Seabirds linking Arctic and ocean

Throughout their annual cycle, migratory animals depend on widely separated areas for reproduction, fuelling migration, moulting and wintering. By migrating, animals link these areas. As conditions in one area can affect the behaviour or state of an individual in a next phase of the annual cycle, knowing the area use of migrants is important for understanding potential drivers of individual performance and population dynamics. For many species, however, basic information on their migration route, stopovers and wintering areas is lacking.

Seabirds are well-represented among animals with the longest and most spectacular migrations. Only a few decades ago much of what was known about seabird migrations was based on land-based sightings or anecdotal at-sea observations, providing only a fragmentary and incomplete picture. Since the development and miniaturization of tracking devices, in particular light-based geolocators, details of seabird migrations are revealed at a fast pace. In this thesis, the migrations of a group of long-distance migratory seabirds breeding in Arctic tundra areas around the North-east Atlantic was studied: Red-necked Phalarope Phalaropus lobatus, Grey Phalarope P fulicarius, Long-tailed Skua Stercorarius longicaudus and Arctic Skua S parasiticus. The main aims are to describe the non-breeding movements of each study species, reveal variation within individuals in migration routes, wintering areas, annual cycles and movement strategies relative to variation between individuals and populations, and link non-breeding movements with other activities in the annual cycle. In international collaborations, individuals were tracked – in some cases over up to five years – using light level geolocators from breeding sites between North-east Greenland and West Siberia.

In chapter 2-3, Red-necked Phalaropes from Fennoscandian and Russian breeding areas were confirmed to migrate to the Arabian Sea in the Indian Ocean – a more or less two-step migration of ca 6000 km, involving a prolonged stopover at around 45°N during both autumn and spring migration. In contrast, Red-necked Phalaropes breeding in Greenland, Iceland and Scotland migrated ca 9000 km westwards to the northern Humboldt Current in the eastern Pacific. The longer migration to the eastern Pacific was associated with longer wings; more pointed wings may reduce flight costs on long migrations. Red-necked Phalaropes following these two flyways face contrasts in availability of saline stopover habitat en route and also in the spatio-temporal variability in upwelling conditions at the two wintering areas, which corresponded to differences in migration strategies as well as wintering movement strategies. Birds wintering in the eastern Pacific migrated generally in more steps and – in spring – at higher speed than those wintering in the Arabian Sea. Moreover, birds wintering in the eastern Pacific stayed at roughly a single site (residency), whereas birds wintering in the Arabian Sea moved considerable distances between sites (itinerancy).

In chapter 4, Grey Phalaropes from North-east Greenland and Iceland were shown to migrate to three broad wintering areas across a wide latitudinal range in the Atlantic. These wintering areas were associated with different autumn and spring migration routes and stopover sites, as well as different degrees of movements within each wintering areas. Interestingly, two out of six individuals with more than one year of data switched from one year to the next between wintering areas ca 6500 km apart, indicating considerable individual flexibility.

Moult (feather replacement) is important for maintaining the main functions of the plumage: insulation and flight. Moult is also energetically demanding and therefore, overlap with other demanding activities, such as breeding and migration, is usually avoided. In chapter 5, start date and duration of primary moult was studied in the four skua species breeding in the northern hemisphere. Among the four skua species, moult was finished at the start of spring migration. Therefore, longer moult durations of the larger species required an earlier start, resulting in temporal overlap with autumn migration. In the most extreme, moult overlapped with the entire autumn migration in Great Skuas. Excluding Great Skuas, the first moult cycle (in 2nd calender-year birds) lasted longer than later moult cycles, which can be attributed to migration limiting the time available for moult in 3rd calendar-year and older birds, but not in 2nd calender-year birds.

Tracking individuals for multiple years allows the investigation of individual consistency in routes and area use. Using geolocator data of Long-tailed Skuas breeding in Greenland and Svalbard (subspecies S l pallescens) and those breeding in Scandinavia (subspecies S l longicaudus), individual consistency in non-breeding movements was quantified in chapter 6. Individuals of both subspecies migrated via a stopover area in the central North Atlantic to winter mainly off South-west Africa, with some individuals venturing into the Indian Ocean. Most individuals closely followed previous years’ movement patterns, but during the wintering period, up to ca 20% of the individuals deviated from earlier routes more than 1000 km. Remarkably, one individual switched mid-winter from the Benguela Current to the Falkland Current in two out of four years, using two distinct itineraries that were both repeated in later years. These results show flexibility in non-breeding movements during a part of the annual cycle, and strongly suggests extensive spatial memory of individual Long-tailed Skuas.

Relying for reproduction on strongly cyclic rodent population sizes, Long-tailed Skuas face large annual variability in feeding conditions upon arrival at the breeding grounds. To survive unfavourable feeding conditions upon arrival at the breeding grounds or to start egg laying soon after arrival, migrants can bring body stores deposited on the wintering grounds or at staging areas along the migration route. In chapter 7, the contributions of distantly-acquired, marine resources (‘capital’) and locally-acquired, terrestrial resources (‘income’) in egg production were investigated in Lapland, Sweden, during a ten-year period with varying rodent densities, using stable isotope data of adult, juvenile and chick down feathers. With higher rodent density, contributions of locally-acquired terrestrial resources were larger whereas the contribution of distantly-acquired marine resources remained constant. Long-tailed Skuas arrive in the breeding area with large body stores, which they likely build up during a prolonged stopover in the central North Atlantic.

Arctic Skuas breeding between Northeast Greenland and the Yamal peninsula (Russia), as shown in chapter 8, winter across the entire Atlantic, and also the Mediterranean and Arabian Seas. Within a single colony (Slettnes in northern Norway), the full range of wintering areas occurred, making this species ideal to quantify the relative effects of breeding and wintering area of the timing of migration and breeding. Beside a strong effect of breeding site on annual schedules, wintering area strongly affected the timing and duration of migration and the duration of the wintering period. Specifically, more distant wintering areas were associated with a substantially longer migration duration, a shorter wintering period, and an earlier departure to arrive in time on the breeding area. Breeding latitude not only shifted annual schedules but also affected the duration of phases in the annual cycle: spring migrations were faster and the time between arrival at the breeding grounds and clutch initiation shorter than at lower breeding latitudes. Whereas most Arctic Skuas wintering in the Atlantic used a spring stopover area in the central North Atlantic, time spent here was shorter for individuals that migrated later, with birds migrating from the Canary Current to breed on Svalbard skipping this stopover entirely. These shorter migrations and pre-laying periods suggest that birds breeding in the high Arctic carry body stores from the wintering areas across 60 latitudinal degrees.

Among the four study species, data on population size and trends are lacking or scant. For Arctic Skuas, population trends and their drivers are relatively well-studied in Scotland, but potential causes of the declines of Arctic Skuas have remained poorly studied elsewhere. In chapter 9, a population decline of ca 50% over the past two decades is documented for the largest European colony at Slettnes, northern Norway. During five recent study years (2014-2018), both bottom-up (food shortage) and top-down (predation) effects negatively affected the reproductive investment and hatching success in this colony. Food shortage in three out of five recent study years was suggested by a high percentage of one-egg clutches, small eggs and low adult female body mass. At the same time, nest predation by Red Fox Vulpes vulpes increased, leading to total breeding failures in recent years, even when food availability appeared good. Clearly, the reproductive output in the study years was far below levels required to sustain a stable population, even with a high adult survival probability.

In the concluding chapter 10, migration patterns and movements within the wintering areas are compared between the four study species. First, I show that the representativeness of the tracking datasets for the studied populations is generally high. By comparing movement behaviour across species and wintering areas, I show how the degree of itinerancy decrease and individual consistency increases with primary productivity and seasonal spatial stability of primary productivity of wintering areas. In the final part of the general discussion, I suggest potential directions for future studies and discuss threats and conservation of seabirds during the non-breeding season, including the question whether the studied seabirds will be able to adjust their non-breeding movements to rapid environmental changes. Considering differences between species in large scale-migration patterns, degree of migratory connectivity and individual flexibility, I argue there is little scope for adjustment in the Red-necked Phalarope, but more scope for adjustment of migration routes and wintering areas to rapid environmental change via developmental plasticity or individual flexibility in the Grey Phalarope, Long-tailed Skua and Arctic Skua.