Acoustic deterrence using startle sounds: long term effectiveness and effects on odontocetes

This report presents the findings of field experiments aimed as testing the long-term effectiveness of a new deterrence method (acoustic startle reflex)on pinnipeds.


Short-term trials on farms with predation problems

a) Ardmaddy (Argyll)

At the beginning of April 2011 predation losses at the Ardmaddy control site began to rise. The site manager reported that food intake was dropping, possibly due to the presence of seals around the cages. The farm was stocked with 2 cages with fish fully grown and awaiting harvest in May 2011. To prevent significant loss, a single loudspeaker deterrence system was placed at the fish farm for the two weeks before harvest.

Methods

Ardmaddy fish farm consists of square steel cages connected by walkways which made deployment of the equipment easy, enabling coverage of the two cages with one loudspeaker. The two stocked cages were positioned adjacent to each other on the landside of the farm. The deterrence system was deployed on 18 May 2011. The equipment could not be connected to mains power so that farm staff had to change batteries every 2-3 days. Predation was monitored in detail by the site manager prior to, and during deployment of our equipment, based on the fish retrieved by divers.

Results

Predation levels at Ardmaddy were high in April and May 2011 with most dive reports revealing losses ranging from 30-70 fish (150-350 kg biomass, Fig 8). Predation levels in both adjacent cages were similar across the one month pre-deployment period (cage 7: 202 fish, cage 8: 203 fish). Overall, 405 fish were killed by seals during the pre-deployment period when monitoring was carried out (Fig 8) with seal inflicted losses accounting for 65% of the overall losses at the farm. The first dive report after deployment of our startle system showed no seal-related kills in cage 8. Cage 7 was harvested a week after deployment of the equipment with no fresh seal kills found in the net. The net contained 10 fish with bite wounds but all were in a state of highly advanced decomposition indicating that these fish were previously missed by the divers. Fish removals by divers are more likely to miss dead fish compared to harvest which reveals the full content of the cage. Cage 7 remained stocked for another week before it was harvested. The full count of all dead fish at harvest revealed only a single fish killed by a seal (Fig 8). No monitoring of seal behaviour was carried out.

Fig 8: Predation losses (seal kills) inflicted by seals on Ardmaddy farm prior to the deployment of the single transducer (white area) and during its operation (pink area). Only one fish was lost due to seal predation when the transducer was in operation.

Figure 8

b) Quanterness (Orkney)

Methods

Seal predation on fish farms in Orkney is understood to be primarily caused by grey seals. Several farms in Orkney suffered heavy predation in winter and spring 2011 but predation was reported to be highly dynamic. On 02 July 2011 we deployed a startle system with 2 transducers and a source level of 174 to 176 dB re 1µPa on a Meridian Salmon (Ltd.) farm with 4 isolated circular cages stocked with smolts. The two loudspeakers were fitted on two separate cages.

Results

Seal predation levels were moderate to low and highly variable prior to deployment of our system (Fig 9). During the first week following deployment no fish mortality attributed to seals was found in the cages but during the second week numbers were comparable to the pre-deployment period. However, the equipment was not operating for parts of the second week due to power failure. The last confirmed operation of the equipment was at the end of week 1 on 19 July 2011. Several severe gales in the following weeks caused the control box to flood destroying all electronic components. The equipment was removed and no additional trials were carried out at this farm.

Fig 9: Predation losses on Quanterness farm prior to the deployment of the single transducer (white area) and during its operation (pink area). 'State unknown' indicates the system was off (battery depleted) for some unknown time during that period.

Figure 9

Contact

Back to top