Demand feeding and growth in Salmonids

Abstract

Summary: Rainbow trout of 35g or more and 1 year old salmon smolts were acclimated to seawater in recirculating systems, or more successfully in floating cages in Dunstaffnage Bay at salinities of 25% to 30%. Rainbow trout learnt within 24 h to press a lever to operate an automatic feeder dispensing dry food pellets. They could be trained by an experimenter firing the feeder to cause food deliveries immediately adjacent to the trigger, but would also learn to use the trigger without any training , particularly when in groups rather that as individuals. Feeds were highly aggregated, bouts of rapid feeding being followed by some hours with very little feeding activity. Feeds within bouts were spaced at a mean of 4 to 8 mins. There was a marked peak of feeding activity at dusk and up to 40% of a days feeds could occur overnight. Individual rainbow trout may waste up to 50% of the food that they delivered but wastage was less marked in groups. In groups of up to 20 trout a single individual was responsible for all of the trigger pressing, though all of the fish present took the food and this reduced the amount wasted. The dominant individual with regard to trigger pressing was also dominant as regards territory. The total daily intake of food was dependent on the reward per trigger press, and a delivery of 0.1 to 0.15% of the aggregate weight of fish present gave the maximum intake with the minimum wastage. This level of intake was greater than that recommended by the food manufacturers for maximum conversion efficiency, but it could be reduced by reducing the reward levels. It is suggested that this could provide a suitable scheme for the operation of demand feeders in fish farming. No improvement in growth rates and conversion efficiency was demonstrated by demand feeding compared to twice daily hand feeding to individual rainbow trout, though there was an improvement in conversion efficiency over once daily hand feeding to groups of trout.