ABSTRACT
The purpose of present two-choice trials was to examine the capacity of groups of juvenile rainbow trout to differentiate between two isolipidic diets containing distinct oils and to detect an eventual preference. The choice was offered by means of two self-feeders per tank. One feeder distributed a standard diet with fish oil (FO), the other a diet containing vegetable oil, either rich in linolenic acid (linseed oil, LO), linoleic acid (sunflower oil, SO), or oleic acid (rapeseed oil, RO). Each 15-day preference test was preceded by a 15-day adaptation period during which both feeders distributed the same diet. The tests were followed by a 10- to 15-day validation period in order to confirm that feeder solicitations were steered by the characteristics of the diets. Preferences were expressed as relative changes in feed demands for a specific feeder. Averaged over all groups, the preference tests demonstrated the capacity of rainbow trout to discriminate between a diet with FO and a diet containing vegetable oil, and indicated a general preference for the diet with FO over the other diets irrespective of whether they received the diet with fish oil (Experiment 1) or with vegetable oil (Experiment 2) prior to the preference test. The tests also indicated a difference in the extent of relative avoidance of each of the three vegetable oil diets. Diet LO was the most avoided, as indicated by the 37-39% decrease in demands for the feeder with diet LO (P<0.05). Diet RO was the best accepted, causing a decrease in feed demands of only 15-17% (P>0.05). The avoidance of diet SO at the end of the preference test was 30% (P>0.05) after an initially higher avoidance of 43% (P<0.05). It is believed that the metabolic consequences of the excess of linolenic or linoleic acid negatively affected the feed acceptances of diets LO and SO. Further work is needed to elucidate a possible interference of differences in palatability. In all groups, the lower demands for the vegetable oil diets were compensated by increased demands for diet FO. Hence, changes in diet selection had no effect on total feed or energy intakes, measured as the sum of both selections.
