The final report from the second Marine Donut production cycle is now available. SalMar summarizes experiences and measurements from the monitoring program carried out in accordance with the Norwegian Directorate of Fisheries’ requirements for development licences.
The monitoring program was carried out to evaluate the technology and document performance in terms of operations, environment, fish welfare, biological results, and technical functionality. Production cycle 2 ran from 4 May to 6 July 2025.
Read the full final report here (Norwegian only)
Water quality and environment-based parameters were continuously monitored throughout the entire production cycle. Oxygen, temperature, salinity, pH, turbidity, CO₂, and algae content were measured in accordance with the monitoring program, in addition to weekly measurements of total gas pressure and suspended solids.
The measurements show that water quality in Marine Donut was good and stable throughout the period. The average oxygen level was 91.5%, and high feeding rates over extended periods did not have a negative impact on oxygen levels. Mixing of fresh water in the fish volume functions well, as confirmed by consistent measurement values and operational experience.
One successful fish transfer was carried out, with a total of 185,196 salmon and an average weight of 3,002 grams. The transfer was completed efficiently and without deviations or fish welfare–related challenges.
The experience confirms the findings from the first production cycle. Water flow velocity along the bottom of the fish volume is still not sufficient to ensure full transport of dead fish to the dead-fish traps. Measures such as lowering the mixers have improved the situation, and during production cycle 2 no dead fish were observed lying stationary on the bottom. Dead fish accumulating on outlet grates remains a challenge, and ROVs have been used to ensure satisfactory dead-fish removal. Further measures are planned ahead of the next stocking.
Feeding in Marine Donut provides good feed distribution, but camera monitoring is challenging at high fish densities. Feed intake can be observed, but stop signals are difficult to define. This affects the ability to optimize feeding control and is reflected in the biological results.
Crowding and harvesting of fish were carried out in accordance with a revised procedure. The procedure was adjusted to reduce structural load when the unit is in a raised position. The use of a bypass system and alternative oxygenation ensured good water quality conditions during the operation. The measures functioned as intended, and all fish were removed from the fish volume at the end of harvesting.
Fish health and welfare were followed up through regular health inspections, welfare scoring, and monitoring of gill health. Health status was assessed as generally good throughout the period, with normal behavior, good appetite, and low to moderate mortality.
Welfare parameters such as fin erosion, wounds, snout injuries, and deformities were recorded weekly. Some skin changes were observed during parts of the period and were, following examination, assessed to be related to waterborne material. These changes subsided toward the end of the production period.
Gill examinations mainly showed minor to moderate, non-specific changes with no signs of known diseases. The results indicate that the fish had satisfactory gill health in Marine Donut.
Stress measurements based on cortisol and other hormones show an expected stress pattern associated with fish transfer, followed by rapid normalization after acclimatization. The results indicate that the fish were in good condition without significant stress-related impacts while in Marine Donut.
A total of 236,207 kg of fish were produced during production cycle 2, with an average weight increase of 1.27 kg. The economic feed conversion ratio was 1.20. Growth performance was good and in line with SalMar’s growth models, despite periods of reduced feeding due to technical conditions.
Mortality during the period the fish were held in Marine Donut was low, at approximately 1.7%. Production data show that the system largely functions as intended, while also highlighting areas for improvement related to feed monitoring and operational practices.
A total of 183,018 fish were harvested with an average weight of 4,285 grams. The superior share was 90.9%. Downgrading was mainly related to biological factors and not directly linked to production in Marine Donut.
The data and communication systems functioned as planned, and measurement data have been stored in accordance with requirements. Full-scale measurements of water flow in the fish volume have been completed. Measurements of hydrodynamic pressure, global movements, and mooring forces have been finalized and published in separate final reports.
The experience from production cycle 2, together with experience from the first cycle, provides a solid factual basis for further optimization of Marine Donut. Identified changes and improvement measures will be incorporated into the continued development of the concept.
