Terms of Reference

Effects of Age Composition Data on Operating Model Performance for Sablefish (Anoplopoma fimbria) in British Columbia

Regional Peer Review - Pacific Region

March 25, 2024

Virtual Meeting

Chairperson: Jessica Finney

Context

The Sablefish (Anoplopoma fimbria) operating model is a key component of the Management Strategy Evaluation (MSE) process that has been used to guide annual harvest management decisions for the BC Sablefish fishery since 2011 (Cox et al. 2023; DFO 2020, 2023a, 2023b). Operating model scenarios are used to (i) characterize stock status relative to reference points and (ii) generate simulated data that represent alternative hypotheses for uncertain future stock and fishery dynamics. The performance of candidate fishery management procedures (MPs) is then evaluated against measurable objectives based on conservation and socio-economic goals for the fishery using the simulated data.

The Sablefish operating model is a two-sex statistical catch-at-age model fit to:

1. fishery-specific landed catch from three fishery gear types (trap, longline hook, and trawl);
2. at-sea releases from each of the fishery gear types;
3. three indices of total abundance (two of which are historic indices that are no longer updated, as well as the current Sablefish stratified random (StRS) trap survey);
4. age composition data from the trap fishery and the StRS survey, as well as from a historic survey for which age compositions are no longer updated; and
5. length compositions from the trawl fishery.

Annual age composition data are a key input to the Sablefish operating model. They inform fishery and survey selectivity within the model, as well as recruitment patterns, natural mortality, and growth. If age composition data are inadequate, operating model estimates of stock status may be biased, and tested MPs will not be exposed to simulated data that are representative of actual stock dynamics. The paucity of age composition data from longline hook and trawl fisheries has resulted in selectivity for these fisheries being heavily informed by Bayesian prior distributions derived from tag release-recovery data (Cox et al. 2023; DFO 2023a; Johnson et al. In prep.^{Footnote 1}). The question of how best to allocate ageing effort to survey and fishery sources is therefore important, as well as considering what risks might be created by incorrectly specifying selectivity for fisheries with low or missing age composition data. Judging the effects of age composition data on model performance can be based on the statistical properties of estimated management parameters, however, it is also important to consider the real-world consequences of bias and variance by evaluating management decision points (e.g., likelihood of being below a limit reference point) and management performance over time.

This Science Response process was initiated by Fisheries and Oceans Canada (DFO) Science in the Pacific Region to increase understanding of how sample size and allocation of ageing effort among survey and fishery gears effects Sablefish operating model performance (bias and precision of management parameters). Outputs will help illustrate trade-offs that result from the allocation of ageing effort. The analytical methods developed for this analysis are applicable to age-structured stock assessments for other species, although the specific results apply to Sablefish.

Objectives

This Science Response is intended to address the objectives listed below:

1. Assess the effects of age composition sample size from survey, trap fishery, and longline fishery sources (i.e., ‘fleets’) on operating model performance by varying (a) the total sample size, and (b) the allocation of ages to each of the fleets. A range of sample sizes that includes historical ageing effort will be considered to determine the value of information.
2. Configure the operating model to represent population scenarios that decline spawning biomass towards the limit reference point of 0.4 B_MSY so that errors in estimated stock status can be characterized for a range of age compositions as in (1).
3. Explore how mis-specified fishery selectivity and future allocation of catch to fleets affect stock assessment error for a range of age compositions as in (1).

Expected Publications

• Science Response

Expected Participation

• Fisheries and Oceans Canada (DFO) (Ecosystems and Oceans Science sector)
• Canadian Sablefish Association

References

• Cox, S.P., Kronlund, A.R., Lacko, L., and Jones, M. 2023. A Revised Operating Model for Sablefish in British Columbia, Canada in 2016. DFO Can. Sci. Advis. Sec. Res. Doc. 2023/023. vii + 127 p.
• DFO. 2020. Evaluating the robustness of candidate management procedures in the BC Sablefish (Anoplopoma fimbria) fishery for 2019-2020. DFO Can. Sci. Advis. Sec. Sci. Resp. 2020/025.
• DFO. 2023a. A Revised Operating Model for Sablefish in British Columbia in 2022. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2023/010.
• DFO. 2023b. Application of the British Columbia Sablefish (Anoplopoma fimbria) Management Procedure for the 2023-24 Fishing Year. DFO Can. Sci. Advis. Sec. Sci. Resp. 2023/009.