Resistance: Understanding the impact of management agreements on reducing evolution of resistance in agriculture - sea lice management as an exemplar

Lead Research Organisation: University of Glasgow
Department Name: School of Computing Science


Sea lice (Lepeophtheirus salmonis) are a parasite that affects the growth and performance of farmed and wild salmon, as well as decreasing their immune response, making them more susceptible to disease. Efforts to combat sea lice cost the salmon industry in Scotland an estimated £40M to control each year. Sea lice in the farmed salmon population in Scotland are managed cooperatively under agreements between managers within an area, though much of this collaboration is on a voluntary basis. Unfortunately, there are indications that sea lice are becoming resistant to current treatments, requiring managers to either increase the amount of treatment used, increase the frequency of treatment, or investigate alternative methods of integrated pest management. This bears significant extra cost and could have serious environmental consequences. In the context of this rising threat of resistance, management programs must be designed to slow the evolution of resistance, but strategic use of control agents can be complicated by the actions of other decision-makers, and by uncertainty about the levels of resistance present.

Compared to terrestrial pests, little is known about how sea lice evolve resistance, but new information (under investigation by one of our team members (Dr. Sturm) under project BB/L022923/1) suggests that the mechanisms may be more complicated than expected, with at least one type of resistance transmitted only through the female line. Because of uncertainty about mechanisms of resistance, tests for resistance in sea lice are not universally reliable, and can be subject to failure if sampling of the lice is not done properly.

Therefore, decision-makers managing sea lice face not only uncertainty about the actions of fellow producers, but also about the level of resistance currently present in the sea lice population.

In this project, we will combine mathematical and game theoretic modelling methods with expert knowledge of sea lice and the salmon industry to build a model of sea lice evolving resistance in a setting of multiple managers making decisions about how to treat sea lice. This combination of an epidemiological model of sea lice spread with a game theoretic model of treatment use will help us understand how management agreements between managers of fish farms can impact the risk of sea lice evolving resistance to treatments, and therefore help industry bodies better build and support these management agreements.

Our project team has close links to industry bodies, and includes experts in the evolution of resistance in sea lice, in game-theoretic and epidemiological modelling, in high-performance computing, and in modelling of farmed fish. Using this expertise, our project will contribute to successful sea-lice management in Scotland, impacting not only the salmon industry, but also potentially wild salmon stocks, and therefore the wider marine environment.

Technical Summary

We will combine mathematical approaches with specialist parasite and industry knowledge to give an overall picture of the impact of joint management schemes on resistance risk in sea lice.

Our first step will be a model of sea lice population growth incorporating population genetics and resistance evolution, implemented for running on a high-performance computing platform. Critically, our model of resistance will include several different possible modes of resistance action and inheritance, including multigenic resistance and maternal inheritance mechanisms.

We will integrate this epidemiological/resistance model with a model of joint treatment decision-making, analysing the impact of management strategies on the evolution of resistance. Our decision-making model will include both a game-theoretic model in which every decision-maker is perfectly rational and an agent-based model with decision-making strategies designed to mimic real-life management. High performance computing will allow many iterations of this model, providing robust results and a clear sensitivity analysis.

We will advance on previous work in several ways: including more complex resistance mechanisms will improve on current sea lice resistance models, which only allows simple single-gene resistance. The computational efficiency of the epidemiological model will allow runs over a larger scale than has been previously possible. Including both game-theoretic and agent-based decision modelling will allow more realistic modelling of treatment decision-making than in the past, along with comparison to a mathematically robust baseline. Combining these models will provide additional novelty and value.

While our work will be achieved using models, our ultimate aim is to build an understanding of how joint decision-making can effect the evolution of resistance. That understanding will advance current knowledge about how cooperation in an epidemiological context can impact resistance.

Planned Impact

This project will have impact on a variety of scales over a broad range of beneficiaries.

The primary non-academic beneficiaries of our project will be the salmon farming industry, e.g. Scottish Salmon Producers Organisation (SSPO), Marine Harvest, Scottish Sea Farms, Fish Vet Group, etc. Our proposed program of work will yield recommendations for the creation and management of farm management agreements, allowing improved efficiency of sea lice treatment and delaying the evolution of resistance. These recommendations will ultimately increase the profits of salmon farmers and minimise the infection stress to both farmed and wild fish.

Bodies that govern the Scottish salmon industry will benefit from our work, including the Scottish government and industry bodies such as CEFAS. Our findings on the implications of differing farm management strategies within a management agreement and the importance of cooperation could be used to inform policy on how management agreements are built. Beyond the salmon industry, our work will have implications for understanding the dynamics of herd health agreements and local animal health cooperatives, such as exist in the cattle and sheep industries throughout Great Britain, and therefore we expect our work will indirectly benefit the terrestrial livestock health management community.

When sea lice are well-controlled in farmed salmon, they will be less likely to affect the wild salmon population, so stakeholders in the health of the wild salmon population such as sport fishermen will also benefit from this work. One of the challenges facing the salmon farming industry is the public perception of the effect on wild salmon. Our work in controlling sea lice population will alleviate the negative perception and the growth of this industry will have a positive socio-economic impact on those communities close to where farms are sited, which are usually rural, isolated and often lack investment.

As our work will support more efficient farmed salmon production, consumers of salmon will benefit from an efficient, and therefore potential lower-cost and higher-quality product. Because fish consumption supports health when it replaces red meat consumption our work has potential to contribute to a healthier populace overall, and increased food security.


10 25 50
Description This ongoing project has been severely impacted by the pandemic.

There were several main achievements of the grant during this first phase of work:
1. The definition and resolution of a mathematical problem that we have called the Firebreak problem - related to questions of how best to defend a network from a contagion given a fixed budget of treatment or vaccination at the beginning of the outbreak.
2. A collated model of sea lice evolution to resistance that incorporates both the epidemiology of sea lice infestation as well as the population genetics of sea lice themselves, and drawing on the network structure of movement of water between salmon farms on the west coast of Scotland.

In the second phase of this project (since resuming after a suspension in summer 2021), we have now implemented a fully-engineered simulation model integrating farm location, hydrodynamic connectivity, various treatment options, and the sea lice lifestyle. Witihn this model framework, we have found:
1. The location of a farm in the loch and hydrodynamic linkage impact on optimal treatment behaviour
2. As expected, repeated treatment use drives an increase in genetic resistance
3. Threshold-based policies (as are currently used within Scottish aquaculture) may be either excessive or unable to limit the epidemic
4. Mosaic treatment in which patches are treated may do more harm than good
5. Altruistic coordination of treatment such that farms use different treatments at the same time may limit resistance.
Exploitation Route The open-source model may be of use, and the findings on treatment coordination may be useful in understanding and limiting the evolution of resistance to treatment
Sectors Agriculture, Food and Drink

Title Model including epidemiology, population genetics, and treatment impacts 
Description We have developed a combination epidemiological and population genetic model of sea lice in a single sea loch. This model is currently in testing, and will become public software when it is comprehensively tested. Update in 2020: This model now includes sea lice epidemiology including hydrology, multiple genetic mechanisms for resistance inheritance, and treatment using chemical therapeutants. It is public and available (, but is still under constant development, and not yet suitable for non-academic use. 
Type Of Material Computer model/algorithm 
Year Produced 2019 
Provided To Others? Yes  
Impact None as yet. 
Title Release of SLIM model for sea lice control 
Description Open-source model written in Python incorporating lice life cycle, management strategies, and resistance evolution 
Type Of Material Computer model/algorithm 
Year Produced 2022 
Provided To Others? Yes  
Impact NA 
Title SLIM release v0.2.3.rc0 
Description Release of SLIM. SLIM is the Sea Lice Model associated with a funded BBSRC project on the evolution to resistance to treatment in sea lice (BBR009309). We aim to integrate an epidemiological and genetic model of sea lice with a model of treatment decision-making by different salmon farms. 
Type Of Technology Software 
Year Produced 2022 
Open Source License? Yes  
Impact This piece of software will underly future publications and be disseminated in itself at upcoming conferences.