Modelling social-ecological systems to improve Atlantic salmon conservation

The use and application of models has become widespread and indispensable in ecology and conservation science. Such models cover a broad spectrum of applications ranging from improved understanding of the likely effects of climate change on biodiversity, to supporting the decisions made in natural resource management. Given the continued rapid global loss of biodiversity, understanding the mechanisms and consequences of conservation decisions is vital for long-term sustainability. Although a number of drivers of biodiversity loss have been identified, one of the most prevalent and widespread is human exploitation of habitats and natural resources (e.g. through hunting, fishing or habitat loss). Because natural resource use is fundamentally driven by humans, accurately predicting future sustainability is reliant as much on understanding human decision-making as it is on understanding ecological dynamics themselves. Thus, the development of social-ecological models that account for the interaction between natural resource dynamics and human decision making is becoming increasingly urgent.

Cutting-edge modelling approaches have made significant progress towards modelling complex social-ecological systems, but their increased complexity poses two interlinked challenges. First, models are often difficult to communicate clearly to non-specialist audiences, such as resource managers and other stakeholders. Thus, frequently the evidence for practical uptake of many models is limited. Second, their complexity implies the need for extensive data to parameterise them effectively and to build the linkages between the ecological and the decision-making component. In terms of social-ecological management systems, while data to parameterise the ecological components are often relatively easily available from the literature, the human decision-making components are often based on limited theory and lack a general empirical basis. To maximise the adoption and use of complex social-ecological models to inform decision making, both appropriate representation of human decision-making, and effective communication, are therefore key.

This PhD project will use management of wild Atlantic salmon populations as a case study with the support of the Atlantic Salmon Trust, a key partner in the Missing Salmon Alliance ( https://missingsalmonalliance.org/ ). The flagship project of this alliance is a programme of works (Bull et al., 2022) to improve the future management of salmon populations and their habitats.

With declines in abundance, and the myriad effects of climate change, managers of wild Atlantic salmon populations face growing challenges when directing their conservation actions and deciding what management actions are most effective in a complex system with many stakeholders, different habitats and under climate change. Often the highly technical outputs from salmon ecological research and modelling do not adequately match the complexity of human decision making models. Developing interlinked social-ecological models of human decision making and ecological dynamics will not only benefit salmon management but will in more general promote the development of effective tools and guidance for helping to tackle complex environmental management challenges.

The aim of this PhD project is to build interlinked social-ecological models of salmon and the relevant stakeholders and evaluate the information needs of salmon stakeholders. The outcome of the project is a co-developed prototype decision support tool for Atlantic salmon management.


The project will use a social-ecological modelling approach that is based on the Generalised Management Strategy Evaluation (GMSE) framework (and GMSE R app) that combines agent-based modelling of human decision making with salmon population modelling.

As a first step the project will determine what kind of information the stakeholders need to adopt models to make better decisions. The PhD will use stakeholder workshops and focus group discussions to produce a compendium of stakeholder needs (sometimes common needs, sometimes conflicting across stakeholder groups) and to create a conceptual GMSE model.

The next step in the PhD project will be to explore the information available, the influence of management and how it interacts with the salmon populations and their habitat in order to make connections between the modelling components of GMSE, i.e. the social and ecological parts of the model. These approaches will tackle the need to link different underpinning models of existing salmon and habitats as well as the social and economic drivers of the system.

The next step for the second part of the PhD will consist of a process of implementing the information into the GMSE model and run the models in order to produce management outputs and predictions for salmon management. Here, the project will use elements of co-development of models together with the stakeholders. This represents a departure from model development conducted in isolation by technical experts, which can subsequently result in poor representation of and uptake by end-users because of complex and inaccessible design, to one of joint ownership in the design of an engaging and user-friendly model. A prototype of this model will take the form of an interactive R Shinyapp.

This project will bring together ecology, social science, modelling and stakeholder engagement in an interdisciplinary way that will improve the link between state-of-the-art scientific modelling and real-world management of natural resources.

Project Timeline

Year 1

Review of the literature on modelling, decision making, social-ecological systems, Atlantic salmon, freshwater and marine resource management. Running first set of stakeholder workshops to identify the information needs and knowledge base for modelling

Year 2

Existing data collation and analysis of data on both the stakeholder system and the salmon system and integration into a conceptual and prototype GMSE model

Year 3

Producing management scenarios and predictions from GMSE model. Evaluating the potential uptake of the model and its outputs by stakeholders

Year 3.5

Thesis writing up and publishing peer reviewed papers

& Skills

Qualitative research methods: stakeholder workshops and focus groups

Modelling: social-ecological modelling, agent-based modelling and Generalised Management Strategy Evaluation modelling (GMSE)

Interactive model development: R Shinyapps

References & further reading

Qualitative research methods: stakeholder workshops, focus groups
Modelling: social-ecological modelling, agent-based modelling and Generalised Management Strategy Evaluation modelling (GMSE)
Interactive model development: R Shinyapps

Apply Now