Global Circumnavigations: Tracking Year-Round Ranges of Nonbreeding Albatrosses

See allHide authors and affiliations

Science  14 Jan 2005:
Vol. 307, Issue 5707, pp. 249-250
DOI: 10.1126/science.1106042


Although albatrosses are paradigms of oceanic specialization, their foraging areas and migration routes when not breeding remain essentially unknown. Our continuous remote tracking of 22 adult gray-headed albatrosses for over 30 bird-years reveals three distinct strategies: (i) Stay in breeding home range; (ii) make return migrations to a specific area of the southwest Indian Ocean; and (iii) make one or more global circumnavigations (the fastest in just 46 days). The consistencies in patterns, routes, and timings offer the first hope of identifying areas of critical habitat for nonbreeding albatrosses, wherein appropriate management of longline fisheries might alleviate the plight of the world's most threatened family of birds.

Albatrosses are the world's most threatened family of birds (with 19 of 21 taxa on the International Union for the Conservation of Nature Red List). Recent catastrophic population decreases are mainly a consequence of incidental mortality in longline and trawl fisheries (1). Knowledge of the at-sea distribution of albatrosses is thus critical to their conservation, yet few data from birds of known status and provenance are available outside the breeding season. Regular successive recoveries have been made off eastern Australia of wandering albatrosses (Diomedea exulans) banded at South Georgia (54°00′S, 38°03′W), showing that some birds migrate great distances, but more recent tracking data from the same species breeding in the Indian Ocean indicate that this population is much more sedentary during the nonbreeding period (24). Satellite tracking of postbreeding birds of other species has confirmed several long-distance migrations (5, 6), but the brevity of these studies and the small samples mean that range and behavior during the nonbreeding season is, by comparison, very little known.

We deployed leg-mounted light level loggers (7) on 47 gray-headed albatrosses (Thalassarche chrystostoma) that were rearing chicks at the end of the austral summer (April 1999) at Bird Island, South Georgia. This species breeds biennially when successful, and 35 of the loggers were retrieved between September and November 2000, of which 22 provided complete data throughout the approximately 18-month interval between breeding events (corresponding to 11,034 bird-days) and 13 failed to download. Subsequent data processing provided two locations per day with a probable mean accuracy ± SD of 186 ± 114 km (7), apart from a variable period around the equinoxes when latitude estimates are unreliable.

Three overall migration strategies were identified: (i) residence in the southwest Atlantic and adjacent areas in an extended version of the breeding season home range (5, 8), also indicated by previous sightings at the colony of banded birds in their sabbatical year; (ii) return migrations to winter in habitats known to be used by other albatrosses in the Indian Ocean; and (iii) one or more global circumnavigations with foraging in areas and habitats used in options (i) and (ii) and also in additional staging areas in the Indian and Pacific Oceans (Fig. 1 and fig. S1). Investigated in more detail, these broad strategies apparently include different combinations of a limited number of consistent seasonal migration patterns [supporting online material (SOM) text]. There was a degree of sexual segregation: Females were much more likely to show a restricted range, whereas most males performed at least one circumpolar navigation.

Fig. 1.

(A to D) Representative migration routes of four gray-headed albatrosses in the 18 months between breeding attempts (SOM text). White, first winter; red, summer; yellow, second winter. Dotted lines link locations obtained before and after the equinox.

We recorded 15 round-the-world journeys by 12 birds (3 of which made double journeys), all in an easterly direction. Any bird that moved beyond the Kerguelen archipelago (49°20′S, 70°20′E) completed such a trip, this area being the eastward limit for a simple return migration to and from South Georgia. There was high consistency in both the timing and choice of route, especially in winter. These circumnavigations included some extraordinary examples of albatross flight performance. Typical journeys from South Georgia to the southwest Indian Ocean took 6.2 days at 950 km per day; the leg to the southwest Pacific lasted 13.2 days at 950 km per day, and the last leg back to South Georgia 10.3 days at 750 km per day. Without stopping, a complete circumnavigation of the Southern Ocean could, in theory, be completed in 30 days; this provides a context for the exceptional performance of the bird that achieved this in just 46 days.

Our results show that (i) there were at least three distinct migration strategies; (ii) individuals consistently exploited the same general staging areas in succeeding winters; (iii) despite the apparent flexibility inherent in a nonbreeding season that lasts 18 months, the timing of migratory journeys was generally well synchronized and of consistent duration, and the routes were typically very similar both at the individual level and within strategies; and (iv) all individuals used the breeding season home range in the Atlantic for at least part of the nonbreeding summer.

Very little is known about environmental, hormonal, or physiological factors influencing the selection of different migration patterns, routes, and staging areas. However, the apparent sexual segregation could result from niche divergence mediated by differences in flight performance (8). In addition, the observed spatial and temporal consistencies suggest that correct timing of events is particularly important, although the cues that trigger movements are unknown. Although all staging areas coincide with habitats either known or likely (given bathymetric and hydrographic features) to be of importance to albatrosses (fig. S1), there are indications that South Georgia birds may use these areas when local albatrosses (especially conspecifics and congeners) are not breeding (and therefore potentially absent). Similarly, tracking studies during the breeding season suggest little or no spatial overlap between adjacent colonies or populations (9, 10).

The relatively small size of the key staging areas and restricted amplitude of the migration routes suggests that it may be possible to identify critical habitats. This might help mitigate interactions with threatening processes (particularly fisheries) and also to identify biodiversity hot spots in pelagic marine systems. Indeed, compared with congeners, the gray-headed albatross has a greater propensity for aggregation at frontal systems and other mesoscale ocean-ographic features, rather than in shelf waters (4), which makes it a potentially valuable indicator species for biodiversity hot spots in pelagic habitats. In comparing its year-round distribution (fig. S1) with data on tuna longline fishing effort (11), only in the southwest Indian Ocean does a core albatross staging area coincide with one of the five or six principal fishing areas exploited over the last decade. However, their migration routes traverse most of the key tuna fishing areas south of 30°S, with the exception of the Tasman Sea. This reinforces the message that protecting albatross and petrel species requires appropriate mitigation measures to be used in longline fisheries throughout all oceans south of 30°S.

Supporting Online Material

SOM Text

Fig. S1


References and Notes

View Abstract

Stay Connected to Science

Navigate This Article