4.1 Primary Performance Metrics
Resolution 24-09 (ICCAT 2024) describes the initial operational management objectives for Southern Atlantic Albacore:
Stock Status. The stock should have a 60% or greater probability of occurring in the green quadrant of the Kobe matrix over a 30-year projection period: i.e. \(SB > SB_{MSY}\) and \(F < F_{MSY}\) simultaneously.
Safety. There should be no greater than 15% probability of the stock falling below \(B_{LIM}\) at any point during the 30-year projection period, using an interim \(B_{LIM}\) of 40% of \(SB_{MSY}\). So that higher values are better for every PM, this is expressed as the probability of the stock remaining at or above \(B_{LIM}\) throughout the projection period, with a target of at least 85%.
Yield. The objective is to maximise overall catch levels. As the Resolution does not specify a particular statistic, the mean annual catch is used.
Stability. Any change in total allowable catch (TAC) between management periods should be 20% or less, noting that asymmetric stability limits (e.g. different thresholds for TAC increases versus decreases) may also be evaluated. This is expressed as the probability that the change in TAC between successive management cycles is less than 20%; i.e. calculated between the TAC values actually set every 3 years.
Table 4.1 summarises these objectives together with the corresponding performance metric and the associated target. The equation for each performance metric is given below the table.
| Objective | Performance Metric | Target |
|---|---|---|
| Stock Status | Probability of occurring in the Kobe green quadrant | ≥ 60% |
| Safety | Probability of remaining at or above \(B_{LIM}\) (40% \(SB_{MSY}\)) throughout the projection | ≥ 85% |
| Yield | Mean annual catch | Maximise |
| Stability | Probability that the change in TAC between management periods is < 20% | Ranked |
Each PM can also be calculated for an individual OM or across a set of OMs (e.g. the Reference Set). In the latter case, the simulations from every OM in the set are first combined into a single pooled MSE object and the summary statistic (e.g. a probability or a mean) is then calculated over that combined set of simulations.
The performance metrics in Table 4.1 are calculated as follows:
Stock Status: \[ \text{Prob}\left(SB/SB_{MSY} > 1 \text{ and } F/F_{MSY} < 1\right) \]
Safety: \[ \text{Prob}\left(\min_y(SB_y/SB_{MSY}) \geq 0.4\right) \]
Yield: \[ \text{mean}_y(C_y) \]
Stability: \[ \text{Prob}\left(\lvert (TAC_c - TAC_{c-1})/TAC_{c-1}\rvert < 0.2\right) \]