Echo sounder technology for appetite-led-feeding and welfare-monitoring of caged salmon (ECHOFEEDING)

Waste feed is a huge economic cost and source of environmental impact in salmon sea-cage farming. Appetite monitoring and feed control is key to reducing waste feed and optimizing fish growth for maximum profits. However, visual-based techniques in current use are labour-intensive, imprecise and prone to overfeeding, and becoming increasingly inadequate in larger modern cages of different designs. In this project, we have explored the technology application of an Echofeeding system to autonomously monitor fish appetite and control feeding. The technology uses an echo sounder transducer to monitor fish and calculate their biomass in the feeding area of a cage. When feeding is performed, it is pre-programmed to continue or stop feeding based on the amount of fish biomass levels in the feeding area. In this way, fish appetite is measured in real-time and dictates feed quantities at each meal. In contrast to visual-based appetite monitoring, the technology of Echofeeding does not require continuous visual observation and can objectively quantify fish appetite. Experimental work conducted in research scale sea cages in both coast and fjord farming environments show that Echofeeding is straightforward to implement for both small and large fish sizes, over different seasons, and with superior feed conversion and growth. While the feeding response as measured with echo sound is predictable in the vertically homogeneous coast environment, it is more challenging to predict and measure in fjords as strong vertical gradients of temperature and salinity occurs. Echofeeding must then be calibrated towards the environment, and thus more frequently. The project has addressed several questions that are of interest to feeding strategies and control in general. Feeding intensity is increased in a step-wise manner for several fish sizes, from smolt to harvest size, and the salmon are found to adapt at a rapid rate. Higher feeding intensity provides a faster response at meal initiation and a more overt response at satiation, as notable from both the change in salmon behaviour and number of pellets trickling down below the fish. For example, we have successfully accustomed groups of 6000 one kilo sized salmon to feed to satiation (~1.5%) at an extreme rate which correspond to only 20 minutes of feeding per day. This provides every fish with 3 pellets per minute and is thus not a very difficult task for them to consume. However, in commercial salmon farming the same amount of food served over two hours would be considered a rather high feeding intensity. Other trials within the project have investigated whether groups of salmon can be conditioned into responding to a sound or light cues signalling food arrival. Salmon reared in tanks can easily be conditioned and respond by swimming into the feeding area before food arrival. Our trials in sea cages were unsuccessful in achieving anticipatory responses. This might be due to that sea caged salmon show stronger social behaviour in sea cages than in tanks, and may not be regarded as inability to learn the association between a signal and feeding. A different learning aspect that is tested is the ability of caged salmon in responding to spatial relocation of feeding. By repositioning the food entrance point from zero to 8 m depth, the fish got rather confused the first meal, but had little problems after that in locating the food, even when the food entrance point was alternated between surface and 8 m depth for every meal. Another relevant observation for submerged feeding was a slower swimming speed during feeding at 8 meter versus at surface, which resulted in less efficient consumption and thus a deeper vertical feeding area.

Prosjektet Echofeeding har satt fokus på fôringsmetodikk i laksenæringen og demonstrert: 1. At ekkolodd kan benyttes som et svært hensiktsmessig verktøy for observasjon av fiskens appetitt, og 2. At laksen kan og bør fôres etter en konsekvent metodikk, og 3. At det er mulig å fôre laksen med autonom kontroll, full vekst og uten fôrspill. Prosjektets resultater inkluderer en kunnskapsbasert protokoll for oppsett og innstillinger av autonom fôringskontroll med Echofeeding, inkludert forventet frekvens av kalibrering av stoppsignal for fôring ettersom fisken vokser, og identifisering av miljøendringer som kan vanskeliggjøre bruken. Dette tilsier at prosjektet har belyst en metodikk for effektiv autonom fôringskontroll til bruk i lakseoppdrett, som også gir et mer objektivt mål av appetitt som velferdsindikator. Robotisering av fôringskontrollen forventes å kunne gi en betydelig reduksjon i fôrspill, som er et vedvarende biologisk og økonomisk problem innen fiskeoppdrett.

Waste feed is a huge economic cost and source of environmental impact in salmon sea-cage farming. Appetite monitoring and feed control is key to reducing waste feed and optimizing fish growth for maximum profits. However, visual-based techniques in current use are labour-intensive, imprecise and prone to overfeeding, and becoming increasingly inadequate in larger modern cages of different designs. In this project, we will explore the technology application of an ECHOFEEDING system to autonomously monitor fish appetite and control feeding. The technology uses an echo sounder transducers to monitor fish and calculate their biomass in the feeding area of a cage. When feeding is performed, it is pre-programmed to continue or stop feeding based on the amount of fish biomass levels in the feeding area. In this way, fish appetite is measured in real-time and dictates feed quantities at each meal. In contrast to visual-based appetite monitoring, the technology of ECHOFEEDING does not require continuous visual observation and can objectively quantify fish appetite. Intensive feeding over shorter intervals is a prerequisite for ECHOFEEDING, leading to longer fish residency times at lower cage depths and reduced contact with surface-dwelling infective salmon lice larvae. We will determine optimal methodology for commercial scale use of the technique, examine factors explaining fish appetite variation in caged salmon and assess the applicability of fish appetite recordings from ECHOFEEDING as an early-warning system for poor fish welfare. Technology innovation achieved in this project has the potential to substantially lower waste feed, maximize economic returns and improve fish welfare in the Norwegian salmon farming industry. The project addresses several of the areas within the Havbruk2 calls for Theme 4 HAVTEK (aquaculture technology), with emphasis on the knowledge and technology for more controlled, efficient feeding.