CCAMLR

Between September 1987 and February 1988, Antarctic fur seals Arctocephalus gazella at Heard Island fed mainly on fish with an average 95.2% of scats in monthly collections containing fish remains. Pelagic myctophids constituted more than 50% of fish taken by fur seals at the beginning of the summer season and again at the end when there was an influx of non-breeding male seals. During the middle period from October to December, fish from the surrounding shelf area comprised the bulk of the diet. These included various benthic nototheniid species, the bentho-pelagic ice fish Champsocephalus gunnari and skate (Bathyraja sp.), the latter being found in over 60%-of scats in October and November. The population of Antarctic fur seals at Heard Island is increasing at about the same rate as at South Georgia. There the increase is thought to be due to the high availability of krill E. superba, but, in the present study no euphausiid remains were found, so the increasing population at Heard Island has been supported on a local diet of fish. Whether this population increase can be sustained on a diet of fish is arguable. Trial fishing around Heard Island indicates that one of the major dietary items of the seals (C. gunnari) is of probable commercial importance and therefore any plans for the establishment of a fishery on Heard Island grounds must be considered in this light.

Satellite imagery of the pack ice edge position in western part of the Atlantic sector of the Antarctic derived on each tenday period basis has been used to analyse its spatial and temporal variation in the area between 65° and 20°W. Amplitude of the seasonal variation of the ice edge position was found to increase from the west to the east. Described is two-year cyclicity of consecutive change of positive and negative anomalies in ice conditions in winter and summer. Velocities of retreat and distribution of the ice edge were calculated for various meridians. Compared to the thirties, the means ice edge position was observed to shift over the current decade to the south to the Scotia Sea in winter, and to the west to the Weddell Sea in summer. 6-7 year cyclicity in the ice edge position is argumented in year to year aspect. Dependencies between dates of ice retreat from the South Orkneys and South Shetlands Islands and the ice edge position during the preceding months have been established. Dependencies are approximately by a linear function and can be used to forecast the dates of ice retreat form the islands.

The effects of a number of harvesting strategies on the Antarctic icefish Champsocephalus gunnari have been simulated for a period of 30 years.
These were
- diffierent levels of constant fishing mortality (F0.1, Fmax, 2xFmax),
- harvesting contantly at 50%F0.1 with an increase of F 3 or 6 years after a good recruitment,
- pulse fishing at an interval of 3 years with no fishing in between;
- a shift in partial recruitment values due to changes in net selectivity.
For the projections recruitment was assumed to fo11wed the historical pattern.
Pulse fishing prove to be the least preferable harvesting alternative. In the absence of regular recruit surveys constant fishing at F0.1 is most likely to be the most profitable and least risky harvesting strategy at present. The establishment of regular recruit survey would offer the possibility to adjust constant levels of fishing mortality to the strength of the incoming year class. An increaseof F, however, should not occur before 4 years after a good recruitment. A forward shift in partial recruitment values would not alter yield significantly when fishing at F0.1 and Fmax but would lead to a higher spawning biomass.