Science Response 2022/001

Update of the 2019 Bocaccio (Sebastes paucispinis) stock assessment for British Columbia in 2021

Context

Bocaccio rockfish (BOR) was assessed in 2019 (DFO 2020) using the Awatea catch-at-age stock assessment model platform, tuned to six fishery-independent trawl survey series (covering 1967-2019), a standardised commercial bottom trawl catch per unit effort (CPUE) series (1996-2012)^{Footnote 1}, annual estimates of commercial catch from two fisheries, and age composition data from the combined commercial trawl fishery and four of the six surveys. The British Columbia (BC) outside coast model started from an assumed equilibrium state in 1935. Three component base runs using a two-sex model were implemented in a Bayesian framework (using the Markov Chain Monte Carlo [MCMC] procedure) under scenarios that fixed natural mortality (M) to three levels (0.07, 0.08, 0.09) with the accumulator age (A) set to 50 years while estimating steepness of the stock-recruit function (h), catchability (q) for surveys and CPUE, and selectivity (µ) for four surveys and the commercial trawl fleet. These three runs were combined into a composite base case which explored the major axis of uncertainty (M) in the stock assessment. A summary of the 2019 stock assessment as well as descriptions of earlier Bocaccio stock assessments (including reviews of this species by the Committee on the Status of Endangered Wildlife in Canada) can be found in the Science Advisory Report 2020/025.

The 2019 stock assessment depicted a coastwide BOR stock experiencing a nearly continuous decline from the start of the population reconstruction in 1935, interrupted only by a period of arrested decline spanning the years 1970-86 resulting from a few moderate recruitment events in 1969, 1976, and 1978. The decline resumed in 1987, continuing until an extremely large recruitment event occurred in 2016, estimated by the model to be 44 times the long-term average recruitment (5% and 95% quantiles: 30 times, 58 times).

The estimated size of the initial biomass was large (median = 32,300 tonnes; 5% and 95% quantiles: 16,500 t, 71,700 t) in order to sustain the catch history while assuming deterministic recruitment during the initial years of the reconstruction. The size of this biomass implied an estimated median maximum sustainable yield (MSY) of 1,461 tonnes (703, 3,623), compared to the average catch over the final five years (2015-2019) of the reconstruction of only 69 tonnes. The size of the spawning biomass at the end of the final year (2019) relative to the unfished equilibrium spawning biomass (B₂₀₂₀/B₀) was very low, estimated at 0.028 (0.013, 0.058), and B₂₀₂₀ relative to equilibrium spawning biomass that would support the MSY (B₂₀₂₀/B_MSY) was 0.096 (0.042, 0.23). Neither of these estimates included a contribution from the 2016 year class because it was assumed that the first four age classes had no mature females. The equivalent ratio at the end of 2019 for the biomass vulnerable to trawl gear (V₂₀₂₀/V₀) was 0.10 (0.050, 0.21). This ratio is higher than for the spawning biomass because the vulnerable biomass was determined by the selectivity function without reference to maturity status and, consequently, included a fractional contribution from the large 2016 cohort.

The estimated exploitation rates (u_t) throughout the time series were low, with the 2019 exploitation rate estimated at u₂₀₁₉ = 0.025 (0.021, 0.044) for the ‘trawl’ fishery and u₂₀₁₉ = 0.0097 (0.0005, 0.0016) for the ‘other’ (non-trawl) fishery. These exploitation rates were much lower than the range of natural mortality rates (0.07 to 0.09) considered plausible based on the available ageing data. Even the maximum exploitation rates estimated by the reconstruction (median for ‘trawl’ = 0.06 and ‘other’ = 0.01) barely entered the plausible range of M. Exploitation at levels substantially less than natural mortality are often thought to be sustainable, but the extremely low status of the 2019 relative biomass estimated by the model attested to the lack of recruitment as well as the success by management to keep exploitation to a minimum level in a mixed species trawl fishery.

The composite base case spawning biomass at the beginning of 2020 was estimated to be above the limit reference point (LRP) with probability P(B₂₀₂₀ > 0.4B_MSY) < 0.01, and above the upper stock reference (USR) point with probability P(B₂₀₂₀ > 0.8B_MSY) = 0 (i.e., no probability of being in the Healthy zone based on the set of MCMC posterior samples). However, a very quick recovery was projected based on the extremely large size of the 2016 cohort.

The 2019 stock assessment predicted that there would be a quick rebound in the spawning biomass starting in 2021 because, even though only a small proportion of the 5-year old fish would be mature, the large size of the 2016 cohort should contribute a detectable increase to the spawning population. Model projections predicted that, by the beginning of 2022 and assuming a constant catch of 200 tonnes/year or a harvest rate of 0.04/year, the BOR stock would move into the Cautious zone (i.e., the median was near the USR of 0.8B_MSY). The stock assessment further projected that the probability of the spawning biomass exceeding the LRP, i.e. P (B_t > 0.4B_MSY), was greater than 95% in one more year (i.e. at the beginning of 2023). As a sensitivity analysis, the projections were repeated based only on the lowest 5% of the posterior distribution of the composite base case, resulting in rebuild predictions that were only delayed by 2-3 years. Finally, the 2019 stock assessment noted that both the Queen Charlotte Sound (QCS) and west coast Vancouver Island (WCVI) synoptic surveys showed good monitoring capability for this species and suggested that corroboration of the size of the rebuild would be available when these surveys were repeated in 2020^{Footnote 2} and in 2021.

Because of the apparent demonstrated capacity of the synoptic surveys to monitor Bocaccio, the 2019 Regional Peer Review (RPR) meeting agreed that it would be feasible to check the progress of the rebuild through an update of the 2019 stock assessment once the 2020 and 2021 information from the synoptic surveys ^{Footnote 3} used in the model became available (DFO 2020). These observations, once processed through the agreed stock assessment, should corroborate the progress of the rebuild. The 2019 RPR also noted that, although no new age data would be available in 2021, it was likely that the 2016 cohort would be identifiable from the length frequency data in the three surveys, given the large relative size of this cohort. It also noted that there was a possiblity that new recruitment, if sufficiently large, would also be identifiable in these data.

This Science Response results from the Science Response Process of October 28, 2021 on the Stock Assessment Update of Bocaccio Rockfish (Sebastes paucispinis) for British Columbia in 2021.

Science Response 2022/001 (PDF, 3,932 KB)

Accessibility Notice

This document is available in PDF format. If the document is not accessible to you, please contact the Secretariat to obtain another appropriate format, such as regular print, large print, Braille or audio version.