Impact zones around Norwegian fish farms from release of bath treatment chemicals

Cage-based marine aquaculture of salmonids is a growing form of global food production, and one of Norway's major export industries. One obstacle to further growth is the naturally occurring sea lice, which may multiply in large numbers in open net-pen salmon farms. The wild salmon population is vulnerable to sea lice in the springtime outwards migration period, and Norwegian farmers are obliged by law to keep the levels of sea lice down. This is often accomplished by bath treatments, where the net pen is enclosed by a tarpaulin and anti-sea lice chemicals are administered. After treatment, the tarpaulin is removed, and the chemicals are released into the surrounding waters.

There has been an increasing awareness of the adverse effects of anti-sea lice chemicals on non-target species. In this work, we investigate how far released chemicals travel before being diluted to ecologically safe concentrations. An offline particle tracking model (LADiM, develped at the Institute of Marine Research, Norway) is used to simulate chemical releases from a large number of actual aquaculture locations, covering both exposed and sheltered locations, for a range of different release dates. Physical forcing fields (3-dimensional ocean currents) are provided by the ocean model system NorKyst800/NorFjords160, which is based on the Regional Ocean Modeling System (ROMS). We report on the extent of the impact zones, the dilution rate, and their sensitivity with respect to location, season, chemical toxicity, spatial resolution of the hydrodynamic model and the random walk component of the particle tracking model.

Our results give valuable insight to regulatory agencies, that may choose to ban certain bath treatment chemicals from being released close to sensitive habitats. We give examples of such habitats, and suggest rule-of-thumb buffer zones around them with respect to three common bath treatment chemicals: Hydrogen peroxide, deltamethrin and azamethiphos.

Due to the large dataset produced, and the multiple ways of displaying and applying the data, this presentation may also be suitable for an eLightning presentation where the audience can interact with the data dynamically.