In addition to characterizing prey remains, data from molecular markers have sometimes been used to identify the feeding grounds of individuals from different populations. In one study, the spatial foraging patterns of two species of bumble bees (Bombus terrestris and B. pascuorum) were analysed using genetic data (Chapman, Wang and Bourke, 2003). These species are common European bumble bees and both live in colonies headed by a single queen who mates only once; therefore, workers from each colony are full-siblings with a relatedness value of 0.75. When foraging, B. pascuorum specializes on flowers with deep corollae that accommodate the long-tongued workers, whereas the short-tongued workers of B. terrestris will visit a wider range of flower types. Workers in both species leave scent marks on flowers that they have visited, presumably to label these as having little remaining pollen, although whether this is for the benefit of themselves or their nest-mates was unknown prior to this study. The authors looked at the foraging patterns for both bumble bee species over three spatial scales in London. At the smallest scale, the workers that visited the same patches of flowers showed no relatedness to one another and therefore must have originated from multiple colonies. Since there should be little advantage to leaving scent marks for unrelated individuals, the marks are more likely to be for the benefit of the individual that is leaving them.
At the intermediate scale, the bumble bee visitors to entire sites were sampled (sites being defined as discrete areas containing numerous flower patches, such as parks or cemeteries). Because all the workers from each colony are full-sisters, the number of full-sisterhoods identified at a site represents the number of colonies that are visiting that site. An analysis of sibships revealed that an estimated 96 B. terrestris and 66 B. pascuorum colonies visited each site. Finally, over a larger spatial scale, total foraging distances were estimated for the workers of each species. The median foraging distance was estimated to be 0.62--2.8km for B. terrestris and 0.51-2.3 km for B. pascuorum, distances that exceed previous estimates based on mark--recapture or radar-tracking studies. These surprisingly long-distance foraging trips will have obvious implications for gene flow among populations of plants that are pollinated by bees.
The search for food is one of the most important reasons for migration, as illustrated by wolves (Canis lupus) in northwest Canada, which are migratory in the areas where their main prey is caribou because of their need to follow the seasonal movements of the herds. A microsatellite analysis of 491 wolves from nine regions in northwest Canada revealed an overall pattern of isolation by distance, but there was an unusually high level of genetic differentiation between wolf packs from either side of the Mackenzie River. Because this river is frozen for 6-8 months out of every year, it should not present an insurmountable physical barrier. The lack of dispersal across the river can, however, be explained by the behaviour of the resident caribou, which undertake an annual migration either side of the river in a north--south direction. In this case the migration patterns of the prey (caribou) seem to be directly influencing the migration patterns of the predators (wolves), because both travel in a north-south direction and neither undertake east--west migrations across the river (Carmichael et al, 2001).
Migration in search of food may also result in temporary 'populations' that consist of individuals that were born hundreds of kilometres apart. This is true of juvenile green turtles (Chelonia mydas), which congregate in foraging grounds that are far away from their nesting grounds. One aggregation frequently forms off the coast of Barbados, even though there are no suitable nesting sites nearby. Mitochondrial haplotypes identified from control region sequences were compared with genetic data previously obtained from nesting beaches to determine where these foraging turtles originated. The group turned out to be ofmixed stock, with 25 per cent originating in Ascension Island, 23 per cent from Aves Island/ Surinam, 19 per cent from Costa Rica, 18.5 per cent from Florida and 10.3 per cent from Mexico (Luke et al., 2004). A foraging group of immature loggerhead sea turtles (Caretta caretta) off Hutchinson Island, Florida, turned out to be less cosmopolitan. Sixty-nine per cent of the members of this group came from south Florida, 20 per cent from Mexico, and 10 per cent from north Florida-North Carolina (Witzell et al., 2002). Many sea turtle populations are declining, and these data highlight the importance of international cooperation to the future management of these species.
Some of the longest distances that are travelled in the search for food and other resources are found in migratory birds. Dunlin (Calidris alpina) breed in the Arctic and spend their winters in Europe, Africa and Asia. A comparison of mtDNA haplotypes from breeding, migrating and overwintering populations revealed that migration was occurring in a general north-south direction, with dunlin that breed in the western Palaearctic overwintering in the western part of their range (Portugal, Morocco), and populations that breed further east overwintering in the Middle East (Wennerberg, 2001).
In the case of dunlins it was possible to match haplotypes between breeding and overwintering grounds because populations were genetically structured, but
this approach has been less successful in some other species. Monarch butterflies (Danaus plexippus; Figure 6.15), for example, are well known for their migration between breeding sites across Canada and the USA and overwintering sites in California (western populations) and Mexico (eastern populations). Although the eastern and western populations do not interbreed, individuals from the two regions show almost no variation in mitochondrial DNA sequence and therefore haplotypes cannot be used to genetically track the Monarch migration routes (Brower and Boyce, 1991). Similarly, North American migratory songbirds tend to show little structuring of mtDNA haplotypes across broad spatial scales. Researchers studying the yellow-breasted chat (Icteria virens), the common yellowthroat (Geothlypis trichas) and the Nashville warbler (Vermivora rufica-pilla), all of which are long-distance migrants, were able to do little more than assign individuals to either eastern or western breeding lineages on the basis of mtDNA haplotypes, and as a result they gained little information about which breeding populations different migrants had originated from (Lovette, Clegg and Smith, 2004).
An alternative and more successful approach for tracking migratory individuals has been based on a combination of microsatellite data and stable isotopes in Wilson's warbler (Wilsonia pusilla; Clegg et al., 2003). The microsatellite data provided some evidence of genetic structure across the North American range of this species, although differentiation among western populations was negligible. The hydrogen isotope ratios from feathers collected on the breeding grounds provided information about the latitude at which the feathers were grown in the previous year. A combination of microsatellite (east--west differentiation) and isotope (north--south differentiation) data allowed researchers to conclude that birds from western Canada and the USA overwinter in Central America, birds that breed near the west coast overwinter in western Mexico, and birds that breed further inland and at higher elevations spend their winters in eastern Mexico. Tracking migration routes with these types of data is becoming an increasingly important conservation tool, since populations of many neotropical migrants are declining and conservation tactics must take into account the possibility of habitat loss at either the breeding or the overwintering ground.
Was this article helpful?