Benchmark Workshop on Atlantic Salmon (Salmo salar) in the North Atlantic (WKBSALMON)

Update:

March 2024: Report updated with additional working document (WD10-Retrospective-patterns-LCM-2024) and explanatory paragraph in Executive Summary

WKBSalmon reviewed the implementation of a Life Cycle Model (LCM) for wild anadromous Atlantic salmon (Salmo salar L.) covering their natal north Atlantic range. The LCM is a time iterative, Bayesian hierarchical model incorporating salmon records of fifteen countries at 25 stock-units. It tracks salmon of two explicit sea-age streams, namely, one-sea-winter (1SW) and multi-sea-winter (MSW), stock unit specific smolt ages, numbers of salmon returning to stock-units, proportions maturing, survival at sea by month and stock unit specific post-smolt survival rates and proportion maturing at 1SW. Mixed-stock catches at West Greenland and Faroes, as well as those in North America, are designated to stock-units based on observed historic tag data, genetic identification and assumed harvest distributions.

The LCM will replace three Pre-Fisheries Abundance (PFA) forecast models, aligned to three management units, one eastern North America and two Northeast Atlantic European complexes of stock-units. The LCM enables a more comprehensive and consistent approach, accounting for migration and maturation of salmon by stock-unit and a hierarchical (over stock-units) modelling of post-smolt survival and proportion maturing in the first year at sea.

The LCM uses outputs from two “Run Reconstruction” models, one for each of eastern North America and Northeast Atlantic origin salmon. These process catch data and exploitation rates and/or returns at stock-unit spatial scales to estimate returning numbers and catches of salmon by sea-age group. The LCM model uses a similar sea-age group structure for all stock-units resulting in a harmonized life cycle for Atlantic salmon from the North Atlantic.

The LCM forecasts estimates of returning salmon by stock-unit based on the post-smolt survival and proportion maturing parameters, forecast forward as a random-walk, from the most recent observations and accounting for “banked” maturing and non-maturing salmon. Forecast returns to stock-units may be compared to Conservation Limit (CL) reference points and “Spawner Escapement Reserves” (SERs – reference points prior to any marine fishing activities) at national and international levels to quantify the risk to the salmon stocks under different mixed-fisheries catch levels.

The LCM was found to provide estimates of stock status and forecasts in line with perceptions and previously used modelling frameworks and to be robust to a range of settings and uncertainties. Subsequent to the meeting, retrospective patterns of the primary model variables (Total PFA, Maturing PFA, Non-maturing PFA, Post-smolt survival and Probability of maturing as 1SW) for all stock units were investigated using Mohn's rho and time series graphing using the full time series and five "peels" each produced by running the assessment model with a further proceeding year's data removed. These also showed the model to be stable, with retrospective patterns with acceptable bounds (ICES, WKFORBIAS; 2020). This work is added to the report as a standalone document listed in Annex 2 (Rivot and Dubost, 2024).