2012外文鱼油中含有有毒物质植物油替代可以减少有毒物质

Aquaculture324–325(2012)118–126

ContentslistsavailableatSciVerseScienceDirect

Aquaculture

journalhomepage:http://wendang.chazidian.com/locate/aqua-online

Completereplacementof shoilwithablendofvegetableoilsaffectsdioxin,

dioxin-likepolychlorinatedbiphenyls(PCBs)andpolybrominateddiphenylethers(PBDEs)in3Atlanticsalmon(Salmosalar)familiesdifferingin eshadiposity

J.G.Bella, ,J.R.Dicka,FionaStrachana,D.R.Guyb,M.H.G.Berntssenc,M.Spraguea

abc

InstituteofAquaculture,UniversityofStirling,StirlingFK94LA,Scotland,UK

LandcatchNaturalSelectionLtd,Thee-Centre,CooperageWay,AlloaFK103LP,Scotland,UKNationalInstituteofNutritionandSeafoodResearch(NIFES),Bergen,Norway

articleinfoabstract

Thebene tsofconsumingadietrichinseafoodarenowwellrespectedandarebasednotonlyonthehighlevelsoflong-chainn 3polyunsaturatedfattyacids(LCn 3PUFA)butalsoduetotherangeofbene cialmacroandmicronutrientspresentin sh.Atlanticsalmoncultureisnowestablishedgloballyandisamajorsourceofhigh-qualityoilrichinLCn 3PUFA.However,salmon eshcanaccumulatepersistentorganicpollutants(POPs),in-cludingdioxinsandfurans(PCDD/Fs),dioxin-likepolychlorinatedbiphenyls(DL-PCBs)andpolybrominateddiphenylethers(PBDEs)thatarederivedfrommarinefeedcomponents,especially shoil(FO).TheaimofthisstudywastogrowsalmonsmoltsonalternativedietswithareducedFMlevel,thatcontainedeither100%ofaddedoilasFOorablendofvegetableoils(VO),overafullproductioncycleandtoassesstheeffectsofthesedietsonPOPdeposition.Threefamiliesofsalmonwereusedwithtwobeingchosenasbeingeither“Lean”or“Fat”,basedon eshadiposityderivedfromabreedingprogramme,whilethethird(CAL)wasamixofnon-pedigreedcommercialfamilies.Fishwereongrownfor55weekswhentheyreached3kgfollowedbyaswitchtoadecontaminatedFOforafurther24weekstorestoreLCn 3PUFAlevelsintheVO-fed sh.Theav-eragesum eshPCDD/Fs,acrossthe3salmonstrains,werereducedfrom1.94±0.01ngTEQ/kginthe shfedFOto0.46±0.02ngTEQ/kginthe shfedVO.ThesumPBDEswerereducedfrom2.82±0.24ng/gintheFO shto0.52ng/gintheVO sh.AveragereductioninsumPCDD/Fs,DL-PCBsandPBDEswas52,79and82%,respectively,intheVOfedsalmon.TherewasevidenceofhigherPBDEretentionintheFat shbutthiswasnotsigni cant.ThisstudyshowsthatsalmoncanbeproducedwithverylowlevelsofPOPsandthatconcentrationscanbereducedsigni cantlybythecarefulselectionofrawmaterials.Theuseofdecontaminated shoilshasanimportantroleinthisprocessalthoughcareshouldbetakentouseoilsthataretreatedwithprotocolsthatreducePCCD/Fs,DL-PCBsandPBDEstoensureverylowlevelsofPOPsincommercialsalmon.

©2011ElsevierB.V.Allrightsreserved.

Articlehistory:

Received24August2011

Receivedinrevisedform3November2011Accepted4November2011

Availableonline11November2011Keywords:

PersistentorganicpollutantsFarmedAtlanticsalmonAlternativedietaryoilsGeneticstrainorfamilyVegetableoilFoodsafety

1.Introduction

Demandforhigh-qualityseafoodcontinuestoincreasedespitethefactthatcapture sherieshavereachedtheirsustainablelimits,orareindecline,whileaquaculturecurrentlysupplies50%ormoreofallsea-foodconsumed(FAO,2009).Atlanticsalmon(Salmosalar)culturehascontinuedtogrowrapidly,with1.5mMTproducedin2006(FAO,2008),althoughthisgrowthhasbeengeneratedbyexploitingmarine shoiland shmeal(FOandFM)whichisnotsustainable,especiallyinthecaseofFO,eveninthemediumterm(TaconandMetian,2008;Taconetal.,2006).Oily sharehoweveramajorsourceoflong-chainn 3polyunsaturatedfattyacids(LCn 3PUFA),includingeicosapen-taenoicanddocosahexaenoicacids(20:5n 3;EPAand22:6n 3,

Correspondingauthor.Tel.:+441786467997;fax:+441786472133.E-mailaddress:g.j.bell@stir.ac.uk(J.G.

Bell).0044-8486/$–seefrontmatter©2011ElsevierB.V.Allrightsreserved.doi:

10.1016/j.aquaculture.2011.11.004

DHA)whichhaveprovenhealthbene tsforconsumers.Whilemoststudieshavebeenconductedoncardiovasculardisease,immunefunc-tionandin ammatorydisorders,bene tshavealsobeenshownforcog-nitivefunctions,developmentaldisordersandinfantdevelopment(Hibbelnetal.,2007;Ruxtonetal.,2005;Siddiquietal.,2009;Yusofetal.,2008).

However,despitetheprovenbene tsofoily shconsumptionthereisconcernregardingthepresenceofthefatsolublepersistentorganicpollutants(POPs)inmarine shoilswhichaccumulateinthewildpe-lagic shspeciesthatisusedintheproductionofFOandFM.Thus,FOderivedfromthesepelagic shareresponsibleformostofthePOPs,in-cludingdioxins/furans(PCDD/F),dioxin-likepolychlorinatedbiphenyls(DL-PCBs)andpolybrominateddiphenylethers(PBDEs),thatarefoundin shfeeds(Berntssenetal.,2007,2010;Hitesetal.,2004a).ThehighlipidcontentofsalmondietsincreasesPOPdepositioninsalmon esh,comparedtoleaner shspecies,andthishasresultedinhealthconcerns(Hitesetal.,2004b;Shawetal.,2008).Itisclearthatthereisatrade-off

J.G.Belletal./Aquaculture324–325(2012)118–126119

betweenthebene tsofoily shconsumption,duetoLCn 3PUFAandotherbene cialnon-lipidnutrients,andthepresenceofpotentiallyharmfulPOPs(Dewaillyetal.,2007;Domingoetal.,2007;Foranetal.,2005a,b)althoughthereisagreementthatthebene tsgenerallyexceedtherisks(EFSA,2005;SACN/COT,2004).

However,asdietaryformulationshaveevolvedinrecentyears,FOinclusionhasreducedandgradualsubstitutionofFOwithterrestrialvegetableoils(VOs)hasbeenshowntoreducePOPconcentrationsinthe eshofbothsalmonandrainbowtrout(Belletal.,2005;Berntssenetal.,2005,2010;Drewetal.,2007).Whilerecentesti-matessuggestthatFOuseinaquacultureisaround88%oftotalpro-duction,andthereforeisthemostpressingcaseforreduction,68%ofglobalFM,comprising~10%FO,isalsousedinaquafeedsandcur-rentdevelopmentswouldseektoreducebothFOandFMinfuturefeeds(TaconandMetian,2008;Torstensenetal.,2008).Thereduc-tionindietaryproteinfromFMwascompensatedusingarangeofplant-basedproteinswhichhavebeendevelopedandintroducedinrecentyears(Drewetal.,2007;Torstensenetal.,2005,2008).

Theaimofthepresentstudywastogrowsalmonsmoltsonalter-nativedietswithareducedFMlevelof25%,comparedtocommercialvaluesin2006of~40%FM,thatcontainedeither100%ofaddedoilasFOorablendofVOs,overafullproductioncycleandtomeasuretheimpactofthesedietsonPOPdeposition.Afurtheraimofthestudywastoestablishifdifferencesin eshfatdepositionmightbere ectedindifferentPOPsaccumulationpatternsacrossthedifferentstrains.Eachofthetwodietswasfedtosmoltsselectedfrom“Lean”and“Fat”salmonfamilies,basedon eshadiposity,aswellasacom-mercialstrain(CAL)usedinpreviousFOreplacementstudies(Belletal.,2004;Torstensenetal.,2005).Moredetailson shstrainsin-cludinggrowthperformanceandlipidcompositionsaredescribedinBelletal.(2010)andintheMaterialsandmethods.Inaddition,a nal nishingdietphase,usingadecontaminatedFO,wasusedtore-storeLCn 3PUFAat nalharvestandtheimpactofthisphaseonPOPsaccumulationwasalsomeasured.Thedietswerepreparedas3,6and9mmongrowingfeedsplusa9mm nishingfeedwhichhadbeensubjectedtoadecontaminationprocess.2.Materialsandmethods

2.1.Diets, shpopulationsandhusbandry

Salmonwerefedoneoftwoexperimentaldietswhichwerepre-paredbySkrettingARC(Stavanger,Norway)in3,6and9mmpelletsizesfor55weekstopermitgrowthfrom60gsmoltsto3kg+athar-vest.TheFMandFOselectedforthe3and6mmdietswereintheupperrangeofPOPconcentrationspermittedbyEUlegislationforusein shfeeds(EC,2006a)whilethoseinthe9mmandwashoutfeedswerelowerinPOPs.The3,6and9mmdietsusedcapelinoilwhilethe nishingdietcontainedablendofdecontaminatedsandeelandspratoils.Alldietswereformulatedtomeetthenutritionalrequirementsofsalmonid sh(NRC,1993).Duplicatepensofthethreegenetic shgroupswerefedthesamebasaldietcontainingei-ther100%ofaNorthernFOorablendofVOcomprisingrapeseed,palmandCamelinaoilsina5:3:2ratio.Camelinasativaoilhasnotbeenusedpreviouslyinaquafeedsalthoughithasbeenusedexten-sivelyasalampoilandincosmeticproducts.Itislowinsaturatedfattyacidsandhasan18:3n 3/18:2n 6ratioof~2:1.Itisunusualinthatitcontainshighlevels(~15%)ofthemonoenoicfattyacid,20:1n 9,whichismorecommonlyfoundinmarineFOandisapre-ferredsourceofdietaryenergy.Camelinaseedscontain>40%oilanditcanbegrowninaridconditions,nottoleratedbyotheroilseedsandrequiresminimaluseoffertilisers,herbicidesorpesticides(INFORM,2007).Thedietscontained25–32%FMand40–45%plantproteins.Afterthe nalsamplingat55weeks,60 sh/penforeachofthe3ge-neticgroups,wereplacedinasinglepenwith shpreviouslyfedtheVOdietidenti edbyclippingtheadipose n.Only shintheweight

rangeof2.0–3.5kgwereselected.Thus,theoriginal12penswerere-ducedto3andallwerefeda nishingdiet,withthesamebasalcom-positionasthe9mmgrowerfeedbutcoatedwithadecontaminatedFO(sandeel/spratoil,9:1v/v;TripleNineProtein,Denmark),forafurther24weeks.Thedietformulationsofthe3and9mmgrowerfeedsandthe9mm nishingdiet,theproximateandthemainfattyacidcompositionsareshowninTable1.Dietswereprovidedbyauto-maticfeederscontrolledbyanautomatedfeedsensorsystem(Arvo-Tec,SternerAquaTechUK,Scotland).Feeduseandmortalitieswererecordedandremoveddaily.

ThreegeneticallycharacterisedgroupsoffarmedAtlanticsalmonpost-smoltswereusedforthefeedtrials.Twogroupscomprisedfull-sibfamiliesselectedfromtheLandcatchNaturalSelectionLtdbreedingprogramme(Argyll,Scotland)basedonestimatedbreedingvalues(EBVs)oftheirparentsforhighorlow eshadiposity.Thepa-rentalEBVswerederivedfromTorryFatmeter(DistellIndustries,WestLothian,UK)lipidassessmentsobtainedonliveandharvestedsibsofthoseparents.Fourunrelatedfull-sibfamilieswerethencreat-edfromtheselectedparentsattheNovember2004stripping,twofamiliesfromtheextremelowerendoftheEBVdistributionfor

Table1

Dietformulations,proximatecompositionsandmajorfattyacids(g/100g)inthe3and9growerdietsand9mm nishingfeeds.Ingredient/diet

3mm3mm9mm9mm9mm nishingFOVOFOVOfeedLT shmeala

3232252525Soybeanmealb(extracted)12.512.5111111Wheatb

77.27.97.97.9Wheatglutenc

1515999Soyaconcentrated55777Maizeglutene

4.34.3555Fishoil(Nordic)f22.8–30––Fishoilg

––––30Rapeseedoilh–11–15–Palmoili

–7–9–Camelinaoilj

–4.5–6–Vitaminsandmineralsk1.51.55.15.15.1Astaxanthin(10%)0.050.050.050.050.05Totalplantproteins

43.84439.939.939.9ProximatecompositionslProtein46.846.741.040.537.7Lipid23.524.034.933.532.6Moisture7.47.24.65.17.6Ash

7.16.96.06.08.7Majorfattyacids16:0

16.017.315.015.917.1Totalsaturatesm26.322.726.121.731.118:1n 9

11.736.610.137.611.9Totalmonounsaturatesn41.644.642.647.534.518:2n 64.616.43.516.53.3Totaln 6o6.217.34.617.15.018:3n 31.49.61.610.00.920:5n 38.61.98.51.212.122:6n 310.32.310.11.58.9Totaln 3p

24.4

14.8

25.3

13.4

26.3

aConsortio,Peru.b

Statkorn,Norway.c

CerestarScandinaviaAS,Denmark.d

ADM,Holland.e

Cargill,USA.f

Nordsildmel,Norway.g

TripleNineFishProtein,Denmark.h

EmmelevAS,Denmark.i

Denofa,Norway.j

TechnologyCropsLtd,UK.k

Vitaminandmineralsupplementationtomeetorexceedrequirementsofsalmon(NRC,1993).l

Valuesaremeansoftriplicateanalyses.m

Includes14:0,15:0,18:0,20:0and22:0.n

Includes16:1n 7,16:1n 9,20:1n 11,20:1n 7,18:1n 7,20:1n 9,22:1n 11,22:1n 9and24:1n 9.o

Includes18:3n 6,20:2n 6,20:3n 6,20:4n 6,22:4n 6and22:5n 6.p

Includes18:4n 3,20:4n 3and22:5n 3.

120J.G.Belletal./Aquaculture324–325(2012)118–126

Torryassessedlipid(“Lean”)andtwofamiliesfromtheextremeupperendofthedistribution(“Fat”).TheaverageEBVofthetwoFatfamilieswas2.00percentageunitshigherthanthatofthetwose-lectedLeanfamilies,representingastandardisedselectiondifferentialof2.33standarddeviations(sds).Athirdgroupwasofnon-pedigreedmixedfamiliesderivedfromtheCaledonian20strain(CAL;MarineHarvestScotlandLtd.),whichhadbeenusedinpreviousreplacementtrials(Belletal.,2004;Torstensenetal.,2005).

Twothousandsmoltsfromeachofthethreegroupswerestockedinto12×5m3netpensattheArdnishFeedTrialsUnit(MarineHarvestScot-landLtd.,Lochailort,PH384LZ,Highland)with500 shperpen.Themean shweights±SDatthestartofthetrialwere52.3±1.5,88.2±7.3and85.1±11.2gfortheCAL,FatandLeangroups,respective-ly.Tosummarise,therewerefourpensallocatedtoeachofthethreegroupswitheachofthetwodietsbeingfedtoduplicatepensforeachgroup.

2.2.Samplecollection

Samplesofthe3,6and9mmfeeds,inboththeFOandVOtreat-mentsandasampleofthedecontaminatedFO nishingdiet,werecol-lectedatthestartofthetrialinMay2006,December2006,June2007andSeptember2007foreachofthefourfeeds,respectively.Thesesam-plingpointsrepresentthe3,6and9mmgrowerfeedsandthe9mm nishingdiet,respectively.Feedsampleswerewrappedinfoilandstoredinsealedpolythenebagsat 20°Cuntilanalysis.Ateachofthetimepointsabove20 shperpenweresampledforcontaminantanalysisproviding2poolsof10 shperpen.ThesamplesweretakenfromtheNorwegianQualityCut(NQC)regionandthe10NQCswerewrappedinfoilandfrozenondryice.Aftertransportationtothelabora-torythesampleswerestoredat 20°Cuntilanalysed.The shsampledwereclosetotheaverage shweightineachpen.Beforeanalysisthepooledsteakswerehomogenisedinafoodprocessorfollowingremovalofskinandbones.POPsanalyseswereconductedusingthesehomoge-nates.Replicatedataarepresentedasmean±SD.

2.3.Analysisofpolychlorinateddibenzo-p-dioxins,polychlorinateddi-benzofurans,dioxin-likepolychlorinatedbiphenyls

Fleshhomogenates(~50g)werefreezedriedfor12hbeforeextrac-tion.Diets(~10g)wereweighedandgroundinacoffeegrinderbeforeextraction.Theentire eshordietsampleswerethenextractedwithiso-hexaneusingacceleratedsolventextraction(ASE ;Dionex,Camberley,Surrey,UK).Sampleswereweighed,mixedwithaninert ller(Hydro-matrix ,VarianInc.,PaloAlto,CA,USA)beforeadditionofa13C-labelledPCDD/FandPCBinternalstandard(EPA-1613-LCS;WellingtonLaborato-ries,Guelph,Canada).Sampleswereextractedfor20minusingASE fortwocycleswithisohexaneunderpressure(1500psi:10,342kPa)at125°C.LipidandorganicmatterwereremovedfromtheASE extractwithsulphuricacidinaseparatingfunnel.Thesulphuricacidwasthendrainedoffandtheorganicphasewashedwith50mlofnanopurewater(MilliporeUKLtd.,Watford,Herts,UK)andleftfor3hatroomtemperature.

Theextractedlipidsthenunderwentaclean-upprotocolusinganautomatedmulti-columnPower-Prep system(FluidManagementSystems,Waltham,MA,USA)usingaseriesofdisposableTe oncol-umnsofmulti-layeredsilica(7.5g),basicalumina(8g)andcarbon(2g).Forlipid-richsamples,high-capacitysilicacolumns(28gacidic,16gbasicand6gneutral)wereused.Theruntimewas150minfol-lowedbya40minpreventativedecontaminationprogramme.Themono-orthoPCBswerecollectedin120mlisohexane/dichloromethane(1:1v/v)andthePCDD/Fandnon-orthoPCBsin120mloftoluene.Thesefractionswereevaporatedto~5mlbyrotaryevaporationinaheatedwaterbathat25°Cand320mbarpressureformono-orthoPCBsand40°Cand70mbarforPCDD/Fandnon-orthoPCBs.Themono-orthoPCBfractionwasdriedunderN2atroomtemperature

and5mleachofisohexaneandconcentratedsulphuricacidaddedfol-lowedbyvortexmixingandcentrifugationat478gfor5min.Theor-ganicphasewasevaporatedto~1mlunderN2beforeadditionofisooctaneandreducedto500μl.FinallythesamplewastransferredtoaconicalGCvialcontaining150μlofnonaneandevaporatedto100μlbeforeanalysisbyGC–MS/MS.

ThePCDD/Fandnon-orthofractionwasevaporatedto500μlunderN2,transferredtoaconicalGCvialwith60μlofnonanebeforeevaporatingto10μlandadding400μlisohexaneand10μlrecoverystandard(EPA-1613-ISSforPCDD/Fand68A-ISforPCB:WellingtonLaboratories,Guelph,Canada)followedbyvortexmixing.Sulphuricacid(35μl)wasadded,vortexedandleftfor1hbeforecentrifuging(5min,1328g).TheorganicphasewastransferredtoacleanconicalGCautosamplervialcontaining10μlnonaneaskeeperandevaporat-edto10μlor50μlbeforeanalysisbyGC–MS/MSforPCDD/Fornon-orthoPCBanalysis,respectively.

Analysisofthe29PCDD/FandDL-PCBcongenershavingWHOTEF1998valueswasconductedusingGC–MS/MSonaTraceGC2000coupledtoaPolarisQiontrapMS/MS(ThermoFisher,HemelHemp-stead,Herts,UK).ThechromatographicseparationsutilisedanRxi-5®msfusedsilicacolumn(ThamesRestekUKLtd,Saunderton,Bucks,UK),30m×0.25mmi.d.×0.25mm lmthickness,withHeascarriergas.Theinjectortemperaturewas250°Candsamplesandstandardswereinjectedinsplitlesswithsurgemode.MSconditionswereinthepositiveelectronionisation(EI+)modeusingautomaticgaincontrolwithelectronenergyof70eVandemissioncurrentof250μA.Thetransferlineandionsourcewerekeptat305and250°C,respectively.Xcalibarversion1.3software(XcalibarSoftwareInc.,Herndon,VA,USA)wasusedfordataacquisitionandprocessingresults.Thelimitofquanti cation(LOQ)wasdeterminedbyusing9×thebackgroundlevel(threetimesthelimitofdetection;LOD)intheblanksofeachcongener.Proceduralblankswererunforeachbatchofupto10sam-ples.Eachbatchofsampleswasrunwithoneproceduralblank,oneexternalqualitycontrol(QC)sampleandoneinternalQCsample.TheexternalQCwascodliveroil(FoodAnalysisPerformanceAssess-mentScheme®;FERA,York,UK:TO623)andtheinternalQCwasin-ternallyvalidatedsalmon esh.Percentagerecoverieswereintherangeof78–114%.Congenerquanti cationwasbasedontheisotopedilutionmethodoftheUnitedStatesEnvironmentalProtectionAgen-cy(USEPA)methods1613and1668(USAEPA,1994,1999).TheLOQrangefor sh eshsampleswas0.03–0.18pg/gwetweightforPCDD/Fsand0.03–0.06pg/gwetweightfornon-orthoandmono-orthoPCBs.Thecongenersanalysedincludedthe17PCDD/Fsand12DL-PCBsassignedWHOestablishedTEFvaluesin1997(VandenBergetal.,1998)whichwererecentlyre-evaluated(VandenBergetal.,2006;Table2).

2.4.Analysisofpolybrominateddiphenylethers(PBDEs)

ForthePBDEanalysis25gofhomogenisedwet eshwasfreeze-driedasdescribedforPCCD/Fs.Thedried eshor10gdietwasgroundinacoffeegrinderand2.5g eshordietplacedintheASE extractioncellcontaining19gsilicagel–sulphuricacidmixture(1:1w/w)andhydromatrixbeforeadditionof20μlofa500ng/mlPBDE119internalstandard(WellingtonLaboratories,Guelph,Canada).SampleswereextractedbyASE for15min×twocycleswithdichloromethane–iso-hexane(4:1,v/v)underpressure(1500psi;10,342kPa)andtempera-ture(40°C).TheASE extractwasreducedto~5mlbyrotaryevaporationat30°Cand320mbarandthenreducedto1mlunderN2atroomtemperature.Iso-hexane(5ml)andconcentratedsulphuricacid(2ml)wereadded,vortexedandcentrifuged(15minat1328g).Theorganicphasewastransferredtoacleantubeandtheprocessre-peated.Thecombinedorganicphaseswerereducedto1mlunderN2and100μlconcentratedsulphuricacidadded,mixedandcentrifuged(4minat212g).TheisohexanelayerwasthentransferredtoasilanisedGCautosamplervialcontaining500μlnonaneasthekeeperandthe

J.G.Belletal./Aquaculture324–325(2012)118–126

121

Table2

Dietaryconcentrations(ngtoxicequivalents/kgwetweight)ofdioxinanddioxin-likepolychlorinatedbiphenyl(PCB)congenersinthe3and9mmgrowerdietsand9mm nishingfeeds.Dietcongener

3mm3mm9mm9mm9mm nishingFOVOFOVOfeed2378-TCDD0.2640.0760.0580.0460.03112378-PeCDD0.1360.0690.0640.0610.031123478-HxCDD0.0050.0020.0040.0030.006123678-HxCDD0.0610.0110.0070.0070.006123789-HxCDD0.0100.0010.0280.0020.0041234678-HpCDD0.0060.0040.0010.0010.001OCDD

ndndndndndSumPCDD0.4820.1630.1620.1200.0792378-TCDF0.8090.2300.0780.0420.02112378PeCDF0.0810.0210.0060.0040.00323478-PeCDF0.2200.0420.1940.0490.038123478-HxCDF0.0340.0140.0050.0030.003123678-HxCDF0.0450.0140.0040.0040.005234678-HxCDF0.0350.0140.0050.0040.005123789-HxCDF0.0250.0080.0020.0020.0031234678-HpCDF0.0110.0090.0010.001nd1234789-HpCDF0.001ndndndndOCDF

ndndndndndSumPCDF1.2310.3520.2950.1090.078PCB770.0060.0030.0070.0020.002PCB81ndndndndndPCB1264.4541.5531.9990.0960.589PCB169

0.0860.0220.0460.0010.009Sumnon-orthoPCBs4.5461.5782.0500.0990.601PCB1050.1860.0520.1090.0130.039PCB1140.0670.0220.0230.0040.019PCB1180.5700.1760.3260.0420.120PCB1230.0170.0020.0100.0020.008PCB1560.2860.0740.1620.0170.064PCB1570.0760.0220.0490.0060.017PCB1670.0040.0010.002nd0.001PCB189

0.0060.0010.003nd0.002Summono-orthoPCBs1.2120.3500.6840.0840.269Sumofdioxins+PCBs

7.471

2.443

3.191

0.412

1.027

TCDD,2,3,7,8tetrachlorinateddibenzodioxin;PCDD,polychlorinateddibenzo-p-diox-in;nd,notdetected.TCDF,2,3,7,8tetrachlorinateddibenzofuran;PCDF,polychlori-nateddibenzofurans.

solventevaporatedunderN2tothenonanemarkbeforeanalysisbyGC/MS.

Analysisofthe7PBDEcongeners(28,47,99,100,153,154and183)wasconductedusingaThermoFisherTraceGCUltraequippedwithaZB5-MScolumn(30×0.25mminternaldiameter×0.25μmphase:Phe-nomenex,Maccles eld,Cheshire,UK)usingHeasthecarriergascoupledtoaTraceDSQMS(Bremen,Germany)inthenegativechemi-calion(Cl-)modeandmethaneasreagentgas(2.0ml/min).TheMSwasoperatedinselectiveionmonitoringmodeatm/z79and81.Proce-duralblankswereasdescribedforPCDD/F+Dl-PCBsabove.Theexter-nalQCwassuppliedbytheNorwegianInstituteforPublicHealthandtheinternalQCwasinternallyvalidatedsalmon esh.Xcalibarversion1.4software(XcalibarSoftware,Inc.,Herndon,VA,USA)wasusedfordataacquisitionandprocessingofresults.Therangeoflimitsofquanti- cation(LOQ)forwhole shsampleswas2–63pg/gwetweightforPBDEs.AsPBDE183wasnotaboveLOQinanydietsor eshsamplesthesumPBDEvaluesarebasedontheother6congenersonly.

2.5.Proximatecompositionsofdiets

Moisturewasdeterminedbythermaldryingtoconstantweightat110°Cfor24h.CrudeproteincontentwasdeterminedbyKjeldhalanalysis(nitrogen×6.25:KjeltecAutoanalyser,Tecator,Hoganas,Swe-den).CrudefatwasdeterminedfollowingacidhydrolysisusingaSoxtecSystem1047hydrolysingunit(Tecatorapplicationnote67/83).Theashcontentwasdeterminedbydryashinginaporcelaincrucibleinamuf e

furnaceat600°Covernight.AllmethodsarebasedonthosedescribedintheAssociationofOf cialAnalyticalChemists(AOAC,2000)andmodi edasdescribedinBelletal.(2001).2.6.Lipidextractionandfattyacidanalysis

Totallipidswereextractedfrom1to2gofpooledmuscleorliverbyhomogenisingin20volumesofchloroform/methanol(2:1v/v)inanUltraTurraxtissuedisrupter(FisherScienti c,Loughborough,UK).TotallipidswerepreparedaccordingtothemethodofFolchetal.(1957)andnon-lipidimpuritieswereremovedbywashingwith0.88%(w/v)KCl.Theweightoflipidswasdeterminedgravimetricallyafterevaporationofsolventandovernightdesiccationundervacuum.Fattyacidmethylesters(FAME)werepreparedbyacid-catalysedtrans-esteri cationoftotallipidsaccordingtothemethodofChristie(2003).Extractionandpuri cationofFAMEwerepreformedasdescribedbyGhionietal.(1996).FAMEwereseparatedbygas–liquidchromatogra-phyusingaThermoFisherTraceGC2000(ThermoFisher,HemelHemp-stead,UK)equippedwithafusedsilicacapillarycolumn(ZBWax,30m×0.32×0.25mmi.d.;Phenomenex,Maccles eld,UK)withhydro-genascarriergasandusingon-columninjection.Thetemperaturegra-dientwasfrom50to150°Cat40°C/minandthento195°Cat1.5°C/minand nallyto220°Cat2°C/min.Individualmethylesterswereidenti edbyreferencetopublisheddata(Ackman,1980).DatawerecollectedandprocessedusingtheChromcardforWindows(version2.00)computerpackage(ThermoquestItaliaS.p.A.,Milan,Italy).2.7.Statisticalanalysis

Thesigni canceofdifferences(Pb0.05)wasdeterminedbyonewayANOVA(initial sh)orbytwo-wayANOVA.Theseanalyseswereusedtodeterminetheeffectsofdiet, shgroupandtheirinter-actionsoncontaminantconcentrationsinsalmon esh.Statisticalan-alyseswereperformedusingGraphpadPrism (version4.0)statisticalpackage(GraphpadSoftware,SanDiego,CA).3.Results

Therewasasigni canteffectof shgroup,andaninteractionbe-tweengroupanddiet,on sh nalweightswiththelowestweightsinCALFOandVO shandtheFatVO shwhilethelargest shweretheLeanandFatFO sh.Theseresultshavebeenreportedindetailprevi-ously(Belletal.,2010).Thelower nalweightintheCAL shmayhavebeenduetothelowerstartingweightofsmoltsbeing52,85and88gfortheCALandLeanandFatgroups,respectively.However,SGRsrangedfrom0.89,inFatVO sh,to1.01inCALVO shand shwithhighergrowthratesmayresultinadilutioneffectthatreducesPOPsloading.

Thefeedsusedinthepresentstudycontainedlowerlevelsofdietary shmealcomparedtocommercialdietsusedforScottishsalmonin2006andrepresented25%ofthedietformulationcomparedtothenormatthattimeof~40%.ToreplacetheFMcomponent,ablendofpro-teinsourcesderivedfromplantproducts,representing40–44%ofthedietformulation,wasusedtoful lsalmonproteinrequirements(Table1).TheoilsourceusedwaseithermarineFOoraVOblendcom-prisingrapeseed,palmandCamelinaoils.POPsbeinglipidsolubleandwidelydistributedinthemarineecosystemtendtoaccumulateupthemarinefoodchainand,thus,theFO'sderivedfromsmallpelagic sh,thatareutilisedinaquafeeds,areamajorsourceofPOPs(Berntssenetal.,2007;Jacobsetal.,2002).The3and6mmFOgrowerfeedswereformulatedtocontainPOPslevelsintheupperrangeoftheper-mittedEUvalues(EC,2006b)whilethe9mmgrowerand9mm nish-ingfeedshadlowerlevelsofPOPsduetotheuseofadifferentFO.ThisallowedustofollowtheimpactofthedifferentdietaryPOPconcentra-tionsonthe nal shproduced,overthe79weektrial.Thus,the3mmFOfeedhadthehighestsumPCCD/Fs+DL-PCBsof7.47ngTEQ/kg(EU

122J.G.Belletal./Aquaculture324–325(2012)118–126

Table3

Dietaryconcentrations(ng/gwetweight)ofpolybrominateddiphenylethers(PBDE)congenersinthe6and9mmgrowerdietsand9mm nishingfeeds.DietcongenerPBDE28PBDE47PBDE99PBDE100PBDE153PBDE154SumPBDE

6mmFO0.254.050.941.000.300.176.71

6mmVO0.0080.190.110.0280.0210.0280.385

9mmFO0.122.510.530.550.0650.203.98

9mmVO0.0110.280.130.0440.0240.0320.521

9mm nishingdiet0.0521.320.250.280.0300.0772.01

nd,notdetected.PBDE183wasnotdetectedinanysamples.

permittedmaximumbeing7ngTEQ/kg).Bycomparison,the9mmFOfeed,whichuseddifferentFOandFMsourcestothe3and6mmfeeds,hadconsiderablylowerPOPcontentof3.19ngTEQ/kg(Table2)whichis57%lowerthanthe3mmfeed.InbothdietsthebiggestcontributiontothesesumvalueswasderivedfromtheDL-PCBsbeing5.76and2.73ngTEQforthe3and9mmFOfeeds,respectively.The9mm nish-ingfeed,whichhadbeensubjectedtoanactivatedcarbondecontami-nationtreatment,hadsumPCCD/Fs+DL-PCBsof1.03ngTEQ/kg,ofwhich0.87ngTEQ/kgwasderivedfromDL-PCBs.Thesumvalueforthe9mm nishingfeedwas86%and68%lowerthanthe3and9mmgrowerFOfeeds,respectively.Thesumof6PBDEcongenersinthe6and9mmFOfeedsfollowedthesamepatternasforthePCCD/Fs+DL-PCBsbeinghighestinthe6mm(6.71ng/g)andlowerinthe9mmfeed(3.98ng/g;Table2).Bycomparison,the3and9mmVOfeeds,whichduetotimeofmanufactureuseddifferentVOsources,hadlowerlevelsofsumPCCD/Fs+DL-PCBsof2.44and0.412ngTEQ/kg,respectively,comparedtotheFOdiets.ThesumPBDEconcentrationsinthe6and9mmVOdietswere0.385and0.521,respectively.The9mm nishingfeedhadasumPBDE

concentrationof2.01ng/gwhichwaslowerthanthetwoFOdietsbuthigherthanthetwoVOdiets(Table3).Inalldiets,includingthe9mm nishingdiet,themajorPBDEcongenerswerePBDE47,100and99.

Thesumof27PCDD/F+DL-PCBconcentrationsintheinitialLean,FatandCAL shgroupsatthestartofthetrialwere0.454±0.054,0.481±0.074and0.612±0.035ngTEQ/kg,respectively(n=3).ThehighervalueintheCAL shwasduetothese shbeingproducedbyMarineHarvestwhiletheFatandLean shwereproducedbyLandcatchwhichwerefeddifferentdietsinfreshwaterresultingindifferentPOPspro les.Theinitialsumof7PBDEconcentrationswere0.403±0.061,0.502±0.072and0.710±0.011ng/g,respective-ly(n=3).Afterfeedingthethreetestdietsoverafullseawaterpro-ductioncycleof55weeksthesalmon eshPCDD/Fs+DL-PCBsinthethree shgroupsfedFOincreasedto1.93–1.95ngTEQ/kgwhilethevaluesin shgroupsfedVOremainedatvaluessimilartoorlowerthantheinitialvaluesi.e.0.442–0.475ngTEQ/kg.Thus,dietwasthemainfactordetermining eshPCCD/F+DL-PCBvaluesbeingsigni cantlydifferentinallcongenerswiththeexceptionof123789-HxCDDand123789-HxCDF(Table4).Whiletherewassomeevidencethat shgroupaffecteddepositionofsomecongenersthisdidnotreachsigni canceandtherewasnointeractionbetweendietand shgroups.

Forthe6PDBEsmeasuredin nalsalmon eshtheorderofthemajorcongenersfollowedthepatternseenindietnamelyPBDE47>100>99inthe shfedFOand47>99>100in http://wendang.chazidian.com-paredtothesumPBDEconcentrationsininitial shtherewasa5-foldincreaseinthisvaluein shfedtheFOdietfor55weekswhilevaluesinthe shfedVOweresimilartotheinitialsmoltvalues.ConsistentlyhighermeanPBDEconcentrationswereseenforall6congenersintheLNSFatgroupfedtheFOdiets,comparedtotheLNSLeanandCALgroups,andtoalesserextentintheFatgroupfedtheVOdiets,although

Table4

Fleshconcentrations(ngTEQ/kgwetweight)ofdioxin(PCDD/F)anddioxin-likepolychlorinatedbiphenyl(DL-PCBs)congenersinsalmon(Salmosalar)afterfeedingeitherthe shoilandvegetableoilblenddietsfor55weeks.FishgroupanddietCongener

2378-TCDD12378-PeCDD123478-HxCDD123678-HxCDD123789-HxCDD1234678-HpCDDSumPCDD2378-TCDF12378PeCDF23478-PeCDF123478-HxCDF123678-HxCDF234678-HxCDF123789-HxCDF1234678-HpCDF1234789-HpCDFSumPCDFPCB77PCB81PCB126PCB169

Sumnon-orthoPCBsPCB105PCB114PCB118PCB123PCB156PCB157PCB167PCB189

Summono-orthoPCBSumdioxins+PCBs

LeanFOMean±SD0.037±0.0110.042±0.0060.002±0.0000.003±0.0010.002±0.000nd

0.086±0.0140.094±0.0070.005±0.0010.115±0.0100.002±0.0010.002±0.0010.003±0.0010.002±0.000ndnd

0.223±0.0090.003±0.000nd

1.087±0.0330.020±0.0011.110±0.0330.075±0.0020.024±0.0010.229±0.0060.011±0.0010.131±0.0070.034±0.0020.002±0.0000.003±0.0000.509±0.0171.928±0.041

FatFOMean±SD0.027±0.0080.037±0.0160.002±0.0010.003±0.0010.002±0.001nd

0.071±0.0150.087±0.0020.006±0.0010.112±0.0090.002±0.0000.002±0.0000.003±0.0010.001±0.000ndnd

0.212±0.0100.003±0.000nd

1.113±0.0460.019±0.0001.136±0.0460.078±0.0030.024±0.0010.234±0.0140.012±0.0010.136±0.0050.036±0.0020.002±0.0000.003±0.0000.525±0.0241.944±0.074

CALFOMean±SD0.041±0.0230.044±0.0010.001±0.0000.004±0.0010.002±0.001nd

0.092±0.0240.087±0.0080.004±0.0010.114±0.0190.002±0.0010.002±0.0010.003±0.0010.002±0.001ndnd

0.215±0.0240.003±0.000nd

1.097±0.0290.019±0.0011.120±0.0280.077±0.0000.024±0.0010.234±0.0030.012±0.0020.136±0.0020.034±0.0010.002±0.0000.003±0.0000.523±0.0051.950±0.040

LeanVOMean±SD0.023±0.0080.018±0.0070.001±0.0000.002±0.0010.001±0.000nd

0.046±0.0150.038±0.0030.002±0.0010.019±0.0050.001±0.0000.001±0.0010.001±0.0000.001±0.001ndnd

0.064±0.0050.001±0.000nd

0.234±0.0030.005±0.0000.239±0.0030.014±0.0010.007±0.0000.042±0.0020.002±0.0000.022±0.0010.006±0.000ndnd

0.093±0.0050.442±0.011

FatVOMean±SD0.022±0.0080.018±0.0030.001±0.0010.002±0.0000.001±0.000nd

0.044±0.0060.036±0.0040.002±0.0010.025±0.0040.001±0.0010.001±0.0010.002±0.0010.001±0.000ndnd

0.069±0.0080.001±0.000nd

0.260±0.0260.005±0.0010.265±0.0260.015±0.0010.007±0.0000.043±0.0040.002±0.0000.022±0.0020.006±0.000ndnd

0.097±0.0080.475±0.048

CALVOMean±SD0.013±0.0020.012±0.0040.001±0.0000.002±0.0000.002±0.001nd

0.031±0.0040.035±0.0010.002±0.0010.023±0.0040.001±0.0000.001±0.0000.002±0.0010.002±0.002ndnd

0.065±0.0040.001±0.000nd

0.252±0.0320.004±0.0010.257±0.0330.015±0.0020.007±0.0010.045±0.0060.002±0.0000.023±0.0040.006±0.001ndnd

0.098±0.0130.451±0.051

Two-wayANOVADiet0.0161b0.00010.03060.00470.0687b0.0001b0.0001b0.0001b0.00010.00320.02330.00920.493

Group0.7300.8440.2620.4930.3960.6120.1900.2620.9561.0001.0000.5650.262

Interaction0.2780.3790.2620.4930.3960.1680.6540.2620.7421.0001.0000.5650.619

b0.0001b0.0001b0.0001b0.0001b0.0001b0.0001b0.0001b0.0001b0.0001b0.0001b0.0001

0.8621.0000.3120.1480.3780.1711.0000.6040.7230.4180.335

0.5291.0000.9600.4930.9670.6341.0000.8870.7230.5460.335

b0.0001b0.00010.3980.6780.7420.864