Testing, validation and use of automated lice counting camera to study depth-dependent lice levels in salmon production cages
This project aims to test the efficacy of an autonomous lice counting camera system for Atlantic salmon in sea cages, and investigate any differences in swimming behaviour and lice levels among individual salmon. The system is able to detect chalimus II, pre-adult and adult stage lice, and if effective in the cage environment, will allow continuous assessment of lice densities on farmed salmon at a range of depths without handling of fish. The project consists of two parts:
Experiment 1 requires 3000 Atlantic salmon divided across three 5x5 m sea cages. 100 fish in each cage will be captured and anaesthetised for manual lice counts and tagged with T-bar Floy tags before being returning to the cages. This will allow comparison of manual lice counts and counts provided by the autonomous system. For 14 days, the camera system will detect lice on fish that pass within ~70 cm of the camera. After 14 days, tagged fish will be caught and euthanised before a second set of manual lice counts are conducted, while untagged fish will be moved to a holding cage for use in Experiment 2.
Experiment 2 requires 7500 salmon in divided across three 12x12 m cages (~2000 fish reused from Experiment 1). Three camera units will count lice on fish at three different depths (e.g. 1 m, 6 m and 11 m) for 4 weeks to reveal relationships between lice loads and swimming depth. 40 fish will be sampled from each cage at the end of each monitoring period for manual lice counts (total 480 fish), using 2 different catch methods. A total of 4 types of catch methods, commonly practiced in research and commercial settings, will be used to determine the bias associated with the methods.
Because we do not expect adverse impacts or confounding factors from Experiment 1, untagged fish are proposed to be reused in Experiment 2 in order to reduce the total number of salmon required. The number of tagged salmon has been carefully considered, taking into account the minimum proportion of the population likely to be observed by the camera system. There will be no tagged salmon in Experiment 2. All salmon will have levels of sea lice naturally acquired from the site over months, however the abundance of lice will not exceed allowable limits and therefore will only incur low-moderate distress. Tagging using T-bar tags has been used extensively in previous studies and has not been shown to have a negative impact on behaviour or welfare.
The verification of lice counting using this camera system provides a passive monitoring program that reduces the need for manual handling and unrepresentative sampling for lice counts. The focus on promoting animal welfare through autonomous examination of fish is increasing in the salmon aquaculture industry, and systems such as this must be tested for accuracy to ensure no under- or overestimation of lice abundance can occur.
Experiment 1 requires 3000 Atlantic salmon divided across three 5x5 m sea cages. 100 fish in each cage will be captured and anaesthetised for manual lice counts and tagged with T-bar Floy tags before being returning to the cages. This will allow comparison of manual lice counts and counts provided by the autonomous system. For 14 days, the camera system will detect lice on fish that pass within ~70 cm of the camera. After 14 days, tagged fish will be caught and euthanised before a second set of manual lice counts are conducted, while untagged fish will be moved to a holding cage for use in Experiment 2.
Experiment 2 requires 7500 salmon in divided across three 12x12 m cages (~2000 fish reused from Experiment 1). Three camera units will count lice on fish at three different depths (e.g. 1 m, 6 m and 11 m) for 4 weeks to reveal relationships between lice loads and swimming depth. 40 fish will be sampled from each cage at the end of each monitoring period for manual lice counts (total 480 fish), using 2 different catch methods. A total of 4 types of catch methods, commonly practiced in research and commercial settings, will be used to determine the bias associated with the methods.
Because we do not expect adverse impacts or confounding factors from Experiment 1, untagged fish are proposed to be reused in Experiment 2 in order to reduce the total number of salmon required. The number of tagged salmon has been carefully considered, taking into account the minimum proportion of the population likely to be observed by the camera system. There will be no tagged salmon in Experiment 2. All salmon will have levels of sea lice naturally acquired from the site over months, however the abundance of lice will not exceed allowable limits and therefore will only incur low-moderate distress. Tagging using T-bar tags has been used extensively in previous studies and has not been shown to have a negative impact on behaviour or welfare.
The verification of lice counting using this camera system provides a passive monitoring program that reduces the need for manual handling and unrepresentative sampling for lice counts. The focus on promoting animal welfare through autonomous examination of fish is increasing in the salmon aquaculture industry, and systems such as this must be tested for accuracy to ensure no under- or overestimation of lice abundance can occur.