The objective of this theme is to develop ways of improving the dialogue between managers and scientists; theme leaders are Graham Pilling and Campbell Davies.
Conducting a Management Strategy Evaluation (MSE) is an iterative process requiring regular dialogue between managers and scientists. First management objectives have to be established and the main uncertainties agreed, then the risks of failing to achieve them evaluated, and finally the management strategy that robustly meet the management objectives identified. The tRFMO, however, have different institutional structures for dealing with communication between commissioners and scientists. This theme will summarises the processes adopted by the tRFMOs when conducting MSE and highlight how the dialogue can be enhanced.
Conditioning of Operating Models
To conduct MSE it is necessary to build a simulation model, i.e. an Operating Model (OM) that represents the uncertainty about system dynamics; theme leaders are Toshi Kitakado,Mark Maunder and Rishi Sharma.
Conducting a requires six steps; namely i) identification of management objectives; ii) selection of hypotheses for the OM; iii) conditioning the OM based on data and knowledge, and possible weighting and rejection of hypotheses; iv) identifying candidate management strategies; v) running a Management Procedure (MP) as a feedback control in order to simulate the long-term impact of management; and then vi) identifying the MP that robustly meet management objectives.
The choice of hypotheses and how to weight them is critical, since once the OMs are chosen the "best" management procedure is determined, the task is to find it. There are many alternative ways to condition an OM. One approach is to use an Integrated stock assessment model like SS or Multifan-CL. The use of an assessment model as the OM implies that they can describe nature almost perfectly. If an assessment model can describe nature why bother with MSE? However, if a management procedure can not perform well when reality is as simple as implied by an assessment model it is unlikely to perform adequately for more realistic representations of uncertainty about resource dynamics. Basing an OM on the current assessment model also has arguably the lowest demands for knowledge and data and allows RFMOs to make a phased transition from the stock assessment paradigm to a risk based approach. There are many important processes, however, that are not modelled in stock assessments and affect the robustness of control systems. Therefore to ensure a control system is robust also requires OMs to be conditioned based on expert beliefs and other a priori information about the processes that may affect the behaviour of management systems in the future. I.e. the focus is on the future, not on fitting historical data as when conditioning an OM on a stock assessment. This is a less data, and more hypothesis-orientated approach.
Under this theme the intention is to collaborate on method development and in particular sharing of code, cloud computing and parallel processing. The theme leaders are Anders Nielsen and Vaughan Pratt.
A lot of fishery science, particularly conducting MSE, requires time spent building and using software, however, most of us have never been formally taught how to do this efficiently. As a result, we may be unaware of tools and practices that would allow us to write more reliable and maintainable code with less effort. If software is not version controlled, readable, and tested, the chances of being able to re-create results are remote. So how can we begin to adopt tools and approaches that will improve both the quality of software and the efficiency with which it is produced?
Albacore Case Study
To help develop practical examples it is intended to focus on albacore stocks worldwide, the leaders are Iago Mosqueira and Laurie Kell.
The work of the group requires various electronic tools and resources, e.g. a glossary of terms, bibliography, a list of publications, etc.
The theme leaders are Abdelouahed Ben Mhamed and Carolina Minte-Vera
One activity of this theme will be to provide an inventory of software used for conditioning OMs, implementing MPs, running and summarising MSE outputs. A summary of software tools can be found at https://rfmo-mse.github.io/. In particular to collaborate on tools for presenting how well different strategies achieve management objectives. This is important since management objectives are similar across the RFMOs and there is a common advice framework. Collaborating on a common tool (e.g. a shiny app shinyapps.io/tunamse/ work require less work and make it easier for scientists and managers to communicate.
As part of the work of the group a folio of collaborative scientific papers will be developed. These will be both for the scientific committees of the tRFMOs and for peer review journals. These will describe case studies, novel methods and applications and ensure independent scrutiny by the scientific community. The folio will initially including tentative individual paper titles, an overview of each paper as well as its goal and a clear statement of the issues being addressed. The papers will be written intersessionally by interested members of the group.
Communication between scientists, policy makers, and stakeholders is vitally important. In order to facilitate dialogue various bodies have developed glossaries that attempts to define and explain the terminology used. A first step in improving dialogue glossaries from the tRFMOs, fisheries management bodies and from other relevant fields (e.g. Risk management and control theory) will be collated.
To help in communication the grey literature from the tRFMO scientific committees will be compiled into an online bibliography.
Data on growth, life history, tagging and catches are important for conditioning Operating Models, a proposal for creating a joint tRFMO database will be prepared.
It was also recognised that it was important to include individuals who could bring a “developing country” perspective to the group, and people from outside the tRFMOs to fill knowledge gaps, e.g. relation to computer science, control theory, risk and ecology.
The experts will provide a review of feedback control systems, modelling and risk management in fields other than fisheries management science for the meeting. Ttey also help to peer-review the work of the group.