CCAMLR

Analysis of seasonal sea ice cover of a 1,250,000 km2 area off the northwestern side of the Antarctic Peninsula indicates that four "ice events" have occurred during the last 1 7 years. The most recent event was first apparent in 1994 and continues through the latest data available. Variability between ice events in seasonal timing, areal extent, seasonal duration and persistence over multiple years is apparent. Annual curves of sea ice cover were integrated over time to produce an annual index of sea ice cover in units of 106 km2-months.

Results from two acoustic and net surveys conducted in the vicinity of Elephant Island during the 1996 austral summer indicate very good krill recruitment from spawning in 1994/5. Areas of high krill density were mapped north of King George and Elephant Islands where water depth was greater than 200m. One-year old juvenile krill numerically dominated catches during the first survey in January and were widely distributed throughout the southern portion of the survey area. Large, sexually mature adult krill were caught during both surveys north of the islands, but were numerically dominate only during the second survey in February-March. Intermediate size krill were caught in very low numbers reflecting poor recruitment from spawning in 1992/93, and 1993/94. Biomass and abundance estimates were the highest since 1992 when the effect of good recruitment from spawning in 1991 was observed. The abundance of salps was similar to 1995 and two to three orders of magnitude less than observed in 1993 and 1994. Following three winters of relatively low sea ice cover off the western side of the Antarctic Peninsula, the winter of 1994 marked the beginning of a period of relatively extensive ice coverage continuing through 1995. These observations support the hypothesized relationships between winter sea ice conditions, the lack of a spring-time salp bloom, the timing of spawning by adult krill, and the success of krill recruitment proposed by Loeb and Siegel (1994a) and Siegel and Loeb (1995).

Four indices of prey availability are calculated for prey surveys conducted in the vicinity of the Seal Island CEMP site during the austral summers of 1990-96. The indices are measures of average prey density, depth, distance from Seal Island, and persistence over time. Acoustic data from two AMLR surveys each year were sub-sampled to include the foraging range of predators breeding at Seal Island. The average depth of the prey field and its average distance from Seal Island were positively correlated; no other relationships between the indices were apparent. Indices of prey availability were compared with indices of predator performance at Seal Island. The depth of the prey field and its distance from Seal Island appear to have a positive effect on the duration of chinstrap foraging trips, but not on breeding success. The distance of the prey field from Seal Island appears to be negatively correlated with both the duration of fur seal foraging trips and pup growth rate.

There is a growing body of evidence that the climate of the Antarctic Peninsula region has been warming over the past 40 years with an associated decreased frequency of winters with extensive sea-ice development. These trends potentially will have a major impact on the structure and function of the Antarctic marine ecosystem. In the Antarctic Peninsula region winter sea-ice coverage is a major factor regulating recruitment and population size of Antarctic krill (Euphausia superba) and population dynamics of salps (Salpa thompsoni). Strong krill recruitment success there follows years of extensive winter sea-ice development and large summertime salp blooms follow winters with relatively little sea-ice. An order of magnitude decrease of krill population size, increased incidence of massive salp blooms, and decreased abundance of krill-dependent Adelie penguins (Pygoscelis adeliae) in the past 15 years suggests that the food web may be affected by climate change. The 1994/95 austral summer season followed the first prolonged winter sea-ice season in three years and provides a strong contrast to the previous years which had little or no winter sea-ice development. The hypothesized relationships between krill and salp population dynamics and winter sea-ice conditions are confirmed and the relative importance of krill and salps within the Antarctic food web are assessed here.

1. Although the development of fisheries for krill in the southern oceans has prompted considerable work on the indirect effects of fisheries on krill predators, to date all work has focused on population level effects. Here, we present for. the first time a model at the level of the foraging trip for the effects of a fishery on krill predators, using the Adelie penguin (Pygoscelis adeliae) as a model organism.
2. The model has four main components: the description of the spatial and•temporal pattern of krill, ii) the effects of the fishery on the krill, iii) the description of penguin breeding and iv) the indirect effects of the fishery on penguin reproduction and survival. As with all models, there is compromise between the level of tractability and the level of biological detail.The objective is to make relative comparisons of penguin reproductive success and adult survival in the absence or presence of a fishery.
3. The biomass of krill appropriate for the predators (and the fishery) fluctuates from one year to the next according to an age-structured, stochastic recruitment model. We use the model to generate the long-term frequency distribution of krill biomass. Furthermore, we assume that there is spatial-temporal structure, determined by diffusion and advection, to krill availability in relation to the location of the penguin breeding colony. Fishing is assumed to change the spatial and temporal distribution of available krill.
4. We assume that after fledging, offspring survival depends in part upon the amount of krill delivered to them during the feeding periods. We use empirical data to estimate parental and offspring needs and a standard life history model to set the upper limits for expected offspring and parental survival. We assume that parental survival subsequent to breeding depends upon the krill deficiency (relative to needs) accumulated while feeding the young.
5. A sensitivity analysis of the breeding model shows that the predictions are robust to parameters about which little is known, to the functional forms relating krill abundance to offspring and parent survival, and to the rules that parents use to allocate krill to their offspring.
6. We evaluate expected reproductive success (offspring survival) and expected parental survival as functions of the amount of krill captured by the fishing fleet. Over the range of catch in our study, the reductions in reproductive success, are essentially linear functions of krill catch, with slope 1.5. Reductions in adult survival are also linear functions of krill catch, but with slopes less than 1; that is, reductions in reproductive success and parental survival are linear functions of krill catch but not 1:1. The reductions in offspring and parent survival are mainly determined by how long the fishing season lasts and the capacity for harvest, rather than when fishing begins.

The purpose of this paper is to synthesize information on the energetics and food requirements of krill predators that may be in the most direct competition with the krill fishery. These krill predators are the Adélie (Pygoscelis adeliae), chinstrap (Pygoscelis antarctica), gentoo (Pygoscelis papua), and macaroni (Eudyptes chrysolophus) penguins and the crabeater (Lobodon carcinophagus) and Antarctic fur (Arctocephalus gazella) seals. Basal, active, and. reproductive energy ' requirements are compiled for each species. Estimates' of the amount of krill needed . to sustain individuals and/or breeding pairs and population requirements are provided.

Acoustic surveys were carried out in January 1996 within two 100 x 80 km boxes located over the shelf-break to the north-east and north-west of South Georgia. Surveys were conducted from RRS James Clark Ross using a Simrad EK500 echo-sounder operating at 38, 120 and 200 kHz. These surveys were the first in a new five-year British Antarctic Survey programme initiated in part to monitor inter-annual variability in the abundance of the Antarctic krill Euphausia superba in the South Georgia region, and to provide krill biomass estimates for management purposes. Survey box 1 was located to the north-east of Cumberland Bay and encompassed Charlotte Bank, an area where historically whale catches were high and where numerous observations of elevated krill abundance have been made. Box 2 was located to the north of Bird Island, within a foraging area of prime importance for breeding krill predators. Together the two boxes therefore provide data of value to fishery scientists, ecologists and environmental modellers. The acoustic surveys were of randomised design and each nominally constituted 10 randomly spaced parallel transects, 80 km in length, running perpendicular to the major direction of shelf-break in the area; poor weather however necessitated some modification of this plan during the 1996 surveys. All transects were steamed during the hours of daylight in order to avoid biases in krill biomass estimation caused by diel vertical migration. Echoes were integrated in 2 m depth bins from 2 to 250 m below the transducers, over 100 second intervals (≈0.5 km at 10 knot survey speed). Two transects were run per day, and at night net hauls were carried out at a location toward the centre of the previous pair of transects to obtain krill for estimates of length frequency distributions. Weighted mean krill lengths of 29.5 mm for box 1 and 32.0 mm for box 2 were determined. Thresholded (mvbs 120 kHz minimum -100 dB) 120 and 38 kHz echo signal pairs were partitioned using a dB difference technique (δmvbs = mvbs 120 kHz - mvbs 38 kHz) into those attributable to krill (δmvbs between 2 and 12 dB), small zooplankton (δmvbs > 12 dB) and nekton (fish/squid, δmvbs