CCAMLR

We provide an update of the Bayesian sex and age structured population stock assessment model for Antarctic toothfish (Dissostichus mawsoni) in the Ross Sea (Subareas 88.1 and SSRUs 88.2A–B), using revised catch, catch-at-age, and tag-recapture data for the 2007 season. The updated reference model resulted in a slightly lower estimate of initial biomass than the 2006 base case. The inclusion of the 2007 recaptures of 2006 tags released had the most substantive impact on the model estimates.
Models using tag-recapture data from all vessels are also presented. Inclusion of all vessels tagging data resulted in a more optimistic assessment. The more optimistic estimate was probably due to the lower recapture rate by non-New Zealand vessels, particularly in the early years. The reason for this is unclear, and may be related to different distributions of fishing effort by different vessels, to poorer survival of tagged fish, or to poorer detection rates. However, if data from all vessels was restricted to 2007 recaptures of 2006 releases, then model estimates were more similar to the 2007 reference model.
Overall, model fits to the data were adequate, and, as in previous assessments, the tag-release and recapture data provided the most information on stock size. Monte-Carlo Markov Chain (MCMC) diagnostics suggested little evidence of non-convergence in the key biomass parameters. MCMC estimates of initial (equilibrium) spawning stock abundance (B0) for the 2007 reference model were estimated as 71 200 t (95% C.I.s 59 570–87 900 t), and current (B2007) biomass was estimated as 81.9% (78.4–85.4%). Estimated yields for the reference case, using the CCAMLR decision rules, were 2700 t.
Similar estimates of initial biomass and yield were obtained for models where the north fishing selectivity was assumed to be logistic (median B0=78 480 t, yield=2988 t), and where the 2007 recaptures from all vessels were added to the New Zealand tag-release and recapture data (median B0=81 100 t, yield=3099 t). If tag-release and recapture data for all vessels in all years were used, then the estimated initial biomass was higher (B0=110 130 t), and corresponding yields were also higher (yield=4200 t).

The current stock assessment in the Ross Sea (statistical Subareas 88.1 and 88.2) is underpinned by a tag and recapture programme, based on Antarctic toothfish Dissostichus mawsoni. The tagging of toothfish in this area is a mandatory requirement (Conservation Measure 41-01 Annex C). The Annex requires vessel to “target toothfish of all sizes in order to meet the tagging requirement”. This document details a number of methods and presents plans of some equipment used by New Zealand vessels operated by Sanford Limited a major New Zealand fishing company in order to satisfy this requirement. These techniques and equipment may have relevance in other tagging programmes for the closely related Patagonian toothfish Dissostichus eleginoides.

Aspects of the reproduction, size distribution, and movements of Antarctic toothfish Dissostichus mawsoni in CCAMLR Statistical Areas 88.1 & 88.2 were reviewed. Based on the presumed location and timing of spawning, and the probable early life history characteristics of toothfish, we investigated models that mimic the drift of eggs and larvae over a 6–24 month period using an oceanic circulation model linked to the high resolution global environmental model (HiGEM). The location of toothfish larvae after an 18–24 month period suggested by the models agreed moderately well with the distribution of the smallest toothfish taken in the toothfish fishery.
Our present hypothesis is that D. mawsoni in CCAMLR Statistical Subareas 88.1 and 88.2 spawn to the north of the Antarctic continental slope, mainly on the ridges and banks of the Pacific-Antarctic Ridge. The spawning appears to take place during winter and spring, and may extend over a period of several months. Depending on the exact location of spawning, eggs and larvae become entrained by the Ross Sea gyres, and may either move west settling out around the Balleny Islands and adjacent Antarctic continental shelf, south onto the Ross Sea shelf, or eastwards with the eastern Ross Sea gyre settling out along the continental slope and shelf to the east of the Ross Sea in Subarea 88.2. As the juveniles grow in size they move west back towards the Ross Sea shelf and then move out into deeper water (greater than 600m). The fish gradually move northwards as they mature, feeding in the slope region in depths of 1000–1500 m, where they gain condition before moving north onto the Pacific-Antarctic ridge to start the cycle again. Spawning fish may remain in the northern area for up to 2–3 years. They then move southwards back onto the shelf and slope where productivity is higher and food is more plentiful where they regain condition before spawning.
We recommend research surveys be carefully planned to test some of these hypotheses.

The ASPM assessment of the Prince Edward Islands toothfish resource by Brandão and Butterworth (2006) that permitted annual fluctuations about a deterministic stock-recruitment relationship is updated to take account of further catch, GLM standardised CPUE and catch-at-length information that has become available for the years 2006 and 2007. The assessment allows for a second fleet to accommodate data from a pot fishery that operated in 2004 and 2005. Biological parameter values adopted for Subarea 48.3 are used. The resource is estimated to be at about 37% of its average pre-exploitation level in terms of spawning biomass. It is suggested that it would be prudent to restrict annual legal catches to 500 tonnes or less, unless a large proportion of the catch is to be taken by pots (which avoid the cetacean predation associated with longlining).

Pilot studies were carried out in 48.3 in 2006 and in 88.1 and 88.2 in 2007 by the Argos Helena, testing 14/0 circle hooks and mackerel bait against hooks and bait routinely used by the vessel (14/0 EZ J-type hooks and squid bait respectively). Catch per unit effort (CPUE) of Macrourus spp. (kg 1000 hooks-1) was lowest with J-type hooks and mackerel bait for both regional trials. Variation in catch rates of Macrourus spp due to bait was significant, but not for hook-type. Bait type had similar effects on the catch rates of Dissostchus spp. and of all other by-catch species combined. However, variation in CPUE for target and non-target species due to geographical location, depth and soak time were also significant and when bycatch to catch ratios were analysed, the variation due to bait was not significant. In order for the benefits of mackerel bait in reducing macrourid (and other species) bycatch to be endorsed by the industry, further mitigation research is required to determine a gear configuration for automated longlines which maintains toothfish catch rates when this bait type is being used.

1. During the 2006/07 fishing season one UK and one New Zealand vessel fished in Subarea 48.4, catching 54 t of toothfish. A total of 291 D. eleginoides and 1 D. mawsoni were tagged and released during fishing operations. This represents a very high tagging rate of 5.4 fish per tonne of catch. Additionally, 100 rajids were tagged and released.
2. D. eleginoides and rajids appear to have different distributions of abundance around the fishing area. There were two tag recaptures in 2007, both of D. eleginoides. These were recaptured 84km and 14km from their points of release.
3. The UK proposes to continue the mark-recapture experiment in Subarea 48.4 in 2007/08. Initial estimates of population size will be made in 2008/09. However, given the currently low level of recaptures, and the difficulties experienced in catching and tagging the anticipated number of toothfish each year, the experiment and CM 41-03 may need to be extended for a few more years.

The UK undertook preliminary trials of the previously described “cone” or “bottle” longline configuration for mitigating cetacean interactions in 2006. This paper should be read in conjunction with the observer’s report of that cruise (Basson 2006). Initial results were promising, and the UK intends to proceed with further, more extensive trials, in the 2007/08 fishing season in Subarea 48.3. Although seabird interactions are expected to be minimal, the trials will stop if 3 seabirds are caught, and sink tests will be undertaken. Although we intend to change gear configurations as described all other conservation measures will apply.

IW longlines are effective in reducing the bycatch of birds in toothfish fisheries, but by putting the line closer to the seabed they increase the bycatch of macrourids and rajids. Experiments reported elsewhere demonstrate that some of this effect can be mitigated by the use of fish rather than squid bait, but even then autolines continue to catch more macrourid bycatch than Spanish systems. We intend to trial two alternative configurations against IW systems in 48.3 in the 2007/08 fishing season: non-IW autolines with spaced weights, and IW systems with alternately spaced weights and floats. All systems will have sink rates of at least 0.3 m.s-1. Comments on the experimental protocol are invited. Although we intend to change gear configurations as described all other conservation measures will apply.

1. The catch-at-length based CASAL model for toothfish at South Georgia is updated with data from the 2007 fishing season. The predicted spawning stock biomass and the yield is slightly higher than was estimated last year.
2. Improvements are made to the fit of tag data through (a) estimating a length-based ogive for tag-induced mortality, based on our 2005 survivorship experiment, in which smaller fish survive tagging better than large fish, and (b) by re-estimating from our tagging data that tag-induced growth retardation is also related to size, smaller fish suffering less growth retardation than larger fish, and that on average it is 1 year or more.
3. A new model is developed which uses estimates of catches at age from 1998 to 2006 (based on random sampling of the catch for age determination). Fits of all data (CPUE; catches at length for the early fishery; catches at age for the later fishery; tag recapture data) are all improved from earlier models, although some poor fits remain. The model estimates year class strength which corresponds, in some years, with estimates made from the South Georgia groundfish survey data.
4. Several of the requests of WG-FSA for developments of the South Georgia toothfish model have now been completed.
5. Note that the data used here are preliminary and lack some of the precaution previously included in assessments. We would recommend that only small changes to the current TAC be considered, and that the toothfish assessment is undertaken at periods of greater than one year.

The exploratory fishery for Antarctic toothfish (D. mawsoni) has been operating for ten years in Subarea 88.1 and for five years in Subarea 88.2. This report summarises the large amount of data collected on toothfish and the associated bycatch by all vessels participating in the fishery. All SSRUs in the two subareas except for 881D and 882C have now been fished. The 2007 D. mawsoni catch was the second highest on record with a total of 3431 t against a combined catch limit of 3579 t. The management of the SSRUs within the two subareas was changed for the 2006 season as part of a 3-year experiment (SC-CAMLR-XXIV). One of the aims of the experiment was to simplify the administration of the fishery by having fewer catch limits. This appeared to be moderately successful, with only one catch limit being slightly exceeded in the 2006 season, and two catch limits exceeded in the 2007 season. Although there was a large overrun of the catch limit in the North region, the overall catch limit for Subarea 88.1 was only exceeded by 2%. The catch limit was under caught in Subarea 88.2. The concentration of effort within smaller spatial areas in Subarea 88.1 has undoubtedly contributed to the large increase in tag recoveries over the past two seasons (Dunn et al. 2007).
The length frequency data from the Ross Sea fishery have been very consistent over the past 3–4 seasons. There is no evidence of any truncation of the overall length frequency distribution, and no evidence for a reduction in fish length in any SSRU over time. Although moderate numbers of small fish are caught in some years (e.g., on the Shelf in 1999 and 2001), these year classes are not seen in large numbers in later years in the fishery. So at this stage there is no evidence for strong variation in year class strength in the fishery.