Modelling assessment of how small-scale vertical movements of infectious sea lice larvae can affect their large-scale distribution in fjordic systems

Abstract

Sea lice are ectoparasites that can be found in high numbers in and around salmon farms, where they are a threat to fish health and can induce high aquacultural costs. Large numbers of suitable hosts facilitate the infection and subsequent release of their planktonic larvae in the surrounding environment where they can potentially infect wild salmonids, including migrating juvenile fish (smolts). Investigating sea lice spatial distribution is generally done using coupled hydrodynamic and particle tracking models. The quality of these numerical tools is critical to identify areas of higher infection risk to valuable wild salmon populations and thus support sustainable salmon aquaculture. While the transport of salmon lice is mainly affected by physical processes, biological behaviours such as vertical swimming can also play an important role. However, a review of previous sea lice studies shows no clear consensus on their swimming abilities and the parameters implemented in sea lice dispersion modelling. Here, we focus on the Diel Vertical Migration (DVM) behaviour, a vertical migration that sea lice perform within a daily cycle. Our sensitivity study of the infectious copepodid phase of sea lice highlights how their retention potential and spatial distribution within a water body are affected by their vertical swimming velocity and maximum swimming depth. In a fjordic system (Loch Linnhe on the West Coast of Scotland), the vertical position of sea lice can affect their horizontal trajectory due to the two-layer exchange flow of the estuarine circulation. At the surface, transport is mainly due to the wind driven circulation, and the residual seaward current. Lower in the water column, the saltier shelf water is drawn into the sea loch. This can potentially increase the retention of sea lice if they dive deep enough to reach those opposing subtidal currents. Therefore, lack of confidence in sea lice DVM parameterisation may introduce inaccuracy in modelled sea lice distribution. Effective sea lice management in aquaculture would benefit from more observational data like farm sea lice count, sea lice swimming laboratory observations, field observation of sea lice vertical distribution and accurate sea lice quantification in the field. This would help to reduce uncertainties derived from sea lice vertical swimming behaviour parameterisation in lice dispersal models.