PATTERNS OF COLONY ATTENDANCE IN THE CAPE PETREL DAPTION-CAPENSE AT NELSON ISLAND, SOUTH-SHETLAND ISLANDS, ANTARCTICA

The annual cycle of colony attendance of the Cape Petrel Daption capen se at Nelson Island, Antarctica, covered about 9 months from the start of September until the end of May. Large numbers of birds attended th e colony continuously for 6 months from arrival until hedging, except for a 2-week incomplete pre-laying exodus when attendance was reduced to 10% of that during incubation. After 1 month (March) of post-hedgin g absence, birds returned to the colony for 2 months in numbers compar able to the level in the breeding season. This was followed by a winte r absence that lasted only 3 months and was interrupted by two short v isits by a small number of birds. Mean attendance was highest during i ncubation, but short peaks of maximum attendance occurred during the p re-breeding period. Except for the incubation and brooding periods, co lony attendance showed high variability both within and between days. Large numbers of visiting birds occurred on the sea and on islets in f ront of the colony until the end of October. Unemployed birds constitu ted an almost constant proportion of up to 50% of colony-attending bir ds during the incubation period. Diurnal variability did not show a co nsistent pattern but may represent preferential feeding during early m orning hours, with additional short departures depending on the length of the daylight period. Colony counts were the most consistent during the second half of incubation, when the lowest counts gave an estimat e of the number of breeding pairs and peak numbers an impression of th e additional number of attending nonbreeders. Pre-breeding counts reve aled the maximum numbers of birds associated with a particular colony. However, for both pre- and post-breeding surveys, several counts a da y over a period of at least a week are recommended because of variabil ity in attendance.