Stock assessment of silky shark in the Western and Central Pacific Ocean 2024

This analysis assesses the southwest Pacific silky shark (Carcharhinus falciformis) stock in the Western and Central Pacific Ocean (WCPO). This is the third attempt at undertaking an assessment of Pacific silky sharks, with one previous attempt covering the entire Pacific (Clarke et al. 2018a), and the other for the WCPO area only (Rice 2013b).

There are no target fisheries for silky sharks in the WCPO, and they are caught as bycatch in longline and purse seine fisheries. Since 2015 all silky sharks caught in fisheries managed by the Western and Central Pacific Fisheries Commission (WCPFC) are required to be released (WCPFC 2013). While silky sharks have been caught in Pacific fisheries since their inception in the 1950s, they have only been reported in catch records since the 1990s (Brouwer et al. 2023, Neubauer et al. 2023a). A catch history is required to be estimated prior to an assessment being undertaken because of the unreliability of both logsheet and observer data, due to: the use of generic reporting codes prior to 2015; the lack of logsheet reporting of bycatch data; and, poor and inconsistent (in time and space) observer coverage for most flags in the Pacific Ocean. However, general data improvements in recent years, along with the availability of biological data and the successful previous stock assessments led Brouwer and Hamer (2020) to conclude that a data rich assessment should be undertaken for this stock.

This assessment took a multi-model approach to assessing silky shark in the WCPO, resulting from large uncertainties in the underlying data and difficulties with fitting of integrated stock assessments for sharks generally. In an effort to understand overfishing risk to silky shark based on different lines of reasoning, a range of models were applied, with varying degrees of complexity and with different data requirements,including a fully integrated stock assessment in Stock Synthesis, and three alternative assessments: length and age structured assessment model (LAM), dynamic surplus production model (DSPM), and a length-based spatial risk assessment (SRA). These approaches were not strictly standardised, with no attempt to use consistent priors between them, and treating each approach as standalone to provide evidence from independent lines of reasoning. However, with a single set of data available, data inputs were standardised across all four assessment approaches.

Results across all assessment methods are largely in agreement that recent fishing mortality was low with respect to biological reference points for sharks. Therefore, while each modelling approach has limitations, we suggest that the weight of evidence is sufficient to conclude that fishing mortality has declined substantially for silky shark in the most recent decade, and that recent stock status is likely improving from previous low levels. We further suggest that the dynamic surplus production models provide the most parsimonious and robust available model for management advice. This suggestion derives from the consistency with other outcomes, but also because of stronger reservations about the robustness of other approaches. The stock synthesis assessment provided a number of challenges and some issues could not be resolved. We consider both the risk assessment and LAM models experimental at this stage, with further work required to use these model for management advice

Main assessment conclusions

• The multi-model approach to assessing silky shark resulted in an uncertain stock status, but high confidence that recent fishing mortality is below levels that would preclude stock rebuilding.

• Based on considerations of model complexity, fit and estimation issues, we suggest that the dynamic surplus production model be used for providing management advice. We further suggest that the model ensemble across initial depletion priors may be over-representing uncertainty, and we suggest output from the intermediate assumption as a candidate model for management advice.

• The largest fishing mortality was estimated to have come from longline fisheries capturing nearly the full size-range of silky sharks, and reductions in interactions as a result of changes in fishing practices over the last decade may have substantially reduced this source of mortality, allowing the stock to rebuild.

• The stock status has been improving since 2010, and the recent fishing mortality rates are below biological reference points for the ensemble (Diagnostic F_recent/F_crash: 0.13 [0.01–0.25]; P(F_recent/F_crash >1)=0; P(F_recent/F_lim >1)=0).

With respect to other shark stocks with lower levels of information:

• Surplus production models should be used when composition data are problematic, and risk assessments should only be used in the context of prioritisation across species. However, alternative, data driven risk assessments could provide an avenue to use a risk assessment approach when time-series are not sufficiently long or reliable to estimate fishing mortality rates relative to biological reference points.

• With non-retention measures leading to sharks and other bycatch species being cut free from longlines, data collection may be problematic. In these cases, the only alternative to provide estimates of risk would be methods such as EASI-fish, which estimate risk from assumptions about fishery overlap with the species distributions and vulnerability to fishing effort.

Given some of the fundamental uncertainties highlighted above, we recommend:

• Additional tagging should be carried out using satellite tags in a range of locations as well as high seas areas to resolve fundamental questions about the species interactions with local oceanography and the dynamics of ENSO. Such tagging may help to resolve questions about the degree of natal homing and limited mixing of the stock, as suggested by genetics. This work is currently scheduled as part of the SRP as work theme 3 (a)vii and is due to begin in 2025 (Brouwer and Hamer 2024), and it is recommended that CCMs prioritise this work and ensure that satellite tagging options are included in the project specification.

• Additional growth studies and validation of aging methods from a range of locations could help build a beĴer understanding of typical growth, as well as regional growth differences. Current growth data are conflicting, with insufficient data to understand the underlying process. This work is currently scheduled as part of the SRP as Project P19X11 and is due to begin in 2025 (Brouwer and Hamer 2024), and it is recommended that CCMs prioritise this work.

• Additional genetic/genomic studies across a broader set of locations could be undertaken to augment the tagging and existing genomics work to help resolve the stock/sub-stock structure patterns. To support this work, a strategic tissue sampling program for sharks is recommended with samples to be stored and curated in the Pacific Marine Specimen Bank. This work is currently scheduled as part of the SRP as work theme 3 (a)xii and is due to begin in 2026 (Brouwer and Hamer 2024), and it is recommended that CCMs maintain this project in the current work stream.