A RISK ASSESSMENT TO ADVISE ON STRATEGIES FOR SUBDIVIDING A PRECAUTIONARY CATCH LIMIT AMONG SMALL-SCALE MANAGEMENT UNITS DURING STAGE 1 OF THE STAGED DEVELOPMENT OF THE KRILL FISHERY IN SUBAREAS 48.1, 48.2, AND 48.3.

We used Foosa and the reference set of parameterizations developed by Watters et al. (2008) to assess the risks and tradeoffs associated with various management strategies for subdividing the precautionary krill catch limit among SSMUs in Area 48. Our methodological approach follows directly from specifications made by the WG-SAM and the WG-EMM. We predict that the tradeoffs inherent in selecting among Fishing Options 2, 3, and 4 are tradeoffs between the risks to predator populations and fishery performance; risks to krill are relatively insensitive to differences among these options. Implementation of Fishing Option 2 (using the subdivisions reported here) will require that the fishery mostly operate in pelagic SSMUs. Up to harvest rates of 0.5 x γ, this subdivision is unlikely to reduce predator populations to 75% or less of the abundances that might occur in the absence of fishing; the risks of such depletion are, however, likely to increase as harvest rates increase beyond 0.5 x γ. Although we predict that catches can be highest and relatively less variable in pelagic SSMUs, the risks that krill densities will fall below thresholds which necessitate involuntary changes in the behavior of the fleet are substantially increased in pelagic SSMUs. We are uncertain about relative catchabilities in pelagic versus coastal SSMUs, and we do not know how much fishing effort might actually be required to catch the SSMU-level quotas that would be allocated to each SSMU given the subdivisions reported here. Implementation of Fishing Option 3 will also require substantive fishery operations in pelagic SSMUs, but, if krill move, to a lesser extent than Fishing Option 2. Up to harvest rates defining the current trigger level (i.e., 0.15 x γ), implementation of Fishing Option 3 is unlikely to reduce predator populations to 75% or less of the abundances that might occur in the absence of fishing. As harvest rates are increased past that defining the current trigger level, the risks of depleting penguin and fish populations increase in some SSMUs. In general, the risks of depleting penguin and fish populations are greater for Fishing Option 3 than for Fishing Option 2 because the former option requires slightly more fishing in coastal SSMUs. Nevertheless, fishery performance under Fishing Option 3 is comparable to that for Fishing Option 2. Implementation of Fishing Option 4 will, relative to the other two options, substantially limit the spatial distribution of the fishery. Furthermore, since Fishing Option 4 would concentrate fishing in a few coastal SSMUs, implementing this option would increase the risks that predator populations will be reduced to 75% or less of the abundances that might occur in the absence of fishing. In fact, in a few SSMUs, such risks even occur at harvest rates near the rate defining the current trigger level. Relative to Fishing Options 2 and 3, fishery performance under Option 4 is also poor, with decreased catches and increased variations in catch... continued