Significant intra-annual variability in krill distribution and abundance at South Georgia revealed by multiple acoustic surveys during 2000/01

Three separate acoustic surveys of Antarctic krill (Euphausia superba) were conducted around South Georgia in the 2000/01 season: in October 2000 (early-season); during late December/early January 2000/01 (mid-season), and in March 2001 (late-season). The surveys were the first in a newly planned five-year series of observations designed to complement and extend an existing time series regularly maintained by the British Antarctic Survey since 1996 (and on a more ad hoc basis since the early 1980s). We hoped that conducting several surveys in one season would provide information on short-term variability that could be used to set data from more restricted once-per-season 'snap-shot' cruises in a broader context. The early- and late-season surveys were associated with logistic support voyages to South Georgia and were restricted to four transects within a box to the northwest of South Georgia. The dedicated mid-season survey covered that box in more detail (twice as many transects) and, in addition, examined boxes to the north, northeast and southwest of the island. Together these surveys provided temporally and spatially extensive coverage around South Georgia. Krill density in the western box in the early-season survey was very low (3.5 g m(-2)) but rose significantly (P = 0.020) by mid-season (to 34.7 g m(-2)). In a pattern that is consistent with observations from previous years, krill density in the western box in the mid-season survey was less than that in the eastern box (80.4 g m(-2)). Analysis of transect data obtained in the western box in the mid-season survey revealed no significant difference in the mean krill density derived from only those four transects surveyed early- and late-season, or from the full eight transects. Our first occupation of a survey box off the central north coast of South Georgia in mid-season revealed a krill density of 47.2 g m(-2) that was intermediate between the eastern and western areas. The size structure of the krill in the central region also reflected a mix of those to the east (generally small) and west (generally large). Krill density to the southwest of South Georgia was 32.1 g m(-2) in mid-season. By March, krill density in the western survey area had fallen significantly (P = 0.037) from the mid-season high to 7.8 g m(-2). Our multiple surveys at South Georgia have revealed major intra-annual changes in krill density at the island and have shown that the timing of the acoustic survey can affect significantly the estimate of krill density. The multiple estimates of krill density will allow reproductive performance indices for top level predators to be compared to prey availability at more appropriate time-scales than previous single 'snap-shot' acoustic survey data have allowed. This is a crucial step in the elucidation of response functions of dependent species to changes in krill abundance, and could be a useful contribution to ecosystem management