Statistical modelling of voluntary feed intake in individual Atlantic salmon (Salmo salar L.)

Precision feeding aims to provide the correct amount of feed to farmed animals for optimal growth and performance and to avoid feed waste. However, knowledge underlying the meal-to-meal variability in voluntary feed intake of farmed species is still limited. This study examined the relationship between meals, feed deprivation time and the feed (pellets) consumed by Atlantic salmon post smolts. The data was collected from individual fish handfed to satiety without social interaction in three independent short-term (6-12 days) experiments. The fixed variables of our model (feed deprivation time (i.e., time between meals), number of pellets provided, day, previous meal size, and fish growth) explained most of the feed intake (number of pellets ingested) (R-2 0.68). Results show that fish ingested more pellets over the course of the trials as they grew, resulting in a positive correlation between feed intake and fish growth (final minus initial fish weight). The time between meals and prior meal size (the number of pellets ingested in the previous meal) significantly affected feed intake in the following meal. Our results suggest that it is possible to optimise meal size by considering the size of the previous meal and the time since it was given.