Migratory birds have the capacity to migrate very long distances between High Arctic breeding areas and more favourable wintering areas in temperate and tropical regions. How birds navigate on these long distance journeys is one of the most fascinating questions to be solved. Today we know a great deal about the compass information that migratory birds can use for orientation. They can select a migratory course with the aid of the sun, the related pattern of skylight polarisation, or the stars and the Earth’s magnetic field (for reviews see, for example, Emlen 1975, Able 1980, Wiltschko and Wiltschko 1995). Many songbirds migrate completely alone between breeding and wintering areas. To be able to find its way the young bird is born with a program including the direction and time it should migrate to reach its species-specific wintering area (Berthold 1996). Adult birds, on the other hand, seem to return to known stop-over sites and many of them use the same wintering and breeding areas year after year (e. g. Perdeck 1958). We know very little about the information they use to locate these geographical sites. One possibility that has been suggested for sea turtles is that they might use a bi-coordinate map based on two components of the Earth’s magnetic field, i.e. the total field intensity and the angle of inclination, to locate their geographic position relative to a goal (Lohmann and Lohmann 1996, see also Åkesson 1996). Experiments with hatchling loggerhead sea turtles suggest that they have an inherited ability to select new courses during migration that is triggered by exposure to both of these parameters of the geomagnetic field. These course shifts coincide with geographic positions where the shifts are relevant to keep the animals within a circular path around the Sargasso Sea in the North Atlantic Ocean (Lohmann and Lohmann 1996). This is an interesting possibility for long-distance navigation, which might also work for birds. However, in other areas of the globe the distribution of geomagnetic field lines might not be as favourable for bicoordinate navigation (cf. Åkesson and Alerstam 1998).

We were interested in investigating if migratory birds are able to use information from the Earth’s magnetic field to locate their geographic position. In particular we studied how the orientation system differs between young inexperienced migrants compared with older ones that have completed at least one migration between the breeding and wintering areas. To study this we performed a long-distance displacement experiment with young and adult migrants during which we registered the birds’ orientation in circular orientation funnels at different sites along the route of displacement. We were also interested in studying the capability of the birds to select a migration course close to and at the geomagnetic North Pole. At this site the magnetic compass of birds, based on the angle of inclination, does not function because the magnetic field lines are completely vertical (inclination +90°; cf. Wiltschko and Wiltschko 1972). Due to the steep angles of inclination, the varying angles of declination, and the rapid passage of time zones during migratory flights, orientation is very complicated at high geographic latitudes (e.g. Alerstam et al. 1990). Therefore, it is of great interest to investigate how birds on migration cope with these difficulties. To do this we have studied the birds’ orientation behaviour in circular cages under as natural conditions as possible: under natural clear skies and when exposed to only the natural geomagnetic field.

By comparing the course selected at each site with available compass cues and the physical character of, for example, the geomagnetic field at the particular site we will be able to analyse:

1. If adult birds might be able to make corrections for displacement away from the normal migratory route and if young birds lack this ability.
2. If earth’s magnetic field might be involved in the birds’ navigation system at these high geomagnetic and geographic latitudes.
3. At what angle of inclination the birds’ magnetic compass fails to give a correct migratory course and in particular if birds are at all capable of selecting a migratory course at the geomagnetic North Pole.

Field-work and methods

We captured a group of 30 white-crowned sparrows Zonitrichia leucophrys in Inuvik (68°21’N, 133°43’W), North-western Territories in Canada at the end of the breeding period (15–31 July 1999). Half of the birds were juveniles and the other half adults (identified according to Pyle et al. 1987; cf.). The birds were kept in cages during the following autumn migration period, 1 August to 2 September 1999 (Leg 2), and transported by the icebreaker Louis S. St-Laurent eastwards and northwards away from their normal breeding area in north-western Canada. We used a similar method during a previous expedition to Russia 1994, studying the orientation of displaced young wheatears (Oenanthe oenanthe, Åkesson et al. 1995 a and b). In addition to the displaced white-crowned sparrows, we captured 15 individuals that were kept in the breeding area in Inuvik, at the Aurora Research Institute, and for which the migratory orientation was registered during the same autumn migration period. White-crowned sparrows breed in northern mainland Canada and winter in the southern United States (Chilton et al. 1995). The population of white-crowned sparrows breeding in Inuvik spends the winter in the south-western United States, resulting in an expected autumn migration course in a geographic south to south-east direction. To reach their wintering areas, white-crowned sparrows migrate alone at night (Chilton et al. 1995).

On board the icebreaker we kept the birds in separate cages in an enclosed container with two windows. The birds were exposed to the natural light regime of the region as we moved and they received food and water ad libitum. At each site we transported them ashore by helicopter and kept them in their cages in a separate tent on the tundra. At midnight (at sites where the sun never sets at this time of the year), or soon after sunset (at sites where the sun went below the horizon), we registered the birds migratory activity in circular orientation cages, so-called Emlen-funnels (Emlen and Emlen 1966), lined with Tipp-Ex© paper. The birds’ jumping activity in the cages, registered as claw scrapes in the Tipp-Ex© pigment, is correlated with the intended migratory direction (Emlen and Emlen 1966). We used diffusing Plexiglas sheets to cover the cages in order to simulate overcast and exclude celestial information (the sun, the pattern of skylight polarisation, and stars) for compass orientation.

Results

We were able to perform orientation cage experiments at 8 out of 9 sites visited during the second half (Leg 2) of the Tundra Northwest 1999 expedition between 22 August and 2 September. At site 11, east of Tuktoyaktuk, we could not perform experiments due to high wind speeds. At all other sites we were able to register the birds’ orientation in one or two experiments under clear and/or simulated overcast skies. At site 17, on north-east Baffin Island, however, snow started to fall when we had initiated the experiments and the activity of the birds in the cages was extremely low. At site 18, near Iqaluit on Baffin Island, the birds were released at night on completion of the orientation cage experiments. We tried to release the birds with small light sticks glued to one of their central tail feathers in order to register the direction of their migratory flight at departure (for method see, Ottosson et al. 1990). However, our white-crowned sparrows were not willing to take off on migratory flights although the weather was clear and wind speed low, and the majority of them landed on the ground nearby. Similar behaviour was observed as regards releases of the same species in the breeding areas in Inuvik, but in the case of savannah sparrows not to the same degree. This release method seems to work well for some species allowing registration of migratory flight direction at departure (e. g. Ottosson et al. 1990, Sandberg et al. 1991), while other migratory birds appear less motivated to depart on migratory flights when released under similar conditions (white crowned sparrows in this study).

Previous experiments with High Arctic breeding passerines have shown their capacity for magnetic compass orientation in magnetic fields with very steep angles of inclination, such as +81 °(wheatears, Sandberg et al. 1991; inclination + 79°, Åkesson et al. 1995) and +89° (snow buntings, Plectrophenax nivalis, Sandberg et al. 1998). In our experiments we were able to study the birds’ magnetic compass orientation, i.e. without celestial compass information, in fields with angles of inclination of +83°, +87°, +88°, +89° and +90°. In addition to the displaced birds, we have studied the orientation of white-crowned sparrows in Inuvik where the angle of inclination is +82°. In two of the conditions, +83° and +88° on north-east Baffin Island (site 17) and Ellesmere Island (site 15) respectively, the orientation experiments were performed in light snowfall or at a different time of the day which resulted in low activity. Therefore, it might be difficult to compare orientation under overcast skies from these two sites with the other sites.

The strength of our experiments is that we can compare the orientation behaviour of the same individual in different geomagnetic conditions and when it is exposed to the Earth’s natural magnetic field. Previous studies at high latitudes have investigated the magnetic compass orientation at one single site at a time and in a magnetic field with one angle of inclination (Sandberg et al. 1991, 1998). Experiments in steep angles of inclination have also been performed in artificial fields in the laboratory (references given in Wiltschko and Wiltschko 1995). Our experiments will be important in revealing the compasses migratory birds use for orientation and navigation at these high geomagnetic and geographic latitudes.