Skip to main content
European Commission logo print header

Researching alternatives to fish oil for aquaculture

Deliverables

The first trials with gilthead sea bream in Gran Canaria involved feeding juvenile bream diets containing 100% FO, 60% LO, 60% RO or 80% LO to triplicate groups for a period of 29 weeks. In the second dietary trials, bream were fed 100% FO or two different 60% blends and a 100% blend of rapeseed/linseed and palm oils (60% VO blends A & B and 100% VO blend) to triplicate groups for 40 weeks from initial weight of 24g. In both trials fish were grown to market size and were then fed a FO finishing diet for a further 14 weeks (trial I) or 20 weeks (trial II). In sea bream, replacement of up to 60% of FO with VO had no detrimental effects on growth or feed conversion. Replacement with 80% linseed oil or 100% of VO blend significantly reduced growth rates. However, in fish > 250-300g this reduction was not observed suggesting that in larger/older fish requirements for essential fatty acids may be lower than in smaller fish. Flesh lipid content was unaffected by dietary lipid although liver lipid was increased in sea bream fed 80% LO and the 100% VO blend. Flesh fatty acid concentrations reflected diet fatty acid concentrations, with DHA tending to be conserved, as was the case in other target species. In sea bream fed 60% VO, DHA & EPA were reduced by ~50% while in fish fed 100% VO the reduction was ~65%. Reduction of DHA and EPA was less in fish fed diets with low PUFA contents, e.g. the VO blend than in fish fed diets high in PUFA, such as LO and RO. DHA and EPA could be restored to > 90% of their values in FO fed fish, sea bream, by feeding a FO diet for 14-21 weeks. The duration of the finishing diet period was dependent on fish size, growth rate, and dietary DHA and EPA contents. These trials suggest that oils most suited as FO substitutes were high in monoenes, contain saturate levels similar to those in the fish being fed and be low in C18 PUFA, especially 18:2n-6, as this fatty acid is poorly oxidised and difficult to remove using finishing diets. High levels of dietary linseed oil increased beta-oxidation in sea bream white muscle but when using blends of VO, no clear effects on beta-oxidation capacity were found. There were no effects of replacing FO with VO on hepatocyte beta-oxidation activity or substrate specificity. There were no differences in NADPH producing enzyme activities in sea bream. Changes of plasma lipoprotein fatty acid compositions in sea bream caused by VO substitution included: Increased FA characteristic of VO (18:1,18:2 n-6 and 18:3 n-3) Decreased FA characteristic of FO (EPA, DHA) These changes occured in plasma lipoprotein fractions in the order: VLDL> LDL> HDL. In sea bream larvae there was evidence of C18-20 elongase activity but not for delta-6 or delta-5 desaturase activities. Delta-6 FAD was cloned and characterised from sea bream and PUFA ELOs were cloned and characterised from salmon and sea bream. Trends were observed for dietary VO to increase delta-6 FAD in sea bream liver but these were not significant. Sea bream showed reduced macrophage phagocytic activity when fed 60% VO compared to fish fed FO. Arachidonic acid - derived prostaglandin production was not significantly affected by feeding VO diets but EPA-derived prostaglandin production was reduced in sea bream fed 100% VO. Sea bream fed linseed oil had significantly increased basal cortisol levels in plasma and levels were also elevated in sea bream fed linseed and rapeseed oil when subjected to a crowding stress.
The first trials with European sea bass in Cadiz, Spain involved feeding juvenile bass of ~90g diets containing 100% FO, 60% LO, 60% RO or 60% OO to triplicate groups for a period of 34 weeks. In the second dietary trials, bream were fed 100% FO or two different 60% blends of rapeseed/linseed and palm oils (60% VO blends A & B) to replicate groups for 64 weeks from initial weight of 5g. In both trials fish were grown to market size and were then fed a FO finishing diet for a further 14 weeks (trial I) or 20 weeks (trial II). In sea bass, replacement of up to 60% of FO with VO had no detrimental effects on growth or feed conversion. Flesh and liver lipid content were unaffected by dietary lipid but flesh fatty acid concentrations closely reflected diet fatty acid concentrations, with DHA tending to be conserved. In sea bass fed 60% VO, DHA & EPA were reduced by ~50%. In general, reduction of DHA and EPA was less in fish fed diets with low PUFA contents, e.g. olive oil, and the VO blend than in fish fed diets high in PUFA. DHA and EPA could be restored to > 90% of their values in FO fish, in sea bass, by feeding a FO finishing diet for 14-20 weeks. The duration of the finishing diet period was dependent on fish size, growth rate, and dietary DHA and EPA contents. Our results suggest that oils most suited as FO substitutes should be high in monoenes, contain saturate levels similar to those in the fish being fed and be low in C18 PUFA, especially 18:2n-6, as this fatty acid is poorly oxidised and difficult to remove using finishing diets. Using blends of VO, no clear effects on beta-oxidation capacity were found in any of the species. In sea bass, hepatocyte HUFA synthesis was very low, ~ 5- to 20-fold lower than in salmon. In sea bass, HUFA synthesis was increased by dietary VO in enterocytes but not hepatocytes. Trends were observed for dietary VO to increase delta-6 FAD in sea bass liver; however, this was only significant for one of the 60% vegetable oil blends tested. Sea bass showed reduced macrophage phagocytic activity when fed 60% VO. Arachidonic acid - derived prostaglandin production was reduced in sea bass fed VO diets, although this was less affected when feeding a VO blend than by single VO replacement diets. When using a single VO to replace FO, a number of immune parameters (haematocrit, leucocyte numbers, macrophage respiratory burst) were altered by dietary treatment, in sea bass. However, when a VO blend was chosen to replace FO, immune functions were not affected up to 60% replacement in sea bass.
The first salmon trials in Scotland involved feeding post-smolts diets containing FO and where FO was replaced by 25, 50, 75 & 100% linseed oil (LO) plus a single replacement with 50% olive oil (OO). In the second dietary trials, salmon were fed FO or a 75% replacement of a blend of rapeseed/linseed and palm oils (75% VO blend) to replicate groups over the whole production cycle from first feeding. In both trials fish were grown to market size and were then fed a FO finishing diet for a further 24 weeks. In salmon, replacement of fish oil (FO) with vegetable oil (VO), up to 100%, did not affect growth and feed conversion. Flesh lipid content was unaffected by dietary lipid although, in liver, lipid was increased in salmon fed > 50% LO but not in salmon fed 75% of the VO blend. Flesh fatty acid concentrations reflected diet fatty acid concentrations, with DHA being highly conserved. However, in salmon fed 100% VO, flesh DHA & EPA concentrations were reduced by ~65%. It was notable that reduction of DHA and EPA was less in fish fed diets with low PUFA contents, e.g. olive oil, and the VO blend than in fish fed diets high in PUFA e.g. RO & LO. DHA and EPA values could be restored to >80% of the values in salmon fed FO by feeding a 100% FO finishing diet for 16-24 weeks with the duration of the finishing diet period being dependent on fish size, growth rate, and dietary DHA and EPA contents. Therefore, the oils most suited as FO substitutes should be high in monoenes, contain saturate levels similar to those in the fish species being fed and be low in C18 PUFA, especially 18:2n-6, as this fatty acid is poorly oxidised and difficult to remove using finishing diets. In salmon hepatocytes enzyme activities decrease along the synthetic pathway pathway such that delta-6 fatty acid desaturase (FAD) = C18-20 fatty acid elongase (ELO) > delta-5 FAD > C20-22 ELO > C22-24 ELO = delta-6*FAD. In salmon hepatocytes HUFA synthesis was increased by dietary VO mainly at delta-6 FAD and C18-20 ELO. DHA synthesis as measured by conversion of labelled 18:3n-3 occurred at a low rate in vivo in trout and salmon. DHA synthesis was increased by feeding VO, certainly in trout and possibly also in salmon. However, DHA synthesis was lower in Atlantic salmon parr than in equivalent sized rainbow trout and DHA synthesis decreased, as fish became larger/older. It is notable that, even in fully induced trout, DHA synthesis was insufficient to maintain carcass levels of DHA. The majority of labelled 18:3n-3 (>95%) fed to trout and salmon was oxidized. The delta-6 FAD, delta-5 FAD and PUFA ELOs were cloned and characterised from salmon. Hepatic delta-6 and delta-5 FAD gene expressions were increased by dietary VO in salmon but ELO expression was increased only in salmon fed LO. Arachidonic acid - derived prostaglandin production was reduced in salmon fed VO diets, although this was less affected when feeding a VO blend than in single VO diets. When salmon were fed a single VO to replace FO, a number of immune parameters (haematocrit, leucocyte numbers, macrophage respiratory burst) were altered by dietary treatment. However, when a VO blend was chosen to replace FO, non-specific immune functions were not affected up to 75% replacement in salmon. Cataract incidence in salmon fed the 75% VO blend was 4 times higher compared to fish fed FO. Vaccination efficacy against furunculosis in salmon was not affected by the dietary VO while salmon challenged with Vibrio anguillarum did not show significant differences in mortality when FO was replaced by a 75% VO blend.
The first salmon trials in Norway involved feeding post-smolts diets containing FO, and where FO were replaced by 25, 50, 75 & 100% rapeseed oil (RO) plus a single replacement with 50% olive oil (OO). In the second dietary trials, salmon were fed FO or a 100% replacement of a blend of rapeseed, linseed and palm oils (100% VO blend) to replicate groups over the whole production cycle from first feeding. In both trials fish were grown to market size and were then fed a FO finishing diet for a further 24 weeks. In salmon, replacement of fish oil (FO) with vegetable oil (VO), up to 100%, did not affect growth and feed conversion. Salmon fed a 100% VO blend had significantly higher final weights than fish fed FO. Flesh lipid content was unaffected by dietary lipid in all trial groups but flesh fatty acid concentrations reflected diet fatty acid concentrations, with DHA tending to be highly conserved. In salmon fed 100% VO, flesh DHA & EPA concentrations were reduced by ~65% although reduction of DHA and EPA was less in fish fed diets with low PUFA contents, e.g. olive oil, and the VO blend than in fish fed diets higher in PUFA such as RO. However, DHA and EPA values could be restored to >70% of the values in salmon fed FO by feeding a 100% FO finishing diet for 16-24 weeks and the duration of the finishing diet period was dependent on fish size, growth rate, and dietary DHA and EPA contents. Therefore, we would suggest that oils most suited as FO substitutes should be high in monoenes, contain saturate levels similar to those in the fish being fed and be low in C18 PUFA, especially 18:2n-6, as this fatty acid is poorly oxidised and difficult to remove using finishing diets. Dietary rapeseed oil added at up to 75% replacement increased- oxidation in salmon white and red muscle. - Hepatic-oxidation increased during smoltification of Atlantic salmon, followed by a decreased --oxidation capacity in red muscle although there were no effects of replacing FO with VO on hepatocyte - -oxidation activity or substrate specificity. There were no differences in NADPH producing enzyme activities in salmon nor were any differences observed in fatty acid synthetase activity. The cholesterol level in salmon plasma during the freshwater stage decreased with VO substitution and the level of plasma LDL lipid decreased with VO substitution. However, VO substitution caused no changes in plasma triglycerides. Changes of plasma lipoprotein fatty acid compositions in Atlantic salmon and caused by VO substitution included: increased FA characteristic of VO (18:1,18:2 n-6 and 18:3 n-3); decreased FA characteristic of FO (EPA, DHA). These changes occured in plasma lipoprotein fractions in the order: VLDL> LDL> HDL. When using a single VO to replace FO, a number of immune parameters (haematocrit, leucocyte numbers, macrophage respiratory burst) were altered by dietary treatment, in salmon. Cataract incidences in salmon fed the 100% VO blend was 5 times higher compared to fish fed FO. However, vaccination efficacy in salmon was not affected by the dietary VO.
The first trials with rainbow trout (RT) in France involved feeding juvenile trout diets containing 100% FO, 100% LO, 100% RO or 50% OO to triplicate groups for a period of 12 weeks. In the second dietary trials, RT were fed 100% FO or 75% and 100% replacement with a blend of rapeseed/linseed and palm oils (75% VO blend and 100% blend) to triplicate groups over the whole production cycle from first feeding. In both trials fish were grown to market size and were then fed a FO finishing diet for a further 12 weeks (trial I) or 24 weeks (trial II). In RT, replacement of fish oil (FO) with vegetable oil (VO), up to 100%, did not significantly affect growth and feed conversion. Flesh lipid content was unaffected by dietary lipid and flesh fatty acid concentrations reflected diet fatty acid concentrations, with DHA tending to be highly conserved. In trout fed 100% VO, flesh DHA & EPA concentrations were reduced by ~50% although reduction of DHA and EPA was less in fish fed diets with low PUFA contents, e.g. olive oil, and the VO blend, than in fish fed diets high in PUFA. DHA and EPA could be restored to > 90% of their values in FO fish, in RT, by feeding a FO finishing diet for 12-24 weeks. The duration of the finishing diet period was dependent on fish size, growth rate, and dietary DHA and EPA contents. This data would suggest that oils most suited as FO substitutes should be high in monoenes, contain saturate levels similar to those in the fish being fed and be low in C18 PUFA, especially 18:2n-6, as this fatty acid is poorly oxidised and difficult to remove using finishing diets. High levels of dietary linseed oil increased beta-oxidation in RT white muscle. However, using blends of VO, no clear effects on beta-oxidation capacity were found in trout or any of the target species. Replacement of FO with VO did not affect tissue lipid uptake in RT. In addition, no differences were seen in lipoprotein lipase activity in muscle and adipose tissue of RT. Delta-6 FAD was cloned and characterised from RT although there was no evidence for dietary VO increasing delta-6 FAD in trout liver.
A primary objective of the RAFOA project was to determine the organoleptic properties of cooked fillets of all target species using expert taste panels in the region of origin of the fish. The studies on salmon and trout also included taste panel evaluation of smoked product and for salmon also assessment of raw fish (sashimi) by Japanese taste panels. A second aim was to relate these subjective measurements in cooked fillets to non-subjective measurements based on electronic detection of volatile compounds in the fillets. Finally, it was proposed to determine shelf life properties and technological properties of fresh fish in response to dietary fatty acid composition. All analyses were conducted on market size fish and the effect of feeding a finishing diet, containing only fish oil, on quality aspects was also included. High levels of VO in the diet had only minor effects on the organoleptic properties of: Salmon, whether raw, cooked or smoked; in some cases fish fed VO diets were preferred to fish fed 100% FO, Trout whether cooked or smoked, Bream, cooked, Bass, although lower scores were observed in very fresh fish. Electronic nose analyses discriminated between fish fed different oils with best separation in trout: A FO finishing diet reduced the discrimination substantially; Replacing FO with VO had no or only minor effects on product quality and storage stability characteristics with: Time on ice being a more important factor for all species tested, than dietary treatment, Feeding a 100% FO-finishing diet prior to harvest eliminating earlier minor effects of feeding VO diets.

Searching for OpenAIRE data...

There was an error trying to search data from OpenAIRE

No results available