bell et al, 2003

Nutrient Metabolism

Altered Fatty Acid Compositions in Atlantic Salmon (Salmo salar) Fed DietsContaining Linseed and Rapeseed Oils Can Be Partially Restored by aSubsequent Fish Oil Finishing Diet

J. Gordon Bell,1 Douglas R. Tocher, R. James Henderson, James R. Dickand Vivian O. Crampton*

Institute of Aquaculture, University of Stirling, Stirling, FK9 4LA, Scotland, UK and *EWOS Innovation,N-4335, Dirdal, Norway

ABSTRACT Atlantic salmon postsmolts were fed a control diet or one of 9 experimental diets containing variousblends of two vegetable oils, linseed (LO) and rapeseed oil (RO), and fish oil (FO) in a triangular trial design, for 50wk. After sampling, fish previously fed 100% FO, LO and RO were switched to a diet containing 100% FO for afurther 20 wk. Fatty acid compositions of flesh total lipid were linearly correlated with dietary fatty acid composi-tions (r 0.991.00, P 0.0001). Inclusion of vegetable oil at 33% of total oil significantly reduced theconcentrations of the highly unsaturated fatty acids, eicosapentaenoate [20:5(n-3)] and docosahexaenoate [22:6(n-3)], to70 and 75%, respectively, of the values in fish fed 100% FO. When vegetable oil was included at 100%of total dietary lipid, the concentrations of 20:5(n-3) and 22:6(n-3) were significantly reduced to 30 and 36%,respectively, of the values in fish fed FO. Transfer of fish previously fed 100% vegetable oil to a 100% FO diet for20 wk restored the concentrations of 20:5(n-3) and 22:6(n-3) to 80% of the value in fish fed 100% FO for 70 wk,although the values were still significantly lower. However, in fish previously fed either 100% LO or RO, concen-trations of 18:2(n-6) remained 50% higher than in fish fed 100% FO. This study suggests that RO and LO can beused successfully to culture salmon through the seawater phase of their growth cycle; this will result in reductionsin flesh 20:5(n-3) and 22:6(n-3) concentrations that can be partially restored by feeding a diet containing onlymarine FO for a period before harvest. J. Nutr. 133: 27932801, 2003.

KEY WORDS: Atlantic salmon linseed oil rapeseed oil polyunsaturated fatty acids fish oil

Almost 30 y ago, it was discovered that native Inuits wereapparently protected from cardiovascular disease by their highfish intake (1,2). The importance of the (n-3) highly unsat-urated fatty acids (HUFA)2 in human nutrition has led toconsiderable research effort in the intervening years. Subse-quently, the efficacy of (n-3) HUFA in the prevention orattenuation of many of the inflammatory conditions that areprevalent in the developed world, including rheumatoid ar-thritis, atopic illness, inflammatory bowel disease and variousneurological conditions, has been established (3,4).

Fish, and particularly those with oily flesh, such as herring,mackerel and salmon, represent a rich, and almost uniquesource of the (n-3) HUFA, especially eicosapentaenoic acid[20:5(n-3), EPA] and docosahexaenoic acid [22:6(n-3),DHA]. Demand for fish products is increasing, yet the tradi-tional capture fisheries are in decline worldwide (5) such thatthe potential shortfall in fish products must be met by aquac-ulture production (6). Aquaculture production currently uses

60% of global fish oil production and by 2010 85% will beconsumed in aquaculture feeds (7). Future expansion of aquac-ulture, and in particular salmon production, can continue onlyif suitable and sustainable alternatives to fish oil (FO) aredeveloped and introduced.

In Atlantic salmon feeds, dietary lipids are a major source ofenergy (8) and also must provide essential PUFA, including18:3(n-3) and 18:2(n-6), and especially EPA and DHA, toallow normal growth and development of cells and tissues(9,10). Measurement of -oxidation capacity in salmonid fishsuggests that saturated and monounsaturated fatty acids(MUFA), especially 16:0, 16:1, 18:1(n-9) and 22:1(n-11), aswell as 18:2(n-6), are the preferred substrates for energy pro-duction (11,12). A number of vegetable oils are rich in 16:0,18:1(n-9) and 18:2(n-6), and in some cases 18:3(n-3) also,whereas, by comparison, Northern hemisphere fish oils are richin the long-chain monoenes, 20:1(n-9) and 22:1(n-11). Interms of providing the essential fatty acids (EFA) required fornormal growth and development, several studies have shownthat salmon can be raised successfully when they consumediets that contain high levels of vegetable oils for periods up to30 wk (1315). In addition, the ability to convert 18:3(n-3)and 18:2(n-6) provided by the diet to their longer-chainHUFA products, including 20:5(n-3), 22:6(n-3) and 20:4(n-6), has been established (1417). However, recent evidence in

1 To whom correspondence should be addressed.E-mail: email [email protected].

2 Abbreviations used: DHA, docosahexaenoic acid; EFA, essential fatty acid;EPA, eicosapentaenoic acid; FO, fish oil; HUFA, highly unsaturated fatty acids;LO, linseed oil; MUFA, monounsaturated fatty acids; RO, rapeseed oil; TAG,triacylglycerol.

0022-3166/03 $3.00 2003 American Society for Nutritional Sciences.Manuscript received 31 March 2003. Initial review completed 25 April 2003. Revision accepted 25 June 2003.

2793

by on March 29, 2007

jn.nutrition.orgD

ownloaded from

rainbow trout suggests that the capacity for endogenous pro-duction of HUFA may not fulfill requirements such that, foroptimal growth and well being of the fish, some dietary 20:5(n-3) and 22:6(n-3) will be required (18). There is alsoevidence that the capacity to synthesize essential HUFA maydecline with age in fish, as it does in humans (18,19).

Atlantic salmon produced using diets containing only ma-rine fish oils are of high nutritional quality; they are rich in(n-3) HUFA and have a high (n-3)/(n-6) PUFA ratio of 4:1(2022). Potential changes in product quality resulting fromvegetable oil inclusion in salmon feeds should be minimized ifthe health benefits of salmon in the human diet are to bemaintained. To achieve these goals, any oil replacing fish oilshould provide sufficient energy in the form of MUFA, andother easily catabolized fatty acids, to maintain high growth,while maximizing endogenous conversion of 18:3(n-3) to 20:5(n-3) and 22:6(n-3). Rapeseed oil (RO) is a potential can-didate for FO substitution because it has moderate levels of18:2(n-6) and 18:3(n-3), and is rich in 18:1(n-9). In addition,the ratio of 18:3(n-3)/18:2(n-6) in rapeseed oil of 1:2 is re-garded as beneficial to human health and should not be det-rimental for fish health, provided EPA and DHA are alsopresent from dietary fishmeal (23). Linseed oil is also a poten-tial candidate for FO replacement because it is rich in -lin-olenic acid [18:3(n-3)], is the substrate for synthesis of (n-3)HUFA and also contains significant levels of 18:2(n-6) suchthat it has an 18:3(n-3)/18:2(n-6) ratio of 34:1. -Linolenicacid also possesses anti-inflammatory properties due to inhibi-tion of arachidonic acid synthesis from 18:2(n-6) at the 6-desaturase step and by inhibition of cyclooxygenase (24,25).

In the present study, Atlantic salmon postsmolts werestocked into 12 seawater pens and fed one of nine experimen-tal diets, containing blends of LO, RO and FO, or a controldiet containing only FO, which was fed in triplicate. The fishwere sampled after 50 wk and the fish previously fed the 100%FO, 100% LO and 100% RO diets were fed for a further 20 wka diet containing only FO to follow the dilution of fatty acidsfrom the vegetable oils.

MATERIALS AND METHODS

Fish and experimental diets. Atlantic salmon postsmolts (n 7200) of initial weight 120 10 g (individual weights of 100

fish/pen) were distributed into 12 cages (5 5 m; 600 fish per cage)in Loch Duich, Lochalsh, Scotland, in February 1999. The temper-ature over the experimental period (February 1999July 2000) rangedfrom 5.5 to 19.6C with a mean temperature of 10.8 3.4C. Thefish were allowed to acclimate for 2 wk before each pen was randomlyassigned one of 10 diets including a control diet with FO alone, whichwas fed to three pens, and nine experimental diets containing blendsof FO, LO and RO. In the period up to 32 wk, feed was distributedmanually, whereas in the period from 32 to 50 wk, the fish wereswitched to automatic feeders controlled by AKVAsmart pelletcounters (AKVAsmart ASA, Bryne, Norway.). Problems with theautomatic feeders meant that 5 pens also received some feed manu-ally. The 10 practical-type extruded diets were formulated (EwosTechnology Centre, Livingston, Scotland) into 3- and 6-mm pellets,to provide 47.0% crude protein, 24.1% crude lipid and 7.6% moistureand 41.8% protein, 30.5% crude lipid and 6.8% moisture, respectively(Table 1). The added oil comprised 100% FO, 100% LO, 100% RO,FO/RO (2:1 and 1:2 w/w), FO/LO (2:1 and 1:2 w/w), RO/LO (2:1 and1:2 w/w) and FO/RO/LO (1:1:1 w/w/w) forming a triangular experi-mental design. The graded substitution of the three dietary oils wasreflected in the fatty acid compositions of the 10 diets (Table 2).Thus, when RO was added, there were increased concentrations of18:1(n-9) and 18:2(n-6) and, to a lesser degree, of 18:3(n-3). Simi-larly, increasing inclusion of LO increased concentrations of 18:3(n-3) and 18:2(n-6), with the precise amount of the latter depend-ing on whether LO was replacing FO or RO. In all cases in which FOwas replaced by a vegetable oil, there was a graded reduction in20:5(n-3) and 22:6(n-3), as well as 16:0, 20:1(n-9), 22:1 and 20:4(n-6). After sampling at 50 wk, the remaining fish in pens fed the 100%FO, RO and LO diets were redistributed into three pens, with eachpen containing 100 fish from each treatment. The fish were markedby fin clipping to identify their previous dietary history. All threepens were fed a diet containing 100% FO, in a 9-mm pellet, for afurther 20 wk. The composition of this diet was essentially similar tothe 6-mm 100% FO diet. The experimental diets differed only intheir oil composition and were formulated to satisfy the nutritionalrequirements of salmonid fish, including (n-3) EFA, which are pro-vided both by the inclusion of dietary fishmeal and by 18:3(n-3)present in LO and RO (26). The experiment was conducted inaccordance with the British Home Office guidelines regarding re-search on experimental animals.

Sampling procedure. After 50 wk, 18 fish were selected atrandom from each cage. Fish were killed by a blow to the head anda sample of muscle, representative of the edible portion, was obtainedby cutting a steak between the leading and trailing edges of the dorsalfin. The samples from each cage were pooled, by weight, in three

TABLE 1

Feed components

Component FOFO/LO(2:1)

FO/LO(1:2) LO

FO/RO(2:1)

FO/RO(1:2) RO

LO/RO(2:1)

LO/RO(1:2)

FO/LO/RO(1:1:1)

g/100g

Fishmeal1 53.8 53.8 53.8 53.8 53.8 53.8 53.8 53.8 53.8 53.8Soya (Hi Pro)2 7.6 7.6 7.6 7.6 7.6 7.6 7.6 7.6 7.6 7.6Wheat3 14.0 14.0 14.0 14.0 14.0 14.0 14.0 14.0 14.0 14.0Fish oil (FO)4 23.6 15.7 7.9 0 15.7 7.9 0 0 0 7.9Rapeseed oil (RO)5 0 0 0 0 7.9 15.7 23.6 7.9 15.7 7.9Linseed oil (LO)5 0 7.9 15.7 23.6 0 0 0 15.7 7.9 7.9Micronutrients6 1 1 1 1 1 1 1 1 1 1

1 Norsemeal, London SW6, UK.2 Grosvenor Grain, Perth, UK.3 Stewarts of Larbert, Larbert, UK.4 United Fish Products, Aberdeen, UK.5 Meade-King Robinson and Company, Liverpool, UK.6 Vitamins, minerals and astaxanthin (Roche Products, Heanor, Derbyshire, England to specification by Ewos, Bathgate, UK. Astaxanthin, as

Carophyll pink, 50 mg/kg diet; vitamin A as retinyl acetate, 2.6 mg/kg; vitamin D as cholecalciferol, 0.1 mg/kg and vitamin E as dl--tocopherolacetate, 250 mg/kg. All other vitamin and minerals provided at levels in excess of nutritional requirements for salmonid fish (26).

BELL ET AL.2794


jn.nutrition.orgD

ownloaded from

pools of six fish each; that is, pool A comprised the six smallest fish,pool B, the next six largest fish and pool C, the largest six fishsampled. The pooled steaks were frozen and stored at 25C untilprocessed. The six steaks were then skinned, deboned and the musclehomogenized in a food processor, after removal of the dorsal fat body.The homogenate was immediately frozen and samples were stored at40C before analysis. After the 20-wk washout period, 15 fish pertreatment (5 per pen) were sampled as described above except thateach sample was an individual fish. Each steak was processed and thehomogenate stored as described above.

Lipid extraction and fatty acid analysis. Total lipid was ex-tracted from 2 g of homogenized muscle by homogenizing in 20volumes of chloroform/methanol (2:1, v/v) in an Ultra-Turraxtissue disrupter (Fisher Scientific, Loughborough, UK). Total lipidwas prepared according to the method of Folch et al. (27) andnonlipid impurities removed by washing with 8.8 g/L KCl. Theweight of lipid was determined gravimetrically after evaporation ofsolvent and overnight desiccation in vacuo. FAME were preparedby acid-catalyzed transesterification of total lipid according to themethod of Christie (28). Extraction and purification of FAMEwere performed as described by Ghioni et al. (29). FAME wereseparated and quantified by GLC (Carlo Erba Vega 8160, Milan,Italy) using a 30 m 0.32 mm i.d. capillary column (CP Wax52CB, Chrompak, London, UK). Hydrogen was used as carrier gasand temperature programming was from 50 to 150C at 40C/minand then to 230C at 2.0C/min. Individual methyl esters wereidentified by comparison with known standards and by reference topublished data (30). Data were collected and processed using theChromcard for Windows (version 1.19) computer package (Ther-moquest Italia S.p.A., Milan, Italy).

Statistical analysis. Significance of difference (P 0.05) be-tween dietary treatments was determined by one-way ANOVA. Dif-ferences between means were determined by Tukeys test. Data iden-tified as nonhomogeneous, using Bartletts test, were subjected to logor arcsin transformation before applying the ANOVA. ANOVA andregression analysis was performed using a GraphPad Prism (version3.0) statistical package (GraphPad Software, San Diego, CA).

RESULTS

Weights of fish sampled after 32 wk did not differ (resultsnot shown, P 0.05). However, at 50 wk, the range of weights(Table 3) had increased such that there were differencesbetween treatments (P 0.0004). Despite these differences inweight, regression analyses showed no correlation betweenfinal weight and any dietary fatty acid [including 16:0, 18:1(n-9), 18:2(n-6), 18:3(n-3), 20:5(n-3) and 22:6(n-3)], indicatingthat dietary treatment was not responsible for the differencesin final weight (results not shown). The differences in finalweight may be explained by feeding methodology. Logisticalproblems encountered with the automatic feeders, at the trialfacility, meant that some treatments had to be fed by hand,which may have affected ration delivery and, thereby, finalweight. There were no differences in final weight among thethree replicate FO pens (2.07 0.56, 2.16 0.72 and 2.10 0.59 kg).

After 50 wk, flesh lipid concentrations did not differ amongthe dietary groups and there was no correlation between fleshlipid concentration and fish weight (Table 3). Similarly, forthe fish from the 100% FO, LO and RO groups that consumedthe FO diet for a further 20 wk, there were no differences inflesh lipid concentration or final weight (Table 3).

The fatty acid compositions of muscle total lipid reflectedthe variation in dietary lipid (Table 4). Plots of fatty acidconcentration (g/100 g) in flesh lipid (Table 4) against fattyacid concentration (g/100 g) in dietary lipid (Table 2) resultedin straight lines; typical examples of these are shown in Fig-ures 1 and 2. Thus, the concentrations of fatty acids in dietarylipid were related linearly to their concentrations in flesh totallipid although the slopes of the plots in Figures 1 and 2 and inTable 5 are different for each fatty acid. This indicates thatthe relationship between the concentration of a fatty acid in

TABLE 2

Fatty acid compositions of the ten experimental diets

FAME/dietFO

(050 wk)FO/LO(2:1)

FO/LO(1:2) LO

FO/RO(2:1)

FO/RO(1:2) RO

LO/RO(2:1)

LO/RO(1:2)

FO/LO/RO(1:1:1)

FO diet(5070 wk)

g/100 g total fatty acids

14:0 5.6 3.6 2.2 1.1 3.7 2.3 1.2 1.2 1.1 1.9 5.216:0 13.6 11.6 9.6 7.6 11.7 9.4 7.2 7.7 7.6 8.8 14.118:0 2.5 2.9 3.4 3.6 2.3 2.0 1.9 3.0 2.4 2.7 2.9

saturates1 24.0 20.2 16.7 13.2 19.9 15.5 11.4 13.0 10.9 14.9 23.116:1(n-7) 5.5 3.5 2.2 1.0 3.8 2.5 1.3 1.2 1.3 1.9 5.518:1(n-9) 15.1 16.1 16.4 16.6 26.7 37.5 48.3 25.9 36.9 28.8 16.720:1(n-9) 9.6 6.0 3.7 1.9 6.5 4.6 3.0 2.2 2.6 3.5 8.822:12 13.2 8.2 4.8 2.3 8.3 5.4 2.7 2.6 2.7 4.3 11.5

monoenes3 44.3 34.6 27.6 22.1 46.2 50.6 55.6 32.3 45.5 39.0 43.118:2(n-6) 4.5 7.9 11.4 13.6 9.5 14.0 17.9 14.6 16.3 13.7 4.920:4(n-6) 0.6 0.4 0.3 0.2 0.5 0.3 0.2 0.2 0.2 0.2 0.6

(n-6)4 5.5 8.7 11.8 13.9 10.4 14.6 18.1 15.0 16.8 14.1 6.018:3(n-3) 1.7 18.6 33.0 45.4 4.7 6.6 8.9 32.9 20.2 22.0 1.818:4(n-3) 2.8 1.8 1.0 0.5 1.9 1.1 0.4 0.4 0.4 0.9 2.720:4(n-3) 0.8 0.6 0.3 0.1 0.6 0.4 0.1 0.2 0.1 0.2 0.920:5(n-3) 7.3 5.3 3.3 1.8 5.7 3.9 1.9 2.2 2.1 3.1 7.022:5(n-3) 1.2 1.0 0.6 0.3 1.1 0.6 0.3 0.4 0.4 0.5 1.322:6(n-3) 10.5 8.4 5.3 2.7 8.7 5.9 2.9 3.4 3.5 4.9 10.8

(n-3) 24.5 35.8 43.6 50.8 22.8 18.6 14.5 39.6 26.7 31.5 24.7(n-3)/(n-6) 4.4 4.1 3.7 3.7 2.2 1.3 0.8 2.6 1.6 2.2 4.1

1 Includes 10:0, 12:0, 17:0 and 22:0.2 Includes 22:1(n-13) and 22:1(n-11).3 Includes 14:1, 17:1, 20:1(n-7), 22:1(n-9) and 24:1.4 Includes 18:3(n-6), 20:2(n-6), 20:3(n-6) and 22:5(n-6).

SUBSTITUTING FISH OIL WITH VEGETABLE OILS IN SALMON 2795


jn.nutrition.orgD

ownloaded from

the diet and the concentration in flesh is different for eachfatty acid. This is demonstrated more clearly in Table 5 fromthe differences ( values) between the concentration of indi-vidual fatty acids in dietary lipid and flesh lipid for fish fed thediets containing 100% FO, 100% LO and 100% RO. Thisdemonstrates that, of all the fatty acids shown in Table 5, only22:6(n-3) was present in flesh at a concentration equal to orgreater than that in the diet for all three treatment groups.This is in contrast to the other (n-3) PUFA, 18:3(n-3) and20:5(n-3), which were always present at lower concentrationsin the flesh compared with the diets. Oleic acid [18:1(n-9)]was the only other fatty acid showing a positive value for FOand LO, and these were negligible. The saturated fatty acidswere also preferentially retained when present at low concen-trations as in the LO and RO treatments. In contrast to22:6(n-3), the fatty acids:18:1(n-9), 18:2(n-6), 18:3(n-3) and22:1 were preferentially discriminated against in flesh relativeto diet when present at high concentrations in dietary lipid.Thus, when a particular fatty acid was abundant in the diet,such as 18:3(n-3) in 100% LO or 22:1 in 100% FO, that fattyacid was not selected for deposition in the flesh but wasselectively utilized for metabolism, probably for energy produc-tion.

The concentrations of 14:0, 16:0, total saturates, 16:1(n-7),20:1(n-9), 22:1, 24:1, 20:4(n-6), 20:5(n-3), 22:5(n-3) and22:6(n-3) in flesh total lipids were significantly greater in fishpreviously fed 100% FO than in those previously fed either100% LO or 100% RO (Table 6). The concentrations of 18:0and total (n-3) PUFA were significantly different for eachtreatment, with the highest levels in fish previously fed the100% LO diet followed by the 100% FO diet and the lowestlevel in fish fed the 100% RO diet. The concentrations of18:1(n-9) and total monoenes were significantly higher in fish

previously fed the 100% RO diet compared with both othertreatments, whereas 20:3(n-3) and total PUFA were signifi-cantly greater in fish fed 100% LO compared with both othertreatments. The concentration of 18:3(n-3) was significantlydifferent for each treatment, with the highest values in fishpreviously fed the 100% LO diet followed by 100% RO andthe lowest values in fish fed 100% FO. The concentrations of18:2(n-6), 20:2(n-6) and total (n-6) PUFA were significantlydifferent for all treatments with the highest values in fishpreviously fed 100% RO, followed by 100% LO, and thelowest concentration in fish fed 100% FO. The concentrationof 20:4(n-3) was significantly greater in fish previously fed100% FO and 100% LO compared with fish fed 100% RO.

After consumption of the experimental diets for 50 wk, the18:2(n-6) concentration in flesh was highest in fish fed the100% RO diet, followed by those fed the 100% LO diet andlowest in those fed the 100% FO diet (Fig. 3A). After con-suming the washout diet for 20 wk, the fish previously fed the100% LO and RO diets still had significantly higher flesh18:2(n-6) concentrations, although the values had declined by40% compared with fish fed the 100% FO diet throughout.A similar result was observed for 18:3(n-3) for which concen-trations in fish previously fed the 100% LO and RO diets werestill significantly higher than those fed 100% FO but haddeclined by 64 and 56% for fish fed LO and RO, respectively,after the 20-wk washout period (Fig. 3B).

After consumption of the experimental diets for 50 wk, theconcentration of 20:5(n-3) in flesh was significantly lower infish fed the 100% RO and LO diets than in those fed 100% FO(Fig. 4A). After consuming the washout diet for 20 wk, the20:5(n-3) concentration in flesh was still significantly lower infish previously fed the 100% LO and RO diets compared withthose that had been fed the 100% FO diet for 70 wk (Fig. 4A).However, the 20:5(n-3) concentration was increased 1.6-fold after the washout period, which was 79% of the value infish fed the 100% FO diet for 70 wk. A similar result was seenfor 22:6(n-3), although after 50 wk, the concentration in thefish fed 100% LO was significantly lower than in fish fed 100%RO (Fig. 4B). Consumption of the washout diet for 20 wkincreased flesh 22:6(n-3) concentrations by 100% and theconcentration in fish previously fed 100% LO or RO was 78%of that in fish fed 100% FO for 70 wk (Fig. 4A).

DISCUSSION

In two previous studies in which FO was substituted witheither rapeseed oil or palm oil (14,15), results obtained sug-gested that substantial replacement of FO could occur withoutdetrimental effects on fish health or product quality. However,these studies lasted for a relatively short time, with fish grownto a final weight of400 g. In the present study, salmon smoltswith an initial weight of 120 g were grown to their harvestweight of2 kg using diets in which the added oil componentcontained between 33 and 100% LO and/or RO. Althoughneither RO and LO contain any (n-3) HUFA, they do containsubstantial levels of -linolenic acid, which is also thought tobe beneficial for human health (23,31).

There were no significant effects of dietary treatment onfish weights when measured after 32 wk (results not shown)but by 50 wk, the weights were significantly different andranged from 1.92 to 2.59 kg. However, regression analysisshowed no relationship between final weight and any dietaryfatty acid [including 16:0, 18:1(n-9) or total monoenes], sug-gesting that dietary treatment was not responsible for thedifferences (results not shown).

A number of earlier studies utilized RO in feed formulations

TABLE 3

Effects of feeding diets containing various levels of fish oil(FO), linseed oil (LO) and rapeseed oil (RO) for 50 wk on fishweights and flesh lipid concentration from Atlantic salmon

before and after a 20-wk period of feeding a dietcontaining 100% FO1

Diet

50 wk 70 wk

Flesh lipid2 Weight3 Flesh lipid4 Weight4

g/100 g kg g/100 g kg

FO 8.6 0.65 2.1 0.6cd 9.9 2.0 3.5 0.5FO/LO (2:1) 7.6 1.9 2.3 0.8bcFO/LO (1:2) 9.5 0.5 2.3 0.6bcLO 7.5 0.5 2.1 0.6cd 9.9 2.3 3.5 0.5FO/RO (2:1) 8.1 0.4 2.0 0.6dFO/RO (1:2) 7.8 0.8 2.4 0.6abRO 7.0 0.6 2.6 0.8a 9.7 2.2 3.3 0.5LO/RO (2:1) 7.0 1.4 2.4 0.7abLO/RO (1:2) 7.5 0.4 2.3 0.7bcFO/LO/RO

(1:1:1) 6.9 0.8 1.9 0.6d

1 Values are means SD. Means in a column with superscriptswithout a common letter differ, P 0.05.

2 Lipid concentration at 50 wk, n 3 (each observation comprisesa pool of 6 individual fish).

3 Weights at 50 wk, n 200 individual fish.4 n 15 individual fish.5 Mean of three replicate pens (8.5 1.05, 8.0 1.33 and 9.2

1.87 g lipid/100 g flesh).

BELL ET AL.2796


jn.nutrition.orgD

ownloaded from

for salmonids, and no changes in growth rates and feed con-version were observed (17,3234). Linseed oil has also beenused, either alone or blended with RO, with no apparentdetrimental effects on growth (16,17,35,36). However, someof these studies used purified or semipurified diets containingrelatively low dietary lipid levels, and they were conducted forrelatively short periods of time compared with the presentstudy. A number of recent studies have used high energy/lipidformulations with various vegetable oils and found no detri-mental effects on growth, although these were also of shorterduration than the present study (1315).

In the present study, flesh lipid content ranged from 6.9 to9.5% although there were no significant differences betweendietary treatments (P 0.099). In previous studies in whichRO or palm oil was substituted for FO, flesh lipid concentra-tion was reduced in fish fed 50% RO and 50 and 100% palmoil compared with fish fed 100% FO (14,15).

In salmonid fish, the fatty acid compositions of tissue lipidsare closely related to that of dietary fatty acid compositions(13,33,37), especially in the flesh in which triacylglycerols(TAG) are the predominant lipid class (20). This relationshipis clearly demonstrated in the present study in which linearcorrelations were observed between flesh fatty acid concentra-tions and their concentrations in dietary lipid. Similar linearrelationships were demonstrated in previous trials in whichgraded inclusion of RO or palm oil was used to replace FO(14,15). However, slopes and correlation coefficients differedamong these studies, suggesting that selective retention or

metabolism of individual fatty acids may vary depending onthe blend of dietary fatty acids, and the size and age of the fish.The correlation coefficients in the present study are particu-larly high, with r values between 0.99 and 1.00, which mayreflect the longer trial period and the larger size of fish atsampling, compared with previous studies (14,15). These lin-ear correlations can be used practically to predict the fleshfatty acid concentration in salmon fed different blends of LO,RO and FO in the later stages of the seawater growth phase. Inaddition, the data in Figures 1 and 2, along with data in Tables4 and 5 reveal how different dietary fatty acids are selectivelyretained or metabolized in flesh lipids depending on theirrelative concentration in each oil blend. For example, in alldietary treatments, 22:6(n-3) was selectively deposited in mus-cle lipid regardless of the concentration in the diet. Thispreferential retention of 22:6(n-3), presumably via specificityof the fatty acyl transferases for incorporation into flesh TAGand phospholipid, is clearly demonstrated (Table 5). Thedifferences ( values) between diet and flesh 22:6(n-3) con-centration (Table 5) give values of 1.52.3 for the 100% FO,LO and RO treatments.

In contrast, all other fatty acids described in Table 5,including EPA, were progressively favored for metabolism,presumably largely for energy production, as their dietary con-centration increased. This is true for the monoenes 22:1 and18:1(n-9) which were selectively utilized in the 100% FO and100% RO treatments, respectively, and presumably reflects theease with which these fatty acids can be catabolized by -ox-

TABLE 4

Fatty acid compositions of flesh total lipid in Atlantic salmon fed diets containing various levels of fish oil (FO),linseed oil (LO) and rapeseed oil (RO) for 50 wk1

FAME/diet FOFO/LO(2:1)

FO/LO(1:2) LO

FO/RO(2:1)

FO/RO(1:2) RO

LO/RO(2:1)

LO/RO(1:2)

FO/LO/RO(1:1:1)

g/100 g total fatty acids

14:0 4.6 0.1 3.4 0.1 2.3 0.1 1.2 0.0 3.4 0.0 2.4 0.2 1.3 0.1 1.2 0.1 1.3 0.0 2.3 0.116:0 14.8 0.1 12.9 0.4 11.0 0.1 9.1 0.3 12.5 0.3 10.8 0.1 8.9 0.2 8.9 0.1 9.1 0.1 11.0 0.218:0 2.8 0.1 3.3 0.1 3.4 0.1 3.8 0.1 2.7 0.1 2.6 0.1 2.6 0.0 3.2 0.0 2.9 0.1 3.0 0.1

saturates3 23.1 0.1 20.4 0.6 17.3 0.3 14.6 0.5 19.4 0.4 16.5 0.3 13.6 0.3 13.8 0.1 14.0 0.3 17.0 0.316:1(n-7) 4.8 0.1 3.3 0.1 2.3 0.1 1.2 0.1 3.4 0.1 2.3 0.1 1.3 0.1 1.1 0.0 1.4 0.1 2.2 0.118:1(n-9) 15.2 0.1 15.4 0.2 16.2 0.3 16.7 0.5 24.9 0.3 34.2 0.1 41.5 0.3 25.5 0.3 33.1 0.2 25.1 0.820:1(n-9) 8.1 0.2 5.8 0.1 3.7 0.1 2.0 0.1 6.3 0.1 5.1 0.1 3.9 0.1 2.5 0.1 3.3 0.0 4.5 0.122:14 10.0 0.1 7.2 0.2 4.2 0.2 1.9 0.0 7.1 0.1 4.3 0.1 2.0 0.1 1.9 0.0 2.2 0.0 4.6 0.0

monoenes5 44.0 0.4 36.3 0.3 29.8 0.2 23.7 0.6 47.3 0.1 51.1 0.2 53.6 0.4 34.0 0.4 43.6 0.3 40.8 1.018:2(n-6) 4.4 0.1 7.1 0.1 9.3 0.1 11.7 0.1 8.5 0.2 11.9 0.1 14.6 0.1 13.1 0.1 13.6 0.1 10.7 0.120:2(n-6) 0.5 0.0 0.6 0.0 0.7 0.1 0.7 0.0 0.7 0.0 1.1 0.1 1.3 0.1 0.9 0.1 1.0 0.1 0.8 0.020:3(n-6) 0.1 0.1 0.1 0.0 0.1 0.1 0.1 0.0 0.1 0.0 0.2 0.0 0.4 0.0 0.2 0.0 0.3 0.1 0.2 0.120:4(n-6) 0.5 0.1 0.4 0.0 0.3 0.0 0.1 0.1 0.4 0.0 0.3 0.0 0.2 0.0 0.2 0.1 0.2 0.0 0.3 0.1

(n-6)6 5.9 0.1 8.4 0.2 10.5 0.2 12.6 0.1 10.0 0.3 13.6 0.3 16.7 0.2 14.4 0.1 15.1 0.1 12.0 0.118:3(n-3) 1.4 0.1 15.0 0.4 26.4 0.1 37.4 0.5 3.8 0.1 5.3 0.1 6.8 0.3 26.4 0.3 16.5 0.3 15.6 0.218:4(n-3) 1.7 0.1 1.2 0.1 1.1 0.1 1.0 0.1 1.2 0.1 0.7 0.0 0.6 0.0 1.1 0.1 0.9 0.1 0.9 0.120:3(n-3) 0.2 0.0 1.0 0.0 2.0 0.1 2.8 0.2 0.3 0.0 0.5 0.0 0.6 0.0 2.0 0.1 1.3 0.1 1.2 0.120:4(n-3) 1.6 0.0 1.5 0.1 1.5 0.1 1.6 0.1 1.3 0.1 0.9 0.0 0.7 0.1 1.3 0.1 1.2 0.1 1.2 0.120:5(n-3) 5.6 0.2 3.9 0.1 2.7 0.1 1.6 0.1 4.0 0.1 2.6 0.1 1.5 0.1 1.7 0.1 1.8 0.0 2.5 0.222:5(n-3) 2.5 0.1 1.8 0.1 1.3 0.1 0.6 0.0 1.9 0.1 1.2 0.0 0.6 0.1 0.6 0.1 0.7 0.1 1.1 0.122:6(n-3) 12.8 0.3 9.5 0.5 6.8 0.3 4.2 0.3 10.0 0.1 7.1 0.1 5.0 0.2 4.6 0.5 4.7 0.1 7.0 0.8

(n-3) 25.9 0.4 34.0 0.8 41.8 0.2 49.1 1.0 22.4 0.1 18.3 0.1 15.9 0.6 37.8 0.4 27.0 0.3 29.6 1.2 PUFA 32.9 0.3 43.3 0.9 53.0 0.4 61.7 1.1 33.3 0.3 32.3 0.2 32.8 0.4 52.2 0.3 42.4 0.3 42.2 1.3(n-3)/(n-6) 4.4 0.3 4.1 0.4 4.0 0.2 3.9 0.8 2.2 0.2 1.4 0.2 1.0 0.3 2.6 0.4 1.8 0.3 2.5 0.9

1 Values are means SD for triplicate analyses.2 FAME, fatty acid methyl esters.3 Includes 15:0, 17:0, 20:0 and 22:0.4 Includes 22:1(n-13) and 22:1(n-11).5 Includes 16:1(n-9), 18:1(n-7), 20:1(n-11), 20:1(n-7), 22:1(n-9) and 22:1(n-11).6 Includes 18:3(n-6), 22:4(n-6) and 22:5(n-6).



jn.nutrition.orgD

ownloaded from

idation (11,38). However, when 18:1(n-9) and 22:1 arepresent in the diet in similar quantities, as in the 100% FOdiet, then 22:1 appears to be the preferred substrate for catab-olism. By contrast, when 18:1(n-9) is present in excess as inthe 100% RO diet then 18:1(n-9) appears to be readily oxi-dized. However, when 18:1(n-9) and 18:3(n-3) are present insimilar concentrations, as in the FO/LO 2:1 diet, then the values between diet and flesh suggest that 18:3(n-3) is selec-tively catabolized in preference to18:1(n-9) ( values were6.6 and 0.2, respectively). There was a similar apparentpreference of 18:2(n-6) over 18:1(n-9) in fish fed the 100%

LO diet ( values of 1.9 and 0.1, respectively), whereas therewas also an apparent preference for catabolism of 18:3(n-3)over 18:2(n-6), when present at similar concentrations, in fishfed the LO/RO 1:2 diet ( values of 3.7 and 2.7, respec-tively). These results support those found in a recent studyusing deuterated 18:3(n-3), which showed that catabolism via-oxidation was preferred over conversion to 22:6(n-3) injuvenile rainbow trout (18). The apparent selection againstmetabolism of 18:1(n-9), and also of the saturated fatty acids,principally 16:0 and 18:0, might be due to their structuralfunction as components of membrane phospholipids at the

FIGURE 2 Relationship between dietary fatty acid concentrationand muscle fatty acid concentration for 22:6(n-3) (A),18:1(n-9) (B) and22:1 (C) in total lipids of Atlantic salmon fed diets containing blends offish oil (FO), linseed oil and rapeseed oil. The additional line is the lineof equality. The symbol with an asterisk represents the 100% FOtreatment that had 3 replicates. The SD for the triplicate FO sampleswere 0.12, 0.49 and 0.06 for A, B and C, respectively.

FIGURE 1 Relationship between dietary fatty acid concentrationand muscle fatty acid concentration for 18:2(n-6) (A), 18:3(n-3) (B) and20:5(n-3) (C) in total lipids of Atlantic salmon fed diets containing blendsof fish oil (FO), linseed oil and rapeseed oil. The additional line is the lineof equality. The symbol with an asterisk represents the 100% FOtreatment that had 3 replicates. The SD for the triplicate FO sampleswere 0.23, 0.61 and 0.12 for A, B and C, respectively.

BELL ET AL.2798


jn.nutrition.orgD

ownloaded from

sn-1 position, with HUFA, particularly 22:6(n-3), being fa-vored in the sn-2 position (39,40). These data suggest that,unless 18:1(n-9) is present in excess, 22:1, 18:3(n-3) and18:2(n-6) appear to be the preferred substrates for oxidation.

In addition to being oxidized for energy, both 18:3(n-3) and

18:2(n-6) are substrates for 6-desaturation and elongation. Itwas shown previously that Atlantic salmon hepatocytes favor18:3(n-3) over 18:2(n-6) as a substrate for desaturation andelongation (16); thus, the apparent differences among mobi-lization of flesh 18:3(n-3),18:2(n-6) and 18:1(n-9), describedabove, may be due in part to the favoring of the former forconversion to HUFA as well as for oxidation (16). A numberof previous studies confirmed that the activity of the desatu-ration and elongation pathways is significantly increased afterinclusion of dietary vegetable oils (1417).

There is now considerable evidence that increased con-sumption of (n-3) HUFA, especially 20:5(n-3) and 22:6(n-3)and reduced consumption of (n-6) PUFA, principally 18:2(n-

TABLE 5

Correlation coefficients, slopes and x- and y-axis intercepts from plots of dietary fatty acid concentrations vs. muscle fatty acidconcentrations including the difference () between diet and muscle fatty acid values for salmon fed 100% fish oil (FO),

100% linseed oil (LO) and 100% rapeseed oil (RO)1

Fatty acidCorrelation

coefficient (r) Slopex-Axis

intercepty-Axis

intercept 100% FO2 100% LO2 100% RO2

saturates 0.99 0.74 0.05 7.0 5.2 0.9 1.4 2.218:1(n-9) 0.99 0.81 0.02 3.8 3.1 0.1 0.1 6.818:2(n-6) 0.99 0.76 0.04 1.2 0.9 0.0 1.9 3.318:3(n-3) 1.00 0.81 0.02 0.4 0.3 0.3 8.0 2.122:13 0.99 0.78 0.05 0.4 0.3 3.2 0.4 0.720:5(n-3) 0.99 0.70 0.04 0.3 0.2 1.7 0.2 0.422:6(n-3) 0.99 1.02 0.06 1.4 1.4 2.3 1.5 2.1

1 Fatty acid concentrations are g fatty acid/100 g total fatty acids in muscle and diet.2 Negative values indicate lower values in muscle compared with diet, whereas positive values indicate accumulation in muscle relative to diet.3 Includes 22:1(n-11) and 22:1(n-13).

TABLE 6

Fatty acid compositions of flesh total lipid in Atlantic salmonfed diets containing either 100% fish oil (FO), 100% linseedoil (LO) or 100% rapeseed oil (RO) for 50 wk, followed by a

100% FO diet for 20 wk1

Fatty acid FO LO RO

g/100 g fatty acids

14:0 4.3 0.2a 3.1 0.2b 3.5 0.4b16:0 14.3 0.5a 12.5 0.7b 12.9 0.9b18:0 3.0 0.1b 3.3 0.2a 2.8 0.1c

saturates2 22.4 0.7a 19.8 1.4b 20.0 1.4b16:1(n-7) 4.4 0.2a 3.2 0.4c 3.5 0.4b18:1(n-9) 16.0 0.5b 16.1 0.4b 23.6 2.2a20:1(n-9) 8.2 0.2a 6.1 0.6c 7.0 0.4b22:13 9.5 0.3a 7.0 0.8b 7.4 0.7b24:1 0.9 0.1a 0.7 0.1c 0.8 0.1b

monoenes4 44.0 0.8b 37.2 2.1c 47.3 1.4a18:2(n-6) 4.8 0.2b 7.2 0.7a 7.7 1.0a20:2(n-6) 0.5 0.1a 0.6 0.1b 0.8 0.1b20:3(n-6) 0.2 0.1 0.2 0.0 0.2 0.120:4(n-6) 0.5 0.1a 0.4 0.1b 0.4 0.1b22:5(n-6) 0.2 0.0 0.2 0.0 0.2 0.0

(n-6)5 6.3 0.2c 8.6 0.7b 9.5 1.0a18:3(n-3) 2.0 0.3c 13.3 3.3a 3.2 0.5b20:3(n-3) 0.3 0.0b 1.1 0.3a 0.4 0.1b20:4(n-3) 1.8 0.1a 1.8 0.2a 1.5 0.1b20:5(n-3) 5.2 0.2a 4.1 0.4b 4.1 0.4b22:5(n-3) 2.5 0.1a 1.9 0.2b 1.9 0.2b22:6(n-3) 12.9 0.8a 10.0 0.8b 10.3 1.1b

(n-3)6 26.5 1.0b 33.8 2.6a 22.7 1.7c PUFA 33.5 1.1b 43.0 3.2a 32.7 1.8b

1 Values are means SD, n 15 individual fish. Values in a rowwithout a common superscript letter differ, P 0.05.

2 Includes 15:0, 17:0, 20:0 and 22:0.3 Includes 22:1(n-11) and 22:1(n-13).4 Includes 16:1(n-9), 20:1(n-11), 20:1(n-7) and 22:1(n-9).5 Includes 18:3(n-6) and 22:4(n-6).6 Includes 18:4(n-3) and 22:4(n-3).

FIGURE 3 Linoleic [18:2(n-6)] (A) and linolenic acid [18:3(n-3)] (B)contents of total lipid from salmon flesh after feeding diets containing100% fish oil (FO), 100% linseed oil (LO) or 100% rapeseed oil (RO) for50 wk and after 20 wk of feeding a 100% FO diet. Means at a timewithout a common letter differ, P 0.05.



jn.nutrition.orgD

ownloaded from

6), can have significant benefits for human health (3,41,42).Increased consumption of oily fish including herring, sardines,mackerel and salmon can provide the necessary (n-3) HUFAin human diets. However, because global capture fisheries havereached their sustainable limit, further demand for fish forhuman consumption will have to be provided by increasedaquaculture production (6). At the present time, salmon pro-duced by aquaculture, using diets containing only marine FO,is rich in (n-3) HUFA and is a valuable component of humandiets (20,21,43). Therefore, any substitution of FO with veg-etable oils should be carried out in such a way as to ensure thatthe qualities that make cultured salmon a healthy food option,i.e., high levels of 20:5(n-3) and 22:6(n-3), are maintained.The present study suggests that substitution of FO with ROand/or LO, up to 66% of total added oil, results in a moderatereduction of (n-3) HUFA but also leads to increased deposi-tion of 18:2(n-6). The second phase of the experiment in-volved returning fish, previously fed 100% LO and RO to dietscontaining only FO for a period of 20 wk, in an attempt towash out the C18 PUFA from the vegetable oils. This actionrestored (n-3) HUFA levels to 80% and lowered 18:2(n-6)levels, although they later remained 50% higher than thosepresent in fish fed 100% FO. These data are similar to aprevious study using smaller salmon smolts in which dietscontaining increasing levels of RO were fed for 16 wk, fol-lowed by a 12-wk washout phase (44). In both studies, thewashout phase restored DHA and EPA levels to values similarto those in fish fed FO throughout, yet 18:2(n-6) valuesremained significantly elevated (44). These results support therationale that vegetable oils used for FO replacement insalmon should ideally contain minimal levels of 18:2(n-6). Ifsalmon is to be regarded as a valuable source of (n-3) HUFAfor humans, then it should also ideally contain low levels of18:2(n-6). It is widely known that the dietary intake of

18:2(n-6) in the Western diet is too high and that a high ratioof (n-6)/(n-3) PUFA is proinflammatory (3,4,45).

In summary, the results from the present study suggest thatAtlantic salmon can be raised on diets in which FO is replacedwith different blends of LO, RO and FO for the entire seawaterculture phase, without apparent detriment to fish growth andhealth. However, for fish in which vegetable oil replaced66% of the added dietary oil, considerable reductions of fleshconcentrations of both 20:5(n-3) and 22:6(n-3) occurred. Re-turning fish previously fed 100% RO or LO to a marine FOdiet for a period before harvest allowed flesh (n-3) HUFAconcentrations to be restored to 80% of values in fish fed FOthroughout the seawater phase although 18:2(n-6) remainedsignificantly higher.

LITERATURE CITED

1. Dyerberg, J., Bang, H. O. & Hjrne, N. (1975) Fatty acid compositionof plasma lipids in Greenland Eskimos. Am. J. Clin. Nutr. 28: 958966.

2. Bang, H. O. & Dyerberg, J. (1980) Lipid metabolism and ischaemicheart disease in Greenland Eskimos. In: Advances in Nutrition Research, Vol. 3.(Draper, H. H., ed.), pp. 122. Plenum Press, New York, NY.

3. Simopoulos, A. P. (1999) Essential fatty acids in health and chronicdisease. Am. J. Nutr. 70: 560S569S.

4. Connor, W. E. (2000) Importance of n-3 fatty acids in health anddisease. Am. J. Clin. Nutr. 71: 171S175S.

5. Williams, N. (1998) Overfishing disrupts entire ecosystems. Science(Washington, DC) 279: 809810.

6. Sargent, J. R. & Tacon, A.G.J. (1999) Development of farmed fish: anutritionally necessary alternative to meat. Proc. Nutr. Soc. 58: 377383.

7. Barlow, S. (2000) Fishmeal and fish oil: sustainable ingredients foraquafeeds. Glob. Aquacult. Advocate 4: 8588.

8. Fryland, L., Madsen, L., Eckhoff, K. M., Lie, . & Berge, R. (1998)Carnitine palmitoyltransferase I, carnitine palmitoyltransferase II, and acyl-CoAoxidase activities in Atlantic salmon (Salmo salar) Lipids 33: 923930.

9. Sargent, J. R., Bell, J. G., Bell, M. V., Henderson, R. J. & Tocher, D. R.(1995) Requirement criteria for essential fatty acids. J. Appl. Ichthyol. 11: 183198.

10. Sargent, J. R., Tocher, D. R. & Bell, J. G. (2002) The Lipids In: FishNutrition, 3rd ed. (Halver, J. E. & Hardy, R. W., eds.) pp. 181257. AcademicPress, New York, NY.

11. Henderson, R. J. & Sargent, J. R. (1985) Chain length specificities ofmitochondrial and peroxisomal -oxidation of fatty acids in rainbow trout (Salmogairdneri). Comp. Biochem. Physiol. 82B: 7985.

12. Henderson, R. J. (1996) Fatty acid metabolism in freshwater fish withparticular reference to polyunsaturated fatty acids. Arch. Anim. Nutr. 49: 522.

13. Torstensen, B. E., Lie, O. & Froyland, L. (2000) Lipid metabolism andtissue composition in Atlantic salmon (Salmo salar L.)effects of capelin oil, palmoil and oleic acidenriched sunflower oil as dietary lipid sources. Lipids 35:653664.

14. Bell, J. G., McEvoy, J., Tocher, D. R., McGhee, F., Campbell, P. J. &Sargent, J. R. (2001) Replacement of fish oil with rapeseed oil in diets ofAtlantic salmon (Salmo salar) affects tissue lipid compositions and hepatocytefatty acid metabolism. J. Nutr. 131: 15351543.

15. Bell, J. G., Henderson, R. J., Tocher, D. R., McGhee, F., Dick, J. R.,Porter, A., Smullen, R. P. & Sargent, J. R. (2002) Substituting fish oil with crudepalm oil in the diet of Atlantic salmon (Salmo salar) affects muscle fatty acidcomposition and hepatic fatty acid metabolism. J. Nutr. 132: 222230.

16. Bell, J. G., Tocher, D. R., Farndale, B. M., Cox, D. I., McKinney, R. W. &Sargent, J. R. (1997) The effect of dietary lipid on polyunsaturated fatty acidmetabolism in Atlantic salmon (Salmo salar) undergoing parr-smolt transforma-tion. Lipids 32: 515525.

17. Tocher, D. R., Bell, J. G., Dick, J. R., Henderson, R. J., McGhee, F.,Mitchell, D. F. & Morris, P. C. (2000) Polyunsaturated fatty acid metabolism inAtlantic salmon (Salmo salar) undergoing parr-smolt transformation and the ef-fects of dietary linseed and rapeseed oils. Fish Physiol. Biochem. 23: 5973.

18. Bell, M. V., Dick, J. R. & Porter, A. E. A. (2001) Biosynthesis and tissuedeposition of docosahexaenoic acid (22:6n-3) in rainbow trout (Oncorhynchusmykiss). Lipids 36: 11531159.

19. Bourre, J. M., Piciotti, M. & Dumont, O. (1990) 6-Desaturase in brainand liver during development and ageing. Lipids 25: 354356.

20. Bell, J. G., McEvoy, J., Webster, J. L., McGhee, F., Millar, R. M. &Sargent, J. R. (1998) Flesh lipid and carotenoid composition of Scottishfarmed Atlantic salmon (Salmo salar). J. Agric. Food Chem. 46: 119127.

21. Aursand, M., Mabon, F. & Martin, G. J. (2000) Characterization offarmed and wild salmon (Salmo salar) by a combined use of compositional andisotopic analyses. J. Am. Oil Chem. Soc. 77: 659666.

22. Refsgaard, H. H. F., Brockhoff, P. B. & Jensen, B. (1998) Biologicalvariation of lipid constituents and distribution of tocopherols and astaxanthin infarmed Atlantic salmon (Salmo salar). J. Agric. Food Chem. 46: 808812.

23. Ackman, R. G. (1990) Canola fatty acidsan ideal mixture for health,

FIGURE 4 Eicosapentaenoic [20:5(n-3)] (A) and docosahexaenoicacid [22:6(n-3)] (B) contents of total lipid from salmon flesh after feedingdiets containing 100% fish oil (FO), 100% linseed oil (LO) or 100%rapeseed oil (RO) for 50 wk and after 20 wk of feeding a 100% FO washout diet. Means at a time without a common letter differ, P 0.05.

BELL ET AL.2800


jn.nutrition.orgD

ownloaded from

nutrition and food use. In: Canola and Rapeseed (Shahidi, F., ed.), pp. 8198. Avi,New York, NY.

24. Garg, M. L., Thomson, A.B.R. & Clandinin, M. T. (1990) Interactions ofsaturated, n-6 and n-3 polyunsaturated fatty acids to modulate arachidonic acidmetabolism. J. Lipid Res. 31: 271277.

25. Huang, Y. S., Smith, R. S., Redden, P. R., Cantrill, R. C. & Horrobin, D. F.(1991) Modification of liver fatty acid metabolism in mice by n-3 and n-66-desaturase substrates and products. Biochim. Biophys. Acta 1082: 319327.

26. National Research Council (1993) Nutrient Requirements of Fish. Na-tional Academy Press, Washington, DC.

27. Folch, J., Lees, M. & Sloane-Stanley, G. H. (1957) A simple method forthe isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226:497509.

28. Christie, W. W. (1982) Lipid Analyses, 2nd ed., pp. 5256. PergamonPress, Oxford, UK.

29. Ghioni, C., Bell, J. G. & Sargent, J. R. (1996) Polyunsaturated fattyacids in neutral lipids and phospholipids of some freshwater insects. Comp.Biochem. Physiol. 114B: 161170.

30. Ackman, R. G. (1980) Fish lipids, part 1. In: Advances in Fish Sciencesand Technology (Connell, J. J., ed.), pp. 86103. Fishing News Books, Farn-ham, UK.

31. De Lorgeril, M., Renaud, S. & Delaye, J. (1994) Mediterranean alpha-linolenic acid-rich diet in the secondary prevention of coronary heart disease.Lancet 343: 14541459.

32. Dosanjh, B. S., Higgs, D. A., Plotnikoff, M. D., Markert, J. R. & Buckley,J. T. (1988) Preliminary evaluation of canola oil, pork lard and marine lipidsingly and in combination as supplemental dietary lipid sources for juvenile fallchinook salmon (Oncorhynchus tshawytscha). Aquaculture 68: 325343.

33. Guillou, A., Soucy, P., Khalil, M. & Adambounou, L. (1995) Effects ofdietary vegetable and marine lipid on growth, muscle fatty acid composition andorganoleptic quality of flesh of brook charr (Salvelinus fontinalis) Aquaculture 136:351362.

34. Polvi, S. M. & Ackman, R. G. (1992) Atlantic salmon (Salmo salar)

muscle lipids and their response to alternative fatty acids. J. Agric. Food Chem.40: 10011007.

35. Sowizral, K. C., Rumsey, G. L. & Kinsella, J. E. (1990) Effects of dietary-linolenic acid on n-3 fatty acids of rainbow trout. Lipids 25: 246253.

36. Yang, X. & Dick, T. A. (1994) Dietary -linolenic and linoleic acidscompetitively affect metabolism of polyunsaturated fatty acids in Arctic charr(Salvelinus alpinus). J. Nutr. 124: 11331145.

37. Lie, ., Waagb, R. & Sandnes, K. (1988) Growth and chemicalcomposition of adult Atlantic salmon (Salmo salar) fed dry silage based diets.Aquaculture 69: 343353.

38. Kiessling, K-H. & Keissling, A. (1993) Selective utilisation of fatty acidsin rainbow trout (Oncorhynchus mykiss Walbaum) red muscle mitochondria. Can.J. Zool. 71: 248251.

39. Bell, M. V. & Dick, J. R. (1991) Molecular species composition ofphosphatidylinositol from the brain, retina, liver and muscle of cod (Gadusmorhua). Lipids 25: 691694.

40. Bell, M. V. & Dick, J. R. (1991) Molecular species composition of themajor diacyl glycerophospholipids from muscle, liver, retina and brain of cod(Gadus morhua). Lipids 26: 565573.

41. De Deckere, E.A.M., Korver, O., Verschuren, P. M. & Katan, M. B. (1998)Health aspects of fish and n-3 polyunsaturated fatty acids from plant and marineorigin. Eur. J. Clin. Nutr. 52: 749753.

42. Horrocks, L. A. & Yeo, Y. K. (1999) Health benefits of docosahexae-noic acid (DHA). Pharmacol. Res. 40: 211225.

43. Schmidt, E. B., Christensen, J. H., Aardestrup, I., Madsen, T., Riahi, S.,Hansen, V. E. & Skou, H. A. (2001) Marine n-3 fatty acids: basic features andbackground. Lipids 36: S65S68.

44. Bell, J. G., McGhee, F., Campbell, P. J. & Sargent, J. R. (2003) Rape-seed oil as an alternative to marine fish oil in diets of post-smolt Atlantic salmon(Salmo salar): changes in flesh fatty acid composition and effectiveness of sub-sequent fish oil wash out. Aquaculture 218: 515528.

45. Ascherio, A., Rimm, E. B., Giovannucci, E. L., Spiegelman, D., Stampfer,M. J. & Willet, W. C. (1996) Dietary fat and risk of coronary heart disease inmen: cohort follow up study in the United States. Br. Med. J. 313: 8490.



jn.nutrition.orgD

ownloaded from

bell et al, 2003

Documents