Risk of impact on nearby fish farms associated with establishment of underwater naval training area at Korsnes fortress

The Norwegian Armed Forces plan to establish a permanent exercise area in Korsfjorden. The planned activity implies detonations of small explosive charges (0.1–2 kg TNT) several times per day, up to 70 detonations per year. There are four salmon fish farms within 3.7–6 km from the exercise area. The risk of direct injury to the fish is small, but pressure waves from the detonations could trigger escape responses which could lead to indirect effects like shell loss, swim bladder compression, stress and loss of appetite. Thresholds for such responses are not well known. As a risk mitigation measure, a bubble curtain has been placed across the bay where the detonation will take place. The purpose of the project is to investigate the risk mitigation effect of the bubble curtain, and collect data on fish behavior and health during a realistic exercise program. 

25 discharges with increasing charge size from 0.1–2 kg were detonated at Korsnes Fortress in March 2021, with and without bubble curtain over a period of 5 days. Fish behavior was observed with camera and echo sounders before, during and after the exposure. The sound levels were measured with hydrophones on all salmon farms. Food consumption and mortality rates per net were recorded by Lerøy Sjøtroll, the owner of the fish farms. Fish samples were taken from the closest nets before and after the detonations, for measurements of cortisol in blood and feces and for pathological examinations focused on barotrauma. 

Broadband peak pressure was reduced by 12–20 dB (re 1µPa) and SEL by 5–12 dB (re 1µPa2·s) when the bubble curtain was used. There was no difference between single and double curtain. Short behavioral responses occurred at the highest recorded peak pressure levels (>163dB re µ1Pa in the hearing band of the fish 20–500 Hz, SEL >147 dB re µPa2·s), but similar responses could also occur during normal operation of the fish farm. We did not find any changes in food consumption or mortality rate during the exposure period. In the fish samples taken immediately after the final detonation, circulatory disruptions and hemorrhages which are consistent with barotrauma was observed, and in the same fish increased cortisol levels were found, but food consumption and mortality rates were normal. Cortisol levels were back to normal in the sample taken 24 hours later. This fish were exposed to pressure levels far from thresholds of injury reported in the literature (>230 dB peak pressure levels). Such injury can occur as sample artifacts, but we cannot fully rule out the possibility of subclinical injury that is only detectable through autopsy and histopathology. Nevertheless, our conclusion is that the fish were not exposed to stress or injury that affected growth, mortality and health at the population level. 

We recommend that bubble curtain is used during exercises to keep the peak pressure levels at the closest fish farm below 160 dB re µPa in the 20–500 Hz band. Between August and November the propagation loss is 10–15 dB lower than in March, and in this period exercise activity should be limited to only one shot per day, or 100 gram charges. For the other farms, the levels will be lower and less of a concern.

The risk to wild fish and marine mammals are very limited, and when the bubble curtain is used, neglectable.