The Secretariat has embarked on a major overhaul of CCAMLR data holdings and associated IT and data infrastructure. This work, which begun in 2013, includes developing an Enterprise Data Model, redeveloping the CCAMLR database, improving data quality assurance, and modernising the data work flow.

The user community can expect to notice significant improvements in data quality, database documentation and ease of use as the new system begins a process of acceptance testing from late 2015. Consequential changes will be required in requested data extracts to reflect the new data model and nomenclature.

This paper provides an update on progress, including work undertaken since WG-SAM-15 and WG-EMM-15.

We presented five sensitive runs of single-sex age-structured CASAL models for stock status assessments of Dissostichus eleginoides in the research block 58.4.4b_1 in the last WG-SAM meeting. IUU catch was estimated within the models using disease mortality functionality. Among them, the plausible model was considered as OB_1b model, which removed tag and CPUE observations in 2008, because the fits to CPUE in 2008 and MPD profiles for tags in the same year was spurious in OB_1a model which included all observations. The Working Group, however, requested the further examination of the sensitivity to the 2008 tag data, along with the IUU selectivity modeled as a double normal function. The Working Group also requested that projections be presented to WG-FSA for this assessment that examine the consequences of different harvest levels for the time to recovery to the target level.

We revised to retain tag data in 2008 for OB_1b and other relevant models following the recommendations. The total IUU catches in the revised OB_1b model were estimated to be 71 tonnes, which corresponded 43 % of that (163 tonnes) in the same model of the last WG-SAM report. The median MCMC estimates of the initial and current biomass were 610 and 330 tonnes in the revised OB_1b model. The CCAMLR yield was calculated at 30 tonnes. Any harvest level lead to the CCAMLR risk 2 value > 50% in the beginning of 35 year projection period, and the CCAMLR yield 30 tonnes lead to the period as 24 years (between 2015 and 2038). We estimated the depredation rates by killer whales in the separated document. When depredation rate of 25 % by killer whales constantly occurs, the CCAMLR yield should be reduced to 22.5 tonnes.

We revised sensitivity runs of age-structured and two fisheries (≥1 300 m and < 1 300 m) CASAL models for stock status assessments of Dissostichus eleginoides in the research block 58.4.3a_1 following the recommendations of the last WG-SAM meeting. Total of four runs were carried out. Among them three runs were carried out using the latest information on age-length relationship in this division: two catch at length models including (ELAN_len model) or excluding (ELAN_tag model) the data from fish tagged in 2012, and a catch at age model including the data with a single age length key applied to all years (ELAN_age model). Another run was a catch at length model as ELAN_len model except using age-length relationship for Ob-Lena Banks (ELAN_lenOB).

The likelihood profiles for tagged fish released in 2012 for ELAN_len, ELAN_age and ELAN_lenOB models reflected a large number of recaptures of the tags released concentrated in the western area of research block. This is why the removal of 2012 tag release data boosted the MPD estimate of B₀ in ELAN_tag model. The selectivity of the right hand limb of deep fishery was poorly estimated in MCMC runs for ELAN_len model.

The median MCMC estimates of the initial and current biomass for ELAN_len model were 330 and 260 tonnes and the CCMLR yield for this stock was calculated at 30 tonnes. The fishing bias to the shallow areas should be avoided because the CCAMLR risk 2 values at 13 tonnes and above for the shallow fishery exceed 50% toward the end of 35 year projection period in the model.

This report arranged the biological information with special reference to bycatch, depredation and spawning dynamics in Division 58.4.4 a & b during 2008-14 seasons in response to the recommendations in the last WG-SAM meeting and their related issues.

D. eleginoides (TOP) was highest both in weight and number, accounting for 93% and 76% of the total catch, respectively, among total of 16 species/genus occurred in Division 58.4.4a & b. Grenadiers (GRV) showed the second highest catch and was widely spread in the division as TOP, but the proportion in number to the whole species/genus (17%) was considerably smaller than that in Kerguelen (Division 58.5.1) and Crozet (Subarea 58.6). The fish assemblage for the three dominant species/genus was mainly divided into three groups using clustering. The three groups were generally divided by depths.

Depredation by whales was analyzed using three methods; the CPUE method and grenadier bycatch method by Gasco et al. (2014) using two size levels of unit cell, and generalized linear model with stepwise process for separating two areas, SSRUs 58.4.4aA & 58.4.4bB and SSRU 58.4.4bC. The depredation rate in weight of whales toward toothfish was estimated to be only 3.8-4.8 % in SSRU 58.4.4bC during 2008-12 using the CPUE method, while the presence/absence of whales was not considered to be a significant parameter for the catch rate of toothfish in the generalized linear model in the same SSRU. The depredation rate in weight was estimated to be 20-25% in SSRUs 58.4.4aA & 58.4.4bB during 2008-14 using the CPUE method. The standardized CPUE in absence of whales in the SSRUs was estimated to be 2 times higher than that in presence of whales in the generalized linear model, and the depredation rate of 25% estimated from this ratio and the occurrence rate of whales of longlines (49%) was identical with the rate estimated using the CPUE method with the 0.5˚x0.25˚unit cells. The grenadier bycatch method by Gasco et al. (2014) is currently rather difficult to be applied in the division because of considerably low proportion in abundance of GRV to sum of GRV and TOP.

Some spawning activity in this division seems to occur in the slope areas during April-August. The GSI values along with macroscopic stages tended to increase with their body sizes for both sexes. Because of some spawning activity along with the evidence of recruitment of small sized fish, the stock is likely to be self-sustainable in each SSRU. Further discussion was made in the separated document of research plan in Division 58.4.3a.

Catches of toothfish, icefish and krill reported in the Convention Area in 2013/14 and 2014/15. Catches reported in 2013/14 were published in the CCAMLR Statistical Bulletin Volume 27 (https://www.ccamlr.org/en/data/statistical-bulletin). Catch totals for the current season (2014/15) are preliminary and based on data reported to 16 September 2015.   

In this document, we revise stock status assessments of D. eleginoides in research blocks 58.4.4b_1 and 58.4.4b_2 using CPUE analogy method, Petersen method and CASAL model. Detailed results of reanalysis using CASAL models in research block 58.4.4b_1 are described in a separated document.

We reset the revised research blocs drawn along the outer margin of grids because the previous block had not contained the eastern grids accurately, and recalculated the sea bed area sizes (600-1 800 m) using the same GIS software (GEBCO_08 Grid, version 20100927) as the latest CCAMLR Statistical Bulletin.

The estimated stock size in block 58.4.4b_1 was 510, 336 and 333 tonnes in CPUE analogy method, Petersen method and CASAL OB_1b model (vulnerable biomass in 2014), respectively.

The stock sizes of D. eleginoides in block 58.4.4b_2 is estimated at 654 tonnes only by using CPUE method, as the catch and tagging data in block 58.4.4b_2 was not enough to be applied to Petersen method and CASAL models.

We propose to continue the current research operation for the next fishing season with the same survey design and total sample size of 60 tonnes in order to further strengthen the stock assessments in the area.

The current season (2014/15) is the final one of the first three-season research which began in 2013 in data-poor fisheries. In the present report, the next three-season (2016-18) research plan in Division 58.4.3a was revised using the latest CCAMLR C2 and Observer data.

The stock sizes for block 58.4.3a_1 was estimated by the Petersen estimator and the CPUE x seabed analogy method.

The stock size using the Petersen estimator was largely different from that using the CPUE method. Predicted numbers of tag recaptures from the estimated stock sizes using the Petersen method were generally consistent with the observed number during 2013-14 seasons, but those using the CPUE method were inconsistent.

We also made analysis on spawning dynamics for D. eleginoides in Division 58.4.3a and compared to the results in other regions in the Southern Indian Ocean. The stock seems to be self-sustainable because of the existence of fish with advanced gonadal indices in the slope area and recruitment of small sized fish mainly in the shallow area, although the stock has a possibility of metapopulation throughout the Southern Indian Ocean as previous genetic studies indicated.

Understanding of the resource structure through clarification of their life history is essential to establish stock assessment and robust stock/ fisheries management of Dissostichus spp. population(s) in data- poor exploratory fisheries. During the second three-season research, we will continue enhanced tagging program, and collection and analysis of biological data including otoliths and gonads to clarify migration route and associated life stages of the fish.

To this end, we propose to follow the current research style in the current research blocks for the second three-season research with the sample sizes estimated following the procedure recommended at the WG-FSA in 2013 in order to maximize the expectation of tag-recapture to the extent possible under the precautionary exploitation rate.

The current season (2014/15) is the final one of the first three-season research which began in 2013 in data-poor fisheries. In the present report, the next three-season (2016-18) research plan in Division 58.4.2 was revised using the updated CCAMLR C2 and Observer data.

The stock sizes for a research block (58.4.2_1) were estimated by the Petersen estimator and the CPUE x seabed analogy method using updated CCAMLR C2 data and reference area sizes.

Understanding of the resource structure through clarification of their life history is essential to establish stock assessment and robust stock/ fisheries management of Dissostichus spp. population(s) in data- poor exploratory fisheries. During the second three-season research, we will continue enhanced tagging program, and collection and analysis of biological data including otoliths and gonads to clarify migration route and associated life stages of the fish.

To this end, we propose to follow the current research style in the current research blocks for the second three-season research with the sample sizes estimated following the procedure recommended at the WG-FSA in 2013 in order to maximize the expectation of tag-recapture to the extent possible under the precautionary exploitation rate.