Table 1. Summary of factors affecting prey availability in the southeastern Bering Sea shelf and basin, observed effects on predators and prey, and expected responses of black-legged kittiwakes breeding at the Pribilof islands (St. Paul & St. George).
| Factors | Effects | Expectations |
| 1. Habitat | ||
| 1.1 Shelf: Food webs affected by cold and warm regimes (1) | In cold years: -Lower abundance and dispersed distribution of juvenile pollock and capelin compared to warm years (2) -Low food availability for piscivorous seabirds (3) | During three cold years (2008-10): - Low biomass of juvenile pollock on shelf (Y) & occurrence of shelf-based prey in bird diets (Y) - High foraging effort and nutritional stress incurred by kittiwakes (Y) |
| 1.2 Basin: Food webs influenced by eddy currents near the Pribilof Islands | - Concentrations of mesopelagic prey increase in areas of high eddy kinetic energy (EKE; 4, 5, 6) | During high-EKE years: - Kittiwakes forage near eddies (Y) - Higher prevalence of oceanic prey (e.g., myctophids) in diets (Y) - Higher incidence of kittiwake foraging trips to the basin (Y) - Higher kittiwake fledging success (Y) |
| 2. Colony | ||
| 2.1 Proximity of seabird colony from alternative foraging habitats in Basin (St. Paul > St. George) | - Lower ability to cope with food shortages in vicinity of St. Paul (7, 8) | St. Paul compared to St. George: - Higher kittiwake usage of shelf habitats (Y) - Longer foraging trips to basin (Y) - Higher nutritional stress (P) - Lower fledging success (N) |
| 2.2 Colony size and population size of red-legged kittiwakes (main prey is myctophids) (St. George > St. Paul) | - Higher density-dependent feedback on seabird populations on St. George (9) | St. George compared to St. Paul: - Longer trips to shelf & basin (N) - Lower myctophid occurrences in diets (N) - Higher nutritional stress (N) |
Support for expectations shown in parentheses (Y = yes, N = no, P = partial).
Stabeno et al. 2012 (1); Hollowed et al. 2012 (2); Satterthwaite et al. 2012 (3); Zainuddin et al. 2006 (4), Drazen et al. 2011 (5); Dell et al. 2011(6); Byrd et al. 2008 (7); Paredes et al. 2012 (8), Hunt 1986 (9).