Lt9 THE ATTRACTION OF FREE-RANGING RAINBOW TROUT TO A FEEDING STATION

The aim of this work was to investigate techniques for controlling the movement of free-ranging rainbow trout. In the first part of the study groups of 50 rainbow trout were successfully conditioned to aggregate at a loudspeaker and feeding point in a 3.5 m x 3.5 m x 3.5 m enclosure in response to a pulsed 140-Hz sound signal played for 10 s prior to and during feeding. The development of the response was quantified using three techniques and a study of the factors involved in conditioning showed that, (1) the number of trials was the important criteria in establishing the conditioned response, (2) other fish and the visual cues associated with the experimental environment, as well as the sound signal, were used to mediate the conditioned response and (3) an increase in the size of the conditioning enclosure adversely affected the rate of conditioning. Additional experiments showed that the sound signal amplitude was positively correlated with the number and activity of fish attracted to feed and that conditioned fish "generalized" to frequencies of between 50 and 380 Hz. Anaesthesia and handling, and 24 days without training, did not adversely 'affect the conditioned response. In the second part of the study naive fish and fish pre-conditioned to a sound signal were released into an 8-acre loch in two separate experiments. These experiments examined the movements of rainbow trout in relation to an artificial feeding station and showed that fish were attracted to and remained close to the feeding station even in the absence of a sound signal. Preconditioned fish were not attracted to feed using •sound although the experimental conditions prevailing at the time of this experiment could have adversely affected the results. A food delivery was, however, a very successful altractant and the possibility of feeding sounds being used in this attraction was investigated, but rejected. The role of other possible cues are discussed.