CCAMLR

About 1300 stomach contents of the icefish Champsocepha1us gunnari caught in the Subarea 48.3 in February 1995, were analyzed using the frequency of occurrence (%) method. Krill (Euphausia superba) (58%) and the amphipod hyperiid Themisto gaudichaudii (55%) occurred similarly at South Georgia and constituted the main food. At Shag Rocks, T. gaudichaudii was the main prey item (52%), followed in importance by the mysid Thysanoessa sp. (35%) and krill (24%). The picture around SG was found to be similar to that observed in the E. Holmberg survey in February/March 1994: common levels of krill availability indicated by frequency of occurrence, but, on the other hand, high proportions of empty stomachs (also at Shag Rocks), close to the values reported for years of krill scarcity. While no krill aggregations were found in the South Georgia shelf during the Holmberg Cruise in Feb-March/1994, an acoustic survey carried out in the same area in March 1995 detected such aggregations. It is suggested that krill availability for C. gunnari around South Georgia in the 1995 summer season, was somewhat below the historical years of high abundance, but above than in 1994.

Due to the detrimental effect longline fisheries have had on seabirds in other parts of the world, concern was raised as to the effect an experimental Hake longline fishery would have on Procellariiform seabirds in South Africa. Carcases of Procellariiform seabirds caught on Hake longlines during setting were retrieved to determine species, sex and stomach contents. Whitechinned petrels constituted 100% of the catch on longlines during setting despite the fact that they were not the most abundant species around fishing vessels either during the day or night. Extrapolations showed that 2 vessels were responsible for 1928 ± 1345 whitechinned petrel deaths a year in South African waters. Stepwise multiple regression suggested that light intensity was the critical factor explaining variation in the number of whitechinned petrels caught during setting. Whitechinned petrels become highly active approximately 2,5 hours before sunrise. If line shooting could be completed prior to Whitechinned Petrels becoming highly active, a great deal of mortality might be avoided. Nine great shearwaters and two pintado petrels were killed during 13 hours of hauling procedures. Hauling mortalities can be avoided by diverting offal outlets to the opposite side of the vessel to where the longline is being hauled. Larger vessels which continuously process offal seem to be more attractive to scavenging seabirds then smaller vessels. Mitigation measures must be introduced into the Hake longline fishery to reduce excessive seabird bycatch. Suggested measures include: (1) the introduction of tori lines on all vessels which catch seabirds; (2) mandatory setting of lines prior to whitechinned petrel activity peaks. Experimentation with levels of artificial lighting and with differentially weighted longlines to ensure that hooks sink immediately when placed in the water are essential to further reduce catch rates of seabirds. Without the implementation of mitigation measures population growth and long term population stability of whitechinned petrels may soon be impaired.

An analysis of the Patagonian toothfish (Dissostichus eleginoides) fishery was carried out using data obtained during the 28th cruise of the SRTM Primorets to the Kerguelen Islands economic zone in the 1994-1995 season.

New catch data reported to CCAMLR indicate that Notothenia rossii was heavily exploited in Subarea 48.3 from 1966- 197 1, rather than just 1970 and 197 1 as previously reported. Previous assessments, based on the incomplete data set, had indicated a 1970 biomass of 600 000 tonnes and an unrealistic stock history considering that this stock was the first to be exploited in the Antarctic. A VPA using the new data suggests that unexploited biomass was over 900 000 tonnes, and declined in a manner to be expected from a new fishery. However, on the basis of historical reports and analyses of stock histories, it is suggested that 50% of the catch of 400 000 tonnes reported from 1970 was Champsocephalus gunnari, rather than N. rossii, and that 25% of the catch in 1969 was also C. gunnari. A VPA run with these catches predicts a more reasonable stock history for a species undergoing first exploitation, and an unexploited biomass of 600 000 tonnes prior to 1966.