15.4.4 Marine ecosystems and their services in the Antarctic

Southern Ocean ecosystems are far from pristine. Over the last 200 years some seal and whale species (e.g., Antarctic fur seals, blue and fin whales) were exploited almost to extinction, then fisheries developed. From the 1960s, fin-fish were exploited, and in the Scotia Sea and surrounding areas stocks of these fish were reduced to low levels and have not yet recovered. In contrast to the Arctic, however, the management of Southern Ocean fisheries is based on an ecosystem approach, within an international convention. The Convention on the Conservation of Antarctic Marine Living Resources (CCAMLR), part of the Antarctic Treaty, was designed to maintain the natural marine ecosystem while allowing sustainable exploitation, and emphasises the need to consider the wider context of the exploitation of individual species, taking account of the entire food web and environmental variations. The CCAMLR applies to areas south of the Antarctic polar front, and management decisions are made by consensus of the Member States (Constable et al., 2000).

The current major fin-fish fishery is for the Patagonian toothfish (Dissostichus eleginoides), and to a lesser extent for the mackerel icefish (Champsocephalus gunnari). The fishery for Antarctic krill (Euphausia superba) developed during the 1970s, peaked in the 1980s at over 500,000 t/yr and now operates at about 100,000 t/yr (Jones and Ramm, 2004), a catch that is well below the precautionary limits set within CCAMLR for maintaining the stock.

During the 20th century there were significant changes in air temperatures, sea-ice and ocean temperatures around the Antarctic Peninsula (see Section 15.6.3) and in the Scotia Sea. Over 50% of the krill stock was lost in the Scotia Sea region (Atkinson et al., 2004), which is the major area for krill fishing. The decline in the abundance of krill in this area appears to be associated with changes in sea ice in the southern Scotia Sea and around the Antarctic Peninsula (Atkinson et al., 2004). Future reductions in sea ice may therefore lead to further changes in distribution and abundance across the whole area, with consequent impacts on food webs where krill are currently key prey items for many predator species and where krill fishing occurs.

For other species the uncertainty in climate predictions leads to uncertainty in projections of impacts, but increases in temperatures and reductions in winter sea ice would undoubtedly affect the reproduction, growth and development of fish and krill, leading to further reductions in population sizes and changes in distributions. However, the potential for species to adapt is mixed, some ‘cold-blooded’ (poikilothermic) organisms may die if water temperatures rise to 5-10°C (Peck, 2005), while the bald rock cod (Pagothenia borchgrevinki), which uses the specialisation of anti-freeze proteins in its blood to live at sub-zero temperatures, can acclimatise so that its swimming performance at +10°C is similar to that at -1°C (Seebacher et al., 2005).

The importance of ocean transport for connecting Southern Ocean ecosystems has been increasingly recognised. Simple warming scenarios may indicate that exploitation effects would be shifted south, but it is also likely that other species may become the target of new fisheries in the same areas. More complex changes in patterns of ocean circulation could have profound effects on ocean ecosystems and fisheries, although not all changes may be negative and some species may benefit. Complex interactions in food webs may, however, generate secondary responses that are difficult to predict. For example, reductions in krill abundance may have negative effects on species of fish, as they become a greater target for predators. Importantly, the impact of changes in these ecosystems will not be confined to the Southern Ocean. Many higher predator species depend on lower-latitude systems during the Antarctic winter or the breeding seasons.

The fundamental precautionary basis for managing exploitation in a changing environment is in place in CCAMLR, but longer duration and more spatially extensive monitoring data are required in order to help identify change and its effects.