Resistance to chemical treatment is a common problem with agricultural pests and human pathogens. Early work on resistance management in agricultural settings identified the “high-dose/refuge” (HDR) strategy, involving an untreated refuge crop planted near a treated crop, allowing treatment-susceptible pests to immigrate and interbreed with treated conspecifics. High treatment doses are intended to induce resistance recessivity in any F1 resistant/susceptible heterozygotes. We explored HDR-related concepts in the context of salmon aquaculture. Globally, open-net-pen salmon farms rely on chemicals to manage infestations by sea lice (Lepeophtheirus salmonis and Caligus spp). In many regions, sea lice have rapidly evolved resistance to the chemicals used, presenting management challenges and endangering sympatric wild salmonids that are susceptible to infestation by farm-origin lice. In the eastern north Pacific, however, treatment of sea lice has proceeded without signs of resistance. We extended a simple host-parasite differential equation model to explore the spread of treatment resistance in sea lice on salmon farms, and how wild hosts could both form an oceanic treatment-free refuge and connect sea louse populations on coastal farms to that refuge. Parameterising our model based on published findings, we investigated the potential effects of population sizes, treatment levels, and resistance constraints.
Results/Conclusions
Our model analyses reveal that the balance among multiple factors could readily affect whether or not treatment resistance evolves in sea lice on salmon farms. First, our model reproduced results consistent with the HDR strategy: both high levels of treatment on farms and a large oceanic wild-host population were required to prevent the evolution of treatment resistance in sea lice. Second, resistance was liable to invade when the resistant phenotype was more resistant to treatment and associated costs were low. Third, the rates of infestation by free-swimming louse larvae and the persistence of the wild-host population spawning near farms (which served to transfer lice between the farm and oceanic environments) were both critical to maintaining treatment susceptibility in sea lice. In our model, treatment both selects for resistance and protects the nearby wild host population, thereby maintaining the service those wild hosts provide in opposing resistance evolution. In recent years, much controversy has surrounded the impact that salmon farms can have on nearby wild salmon populations. Our work highlights the potential for those same wild salmon to provide a direct benefit to farms, giving salmon farmers a reason that protecting wild salmon may well be in their own best interest.