Utilisation of the oceanic habitat by king penguins over the annual cycle

The distribution and behaviour of foraging seabirds depend on the physical features of the ocean at different time and space scales, but little is kno wn for penguins. We investigated the foraging behaviour of king penguins in relation to oceanographic features over the birds' complete annual cycle. A total of 44 birds was followed between 1994 and 1997 at the Crozet Island s to monitor foraging habitat, diving behaviour, and sea temperature of the water column, using satellite-tracking and time-temperature-depth recorder s (TDR) carried by the birds, The study included breeding in summer, the wi nter period of chick raising, and the post-moult period in spring. King pen guins foraged in 2 specific regions in response to the seasonal changes in local prey availability. In summer, satellite-tracked birds during the incu bating and brooding stages (n = 14) preferentially exploited the polar fron t located 340 to 450 km to the south of their breeding site. TDR-equipped b irds (n = 12) also foraged at the polar front in summer as indicated by the vertical temperature profiles. In autumn and winter, satellite tracks (n = 8) and sea temperature measurements of TDR-equipped birds (n = 8) showed t hat birds with creching chicks instead foraged in antarctic waters, with 70 % of individuals reaching the latitude of the pack-ice limit (1600 km from the colony). This suggests better prey availability than in the polar front al zone at that time. When the birds were at the latitude of the polar fron t, the thickness of the surface mixed layer (SML) ranged from 80 m in summe r to 140 m in winter, the SML temperature was similar to4 degreesC and the thermocline had a mean maximum gradient of similar to0.5 degreesC for each 10 m depth. When the birds were at their most southerly position, the depth of the SML ranged from 100 m in autumn to similar to 150 m in winter, whil e its temperature ranged from -0.8 to 2 degreesC. The temperature gradient of the thermocline showed an inversion in autumn, and this gradient was pos itive in winter (mean maximum gradient of 0.3 degreesC for each 10 m depth) . Except for spring birds (n = 4) and for 1 winter bird, where the SML exce eded the diving range, all TDR-equipped penguins (n = 19) dived preferentia lly in and below the depth of the thermocline, thereby minimising diving in the SML. Therefore, their prey may have been predictably concentrated belo w the SML through oceanographic processes. In king penguins, the strategy o f going south could then have evolved in relation to the thinning of the SM L towards the south at any time of the year.