Application of a low oil pre-harvest diet to manipulate the composition and quality of Atlantic salmon, Salmo salar L.

Abstract A trial was undertaken to evaluate a low oil pre-harvest feed (LOW, 50% crude protein, 18% crude fat) as a means of manipulating the quality of Atlantic salmon growing between 2.5 and 4.5 kg in the spring/summer. For the preceding 14 months, the salmon consumed commercial-style feeds, which varied in protein and oil content (CON); feeds with variable but relatively lower lipid/protein ratios (RET); feeds with variable but relatively higher lipid/protein ratios (ADV) or feeds with a fixed, low oil/protein ratio (LOW). During the 70-day pre-harvest period, salmon previously given the CON, RET and LOW feeds were fed the LOW diet while those fed the ADV feeds were fed a high energy diet (HIGH; 40% crude protein, 36% crude fat). Salmon fed the HIGH feed achieved an SGR (0.34% day −1 ) higher than that of salmon switched to the LOW feed (CON/LOW and RET/LOW). Fish on the LOW/LOW regime were initially smaller than the others but underwent a period of catch-up growth (SGR=0.40% day −1 ) and at harvest, these fish were not significantly smaller than salmon fed the CON feeds for 14 months followed by the LOW feed for 70 days. Irrespective of the feeds fed in the preceding 14 months, feeding the LOW feed for 70 days resulted in condition factors (1.13–1.14) that were significantly lower than for the fish fed the HIGH feed (1.20). With the exception of fish on the LOW/LOW regime, carcass yields improved significantly during the pre-harvest period irrespective of the feed fed and there were no significant differences in carcass yield at harvest. Salmon fed LOW feeds from smolt input until harvest were the leanest (9.3% fat in fillet portion). There was a significant reduction in the fillet fat percentage of fish on the CON/LOW regime (reduced from 12.9 to 11.2%) but no significant effect amongst fish on the RET/LOW regime (11.8%). Salmon switched to the LOW feed pre-harvest were significantly leaner than those fed the HIGH feed (13.8%). There were no significant differences in perceived flesh colour (Roche SalmoFan™) at harvest and using the low oil, pre-harvest feed did not significantly improve instrumentally measured colour. By applying a low oil, pre-harvest feed, an adequate growth rate and FCR could be achieved for the majority of the production cycle and by manipulating the quality of the fish nearer to harvest, greater flexibility would be achieved, allowing the producer to tailor the fish to the specific fat/condition factor requirements of the market at the time of harvest.