catching tired migrants arriving low over water (Macdonald & Mason 1973, Riddiford 1978). Moreover, in the Mediterranean-North African region, Eleonora's Falcons Falco eleonorae breed during the autumn migration season and specialise on passage birds caught by day while on migration (Chapter 24). In the deserts of the Middle East, Sooty Falcons Falco concolor fill a similar role but, being smaller than Eleonora's Falcon, they take a narrower range of prey species. By flying at night, migrants can avoid this onslaught for part of their journey (owls not being known to hunt high-flying migrants), although in the Mediterranean region some birds may be taken by bats (Chapter 24).
Parasites are also likely to affect the migratory performance of birds, not only because their effects can be debilitating, but also because, when abundant (especially gut parasites), they can absorb a substantial part of the host's food intake. Their effects are tantamount to lowering the feeding rates of their hosts. Birds might pick up parasites at stopover sites that affect them later in the year. Migratory birds have sometimes been found to contain a greater range of parasites, such as haematozoa, than closely related resident species, a difference attributed to the exposure of migrants, as they pass through different regions, to a wider range of parasite species and their vectors (Bennett & Fallis 1960, Figuerola & Green 2000). Correspondingly, organs concerned with immune defence were found to be larger in migratory than in closely related non-migratory species (Moller & Eritzoe 1998). Many birds caught on migration have been found to harbour ticks, which can pass readily from bird to bird, and act as both reservoirs and vectors of pathogens. Interspecies transfer of avian haemosporidian parasites (Haemoproteus and Plasmodium) has been found to occur between resident and migratory species wintering in Africa (Waldenstrom et al. 2002), and the reactivation of latent Borrelia infections among Redwings Turdus iliaca was attributed to the stress of autumn migration (Gylfe et al. 2000). Similarly, in some North American species sampled on migration, individuals with blood parasites (mainly Haemoproteus and Plasmodium) had lower body weights or fat levels than unin-fected ones (Garvin et al. 2006). These various parasites do not necessarily kill their hosts, but may do so in particular circumstances, or reduce their breeding success, and some of their effects could therefore be regarded as additional costs of migration. Migratory birds may also help to disperse disease organisms, transferring them from one region to another to infect different populations (Ricklefs et al. 2005). Avian flu (strain H5N1) is a topical example (Olsen et al. 2006), but others have included Lyme disease (Scott et al. 2001) and West Nile virus (Rappole et al. 2000, Owen et al. 2006).
Another known source of mass mortality among migratory waterbirds is botulism, caused by a neurotoxin produced by the bacillus Clostridium botulinum. This anaerobic bacterium grows well on rotting organic matter in shallow stagnant waters or mud during warm weather, and occurs in most parts of the world. Affected birds become paralysed and limp, and die from respiratory failure or drowning, but may otherwise appear in good condition (Locke & Friend 1987). Year-to-year losses from botulism are highly variable, but can be spectacular, affecting migratory waterfowl at moulting and stopover sites, especially in western North America. An estimated one million birds died from botulism at a lake in Oregon in 1925, 1-3 million at Great Salt Lake in Utah in 1929, and 250 000 at the northern end of Great Salt Lake in 1932 (Jensen & Williams 1964). Many other outbreaks have involved smaller numbers of birds, but totalled over wide areas or over a period of years they can be substantial, for example the 4-5 million waterfowl deaths attributed to botulism in the western USA in 1952 (Smith 1975, Locke & Friend 1987), or the 1.5 million in California alone during 1954-1970 (Hunter et al. 1970).
Although botulism is not infectious, its effects are exacerbated by droughts which cause waterbirds to concentrate in larger than usual numbers on remaining shallow wetlands. The birds become poisoned when they ingest toxin-producing bacteria in the bodies of invertebrates that form their food, including the maggots living in rotting vegetation and carcasses. Dead waterfowl themselves become host to more maggots and perpetuate the outbreak. Mortality ends when the weather cools, when the birds leave the site or switch to other foods, when flies stop breeding, or when water levels rise (Wobeser 1981). Although botulism hits waterfowl seriously only in relatively arid areas, and only in occasional years, its impacts can be so substantial that affected populations could take several years to recover from each serious outbreak, constituting 'the greatest drain upon western waterfowl due to any single natural agency' (Kalmbach & Gunderson 1934).
Various other bacterial and other disease agents and neurotoxins have also caused occasional mass mortality incidents among migrating waterbirds concentrated at stopover sites (for review, see Newton 1998b). Examples include the erysipelas that killed 5000 Eared (Black-necked) Grebes Podiceps nigricollis on Great Salt Lake in 1975, and the streptococcal infection that killed another 7500 grebes there in 1977 (Jensen & Cotter 1976, Jensen 1979). Another more recent example was an outbreak of the H5N1 strain of avian influenza that killed many waterfowl, including rare Bar-headed Geese Anser indicus, on Qinghaihu Lake in China in 2005 (Liu et al. 2005). Influenza viruses are found in a wide range of bird species, especially aquatic ones, including ducks, geese and swans, and gulls, terns and waders. The viruses are excreted into the water, so can easily pass from bird to bird. Some bird species are likely to be little affected, and so act as reservoirs and transmitters of the virus, while others are extremely vulnerable and quickly die.
As with predation, parasitism has been suggested as a selective factor influencing the timing of migration, the habitats used, and the routes taken. In particular, by migrating long distances, many wader species are able to remain year-round in habitats (such as arctic tundra and lower latitude coastlines) which have relatively few parasites (Piersma 1997). As yet, however, the evidence for migration-related effects of either predation or parasitism on the migratory timing, routes and strategies of birds is little more than suggestive.
Was this article helpful?