--- ## Modelling the evolution of sea lice to resist treatment using SLIM :fish::bug: :pill: :chart_with_upwards_trend: The SLIM team: Sara Jakubiak, Debby Lipschutz, Enrico Trombetta, Sara Kutkova, Anthony O'Hare, Jess Enright --- Sea lice are parasites of salmonids, and can be an economic and welfare problem for salmon aquaculture: sea lice | cages :-------------------------:|:-------------------------:  |  --- Sea lice have a number of life stages. At some they move in the water and at some they are attached to fish.  --- | One sea loch may have a number of farms, and we know about the water flow between them. | | |-|-| Farms within one management area are meant to coordinate treatment. --- We modelled - Fish growth :fish: - Lice life cycle, infestation, reproduction :bug: - Treatment with several treatment options :pill: :fish: :bath: - Evolution to treatment with EMB (a common chemical treatment) 🧬 - Coordination of treatment in several pre-set ways :handshake: - An external lice reservoir that changes only in genetics --- Treatment with several options: emamectin benzoate  (https://www.msd-animal-health-hub.co.uk/) --- Mechanical treatment: thermolicer  (https://www.salmonscotland.co.uk/facts/faqs/innovation/what-is-a-thermolicer) https://www.youtube.com/watch?v=XnT6ihN1TaY --- Cleaner fish - wrasse and lumpfish:  --- Modelling treatment effectiveness: - chemical: 1 - resistance - mechanical: number fit from government reports - cleaner fish: inverse exponential from a paper on Norwegian sea lice modelling Important parameters from Aldrin et al. 2017 **We only model resistance to chemical** --- Genetic model for resistance: - simple heterozygous dominant resistance mechanism - no fitness cost to resistance - Homozygote resistant *more* resistant than heterozygote --- Treatment coordination policies: Overall idea is that when any farm in a management area exceeds a lice threshold, then all farms in the area treat. We model: - Far too much/little treatment - simple defection: a probability $p$ that each farm ignores instruction - different styles of rotation - fixed order - random choice --- Results: too/little much treatment  --- Results: high/low cooperation  --- Results: high/low cooperation  --- Results: fixed vs random rotation  **Very little difference** --- Results: different positions in network:  Left to right: most to least connected --- Results: impact of network shape  --- Takeaway messages: - treatment coordination and network characteristics can impact lice numbers and resistance - full cooperation not necessarily best if it leads to more overall treatment We should consider these when devising treatment systems! --- We have many ideas for future work: - climate change - more complex genetics - more complex coordination - strategy optimisation - larger scale --- Future ideas: Resistance to more treatments  --- Future ideas: larger scale  (https://blogs.gov.scot/marine-scotland/2020/03/30/new-reports-looking-at-sea-lice-dispersal-around-scotland/) --- Thanks! ---  https://github.com/resistance-modelling/slim ---