ORIGINAL RESEARCH ARTICLE
Tracking trends in eutrophication based on pigments in recent coastal sediments
Ma ł gorzata Szymczak- Ż y ł a
a, Magdalena Krajewska
a,
Aleksandra Winogradow
a, Agata Zaborska
a, Gijs D. Breedveld
b, Gra ż yna Kowalewska
a,*
aInstituteofOceanology,PolishAcademyofSciences,Sopot,Poland
bNorwegianGeotechnicalInstitute,Oslo,Norway
Received15June2016;accepted22August2016 Availableonline29September2016
KEYWORDS Eutrophication;
Pigments;
Markers;
Sediments;
BalticSea;
Norwegianfjords
Summary Eutrophicationintwodifferentcoastalareas—theGulfofGdańsk(southernBaltic) andtheOslofjord/Drammensfjord(Norway)—bothsubjecttohumanpressureandwithrestricted waterexchangewithadjacentseas,wasinvestigatedandcompared.Sedimentcores(upto20cm long)werecollectedat12stationsusingacoresampler,6ineachofthetwoareas,anddivided intosub-samples.Thephysicochemicalparameterscharacterizing theadjacentwatercolumn andnear-bottomwater,i.e.salinity,oxygenconcentrationandtemperature, weremeasured duringsamplecollection.Chlorophylls-a,-band-c,theirderivativesandselectedcarotenoids weredeterminedforallthesamples,aswereadditionalparameterscharacterizingthesedi- ments,i.e.Corg,Ntot,d13Candd15N,grainsize.210Pbactivitywasalsodeterminedandonthat basissedimentmixingandaccumulationrateswereestimated.Thedistributionofpigmentsin sedimentswasrelatedtoenvironmentalconditions, thesamplingsite location and sediment characteristics.Theresultsareinagreementwithotherobservationsthateutrophicationinthe GulfofGdańskhasincreased,especiallysincethe1970s,whereasintheOslofjorditdecreased duringthesameperiod.ThepigmentsarebetterpreservedininnerOslofjordsedimentsthanin thosefromtheGulfofGdańsk.Theresultsdemonstrate thatthesumofchloropigments-ain
PeerreviewundertheresponsibilityofInstituteofOceanologyofthePolishAcademyofSciences.
* Correspondingauthorat:InstituteofOceanology,PolishAcademyofSciences,ul.PowstańcówWarszawy55,81-712Sopot,Poland.
Tel.:+48587311615.
E-mailaddresses:[email protected](M.Szymczak-Żyła),[email protected](M.Krajewska),[email protected]
(A.Winogradow),[email protected](A.Zaborska),[email protected](G.D.Breedveld),[email protected](G.Kowalewska).
Availableonlineatwww.sciencedirect.com
ScienceDirect
jo u rn al ho m e p age : w w w. jo ur na ls .e l se v i er.c o m / o ce an o lo g i a/
http://dx.doi.org/10.1016/j.oceano.2016.08.003
0078-3234/#2016InstituteofOceanologyofthePolishAcademyofSciences.ProductionandhostingbyElsevierSp.zo.o.Thisisanopen accessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
Eutrophicationisoneofthemostimportantproblemsaffect- ingmanycoastalareasworldwide(e.g.Bianchietal.,2010;
Chenetal.,2001;Fleming-Lehtinenetal.,2015;HELCOM, 2007;Lietal.,2013;Oriveetal.,2002).Itoccursinaquatic basinsofhighprimaryproductioncausedbyelevatednutrient concentrations(Edlundetal.,2009;Harmonetal.,2014).The intensivebloomsofalgaeandcyanobacteria(includingtoxin- producingphytoplanktonspecies),followedbyhighratesof sedimentation and accumulation, in conjunction with restrictedwaterexchangeresultineutrophication,whichis manifested by hypoxia/anoxia in the sediments and near- bottomwater (Conley etal.,2011; HYPOX,2016).Oxygen depletioninhibitsthegrowthofbenthicorganisms—thisis reflectedintheformationoflaminarsediments(Reussetal., 2005;Zhaoetal.,2012).Likedarknessandlowtemperatures, anoxia prevents the remineralization of organic matter in sediments(HedgesandKeil,1995).
Despitethelargebodyofknowledgerelatingtoeutrophi- cationanditsimprintonbottomsediments,itisstillnoteasy toevaluateitquantitatively,analyzeitstrendsinabasinand compare it in different locations. Numerous proxies have been applied to thisphenomenon, including organic com- pounds — principally pigments. These are chlorophyll-a, carotenoidsandtheirderivatives.Chlorophyll-ainwateris wellknownasamarkerofprimaryproductionandhasbeen usedforthispurposeinoceanographyforover50years(e.g.
BianchiandCanuel,2011;JeffreyandMantoura,1997);the sameappliestoitsderivatives(Bianchietal.,1997,2002a,b;
Carpenteretal.,1988).However,chlorophyll-a concentra- tionsinwaterchangefrequentlyintimeandspace,whereas chloropigments-a(chlorophyll-aanditsderivatives)insedi- mentshavebeenshowntobegoodindicatorsoftheaverage primaryproductioninabasin(Bianchietal.,2002a,b;Harris etal., 1996;Stephens etal., 1997; Szymczak-Żyła etal., 2011).Particularsedimentarychlorophyll-aderivativesmay betakenasmarkersofsyn-andpost-depositionalenviron- mental conditions (Szymczak-Żyła et al., 2011). Not only chloropigmentsbutalsocarotenoidsaremonitoredinsedi- mentsaschemotaxonomicandbiomassmarkers;indeed,b- caroteneisconsideredanevenbetterproxyfortotalalgal biomassthanchlorophyll-a(Dixitetal.,2000;Leavitt,1993;
Schüller et al., 2013). Numerous papers have focused on chloropigmentsandcarotenoidsinrecentandoldsediments, mainlyin lakes (e.g.Hodgson etal., 2004;Leavitt etal., 1997;McGowanetal.,2012;Moorhouseetal.,2014;Pienitz etal.,1992).Pigmentshavealsobeentrackedinshelfareas (Chenetal.,2001;Lietal.,2012,2013;Loudaetal.,2000;
Shankle etal.,2002;Sampereetal.,2008), inlarge river estuariesinAmerica(Canueletal.,2009;Chenetal.,2005;
Edlundetal.,2009;Wysockietal.,2006)andAsia(Lietal., 2011;Zhaoetal.,2012),inNewZealandfjords(Schüllerand Savage,2011)andoffthecoast ofAntarctica(Sañéetal., 2013).In contrast,notmany papershavebeen writtenon pigmentsinEuropeancoastalzonesediments(Bianchietal., 1996;Bourgeoisetal.,2011;Reussetal.,2005;Tselepides etal., 2000)and even feweron Baltic sediments (Bianchi etal.,2002a,b;Kowalewska,1997;Kowalewskaetal.,2004;
Reuss etal.,2005;Savageetal.,2010;Szymczak-Żyłaand Kowalewska,2007),despitethefactthateutrophicationand hypoxiawereidentifiedasproblemsinthisseaalreadymany yearsago(Conleyetal.,2009;HELCOM,2007).
Pigmentconcentrationsinsedimentsdependondifferent factors,associatedwith(1)primaryproductionandsedimen- tation,(2)pigmentstability and(3)post-depositionalcon- ditionsinsediments.Pigmentdegradealreadyinthewater columnandafterdepositioninthesedimentsasaresultof senescence,oxidation,herbivoregrazingorbacterialdegra- dation(e.g.Bianchietal.,1988;Loudaetal.,1998,2002;
Spooneretal.,1994a,b;Szymczak-Żyłaetal.,2006;Welsch- meyer and Lorenzen, 1985). The influence of particular factors on pigment content may bedifferent at different sites, so it is not an easy task to compare the extent of eutrophicationindifferentareasbasedonpigmentproxiesin sediments, or to make judgements about eutrophication trends(Leavitt,1993;Reussetal.,2005).
The aimofthis workwas tocompare eutrophicationin different water basins,exemplifiedby theGulf of Gdańsk (southern Baltic) and the Oslofjord/Drammensfjord (Nor- way), andits trendsin each one.Thesetwo waterbodies differ in salinity, geomorphologyand the extent of water mixing,butbothexperiencelimitedexchangeofwaterwith theadjacentseaandbothare subjectto humanpressure.
The aimwas realized byanalysingthe pigmentcontent in recentsedimentsinrelationtoenvironmentalconditionsin thenear-bottomwateraswellassedimentcharacteristics, including accumulation rate, sediment mixing, grain size distribution,carbonand nitrogencontent,i.e. parameters andfactorsassociatedwitheutrophication.
2. Material and methods
2.1. Studyareas2.1.1. GulfofGdańsk
The Gulf of Gdańsk (Fig. 1, area 4940km2) is part of the southern Baltic Sea (Majewski, 1990). The adjacent sedimentscalculatedperdryweightofsedimentsisavaluablemeasureofeutrophication,providing thatthemonitoringsite isselectedproperly,i.e.sedimentsarehypoxic/anoxicand non-mixed.
Besides,theresultsconfirmpreviousobservationsthatthepercentagesofparticularchlorophyll-a derivativesinthesumofchloropigments-aareuniversalmarkersofenvironmentalconditionsina basin.Theratiosofchloropigments-bandchlorophylls-ctothesumofchloropigments-a(SChlns-b/
SChlns-a;Chls-c/SChlns-a)maybyappliedascomplementarymarkersof freshwaterand marine organicmatterinput,respectively.
#2016InstituteofOceanologyofthePolishAcademyofSciences.ProductionandhostingbyElsevier Sp.zo.o.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.
org/licenses/by-nc-nd/4.0/).
Gdańsk-Sopot-Gdyniaconurbation(>1.2millioninhabitants) hasaconsiderableanthropogenicinfluenceon theGulf.In addition,theGulfofGdańskannuallyreceivessome39km3 (average flow rate 1240m3s1) of freshwater from the River Wisła (Vistula) (Pastuszak and Witek, 2012), which correspondsto13%ofthetotalvolumeofGulfofGdańsk waters.ThecatchmentareaoftheRiverWisłacoversabout 54%oftheareaofPoland(170000km2)andisinhabitedby almost60%ofthecountry'spopulation(27%oftheBaltic catchmentarea'spopulation), andtheriveritself accumu- latespollutantsfrom industrial,municipal andagricultural effluents;nutrientsareespeciallyimportant.Thenutrients and organic matter entering the Gulf with Wisła waters causes high primary production, which, together with the limited water exchange in the Gulf of Gdańsk, results in eutrophication(IMGW,2013;Witeketal.,1999).
ThedepthoftheGulfofGdańsk(average—59m;max- imum—118m)increasesseawardsfromtheshallowcoastal zone.The salinity oftheGulf's surfacewaters variesfrom 4.5neartheWisłamouthto8inthenorthern,deeppart ofthebasin.Initsdeeperregions(theGdańskDeep),vertical stratificationofthewateroccurs,resultinginahaloclineata depthof60—80mthatseparatesthemoresalinedeepwater (12.5)fromthelesssalinewateratthesurface.Asvertical waterexchangeislimited,thebottomwaterinthedeepest areasishypoxic/anoxic(Conleyetal.,2009;IMGW,2013).
The oxygenconditions improve during intermittent inflow eventsofcoldandwell-oxygenatedseawaterfromtheNorth SeathroughtheDanishStraits.Suchstronginflowsintothe Baltictakeplaceonce everyfewyears,usually duringlate autumnorwinter(HELCOM,2013;MälkkiandPerttilä,2012;
Mohrholz et al., 2015). The hydrological conditions and bottomtopographystronglydifferentiatethisshallowbasin asfarassedimentcharacteristicsareconcerned.Thus,the
sedimentsintheGulfvaryfromcoarsesandsnearthecoastto siltyclayintheGdańskDeep.Thisandthelackoftidesmake the Gulf an exceptional natural model basin for studying eutrophicationina marineenvironmentstrongly impacted byfreshwaterinput(IMGW,2013).
2.1.2. Oslofjord/Drammensfjord
TheOslofjord(southernNorway)isanapproximately100km long northward extension of the Skagerrak (Fig. 1). It is dividedintoaninnerandouterfjordbyanarrowsoundwith a sill at 19.5m water depth (Drøbak Sound). The inner Oslofjord consists of two main basins: the Vestfjord and theBunnefjord,bothwithamaximumwaterdepthofabout 160m,separatedbyasillatabout50mwaterdepth.Both fjord basins contain a number of smaller, semi-enclosed basins. The bottom topography of the Oslofjord restricts deeper-water exchange and renewal in the inner fjord.
Thewatermassesarestratifiedwithbrackishsurfacewater andmarinebottomwater.Deepwaterrenewalstakeplacein winterwithstrongnortherlywinds(Hessetal.,2014).The innermicrotidalarea,whichbordersthemostdenselypopu- latedandindustrializedareainNorway,hasreceivedlarge amountsofwastewatersandnutrients,particularlyduring thelastcentury.Asaconsequenceofthatandthelimited water-exchange between the different basins, oxygen- depletedbottomwaterconditionshavedevelopedinseveral basins (Dale et al.,1999).The nutrientload to the inner Oslofjordreachedamaximumaround1970;sincethenithas decreasedconsiderably,andtheoxygenconditionsareslowly improvinginmostbasins(Hessetal.,2014).
TheouterOslofjordisconnectedtotheDrammensfjord, whichhasalengthof20kmandawidthof1.6—3.0km.Itis separated from the Greater Oslofjord by a sill at Svelvik, whichwasdredgedfrom6to8mdeptharound1900,andto Figure1 Locationofthesamplingsites:(a)GulfofGdańskand(b)Oslofjord/Drammensfjord.
10min1951(Smittenbergetal.,2005).Sincethe1800s,the redoxclineinDrammensfjordhasmovedtoshallowerwater depths following the increased influx of organic material from thepulp and paper industry,and from an increasing populationand moreintensive agriculture in the drainage area (Alve,1991). Oxygen depletion was first detected in 1899,andthepresenceofH2SinJune1933(Öztürk,1995;
Smittenberg etal., 2005). An incursion of oceanicwaters currently occurs once every 3—5 years, mainly between November andMay, displacing someof the anoxic bottom watersupwards(Alve,1995a;Richards,1965).Theshallow- estpositionoftheredoxclineoccurredinthelate1970sto early1980s,whenitlayat30—35mwaterdepth(Magnusson and Næs, 1986). In 1988, following the closing down of industry and the implementation of governmental regula- tions,itwasrecordedatabout35mwaterdepthinthenorth andabout60min thesouth with anoxic sediments inthe deeperpartsandoxicsedimentsintheshallowerparts(Alve, 1995b).Recentstudiesconductedbylocalauthorities(Mon- itoringofDrammensfjord2008—2011,NGI,2010)showthat the deep-water dissolved oxygen levels increased from anoxic to a period of oxic conditions following the 2004/
2005-dredgingofthesillto12mdepth.Themixingoffresh- waterfromtheRiverDrammen(drainagearea17000km2, averageflowrate300m3s1),whichentersthefjordatits head,givesrisetoabrackishsurfacewaterlayer(salinity1—
10,dependingontheseason)thatisseparatedfromsaline bottomwater(30.5)below40mdepth.Riverregulationhas
smoothedtheannualfreshwatersupplytothefjordoverthe last 60 years, reducing the effect of spring flooding and increasingthewintersupply.Thishasshiftedtheminimum freshwatersupplyfromwintertolatesummer,thusincreas- ingtheresidencetimeofthesurfacewaterduringsummer (Smittenbergetal.,2005).
2.2. Samplecollection
SedimentswerecollectedatsixstationsintheGulfofGdańsk and six stations in the Oslofjord/Drammensfjord (Fig. 1;
Table1).Thestations intheGulfofGdańskwereselected soastocoverawiderangeofenvironmentalconditionsinthe studyarea,i.e.differentwaterdepth,salinity,oxygencon- centration,sedimenttypeanddistancefromthecoastline.
Thefourstationsselected—P110,P116,M1andP1—were positionedalongthewayofWisławaterinflowintheGulfas farastheGdańskDeep.Theothertwostationswerelocated inPuckBay,theshallow,westernpartoftheGulf(Table1):
stationBMPK10inthemiddleoftheBayandP104closetothe townof Hel,arecreationaland sportscentre,nearinten- sivelyusedshippinglanesandstrongwatercurrentsaround thetipoftheHelPeninsula.
The stations in the Oslofjord/Drammensfjordwerealso selectedsuchastoobtainmanydifferentlocationsconcern- ingdistancefromtheopensea,salinity,oxygenconcentra- tionandhumanpressure(Fig.1;Table1).Twostationswere locatedinDrammensfjord:AinthemiddleofthefjordandB
Table1 Characteristicsofthesamplingstations.
Station Coordinates Water depth[m]
Parametersofnear-bottomwater Sediment accumulation rate[cmy1]
Sedimentmixing depth[cm]
Salinity Temp.[8C] Oxygen[mgL1] GulfofGdańsk
P1 54850.0420N 19819.6830E
112 12.0 6.3 3.4 0.160.01 0
M1 54844.9120N 19817.6620E
95 11.7 6.4 3.9 0.160.01 0
P116 54839.0910N 19817.5750E
92 10.8 6.4 0.5 0.140.01 0
P110 54829.9860N 19806.9020E
72 8.6 5.0 5.4 0.170.02 3
BMPK10 54833.5450N 18840.9500E
31 7.5 4.9 11.1 Noaccumulationor
max0.070.01
12 P104 54834.9440N
18847.3700E
55 7.6 4.5 12.1 Noaccumulation 12
Oslofjord/Drammensfjord A 59841.2760N
10822.7450E
113 31.2 8.0 0.3 0.110.01 0
B 59838.8620N 10824.8040E
122 31.0 8.2 0.5 0.270.02 0
C 59845.0660N 10834.4290E
154—158 32.3 8.3 9.2 0.200.02 4
D 59847.3860N 10843.1540E
152 32.6 9.2 1.7 0.100.03 5
E 59850.6430N 10843.5570E
77 33.2 8.5 0.2 0.180.01 0
F 59851.4700N 10841.7100E
78 33.6 8.3 1.7 0.050.01 3
nearerSvelvik.Theotherfourstationswerelocatedinthe innerOslofjord:stationCinthedeepest,southernmostpart oftheVestfjord,andstationsD,EandFintheBunnefjord.
StationDwasinthedeepestandsouthern-mostpartofthe Bunnefjord,stationEinitsshallowerpart,andstationFwas closesttotheportofOslo.
Thesediments werecollectedduringtwo cruisesofr/v 'Oceania':inApril2014(intheGulfofGdańsk)andinJune 2014(intheNorwegianfjords).Sedimentsamplesweretaken withaNiemistöcoresamplerintheGulfofGdańskandwitha GEMAXtwin-core samplerintheNorwegianfjords;in both cases the core diameter F=10cm.Eight cores were col- lected at each station. All the cores were photographed immediatelyaftercollection(Appendix1).Aftercollection thefollowinglayersweretakenfromeachsedimentcore:0—
1,1—5,5—10,10—15and15—20cm;sub-samplesfromallthe corescollectedatthesamestationwerepooled.Theaddi- tionalcorefor210Pbanalysiswasdividedintothinner,1cm thicklayersfrom0to10cmand2cmthicklayersfrom10to 30cmsedimentdepth.Allthesub-samples werefrozenon boardimmediatelyaftercollection.
2.3. Physicochemicalseawater parameters
The salinity, temperature and oxygen concentration were measuredwithaSBE19probe(verticalprofilesinthewater column)andwith aProfiLine Multi197iWTW meter(near- bottomwater).
2.4. Analyses
2.4.1. Pigmentanalysis
Theconcentrationsofpigmentsweredeterminedinallsam- pleswith HPLCusing a proceduredescribed indetail else- where(Kowalewskaetal.,1996;Szymczak-Żyłaetal.,2008).
2.4.1.1. Pigmentextractionfromsediments. Afrozensedi- mentsample(3—5g)wasplacedinaglasscentrifugetubeand leftto thaw.Aftercentrifugation(10min,2500rpm)water wasremovedandthesampleflushedwithacetone,stirred, sonicated(2—3min),centrifugedagain,andtheextractdec- anted.Theextractionwasrepeateduntilthesupernatantwas colourless(max.3times).Theacetoneextractsweretrans- ferredtoaseparatefunnelinwhichliquid—liquidextraction wasperformedintheacetoneextract:benzene:watersystem.
Thebenzenelayerwasthentransferredto aglassvialand evaporatedtodrynessinastreamofargonandstoredfrozen (208C) until HPLC analysis. The extracted sediment was driedat608Candweighed.Thepigmentcontentwascalcu- latedperdrysedimentweight.
2.4.1.2. HPLCpigmentanalysis. Thesedimentextractpre- paredasabovewasdissolvedinacetoneandinjectedintothe HPLCset(Knauer,Germany)withtwodetectors:diodearray (DAD 2800 Knauer) and fluorescence detector (RF-20Axs, Shimadzu,Japan),autosampler(KnauerOptimas,Germany), then into a Lichrospher 100RP-18 endcapped column (250mm4mm,5mm;Merck,Germany) through aguard column (Lichrospher 100RP-18 endcapped, 4mm4mm;
Merck,Germany).
Thefollowingpigmentsweredetermined:chloropigments- a (chlorophyll-a and its derivatives: pheophorbides-a,
pyropheophorbides-a,chlorophyll-a-allomers,chlorophyll-a- epimer,pheophytin-a,pheophytin-a-epimer,pyropheophytin- aandsumofsterylchlorinesters);chloropigments-b(chlor- ophyll-b andpheophytin-b); sum of chlorophylls-c; carote- noids (fucoxanthin, alloxanthin, diatoxanthin, lutein, zeaxanthin,cantaxanthin,echinenoneandb-carotene).
Separationsofchloropigments-aand-bwerecarriedout usinganHPLC/DADset,intheA(acetone):B(80:20,acetone:
water,v/v)gradientsystemat aflowrateof1.0mLmin1 . The mobile phase and gradient system was a modified versionofthatusedbySzymczak-Żyłaetal.(2008).Absorp- tion spectra were measured over the 360—750nm range.
Pigmentconcentrations were determinedaccordingto the proceduredescribedbySzymczak-Żyłaetal.(2008).
Chlorophylls-cwereanalyzedusinganHPLC/FLset,inthe A(acetone):B(80:20,acetone:water,v/v)gradientsystemat a flow rate of 0.5mLmin1. The excitation and emission wavelengthswere440and630nm,respectively.Quantita- tivedataofchlorophylls-ccontentwereobtainedaccording totheproceduredescribedbyKowalewskaetal.(1996).
CarotenoidseparationswerecarriedoutusinganHPLC/
DADset,intheA(85:15,methanol:0.5Mammoniumacetate, aq.v/v):B(90:10,acetonitrile:water,v/v):C(ethylacetate) gradientsystemataflowrateof1.0mLmin1.Themobile phase andgradient systemwas a modifiedversion ofthat usedbyChenetal.(2001).Carotenoidconcentrationsinthe sampleswerecalculatedinthesamewayasthoseforchlor- opigments-aandb(Szymczak-Żyłaetal.,2008).
Pigmentswereidentifiedonthebasis ofretentiontime andabsorbancespectra compared withpigment standards (DHI,Denmark).
2.4.2. Additionalanalyses
2.4.2.1. 210Pbanalysis—sedimentaccumulationrate. The
210Pbdatingmethod(Goldberg,1963)wasusedtodetermine thesedimentaccumulationrate.Sedimentsamplesfor210Pb datingwerefreeze-driedandgroundinthelaboratory.Sedi- ment moisture and porosity were calculated. The 210Pb activityconcentrationwasmeasuredindirectlybythealpha spectrometry counting of its daughter nuclide 210Po (Zaborska et al., 2007). In brief, sediment samples were spikedwith209Po(chemicalyieldtracer)anddigested.Polo- nium isotopes were spontaneously deposited onto silver discs.Thesewereanalyzedfor210Poand209Poactivitycon- centrationinamulti-channelanalyser(Canberra)equipped withSi/Lidetectors.Thesamples werecountedfor 1day.
Theactivityconcentration of210Poinasamplewas deter- minedon thebasisofchemicalrecoverybycomparingthe measuredandspikedactivityconcentrationof209Po.Blanks andstandardsweremeasuredtoverifytheefficiencyofthe separation procedure and detection. Standard reference materials(IAEA-326)wereusedtoverifythemeasurements.
Oneblanksample(withoutthesediment)wasmeasuredwith every7 sedimentsamples. The environmental background wasnegligible.Thelinearaccumulationrate(LAR,cmy1) wascalculatedassuminganexponentialdecreasein210Pbex
withsedimentdepth(Zaborskaetal.,2007).
2.4.2.2. Grain size analysis. The fine fractions from sam- plingstations BMPK10 andP104andthe sedimentsamples fromthe otherstations wereanalyzedby laserdiffraction usingaFritschLaserParticleSizerAnalysette-22(Kramarska
etal., 1996)and recorded at a resolutionof 1f. Sodium pyrophosphatewasusedtopreventaggregatesformingdur- ingmeasurement. Becausethesedimentsamples fromthe Oslofjordcontainedcarbonates,theywerepre-treatedwith 10%HCl. Allsedimentsamplesweretreatedwith30%H2O2
beforeanalysisinordertoremoveorganicmatter.Sediments from stations BMPK10 and P104 were first passed wet (Myślińska, 1992) through a sieve of mesh diameter of 0.063mm.
2.4.2.3. Carbon and nitrogen analyses. Organic carbon (Corg),totalnitrogen(Ntot),stablecarbon(d13C)andnitrogen (d15N)isotopeanalysesweredoneinaFlashEA1112Series ElementalAnalyzercombinedwithanIRMSDeltaVAdvantage Isotopic RatioMass Spectrometer (ThermoElectron Corp., Germany).Dry,homogeneoussamplesofthesedimentswere weighed(2—4mgfortheBalticSeasedimentsamples,20—
25mgfortheOslofjord/Drammensfjordsedimentsamples) intosilvervialsandacidifiedwith2MHCl(Changetal.,1991;
HedgesandStern,1984).TheCorgandNtotconcentrationsare statedas percentagesof thebulk of thedry sample after removalofcarbonates.Qualitycontroloftheorganiccarbon measurementswascarriedoutwithstandards(ThermoElec- tronCorp.). Theaccuracyandprecision(averagerecovery 99.12.0%)ofthemethodologyweresatisfactory.Isotopic ratiosd13Candd15Nwerecalculatedusinglaboratoryworking pure reference gases CO2 and N2 calibrated against IAEA standards:CO-8 andUSGS40for d13C andN-1 and USGS40 ford15N.Thed13Cresultsaregivenintheconventionaldelta notation,i.e.versusPDBford13Candversusairford15N.
2.5. Statistical analysis
The results were statistically processed using STATISTICA 12.5software(StatSoft,Poland):correlationanalysis,cluster analysisandprincipalcomponentanalysis(PCA)wereused.
Non-parametricmethods(e.g.R-Spearmancorrelationana- lysis) were applied to cases where the basic conditions necessaryfor using parametric methods werenot fulfilled (tested with the Shapiro—Wilk and Brown-Forsyth tests).
Correlationanalysiswas usedtoevaluatetherelationships betweenthepigmentcontents,pigmentratiosinthesedi- mentand the environmental parameters. Acorrelation of
p<0.05wasregardedassignificant.Clusteranalysis(Ward's method,Euclideandistance)wasusedtoproduceaclassifi- cationofthesamplingstationstakingintoaccountsediment characteristics(pigmentcontent,grainsize,organiccarbon andnitrogencontent) andnear-bottomwaterparameters.
Relationships between the content of pigments, ratio of particularpigmentsinthesedimentsamplesandothermea- suredparameterswerealsocheckedusingPCA.
3. Results
3.1. Pigmentdistribution
Thehighestpigmentconcentrationswerefoundinsediments collectedfromtheGdańskDeep,wheretheconcentrationof e.g. SChlns-a in the surface (0—1cm) layer of sediments rangedfrom350nmolg1d.w.(dryweightofsediment)at station P110 to 800nmolg1 d.w. at station P1 (Fig. 2).
RelativetotheGdańskDeeparea,thesedimentsfromPuck Bay(stationsBMPK10,P104)containedmuchloweramounts of pigments (e.g. SChlns-a in 0—1cm layer 80nmolg1 d.w.). From the Oslofjord/Drammensfjord, only the sedi- ments from the Bunnefjord were rich in pigments, where theconcentrationofSChlns-ainthesurface(0—1cm)layer ofsedimentsrangedfrom130nmolg1d.w.atstationFto 320nmolg1d.w.atstationE(Fig.2).Aconsiderablylower pigment content was determined in Drammensfjord (35nmolg1 d.w. at station Aand 60nmolg1 d.w.at stationB)andinVestfjord(stationC35nmolg1d.w.).
The concentrationrangesof parentchloropigments and carotenoidsinthesurface(0—1cm)sedimentlayerwereas follows:4—227nmolg1d.w.(chl-a),0.2—15nmolg1d.w.
(chl-b), 0.2—8.5nmolg1 d.w. (chls-c), 3.5—387nmolg1 d.w. (fuco), 0.5—45nmolg1 d.w. (diato), 1—82nmolg1 d.w. (lut) and 0—256nmolg1 d.w.(b-car)(Table 2). The profilesofallpigmentconcentrationsresembledthatofthe sumofchloropigments-a.
Pigmentconcentrationsweredistinctlyhigherinthesur- face(0—1cm)layerthaninthedeepersedimentlayersofthe GulfofGdańsk(Fig.2).Thiswasnotobservedinthesamples fromtheOslofjord/Drammensfjords,wherethesurfacelayer pigment concentrations are comparable to or even lower thaninthedeeperlayers.
Figure2 Averagecontents[nmolg1d.w.]ofthesumofchloropigments-a(SChlns-a);n=2.
3.2. Physicochemicalparametersofseawater
AllthestationsintheOslofjord/Drammensfjordwerechar- acterizedbyamuchhighersalinitybothinthewatercolumn andin thenear-bottom water,and thehalocline was at a shallowerdepth,thanintheGulfofGdańsk(Table1,Appen- dix2).Thetemperatureofthenear-bottomwaterwashigher andmoreevenlydistributedintheNorwegianfjordsthanin theGulfofGdańsk.IntheGulfofGdańskhighertemperatures wererecordedatthestationsintheGdańskDeepthanatthe shallowerstations,inPuckBay.Therewasanoxiainthenear- bottomwateratonestation(P116,oxygenconcentration— 0.5mgL1)intheGdańskDeep,whilethreeotherstationsin theGdańskDeep(P1,M1andP110)hadlowoxygenconcen- tration (3.5—5.4mgL1). Only two shallow-water stations (BMPK10andP104)hadahighoxygenconcentration.Allbut oneofthestationsintheNorwegianfjordsexhibitedanoxia/
hypoxia.The exception was stationC (Vestfjord) with the higheroxygenconcentration(9.2mgL1)(Table1).
3.3. Sedimentcharacteristics
3.3.1. 210Pb—sedimentaccumulationrates
210Pbtot in the surface (0—1cm) layer varied significantly dependingonsamplingregion—from82to650Bqkg1 intheGulfofGdańskandfrom120to280Bqkg1inthe Oslofjord(Appendix 3).The 210Pbtotvalues for theGulf of Gdańskcoresconcurwiththoserecentlyreportedbyseveral authors (SuplińskaandPietrzak-Flis,2008;Zaborskaetal., 2014;Zalewskaetal.,2015).The210PbtotresultsforNorwe- gianfjordsedimentsarealsocomparabletothoseobtained byotherauthors(DolvenandAlve,2010;Smittenbergetal., 2005;Zegersetal.,2003).
Mostofthe210Pbtotactivityconcentrationprofilesshowed anexponentialdecreasealongthecore,althoughsediment mixingand/orextremelylownetsedimentationwasappar- entatsomestations.The210PbtotprofilesofNorwegianfjord sediments (C, D, F) indicated disturbances in the surface sediments(3—5cm)causedbyhumanoranimalactivityand/
orbycurrents(Table1;Appendix3).Similar210Pbtotprofiles werereported byDolven andAlve(2010) andSmittenberg
etal.(2005)fortheOslofjordsediments.Accordingtothose authorsthesesedimentsexhibitednumerousdisturbancesas wellasmixinginboththeupperandlowerpartsofthecores.
Inthefjords(thiswork)therewasnomixingatthreestations:
A,BandE.IntheGdańskDeepnosedimentmixingwasfound tohaveoccurredatthreestations(P1,M1,P116),whilethere wasslight(3cm)mixingatstationP110andintensivemixing at the shallow-water stations BMPK10 and P104 down to 12cmdepth(Table1).
Thestations intheGdańskDeep werecharacterizedby intermediatelinearaccumulationrates(LARs)rangingfrom 0.14cmy1 (station P116) to 0.17cmy1 (station P110) (Table 1). The LAR was not calculated for station P104 in PuckBay,sincethe210Pbtotactivityconcentrationprofiledid not decrease with depth there. At station BMPK10, the
210Pbtot profile showed very little decrease with depth, anditwas difficult to calculatetheLAR (no accumulation atall orno more than0.07cmy1).The lasttwo stations werelocatedinrelatively shallowareas(30—50m),where thecoarsersedimentfractionprevailedasaresultoflocal currentstransferringfineparticulatemattertodeeperareas.
LARs ranging from 0.05cmy1 (station F) to 0.18cmy1 (stationE)weremeasuredforeasternOslofjord(Bunnefjord) sediments. Station C in the western Oslofjord (Vestfjord) displayed the second highest LAR of 0.20cmy1. In the Drammensfjord,theLARswere0.11cmy1(stationA)and 0.27cmy1(stationB)(Table1).
The LARs obtained in this study for the Gdańsk Deep stations(0.14—0.17cmy1)agreewiththevaluesreported for this region (0.1—0.24cmy1, Pempkowiak, 1991;
Suplińska and Pietrzak-Flis, 2008; Zaborska et al., 2014;
Zalewskaetal.,2015).Fewdataonsedimentaccumulation ratesintheOslofjordhavebeenreported.PauandHammer (2013)estimatedaccumulationratesinthenorthernpartof theVestfjord. TheyreportLARsfrom 0.04to 0.18cmy1, dependingonthebottomdepth(withhigherratesindeeper areas). Their results generally agree well with the LARs obtained in this work (from 0.5 to 0.20cmy1). Dolven andAlve(2010)andDolvenetal.(2013)studiedbothparts oftheOslofjord.Theyfoundverylarge andvariableaccu- mulationratesof0.1—0.3cmy1inthenorthernBunnefjord andevenlargerratesofca0.4cmy1inits southernpart.
ExtremelylargeLARsof1.3—2.5cmy1andveryhighfluxes of 210Pb have been reported for the southern Vestfjord (Dolven andAlve, 2010).In the Drammensfjord,sediment accumulationratesfrom0.15to0.25cmy1weremeasured bySmittenbergetal.(2005)andHuguetetal.(2007),which isinagreementwiththeresultsofthiswork.
3.3.2. Grainsize
The grainsize fractions are presented as thesumof sand (>0.063mm), silt (from 0.063 to 0.004mm) and clay (<0.004mm).Thesedimentsrichestinsandwereatstations P104andBMPK10 (Table3; Appendix4).The silt fractions werethelargestatthestationsintheGdańskDeepandthe easternOslofjord(Bunnefjord).Thecontentofthesmallest grain size (<0.004mm) fraction was the highest in the DrammensfjordandtheVestfjord(stationC).
3.3.3. Carbonandnitrogen
Organic carbon concentrationsin sediments of the Gdańsk Deep were by far the highest of all the samples studied Table2 Averagecontentsofselectedpigments[nmolg1
d.w.]inthesurface(0—1cm)sedimentlayer;n=2.
Station Chl-a Chl-b Chls-c Fuco Diato Lut b-car GulfofGdańsk
P1 227.4 14.8 8.5 386.6 45.0 81.7 255.8 M1 164.3 12.3 5.7 204.1 36.2 76.2 207.9 P116 194.3 13.1 6.7 268.3 34.9 67.3 184.7 P110 176.9 7.4 6.5 144.4 12.8 29.1 67.6 BMPK10 35.3 1.4 1.8 27.6 1.4 4.3 14.3
P104 32.9 1.4 1.8 41.1 4.4 3.4 13.1
Oslofjord/Drammensfjord
A 9.3 1.1 0.2 3.5 0.5 4.6 8.0
B 22.3 1.6 0.5 12.6 3.5 10.0 19.5
C 4.3 0.2 0.3 8.9 0.9 1.3 0.0
D 83.9 5.1 3.1 79.3 23.0 24.5 89.0
E 79.4 5.3 5.1 90.8 34.3 27.2 119.6
F 32.5 1.5 2.3 23.5 18.3 21.3 104.1
(5—8%)(Table3;Appendix4).Thelowestconcentrationswere inPuck Bay (BMPK10, P104) (1—3.4%). The organic carbon contentinthe samplesfrom theNorwegian fjords were in betweenthosevalues:thehighestcontentwasforstationsin theBunnefjord—from1.5to4.5%atDandE,and7%inthe1—
5cmlayerofsediments atF.The valueswerelowerinthe Drammensfjord (1.4—2.5%). The d13C values measured for sedimentcores ranged from 26.9% to 20.5%.The dis- tinctly higher values were measured for four (C, D, E, F) sedimentcorescollectedfromtheOslofjords(from22.3%
to 20.5%). The d15N values measured for the Baltic Sea sedimentsrangedfrom2%to3.7%,whilefortheNorwe- gianfjordsedimentsfrom1.6%to3.5%(Appendix4).
4. Discussion
4.1. Environmentalconditionsandthepigment recordinsediments
Chloropigments-acontentinrecentsedimentsisagood,well- documentedindicatorofproductivityinbothlacustrineand marinesediments(e.g.Kowalewskaetal.,2004;Leavittand Hodgson,2001;Loudaetal.,2000;Szymczak-ŻyłaandKowa- lewska,2007;Villanueva and Hastings,2000).The method basedondeterminingchloropigments-ainsediments(calcu- latedperdryweightofsediment,notnormalizedtoorganic carbon)yieldstheaverageeutrophicationpictureforanarea.
Thepigmentcontentsinthesedimentsdiscussedinthispaper (Fig.2,Table2)areinaccordancewithpreviousstudiesofthe GulfofGdańsk.Szymczak-Żyłaetal.(2011)reportedthatthe averagesum of chloropigments-a(SChlns-a) in theGdańsk Deepsedimentswas400nmolg1d.w.(in0—1cm).Inthat workthehighestvalue(900nmolg1d.w.)wasrecordedin May2003.Thiswasaconsequenceoftheintensealgalbloom (Chl-a20mgm3)thathadtakenplaceinApril2003(IMGW, 2009).Thehighvalueofpigmentsdeterminedinthesurface (0—1cm)sediment layer(SChlns-a 800nmolg1 d.w.) of
the Gdańsk Deep in April 2014 (this paper) indicates that productivity in thisarea is still high.The concentration of pigmentsinsedimentsdependsnotonlyonprimaryproduc- tionbutalsootherfactorsinfluencingsedimentation,hydro- logicalanddepositionalconditions.Takingintoaccountthe concentrationofpigments(SChlns-a)inthesurface(0—1cm) sediment layer and sediment characteristics, including organiccarbonandtotalnitrogencontent,grainsizedistribu- tion andnear-bottomwater parameters (salinity, tempera- tureandoxygenconcentration),hierarchicalclusteranalysis (dendrogramofthesamplingstations—Fig.3a)showedthat sedimentsintheGdańskDeep(stationsP1andP110,M1and P116)differfromthoseattheotherGulfofGdańskstations, i.e.thoseinPuckBay(BMPK10,P104).Furthermore,theGulf ofGdańskstationsdiffersignificantlyfromthoseintheOslof- jord/Drammensfjord area (Fig. 3a). Principal Component Analysis(PCA)wasappliedtochecktheseresultsandtofind themostsignificantfactoraffectingthepigmentconcentra- tioninthesedimentsofthestudyareas(Fig.3bandc).The PCAdatamatrixmodelexplainsover90%ofthetotalvaria- tionswiththefirsttwoprincipalcomponents.Theconsider- ably lower pigment content in the Puck Bay sediments correspondswiththegoodoxicconditionsinthenear-bottom waterandthehighcontentofsandfraction(Fig.3bandc), whichisduetothestrongwatercurrentsinthisarea.Ahigh positiveandsignificantcorrelationofSChlns-awiththeper- centageoforganic carbon(r=0.92,p<0.05)andtheper- centageofthefinesedimentfraction<0.063mm (r=0.88, p<0.05)wasobtainedforallthesamplesfromtheGulfof Gdańsk. Shankle et al.(2002) has observed that sediment grainsizeisanimportantfactoraffectingpigmentcontentin sediments.Thepigment-richGdańskDeepsedimentscontain large quantities of organic carbon, total nitrogen and silt fraction (Fig. 3b and c). Anoxia in the near-bottom water andsedimentspreventsthedecompositionoforganicmatter andpreservespigmentsintheGdańskDeepsediments.The pigmentconcentrationcorrelatesnegativelywiththeoxygen contentinthenear-bottomwater(r=0.9,p<0.05),which
Table3 Sedimentcharacteristics:grainsize,organiccarbon(Corg),totalnitrogen(Ntot),stablecarbon(d13C)andnitrogen(d15N) contentinthesurface(0—1cm)sedimentlayer.
Station Grainsize[%] Corg[%] Ntot[%] d13C[%] d15N[%]
Sand
>0.063mm
Silt
0.063—0.004mm
Clay
<0.004mm GulfofGdańsk
P1 0.8 75.8 23.4 8.13 1.14 25.3 3.72
M1 0 73.5 26.5 6.90 0.94 25.4 2.51
P116 0 73.2 26.8 7.22 1.02 25.2 2.21
P110 1.4 76.5 22.1 6.65 0.98 25.2 2.52
BMPK10 53.9 35.4 10.7 3.36 0.37 25.2 2.46
P104 64.1 28.7 7.2 1.62 0.22 25.2 3.52
Oslofjord/Drammensfjord
A 0 51.9 48.1 2.03 0.15 25.9 2.75
B 0 54.8 45.2 1.60 0.15 25.2 1.23
C 0 54.1 45.9 2.97 0.23 21.7 1.02
D 0 67.5 32.5 2.58 0.33 22.0 2.29
E 0 63.8 36.2 1.58 0.19 22.0 2.99
F 0 63.9 36.1 3.74 0.33 21.5 2.32
isinagreementwiththeobservationsofotherauthorsthat oxygen is one of themost significant factors affecting the pigment concentration in sediments (Bianchi et al., 2000;
Reussetal., 2005;Villanuevaand Hastings,2000). Gulfof Gdańsk sediments differ significantly from the Oslofjord/
Drammensfjordsediments(Fig.3aandc),whichhaveahigher clayfractioncontentandhighernear-bottomwatersalinity andtemperature. The near-bottom watersof mostof the Norwegianfjordareasstudiedhere,exceptVestfjord,were anoxic/hypoxic (Table 1). In the Oslofjord/Drammensfjord areas studied, only the sediments of the Bunnefjord were richin pigments, whilein theDrammensfjord,despite the anoxic/hypoxicconditions,therewerefarfewerpigmentsin thesediments(Fig.2,Table2).Thisisindicativeofthisarea's lowproductivity, or thata large proportion of pigmentsis degradedtocolourlessproductsinthewatercolumn.Allthis demonstratesthatchloropigments-ainsurfacesedimentsare areflectionoftheeutrophicationofthebasin.
Pigmentconcentrationsaredistinctlyhigherinthesurface (0—1cm)layer,whichmighthavebeenformedinthepast5—7 years,thanindeepersedimentlayersoftheGulfofGdańsk (Fig.2).Thistallieswithourpreviousobservations(Szymczak- Żyła and Kowalewska, 2007) and those of other authors (Stephens etal., 1997;VillanuevaandHastings,2000) that the pigment content decreases the fastest in the surface sedimentlayer.Inthedeeperlayerspigmenttransformation proceeds much more slowly, especially in anoxic basins.
However, this was not observed in the samples from the Oslofjord/Drammensfjord,wherepigmentconcentrationsin thesurfacelayerarecomparabletoorevenlowerthaninthe deeperlayers(Fig.2).Thissuggestsdecreasingprimarypro- ductionandthatconditionsintheNorwegianfjordsaremore propitioustopigment preservationinsediments.The topo- graphyofthefjordsbeingwhatitis(seeSection2.1.2),the pigmentsremainundecomposedformanyyears.IntheGulfof Gdańskverticalmixingandbottom-waterexchangeislimited butnotwithinthesamerangeasinthefjords.
Sedimentsdifferednotonlyinthecontentofpigmentsbut alsointhepercentageofparticularchlorophyll-aderivatives intheirsum.OurpreviousobservationsfromGulfofGdańsk studiesshowed thatthepercentageofparticularsedimen- tarychlorophyll-aderivativesmaybetakentobemarkersof syn-andpost-depositionalenvironmentalconditions(Szymc- zak-Żyłaetal.,2011).Chlorophyll-aallomersarecharacter- istic of sediments originating from an oxygenated coastal zone.Chlorophyll-aandpheophorbides-aindicate thepre- senceof comparativelyfreshmaterial(Loudaetal.,1998, 2002).Pyropheophorbides-aaremainlyamarkerforgrazing by zooplankton and/or zoobenthos (Bianchi et al., 1998, 2002a, b; Head and Harris, 1996; Szymczak-Żyła et al., 2006).Finally,pyropheophytin-aandthe sterylderivatives occurmainlyinanoxicsediments(Chenetal.,2001;Louda etal.,2000;Shankleetal.,2002;Szymczak-Żyłaetal.,2011;
VillanuevaandHastings,2000).Basedontheseobservations, we named indicators of processes ('grazing') and environ- mentalconditions('oxic'and'anoxic')(Fig.4a—c).Therewas a high (almost 20%) percentage of 'grazing' chlorophyll-a derivatives(pyropheophorbides-a)inthePuckBaysediments (stations BMPK10 and P104), which corresponds with the sedimentmixingresults.The210Pbtotprofileinthesediments indicated intensive mixing at the shallow-water Puck Bay stationsdownto12cmdepth(Table1).Thepigmentresults Figure3 Results of the statistical analysis: (a) hierarchical
dendrogram of sampling stations (cluster analysis — Ward's method, Euclidean distance) based on characteristics of the surface (0—1cm) sediment layer: chloropigments-a content (SChlns-a),organiccarbon(Corg)andtotalnitrogen(Ntot)con- tent, grain size (%sand, %silt, %clay) and near-bottom water parameters (salinity,temperature and oxygen concentration);
(b) scatter plot of principal component loadingby individual variables;(c)scatterplotofprincipalcomponentobjectscores bysamplingsites.
presented in this paper suggest that the sediments there wereprobablydisturbedbyanimalactivity.IntheNorwegian fjordsthesedimentsatstationsC,D,Fwerealsomixed.The highpercentageof'grazing' derivativesat stationsCandF suggests that benthic activity is responsible for sediment mixing in these areas. Indeed, a largebiomass of benthic animals(Polychaetes)wasobservedalreadyduringsampling atstationF(Appendix1).However,atstationA,wherethe sedimentswerenotmixed,thehighpercentageof'grazing' derivatives suggests intense zooplankton activity in the water column. Two stations with mixed sediments (P110 andD)didnothaveahighproportionof'grazing'indicators.
This can be explained by mixing by abiotic factors (e.g.
currents)atthesesamplingsites.
The maximum percentage of chlorophyll-a-allomers, whichform under oxic conditions, ('oxic' indicator) was in thesedimentsfrom thePuckBay stationsandin Vestfjord (stationC),whilederivativescharacteristicofanoxiccondi- tions (pyropheophytin-a and the steryl chlorin esters — 'anoxic'indicator)werefoundinsedimentsfromtheGdańsk DeepandtheremainingNorwegianfjordstations(A,B,D,E, F)(Fig.4bandc).Thisisindicativeofoxic/anoxicconditions inthenear-bottomwater(Table1).
4.2. Freshwaterand seawaterinfluences
Eutrophicationoccursincoastalareasimpactedbyinorganic nutrient loads, by nutrients bound to organic matter and restrictedwaterexchange.Thetraditionalproxiesusedfor determiningorganicmattersourcesarethecarbon-to-nitro- genratio(CN1)andthestableisotopesofthese twoele- ments (d13C, d15N) (Fontugne and Jouanneau, 1987;
Maksymowska etal.,2000; Szczepańska etal., 2012),but therangescharacteristicofdifferentorganicmattersources arebroadandfrequentlyoverlap(Cravotta,1997;Schulzand Zabel,2006).Theuseofthesethreeproxiesforsediments oftenyields differentinformation,sothatdeterminingthe originof organicmatteris difficult.The useof d15Nas an indicator forsediments is particularly ambiguous,because nitrogenisfractionatedinthetrophicnetworkratherthan duringphotosynthesis.Thetypicalnitrogenisotopiccompo- sition (d15N) of marine phytoplankton in temperate seas variesfrom3.0%to12.0%.Freshwaterphytoplanktoniso- topicsignaturesmentionedintheliteraturehaved15Naround 5%(SchulzandZabel,2006).
TheCN1resultsforfive(P1,M1,P116,P110,P104)ofthe sixsedimentcoresfromtheGulfofGdańsk(rangefrom8to 10)indicateamixedmarineandterrigenousoriginofthe sedimentaryorganicmatter(Szczepańskaetal.,2012).The exceptionisstationBMPK10,forwhichthemeasuredvalues arehigher(from10.5to13.4),indicatingamoreterrigenous origin.Thed13Cvalues(from25.7%to25%)indicatethat theorganicmatterinallthecoresconsistsofamixtureof terrestrialandmarinematter.Thisisinagreementwiththe conclusiondrawn from theCN1 ratio anddatapublished earlierfor thisarea (Szczepańska etal.,2012;Vossetal., 2000).Thed15NvaluesmeasuredfortheBalticSeasediments rangedfrom 2% to 3.7%. Thestable nitrogen signature recordedinthisstudyissimilartothatofVossetal.(2000), whonotedthatthissignatureisindicativeofmarinephyto- plankton.ThisobservationthuscontradictstheCN1ratio andd13Cresults.
TheCN1 ratiosencounteredinthecoresfromtheNor- wegianfjordsrangedfrom9to19.Thesevalueswereclose tothoseobtainedbyotherauthorsforthesamearea(Smit- tenbergetal.,2005).Thehighestvaluesweremeasuredfor thesedimentcorecollectedfromstationAandareindicative of the terrigenous provenance of the organic matter. The CN1ratioforthesedimentcoresfromstationsBandCranges from 12.6 to 14.8, indicating thatthe organic matter is a mixtureofterrestrialandmarinematerial.Thelowestratios (from9to12)wereforsedimentsatstationsDandEandwas closertothevaluesforfreshlydepositedmarinephytoplank- ton (Kenney et al., 2010). The d13C values measured for Figure4 Averagepercentageofparticularchlorophyll-aderi- vativesinthesumofchloropigments-a:(a)pyropheophorbides-a ('grazing'indicator);(b)chlorophyll-a-allomers('oxic'indicator);
(c) sumofpyropheophytin-a andsterylchlorinesters('anoxic' indicator);n=2.
sedimentcorescollectedfromtheOslofjord(StationsC,D,E, F)(from22.3%to20.5%;Appendix4)pointtothemarine provenanceoftheorganicmatter.Thevaluesforthesediment coresfromDrammensfjord(AandB)arelower,from26.9% to25.0%,whichindicatesthattheorganicmatterinthose sediments isamixture ofterrestrialandmarine materials.
The results for sediment cores A and Bare similar to the resultsobtainedbyotherauthors(Huguetetal.,2007;Smit- tenberg et al., 2005), but differ from the conclusion for stationAbasedontheCN1ratio.Thed15Nvaluesmeasured fortheNorwegianfjordsedimentsrangedfrom1.6%to3.5%
(Appendix4),indicatingthattheorganicmatterthereorigi- natedfromphytoplanktonorterrestrialorganisms(Schulzand Zabel,2006).All thecoresdisplayagreatvariationofd15N with depth, demonstrating the varying origin and fate of organicmatter,especiallyintheNorwegianfjord.Generally, inourworkd15Nvaluesaresmallerin thefjord sediments, especiallyintheDrammensfjord,thanintheGulfofGdańsk sediments;thiscanbeexplainedbythedifferentmainsources oforganicmatterforthesetwowaterareas,maybebecause ofthehigherinputofsewage(Cravotta,1997)orundecom- posedmacrophyta(Bucholcetal.,2014;Kenneyetal.,2010) whichcharacterizethehigherd15N,intheGulfofGdańsk.
Havingthe above inmind, the chlorophyll-bandchlor- ophylls-ctosumofchloropigments-aratios(Chl-b/SChlns-a;
Chls-c/SChlns-a)werevalidatedasmarkersofriverineand marine provenance of organic matter, respectively. These ratios depend on the initial concentrations of the parent (Chl-bandChls-c)pigmentsinthematterundergoingsedimen- tationandtheirstability.Eventhoughchlorophyll-band-care less stablethanchlorophyll-a (Leavittand Hodgson,2001), thesemarkerscanenrichourknowledgeabouttheoriginof organicmatter.PreviousstudiesofsouthernBalticSeasedi- mentshaveshownthattheratiosofchlorophylls-cand-bto chlorophyll-adependsontheproportionsofdiatomsandgreen algae in the total Baltic phytoplankton. Chlorophyll-c/
chlorophyll-aratioswerehigherinsamplesfrommarineareas (Kowalewska et al.,1996). Theauthors suggestedthat the aboveratiosinthesedimentcanbeusedasanindicatorfor freshandmarineorganicmatterintheadjacentwaters.
TheresultsofthisworkindicatethattheChl-b/SChlns-a ratiowashigherintheGdańskDeep(stationsP1—P116)and Drammensfjord(stationsAandB)thanintheothersediments studied (Fig. 5a). Both areas are strongly influenced by riverinewater.TheGdańskDeepisasinkfororganicmatter carriedbytheRiverWisłaintotheGulfofGdańsk(Jankowski
Figure5 Pigmentratios:(a)chlorophyll-btochloropigments-aratio(Chl-b/SChlns-a);(b)sumofchloropigments-btochloropig- ments-aratio(SChlns-b/SChlns-a);(c)percentageofluteininthesumofcarotenoids(%lut);(d)chlorophylls-ctochloropigments-a ratio(Chls-c/SChlns-a);(e)percentageoffucoxanthininthesumofcarotenoids(%fuco);(f)percentageofdiatoxanthininthesumof carotenoids(%diato).
andStaśkiewicz,1994),whileDrammensfjordaccumulatesin sedimentsfreshwaterorganicmatterfromtheRiverDram- men(Fig.1).Thehighestchl-b/SChlns-aratiowasrecorded insedimentsatstationA(Drammensfjord),locatedcloseto themouthoftheriver,wherebloomsofgreenalgaeBotryo- coccusbrauni(Kützing)wereobserved(Smittenbergetal., 2005).Thehighcorrelationbetweenthepercentageoflutein in the sum of carotenoids (%lut) and chl-b/SChlns-a (r=0.7and0.87,p<0.05)fortheGulfofGdańskandthe Norwegian fjords, respectively, showed that even though luteinis more stable thanchlorophyll-b andis recognized asamarkerofgreenalgae(LeavittandHodgson,2001),the Chl-b/SChlns-a ratio is also a good marker of green algal biomass,whichismoreabundantinfreshwaterenvironments thanthatofdiatomsanddinoflagellates(BellingerandSigee, 2010).Insedimentsampleswithalargeproportionofchlor- ophyll-bderivatives,suchastheGulfofGdańsk(Fig.5band c),thehighercorrelationbetween%lutandtheSChlns-b/
SChlns-a(thesumofchlorophyll-banditsderivativestothe sumofchloropigments-a)ratio(r=0.98,p<0.05)suggests thatitisanevenbettermarkerthanchl-b/SChlns-a.
The highestChls-c/SChlns-a ratio was recorded at sta- tions BMPK10 and P104 in the Gulf of Gdańsk and in the Oslofjordsamples(stations C—F)(Fig. 5d).Thereis ahigh correlationbetweenChls-c/SChlns-aandthepercentageof fucoxanthin(thepigmentoccurringmainlyindiatoms(Jef- freyandVesk,1997))inthesumofcarotenoids (%fuco)for theGulfofGdańsksediments(r=0.92,p<0.05)andahigh correlationwiththepercentageofdiatoxanthin(%diato)for theOslofjord(r=0.87,p<0.05)(Fig.5eandf).Thissuggests thatthediatomsarethemainsourceofchlorophylls-cinthe Gulfof Gdańsk sediments, while in the Oslofjord,besides diatoms,also othermarinespeciescontainingdiatoxanthin suchas dinoflagellates.Theexceptionthere was stationC (Vestfjord)withthehighpercentageoffucoxanthin(60%, Fig. 5e). Both phytoplankton groups(diatoms and dinofla- gellates)containnotonlychlorophyll-abutalsochlorophylls- c(JeffreyandVesk,1997).Thesegroupsofphytoplankton, though present in both sea- and freshwater, occur in the greatestproportionsinwatersofseasandoceans,especially inthetemperatezones(SverdrupandArmbrust,2008).The Chls-c/SChlns-aratioinsediments,proposedinthispaper, couldbeamarkerofseawaterinfluences.
PrincipalComponentAnalysis (PCA)wasapplied tovali- date these results. The proxies for the 0—1cm sediment layer,boththatproposedinthiswork(SChlns-b/SChlns-a;
Chls-c/SChlns-a)andthetraditionalone(CN1;d13C;d15N, asadditionalvariables),wereconsidered(Fig.6aandb).The PCAdatamatrixmodelexplainsover90%ofthetotalvaria- tions with the first two principal components. The first principalcomponentexplains66%ofthevariationsandis strongly correlatedwith four of thevariables. Itincreases withincreasingofSChlns-b/SChlns-aratioandthepercen- tageofluteininthesumofcarotenoids,andwithdecreasing ofChls-c/SChlns-aratioandthepercentageoffucoxanthin.
Thesecondprincipalcomponent(27%ofthetotalvariance) increaseswith increasing of percentage ofdiatoxanthin in thesumofcarotenoids.Thegreatestinputtoorganicmatter inthesedimentsatthestationsinDrammensfjord(stations A,B)isfromfreshwaterorganisms,containingchlorophyll-b (greenalgaeandhigherplants).Thegreatestinputtoorganic matter in the sediments at the Oslofjord stations is from
marineorganismscontainingchlorophylls-c,suchasdiatoms anddinoflagellates(Fig.6aandb).Thed13Cvaluesmeasured forsedimentcorescollectedfromtheOslofjordareinagree- mentwiththesepigmentmarkers.Theresultsofthiswork have confirmed the previous observations for the Gulf of Gdańsk(Kowalewskaetal.,1996).Chls-c/SChlns-amaythus be asimple marker of seawater organicmatter input and SChlns-b/SChlns-a amarker of freshwaterorganic matter Figure6 Resultsofthestatisticalanalysis:scatterplotof(a) principalcomponentloadingbyindividualvariables;(b)princi- palcomponentobjectscoresbysamplingsitesbasedonproposed indicatorsofthesourceoforganicmatter(SChlns-b/SChlns-a;
Chls-c/SChlns-a), percentagesofparticularcarotenoidsinthe sumofcarotenoids(%fuco,%lut,%diato)andtraditionalproxies asadditionalvariables(CN1;d13C;d15N)forthe0—1cmsedi- mentlayer.
input, complementing the traditional proxies, which are oftenambiguousandcontradictory,especiallyd15N(Kenney etal.,2010;SchulzandZabel,2006;Torresetal.,2012).
4.3. Eutrophicationtrends
Sixstationswereselectedfortrackingeutrophicationtrends inthe twobasins.These werestationswith unmixedsedi- mentsaccordingtothe210Pbanalysis(Table1):threeinthe GdańskDeep(P1,M1,P116),twointheDrammensfjord(A andB)andoneintheOslofjord(E).Lookingatthesumof chloropigments-aprofilesforthesestations(Fig.7a),onecan seethatthereisatemporalincreaseinthechloropigments-a contentinthesedimentsatalltheGdańskDeepstations;this isparticularlyobviousinthesurfacesediments,whichwere formedinthelast30—40years.Thepigmentcontentswere averaged for0—5cmbasingon weightedmeanfor0—1cm and1—5cmlayers.Thisisinagreementwithotherobserva- tionsthateutrophicationofthesouthernBalticincreasedin the1970s(Fleming-Lehtinenetal.,2015).IntheOslofjord/
Drammensfjordmaximumofchloropigments-acontentcor- respondstodeeperlayers(A,5—10cm;B,10—15cm;E,5—
10cm),formedwhenthenutrientloadwasatitshighestin the1970s (Hessetal.,2014).Thisindicates adecrease in eutrophicationinthepast30—40years.Moreover,onecan seethatallo-chlorophyll-aderivatives forminginoxic con- ditions (Fig. 7b) take minimum values in the sediments accumulatedduringthe1970s,andingeneralarelower in
theGulfofGdańskthanintheOslofjord/Drammensfjord.The derivativespreservedinold sediments—'anoxic'indicators (pyropheophytin-aandsterylchlorins)—haveoppositepro- filestoallo-chlorophyllsandingeneralaremoreabundantin sediments of the Gulf of Gdańsk than in the Oslofjord/
Drammensfjord, although the chloropigments-a at station Eare therichest inthese compounds(Fig. 7c), indicating ahighlevelofanoxia.Thesumofchloropigments-anormal- izedto theorganiccarboncontent (Fig. 7d)demonstrates similar trends to those of chloropigments-a calculated to dried weight of sediments, with the exception of station E,wherethereisanincreaseinthesurface0—5cmlayerin comparison with the next 5—10cm. This difference is obvious,asnormalization of thechloropigments-a content mirrorsastageinthedecompositionofpigments,nottheir contentinthesediments.Theb-caroteneprofilesaresimilar tothoseforchloropigments-a(Fig.7e),whichindicatesthat chloropigments-a,despitebeinglessstablethanb-carotene (LeavittandHodgson,2001),aregoodmarkersfortracking eutrophicationtrends.
5. Conclusions
Comparisonofpigmentsinthesedimentsinthetwo basins showsthateutrophicationintheGulfofGdańskhasincreased overthelast30—40yearsbuthasdecreasedintheOslofjord/
Drammensfjordduringthesameperiod.Theresultspresented in this paper prove that the sum of chloropigments-a in Figure 7 Pigment content profiles for selected non-mixed sediments: (a) SChlns-a [nmolg1 d.w.]; (b) the percentage of chlorophyll-a-allomersinthesumofchloropigments-a('oxic'indicator);(c)percentageofthesumofpyropheophytin-aandsteryl chlorinestersinthesumofchloropigments-a('anoxic'indicator);(d)SChlns-a[nmolg1Corg];(e)b-carotene[nmolg1d.w.].
