Effect of dietary canola oil level on the growth performance and fattyacid composition of juvenile red sea bream, Pagrus major
S.S.Y. Huang a,c, A.N. Oo b, D.A. Higgs c, C.J. Brauner a, S. Satoh b,⁎
a Department of Zoology, University of British Columbia, Vancouver, BC, Canada V6T 1Z4b Department of Marine Biosciences, Tokyo University of Marine Science and Technology, 4-5-7 Konan, Minato-City, Tokyo 108-8477, Japan
c Department of Fisheries and Oceans/University of British Columbia, Centre for Aquaculture and Environmental Research,4160 Marine Drive, West Vancouver, BC, Canada V7V 1N6
Received 10 April 2007; received in revised form 5 June 2007; accepted 5 June 2007
Fish oil (FO) is an invaluable dietary component forfish because it furnishes the essential fatty acids (EFA)such as eicosapentaenoic acid (EPA), docosahexaenoic
acid (DHA) and arachidonic acid (AA) which are neededfor optimal growth and development. The constituentfatty acids in FO also influence biological membranestructure and function. Moreover, FO is an importantsource of lipid soluble vitamins and highly digestibleenergy (Higgs and Dong, 2000). Indeed, marine fishingest little carbohydrate as part of their natural diet.Consequently, they are incapable of metabolically
421S.S.Y. Huang et al. / Aquaculture 271 (2007) 420–431
utilizing high dietary levels of digestible carbohydrate.Thus, lipids are the favoured form of non-protein metab-olic energy (Sargent et al., 2002).
Fish species, especially those of marine origin, pro-vide excellent dietary sources of high quality protein,vitamins and minerals. Fish, in particular, contain arelatively high proportion of n-3 highly unsaturated fattyacids (HUFA), which have a wide range of health ben-efits in humans such as prevention of cardiovasculardisease, Alzheimer's disease and abnormal neurologicaland ocular development (Mozaffarian and Rimm, 2006;Schaefer et al., 2006; Ikonomou et al., 2007). Conse-quently, global consumer demands for food fish havegrown tremendously and aquaculture is playing anincreasingly important role to maintain the percent percapita consumption of fish (FAO-FD, 2004; Tacon,2004). At the same time, global demand for FO foraquafeeds has grown in direct proportion to the increasein aquaculture output. This is particularly true for car-nivorous species such as salmon which require highenergy (lipid) content diets (lipid≤40% in grower diets)for optimal growth, feed efficiency and minimization ofsea water culture time (Steffens, 1993; Hardy et al.,2001). However, the global supply of FO is limited andthe prices of FO have been forecasted to rise (Hardyet al., 2001; New and Wijkström, 2002). Plant oil pricestoday are generally less than those of FO and for reasonsrelated to sustainability, it is not possible to furtherincrease the annual global harvest of pelagic fish stocksand consequently the supply of FO (Barlow, 2000).
Plant oils are rich in C18 fatty acids but unlike FO, then-3 HUFA (EPA and DHA) are absent. Nevertheless,plant oils are rich in unsaturated fatty acids and thereforeare highly digestible (Opsahl-Ferstad et al., 2003). Fish,like other vertebrates, lack the enzymes necessary tosynthesize the parent acids of the n-3 and n-6 families offatty acids namely, linolenic acid (18:3n-3; LNA) andlinoleic acid (18:2n-6; LA), respectively (Bell et al.,1986; Tocher et al., 1989; Mourente and Tocher, 1993a,b). These parent acids are abundant in plant oils. How-ever, marine fish species have dietary requirements forboth EPA and DHA since they have limited capacity tosynthesize these compounds from LNA (Ghioni et al.,1999; Tocher and Ghioni, 1999). Some marine speciesappear to have a dietary requirement for AA and theremay be a need for an optimal balance between EPA,DHA and AA for good growth and normal development(Higgs and Dong, 2000). In Atlantic salmon (Salmosalar), alterations between the dietary proportions ofsaturated fatty acids (SFA), n-3 and n-6 polyunsaturatedfatty acids (PUFA) can affect aerobic swimming per-formance and gill function (McKenzie et al., 1998;
Wagner et al., 2004). A study on the European seabass (Dicentrarchus labrax), a marine species, has alsoshown that dietary inclusion of plant oil (canola oil orpalm oil) can positively influence cardio-respiratory andswimming performance (Chatelier et al., 2006).
Recent studies aimed at assessing the merits of usingplant oils as partial replacements for FO in diets formarinefish have demonstrated promising results with respectto gilthead sea bream (Sparus aurata; Montero et al.,2003; Izquierdo et al., 2003, 2005; Menoyo et al., 2004),European sea bass (Izquierdo et al., 2003;Mourente et al.,2005) and turbot (Psetta maxima; Regost et al., 2003).Indeed, short and long-term studies have shown thatrapeseed oil can replace up to 60% of the supplementalFO in diets for gilthead sea bream and European sea basswithout adverse effects on fish growth, feed efficiencyand survival (Izquierdo et al., 2003, 2005;Mourente et al.,2005). More recently, Glencross et al. (2003a) demon-strated that refined canola oil could comprise 40% of thetotal dietary lipid of juvenile Australian pink snapper(Pagrus auratus) without compromising their growthover a 54-day period. Also, in a second study, Glencrosset al. (2003b) found that canola oil could comprise upto 68% of the dietary lipid for this species without anyadverse effects on fish growth over a 32-day period.Nonetheless, plant oils referred to as canola oil (CO), apartfrom the aforementioned studies of Glencross et al.(2003a,b) and Chatelier et al. (2006) on European seabass, have received little attention in marine fish diets.
Canola is the trademark name given to rapeseed cul-tivars that have been selected genetically to contain lowerucic acid content in the oil (b2%) and glucosinolates orantithyroid compounds in the meal (now b30 μmol/g ofair dry oil-free meal; Canola Council of Canada, 2004).Like rapeseed oil, CO has an excellent balance between18:1n-9 (oleic acid), 18:2n-6 and 18:3n-3. Canola is thedominant oilseed crop of Canada, and the oil, alongwith low erucic acid rapeseed oil, has been shown to bean excellent source of supplemental dietary lipid forfinfish, provided that their respective EFA needs havebeen met (Dosanjh et al., 1984, 1988, 1998; Bell et al.,2001, 2003; Rosenlund et al., 2001; Grant, 2006; Higgset al., 2006).
The red sea bream (Pagrus major), a strictly carniv-orous marine fish, is a major finfish species cultured inJapan. In fact, the demand for red sea bream has growntremendously within the last decade primarily because itis a high-quality sashimi grade fish with high marketvalue (Watanabe and Vassallo-Agius, 2003). Whileextensive research has been done to define the basicnutritional and husbandry requirements for this species,the merits of including alternative protein and lipid
Table 1Ingredient compositions of the experimental diets
Numerical values after CO refer to the percentage of CO expressed inrelation to the total dietary lipid content.a Supplemented with 200 mg/kg BHT.b Supplied by Hayashichemical Co. Ltd., Tokyo, Japan with no
anti-oxidant added.c Mineral supplement supplied (mg/kg diet): Na (as NaCl) 197; Mg
(as MgSO4 ·7H2O) 735; Fe (as FeC6H5O7·5H2O) 258; Zn (asZnSO4·7H2O) 40; Mn (as MnSO4 ·5H2O) 18; Cu (as CuSO4 ·5H2O)3.9; Al (as AlCl3 ·6H2O) 0.56; Co (as CoCl2 ·6H2O) 0.15; I (as KIO3)0.89; α-cellulose carrier.d Vitamin supplement supplied (amount/kg diet): thiamin hydro-
422 S.S.Y. Huang et al. / Aquaculture 271 (2007) 420–431
sources to fish meal and FO have only been exploredrecently (Watanabe, 2002). Since Japan is the largestimporter of Canadian canola seed (Canola Council ofCanada, 2005), there is a strong incentive to consider thissource of CO as a potentially suitable alternative dietarylipid source for this species, especially given the successof using low erucic acid type rapeseed oil in the diets ofother marine species and of canola oil on P. auratus, aspreviously mentioned.
Based on taxonomical analyses, Paulin (1990) pro-posed that P. major and P. auratus are independent andreproductively isolated populations of the same speciesfound in Japan and Australiasia. However, P. majorhas been traditionally treated as an independent speciesin Japan despite Paulin's recommendations (Nakabo,1993). Tabata and Taniguchi (2000) partly supportPaulin's finding in that P. major and P. auratus may notdiffer enough to be classified as different species. How-ever, significant differences at the mitochondrial DNAcontrol region and morphology of the head bump sug-gest that the relationship of the two populations is at thesubspecies level (Tabata and Taniguchi, 2000). Further-more, the same researchers did not find any data thatshowed interbreeding between the populations.
Therefore, the present study was undertaken to eval-uate the efficacy of using refined CO as a partial or totalsubstitute for the supplemental FO in a practical diet forjuvenile Japanese red sea bream fingerlings that were in arapid growth phase over a 12-week (84-day) period. Theassessment criteria included not only the possible dietaryCO concentration effects on fish growth performanceand survival, but also on whole body proximate and lipidcomposition as well as liver polar lipid composition.
2. Materials and methods
2.1. Experimental diets
Four diets of equivalent crude protein (∼46%), ener-gy (∼21.9 MJ/kg) and lipid (∼15%) concentration on adry weight basis were formulated. The diets hadidentical ingredient compositions, except CO replacedeither 0%, 33%, 67%, or 100% of the supplementaldietary lipid content (100 g/kg diet) with the remainderoriginating from FO (pollock liver oil). Thus, CO com-prised either 0% (diet FO), 25% (CO25), 48% (CO48),or 70% (CO70) of total dietary lipid content (Table 1).The diets were cold-pelleted with a laboratory pellet mill(AEZ12 M, Hiraga-Seikakusho, Kobe, Japan). There-after, the pellets (2.5 mm) were dried using a vacuumfreeze-drier (RLE-206, Kyowa Vacuum Engineering,Tokyo, Japan), and stored at 4 °C until use.
2.2. Fish maintenance and experimental design
Red sea bream (Pagrus major) fingerlings were ob-tained from Seiho Suisan Co. Ltd. (Mie, Japan) and werefed commercial larval feed (Amblose, Nippon FeedMfg.Co. Ltd., Tokyo, Japan) for 3 weeks while they wereacclimated to the culture conditions at the Laboratory ofFish Nutrition located at the Tokyo University of MarineScience and Technology, Tokyo, Japan. Following this,25 fish (average weight 3.61±0.12 g (1SD)) weredistributed randomly into 60-L glass tanks that were eachsupplied with 700–800 ml/min of aerated, re-circulated,25±1.2 °C artificial sea water (Sea Life®, Tokyo, Japan;salinity, 30 g/L). Moreover, a 12-h light/12-h dark photo-period regime was in effect during the study. Half of thewater was renewed weekly to maintain the acceptablewater quality limits for the preceding parameters andwater quality was monitored daily. Triplicate groups offish were each fed one of the four aforementioned dietsby hand to apparent satiation three times daily (0900 h,1200 h and 1600 h) for 84 days and the diet treatmentswere assigned using a randomized block design.
Table 2Mean concentrations (% dry weight basis except for % moisture±1SD) of proximate constituents and gross energy (MJ/kg dry weight)in the experimental diets
The supplemental lipid stemmed from either pollock liver oil (FO),different blends of canola oil (CO) with FO, or CO. Numerical valuesafter CO refer to the percentage of CO expressed in relation to the totaldietary lipid content.
423S.S.Y. Huang et al. / Aquaculture 271 (2007) 420–431
2.3. Fish weighing and sampling
All fish in each group were anaesthetized (2-phenoxyethanol at 0.5 ml/L) and then weighed individ-
Table 3Percent fatty acid contents (g/100 g fatty acids) in pollock liver oil (FO) and
Numerical values after CO refer to the percentage of CO expressed in relati
ually, after removal of excess surface moisture, to thenearest 0.01 g at 21-day intervals during the study. Onday 0, 15 fish from a common pool of fish were sam-pled randomly and stored at −30 °C for subsequentdeterminations of their initial proximate and lipidcompositions (the analyses were conducted on 3composite samples of 5 fish each; n=3). On day 84,3 fish were sampled randomly from each replicategroup (tank) per diet treatment for subsequent deter-minations of whole body proximate and lipid composi-tions and 5 fish were sampled randomly from eachgroup for assessment of liver polar lipid composition.Whole body samples were ground to a homogeneousconsistency using a centrifugal mill (Retsch ZM 100,Haan, Germany) fitted with a 0.25 mm screen. Thehomogenate from each replicate tank was pooled(n=3/diet treatment) and stored at −30 °C undernitrogen pending analysis. Liver samples were groundby hand and analyzed immediately following terminalsampling.
Table 4Mean (±1SD) initial (IBW, g) and final (FBW, g) body weight, weightgain (WG, g), specific growth rate (SGR, %/day), dry feed intake (DFI,g/fish), feed efficiency (FE, g/g), protein efficiency ratio (PER, g/g),percent protein deposited (PPD, %), percent lipid deposited (PLD, %),gross energy utilization (GEU, %), and survival (S, %) of red seabream in relation to diet treatment
Data for each parameter (n=3) were analyzed by randomized blockANOVA. No significant differences were found for any of theperformance parameters due to diet treatment. The supplementaldietary lipid stemmed from either pollock liver oil (FO), differentblends of canola oil (CO) with FO, or CO. Numerical values after COrefer to the percentage of CO expressed in relation to the total dietarylipid content.
424 S.S.Y. Huang et al. / Aquaculture 271 (2007) 420–431
2.4. Chemical analyses
2.4.1. Moisture, ash, protein and gross energydeterminations
Determinations of moisture, ash, protein, and grossenergy concentrations in the diets and fish samples wereconducted as described below. Percent moisture wasmeasured by oven drying each sample at 110 °C for 4-hand then weighing each sample at hourly intervals untilconstant weight was obtained. Ash content was deter-mined by ashing each dried sample in a porcelain crucibleusing a muffle furnace at 600 °C overnight (Woyewodaet al., 1986). Crude protein concentration was determined
Table 5Initial (n=3) and final mean (n=3) concentrations (% of wet weight±1SD) obodies of red sea bream in relation to diet treatment
Percentages for each proximate constituent were arcsine square root transfappropriate, differences among treatment means were detected using Tukesignificant differences among the means (Pb0.05). The supplemental dietarcanola oil (CO) with FO, or CO. Numerical values after CO refer to the per
by the Kjeldahl procedure using a Kjeltec Auto SamplerSystem 1030 Analyzer (Foss Ltd., Tokyo, Japan). Percentnitrogen was multiplied by 6.25 to obtain an estimate ofpercent protein. Dietary gross energy concentrations weredetermined by bomb calorimetry (IKA CalorimeterSystem C5001 duo control, IKA® Werke GmbH & Co.KG, Staufen, Germany) whereas whole body energyconcentrations were estimated by ascribing 23.64 kJ/gand 39.54 kJ/g for protein and lipid, respectively.
2.4.2. Lipid extraction and fatty acid analysisTotal lipids were extracted from homogenized (Nissei
250 Ace Homogenizer, Nihonseiki Kaisha Ltd., Tokyo,Japan) whole bodies and livers using chloroform/metha-nol (2:1, v/v) according to the methods of Folch et al.(1957) with minor revisions. Total lipids extracted fromthe livers were separated into neutral and polar lipidfractions via silica gel cartridges (Sep-Pack, WatersCo., Millford, U.S.A.), as described by Juaneda andRocquelin (1985). The total lipids from the diets andwhole bodies and the polar lipids from the livers weretrans-methylated (Christie, 1973, modified), and then thefatty acid methyl esters (FAME) were separated andquantified by a gas chromatograph (GC14B; ShimadzuCo., Tokyo, Japan) equipped with a Supercowax-10fused silica wall coated 30 m×0.32mm×0.22 μm capil-lary column (Supleco Inc., Pennsylvania, U.S.A.) and ahydrogen flame ionization detector. FAME were elutedfrom the column using helium as the carrier gas. Theinitial and final temperatures of the column were 170 °Cand 250 °C, respectively. The gradient increment was setat 2 °C per minute. The injector and detector tempera-tures were set at 250 °C. The individual fatty acids wereidentified with known standards. Subsequently, individ-ual FAME concentrations were expressed as a percent-age of the total of the identifiable fatty acids (N95%).Area percentage normalized values for the fatty acidswere considered to be equivalent to weight percentage
f proximate constituents and gross energy contents (kJ/g) in the whole
ormed before being subjected to randomized block ANOVA. Wherey's test with P=0.05. Different superscripts within a column denotey lipid stemmed from either pollock liver oil (FO), different blends ofcentage of CO expressed in relation to the total dietary lipid content.
425S.S.Y. Huang et al. / Aquaculture 271 (2007) 420–431
values since there were insignificant amounts of fattyacids with less than 12 carbon atoms (AOAC, 2000).
2.5. Data calculation and statistical analyses
The effect of diet treatment on the growth perfor-mance of the fish was assessed by the following:
(1) Wet weight gain (WG) (g)=(final mean wet weight(FW) (g)− initial mean wet weight (IW) (g))
(2) Specific growth rate (SGR) (% body weight/day)=[(ln FW (g)− ln IW (g))/time (days)]×100
(3) Dry feed intake (DFI) (g/fish)= total daily dry feedintake/fish over 84 days
(4) Feed efficiency (FE) (g/g)=WG (g)/DFI (g/fish)(5) Protein efficiency ratio (PER) (g/g)=WG (g)/pro-
tein intake (g)
Table 6Final mean (±1SD) fatty acid contents (g/100 g total fatty acids) in the who
DietFatty acid Initial
FO
14:0 4.17 3.04±0.83a
16:0 23.4 17.5±2.45a
18:0 5.60 3.81±0.1020:0 0.14 0.14±0.02d
ΣSFA 33.30 24.5±3.32a
16:1n-7 6.00 6.00±0.96a
18:1n-(9+7) 23.5 22.3±1.06d
20:1n-(11+9) 2.45 8.88±1.14a
22:1n-(11+13+9) 0.96 8.59±1.92a
ΣMFA 32.9 45.8±1.17d
18:2n-6 3.83 4.24±0.08d
20:2n-6 0.18 0.20±0.03b
20:3n-6 0.07 0.06±0.0020:4n-6 0.80 0.50±0.04a
22:4n-6 0.24 0.19±0.05a
22:5n-6 0.29 0.18±0.02a
Σn-6 PUFA 5.41 5.36±0.21d
18:3n-3 0.56 0.62±0.02d
18:4n-3 0.49 0.89±0.03a
20:4n-3 0.04 0.62±0.05a
20:5n-3 5.49 5.56±0.53a
22:5n-3 1.28 1.71±0.21a
22:6n-3 15.8 11.4±1.42a
Σn-3 HUFA 21.3 17.0±2.01a
Σn-3 PUFA 23.7 20.8±2.24a
22:4n-9 0.16 0.25±0.02a
Others 8.55 3.65±0.05a
Total PUFA 29.3 26.4±2.43n-3/n-6 4.38 3.87±0.29a
The supplemental dietary lipid stemmed from either pollock liver oil (FO), difCO refer to the percentage of CO expressed in relation to the total dietary liPercentage data of each fatty acid (n=3/diet treatment with each mean basedanalyzed by randomized block ANOVA and, where appropriate, Tukey's tesdifferences among the means (Pb0.05).
(6) Percent protein deposited (PPD) (%)=proteingain (g)×100/protein intake (g)
(8) Gross energy utilization (GEU) (%)=gross energygain (MJ/fish)×100/gross energy intake (MJ/fish)
(9) Survival (S) (%)=(number of fish in each groupremaining on day 84/initial number of fish)×100
All data were subjected to randomized block An-alyses of Variance (ANOVA; SigmaStat 3.0, SPSS,Chicago, U.S.A.) to test for possible diet and blockeffects. Arcsine square root transformations of percentagedata were conducted to achieve homogeneity of variancebefore statistical analysis. Tukey's test with P=0.05 wasused to detect significant differences among means whereappropriate. Graphical relationships between selected
le bodies of red sea bream in relation to diet treatment
ferent blends of canola oil (CO) with FO, or CO. Numerical values afterpid content.on the analysis of 3 fish) were arcsine square root transformed and thent with P=0.05. Different superscripts within a row denote significant
Table 7Final mean (±1SD) fatty acid contents (g/100 g total fatty acids) in theliver polar lipids of red sea bream in relation to diet treatment
Percentage data of each fatty acid (n=3/diet treatment with each meanbased on the analysis of 5 fish) were arcsine square root transformed andthen analyzed by randomized block ANOVA and, where appropriate,Tukey's test with P=0.05. Different superscripts within a row denotesignificant differences among the means (Pb0.05). The supplementaldietary lipid stemmed from either pollock liver oil (FO), different blendsof canola oil (CO)with FO, or CO.Numerical values after CO refer to thepercentage of CO expressed in relation to the total dietary lipid content.
426 S.S.Y. Huang et al. / Aquaculture 271 (2007) 420–431
dietary fatty acid concentrations and their respective fattyacid concentrations in fish lipids were examined accord-ing to Bell et al. (2002) to gain further insights into themetabolic fates of these fatty acids.
3. Results
All test diets contained similar concentrations of theproximate constituents that were in each case similar tothe expected values (Tables 1 and 2). Dietary fatty acidcompositions reflected the concentrations of FO and COin the supplemental lipids and their respective fatty acidcompositions (Table 3). Thus, dietary CO concentra-tions were directly related to percentages of 20:0, 18:1n-(9+7), Σmonounsaturated fatty acids (MFA), 18:2n-6, Σn-6 PUFA, and 18:3n-3, and inversely related to14:0, 16:0, 18:0, Σsaturated fatty acids (SFA), 16:1n-7,20:1n-(11+9), 22:1n-(11+13+9), 20:2n-6, 20:4n-6,22:4n-6, 22:5n-6, 18:4n-3, 20:4n-3, 20:5n-3, 22:5n-3,22:6n-3,Σn-3 HUFA,Σn-3PUFA and ratio of n-3 to n-6fatty acids (Table 3).
Diet treatment did not influence growth performanceof the red sea bream in this study. For instance, all groupsirrespective of diet fed had similar initial mean weightand values for FW,WG, SGR, DFI, FE, PER, PPD, PLD,andGEU and%S at the end of the 84-day feeding period.Furthermore, it is noteworthy that the final mean weightsof all groups were about 13 fold higher than their re-spective mean initial weights (Table 4). In addition,except for percentages of whole body moisture, whichwere higher in FO andCO25-fed fish relative to those feddiets CO48 and CO70, diet treatment had no influenceon the terminal concentrations of ash, crude protein,lipid, and gross energy in the fish (Table 5).
Total body lipid compositions of the juvenile red seabream following the 12-week feeding trial mirrored thetrends described above for dietary fatty acid composi-tions (Tables 3 and 6). Dietary CO level was positivelycorrelated with fish 20:0, 18:1n-(9+7),ΣMFA, 18:2n-6,20:2n-6, Σn-6 PUFA (mainly due to step-wise eleva-tions in 18:2n-6), 18:3n-3, and negatively correlatedwith 14:0, 16:0, ΣSFA, 16:1n-7, 20:1n-(11+9), 22:1n-(11+13+9), 20:4n-6, 22:4n-6, 22:5n-6, 18:4n-3, 20:4n-3, 20:5n-3, 22:5n-3, 22:6n-3, Σn-3 HUFA, Σn-3 PUFA(primarily due to progressive reductions in n-3 HUFA)and the ratio of n-3 to n-6 fatty acids. Key fatty acids ofphysiological importance in body lipids, namely, 20:4n-6 (AA), 20:5n-3 (EPA) and 22:6n-3 (DHA) were re-duced respectively by 60%, 83%, and 54% in fish in-gesting diet CO70 versus those consuming diet FO.Total levels of PUFA in body lipids were not influencedsignificantly by diet treatment largely because the pro-
gressive declines in concentrations of n-3 PUFA wereoffset by the incremental rises in n-6 PUFA (Table 6).
Fatty acid compositions of liver polar lipids of the fishon day 84 (Table 7) generally exhibited different profilesrelative to respective dietary and whole body lipid fattyacid compositions (Tables 3 and 6). Concentrations ofSFA (mainly due to elevations of 16:0 and 18:0), n-3HUFA (especially 22:6n-3), n-3 PUFA, and total PUFAwere markedly higher than noted in the dietary and bodylipids; however, those forΣMFAwere lower (Tables 3, 6and 7). Further,ΣSFA did not follow a consistent trend inrelation to the dietary CO concentration and werehighest for CO25-fed fish followed by those fed diets
Fig. 1. Relationship (±1 SD) between whole body (WB) and liver polar lipid (LP) fatty acid concentrations and dietary concentrations of (a).18:2n-6,(SlopeWB=0.8305, R2
WB=0.9928; SlopeLP=0.2099,R2LP=0.9884), (c). AA (SlopeWB=0.7333, R2WB=0.9942; SlopeLP=1.2362, R2LP=0.7568), (d). EPA (SlopeWB=0.6464, R2WB=0.9984;SlopeLP=0.3237, R
2LP=0.8827) and (e). DHA (SlopeWB=0.9834, R
2WB=0.8168; SlopeLP=1.1128, R
2LP=0.9114) in red sea bream juveniles fed
either FO, CO25, CO48 or CO70 (n=3/diet treatment based on 3 (WB) or 5 (LP) fish per replicate). The grey line in each graph indicates the line oflinearity. Standard deviations are plotted but are within the boundaries of the data points.
427S.S.Y. Huang et al. / Aquaculture 271 (2007) 420–431
FO, CO48 and CO70. Moreover, within MFA, key fattyacids such as 16:1n-7 and 20:1n-(11+9) did not show aconsistent relationship with dietary CO concentration.Thus, incremental changes inΣMFA among treatmentswere not found, unlike in body lipids.Within n-6 PUFA,levels of 20:4n-6 and 22:4n-6 in fish fed diets CO25,CO48 and CO70 were relatively maintained whereas22:5n-6 was found to be higher in fish fed diets CO48and CO70 relative to fish fed diet FO, which differedfrom what was seen in the body lipids. With respect ton-3 PUFA, levels of 22:6n-3 were maintained and in allcases exceeded 29% of total fatty acids and did notdiffer among the fish fed diets with varying levels ofCO. Moreover, liver polar lipid concentrations of EPAand 22:5n-3, although inversely related to dietary COconcentration, showed less of a stepwise decline acrosstreatments than observed in the body lipids (Tables 6and 7). Also, ratios of n-3 to n-6 fatty acids and of 18:1to Σn-3 HUFA in liver polar lipids were found to be
inversely and directly related to dietary CO concentra-tion, respectively.
To better illustrate the findings for fatty acid utilizationand retention, whole body and liver polar lipid concentra-tions of 18:2n-6, 18:3n-3, AA, EPA, and DHA weregraphed as functions of their respective dietary levels(Fig. 1). The retentions of these fatty acids were found todiffer between the body and liver polar lipids. In bodylipids, 18:2n-6, 18:3n-3 and EPA were relatively well-conserved, and their rates of utilization increased slightlyas their respective dietary levels were raised. However, inliver polar lipids, these fatty acids were noted to bepreferentially utilized, particularly at high dietary levels.Nonetheless, levels of AA and DHAwere maintained inboth body and liver polar lipids regardless of dietary level.These fatty acids also appeared to be produced in the fishas indicated by the slopes of 0.73 and 1.24 for AA and0.98 and 1.12 for DHA obtained for whole body and liverpolar lipids, respectively (Fig. 1c, e).
428 S.S.Y. Huang et al. / Aquaculture 271 (2007) 420–431
4. Discussion
The present study indicates that all of the supplementalpollock liver oil (10% of diet) in a premium qualitypractical diet for red sea bream can be replaced by refinedcanola oil without any adverse effects on their growth,general health (survival) or whole body proximatecomposition. Indeed, our results showed that CO couldcomprise up to 70% of total dietary lipid content providedthat some fish oil (residual oil from fish meal) was presentconcurrently to furnish adequate quantities of EFA,specifically EPA, DHA and likely AA, for good growthand health. In this regard, diet CO70 contained about0.75%n-3HUFA and 0.03%AAand it is conceivable thatthese dietary concentrations were just adequate ormarginal. This is because the growth rate, feed intakeand feed and protein utilization of the red sea bream feddiet CO70 tended to be numerically lower than respectivevalues noted for fish fed the other diets; however, thedifferences were not statistically significant. Interestingly,the trends described above were not seen for fish fed dietCO48 which contained about 1.28% n-3 HUFA and0.045% AA. Takeuchi et al. (1990, 1992) proposed thatthe EFA needs of juvenile red sea bream are satisfiedwhen the diet contains 1% EPA, 0.5% DHA or 2.8% n-3HUFA, when dietary lipid is at the optimal level of 15%(Takeuchi et al., 1991). All of the test diets in this studymet the 0.5% DHA but not the 1% EPA requirement.Hence, our results support the notion that DHA has higherefficacy thanEPA in red sea bream (Takeuchi et al., 1990).Nevertheless, chronic assessments should be conducted toconfirm the acceptable dietary CO concentration found inthis study in view of the trends noted above.
While our results did not clearly indicate that therewas an optimal balance between dietary levels of EPA,DHA and AA, it is interesting to note that the red seabream fed diet CO25 exhibited a trend toward improvedgrowth, feed intake and feed utilization relative to thoseconsuming other test diets. However, again, the differ-ences between the groups for each of the foregoingperformance parameters were not significant. A study oflonger duration may provide more definitive evidencewith respect to this point.
In any case, the present findings agree with and extendthose obtained in short and long-term studies on giltheadsea bream, European sea bass, and pink snapper in whichgrowth performance measures were found to be uncom-promised when rapeseed oil or refined canola oil was usedto replace 60% (former two species) to 100% (latter spe-cies) of the supplemental fish oil (Glencross et al., 2003a;Montero et al., 2003; Izquierdo et al., 2005;Mourente et al.,2005). Furthermore, the present study exceeded the∼40%
total dietary lipid substitution with refined CO achieved byGlencross et al. (2003a) who worked with older juveniles(IBW=28 g) and over a shorter time period (54 days).Although the second study by Glencross et al. (2003b) didobserve that CO could comprise 68% of the dietary lipid ofpink snapper, it is noteworthy that the fish were in arelatively slow growth phase (final body weights were∼1.5 fold higher than the initial body weights) and thestudy duration was short i.e., only 32 days. By contrast, thered sea bream in this study exhibited a 13 fold increase overtheir initial bodyweight over an 84-day period, irrespectiveof diet treatment. Moreover, all groups in the present studyshowed excellent values for feed and protein utilization thatwere generally markedly improved relative to the valuesnoted for the same parameters in the studies by Glencrosset al. (2003a,b). Our results also agree with the findings ofstudies on Pacific and Atlantic salmon which have shownthat canola oil is an excellent source of supplemental die-tary lipid provided that the diets contain sufficient concen-trations of essential fatty acids for good growth and health(Dosanjh et al., 1984, 1988, 1998; Rosenlund et al., 2001).
Whole body proximate compositions, with the excep-tion of percent moisture, were not affected by diet treat-ment. Although differences in whole body moisturepercentages are often accompanied by reciprocal changesin whole body lipid content (Bendiksen et al., 2003), thelatter effect was not found in this study. Since feed intake,feed efficiency, and the ratio of dietary protein to lipid orenergy also did not vary among treatments and thesevariables are known to influence whole body and filletproximate constituents in other fish species (Higgs et al.,1995; Rasmussen, 2001), it is likely that the dissimilarmoisture contents found between groups in this study didnot represent an effect of biological significance.
Whole body fatty acid compositions of the fish gen-erally reflected the trends that were observed in the fattyacid concentrations of the diet treatments. In this regard,CO is a richer source of unsaturated fatty acids, 18:1n-9,18:2n-6, and 18:3n-3 relative to pollock liver oil. Also,unlike the latter lipid source, CO is devoid of n-3 HUFA.Furthermore, marine finfish species exhibit limited or noability to desaturate and elongate 18:3n-3 to n-3HUFA andsometimes EPA to DHA as well as 18:2n-6 to 20:4n-6 dueto low or absent Δ6- and Δ5- desaturase activities (Bellet al., 1994; Montero et al., 2004; Izquierdo et al., 2005).Thus, alterations in the whole body fatty acid compositionsof the red sea bream were strongly influenced by thedissimilar fatty acid compositions of pollock liver oil andcanola oil and their respective dietary levels as well as theinherent abilities of the red sea bream to metabolize fattyacids. In regard to the latter, our findings revealed that thered sea bream is similar to other marine finfish species
429S.S.Y. Huang et al. / Aquaculture 271 (2007) 420–431
investigated since they exhibited little ability to utilizesaturated fatty acids for energy purposes. Lipolytic activi-ties and fatty acid adsorption in red sea bream may beinversely related to the melting-point of the fatty acids ashas been observed in some other studies on fish (Lie et al.,1987; Johnsen et al., 2000). Thus, whole body concentra-tions of SFA in red sea bream were strongly negativelyrelated to the dietary CO concentration. Values of ΣMFA,on the other hand, were positively influenced by dietarycanola oil concentration.
Other researchers have also observed similar resultswhen studying the effects of rapeseed or canola oil inclu-sion in diets of salmonids and marine species (Bell et al.,2001, 2003; Glencross et al., 2003a;Montero et al., 2003).Further, our results suggest that one of the dietary MFA,namely, 22:1n-(11+13+9) was used preferentially as asource of energy by the red sea bream. The others that mayhave been utilized similarly as sources of energy includedLA and LNA and possibly EPA, although the latter fattyacid appeared to be metabolically converted to 22:6n-3 aswell. This conclusion is based on our observation of lowerconcentrations of EPA in body lipids than in respectivedietary lipids of fish given the different treatments and thelower decline of DHA relative to EPA in the body lipids offish fed diet CO70. Collectively, the preceding findingsexplain some of the departures that we observed in wholebody fatty acid compositions of the red sea bream relativeto their dietary fatty acid compositions. This is furtherillustrated in Fig. 1, in which at higher dietary concentra-tion, EPA is preferentially oxidized or bioconverted inbody lipid. Furthermore, the percent decline of 22:6n-3 inthe body lipids was not as great in fish fed diet CO70 asthat seen for 20:5n-3 (Tables 3 and 6). Moreover, thesefindings are in accord with the fatty acid metabolismabilities of marine finfish species in general as describedabove. Our observation of limited bioconversion of 18:2n-6 in the red sea bream of this study, based upon therelationship between the respective diet and terminalwhole body lipid levels of LA and 20:4n-6 (Fig. 1a, c), alsoagrees with the findings of others who have shown somedegree ofΔ6-desaturation inmarine species such as turbot(Linares and Henderson, 1991), gilthead sea bream, andgolden grey mullet (Liza aurata; Mourente and Tocher,1993a,b). This is further supported by terminal levels ofbody 18:3n-6 which were not present in dietary lipids.
As mentioned previously, liver polar lipid compositionsgenerally exhibited different trends to those observed inwhole body and dietary lipids. Signs of selective utilizationand retention of specific fatty acids were much more pro-found in liver polar lipids. In this regard, ΣSFA werehighest and lowest in CO25 and CO70-fed fish, respec-tively. However, those for MFA did not differ among fish
fed diets with varying CO levels and they were all sig-nificantly higher than found in fish fed diet FO (controldiet). However, SFAwere accumulated to a greater extentand MFA to a lesser extent in liver polar lipids relative tobody lipids. Likely, this reflected the preferential retentionof saturated as opposed to monounsaturated fatty acids onposition 1 of the phospholipids. Moreover, the red seabream were also noted to accumulate and retain far more22:6n-3 (all treatments) and to a lesser extent 20:5n-3 (fishfed diets with CO) and 22:5n-3 (all treatments) in polarlipids versus body lipids. The preceding findings agreewith those of Yone and Fujii (1975), who observed similarincreases in liver phospholipid DHA levels as more LNAwas added to corn oil-based diets for juvenile red seabream. Further, the accumulation of increased quantities ofn-3 HUFA in polar lipids versus body lipids of the red seabream explains whywe notedmuch higher levels of PUFAin liver polar lipids of all groups relative to body lipids,which are mostly comprised of triacylglycerols in fish(Higgs et al., 1995). The elevated levels of PUFA in polarlipids versus body lipids were not due to elevations in n-6PUFA, as these showed a narrower range in the formerlipids than in the latter, even though the predominant n-6fatty acid, 18:2n-6, was positively correlated with dietaryCO concentration in each case. However, we notedmarkedincreases in all of the 20 carbon fatty acids of the n-6 familyin liver polar lipids relative to levels in dietary and bodylipids, particularly 20:3n-6, which correlated to dietary COlevel. Thus, there may be bioconversion of 18:2n-6 to20:4n-6 in liver polar lipids of the red sea bream and/orselective incorporation of the 20 carbon n-6 fatty acids intothe membranes of hepatocytes (Seliez et al., 2003).
Undesirable elevations in liver lipid content havebeen seen in red sea bream when the diets have con-tained less than 0.5% of EPA and/or DHA (Takeuchiet al., 1990). Ratios of 18:1n-9/n-3 HUFA in excess of 1have also been associated with poor growth perfor-mance and feed efficiency in fingerlings (Fujii et al.,1976; Yone, 1978). The latter effects were not observedin this study and it is noteworthy that all of our test dietscontained adequate concentrations of DHA for red seabream and the ratios of 18:1 to n-3 HUFA were notgreater than 0.51 in the liver polar lipids. Hence the redsea bream in this study did not exhibit any signs of EFAdeficiency even when dietary CO concentration was ashigh as 70% of the total lipid content.
In conclusion, our findings indicate that refined ca-nola oil is a potentially suitable dietary lipid source forjuvenile red sea bream under our test conditions. How-ever, the long-term effects of using canola oil as themain source of supplemental dietary lipid still need to beassessed before definitive recommendations can be
430 S.S.Y. Huang et al. / Aquaculture 271 (2007) 420–431
made. Future work should also include evaluation of thehaematological and immunological responses of the fishin relation to the dietary canola oil level and the con-comitant reductions in the levels of the essential highlyunsaturated fatty acids of the n-3 and n-6 families.
Acknowledgements
The authors are grateful to AquaNet for the researchgrant extended to S.S.Y. Huang. We also thank K.Yamada, K. Kamada, H. Kudoh, and T. Kaneko forassistance with fish sampling and technical support.Lastly we express our thanks to Seiho Suisan Co. Ltd.for their generous supply of the experimental fish.
References
AOAC, 2000. In: Horwitz, W. (Ed.), Official Methods of Analysis ofAOAC International. AOAC International, Gaithersburg, MD.
Barlow, S., 2000. Fishmeal and fish oil: sustainable feed ingredientsfor aquafeeds. Glob. Aquacult. Advocate 4, 85–88.
Bell, M.V., Henderson, R.J., Sargent, J.R., 1986. The role of polyunsatu-rated fatty acids in fish. Comp. Biochem. Biophys. 967, 345–363.
Bell, J.G., Tocher, D.R., MacDonald, F.M., Sargent, J.R., 1994. Effectsof diets rich in linoleic (18:2n-6) and alpha-linolenic (18:3n-3)acids on the growth, lipid class and fatty acid compositions andeicosanoid production in juvenile turbot (Scophthalmus maximusL.). Fish Physiol. Biochem. 13, 105–118.
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 dietsofAtlantic salmon (Salmo salar) affects tissue lipid compositions andhepatocyte fatty acid metabolism. J. Nutr. 131, 1535–1543.
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 oilwith crude palm oil in the diet of Atlantic salmon (Salmo salar)affects muscle fatty acid composition and hepatic fatty acidmetabolism. J. Nutr. 132, 222–230.
Bell, J.G., McGhee, F., Campbell, P.J., Sargent, J.R., 2003. Rapeseedoil as an alternative to marine fish oil in diets of post-smolt Atlanticsalmon (Salmo salar): changes in flesh fatty acid composition andeffectiveness of subsequent fish oil “wash out”. Aquaculture 218,515–528.
Bendiksen, E.A.,Arnesen,A.M., Jobling,M., 2003. Effects of dietary fattyacid profile and fat content on smolting and seawater performance inAtlantic salmon (Salmo salar L.). Aquaculture 225, 149–163.
Canola Council of Canada, 2004. Canola, Standards and Regulations.Canola Council of Canada Publications. 8 p. http://www.canola-council.org/pubs/standards.html.
Canola Council of Canada, 2005. Canola Facts: A Major CanadianExport. http://www.canola-council.org/facts_export.html.
Chatelier, A., McKenzie, D.J., Prinet, A., Galois, R., Robin, J.,Zambonino, J., Claireaux, G., 2006. Associations between tissuefatty acid composition and physiological traits of performance andmetabolism in the seabass (Dicentrarchus labrax). J. Exp. Biol.209, 3429–3439.
Christie, W.W., 1973. Lipid Analysis. Pergamon of Canada Ltd,Toronto, Canada, p. 90.
Dosanjh, B.S., Higgs, D.A., Plotnikoff, M.D., McBride, J.R., Markert,J.R., Buckley, J.T., 1984. Efficacy of canola oil, pork lard and
marine oil singly and in combination as supplemental dietary lipidsources for juvenile coho salmon (Oncorhynchus kisutch).Aquaculture 36, 333–345.
Dosanjh, B.S., Higgs, D.A., Plotnikoff, M.D., Markert, J.R., Buckley,J.T., 1988. Preliminary evaluation of canola oil, pork lard andmarine lipid singly and in combination as supplemental dietarylipid sources for juvenile fall chinook salmon (Oncorhynchustshawytscha). Aquaculture 68, 325–343.
Dosanjh, B.S., Higgs, D.A., McKenzie, D.J., Randall, D.J., Eales, J.G.,Rowshandeli, N., Rowshandeli, M., Deacon, G., 1998. Influenceof dietary blends of menhaden oil and canola oil on growth, musclelipid composition, and thyroidal status of Atlantic salmon (Salmosalar) in sea water. Fish Physiol. Biochem. 19, 123–134.
Food and Agriculture Organization – Fisheries Department. (FAO-FD),2004. State of World Fisheries and Aquaculture. http://www.fao.org/DOCREP/007/y5600e/y5600e00.htm.
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, 407–509.
Fujii, M., Nakayama, H., Yone, Y., 1976. Effect of ω3 fatty acids ongrowth, feed efficiency and fatty acid composition of red sea bream(Chrysophrys major). Rep. Fish. Res. Lab. Kyushu. Univ. 3, 65–86.
Ghioni, C., Tocher, D.R., Bell, M.V., Dick, J.R., Sargent, J.R., 1999.Low C-18 and C-20 fatty acid elongase activity and limitedconversion of stearidonic acid, 18:4(n-3), to eicosapentaenoic acid,20:5(n-3), in a cell line from the turbot, Scophthalmus maximus.Biochim. Biophys. Acta. 1437, 170–181.
Glencross, B., Hawkins, W., Curnow, J., 2003a. Evaluation of canolaoils as alternative lipid resources in diets for juvenile red seabream,Pagrus auratus. Aquacult. Nutr. 9, 305–315.
Glencross, B., Hawkins, W., Curnow, J., 2003b. Restoration of thefatty acid composition of red seabream (Pagrus auratus) using afish oil finishing diet after grow-out on plant oil based diets.Aquacult. Nutr. 9, 409–418.
Grant, A.A.M., 2006. Growth, fatty acid composition and Na+/K+-ATPase isoform physiology of juvenile chinook salmon(Oncorhynchus tshawytscha) fed diets supplemented withanchovy or blends of anchovy and canola oil. MSc. thesis,University of British Columbia, Vancouver, Canada, pp. 35-61.
Hardy, R.W., Higgs, D.A., Lall, S.P., Tacon, A.G.J., 2001. Alternativedietary protein and lipid sources for sustainable production ofsalmonids. Fisken Og Havet NR, 8-2001. Havforskningsinstituttet(Institute of Marine Research), Bergen, Norway, p. 58.
Higgs, D.A., Dong, F.M., 2000. Lipids and fatty acids. In: Stickney,R.R. (Ed.), The Encyclopedia of Aquaculture. John Wiley andSons, New York, U.S.A., pp. 476–496.
Higgs, D.A., Macdonald, J.S., Levings, C.D., Dosanjh, B.S., 1995.Nutrition and feeding habits in relation to life history stage. In:Groot, C., Margolis, L., Clarke, W.C. (Eds.), PhysiologicalEcology of Pacific Salmon. UBC Press, Vancouver, pp. 159–315.
Higgs, D.A., Balfry, S.K., Oakes, J., Rowshandeli, M., Skura, B.J.,Deacon, G., 2006. Efficacy of an equal blend of canola oil andpoultry fat as an alternate dietary lipid source for Atlantic salmon(Salmo salar L.) in sea water. I: Effects on growth performance,and whole body and fillet proximate and lipid composition.Aquacult. Res. 37, 180–191.
Ikonomou, M.G., Higgs, D.A., Gibbs, M., Oakes, J., Skura, B.,McKinley, S., Balfry, S.K., Jones, S., Withler, R., Dubetz, C., 2007.Flesh quality of market-size farmed and wild British Columbiasalmon. Environ. Sci. Technol. 41, 437–443.
Izquierdo, M.S., Obach, A., Arantzamendi, L., Montero, D., Tobaina,L., Rosenlund, G., 2003. Dietary lipid sources for seabream and
Izquierdo, M.S., Montero, D., Tobaina, L., Caballero, M.J., Rosen-lund, G., Gines, R., 2005. Alterations in fillet fatty acid profile andflesh quality in gilthead seabream (Sparus aurata) fed vegetableoils for a long term period. Recovery of fatty acid profiles by fishoil feeding. Aquaculture 250, 431–444.
Johnsen, R.I., Grahl-Nielsen, O., Roem, A., 2000. Relative absorption offatty acids by Atlantic salmon Salmo salar from different diets, asevaluated by multivariate statistics. Aquacult. Nutr. 6, 255–261.
Juaneda, P., Rocquelin, G., 1985. Rapid and convenient separation ofphospholipids and nonphosphorus lipids from rat heart using silicacartridges. Lipids 20, 40–41.
Lie, O., Lied, E., Lambertsen, G., 1987. Lipid digestion in cod (Gadusmorhua). Comp. Biochem. Physiol. 88B, 697–700.
Linares, F., Henderson, R.J., 1991. Incorporation of 14C-labelledpolyunsaturated fatty acids by juvenile turbot, Scophthalmusmaximus (L.) in vivo. J. Fish Biol. 38, 335–347.
Menoyo, D., Izquierdo, M.S., Robaina, L., Gines, R., Lopez-Bote, C.J.,Bautista, C.J., 2004. Adaptation of lipid metabolism, tissue composi-tion and flesh quality in gilthead sea bream (Sparus aurata) to the fishoil replacement by linseed and soybean oils. Brit. J. Nutr. 92, 41–52.
Montero, D., Kalinowski, T., Obach, A., Robaina, L., Tort, L.,Caballero, M.J., Izquierdo, M.S., 2003. Vegetable lipid sources forgilthead seabream (Sparus aurata): effects on fish health.Aquaculture 225, 353–370.
Montero, D., Socorro, J., Tort, L., Caballero, M.J., Robaina, L.E., Vergara,J.M., Izquierdo, M.S., 2004. Glomerulonephritis and immunosuppres-sion associated with dietary essential fatty acid deficiency in giltheadsea bream, Sparus aurata L., juveniles. J. Fish Dis. 24, 297–306.
Mourente, G., Tocher, D.R., 1993a. Incorporation and metabolism of14C-labeled polyunsaturated fatty-acids in juvenile gilthead seabream Sparus aurata, in vivo. Fish Physiol. Biochem. 10, 433–453.
Mourente, G., Tocher, D.R., 1993b. Incorporation and metabolism of14C-labeled polyunsaturated fatty-acids in juvenile wild-caughtjuveniles of golden grey mullet, Liza aurata, in vivo. Fish Physiol.Biohem. 12, 119–130.
Mourente, G., Good, J.E., Bell, J.G., 2005. Partial substitution of fish oilwith rapeseed, linseed and olive oils in diets for European sea bass(Dicentrarchus labrax L.): effects on flesh fatty acid composition,plasma prostaglandins E2 and F2α, immune function and effective-ness of a fish oil finishing diet. Aquacult. Nutr. 11, 25–40.
Mozaffarian, D., Rimm, E.B., 2006. Fish intake, contaminants, and humanhealth-evaluating the risks and benefits. JAMA 296, 1885–1898.
Nakabo, T., 1993. Fishes of Japan with Pictorial Keys to the Species.Tokai University Press, Tokyo (in Japanese).
New, M.B., Wijkström, U.N., 2002. Use Of Fishmeal And Fish Oil InAquafeeds: Further Thoughts On The Fishmeal Trap. FAOFisheries Circvol, vol. 975.
Paulin, C., 1990. Pagrus auratus, a new combination for the speciesknown as “snapper” in Australasian waters (Pisces: Sparidae).N. Z. J. Mar. Freshw. Res. 24, 259–265.
Rasmussen, R.S., 2001. Quality of farmed salmonids with emphasis onproximate composition, yield and sensory characteristics. Aqua-cult. Res. 32, 767–786.
Regost, C., Arzel, J., Robin, J., Rosenlund, G., Kaushik, S.J., 2003. Totalreplacement of fish oil by soybean or linseed oil with a return to fishoil in turbot (Psetta maxima): 1. Growth performance, flesh fattyacid profile, and lipid metabolism. Aquaculture 217, 465–482.
Rosenlund, G., Obach, A., Sandberg, M.G., Standal, H., Tveit, K.,2001. Effect of alternative lipid sources on long-term growthperformance and quality of Atlantic salmon (Salmo salar L.).Aquacult. Res. 32 (Suppl 1), 323–328.
Sargent, J.R., Tocher, D.R., Bell, J.G., 2002. The lipids. In: Halver,J.E., Hardy, R.W. (Eds.), Fish Nutrition. Academic Press, SanDiego, pp. 181–257.
Seliez, I., Panserat, S., Corraze, G., Kaushik, S., Bergot, P., 2003.Cloning and nutritional regulation of a Δ6-desaturase-like enzymein the marine teleost gilthead seabream (Sparus aurata). Comp.Biochem. Physiol. 135B, 449–460.
Steffens, W., 1993. Significance of extruded feeds for trout nutritionand water protection. Arch. Tierenahr. 4, 189–210.
Tacon, A.G.J., 2004. Use of fish meal and fish oil in aquaculture: aglobal perspective. Aquat. Res. Cult. Dev. 1, 3–14.
Tabata, K., Taniguchi, N., 2000. Differences between Pagrus majorand Pagrus auratus through mainly mtDNA control regionanalysis. Fish. Sci. 66, 9–18.
Takeuchi, T., Toyota, M., Satoh, S., Watanabe, T., 1990. Require-ment of juvenile red seabream Pagrus major for eicosapentae-noic and docosahexaenoic acids. Nippon Suisan Gakkaishi 56,1263–1269.
Takeuchi, T., Shiina, Y., Watanabe, T., 1991. Suitable protein and lipidlevels in diet for fingerlings of red sea bream Pagrus major.Nippon Suisan Gakkaishi. 57, 293–299.
Takeuchi, T., Shiina, Y., Watanabe, T., 1992. Suitable levels of n-3highly unsaturated fatty acids in diet for fingerlings of red seabream. Nippon Suisan Gakkaishi 58, 509–514.
Tocher, D.R., Ghioni, C., 1999. Fatty acid metabolism in marine fish:low activity of fatty acyl delta 5 desaturation in gilthead sea bream(Sparus aurata) cells. Lipids 34, 433–440.
Tocher, D.R., Carr, J., Sargent, J.R., 1989. Poly-unsaturated fatty-acidmetabolism in fish cells differential metabolism of (n-3) and (n-6)series acids by cultured-cells originating from a fresh-water teleostfish and from a marine teleost fish. Comp. Biochem. Physiol. 94,367–374.
Wagner, G.N., Balfry, S.K., Higgs, D.A., Lall, S.P., Farrell, A.P., 2004.Dietary fatty acid composition affects the repeat swimmingperformance of Atlantic salmon in seawater. Comp. Biochem.Physiol. 137, 567–576.
Watanabe, T., 2002. Strategies for further development of aquaticfeeds. Fish. Sci. 68, 242–252.
Watanabe, T., Vassallo-Agius, R., 2003. Broodstock nutrition researchon marine finfish in Japan. Aquaculture 227, 35–61.
Woyewoda, A.D., Shaw, S.J., Ke, P.J., Burns, B.G., 1986. Recom-mended laboratory methods for assessment of fish quality.Canadian Technical Report of Fisheries and Aquatic Sciences,No. 1448.
Yone, Y., 1978. Essential fatty acids and lipid requirements of marinefish. In: Japan. Soc. Sci. Fish. (Ed.), Dietary Lipids in Aquaculture.Koseisha-kosei-kaku, Tokyo, pp. 43-49.
Yone, Y., Fujii, M., 1975. Studies on nutrition of red sea bream— XII:effects of ω3 fatty acid supplement in a corn oil diet on fatty acidcomposition of fish. Bull. Jpn. Soc. Sci. Fish. 41, 79–86.
