Optimization of feed distribution to sea caged fish with an emphasis on Atlantic salmon (Salmo salar L.)

The aquaculturist is faced with a number of difficulties when feeding fish. Apart from the human time component involved in feeding a large number of fish to appetite, it is important to feed an appropriate quantity at a suitable frequency to ensure that no food is wasted and fish are satiated. Fish display preferential feeding patterns that relate to endogenous rhythms and changing biological and environmental factors. This study describes a new technology called the \Adaptive\" Feeding System designed to automatically feed fish by regulating feed input based on the levels of waste food detected beneath a feeding zone. A series of trials with the system are also discussed. The system consisted of a surface mounted microprocessor linked to an underwater sensing device capable of resolving a single feed pellet. An internal algorithm controlled the operation of the system. Feeding data was stored by the microprocessor and down loaded via a data-logger to an IBM-compatible computer for analysis with specific software. Atlantic salmon Salmo salar in sea-cages exhibit feeding patterns which vary both diurnally and seasonally. Hitherto there are no data reporting feed rate and its variation through a complete annual cycle. Here we present data from Scotland showing diurnal and inter-seasonal variation in feeding patterns and feeding rates of Atlantic salmon fed daily to satiation from shortly after transfer to seawater until harvest about 11 months later. A major feeding peak regularly occurred soon after dawn and feeding rates remained high for approximately one hour. Over the remainder of the day the fish fed at a lower but steady rate. Relative feed intake varied over the trial being initially high in summer followed by a sharp decline in autumn and then further declining until fish reached harvest size at the beginning of the following summer. Further investigations of the relationship between variation in circannual feeding patterns and environmental parameters should now be carried out to improve the understanding of the mechanism behind these patterns. Tasmanian Atlantic salmon (2-3kg) fed daily to satiation for four months over winter displayed a diurnal pattern of feed intake. The first peak of intake commenced just after dawn for 2-3 hours during which up to 60% of the total daily feed intake occurred. Some feeding occurred during the middle of the day but this was eclipsed by a significant feeding bout approximately 30 minutes before total darkness in the evening. This typical diel pattern often disintegrated due to changes in environmental factors but more significantly due to suspected disturbance by human activity or the presence of predators. Surface activity of the salmon in response to feed input and to a lesser extent swimming speed of the fish were found to be reasonable indicators of feed intake. A better indicator of feed intake was measured by monitoring small quantities of waste feed sensed by the adaptive feeding systems' sensor. A further study investigating the effect of restricted feeding periods followed by refeeding to satiation was carried out on Atlantic salmon (approx. 1.3kg) during winter/spring on growth. Four treatments included those fed daily to satiation (A) those fed for 5 days then starved for 2(B) those fed for 10 days then starved for 4(C) and those fed for 7 days then starved for 7(D). Every 28 days growth was measured by weight. Group A showed significantly higher growth (p<0.05) in weight than the other treatments over 4 months. Groups B and C showed similar growth and Group D displayed the poorest growth. The ability of periodically starved salmon to catch up in size to continually fed salmon was not apparent from this experiment. The results differ from other studies that have shown finfish can compensate totally for lost growth. The application of the \"Adaptive\" feeding technology to other fish species including Atlantic salmon (Salmo salar) in Scotland Chinook salmon (Oncorhynchus tshawytscha) in New Zealand barramundi (Lates calcarifer) in Northern Australia and yellow-tail (Seriola quinqueradiata) in Japan was undertaken. Improvements in production performance due to satiation feeding and the use of the \"feedback\" system were noted which included a reduction in FCR 5-20% and a 10-40% reduction in production time for similar harvest weights."