Science Response : Rapid Status Assessments for Pacific Salmon
Bibliographic record
Abstract
Regular tracking of the state and distribution of salmon biodiversity is increasingly important in a changing climate. Broad declines in Canadian Pacific salmon abundances have been linked to global climate change and other factors such as deteriorating habitats, increased fish disease, and invasive species (Grant et al. 2019). To track salmon biodiversity change, we present a Wild Salmon Policy (WSP) rapid status assessment approach for Pacific salmon. This approach can assign a Red, Amber, or Green status, with High, Medium or Low confidence to salmon conservation units (CUs) with applicable data. Pacific salmon adaptive diversity occurs at a range of scales that include the species, CU, population and deme. The WSP identifies diversity at the scale of CUs, which are fundamental units that cannot be recolonized if lost (DFO 2005; Holtby and Ciruna 2007; Wade et al. 2019). Fisheries and Oceans Canada (DFO)’s WSP covers five species of Pacific salmon: Sockeye (Oncorhynchus nerka), Chinook (O. tshawytscha), Coho (O. kisutch), Pink (O. gorbuscha) and Chum Salmon (O. keta). DFO has the authority to manage these salmon under the Fisheries Act (2019). Steelhead (O. mykiss) are managed provincially, and therefore are not included in WSP rapid status assessments. This Canadian Science Advisory Secretariate (CSAS) review of the WSP rapid status assessment approach was requested by DFO Science Branch to support the evaluation of Pacific salmon Stock Management Unit (SMU) statuses relative to their Limit Reference Points (LRPs). An SMU defines a group of one or more Pacific salmon CUs that are managed together with the objective of achieving a joint status. The LRP represents the status below which serious harm is occurring to the stock, based on biological criteria established by DFO Science through peer review. An SMU below its LRP triggers a rebuilding plan. A recent CSAS process recommended that LRPs for SMUs be defined as a percentage, with the objective being that 100% of all CUs in the SMU are above the WSP Red status zone (DFO 2023; Holt et al. 2023a, 2023b). An SMU falls below the LRP if one or more CUs in an SMU are in the WSP Red status zone. The WSP rapid status approach was recommended for assessing LRP status (DFO 2023; Holt et al. 2023a). Subsequently through the current report’s CSAS process, a recommended next step is the vetting of the individual CU WSP rapid status results, and LRP status determination, by experts in a structured process. Existing WSP integrated status assessments provide a foundational approach to tracking annual salmon CU status. This approach uses an expert decision-making process to combine statuses across individual WSP metrics, and additional related information, into a single integrated status. However, the WSP integrated status assessment approach only gets us part way to tracking annual CU status, since it is time- and labor-intensive, and as a result, has only been completed for 11% of the current 377 CUs, and is 5–10 years out of date. To expand the number of CUs assessed, and provide annual CU status updates, this paper presents a new WSP rapid status approach that approximates the expert decision-making process used in the integrated status assessments. Annual WSP rapid statuses are estimated using an algorithm implemented with computer code for British Columbia (BC) and Yukon CUs with applicable data. The WSP rapid status approach provides more complete coverage of WSP statuses across CUs. Expanding the number of assessed CUs will require input from stock assessment experts to select appropriate escapement enumeration sites and years, and to perform data treatments such as gap filling as applicable. Experts would work iteratively to explore specifications for use with the WSP rapid status algorithm, such as identifying applicable WSP rapid status metrics for these data, and reviewing the WSP rapid statuses generated by the algorithm to finalize the approach for their CUs. The establishment of a governance strategy for this work is recommended, including the identification of roles and responsibilities, to ensure the inclusion of new CUs, and annual updates across CUs. The WSP rapid status approach is integrated into DFO’s Pacific Salmon Status Scanner. DFO’s Salmon Scanner is an interactive data visualization tool specifically designed for experts to support scientific exploration and help them incorporate science into decision-making processes. Experts are those with expertise on Pacific salmon including stock assessment biologists, Indigenous technical experts, research scientists, habitat, harvest, and hatchery management biologists, etc. The objectives of this Science Response are to: 1. Summarize the methods, results, and conclusions of the WSP rapid status approach. The development of this approach included three key components: a. a performance evaluation of candidate WSP rapid status algorithms against existing CSAS reviewed WSP integrated statuses; b. an evaluation of the application of the rapid status algorithm to years and CUs that currently do not have WSP integrated statuses completed; c. a measure of confidence in WSP rapid status results. 2. Document the review processes that have occurred to develop the rapid status algorithm. 3. Provide advice on next steps and future work. This Science Response Report results from the regional peer review of November 18, 2022 on the Rapid status approximations for Pacific salmon derived from integrated expert assessments under Fisheries and Oceans Canada Wild Salmon Policy.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.018 | 0.041 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.005 | 0.002 |
| Scholarly communication | 0.004 | 0.003 |
| Open science | 0.004 | 0.005 |
| Research integrity | 0.017 | 0.016 |
| Insufficient payload (model declined to judge) | 0.067 | 0.029 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".