• Predict and Act: Walters and Hilborn (1976) pointed out there was, and still is, a tendency to observe disturbed systems, synthesise elaborate models, and conduct optimisation exercises which pretend that our actions will not affect the way we observe and learn in the future.
• Learn and Adapt: In natural resource management the importance of adaptive management to reduce the risk of failing to achieve management objectives due to uncertainty has long been recognised (Walters and Hilborn 1978). Adaptive management learns by doing, so that policies evolve as new observations and information become available (Walters and Holling 1990)

• Stock assessment requires providing robust advice on the status of stocks and their response to management.
• Ideally, advice should follow the principle of risk-equivalence so that the risk of depleting a stock below a limit or not achieving a target is independent of the stock, assessment method or the amount of data and knowledge available.

• Risk equivalence means that the higher the uncertainty, the lower should be the level of exploitation for a given probability of achieving targets and avoiding limits.
• There is, therefore, an economic, and quantifiable, benefit in investing in better data collection, scientific study and monitoring control and surveillance.

The six steps of Adaptive Management are

1. Problem Assessment
2. Design
3. Implementation
4. Monitoring
5. Evaluation, and
6. Adjustment.

A simulation approach that helps in the Design of robust management strategies to ensure sustainability based on ecological, social and economic objectives (Sharma et al. 2020).

• MSE (Punt and Donovan 2007) models feedback control rules, where as a system responds to management the advice also changes.
• Strategies are modelled as Management Procedures (MPs), simplified representations of the processes of collecting and analysing data and harvest control rules (HCRs) used to set catch limits.
• Instead of attempting to solve for an optimal management policy in the space of all possible sequences of actions a manager could take, MSE evaluates the consequences of a pre-determined set of strategies evaluated based on management objectives (Memarzadeh and Boettiger 2018).

• The first step is Problem Assessment i.e. define the aims and the objectives
• MPs are then developed, where each component is formally specified that represent the monitoring data, analysis method, harvest control rule and management measures.
• MPs are then simulation tested using a Operating Model that represents the main uncertainties about stock and fleet dynamics, prior to Implementation

Six steps

1. Identification of management objectives;
2. Selection of hypotheses for the OM;
3. Conditioning the Operating Model based on data and knowledge;
4. Identifying candidate management strategies;
5. Running the Management Procedure as a feedback controller to simulate the long-term impact of management; and then
6. Identifying the Management Procedures that robustly meet management objectives.

Adaptive management and MSE both consider feedback, and MSE helps in the first two steps of Adaptive Management, i.e. 1) Problem Assessment and 2) Design. By identifying objectives, and designing strategies to achieve them.

Once implemented an MP requires less effort than conducting a stock assessment each time advice is required. This should allow time to be dedicated to a better understanding of resource dynamics, and so helping to move away from Predict and Act towards Learn and Adapt.

After an MP has been implemented a review should be conducted to evaluate whether objectives have been achieved, has MP performed as designed, and how can improvements be made.

Screen potential Emprirical Indicators

• Condition Operating Model
• Run Operating Model for a variety of historical scenarios
• Generate empirical indicators using Observation Error Model
• Use ROC curves as a filter for indicator & reference levels

1. Develop Operating Model

2. Run OM without feedback to generate empirical indicators using OEM, use ROC curves to identify best indicator

3. Run MSE to evaluate MP based on best empirical indicator

4. Perform implementation review, e.g. based on an assessment

5. Compare OM under MP and assessment results

6. Evaluate risk equivalence and value-of-information i.e. for

   • Different lengths of time series
   • Quality of data
   • Quality of priors, i.e. fishbase, stock specific

Run for a range of trajectories, i.e. sustainable, declining, recovering stocks

Generate empirical indicators

Effort has been set by licences, so we set catches consistent with the implied F using a HCR

• Monitoring
• Evaluation, and
• Adjustment.

After implementation need to check that the MP has achieved objectives

• Therefore, after 5, 10 years assess the stock using JABBA
• Compare OM and assessment estimates

• Different lengths of time series, i.e. years after implementation
• Quality of data
• Quality of priors, i.e. fishbase v. stock specific