Predictionoffeedintakebythesalmonsmoltwasestimated accordingtofishsizeandwatertemperatureusingalgorithms pro-videdbythefeedsupplier.Thedayafterbeginningeachtreatment, theaverageweightandn
Trang 1j ou rn a l h o m epa g e : w w w e l s e v i e r c o m / l o c a t e / a q u a - o n l i n e
a The Conservation Fund Freshwater Institute, 1098 Turner Road, Shepherdstown WV 25443, USA
b Nofima, NO-6600 Sunndalsøra, Norway
c Faculty for Agricultural and Environmental Sciences, University of Rostock, Jusus-von-Liebig-Weg 6, 18059 Rostock, Germany
d AVS Chile SA, Imperial 0655, Of 3A, Puerto Varas, Chile
a r t i c l e i n f o
Article history:
Available online 24 November 2014
Keywords:
RAS
Salmon
Alkalinity
Nitrification
CO 2
Biofilter
a b s t r a c t Whenoperatingwaterrecirculatingsystems(RAS)withhighmake-upwaterflushingratesinlocations thathavelowalkalinityintherawwater,suchasNorway,knowledgeabouttherequiredRAS alkalin-ityconcentrationisimportant.FlushingRASwithmake-upwatercontaininglowalkalinitywashesout valuablebaseaddedtotheRAS(asbicarbonate,hydroxide,orcarbonate),whichincreasesfarm operat-ingcostswhenhighalkalinityconcentrationsaremaintained;however,alkalinitymustnotbesolow thatitinterfereswithnitrificationorpHstability.Forthesereasons,astudywasdesignedtoevaluate theeffectsofalkalinityonbiofilterperformance,andCO2strippingduringcascadeaeration,withintwo replicatesemi-commercialscaleAtlanticsalmonsmoltRASoperatedwithmovingbedbiologicalfilters Alkalinitytreatmentsofnominal10,70,and200mg/LasCaCO3weremaintainedusingapHcontroller andchemicaldosingpumpssupplyingsodiumbicarbonate(NaHCO3).Eachofthethreetreatmentswas replicatedthreetimesineachRAS.BothRASwereoperatedateachtreatmentlevelfor2weeks;water qualitysamplingwasconductedattheendofthesecondweek.Aconstantfeedingof23kg/day/RASwas providedevery1–2h,andcontinuouslighting,whichminimizeddiurnalfluctuationsinwaterquality RAShydraulicretentiontimeandwatertemperaturewere4.3daysand12.5±0.5◦C,respectively,typical
ofsmoltproductionRASinNorway
Itwasfoundthatalownominalalkalinity(10mg/LasCaCO3)ledtoasignificantlyhighersteady-state TANconcentration,comparedtowhen70or200mg/Lalkalinitywasused.Themeanarealnitrification ratewashigheratthelowestalkalinity;however,themeanTANremovalefficiencyacrosstheMBBRwas notsignificantlyaffectedbyalkalinitytreatment.TheCO2strippingefficiencyshowedonlyatendency towardshigherefficiencyatthelowestalkalinity.Incontrast,therelativefractionoftotalinorganiccarbon
comparedtothehigheralkalinities(70and200mg/LasCaCO3).Despitethis,whencalculatingthetotal lossofinorganiccarbonfromRAS,itwasfoundthatthedailylosswasaboutequalat10,and70mg/L, whereasitwashighestat200mg/Lalkalinity.pHrecordingsdemonstratedthatthe10mg/Lalkalinity treatmentresultedinthelowestsystempH,thelargestincreasein[H+]acrossthefishculturetanks,as wellasgivinglittleresponsetimeincaseofalkalinitydosingmalfunction.RapidpHchangesunderthe relativelyacidicconditionsat10mg/Lalkalinitymayultimatelycreatefishhealthissuesduetoe.g.CO2or
ifaluminiumorothermetalsarepresent.Inconclusion,Atlanticsalmonsmoltproducersusingsoftwater make-upsourcesshouldaimfor70mg/Lalkalinityconsideringtherelativelylowlossofinorganiccarbon comparedto200mg/Lalkalinity,andtheincreasedpHstabilityaswellasreducedTANconcentration, comparedtoloweralkalinityconcentrations
©2014TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense
(http://creativecommons.org/licenses/by/3.0/)
Abbreviations: HRT, hydraulic retention time; MBBR, moving bed bioreactor; TSS, total suspended solids; TAN, total ammonia nitrogen; TIC, total inorganic carbon; RAS, tecirculating aquaculture system.
∗ Corresponding author Tel.: +1 304 870 2211; fax: +1 304 870 2208.
E-mail address: s.summerfelt@freshwaterinstitute.org (S.T Summerfelt).
http://dx.doi.org/10.1016/j.aquaeng.2014.11.002
0144-8609/© 2014 The Authors Published by Elsevier B.V This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/3.0/ ).
Trang 21 Introduction
Waterrecirculatingaquaculturesystems(RAS)areincreasingly
used toproduce Atlantic salmonsmolt (Bergheim et al., 2009;
Dalsgaardetal.,2013;Kolarevicetal.,2014).Thesesystemsare
oftenintensive, operating with low systemflushing rates, high
stockingdensities,pureoxygensupplementation,biofiltrationto
removeammonia,and variousformsofaerationtoremove
dis-solvedcarbondioxide(CO2).Pureoxygensupplementationisused
tosupporthigher feedloads andincreasedfishproductionin a
given RAS, but this alsocreate conditions where elevated
lev-elsof dissolved CO2 canaccumulate if inadequate air-to-water
contacting is not provided (Summerfelt et al., 2000) Fish can
sense and will avoid areas of high dissolved CO2 (Clingerman
etal., 2007), when possible.However,chronicexposureto
ele-vatedconcentrationsofdissolvedCO2 hasbeenassociated with
reducedgrowth(Danleyetal.,2005;Fivelstadetal.,2007),reduced
conditionfactor(Fivelstadetal.,1998,2003a,2003b),and
nephro-calcinosis (Landolt, 1975; Fivelstad et al., 1999; Hosfeld et al.,
2008)insalmonids.Inaddition,freeacidproducedduring
nitri-ficationreactswithbicarbonatealkalinityinthewaterreleasing
morecarbondioxidethantheautotrophicnitrifyingbacteria
con-sume(U.S.EPA,1975).Theconversionoftotalammonianitrogen
(TAN) tonitratenitrogen(NO3–N)withnitrifyingbacteria
con-sumesapproximately0.15–0.19kgsodiumbicarbonate(NaHCO3)
forevery1kgoffeedconsumedbythefish(Davidsonetal.,2011)
And,ifthisalkalinitylossisnotcompensatedforby
supplemen-tationwitha base(such assodium hydroxide orNaHCO3),the
alkalinityandpHofthesystemwilldecrease(LoylessandMalone,
1997).The loss of alkalinityand the increase of dissolved CO2
arebothconditionsthatreducethepHoftherecirculatingwater
accordingtoacid–baseequilibriumofthecarbonatesystem(e.g
Loyless andMalone,1997; Colt,2006).Asstockingdensity and
systemhydraulicretentiontimeinRAShaveincreasedinrecent
years,applicationoftechnologiestocontrolalkalinity,pH,and
dis-solvedCO2havebecomesignificantlymoreimportant.Preventing
largedropsinpHcanbecriticaltopreventsolubilizingmetals,such
asaluminium,becauseoftheirtoxiceffectonfish(Skogheimand
Rosseland,1986;Fivelstadetal.,2003b)
Carbondioxideisexcreted(alongwithammonia)throughthe
fish’sgillsinproportiontoitsfeedandoxygenconsumptionrate
AconsiderableamountofdissolvedCO2 isalsoproducedinthe
biofilter(SummerfeltandSharrer,2004).ControllingdissolvedCO2
fromaccumulatingtodetrimentallevelsisparticularlyimportant
in fishfarmsthat useintensivewater recyclingsystems These
systemsuseoxygenationunitstocreatehighlevelsofavailable
dis-solvedoxygenintheculturetanks,butoxygenationunitsprovide
insufficientgasexchangetostripmuchdissolvedCO2.Inaddition,
theconcentrationofdissolvedCO2 produced withintheculture
tankcanbequitelargewhenpureoxygenationisused,withup
8–12mg/L of dissolved CO2 produced in a single pass at high
stockingdensities.DissolvedCO2 isstrippedfromthe
recirculat-ingwater,typicallyafterthebiofilterandbeforetheoxygenation
process(Summerfeltetal.,2000;SummerfeltandSharrer,2004)
DissolvedCO2strippingisbasedontheprinciplethatthepartial
pressureofCO2inaircontactedwithwaterislessthanthepartial
pressureoftheCO2dissolvedinthewater.ThedissolvedCO2
there-forecomesoutofsolutionandisstrippedoffasagas.Increasing
thevolumeofairflowthatiscontactedwiththewaterflowwill
increasethedissolvedCO2thatcanberemoved
Maintaining adequate alkalinity concentrations has been
reportedtobecriticalforsustainingnitrification,i.e.,the
wastewa-terliteraturereportsthat40–80mg/L(asCaCO3)istheminimum
alkalinityrequiredtosupportnitrification(Paz,1984;Biesterfeld
et al.,2003).Villaverde et al (1997)reported a linear increase
in nitrification efficiency of 13% per unit pHincrease frompH
5.0 to8.5 Rusten etal (2006)foundthat thenitrification rate droppedtoonlyhalfoftheoriginalratewhenalkalinitydropped fromapproximately115mg/LasCaCO3 (pH7.3)to57mg/L(pH 6.7)inabench-scaleexperiment performedusingbiofilm carri-erscollectedfromaturbotfarm’smovingbedbiologicalreactor (MBBR).Moreover,Colt(2006)warnsthatthenitrificationprocess slowsdownatlowpHandChenetal.(2006)recommend main-taininganalkalinityof200mg/LasCaCO3tosupportnitrification whenwaterexchangerateisminimal.RASoperatedatsuboptimal alkalinitycouldtheoreticallyencounterlargerpHswings,higher concentrationsofTANandNO2–Nifnitrificationefficiencydrops, andmicrobialcommunityinstability(Mydlandetal.,2010),which maybeharmfultothefish.However,theconsequencesof oper-atingaRASwithoutadequatealkalinityhavebeenlittlestudied, particularlyforsystemsusedtoproduceAtlanticsalmon.Thisis
aspecieswhichissensitivetoelevatedconcentrationsofnitrite nitrogenwithoutconcurrentchlorideadjustments(Gutierrezetal.,
2011),torelativelylowlevelsofNH3–N(Kolarevicetal.,2013),and
CO2 (Fivelstad,2013).Researchisneededtodetermineif main-taininganalkalinityof80–200mg/LasCaCO3 isreallybeneficial, becauseoperatingathighalkalinitylevelswillincreasethecostof supplementationwithbase.Interestingly,highnitrificationratesat lowpHandalkalinityhavebeenreportedpreviouslyinlaboratory scalereactorexperiments(TarreandGreen,2004).When operat-ingRASwithhighmake-upwaterflushingratesinlocationssuchas Norwaythathavelowalkalinityintherawwater(Kristensenetal.,
2009),knowledgeaboutrequiredRASalkalinitywillbe particu-larlyimportant.Furthermore,sinceithasrecentlybeenproposed thatlargertankscalesincreaseperformanceinsalmon(Espmark
etal.,2013),wewantedtostudyeffectsofalkalinityinlargerscale RAS.Forthesereasons,astudywasdesignedtoevaluatetheeffects
ofalkalinityonCO2strippingduringcascadeaeration,plus biofil-terperformancewithinsalmonsmoltsemi-commercialscaleRAS operatedwithmovingbedbiologicalfilters.Thegoalofthestudy wastotestthefollowinghypotheses,thatincreasingalkalinity con-centrationsfrom10to200mg/LinaRASwill(1)stabilizesystem
pH,(2)decreaseNO2–NandTANconcentrationsandvariability, and(3)increaseTANremovalefficiencyandremovalrateacross theMBBR.Finally,wehypothesizethat(4)ahigheralkalinitywill decreaseCO2removalefficiency,andincreasetotalinorganic car-bon(TIC)removal,acrossforced-ventilationcascadedegassersand thusleadtoelevatedcostsassociatedwithbicarbonatedosing
2 Materials and methods
ThestudieswereconductedattheNofimaCentrefor Recircula-tioninAquacultureatSunndalsøra,Norway,describedinTerjesen
etal.(2013) 2.1 Experimentaltreatments Alkalinityoftherecirculatingwaterwasmaintainedatthree treatmentlevels,i.e.,atnominal10,70,and200mg/LasCaCO3, usinganonlinepHelectrode(Sensorex8000CD-pHwithsolution groundandamplifier,GoldenGrove,USA)locatedinthesumpat thebaseoftheCO2strippingcolumn.EachpHprobewasequipped withanautomaticcleaningsystem;a waterjetprogrammedto flush the probe each tenth minute (Storvik Aqua, Sunndalsøra, Norway).Thisautomaticcleaningwasfoundtobeclearlynecessary
tomaintainstablepHduringthe14daysofeachtreatmentreplicate (Kolarevi ´cetal.,2011).ThepHprobeswerecalibratedusingMerck two-pointbuffers(Merck,Darmnstadt,Germany),eachweek,right aftera treatmentperiod ended, andafter sevendays intoeach treatment.AdedicatedpHcontroller(ModelWPD320,Walchem, Holliston,MA)wasusedtocontrolachemicaldosingpump(Model
Trang 3tomaintaina pHset-point in each RAS Ground well water at
Nofimain Sunndalsøracontains around 6–20mg/Las CaCO3 of
alkalinity,dependingonseasonandthepumpwellinuse(Terjesen
etal.,2013)
Thus,apHcontrolsystemwasusedtocontrolthetreatment
alkalinitybecause,accordingtoacid–baseequilibriumin
freshwa-terwithlowammonialevels,waterpHisanapproximatemeasure
ofthealkalinitywhenthewatertemperatureanddissolvedCO2
concentrationremainconstant(Summerfeltetal.,2001).Changes
indissolvedCO2donotaffectalkalinity,perthedefinition,butdo
affectthewaterpH.Thus,strippingdissolvedCO2increasesthepH
ofwaterasitdecreasesthetotalinorganiccarbonconcentration,
butitdoesnotchangethealkalinityconcentration.Maintaininga
constantdissolvedCO2concentrationrequiredmaintaininga
con-stantCO2productionrateandremovalrateforagivenalkalinity
treatment.ArelativelyconstantCO2productionratewasachieved
bymaintainingacontinuousphotoperiodforthefish,constantdaily
feedrateof23kg/day/RAS,andfeedingevery1–2h, 24haday
Continuouslightingandfeeding24hdailyhavebeenfoundto
min-imizediurnalfluctuationsinwaterquality,i.e.,thechangeinTAN,
CO2,O2,andTSS,concentrationsacrosstheculturetank(Davidson
etal.,2009;Kolarevicetal.,2009).Watertemperaturewaslogged
eachfifthminuteintheCO2-degassersump,andwasmaintained
at12.8±0.4◦C(SD)(RAS1)and12.7±0.4◦C(RAS2)throughoutthe
study
EachtreatmentwasreplicatedthreetimesintworeplicateRAS,
i.e.,sixquasi-replicateswereprovidedforeachtreatment(Table1)
BothRASwereoperatedateachtreatmentlevelfor2weeks;water
qualitysamplingwasconductedattheendofeachsecondweek
Theexperimentaldesign,bothintermsofreplicatesystems,and
lengthofperiods,wasnecessary tomakethis studyat a
semi-commercialscalepossible.Thestudylasted20weeks,becausethe
firsttreatmenttestedhadtoberepeatedduetopHinstabilities
duringthestartofthetrial(Table1)
2.2 Waterrecirculatingsystems
TwoRASwereusedinthisstudy(Fig.1).EachRAScontained
56.2m3oftotalwatervolume,includingten3.3m3culturetanks,
abeltfilter(ModelSFK400with60mbeltsieveopening,
Sal-snes,Namsos,Norway),threecentrifugalpumps(1.5kW/pump,ITT
Flygt,Trondheim,Norway)tolift0.75m3/mineach,at6mhead,
froma sump followingthe beltfilter, a movingbed bioreactor
(MBBR), a forced ventilated cascade aeration column for
strip-pingCO2 fromthewaterflowingbygravity outoftheMBBR, a
pumpsumpbelowthedegasserof1.9m3volume,threecentrifugal
pumps(3kW/pump,ITTFlygt),toliftanominalflowof0.75m3/min
eachagainst12–13mheadfromasumpatthebaseofthestripping
column,andadownflowbubblecontactor(AquaOptima,
Trond-heim,Norway)toaddpureoxygengastothewaterjustbefore
itenteredeachculturetank.EachMBBRcontainedthree7.0m3
chambersthat each contained3.5m3 of media(BiofilmChip P,
KrügerKaldnes,Sandefjord,Norway)witha specificsurfacearea
of900m2/m3(manufacturersstatement)
ToensurethatthewasteloadontheMBBRapproachedmaximal
capacityatarelevantculturetankwaterqualityforAtlanticsalmon
smolt(0.2–0.7mg/LTAN)(Dalsgaardetal.,2013;Terjesenetal.,
2013),datafromTerjesenetal.(2013)wasusedtocalculatethe
requiredmediaareainthepresentstudy.Theseauthorsfoundthat
thecapacityofallthreechamberscombinedintheRAS1(orRAS2)
MBBRwas 61kgfeed/day,calculated toequal 2083gTAN
pro-duced/day,at50%feedprotein,tomaintainmaximal0.7mg/LTAN
inthereturnflowfromtheculturetanks,at14◦Cand1.9m3/min
flow.Inthepresentstudyareducedfeedloadof23kg/day/RAS,
was used Based on Terjesen et al (2013) this feed load
(47%proteininfeed,seebelow)shouldresultin700gTAN/day/RAS releasedtotheculturetankwater(nitrogenretentionof53%for Atlanticsalmonparr,Grisdale-HellandandHelland,1997).Atthe temperatureof12.8◦Cusedinthepresentstudy,thisTAN produc-tionwouldrequireone-thirdofthetotalMBBRarea(i.e.3140m2), givinga68%TANremovalto0.22mg/Lintheeffluent,andrequire
a 1.1m3/minsystemflow rate.Thus, therecirculatingwater at thisflowratewasonlypumpedthroughoneofthethreeMBBR chambers(thelast)perRASduringthepresentstudy(Fig.1) TheCO2strippingcolumnusedcontained1.6mdepthofrandom packing(2.4m3ofNor-Pacrings5cmdiam.,JaegerEnvironmental Products,Houston,TX,USA) andusedafantoventilate10 vol-umesofairforeveryonevolumeofwaterpassingcounter-current throughthecolumn.Thedownflowbubblecontactorswere con-nectedtoonecontrolcircuitandonlineO2probe(ModelDO6441, Sensorex)perculturetank.Theoxygencontrollerwasusedto main-taindissolvedoxygenintheculturetanksatbetween85%and90%
ofsaturation.AlthoughtheseRASareequippedwithsystemsfor ozonation,therecirculatingwaterwasnotozonatedduringthe study,toensurethatanynitriteaccumulationduetothealkalinity treatmentscouldoccurfreely,withoutbeingoxidizedbyozone WaterflowratethrougheachMBBRandmake-upwaterflow rate were measured using magnetic flow meters (Sitrans FM Magflo, Siemens, Munich, Germany) with continuous data log-ging(eachfifthminute).Waterflowthrougheach MBBR(RAS1:
1147±8l/min, RAS2; 1148±7l/min) and make-up water flow (RAS1:9±0l/min,RAS2;9±0l/min)wereheldconstant through-outthestudy,andwassimilarbetweentheRAS Thus,99.2%of theflowwasreusedeachpassthroughtheRAS,andthedaily sys-temwaterexchangeofeachRASwas23%,i.e.themeanhydraulic retentiontimewithineachRASwasestimatedtobe4.3day.The meanhydraulicretentiontime(HRT)throughthe3.3m3fish cul-turetanks,7m3MBBRchamber,andthe1.9m3 sumpbelowthe CO2-degasserwere29min,5.3min,and1.9min,respectively.Note,
pHwascontinuouslymonitoredimmediatelyafterthecascade col-umninthesumpfordosingsodiumbicarbonateandcontrolling pH/alkalinity.The approximately 1.9minHRT within thesump allowedthedehydrationofcarbonicacid,i.e.,theratelimitingstep witha22shalf-life,andreallocationofbicarbonateand carbon-ate(bothnearlyinstantaneous)toreturntoacid–baseequilibrium (Kern,1960;GraceandPiedrahita,1994)andachieveastablepH immediatelyfollowingCO2stripping
Thebiofiltershadbeenconnectedtotankswithfeedingsalmon parrforaleastsixmonthspriortothebeginningofthestudyto thoroughlyestablishthebiofilterswithbacteria.Tenweeksprior
totheexperiment,allpipeswerecleaned,andtheCO2degasser mediataken outandthoroughlycleaned.Thebiofiltermediain eachRASweretakenout,mixedbetweenRAS1/2toensuresimilar biofilmandMBBRmicrobiota,andthenreturnedinequalvolume (3.5m3)totheRASs.Subsequently, each RASwasstocked with similarbiomass,andmaintainedonequalfeedload(seebelow) untilstartofthetrial.AllroutineworkonbothRASwasconducted equally,andatapproximatelythesametimeeachday.Inaddition, alldailyflushing/scrubbingofsediments(i.e.,thebeltfilter)and flushingoftankswasdoneafterwaterqualitysamplinghadbeen completedfortheday
2.3 Atlanticsalmon Atlantic salmon parr of SalmoBreed strain (SalmoBreed AS, Bergen,Norway)wasused.Thefishwerestockedintofive3.3m3 culturetanksineachRASsevenweeksbeforeinitiatingdata col-lection,leavingfivetankswithwatercirculatingatsamerateas
intheothertanks,butwithnofish.Toavoidtoohighfishdensity, rightaftersamplingperiod8,halfthenumberoffishofallfivetanks weremovedintothefivepreviouslyemptytanks.Atthestartofthe
Trang 4Table 1
Experimental treatment schedule Both RAS were operated at each nominal treatment level for 2 weeks and each treatment was replicated three times in each of the two replicate RAS.
Last treatment day Treatment period Nominal RAS#1 alkalinity (as CaCO 3 ) Nominal RAS#2 alkalinity (as CaCO 3 )
a Treatment period 1 (10 and 70 mg/L alkalinity) was repeated after the final treatment, due to pH instabilities at the start of the study.
experiment(startofperiod1)eachoffivetanksperRAScontained
3935salmonparrofinitialsizeof61.1±1.6g(SD)
The salmon parr were fed commercial diets (EWOS Micro,
Bergen, Norway)with 3mm pelletsizes throughout thestudy
Whenchangingfeedlots (i.e.differentshipmentsof same feed
type),samplesweretakenandanalyzedforproximate
composi-tion,accordingtoTerjesenetal.(2013),toensurethatminorfeed
changesdidnotinfluencethesystemloads.Theaverageproximal
chemicalcomposition(%w/woffeed“asis”)was94.0±0.7(SD)dry matter,46.9±0.8crudeprotein,22.4±0.5fat,and8.3±0.1ash Predictionoffeedintakebythesalmonsmoltwasestimated accordingtofishsizeandwatertemperatureusingalgorithms pro-videdbythefeedsupplier.Thedayafterbeginningeachtreatment, theaverageweightandnumberofindividualfishpertankwere recordedinordertoquantifythebiomassoffishineachtank.Based
ontheseinputs,thenumberoffishthathadtoberemovedfrom
Legends:
Belt fi lter w/sludge discharge
Degasse r
Pump sump withpH probe
WRAS
1 or 2
Central water treatment room
hall 1 or 2 Oxygenaon
Pump sump with alka-linitysupply
Fis h tank Sidewall drain Swirl
separator w/sludge discharge
Fish tank
Fis h tank
Movi ng bed bioreactor
Make-up water
pH con trol
NaHCO3 reservoir
RAS flow Make-up flow MBBR inlet not in use
pH control data Alkalinity flow
Fig 1.Process flow drawing of the two RAS used in this study Only three fish tanks are shown, out of the 10 culture tanks that were used in each RAS during this study Refer to section 2.2 for description of components Note that the moving bed bioreactor contains three chambers, but flow was only added at the head of the chamber closest
Trang 5Table 2
Mean pH and alkalinity (±s.e.) measured in the sump below the CO 2 stripping column for the three nominal alkalinity treatments, i.e., 10, 70, and 200 mg/L Values represent analyses of samples collected each 14th day of each treatment (n = 6 per treatment).
Nominal alkalinity (as CaCO 3 ) RAS #1 alkalinity (as CaCO 3 ) RAS #2 alkalinity (as CaCO 3 ) RAS #1 pH RAS #2 pH
eachsystemtomaintainthe23kg/day/RASfeedratewas
calcu-lated.Thefishbiomasswasadjustedaccordinglythatsameday
2.4 Waterqualitymonitoring
Waterqualitysampleswerecollectedafter14daysintoeach
treatmentreplicate.Eachsamplingeventwasdoneatthesame
timeeach14thday,startingat08:00andcompletedintwohours,
at10:00AM.Water samplesfor TAN(total ammonia nitrogen),
NO2–N(nitritenitrogen),NO3–N(nitratenitrogen),TSS(total
sus-pendedsolids),CO2,alkalinity,pHandTIC(totalinorganiccarbon)
werecollectedbeforeandaftertheMBBR.Inaddition,CO2,
alkalin-ity,pHandTICwerealsosampledaftertheCO2strippingcolumn,
i.e.,inthesumpbelowtheCO2strippingcolumn.Handheldmeters
wereused forpH determination (HachHQ40D withPHC10101
electrodes,HachLange,Düsseldorf,Germany),andthesewere
two-pointcalibratedusingNBSbufferseachdayofuse.Insomecases,
dissolvedCO2wasalsoestimatedusinganOxyguardportableCO2
analyzer(Oxyguard,Birkerød,Denmark).Alkalinitywasmeasured
bytitrationaccordingtoStandard Methods(APHA,2005),using
aHACHDigitalTitratorModel16900(Hach,Loveland,Colorado,
USA),andaOrion720ApluspHmeter.TAN,NO2–N,andNO3–N
sampleswereanalyzedusinganautoanalyzer(FlowSolutionIV,OI
Analytical,CollegeStation,TX,USA),accordingtoU.S.E.P.AMethod
350.1(U.S.EPA,1983)forTANandU.S.E.P.AMethod353.2(U.S
EPA,1983)forNO3–NandNO2–N.TICwasanalyzedonfresh
sam-pleskeptonice,collectedwithsiphonsandwithoutair-bubbles
intoglassflaskswithtaperedstoppers,accordingtomethod6/93
Rev.B(PerstorpAnalytical,Perstorp,Sweden);furtherdetailsare
providedbyTerjesenetal.(2013).TICwasalsousedtocalculate
dissolvedCO2,usingpH,andtemperaturemeasuredatthesame
timeandlocationaswhencollectingwatersamplesforTIC
anal-ysis.ThecarbonatesystemconstantsinSummerfeltetal.(2001)
wereusedinthecalculations
2.5 Statisticalanalyses
Dataarepresentedasthetreatmentmeans±s.e.unless
other-wisenoted.SPSS(Chigaco,IL,USA)syntaxwrittenforrandomized
designs was used to assign alkalinity treatments randomly to
experimentalperiodandtoRAS#,withtheconstraintthatno
treat-mentwasallowedtofollowdirectlyafterasimilartreatment,in
thesameRAS(Table1).Priortostatisticalanalysesallpercentage
datawerearcsinesquareroottransformed.Analysesontheeffects
ofalkalinityweredoneusingone-wayANOVAsinSPSS.If
signifi-cant(p<0.05),Tukey’sposthoctestsweresubsequentlyappliedto
determinebetween-treatmentsignificantdifferences
3 Results and discussion
ThepHcontrollermaintainedrelativelyconstantalkalinityfor
eachtreatmentreplicate(i.e.,for10,70,and200mg/L)throughout
thestudy(Table2).Forexample,tomaintainthelowdose
alkalin-itytreatment,thepHwascontrolledat6.66±0.13and6.43±0.01
inthesumpbelowtheCO2strippingcolumninRAS#1andRAS
#2,respectively,while alkalinityinthesame locationaveraged
11±1mg/Land9±1mg/LasCaCO3,respectively.Themediumand
highdosealkalinitytreatmentsweremaintainedsimilarlyusingpH control(Table2)
Significantdifferenceswerefoundbetweenthequasi-steady stateTANconcentrationmeasuredbetweenalkalinity concentra-tionsattheMBBRinlet,MBBRoutlet,anddegassersump(Table3) The mean TAN concentration entering the MBBR was reduced from 0.65±0.08mg/L at the nominal 10mg/L alkalinity treat-mentcomparedto0.43±0.04mg/Land0.39±0.05mg/LofTANat thenominal70mg/Land200mg/Lalkalinitytreatments, respec-tively.The mean TANconcentrationexiting theMBBR dropped from 0.39±0.06mg/Lat the 10mg/L as CaCO3 alkalinity treat-mentcomparedto0.22±0.03mg/Land0.23±0.04mg/LofTANat the70mg/Land 200mg/LasCaCO3 alkalinitytreatments.These resultssuggestthat theMBBR wasabletomaintainlowerTAN concentrationswhenalkalinityconcentrationsenteringtheMBBR weremaintainedatnominal70or200mg/LasCaCO3 compared
toa sustainedalkalinityofonly10mg/LasCaCO3 alkalinity.No differencesinsteadystateTANconcentrationweredistinguished betweenalkalinitytreatmentsof70mg/Land200mg/LasCaCO3 Hence,ourhypothesisthatincreasedalkalinityreducesTAN con-centrationsinRASforAtlanticsalmonsmoltswassupportedbythe results.InRASforAtlanticsalmonsmolts,arelativelylowTAN con-centrationismaintained,evenaslowas0.2mg/L(Dalsgaardetal.,
2013).HenceinsuchRAS,theTANsubstrateconcentrationwillnot
besaturating.ItmustthereforebenotedthatinRASoperatedfor otherfishspecies,athigherintensitiesandmuchhigher steady-stateTANlevels,thealkalinityrequirementmaybedifferentfrom thatfoundinthepresentstudy
Regardingnitriteconcentrations,unlikeforTAN,nosignificant differenceswereobservedbetweentreatments,whichaveraged 0.42–0.58mg/L,andwerequitevariable(Table3).Inthesametype
ofRAS,wehaveearliershownthatalthoughTANandCO2removal capacitywashigherthananticipatedusingcommondimensioning rules,nitriteremovaldidnotmeetspecificationsbecausehigher than0.1mg/LNO2–Nwasobserved.Inastudycomparingfixedbed andmovingbedreactorsSuhrandPedersen(2010)foundno differ-encesbetweenthebioreactorsystemsinaccumulationofNO2–N Thefindingthattheexperimentaltreatmentsdidnotaffectnitrite
inthepresentstudy,suggestthatinthisset-up,nitriteremovalwas notlimitedbyalkalinity.ThemeanNO3–Nconcentrationexiting theMBBRrangedfrom40to42mg/Lforthethreetreatments.No significantchangecouldbedistinguishedeitherinNO3–N concen-trationacrosstheMBBR,whichindicatesthatequalTANconversion
toNO3–Noccurredforthethreetreatmentsaswasintendedwith theuseofequalfishfeedloadingthroughoutthestudy
ThemeanarealnitrificationrateacrosstheMBBRrangedfrom 0.09to0.14g/d/m2 (Table4).Thereweresignificantdifferences betweenthemeanarealnitrificationratesacrosstheMBBR,i.e., 0.14±0.02g/d/m2and0.09±0.02g/d/m2,calculatedatthe alka-linitytreatmentsof10mg/Land200mg/LasCaCO3,respectively
NodifferencesweresuggestedinTANremovalefficiencyacrossthe MBBRinasinglepass,withremovalefficienciesrangingfrom41
to50%removalforeachtreatment(Table4).Interestingly,these resultssuggestthatthemeanarealnitrificationrateswere signifi-cantlyhigheratanominalalkalinityof10mg/LasCaCO3compared
to200mg/LasCaCO3,whichisoppositeofourhypothesis Com-paredtopreviousstudies,thearealnitrificationratesreportedhere
Trang 6Table 3
TAN concentration (±s.e.) measured at the MBBR inlet, MBBR outlet, and degasser sump, plus NO 2 –N concentration measured at the MBBR outlet, at the low (10 mg/L), medium (70 mg/L), and high (200 mg/L) alkalinity concentrations Values represent analyses of samples collected each 14th day of each treatment (n = 6 per treatment).
Nominal alkalinity (as CaCO 3 ) TAN MBBR inlet (mg/L) TAN MBBR outlet (mg/L) TAN Degasser sump (mg/L) NO 2 –N MBBR outlet (mg/L)
a,b Treatment means not sharing a common letter are significantly different (p < 0.05).
weresimilartothearealratesmeasuredacrossthesameorlarger
MBBRsinthesamefacility,exceptinthisotherstudyallthreeMBBR
chamberswereutilized(Terjesenetal.,2013).PfeifferandWills
(2011)reportedanarealnitrificationrateapproximatelytwicethat
measuredinthepresentstudy;however,inthatstudythe
temper-aturewashigherat24–25◦C.Twofactorsmayexplainthehigher
arealremovalrateatloweralkalinity.First,weshouldnotethatthe
TANproductionrateshouldbethesameforalltreatmentsbecause
thesameamountoffeedwasconsumeddailythroughoutthestudy
Thus,thedifferenceintheMBBRarealnitrificationratesbetween
treatmentscouldbeexplainedbyTANremovalonsurfacesoutside
oftheMBBR,i.e.,nitrificationcouldhaveoccurredinbiofilmsthat
hadformedonpipe,sump,tankandCO2strippingcolumnpacking
surfaces.Thehigheralkalinitytreatment(200mg/L)mayhave
sup-portedmorenitrificationactivityoutsideoftheMBBRthanthelow
alkalinitytreatment,whichwouldexplainthelowerareal
nitrifica-tionratewithintheMBBRatthehigheralkalinity.Thishypothesis
tendstobesupportedbythehigherTANconcentrationsmaintained
duringtheloweralkalinityconditions,possiblybecauselessTAN
wasremovedonsurfacesoutsideoftheMBBRunderlowalkalinity
conditions.AsecondexplanationisalsobasedontheinletTAN
con-centrationthatwashigheratthe10mg/Lnominalalkalinity;earlier
nitrificationkineticstudiesdemonstratesthatincreasedsubstrate
concentrationresultinhigher removalrates (Chenet al.,2006;
Rustenetal.,2006),asindeedfoundinthepresentstudy
How-ever,atthelowestalkalinitytreatment,10mg/L,theMBBRoutlet
concentrationwashigher(Table3 i.e.atthislowalkalinitylevel
thesamelowoutletTANconcentrationwasnotmaintained.Thisis
indicatedalsobytheunchangedremovalefficiency(Table4 and
theeffectlikelyledtothehighersteady-stateTANlevelat10mg/L,
asdiscussedabove.Theresultsinvitetoamoredetailed
investi-gationabouttheremovalratesatdifferentinletconcentrations,
atthethreealkalinitylevels.Possibly,atabove70mg/Lalkalinity,
theremovalrateatlowerinletconcentrationsarehigher,thanat
lowalkalinity,consideringourfindingsofanunchangedefficiency
despiteasignificantlylowerinletTANconcentration.Thepresent
arealnitrificationrateswerecomparabletotheratesreportedby
Rustenetal.(2006).However,asshowninFig.8BofTerjesenetal
(2013),themeasuredremovalratesinthesametype ofRASas
inthepresentstudy,werehigheratthelowesteffluentTAN
con-centrations,thancouldbepredictedfromRustenetal.(2006).In
conclusion,futurestudiesshouldinvestigatethehypothesisthat
thisMBBRisparticularlyefficientatverylowinletTAN
concen-trations,butthatatsuchlowsubstrateconcentrationsalkalinity
shouldbeatorabove70mg/LasCaCO3
ThedissolvedCO2concentrationenteringthedegasserdidnot
differsignificantly betweenalkalinity treatments,and averaged
6.8±0.7mg/L(calculatedfromTIC),and7±0mg/L(measuredwith
probe).TheabsoluteremovalofTIC(Table5)wasgreateratthe
highestalkalinity(1.65mg/LTICremoved)thanatthetwolower
alkalinitytreatments(0.97and0.93mg/LTICremoved).Thisis
sup-portedbythefindingsof Coltetal.(2012),thatthedifferences
betweentrueCO2removal(equaltothedecreaseinTIC)andthe
apparentCO2removal(equaltothedecreaseinCO2concentration
afterre-equilibration) aresmall for lowalkalinitiesin
freshwa-terbut arelargerin highalkalinities.A tendency(p=0.11)was
Table 4
The mean areal nitrification rate (±s.e.), and mean TAN removal efficiency (±s.e.) across the MBBR for the low (10 mg/L), medium (70 mg/L), and high (200 mg/L) alkalinity treatments Values represent analyses of samples collected each 14th day
of each treatment (n = 6 per treatment).
Nominal alkalinity (as CaCO 3 )
Areal nitrification rate a
(g/d/m 2 )
TAN removal efficiency (%)
10 mg/L 0.14 ± 0.02 B 41 ± 5
70 mg/L 0.11 ± 0.01 AB 50 ± 3
200 mg/L 0.09 ± 0.02 A 43 ± 3
A,B Treatment means not sharing a common letter are significantly different (p < 0.05).
a Areal nitrification rate was calculated by multiplying the flow rate of recirculat-ing water passing through the biofilter with the change (i.e., inlet minus outlet) in the combined TAN plus NO 2 –N concentration across the biofilter; this product was then divided by the total estimated surface area of media in the biofilter.
observed however,suggesting a higherCO2 strippingefficiency
atthe10mg/Lnominalalkalinitytreatmentwhenmeasuredwith probe(Fig.2 butthiswasnotfoundwhenCO2concentrationwas calculatedfromtheTICconcentration,pH,andtemperaturedata ThemeanCO2removalefficienciesrangedfrom54to63%acrossthe 1.6mtallforced-ventilatedaerationcolumns.Thisisquite effec-tiveconsideringthatthemeanCO2inletconcentrationwasunder
7mg/L.Incomparison, Moran(2010)reports thatCO2 stripping efficiencyaveraged75–77%acrossa1.65mpackingdepth(similar
tothepresentstudy)ataninletconcentrationof10mg/L(slightly higherthanthepresentstudy)infreshwaterwhenmeasuredwith
aCO2probe.Theresultsonthemeanremovalefficiencyare com-parabletoearlierstudiesinthisRASwhichwasdesignedtoavoid
Nominal alk alinity
40 45 50 55 60 65
70
CO2 from TIC CO2 from probe
Fig 2.Mean (± s.e.) CO 2 stripping efficiency recorded across the forced ventilated cascade aeration columns at the three alkalinity treatments, i.e., 10 mg/L, 70 mg/L, and 200 mg/L as CaCO 3 Note that the white hatched bars and grey hatched bars represent stripping efficiency where CO 2 concentrations were calculated from TIC measurements or probe measurements, respectively Values represent analyses of samples collected each 14th day of each treatment (n = 6 per treatment) Strip-ping efficiency was calculated from the difference in the column inlet and outlet
Trang 7Table 5
Total inorganic carbon (TIC) loss at the three alkalinity treatments, divided into loss through nitrification, degasser, gain from make-up water, and loss through the overflow
in the degasser sump.
a Calculated using TAN removal for the three treatments ( Table 3 system flow rate, and an alkalinity consumption of 7.07 g CaCO 3 per g TAN removed in nitrification ( Chen
et al., 2006 ).
b The make-up water TIC concentration refers to measurements on samples collected after period 10.
CO2levelsabove10mg/L(Terjesenetal.,2013),concentrationsthat
mayhaveadverseeffectsonAtlanticsalmonperformance,health
andwelfare(Fivelstad,2013)
Results from the present study did not support our fourth
hypothesis, i.e., that increased alkalinity would decrease CO2
removalefficiency.WehadassumedthatdissolvedCO2 removal
efficiency would decrease across the cascade column at high
alkalinitiesbecause the CO2 concentration at the outlet of the
columnwouldbepartlyreplenishedbyashiftinacid–base
equi-librium from this largepool of carbonate.However, Colt et al
(2012)suggests that in freshwater,there is no practical
differ-ence(i.e.,<1%difference)betweentheapparentCO2removalrate
andthetrueCO2removalratewhenthealkalinityis100mg/Las
CaCO3 (2meq/L)orless.Theynotethatthedifferencein
appar-ent CO2 removal rate and the true CO2 removal rate would
beless than or equal to5%when the alkalinityis 200mg/L as
CaCO3(4meq/L).Thus,forpracticalpurposes,CO2strippingisnot
impactedbyacid–base equilibriumatthehigheralkalinity
con-centrationstested.Fromthe perspectiveof asteady-state mass
balance,this suggeststhattheconcentrationof CO2
accumulat-inginidenticalRASthatarereceivingthesamefeedrateandare
operatedatthesamewaterflowratewouldhavethesame
steady-stateCO2concentrationexitingtheculturetank(andexitingthe
biofilter),becauseCO2productionandremovalarenotimpacted
byalkalinityunderthealkalinityrangetypicallyusedinRAS.Thus,
dissolvedCO2strippingefficiencyispredictedtobeafixed
prop-ertyofadegassingunit,irrespectiveofalkalinityunderfreshwater
conditions,whenalkalinityis≤200mg/LasCaCO3
TheTICconcentrationincreasedwithalkalinity,andwas
sig-nificantlydifferent(p<0.001) betweenalltreatments,as would
beexpected.Inthedegasserinlet,TICaveraged2.5±0.8mg/Lat
10mg/Lnominalalkalinity,21.0±1.3mg/Lat70mg/Lalkalinity,
and41.3±5.1mg/Lat200mg/Lnominalalkalinity.Atthelowest
alkalinitytreatment,nominal10mg/L,therelativeCO2-fractionin
TICincreasedbecausethepHwasalsoreduced(Table2).This
dif-ferenceinTICcompositioninfluencedthefateofthecarbonwhen
passingtheCO2 degasser.WhentheRASwaterenteredtheCO2
strippingcolumnat10mg/Lalkalinity,asmuchas38%ofthe
sys-temcarbonwasremoved.Incontrast,atthetwohigheralkalinities
only4%oftheTICwaslostwhenpassingthedegasser(Fig.3)
Overalltreatments,TICremovalfromtheRASaveragedonly
0.2–0.3g/min due tonitrificationand 1.1–1.9g/mindue toCO2
stripping (Table 5) We also estimate that TIC removal due to
thenitrificationwithintheCO2strippingcolumn(probably<10%
ofnitrificationoccurringintheMBBRbasedonsurfaceareaand
concentration),wouldamounttoonly0.02–0.03g/min.Thus,loss
ofTICtonitrificationintheCO2strippingcolumnisnegligiblein practicecomparedtothelossduetoCO2stripping
TheRASpHwassignificantlyloweratanalkalinityof10mg/L thanatthetwohigheralkalinitytreatments(Table2).Itwasalso observedthatpHoscillationsweremorepronouncedatthelowest alkalinity,sincelittlebuffercapacityexistedandthedosing con-trollerthereforeproduced moreunder-and overshootingofthe set-point.Furthermore,wefoundasignificantinverselinear corre-lation(r2=0.96,p<0.001,n=20,datalog10transformed)between alkalinity and the difference between degassersump [H+] and MBBRinlet[H+].Thus,asexpectedtheH+concentrationincreased morethroughtheculturetanksatloweralkalinity.Whenswitching alkalinitytreatments,therateofthepHdeclineacceleratedfrom
200mg/Lalkalinity,to70mg/L,andwasespeciallyrapiddowntoa
pHtypicalofthe10mg/Ltreatment(Fig.4).Thistrendisanalogous
tothesituationthatwilloccurduringalkalinitydosingmalfunction Hence,ina RASoperatedatvery lowalkalinity,thetime avail-ableforreplacingthedosingequipment,beforeadverseeffectson thefishoccurfromhighCO2,willbeconsiderablylessthanwhen operatingat70or200mg/Lalkalinity
Nominal alkalinity
0 10 20 30 40
50
B
Fig 3. Mean (± s.e.) of the relative amount of total inorganic carbon (TIC) removed across the forced ventilated cascade aeration columns at the three nominal alka-linity treatments, i.e., 10 mg/L, 70 mg/L, and 200 mg/L as CaCO 3 Values represent analyses of samples collected each 14th day of each treatment (n = 6 per treatment).
Trang 86.0
6.5
7.0
7.5
8.0
8.5
200 mg/L alk, pH 7.98 Time day 0, end dosing
pH 7.68, 2.5 days,
-0.12 pH /day
pH 6.4, 5 days,
-0.5 pH/day
Dosing re-establi shed
Fig 4. Effect of stopping alkalinity dosing on RAS pH, at a constant feed loading to
the system Note the accelerating decline in pH with time Alkalinity dosing was
re-established at pH 6.4 The pH values of 7.98, 7.68, and 6.4 were typical of alkalinities
of nominally 200 mg/L, 70 mg/L, and 10 mg/L as CaCO 3 , respectively.
AlkalinitydosingconstitutesacostfortheRASfarmer.Thisis
especiallyevidentwhenusingsoftmake-upwatersourcesthatare
lowinalkalinity,whichisthecaseatmanylocationsinNorway
(Kristensenetal.,2009)butnotine.g.,EasternU.S(Davidsonetal.,
2011).Whenthemake-upwatersourceislowinalkalinity,dosing
ofe.g.,bicarbonatemustbeused.Thedecisiononthecontrolled
alkalinityconcentrationmustbebasedoneffectsonthefish,
nitri-fication,degassing,pHstabilityandlossofcarbonthroughtheRAS
loop.Aninorganiccarbonbudgetwasmadeforthethree
exper-imentaltreatmentsinthepresentstudy(Table5 incorporating
componentsthatdifferedbetweentreatments.Themostsignificant
componentwasthelossofTICthroughthedegasser.The200mg/L
alkalinitytreatmentshowedthehighestabsolutelossthroughthe
degasser,aswellasfortotalinorganiccarbonlossoutofthe
sys-tem(includingTIClosttowaterflushing;Table5).Althoughonly
4%oftheTICenteringthedegasserisremovedinthe200mg/L
alka-linitytreatment,theconsiderablyhigherTICconcentrationinthis
treatmentmadetheremovalsignificant
Inconclusion,Atlanticsalmonsmoltproducersusingsoftwater
make-upsources,andwishingtokeeparelativelylowsteady-state
TANandCO2concentrationinlinewithsalmontolerances,should
aimfor70mg/Lalkalinityconsideringtherelativelylowlossof
car-boncomparedto200mg/Lalkalinity,andtheincreasedpHstability
aswellasreducedTANconcentration,compared towhenusing
10mg/Lalkalinity.Inthepresentstudy,theexperimentaldesign
precludedstudyingeffectsofalkalinityonsalmonsmolt
perfor-mance,physiology, andwelfare, andsuchinvestigationsshould
thereforebeundertakeninfuturestudies
4 Conclusions
Thisstudy wasconducted in semi-commercialscale RAS, at
typicalwaterqualitiesfoundinAtlanticsalmonsmoltproduction
facilities At these conditions, it was foundthat a low
alkalin-ity(10mg/LasCaCO3)ledtoasignificantly highersteady-state
TANconcentration,comparedwhen70or200mg/Lalkalinitywas
used.Themeanarealnitrificationratewashigheratthelowest
alkalinity;however,themeanTANremovalefficiencyacrossthe
MBBRwasnotsignificantlyaffectedbyalkalinitytreatment.The
CO2 strippingefficiencyshowedonlyatendencytowardshigher
efficiencyatthelowestalkalinity,butdifferenceswerenot signif-icant.Thus,dissolvedCO2strippingefficiencyacrossadegassing unitisindependentofalkalinityunderfreshwaterconditions,when alkalinityis≤200mg/LasCaCO3.Incontrast,therelativefractionof totalinorganiccarbonthatwasremovedfromtheRASduringCO2 strippingwasmuchhigheratalowalkalinity(10mg/L)compared
tothehigheralkalinities(70and200mg/LasCaCO3).However, whencalculatingthetotalabsolutelossofinorganiccarbonfrom RAS,itwasfoundthatthedailylosswasaboutequalat10,and
70mg/L,whereasitwashighestat200mg/Lalkalinity.pH recor-dingsdemonstratedthatthe10mg/Lalkalinitytreatmentresulted
inthelowestsystempH,thelargestincreasein[H+]acrossthefish culturetanks,aswellasgivinglittleresponsetimeincaseof alka-linitydosingmalfunction.RapidpHchangesundertherelatively acidicconditionsat10mg/Lalkalinitymayultimatelycreatefish healthissuesduetoCO2orifaluminiumorothermetalsarepresent
Inconclusion, Atlanticsalmonsmoltproducersusingsoftwater make-up sources should aimfor 70mg/Lalkalinity considering therelativelylowlossofinorganiccarboncomparedto200mg/L alkalinity,andtheincreasedpHstabilityaswellasreducedTAN concentration,comparedtowhenusing10mg/Lalkalinity
Acknowledgements
The authorswish toexpress special thanksto BrittSeljebø, KristinNerdal,andDagEgilBundgaardforwaterchemistry analy-sis.Wealsothanktworeviewersfortheirvaluablecomments.The authorsalsowishtothankthereviewers,whoprovided insight-fulcommentsandsuggestions.Thisresearchwassupportedbythe ResearchCouncilofNorwaythroughtheStrategicInstitute Pro-gram(projectno.186913/I30)“Fishwelfareandperformancein recirculatingaquaculturesystems”.Allexperimentalprotocolsand methodswereincompliancewiththeanimalwelfarerequirements
bytheNorwegianAnimalResearchAuthority.Useoftradenames doesnotimplyendorsementbyNofimaortheauthors
References
APHA, 2005 In: Eaton, A., Clesceri, L., Rice, E., Greenberg, A (Eds.), Standard methods for the examination of water and wastewater American Public Health Associa-tion, Washington, USA.
Bergheim, A., Drengstig, A., Ulgenes, Y., Fivelstad, S., 2009 Production of Atlantic salmon smolts in Europe—current characteristics and future trends Aquac Eng.
41, 46–52.
Biesterfeld, S., Farmer, G., Russell, P., Figueroa, L., 2003 Effect of alkalinity type and concentration on nitrifying biofilm activity Water Environ Res 75, 196–204.
Chen, S., Ling, J., Blancheton, J.-P., 2006 Nitrification kinetics of biofilm as affected
by water quality factors Aquac Eng 34, 179–197.
Clingerman, J., Bebak, J., Mazik, P.M., Summerfelt, S.T., 2007 Use of avoidance response by rainbow trout to carbon dioxide for fish self-transfer between tanks Aquac Eng 37, 234–251.
Colt, J., 2006 Water quality requirements for reuse systems Aquac Eng 34, 143–156.
Colt, J., Watten, B., Pfeiffer, T., 2012 Carbon dioxide stripping in aquaculture Part 1: terminology and reporting Aquac Eng 47, 27–37.
Dalsgaard, J., Lund, I., Thorarinsdottir, R., Drengstig, A., Arvonen, K., Pedersen, P.B.,
2013 Farming different species in RAS in Nordic countries: current status and future perspectives Aquac Eng 53, 2–13.
Danley, M.L., Kenney, P.B., Mazik, P.M., Kiser, R., Hankins, J.A., 2005 Effects of carbon dioxide exposure on intensively cultured rainbow trout Oncorhynchus mykiss: physiological responses and fillet attributes J World Aquac Soc 36, 249–261.
Davidson, J., Good, C., Welsh, C., Brazil, B., Summerfelt, S., 2009 Heavy metal and waste metabolite accumulation and their potential effect on rainbow trout per-formance in a replicated water reuse system operated at low or high system flushing rates Aquac Eng 41, 136–145.
Davidson, J., Good, C., Welsh, C., Summerfelt, S., 2011 The effects of ozone and water exchange rates on water quality and rainbow trout performance in replicated water recirculating systems Aquac Eng 44, 80–96.
Espmark, Å., Kolarevic, J., Åsgård, T., Willumsen, F., Lange, G., Alfredesen, J., Alver, M., Terjesen, B., 2013 Tank size in experimental design matters In: Aquaculture Europe 2013, Trondheim, Norway, p 130.
Fivelstad, S., 2013 Long-term carbon dioxide experiments with salmonids Aquac Eng 53, 40–48.
Trang 9Fivelstad, S., Haavik, H., Løvik, G., Olsen, A.B., 1998 Sublethal effects and safe levels
of carbon dioxide in seawater for Atlantic salmon postsmolts (Salmo salar L.):
ion regulation and growth Aquaculture 160, 305–316.
Fivelstad, S., Olsen, A.B., Asgard, T., Bæverfjord, G., Rasmussen, T., Vindheim,
T., Stefansson, S., 2003a Long-term sublethal effects of carbon dioxide on
Atlantic salmon smolts (Salmo salar L.): ion regulation, haematology,
ele-ment composition, nephrocalcinosis and growth parameters Aquaculture 215,
301–319.
Fivelstad, S., Olsen, A.B., Kløften, H., Ski, H., Stefansson, S., 1999 Effects of carbon
dioxide on Atlantic salmon (Salmo salar L.) smolts at constant pH in bicarbonate
rich freshwater Aquaculture 178, 171–187.
Fivelstad, S., Waagbø, R., Stefansson, S., Olsen, A.B., 2007 Impacts of elevated water
carbon dioxide partial pressure at two temperatures on Atlantic salmon (Salmo
salar L.) parr growth and haematology Aquaculture 269, 241–249.
Fivelstad, S., Waagbø, R., Zeitz, S.F., Hosfeld, A.C.D., Olsen, A.B., Stefansson, S., 2003b.
A major water quality problem in smolt farms: combined effects of carbon
dioxide, reduced pH and aluminium on Atlantic salmon (Salmo salar L.) smolts:
physiology and growth Aquaculture 215, 339–357.
Grace, G., Piedrahita, R., 1994 Carbon dioxide control In: Timmons, M.B., Losordo,
T.M (Eds.), Aquaculture Water Reuse Systems: Engineering Design and
Man-agement Elsevier Science, New York, NY, pp 209–234.
Grisdale-Helland, B., Helland, S.J., 1997 Replacement of protein by fat and
carbo-hydrate in diets for Atlantic salmon (Salmo salar) at the end of the freshwater
stage Aquaculture 152, 167–180.
Gutierrez, A., Kolarevic, J., Sæther, B., Bæverfjord, G., Takle, H., Medina, H., Terjesen,
B., 2011 Effects of sub-lethal nitrite exposure at high chloride background
dur-ing the parr stage of Atlantic salmon In: Aquaculture Europe 2011 Proceedings,
Rhodes, Greece, pp 1080–1081.
Hosfeld, C.D., Engevik, A., Mollan, T., Lunde, T.M., Waagbø, R., Olsen, A.B., Breck, O.,
Stefansson, S., Fivelstad, S., 2008 Long-term separate and combined effects of
environmental hypercapnia and hyperoxia in Atlantic salmon (Salmo salar L.)
smolts Aquaculture 280, 146–153.
Kern, D., 1960 The hydration of carbon dioxide J Chem Educ 37, 14–23.
Kolarevic, J., Baeverfjord, G., Takle, H., Ytteborg, E., Reiten, B.K.M., Nergård,
S., Terjesen, B.F., 2014 Performance and welfare of Atlantic salmon smolt
reared in recirculating or flow through aquaculture systems Aquaculture,
http://dx.doi.org/10.1016/j.aquaculture.2014.03.033 (in press).
Kolarevic, J., Bæverfjord, G., Åsgård, T., Oehme, M., Takle, H., Ytterborg, E.,
Grisdale-Helland, B., Helland, S., Terjesen, B., 2009 Effects of light intensity and period
on feed intake, growth and respiration in Atlantic salmon (Salmo salar) in
fresh-water In: Aquaculture Europe 2009, European Aquaculture Society, Trondheim,
Norway, pp 323–324.
Kolarevi ´c, J., Ciri ´c, M., Zühlke, A., Terjesen, B.F., 2011 On-line pH measurements in
recirculating aquaculture systems (RAS) In: Aquaculture Europe 2011, European
Aquaculture Society, Trondheim, Norway, pp 570–571.
Kolarevic, J., Selset, R., Felip, O., Good, C., Snekvik, K., Takle, H., Ytteborg, E.,
Bæver-fjord, G., Åsgård, T., Terjesen, B.F., 2013 Influence of long term ammonia
exposure on Atlantic salmon (Salmo salar L.) parr growth and welfare Aquac.
Res 44, 1649–1664.
Kristensen, T., Åtland, Å., Rosten, T., Urke, H., Rosseland, B.O., 2009 Important influent–water quality parameters at freshwater production sites in two salmon producing countires Aquac Eng 41, 53–59.
Landolt, M., 1975 Visceral granuloma and nephrocalcinosis in trout In: Ribelin, W., Migaki, G (Eds.), The Pathobiology of Fishes University of Wisconsin Press, Madison, pp 793–799.
Loyless, J.C., Malone, R.F., 1997 A sodium bicarbonate dosing methodology for ph management in freshwater-recirculating aquaculture systems Progressive Fish-Culturist, vol 59., pp 198–205.
Moran, D., 2010 Carbon dioxide degassing in fresh and saline water I: degassing performance of a cascade column Aquac Eng 43, 29–36.
Mydland, L., Rud, I., Rudi, K., Ulgenes, Y., Ibieta, P., Gutierrez, X., Reiten, B., Sum-merfelt, S., Terjesen, B., 2010 Water quality and microbial community shifts during start-up, disturbances and steady-state in a new moving bed bioreactor In: Aquaculture Europe 2010, Porto, Portugal.
Paz, J., 1984 The Effects of Borderline Alkalinity on Nitrification in Natural Water Systems Polytechnic Institute of New York, New York, pp 125.
Pfeiffer, T.J., Wills, P.S., 2011 Evaluation of three types of structured floating plastic media in moving bed biofilters for total ammonia nitrogen removal in a low salinity hatchery recirculating aquaculture system Aquac Eng 45, 51–59.
Rusten, B., Eikebrokk, B., Ulgenes, Y., Lygren, E., 2006 Design and operations of the Kaldnes moving bed biofilm reactors Aquac Eng 34, 322–331.
Skogheim, O., Rosseland, B., 1986 Mortality of smolt of Atlantic salmon, Salmo salar L., at low levels of aluminium in acidic softwater Bull Environ Contam Toxicol.
37, 258–265.
Suhr, K.I., Pedersen, P.B., 2010 Nitrification in moving bed and fixed bed biofilters treating effluent water from a large commercial outdoor rainbow trout RAS Aquac Eng 42, 31–37.
Summerfelt, S.T., Sharrer, M.J., 2004 Design implication of biofilter carbon diox-ide production within recirculating salmonid culture systems Aquac Eng 32, 171–182.
Summerfelt, S., Bebak-Williams, J., Tsukuda, S., 2001 Controlled systems: water reuse and recirculation In: Wedermeyer, G (Ed.), Second Edition of Fish Hatch-ery Management American Fisheries Society, Bethesda, pp 285–395.
Summerfelt, S.T., Vinci, B.J., Piedrahita, R.H., 2000 Oxygenation and carbon dioxide control in water reuse systems Aquac Eng 22, 87–108.
Tarre, S., Green, M., 2004 High-rate nitrification at low ph in suspended- and attached-biomass reactors Appl Envrion Microbiol 70, 6481–6487.
Terjesen, B.F., Summerfelt, S.T., Nerland, S., Ulgenes, Y., Fjæra, S.O., Megård Reiten, B.K., Selset, R., Kolarevic, J., Brunsvik, P., Bæverfjord, G., Takle, H., Kittelsen, A., Åsgård, T., 2013 Design, dimensioning, and performance of a research facility for studies on the requirements of fish in RAS environments Aquac Eng 54, 49–63.
U.S EPA, 1975 Process Design Manual for Nitrogen Control Office of Technology Transfer Environmental Protection Agency, Washington, DC, USA.
U.S EPA, 1983 Methods for chemical analysis of water and wastes Environmental Protection Agency, Cincinnati, OH, USA, pp 491.
Villaverde, S., García-Encina, P.A., Fdz-Polanco, F., 1997 Influence of pH over nitri-fying biofilm activity in submerged biofilters Water Res 31, 1180–1186.