Is food availability limiting African Penguins Spheniscus demersus at Boulders? A comparison of foraging effort at mainland and island colonies

The African Penguin Spheniscus demersus (Vulnerable) formed three new colonies during the 1980s, two on the South African mainland (Stony Point and Boulders) and one on Robben Island. One of the mainland colonies, at Boulders, Simon's Town, is in a suburban area, resulting in conflict with humans. Growth of the Boulders colony was initially rapid, largely through immigration, but has since slowed, possibly as a result of density‐dependent effects either on land (where there has been active management to limit the spread of the colony) or at sea. We test the latter hypothesis by comparing the foraging effort of Penguins feeding small chicks at island and mainland sites, and relate this to the foraging area available to birds. Three‐dimensional foraging paths of African Penguins were reconstructed using GPS and time–depth loggers. There were no intercolony differences in the rate at which birds dived during the day (33 dives/h), in diving depths (mean 17 m, max. 69 m) or in travelling speeds. The maximum speed recorded was 2.85 m/s, with birds travelling faster when commuting (average 1.18 m/s) than when foraging (0.93 m/s) or resting at sea (0.66 m/s during the day, 0.41 m/s at night). There were strong correlations between foraging trip duration, foraging range and total distance travelled. Foraging effort was correlated with chick age at Robben Island, but not at Boulders. Contrary to Ashmole's hypothesis, birds from Boulders (c. 1000 pairs) travelled further (46–53 km) and foraged for longer (13.2 h) than did birds from Robben Island (c. 7000 pairs) and Dassen Island (c. 21 000 pairs) (33 km, 10.3 h). The mean foraging range also differed significantly between mainland (18–20 km) and island colonies (9 km). The area available to central‐place‐foraging seabirds breeding on the mainland is typically less than that for seabirds breeding on islands, but the greater foraging range of Boulders birds results in an absolute foraging area roughly twice that of island colonies, and the area per pair is an order of magnitude greater for the relatively small Boulders colony. Ashmole's hypothesis assumes relatively uniform prey availability among colonies, but our results suggest this does not apply in this case. The greater foraging effort of Boulders birds probably reflects reduced prey availability in False Bay, and thus the recent slowing in growth at the colony may be the result of differential immigration rather than management actions to limit the spatial growth of the colony.