Homepage for DFO's State of the Salmon Program.
WSP Rapid Status Assessments
The WSP rapid status assessment approach uses a decision tree to assess status (Pestal et al. 2023, DFO 2024). The decision tree is a set of consecutive decision rules that were commonly used by experts in previous WSP integrated status assessment processes to assess WSP statuses. The decision tree was built using Classification and Regression Tree (CART) analyses, in combination with common rationale or rules extracted from the CU-specific status narratives from previous WSP integrated status processes (Grant & Pestal 2012, DFO 2015, DFO 2016, Grant et al. 2020).
The resulting decision tree thus approximates the expert decision-making process to assess WSP statuses. Decision rules are sequentially applied to the WSP metric values calculated for each CU being assessed. The decision tree pathway taken by each CU in the year being assessed depends on which metrics are available and how the CU values compare to the metric’s thresholds. At the end of each pathway, the decision tree assigns a Red, Amber, or Green status, with High, Medium or Low confidence.
Confidence is determined based on the metrics used assess status, and their values. Statuses determined using abundance-based metrics (relative abundance and absolute abundance) are generally associated with higher confidence than those that are assigned using only trend-based metrics (long-term trend and percent change).
The WSP rapid status assessment approach requires expert input to prepare data, determine how metrics are calculated, and review resulting statuses. However, metric calculation, and the labour-intensive process of integrating metrics into statuses are automated, significantly streamlining the overall process.
.png)
Figure above: The WSP rapid status decision tree.
-To assess a CU’s WSP rapid status, the first step is for CU experts to determine whether or not there are applicable and sufficient data available to assess status using relative abundance metrics (that have thresholds estimated for the CU) or trend metrics (shared thresholds are already available across CUs for each metric). If the data are not considered applicable or sufficient, the CU status is classified as data deficient.
-If there are applicable and sufficient data to assess WSP rapid statuses, then each of the annual metric values are compared to thresholds presented at each decision point. Yes or No answers split each path of the decision tree, terminating at WSP rapid status assignments of Red, Amber or Green status, with High, Medium or Low confidence rating. The different splits are identified as nodes: 1 to 65.
-The first question (node 1) in the WSP status decision tree is Can the absolute abundance metric be used and is the current generation average less than 1,500. If the answer is yes to both these questions (two green checkmarks), then the WSP status is Red with High confidence (node 3).
-If the answer is no to one or both of these questions (one or two red X’s) then you move to the next question (node 2): Can the absolute abundance metric be used and is the current generation average less than 10,000. If the answer is yes to both these questions (two green checkmarks), then the next group of questions is addressed in Pathway 1 (starting at node 4), where WSP statuses are restricted to Red or Amber statuses. If the answer is NO to one or more of these questions (or or two red X’s), then the next group of questions is addressed in Pathway 2 (starting in node 5), where WSP statuses can include Red, Amber or Green statuses.
-In both Pathway 1 or Pathway 2, questions on whether you have unique relative abundance thresholds (Sgen/Smys; percentile; carrying capacity) for the CU, and applicable data that can be compared to these thresholds, then the relative abundance metric values are compared to metric thresholds to determine WSP rapid status, ranging from Medium to High Confidence. If there are no realative abundance thresholds, then trend metrics are applied to evaluate WSP rapid status. Trend metrics produce lower confidence WSP rapid statuses.
Three Core Principles of the WSP Rapid Status Approach
-
Methods to identify Conservation Units and assess rapid statuses comply with conservation biology principles (Mace and Lande 1991; Mace et al. 1992, 2008; Caughley 1994; National Research Council (US) Committee on Scientific Issues in the Endangered Species Act 1998; McElhany et al. 2000; Rodrigues et al. 2006), and are aligned with scientific peer-reviewed publications (Holtby and Ciruna 2007; Holt 2009; Holt et al. 2009; Holt 2010; Grant et al. 2011; Grant and Pestal 2012; DFO 2015, 2016; Brown et al. 2019; Grant et al. 2020). This ensures that Pacific salmon statuses are scientifically objective, consistent, and comparable across CUs.
-
Data are vetted and WSP rapid statuses are evaluated by CU experts. DFO stock assessment leads work in collaboration with Indigenous groups, consultants, and others that support or lead salmon stock assessment programs to fine tune the CU data used (determining appropriate escapement sites, years, data treatment, etc.), select applicable WSP metrics, and determine specifications for metric calculations.
-
The WSP rapid status decision tree is continually refined as new information becomes available. This means that data sets and status metrics for each CU will be regularly reviewed and updated, and the decision tree will be reviewed through on-going work with CU experts. Experts can identify where decision rules may require revision, or where additional decision rules may be needed as new CUs are assessed. In any such case, decision tree performance will be tested for overall improvements.
Benefits
The WSP rapid status approach has the ability to provide more complete coverage of WSP statuses across CUs in B.C. and the Yukon, and can be regularly updated. Expanding the number of assessed CUs will require input from stock assessment experts to prepare input data and review/verify the results. However, overall the time commitment required to assess WSP statuses has been greatly reduced through implementation of this decision tree and the data processes that surround it. The WSP rapid status approach, similar to the WSP integrated status processes, ensures WSP status assessments are science-determined (objective), consistent, and comparable between CUs and years.
WSP Rapid Statuses are uploaded into DFO’s Pacific Salmon Status Scanner once fully reviewed. The Scanner is an interactive data visualization tool specifically designed for experts to support their work on Pacific salmon.