CCAMLR

Interactions between seabirds and longline fishing can cause incidental bird mortality and reduced gear efficiency. The potential for solving these problems by using a bird-scaring streamer line and a line shooter was investigated in commercial longlining in the northern Atlantic off the coast of Norway. We compared the bycatch rate of northern fulmars Fulmarus glacialis, the loss rate of baits to fulmars and the catch rates of fish target species among lines set with either of these mitigation measures, a combination of both and no mitigation measure. A total of 58,420 hooks were set in each of the four treatments. No birds were caught using the bird-scaring line alone and a single fulmar was caught when the bird-scaring line was used in combination with the line shooter. In contrast, 32 fulmars were caught in sets with no mitigation device and thirteen in sets with the line shooter alone. Losses of mackerel bait were reduced when the bird-scaring line was used, but not by using the line shooter alone. Longlines set with the line shooter reached 3 m depth 15% faster then lines set without the line shooter; beyond this depth sinking rates were similar (about 15 cm s-1). Fish catch did not vary significantly among setting methods. These results should also be applicable to the bycatch of Fulmarus spp. in demersal longline fisheries worldwide.

This paper considers a class of survey patterns for a toothfish long lining fishery. These survey patterns are characterised by a minimum mutual separation distance and a shot selection criteria. The shot selection criteria can be either that locations of known good quality are preferentially selected or that locations are selected at random. The landscape and fishery are simulated in a spatially explicit Monte Carlo model and the use of such models to study these issues is considered along with their ability to answer such questions as: How well do different survey restrictions capture information of interest about the fished population? If and when can the catch/effort data be used as a measure of the relative abundance of the population? How well can this method detect changes in the abundance of the population? How sensitive are the results to the number of research shots used?
This paper shall show the strength of the simulation method in answering these questions and in exploring the efficacy of survey strategies in general as well as a generator of ideas for survey design.

Revisions to the Generalised Yield Model provide for re-estimating the recruitment series from survey data for each value of natural mortality used in the assessments, whenever it is altered over the range of uncertainty in M factored into the assessment process. Data from outputs from mixture analyses from surveys are input to the GYM in raw form. An example file is provided. Consequently, endeavouring to predetermine the recruitment series with an average value of M is no longer necessary for inputting to the GYM.
In addition, provision is now made for inputting different fishing selectivity functions for different years in the assessment model. This is done in the usual way but with repeated input sections for each year at which the selectivity is altered.