Green lamps as visual stimuli affect the catch efficiency of floating cod Gadus morhua pots in the Baltic Sea

(1)

ContentslistsavailableatScienceDirect

Fisheries

Research

jo u r n al h om ep a g e :w w w . e l s e v i e r . c o m / l o c a t e / f i s h r e s

Green

lamps

as

visual

stimuli

affect

the

catch

efficiency

of

floating

cod

(

Gadus

morhua

)

pots

in

the

Baltic

Sea

Andreas

C.

Bryhn

∗

,

Sara

J.

Königson,

Sven-Gunnar

Lunneryd,

Mikaela

A.J.

Bergenius

SwedishUniversityofAgriculturalSciences,DepartmentofAquaticResources,Skolgatan6,74242Öregrund,Sweden

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received11June2013

Receivedinrevisedform9April2014 Accepted21April2014

HandlingEditorDr.P.He Availableonline16May2014

Keywords: Atlanticcod Gadusmorhua Visualstimuli Codpot Greenlight

a

b

s

t

r

a

c

t

Thisstudycomparedcatchesinnumbers(catchperuniteffort,CPUE)andbiomass(weightperuniteffort, WPUE)ofAtlanticcod(Gadusmorhua)intwo-chamberedfloatingcodpotsequippedwithandwithout agreenlampasaformofvisualstimuliintwoareas.AgreenlampinsidethepotincreasedtheCPUE andWPUEofcodabovethemaximumlegalsize(>38cm)by74and80%,respectively.Therewereno differencesinCPUEandWPUEofsmall(<38cm)codcaughtinpotswithandwithoutagreenlamp.By increasingthecatchefficiencyofcodgreenlampscouldconsequentlybehighlybeneficialfortheBaltic commercialcodpotfishery.

1. Introduction

Seal-inflicted damage to fishing gear and catch losses have

increasedrapidlyin many typesof small-scale coastalfisheries

alongtheBalticSeacoastandhaveseverelyaffectedtheinshore

fishery (Westerberg et al., 2008; Hemmingsson et al., 2008;

BruckmeierandHöjLarsen,2008).ThegillnetfisheryforAtlantic cod(Gadusmorhua;hereaftersimplyreferredtoascod)has,for instance,experiencedanextensivesurgeindamagecausedbygrey seal(Halichoerusgrypus)sinceyear2005(Königsonetal.,2009).In responsetotheconflictbetweentheincreasingsealpopulationand acontinuedviablecoastalfisheryalternative,seal-safegear,such

asbaitedcodpots,isnowunderdevelopment.Potsmayserveas

analternativefishingmethodwithlesssealdamageandfishloss thangillnetsandlonglines(Königson,2011).

TheCommonFisheriesPolicy(CFP)aswellastheHelcomBaltic

SeaActionPlanpromotethedevelopmentofanecosystem-based

managementofcoastalfisheries.Theuseofpotsinthecod

fish-erywouldcontributetowardsthisdevelopment.Incomparisonto

trawlsandotheractivefishinggears,potscauselimitedharmtothe marineenvironment(Jenningsetal.,2001;Thomsenetal.,2010).

Suuronenetal.(2012)includedpotsinacompilationofLIFE(Low ImpactandFuelEfficient)fishinggearduetotheirlowenergyuse,

∗Correspondingauthor.Tel.:+46104784152. E-mailaddress:andreas.bryhn@slu.se(A.C.Bryhn).

effectivespeciesselectivity,betterprotectionagainstmarine mam-mals,andlowbycatchofmarinemammalsandseabirds,inaddition

tolow gear constructioncosts.Anotheradvantage withpotsis

thattheycanbedesignedtocapturecod aboveacertainlength

(Königson,2011; Ovegårdet al.,2011).Anotherequally impor-tantreasonfordevelopingalternativefishinggear,suchaspots, isthatthesmall-scalecoastalfisheriessufferfromlowprofitability

andscantadditionofyoungfishers.Apositivedevelopmentsuch

asusingalternativefishinggearcouldincludeecolabellingfishor

marketingthefishaslocallycaughtwhichinturncouldgivethe

fishersahighercatchvalue.Comparableorhighercatchesfrom

anequalamountofeffortprovidesincentivesforchangingfroma traditionaltoanalternativefishery.Generally,thecatchefficiency ofpotshasbeenfarlowerthanthatofotherfishinggear,suchas trawls(Suuronenetal.,2012).Presentlythere isnocommercial potfisheryintheBaltic.However,experimentalfishingconducted

incooperationwithcommercialfishermenhasshowedthat

dur-ingcertaintimeperiodspotscanbeasefficientasgillnetandhook fisheriesinthesamearea(KönigsonandLunneryd,2013).The fish-ingefficiencyofthepotsisusuallymaintainedbyattractingfishto thepotusingbait(FurevikandLøkkeborg,1994;Løkkeborg,1998), butdiscussionsabouthowtofurtherimprovefishingefficiencyare ongoing.

Thefishingefficiencyofpotsistoagreatextentrelatedtofish behaviourwhencomparedtoothertypesoffishinggear.Potsmust havetherightcharacteristicstolurethefishtoenterthepot.Cod generallyapproachpotsslowly(Furevik,1994)andfishsometimes

(2)

Fig.1. Locationoftestfishingconcerninggreenlampsasvisualstimuli.(1)Bayof Hanö.(2)KarlskronaArchipelago.Theinsetmapshowsthelocationofthetestsites insouthernSweden.

swimintopotsforotherreasonsthanforaging,e.g.,curiosity, move-mentofotherfishinsidethepot,orprotection(HighandEllis,1973; Furevik,1994).

Ithasbeenknownforcenturiesthatfishcanbeattracted,or otherwiseaffected,byartificiallight(Hasegawa,1993;Marchesan etal.,2005)andbrightobjects(Slack-Smith,2001).Intensivelight enablesmanyfishspeciestodistinguishfoodmoreclearly( Ben-Yami,1988).Behaviourinrelationtovisualstimuliis,however,to

someextentspeciesdependent(Ben-Yami,1988).Studiesoncod

haveshowedthatfarmedcodmaygrowmorerapidlywhen

stimu-latedbyartificiallight(Tarangeretal.,2005)andcodincaptivityuse visiontoalargedegreetodetectfoodinmid-waterbutuseamore complexcombinationofsensestofindfoodinbottomareas(Brawn, 1969).Moreover,codvisionreactsdifferentlytodifferent wave-lengthsofthephotons;aprimarysensitivitypeakoccursat490nm

(blue/greenlight)andasecondarypeakat550nm(green/yellow;

AnthonyandHawkins,1983).Thesestudiesindicatethatthereis reasontohypothesisethatitcouldbepossibletoincreasefishing

efficiencyofpotsbyattractingcodintothepotchambersusing

greenlightasavisualstimulus.

Thepresentstudyinvestigatedwhethergreenlampmounted

insidefloatingcodpotswouldincreasethefishingefficiency (mea-suredasthecatchperuniteffort,CPUE,andweightperuniteffort,

WPUE)ofcodaboveandbelowthemaximumlegalsizelimitintwo

differentareas.Thecatchefficiencyfromexperimentalcodpots

wascomparedtotheefficiencyofcontrolpotswithoutlightsthat weredeployedatthesamesites.Auniteffortisdefinedhereasone

soakingeventincludingonetwo-chamberpot.Thedifferencein

CPUEandWPUEdependingonsoakingtimewasalsostudiedand

discussed.

2. Materialsandmethods

Experimental fishing was conducted at two locations (Bay

ofHanö; 55◦₄₂′_–56◦₀₀′ _N _and₁₄◦₄₁′_–15◦₀₈′ _E,_and _Karlskrona

Archipelago;56◦₀₃′_–56◦₀₆′_N_and₁₅◦₃₀′_–15◦₃₄′_E)_in_the

south-ernBaltic Seain Swedishterritorialwaters betweenMay2008

andNovember2010(Fig.1).Experimentswerecarriedoutin

col-laborationwithtwoprofessionalfishersusingatwo-chambered

single-entrancefloatingcodpotdescribedbyFureviketal.(2008)

andOvegårdetal.(2011).InKarlskronaArchipelago,fishingwas

conductedclosetotheshoreatanaveragedepthof20m.Tests

withcontrolpotsandpotsmodifiedwithgreenunderwaterlamps

werecarriedoutfromApril2009toJuly2010.Duringthefirsttwo monthsfromApriluntilMay,2009,groundlineswithfivecontrol

potswereset,atthesametimeandsidebysidetogroundlines

withfivetestpots,identicallydesignedbutincludingvisualstimuli. Thereafter,potsweredeployedingroundlinesoffouroreightpots alongthesamegroundline.Onetofourgroundlineswereset dur-ingthesameday(Table1).Groundlinesincludedthesamenumber ofrandomlyplacedcontrolpotsaswellasonetypeoftestpots.In theBayofHanö,testswith4or8potspergroundlinewerecarried

outfromJuneuntilNovember,2010.Atthissite,thefishingwas

conductedatopenseaatanaveragedepthof43m.Foran

illus-trationofhowgroundlinesandpotswereused,seeOvegårdetal. (2011).Ifanytypeofseriousdisturbancewasnotedwhenthepot wasemptied,suchasifthegreenlampwasnotfunctioning,ifthe catchincontrolandtestpotsweremixedup,iftherewasaholein thepot,orifthepothadbeentwistedaroundthegroundline,the specificpotwasexcludedfromtheanalysis.In5%oftheemptied

potssucherrorswererecorded.

Green lamps (electric fishing light) were acquired from

www.artisanalfish.com. Each lamp consisted of two LED green

lights, with a peak wavelength of 523nm (linewidth at

Ee/2=26nm). Measured maximum output intensity (Ee) was

124␮W.Thesizeofeachlampwas120mm×43mmwithapower

supplyof3VLR06(2AA).Thelampswereplacedbythebaitbagin themiddleofthepot.Onlypotswithfunctioninglampsattheend ofthesoakingperiodwereincludedinthedatasetforstatistical analyses.

Inbothareas,eachpotwasbaitedwithabout250gofcutfresh

herring.Potsoakingtimewasonaverage3.9days,varyingfrom

onedaytonineteendays,dependingonfactorssuchaslocation,

weather conditions,and fishingroutines.Cod aboveand below

38cm(minimumlegalsizeforBalticSeacod)wereanalysed sepa-rately.FromMarch,2010,allpotswerefittedwitha45mmsquare meshpanelinordertotestthedifferenceincatchofundersized cod(resultsfromtestsusingpanelsarepresentedinOvegårdetal., 2011,althoughtestsconcerninggreenlampsinpotswerenot

ana-lysedinthatstudy).CPUEandWPUE(catchandweightperunit

effortinnumberandkgcodperpot)of<38cmand>38cmcodfrom potswithoutselectionpanels(328potsinKarlskronaArchipelago)

wereseparatedintotwodatasets(smallcodandlargecod).From

potswhichwereequippedwithaselectionpanel(allpotsdeployed intheBayofHanöand261ofthepotsinKarlskronaArchipelago),

CPUEandWPUEofcodabove38cmwereaddedtothelargecod

datasetwhile<38cmcodwerenotusedintheanalysisduetothe

lownumberofsmallcodcaughtinsuchpots.

Thetwofishersreportedcatches,typesofvisualstimuliapplied toeachpot,location,andsoakingtimeusingstandardised proto-cols.IntheBayofHanö,thefisherreporteddatafromeachpotin

everygroundline.InKarlskronaArchipelago,datawererecorded

foreverygroundlinewitheithercontrolortestpotsaswellasfor eachpotineachgroundline(Table1).On40%ofthefishingtrips

inBayofHanö,on-boardobserversjoinedthefisherstomeasure

thelengthandweightoffishandtoensuremaintaineddata qual-itybycomparingtherecordingsbythefishermentothosetakenby theobservers.Therewerenodeviationsbetweenthereportstaken

byfishermenandobservers.InKarlskronaArchipelago,observers

werepresentlessfrequently,althoughregularcontactwas

main-tained between researchers, observers and fishers in order to

ensurehighdataquality.

AccordingtoFurevik(1994),itisthedurationoftheodourplume ofthebaitwhich attractsthefishtothepotandinfluencesthe fishingtimeofapot.Sincethebaitiseitherconsumedorceases toemitasufficientamountofolfactorystimulusafteroneortwo days(Furevik,1994),potswithsoakingtimeofonetotwodays wereanalysedseparatelyasasubsetofthetotaldataset.The fish-ingdepthanddatewererecordedforeachtrip/effort.Associations

betweencatchandwaterdepth,season(months)andlunarcycle

(3)

Table1

Detailsaboutgreenlamptests.Location1:BayofHanö.Location2:KarlskronaArchipelago.Selectionpanelssharplydecreasethecatchofcod<38cm(Ovegårdetal.,2011).

Location Fishingperiod Datareported Groundlinespertrial Potspergroundline Selectionpanel Numberof emptiedpots Test Control 1 22/06/2010–07/11/2010 Perpot 1–4 8 Yes 219 219 2 23/04/2009–21/07/2010 Pergroundline/perpot 1–3 4–8 Yes/no 296 293

Non-parametricMann–WhitneyUtestswereusedtocompare CPUEandWPUEbetweencontrolandexperimentalpots.Asparts ofthedatawererecordedpergroundlineofpots(insteadof indi-vidualpots)intheKarlskronaArchipelago,CPUEandWPUEwere calculatedastheaveragecatchperpotinagroundline.The non-parametricMann–WhitneyUanalysiswasselectedasnoneofthe comparablesamplesdisplayedanormaldistribution.Furthermore, aforwardstepwisemultipleregressionwasperformedto investi-gatemultipledeterminantsofCPUEandWPUE.Theseregressions wereusedasabasistoconstructnon-parametricgeneralised lin-ear models (GLMs) withPoisson distributions. Only significant (p<0.05)explanatoryvariables werekeptin theGLMs,andthe softwareusedwasBrodgar(www.brodgar.com).Inregressionsand

models,CPUEandWPUEwereBox%3Fndash;Coxtransformedinto

(CPUE)0.25_and_(WPUE)0.25_in_order_to_attain_normal_{distributions.}

The95%confidencelevelwasusedthroughoutthestudy.

3. Results

Fig.2showstheCPUEof>38cmcodinpotswithandwithout

greenlamps,andfordifferentsoakingtimes.Nooptimalsoaking

timecouldbeobservedinthedataanditshouldbenotedthatpots inbothcategories(lampandnolamp)continuedtoretaincodlong aftertheolfactoryeffectfromtheherringbaithadvanished(after approximately1–2days,Fig.2).Thetwoareaswerefishedduring

differentmonthsandthecatchwasparticularlyhighin

August-November(Fig.3),althoughdatawereonlycollectedfromtheBay

Fig.2.Soakingtime(days)andCPUEofcod>38cminpotsdeployedintheBayof HanöandtheKarlskronaArchipelago.Upperpanel:testpots(N=512)withgreen lamps.Lowerpanel:controlpots(N=515)withoutlamps.

Fig.3.CPUEofcod>38cmduringdifferentmonths.Upperpanel:site1(Bayof Hanö).Lowerpanel:site2(KarlskronaArchipelago).

ofHanö(site1)duringthesemonths.DifferencesinCPUEandWPUE betweensite1andsite2aredisplayedinFig.4whiledifferencesin

CPUEandWPUEbetweencontrolpotsandtestpotsaredisplayed

inFig.5.

(4)

Table2

Resultsfromgreenlamptests.p-ValuesdescribestatisticaldifferenceuncertaintyaccordingtoMann–WhitneyUtests.Location1:BayofHanö.Location2:Karlskrona Archipelago.N.A.:datanotavailable.

Soakingtime(days) Location Sizeofcod(cm) Pot category

Numberofpots MeanCPUE (number)

p-Valueof difference

Mean WPUE(kg)

p-Valueof difference 1–14 1+2 >38 Test 512 3.4 <0.001 2.8 <0.001

Control 515 2.0 1.6

1 >38 Test 219 6.2 <0.001 4.8 <0.001

Control 219 3.2 2.3

2 >38 Test 293 1.3 0.024 1.4 0.030

Control 296 1.1 1.0

2 <38 Test 164 1.6 0.61 N.A. N.A. Control 164 1.3 N.A.

1–2 1+2 >38 Test 154 3.1 0.009 2.6 0.014

Control 154 1.7 1.4

1 >38 Test 87 4.4 0.021 3.4 0.033

Control 87 2.3 1.7

2 >38 Test 67 1.3 0.11 1.4 0.17

Control 67 1.0 1.0

2 <38 Test 39 2.2 0.28 N.A. N.A.

Control 40 2.6 N.A.

Resultsfromtheexperimentwithgreenlampsandwith differ-entsoakingtimearedisplayedinTable2.Potsequippedwitha

greenlampcontainedanaverageCPUEof3.4(standarddeviation,

SD:5.8)concerningcod>38cmandtheaverageWPUEofsuchcod was2.8kg(SD:4.8kg).Thesevaluesweresignificantly(p<0.001)

higherthanthoseofthecontrolpotswhichhadanaverageCPUE

of2.0(SD:3.4),andaWPUEof1.6kg(SD:2.8kg).

Whentestdatawereanalysedperarea,significantdifferences in>38cmcodCPUEandWPUEremained(Table2).AmongtheBay

ofHanödata,theaverageCPUEinpotswithgreenlampswas6.2

individuals>38cmandtheaverageWPUEof>38cmcodwas4.8kg, comparedtotheaveragecontrolpotwithaCPUEof3.2andaWPUE of2.3kgconcerninglargecod(numberofpots:98).Mann–Whitney

UtestsontheseBayofHanödataindicateddifferencesinCPUE

andWPUEatp<0.001.InKarlskronaArchipelago,meanCPUEand

WPUEintestpotswere1.3and1.4kg,respectively,comparedto

controlpotswithameanCPUEof1.1andameanWPUEof1.0kg

(differenceatp=0.024forCPUEandp=0.030forWPUE).

Green lampsmadeno significantdifference(p=0.61) tothe

CPUEofsmallcod(<38cm)intheKarlskronaArchipelagowhentest

Fig.5.CPUEandWPUEofcod>38cmincontrolpotsandtestpots.

andcontrolpotswithoutsizeselectionpanelswereused(Table2). TheaverageCPUEof<38cmcodwas1.4(SD:1.4).

Theyieldinpotswith1–2daysofsoakingtimewasgenerally

lowerthantheaverageyieldinthetotaldatasetconcerningcod

>38cm(Table2).Smallcod(<38cm)were,however,particularly

numerousinpotswithasoakingtimeof1–2days.Thereduced

datasetwith1–2daysofsoakingtimelargelycorroboratedpatterns inthetotaldataset,althoughthenumberofpotswas,asexpected,

lowerandpvalueswereinmanycaseshigher.

DifferencesinCPUEandWPUEofcod>38cmbetweentestpots andcontrolpotsremainedsignificantinthereduceddataset(with

soakingtime:1–2days)whendatafrombothareaswereanalysed

(p=0.009andp=0.014,respectively;Table2).CPUEandWPUEof >38cmcodwerealsosignificantlyhigherintestpotsdeployedfor 1–2daysintheBayofHanö(atp=0.021andp=0.033,respectively) comparedtocontrolpots.IntheKarlskronaArchipelago,the

dif-ferencein>38cmcodCPUEandWPUEbetweenpotsusedfor1–2

dayswasnotsignificant(p=0.11and0.17,respectively).No signif-icantdifference(p=0.28)inCPUEcouldbedistinguishedbetween testpotsandcontrolpotsconcerningcod<38cmcaughtinpots withoutsizeselectionpanelswhenthesoakingtimewas1–2days.

TherewasnosignificantcorrelationbetweenCPUEorWPUE

andsoakingtime,paneluseinpots,waterdepth,season(month

oftheyear)orlunarcycle.However,thereweresignificant

corre-lationsbetweenCPUEandconsecutivemonth(asFig.3suggests),

betweenWPUEandconsecutivemonth,betweenCPUEandsite,

WPUEandsite,CPUEandtreatmentand,finally,WPUEand

treat-ment. Forward stepwise multiple regressions (Tables 3 and 4)

excludedconsecutivemonthasasignificantexplanatoryvariable

concerningCPUE andWPUEasdependentvariables. GLMswith

Poissondistributionsconfirmedthatsiteandtreatmentwere

sig-nificantdeterminantsofCPUEandWPUEbutcouldnotsupport

theuseoftheinteractionproductbetweensiteandtreatmentas

determinants(Table5).

4. Discussion

ThisstudyshowsthattheAtlanticcodresemblesmanyother

pelagicfishspecies (e.g.,herring, anchoviesandmackerel;

Ben-Yami,1988)inthatitisattractedtolightandthatthecodcatch

efficiencyofpotsequippedwitha greenlampwassignificantly

higherthanthoselackingalampintwoareasontheeastcoast

ofSweden.Resultssuggestedthatthecodcatchrateintermsof

weightperpotcouldincreasebyonaverage80%(andby108%in

(5)

Table3

ForwardstepwisemultipleregressionconcerningCPUE.Site:1forBayofHanöand2forKarlskronaArchipelago.Treatment:1forgreenlampand0fornolamp.Y=CPUE0.25_.

StepandninXn Xn Equation R2 p

1 (Site–1) Y=1.00−0.286·X1 0.045 <0.001forallcoefficients

2 Treatment Y=0.924−0.285·X1+0.162·X2 0.059 <0.001forallcoefficients

3 (Site–1)·Treatment Y=0.864–0.181·X1+0.282·X2−0.209·X3 0.065 0.018forcoefficientregarding

X1;0.011forcoefficient

regardingX3;<0.001for

remainingcoefficients

Table4

ForwardstepwisemultipleregressionconcerningWPUE(inkg).Site:1forBayofHanöand2forKarlskronaArchipelago.Treatment:1forgreenlampand0fornolamp.

Y=WPUE0.25.

StepandninXn Xn Equation R2 p

1 (Site–1) Y=0.918−0.206·X1 0.025 <0.001forallcoefficients

2 Treatment Y=0.838−0.206·X1+0.161·X2 0.041 <0.001forallcoefficients

3 (Site–1)·Treatment Y=0.784−0.112·X1+0.269·X2−0.188·X3 0.047 0.045forcoefficientregarding

X1;0.018forcoefficient

regardingX3;<0.001for

remainingcoefficients

Table5

Statisticsfromgeneralisedlinearmodels(GLMs)forCPUE0.25_and_WPUE0.25_;_WPUE_in_kg._Site:₁_for_Bay_of_Hanö_and₂_for_Karlskrona_Archipelago._Treatment:₁_for_green

lampand0fornolamp.InteractionsbetweenSiteandTreatmentwereinsignificant(p>0.05).

Explanatoryvariable Coefficientsigns p-ValueinGLMwithCPUE0.25 _p_-Value_in_GLM_with_WPUE0.25

(Site–1) Negative <0.001 <0.001

Treatment Positive 0.0069 0.0075

Lightattractionamongcodhasnotbeentestedbeforein experi-mentsusingcontrolgroups.However,somesignificantresultsin thisstudycanbesupportedbyearlierfindingsintheliterature,e.g., thatcodcanusevisionwhenlocatingfood(Brawn,1969)andthat codarereceptivetogreenlight(AnthonyandHawkins,1983).

Theolfactorystimulusfromherringbaitinpotsiseffectivefor 1–2days(Furevik,1994).However,thepropensitytoretaincod

persistedinthepotsformuchmorethantwodays,whetherthere

wasagreenlampinsideornot(Fig.2).Königsonetal.(2014)used

alargerdatasetofpotsfromtheBayofHanöandtheKarlskrona

Archipelagotoinvestigatetheoptimalsoakingtimeforpotsinthe

area.Theydistinguishedapeakincodyieldafterapproximately

6daysofsoakingtime.Presumably,theherringbaitisimportant duringthefirsttwodaysandafterthat,thefisheithertendtostay

inpotsandcannotfindtheirwayoutorareaccompaniedbynew

individualswhoareattractedbycodaggregatinginside.

The only inconsistency in the test results between the full

dataset (soaking time: 1–14 days) and the reduced dataset

(soaking time: 1–2 days) concerned cod >38cm in the

Karl-skrona Archipelago. p-values from the full dataset indicated a

significant difference in CPUE and WPUE while p-values from

the reduced dataset pointed at a lack of significant difference

(Table2).However,differencesweremarginalinabsolute num-bers(CPUE=1.3andWPUE=1.4kgintestpotsandCPUE=1.0–1.1 andWPUE=1.0kgincontrolpots;Table2).

Thenear-fieldandingressbehaviourisimportantforthe fish-ingefficiencyofthepots.AlongNorway’sAtlanticcoast,Furevik (1994)foundthatalargeproportionofcodbeingattractedtothe

potdidnotactuallyenterwhenapproachingtheentranceofthe

pot.Theincreaseinthenumberofcodinpotswithgreenlamps

insideindicatesthatvisualstimuliinsidethepotmayindeedaffect

thepropensity for cod to enter the potand thereby affectthe

ingressbehaviourofthefish.However,lightcould,alternatively, indirectlylurecodtoenterthepotsbyattractingpotentialcodprey speciessuchassmallerfishorcrustaceans.Thepotential differ-encesinpreyspeciescouldexplainthemarkeddifferenceincatch betweenthetwoareas.TheBayofHanöhasaslightlyhighersalinity

thanKarlskronaArchipelago,whichmaypointtoahigher

num-berofcrustaceans.Thedensityofinvertebratesincreasesinthe

BalticSeaalonganincreasingsalinitygradient(Bonsdorff,2006).

Anotherfactor thatcan influencetheeffect fromvisualstimuli

whichmightdifferinthetwoareasisthevisibilityinthewater.In

KarlskronaArchipelago,experimentswereconductedclosertothe

shorewherewatermighthavebeenmoreturbidcomparedtothe

BayofHanöwherefishingwascarriedoutfurtherawayfromthe

shore.MultipleregressionandGLMsshowedthatdepth,paneluse

inpots,seasonandconsecutivemonthwereredundantas

deter-minantsofCPUEandWPUEofcod>38cmwhensitewasusedas

anexplanatoryvariable(Tables3–5).

StandarddeviationsofCPUEandWPUEweresimilartomean

values,indicatingalargedifferencebetweencatchesindifferent

pots.Manypotswereequippedwithselectionpanels(Table1).

AccordingtoOvegårdetal.(2011),selectionpanelssignificantly decreased catchesofundersized cod(<38cm)inthetwo

inves-tigatedareas.Ifpanelswouldnothavebeenusedinthepresent

study,catchesofundersizedcodwouldprobablyhavebeenhigher

andthesecatchescouldhavebeenanalysedinmoredetail. Futurestudiesontheoptimalluminousintensityandanoptimal

photonwavelengthinlampsdeployedforattractingcodintopots

arerecommended,incombinationwithstudiesonhowpotential

preyforcod(smallfishandcrustaceans)maybeattractedtolight inapotandtherebyinfluencesthecatchrateofcod.

Toconclude,thisstudyshowsthatthepropensityofcodtobe

caughtincodpotsmaybeaffectedbyvisualstimuli.Agreenlamp insidethepotincreasedthenumberandweightoflarge(>38cm)

cod.Resultsshowedthatgreenlampsmaybeusedinthe

commer-cialpotfisheryasthelampincreasedthemeancatchweightoflegal sizedcodby80%.

Acknowledgements

We would like to thank the participating fishermen Göran

andKurtOlofsson,andPerMånsson,fortheireffortwhichmade

(6)

acknowledged for contributing with field work, and Christer

Blomquistforprovidinglightanalyses.LenaBergströmprovided

valuableadviceregardingthestatisticalanalysis.PeterLjungberg

designedFig.1.Twoanonymousreviewershavecontributedwith

helpfulcommentsonearlierversionsofthemanuscript.Thisstudy

wasfunded bythe EuropeanStructural Fund for Fisheries and

theProgramSealandFisherywitheconomicalsupportfromthe

SwedishAgencyforMarineandWaterManagement.

References

Anthony,P.D.,Hawkins,A.D.,1983.SpectralsensitivityofthecodGadusmorhuaL. Mar.Behav.Physiol.10,145–166.

Ben-Yami,M.,1988.AttractingFishwithLight.FAO,Rome.

Bonsdorff,E.,2006.Zoobenthicdiversity-gradientsintheBalticSea:continuous post-glacialsuccessioninastressedecosystem.J.Exp.Mar.Biol.Ecol.333, 383–391.

Brawn,V.M.,1969.Feedingbehaviorofcod(Gadusmorhua).J.Fish.Res.BoardCan. 26,583–596.

Bruckmeier,K.,HöjLarsen,C.,2008.Swedishcoastalfisheries–fromconflict miti-gationtoparticipatorymanagement.Mar.Policy32,201–211.

Furevik,D.,1994.Behaviouroffishinrelationtopots.In:Fernö,A.,Olsen,S.(Eds.), MarineFishBehaviourinCaptureAndAbundanceEstimation.FishingNews Books,Oxford.

Furevik,D.M., Løkkeborg, S.,1994.Fishing trialsinNorway fortorsk (Brosme brosme)andcod(Gadusmorhua)usingbaitedcommercialpots.Fish.Res.19, 219–229.

Furevik,D.M.,Humborstad,O.B.,Jørgensen,T.,Løkkeborg,S.,2008.Floatedfishpot eliminatesbycatchofredkingcrabandmaintainstargetcatchofcod.Fish.Res. 92,23–27.

Hasegawa,E.,1993.History,developmentandpresentconditionoffishingwithlight. Bull.Fac.Bioresour.MieUniv.10,131–140(inJapanesewithEnglishabstract). Hemmingsson,M.,Fjälling,A.,Lunneryd,S.-G.,2008.Thepontoontrap:description

andfunctionofaseal-safetrap-net.Fish.Res.93,357–359.

High,W.L.,Ellis,I.E.,1973.Underwaterobservationsoffishbehaviorintraps. Hel-goländerWissenschaftlicheMeeresuntersuchungen24,341–347.

Jennings,S.,Kaiser,M.J.,Reynolds,J.D.,2001.MarineFisheriesEcology.Blackwell, Malden,MA.

Königson,S.,(PhDdissertation)2011.SealsandFisheries–AStudyoftheConflict andSomePossibleSolutions.GothenburgUniversity.

Königson,S.,Lunneryd,S.-G.,2013.Developmentofalternativefishinggearin theSwedishsmall-scalecoastalfisheries.ProgressinMarineConservationin Europe.In:ProceedingsoftheSymposiumStralsund,Germany18–22June2012, pp.217–227.

Königson,S.,Lunneryd,S.-G.,Sundqvist,F.,Stridh,H.,2009.Greysealpredationin codgillnetfisheriesinthecentralBalticSea.J.NorthwestAtlanticFish.Sci.42, 41–47.

Königson,S.,Fredriksson,R.,Lunneryd,S.-G.,Strömberg,P.,Bergström,U.,2014. CodpotsintheBalticinshorefishery:aretheyeffectiveandwhataffectstheir efficiency?ICESJ.Mar.Sci.(inpress).

Løkkeborg,S.,1998.FeedingbehaviourofcodGadusmorhua:activityrhythmand chemicallymediatedfoodsearch.Anim.Behav.56,371–378.

Marchesan,M.,Spoto,M.,Verginella,L.,Ferrero,E.A.,2005.Behaviouraleffects ofartificiallightonfishspeciesofcommercialinterest.Fish.Res.73,171– 185.

Ovegård,M.,Königson,S.,Persson,A.,Lunneryd,S.-G.,2011.Sizeselectivecapture ofAtlanticcod(Gadusmorhua)infloatingpots.Fish.Res.107,239–244. Slack-Smith,R.J.,2001.FishingwithTrapsandPots.FAO,Rome.

Suuronen,P.,Chopin,F.,Glass,C.,Løkkeborg,S.,Matsushita,Y.,Queirolo,D.,Rihan, D.,2012.Lowimpactandfuelefficientfishing–lookingbeyondthehorizon. Fish.Res.119–120,135–146.

Taranger,G.L.,Aardal,L.,Hansen,T.,Kjesbu,O.S.,2005.Continuouslightdelays sex-ualmaturationandincreasesgrowthofAtlanticcod(GadusmorhuaL.)insea cages.ICESJ.Mar.Sci.63,365–375.

Thomsen,B.,Humborstad,O.-B.,Furevik,D.M.,2010.Fishpots:fishbehaviorcapture processesandconservationissues.In:He,P.(Ed.),BehaviorofMarineFishes: CaptureProcessesandConservationChallenges.Blackwell,Hoboken. Westerberg,H.,Lunneryd,S.,Fjälling,G.,Wahlberg,A.M.,2008.Reconcilingfisheries