This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
AbstractQuestions:Weinvestigatedthechangingroleofclimate,forestfiresandhumanpopu-lationsizeinthebroad-scalecompositionalchangesinHolocenevegetationdynamicsbeforeandaftertheonsetoffarminginSweden(at6,000calyrBP)andinFinland(at4,000calyrBP).Location:SouthernandcentralSweden,SWandSEFinland.Methods:HoloceneregionalplantabundanceswerereconstructedusingtheREVEALSmodelonselectedfossilpollenrecordsfromlakes.Therelativeimportanceofclimate,firesandhumanpopulationsizeonchangesinvegetationcompositionwasassessedusing variationpartitioning. Past climate variablewasderived from the LOVECLIMclimatemodel. Fire variablewas reconstructed from sedimentary charcoal records.Estimatedtrendinhumanpopulationsizewasbasedonthetemporaldistributionofarchaeologicalradiocarbondates.Results:Climateexplainsthehighestproportionofvariationinvegetationcomposi-tionduringthewholestudyperiodinSweden(10,000–4,000calyrBP)andinFinland(10,000–1,000calyrBP),andduringthepre-agriculturalperiod.Ingeneral,firesex-plainarelatively lowproportionofvariation.Humanpopulationsizehassignificanteffectonvegetationdynamicsaftertheonsetoffarmingandexplainsthehighestvari-ationinvegetationinSSwedenandSWFinland.Conclusions:Mesolithichunter-gathererpopulationsdidnotsignificantlyaffectveg-etationcompositioninFennoscandia,andclimatewasthemaindriverofchangesatthat time.Agricultural communities,however,hadgreatereffectonvegetationdy-namics,andtheroleofhumanpopulationsizebecameamoreimportantfactorduringthelateHolocene.Ourresultsdemonstratethatclimatecanbeconsideredthemaindriveroflong-termvegetationdynamicsinFennoscandia.However,insomeregionsthe influenceofhumanpopulation sizeonHolocenevegetationchangesexceededthatofclimateandhasalongevitydatingtotheearlyNeolithic.
Pollen-based reconstructions of Holocene vegetation dynamicshavedemonstratedtheinfluenceofanthropogenicactivityonfor-estcompositionandlandscapeopennesssincethemid-HoloceneinCentralandNorthernEurope(Fyfe,Woodbridge,&Roberts,2015;Marquer etal., 2014, 2017). Fennoscandian pollen data demon-stratetheregionaldifferencesinforestcompositionthroughouttheHolocene, and thisvariation has been generally connected to thechangesinclimate(Birks,1986;Heikkilä&Seppä,2003;Milleretal.,2008).Inaddition,fireisanimportantdisturbancefactorinborealforests that profoundly affects forest age, structure, compositionandsuccessiondynamics (Bradshaw,Lindbladh,&Hannon,2010).AlthoughMesolithichunter-gatherersmayhavealteredlocal-scalevegetation composition through fires and favouring food plants(Hörnbergetal.,2005;Regnell,Gaillard,Bartholin,&Karsten,1995),the regional scalevegetation change in Fennoscandia presumablyremained under natural drivers longer than in other European re-gions.Thecurrentintensiveanthropogenicinfluenceandpredictedacceleratedwarming of the boreal biome (Christensen, Kumar, &Aldrian,2013)maycausecompositionalandstructuralchangestothe presentvegetation.Therefore, understanding the complex in-teractionsbetweennatural-andhuman-inducedchangesinthepastregionalvegetationdynamics can shed light on the future effectsofachangingclimateonecosystemsthatareheavilyinfluencedbyhumanactivity.
The strongest effect of human impact on vegetation inFennoscandia is connected to the onset of agriculture, which wasoftenaccompaniedby increasedhuman-induced fires (Granström&Niklasson 2008; Huttunen, 1980). The earliest signs of agriculturearefoundatca.6,000calibratedyearsbeforepresent(calyrBP)inSSweden(Sørensen&Karg,2012;Welinder,2011).InSWFinlandthefirstunambiguousevidenceofcultivationisdated4,000–3,500calyrBP (Lahtinen,Oinonen,Tallavaara,Walker, & Rowley-Conwy, 2017;Taavitsainen,Simola,&Grönlund,1998).InEFinland,earliestsignsofcereal cultivation inpalynological recordsare reportedalready fromthe early Neolithic ca. 6,400–5,200cal yr BP (Alenius, Mökkönen,Holmqvist, & Ojala, 2017; Alenius, Mönkkönen, & Lahelma, 2013),whereasfirstarchaeologicalevidenceofcultivation,alongsideincreas-ingpalynologicalevidence,isdetectedafter3,200calyrBP(Lavento,2001;Taavitsainenetal.,1998).
Studies concerning the role of human impact and climate onvegetation changes during themid- and lateHolocene inNEuropehavebeenmainlyqualitative,basedon interpretationofpollendataorarchaeologicalfindings.Recently,Reitaluetal. (2013),Kuosmanenetal. (2016)andMarqueretal. (2017)haveaddressed thisquestionbymeansofvariationpartitioning,usingdifferentproxiestoassesstheroleplayedby the anthropogenic influence.Reitalu etal. (2013) de-rived thehuman impactvariable frompollen records,Marqueretal.(2017) used the anthropogenic land-cover change scenario (ALCC;Kaplanetal.,2010),whileKuosmanenetal. (2016)utilizedahumanpopulation sizeproxy,basedon the temporal frequencydistribution
Here we employ this independent proxy of human populationsize together with climate and regional fires to statistically assesstheir relative importance on the variation in the Holocene vegeta-tioncomposition inFennoscandiaduring three timeperiods: (1) thewholestudyperiodinSweden(10,000–4,000calyrBP)andFinland(10,000–1,000calyrBP); (2)beforetheonsetoffarminginSweden(at10,000–6,000calyrBP)andinFinland(at10,000–4,000calyrBP);and (3)after theonsetof farming inSweden (at6,000–4,000calyrBP)andinFinland(at4,000–1,000calyrBP).ThelengthofthestudyperiodsbetweenSwedenandFinlanddiffersduetothelimitationsofdataonhumanpopulationsize.
2 | METHODS
2.1 | Study area
ThestudyareaislocatedinFennoscandiaandisdividedintofourre-gions;(1)southern(S)Sweden,(2)central(C)Sweden,(3)southwest(SW)Finlandand (4)southeast (SE)Finland (Figure1).Thenorthernlimitof thestudy region inSweden is locatedat61.5°N,while theborderbetweenSandCSwedenisat58.3°N.InFinland,thenorth-ern limitof thestudy region isat62.5°N,and theborderbetweenSWandSEFinland isat25.5°E (alongLakePäijänne).Thesestudyregionswerechosenbecausetheavailabilityofahighamountofdataforpollen-basedREVEALSestimates,humanpopulationsizedataandcharcoaldata.
2.2 | Regional plant abundances
Weusedpollen records from33 lakes selected from theEuropeanPollen Database (Fyfe etal., 2009; Giesecke etal., 2014), theLANDCLIMpollendataarchive(Marqueretal.,2017;Trondmanetal.,2015,2016)orprovideddirectlybydatacontributors(AppendixS1).An important criterion for the selected pollen records was robustchronological control and adequate Holocene time resolution. TheREVEALSmodel(Sugita,2007)wasrunseparatelyforeachstudyre-gionusingfivepollenrecordsforCSweden,17forSSweden,fiveforSEFinlandandsixforSWFinland(Figure1).
Withineachregion,theREVEALSmodelwasusedtoconvertpol-lenpercentagesofthe23mostimportantpollentaxaintoproportionalvegetationcover(AppendixS4)for200-yeartimewindowsfollowingMazieretal.(2012),Trondmanetal.(2015)andMarqueretal.(2017).Descriptions ofHolocene changes invegetation cover are depictedinFigure2andAppendixS4.Thestudyperiodcovers10,000–1,000calyrBPcorrespondingwiththetimeofavailablehumanpopulationsizedata.Additionallytotaxa-specificREVEALSestimates,nineplant
| 3Journal of Vegetation Science
KUOSMANEN Et Al.
functionaltypes(PFTs;AppendixS5),groupedintermsofbioclimaticlimits afterTrondman etal. (2015),were defined in order to assessbroad-scaletrendsindifferentplantecosystemsduringtheHolocene.
2.3 | Explanatory variables of Holocene vegetation changes
2.3.1 | Climate
Theclimatevariables for thestudy regionswereobtained fromtheLOVECLIMclimatemodelaspresentedbyZhang,Renssen,andSeppä(2016).Theversionappliedinthisstudyexplicitlyrepresentscompo-nents of the atmosphere, oceans (including sea ice) and vegetationin the climate system with intermediate complexity (Goosse etal.,2010).Thespatialresolutionofthemodelis5.6°×5.6°andthreegridcellswereused toprovide regional temperaturedata for the studyregions.SandCSwedenfellintothesamelatitudinalband,andthusthe climatedata are the same for these regions.The climate varia-bles includewinterDecember–January–February (DJF)andsummer
June–July–August (JJA)temperaturesexpressedas30-yearaverageand as anomalies from the preindustrialmean (550–250cal yrBP).Climateparameterswereaveragedover200-yearbinsforthestatisti-calanalyses(Figure2;AppendixS4).Precipitationdatawereexcludeddue to the largeuncertainty in the simulations.Theappliedclimatesimulationprovidesrealistictrendsonlargesub-continentalscale,es-peciallyforFennoscandia,asdemonstratedbyinter-modelandmodeldatacomparisons(Zhang,Renssen,Seppä,&Valdes,2017a,b).
2.3.2 | Forest fires
The regional fire variable was reconstructed based on sedimentarycharcoalrecordsfrom19lakes.Charcoaldatacoveringpartorallofthelast10,000yearswereacquiredfromthelatestversionoftheGlobalCharcoal Database (GCD v3; Marlon etal., 2016), from previouslypublishedsyntheses(Clear,Molinari,&Bradshaw,2014)orprovideddirectlybytheauthors(AppendixS2).ToretrieveregionalcompositesofchangesinfireactivityovertheHolocene,charcoalaccumulationinsedimentswastransformedandstandardizedaccordingtoaprotocol
F IGURE 1 Locationofthefourstudyregions.Theradiocarbon-datedarchaeologicalsitesforhumanpopulationsizeproxyareshownasgreycircles,siteswithpollenrecordaremarkedasgreencirclesandsiteswithcharcoalrecordaremarkedwithblackpolygons.Thecorrespondencebetweenthelabelofeachsite,thesitenameandthecharacteristicsofeachsitearegiveninAppendicesS1andS2
describedbyPower,Marlon,Bartlein,andHarrison(2010)andwidelyused within the paleofire community (Daniau etal., 2012; Molinarietal.,2013).Thesmoothcurveswereconstructedbydeterminingfit-tedvaluesat200-year intervals (Figure2;AppendixS4).Compositecurveswereproducedwith theRpackagepaleofire (Blarquezetal.,2014;RFoundationforStatisticalComputing,Vienna,Austria).
2.3.3 | Human population size
Changes in human population size were reconstructed using thetemporalfrequencydistributionofarchaeologicalradiocarbondates(Shennan&Edinborough,2007;Shennanetal.,2013).TheSwedishdata were extracted from the EUROEVOL data set that containsover 14,000 archaeological radiocarbon dates from the NeolithicinNWEurope,spanningtheLateMesolithicandEarlyBronzeAge(Shennan etal., 2013; Timpson etal. 2014; Manning, Colledge,Crema, Shennan, & Timpson, 2016). The Swedish samples weresupplementedbyMesolithicdates(Edinborough,2005)anddividedintosouthern(414dates)andcentral(352dates)subsetsaccordingtothestudyregions.DuetotheNeolithicfocusoftheEUROEVOLdata, the Swedish sampleswere restricted to cover the period of10,000–4,000cal yr BP. The Finnish data set (Oinonen, Pesonen,& Tallavaara, 2010; Tallavaara etal., 2010) covers the period of10,000–1,000calyrBPbothinSW(692dates)andSE(315dates)regions.
Summedprobabilitydistributionsofcalibratedradiocarbondates(SPDs)werecreatedforeachregion,andevaluatedforrandomsam-pling bias and calibration effects using a Monte-Carlo simulationapproach (Shennanetal., 2013;Timpsonetal. 2014) andby imple-menting the methodology used in Crema, Habu, Kobayashi, andMadella(2016;AppendixS6).Toreducethepotentialbiasofoversam-plingwithinsites,multipledates from individual sitesweregroupedintonon-overlappingbins,suchthatafterorderingthedatesatonesite,thenextdatewasonlyassignedtoanewbinifthereweremorethan200radiocarbonyearsbetweenitandthepreviousdate(Shennanetal., 2013;Tallavaara etal., 2010). Finally, SPDswere binned intoconservative200-yearbinsby averaging summedprobabilityvalueswithineachbin(Figure2,AppendixS4).
2.4 | Statistical analysis
Variationpartitioning (Borcard,Legendre,&Drapeu,1992)wasper-formed to assess the relative importance of climate, regional firesandhumanpopulationsizeonvariationinHolocenevegetationcom-position. In the analysis, vegetation composition derived from theREVEALSestimateswereusedasresponsematrix,whileclimate,re-gionalfiresandhumanpopulationsizewereusedasexplanatoryvari-ables. REVEALS estimates were Hellinger-transformed (Legendre &
Variationpartitioningwas first performed for thewhole studyperiod inallregions,namelyat10,000–4,000calyrBPinSandCSweden, and at 10,000–1,000calyrBP in SWandSEFinland. Inorder toevaluate theeffectof theexplanatoryvariablesonvege-tationbefore theonsetof farming,variationpartitioningwasper-formedfortheperiodof10,000–6,000calyrBPforCandSSweden,andfortheperiodof10,000–4,000calyrBPforSWandSEFinland.To assess the relative importance of the explanatoryvariables onvegetationchangesafterthebeginningofagriculture,variationpar-titioningwasperformedfortheperiodof6,000–4,000calyrBPforS andCSweden, and for theperiodof4,000–1,000calyrBP forSWandSEFinland.Toseparatelyevaluatetherelativeimportanceofsummerandwintertemperaturesonvegetation,variationparti-tioningwithsummerandwintertemperatureasindividualexplana-toryvariableswasperformedforallthreestudyperiods,thewholeHolocene,thepre-agricultureandtheagriculturalperiods,forallthestudyregions.
Inordertoassessthechangeintheroleofclimate,regionalfiresandhumanpopulation size during theHolocene, amovingwindowapproach(Reitaluetal.,2013)wasemployed.Thisapproachenablesperformingvariationpartitioning for subsetsofdata in shorter timeperiods, hence providing information on how the relative roles ofclimate and human population size change over time.Movingwin-dowanalysiswasperformedforallstudyregionsin2,000-yeartime windows,withten200-yearbinsineachtimewindow.
F IGURE 2 Holocenechangesintheexplanatoryvariables:simulatedsummer(SumT)andwinter(WinT)temperatures(fromtheLOVECLIMmodel),fires(charcoalinfluxz-scores)andhumanpopulationsizedataexpressedassummedprobabilitydistributions(SPD)ateachstudyregion(a)SSweden,(b)CSweden,(c)SWFinlandand(d)SEFinland(seeFigure1)andthemeanofREVEALSestimatesof23taxa
6 | Journal of Vegetation Science
KUOSMANEN Et Al.
3.1.2 | Pre- agriculture
Beforetheonsetoffarmingclimatealoneexplainedmostofthevari-ationinvegetation,withwintertemperaturesexplainingahigherpro-portionofvariation than summer temperatures (Figure3;AppendixS7). Fires had a significant effect on vegetation composition inFinland,explaining3%of thevariation invegetation inSWand4%inSEFinland.Humanpopulationsizeexplained5%of thevariationinvegetationinSWFinland,butwasnotsignificantinotherregions.Climate,forestfiresandhumanpopulationsizeexplainedatleast65%ofthevariationinvegetationcompositionbeforetheonsetofagricul-tureinallregions.
respectively;Figure3).Individually,climateexplainedarelativelylowfractionofvariationinvegetationinallregions,withsummertemper-aturedrivingahigherproportionofvariationthanwintertemperature(AppendixS7).Forest firesdidnotexplain individuallyanyvariationin vegetation.However, togetherwithhumanpopulation size, theyexplained 6% of variation in vegetation dynamics in SE Finland. Ingeneral,arelativelyhighproportionofvariationinvegetationwasleftunexplainedinallregions.
3.1.4 | Moving window approach
Results of the moving window approach (Appendix S9: Figure S1)demonstrate the strong impact of climate on vegetation dynamicsfrom the early to themid-Holocene in S andC Sweden, especiallybetween8,600–6,500cal yrBP,when climate explained40%–60%ofthevariationinvegetation.Inbothregionstherelativeimportanceofclimatewasnotably lowerat6,500–5,000calyrBP.Therelative
InFinland,theresultsofthemovingwindowapproach(AppendixS9:FigureS2)demonstratenotablefluctuations intheimportanceofclimateinbothregions.InSWandSEFinlandclimateexplained20%–60% of the variation during ca. 1,000-year periods around8,000,5,700and3,800calyrBPwithlowerimportanceinbetween.In SW Finland the explanatory value of fires was higher (20%)around6,000calyrBP,withlowervaluesduringthelateHolocene.In SE Finland fires explained the highest amount of variation invegetation during the early Holocene. In general, the importanceofhumanpopulationsizewas low,exceptaround5,000calyrBP,when it explainedover20%of thevariation invegetation. In SWFinlandtherelativeimportanceofhumanpopulationsizeincreased
from 3,500cal yr BP exceeding the explanatory value of climateby explaining 20% of the variation in vegetation (Figure4). In SEFinlandtherelativeimportanceofhumanpopulationsizeremainedlowduringtheHolocene.
4 | DISCUSSION
4.1 | Role of climate in Holocene vegetation dynamics
Theimportanceofclimateasadriverofvegetationchangewashigh-estduringthebeginningoftheHoloceneThermalMaximum(HTM)ca.8,000–7,000calyrBP,andagainaround6,000–5,000calyrBP,corresponding to themain temperature shiftsduring theHolocene.Thiscanbeseenespecially inSSwedenandSWFinland (Figure4).Asimilar trendofhigher relative importanceofclimate in theearlyHolocene is also suggested byMarquer etal. (2017; Figure4). ThestrongroleofclimateinvegetationdynamicsatthebeginningoftheHTMisinagreementwiththewell-establishednorthwardrangeshift
F IGURE 4 Relativeimportanceofclimate,firesandanthropogenicinfluenceonvariationinvegetationdynamicsthroughtheHoloceneinNEurope.Thefractionofvariationinvegetationexplainedbyhumanpopulationsize,regionalfiresandclimatein(a)SSwedenand(b)SWFinlandforten200-yearsubsetsofdatain2,000-yearmovingtimewindows.(c)Thefractionofvariationexplainedbyclimateandhumanimpactfor14subsetsofdatain1,200-yearmovingtimewindowinEstonia(Reitaluetal.,2013).(d)ThefractionofvariationexplainedbyclimateandlanduseaverageacrosssouthernScandinaviaandBalticcountries(Marqueretal.,2017).ForSSweden,SWFinlandandEstoniatheresultsofeachtimewindowareshownfromthemidpointofthetimeperiod.TheusedtimewindowsareshowninAppendixS9:TablesS1,S2
20 40 60 80 1000
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
Tim
e pe
riod
(Cal
yea
r BP
)
No data
20 40 60 80 1000
No data
Southwest FinlandSouth Sweden Estonia Scandinavia and Baltic
population sizeHumanClimate Fires% of RV variation explained
oftemperatetreetaxasuchasCorylus,Tilia,Ulmus and Quercusduringthisperiod (Milleretal.,2008;Seppäetal.,2015), andcanbeseenasanincreaseoftemperatebroad-leafforestcoverinFennoscandia.Thiscorrespondswiththepresentecologicalunderstandingthattem-peratureisoneofthemainlimitingfactorsforthenorthernlimitsofthe temperate tree species (Jackson, Betancourt, Booth, & Grayd,2009;Woodward,Lomas,&Kelly,2004).
Althoughtheimportanceofhumanpopulationsizeonvegetationdynamics increases from the early Neolithic in Sweden and duringthe late-Holocene (ca 3,500cal yr BP) in Finland, climate remainsanimportantdriverbehindvegetationcoverchanges,especiallyinCSwedenandSEFinland(AppendixS9:FiguresS1,S2).Similarly,there-sultsofReitaluetal.(2013)andMarqueretal.(2017)demonstratethestrongeffectofclimateonvegetationdynamicsinEstoniaandsouth-ernScandinaviaduringthelast1,000years.Ifthiscontrolofclimateonboreal forest compositionwill prevail, theprojected acceleratingfuturewarminginnorthernregions(Christensenetal.,2013)willactasthemaindriveroffuturechangesintheborealecosystems,atleastatregionalscale.
4.2 | Role of fires in Holocene vegetation dynamics
The change from a natural to amore human-induced fire impactonvegetationdynamicsisnotclearlydefinedinourdata.Ahigherinfluenceofforestfiresonvegetationcompositionduringthepre-agriculture periodwas detected for both Sweden and Finland. InSSweden,theslightincreaseintheroleoffiresduringthebegin-ningoftheagriculturalperiodcoincideswiththeincreasingeffectof human population size (Figure4.),while in C Sweden fires didnotshowanyeffectonvegetationcomposition.Itisplausiblethatduring this time human-induced fires were local fires, which didnothavenotableimpactonregionalvegetationcomposition.Sinceslash-and-burnpracticeswereoftenusedinconnectionwithearlyfarming (Alenius, Haggrén, Koivisto, Sugita, & Vanhanen, 2017;Huttunen,1980),thenotablylowimportanceofforestfiresonveg-etationduringtheagriculturalperiodinSWFinlandwassurprising.Although the charcoal dataused in this study reconstruct the re-gionalfirehistory,theunevendistributionofthesedimentarychar-coalrecords inthestudyregionsmightbiastheresults.However,changes in vegetation dynamics are partly explained by the jointeffectoffiresandhumanpopulationsize,whereaspre-agriculturefires and climate have a notable joint effect on variation in veg-etation.Therefore, it isplausiblethatthe increased jointeffectofhumanpopulationsizeandfiresindicatethechangefromnatural-tomorehuman-inducedfiresaffectingvegetationdynamicsduringthelateHolocene.
4.3 | Increasing anthropogenic influence from early Neolithic
The comparison between our results and those of Marquer etal.(2017) for S Scandinavia and the Baltic region demonstrate an in-creasingroleofhumanpopulationsizeonvegetationdynamicsfrom
themid-HoloceneinSSwedensuggestingincreasinglandusebythegrowingpopulation(Figure4).ItisplausiblethattherapidincreaseinhumanpopulationsizeinSSwedenwasduetosuitableclimatecon-ditions during theHTM (Tallavaara& Seppä, 2012) and the spreadofmore intensive land-usepracticesfromCentralEurope.Theshifttoagrariansocietygraduallychangedthelandscape,whenNeolithiccommunitiesstartedtomigrate inland.Broad-leaf forestswere firstusedaswoodlandpasturesandthenclearedforcultivation(Lagerås,1996;Sköld,Lagerås,&Berglund,2010).
The low human impact on vegetation dynamics in SW Finlanduntil3,500calyrBP is in linewith thesmallhumanpopulationsizeandwith the lateranthropogenic influenceonvegetation inFinlandcompared to S Scandinavia and theBaltic region (Fyfe etal., 2015;Marquer etal., 2017). The increasing importance of human impactin SWFinland correspondswith theonset of agriculture ca4,000–3,500cal yr BP (Taavitsainen etal., 1998), suggesting that humanpopulation sizewas themajordriverofvegetationdynamicsduringthelateHolocene,especiallyduringtheperiodof3,000–1,000calyrBP.Growingagrariancommunitiesclearedtheforestforpasturesandarableland.Cultivatedfieldswerecommonlyleftasfallowforafewyearsinordertomaintainsoilproductivity.ThismayhavecausedanincreaseintheproportionofBetula and Alnustogetherwithshrubsinthevegetationcover(Reitaluetal.,2013).
The lower importance of human population size on regionalvegetationchangeinCSwedenandSEFinlandcouldbeexplainedby the fact that these regionsweremore forested and less pop-ulatedduring the studyperiod.Alenius etal. (2013) showed that,forexample, inEFinlandforestclearancesandcultivation intensi-fiedonlyfromAD600onwards,coincidingwith increasinghumanpopulationsize.Duetothe landuplift,substantially largeareasofSEFinlandwereraisedrapidlyabovetheBalticsea levelafter thedeglaciation,incontrasttoSWFinland(Tikkanen&Oksanen,2002)andthereforeSEFinlandlacksthick,fine-grainedmineralsoillayers,such as clay and silts,whichprovide fertile soils for cultivation inSW Finland. It is plausible that, due to the lack of suitable areasformoreintensivefieldcultivation,humanpopulationsizeremainedrelativelylowinSEFinlandandanimalhusbandrytogetherwithspo-radicslash-and-burncultivationwerepracticed longer than inSWFinland(Taavitsainenetal.,1998).
4.4 | Issue of causality in the variation partitioning analysis
It is also important to note that variation partitioning as amethoddoesnot consider the causality, but only the co-variationbetweenvariablesthroughtime.Thecausalitybetweenclimateandvegetationisrelativelyclear,whilethecausalitybetweenchangesinvegetationand inhunter-gathererpopulation sizeand forest fires ismoredif-ficult todeduce.Hunter-gathererpopulationswere small and scat-tered,affectingmainlythelocal-scalevegetationdynamics(Hörnbergetal., 2005;Regnell etal., 1995). In Finland, thedecline in hunter-gatherersalongwiththedeclineoftemperateforestscoincideswiththe expansion of spruce and consequently the boreal ecosystem
| 9Journal of Vegetation Science
KUOSMANEN Et Al.
(Seppä etal., 2009). Boreal forests provide lower food abundancecompared to temperatemixed forests, and ithasbeenargued thatthisecosystemchangecausedthedeclineinhunter-gathererpopula-tionsize(Tallavaara&Seppä,2012).Itisalsoprobablethatinsteadofacausalrelationshipbetweenhumanpopulationsizeandvegetation,thesevariablesco-varybecauseofthechangesinclimate.Similarly,thecausalitymaybedifficulttodistinguishbetweenfiresandvegeta-tion.Ohlsonetal.(2011)showedthatfirefrequencyinFennoscandiadecreasedafterthespruceexpansion.However,itisstillspeculativeifspruceexpandedbecauseoffirereductionorthefiresdecreasedduetothespruceexpansion.Sincedisentanglingtheseissuesofcau-salitywithvariationpartitioningischallenging,futurestudieswouldbenefitfromtheuseofothermethodsthatprovidemoreinsightintothedirectionof the interactionbetween response andexplanatoryvariables.
5 | CONCLUSION
Hunter-gathererpopulationsdidnotsignificantlyaffectvegetationdynamics inFennoscandia.ClimatewasthemaindriverofchangeduringtheHoloceneandlarge-scaleshiftsinvegetationwerelikelydriven by climate.However,with relatively stable climatic condi-tions, theroleofclimatemaydecreaseandother factors, suchasdisturbances, site characteristics, competition and human impactmayhaveagreatereffectonvegetationcomposition.Agriculturalpopulations,however,highlyaffectedvegetationcomposition,andtheroleplayedbyhumanpopulationsizebecamemoreimportantduringthelateHolocene.Thereisaclearregionaldependencere-latedtohumanpopulationsize,i.e.theeffectofhumanimpactonvegetationchangewasnotablyhigherinSSwedenandSWFinland,where land use was more intensive, than in C Sweden and SEFinland.
Thevariationpartitioningapproachprovidesimportantinsightsintothedriversofpastvegetationdynamicsandprovidesasteppingstoneforfindingmoreprecisemethodstoassesstheprocessesbe-hindpastchangesinlong-termvegetationdynamics.TherelativelyhighimportanceofclimatethroughouttheHolocenesuggeststhat–evenafter theonsetof farming–climate, togetherwithhumanimpact, remained an important driver of changes in broad-scalevegetationdynamics.Althoughtheanthropogenicimpactmayhaveperiodically and regionally overruled the effect of climate alreadyfromtheearlyNeolithic,climatecanbeconsideredthemaindriverof long-termregionalvegetationdynamics inFennoscandiaduringtheHolocene.
ACKNOWLEDGEMENTS
Financial support was provided by the PEDECO project (number16-23183Y) funded by the Czech Science Foundation (GAČR) andEBOR project, funded by the Academy of Finland. We thank theLANDCLIMprojectcoordinatedbyM.J.Gaillardfordatacontributions.
Alenius, T., Haggrén, G., Koivisto, S., Sugita, S., & Vanhanen, S. (2017).LandscapedynamicsinsouthernFinlandduringtheIronAgeandtheMiddleAgesusingpollenbasedLandscapeReconstruction(LRA)andmacrofossildata.Journal of Archaeological Science: Reports,12,12–24.https://doi.org/10.1016/j.jasrep.2016.11.041
Alenius,T.,Mökkönen,T.,Holmqvist,E.,&Ojala,A.(2017).NeolithiclanduseinthenorthernBorealzone:High-resolutionmultiproxyanalysesfromLakeHuhdasjärvi,south-easternFinland.Vegetation History and Archaeobotany,26,495–486.https://doi.org/10.1007/s00334-017-0606-2
Alenius,T.,Mönkkönen,T.,&Lahelma,A.(2013).Earlyfarminginthenorth-ernborealzone:Reassessing thehistoryof landuse insoutheasternFinland through high-resolution pollen analysis.Geoarchaeology, 28,1–24.https://doi.org/10.1002/gea.21428
Birks,H.J.B.(1986).Late-Quaternarybioticchangesinterrestrialandla-custrineenvironments,withparticularreferencetonorth-westEurope.InB.E.Berglund(Ed.),Handbook of Holocene palaeoecology and palaeo-hydrology(pp.3–65).Chichester,UK:Wiley.
Blarquez,O.,Vanniére,B.,Marlon,J.R.,Daniau,A.L.,Power,M.J.,Brewer,S.,&Bartlein,P.J.(2014).paleofire:AnRpackagetoanalysesedimen-tary charcoal records from the Global Charcoal Database to recon-structpastbiomassburning.Computers & Geosciences,72,255–261.https://doi.org/10.1016/j.cageo.2014.07.020
Borcard,D., Legendre,P.,&Drapeu,P. (1992).Partiallingout the spatialcomponent of ecological variation. Ecology,73, 1045–1055. https://doi.org/10.2307/1940179
Bradshaw,R.H.W.,Lindbladh,M.,&Hannon,G.E.(2010).TheroleoffireinsouthernScandinavianforestsduringthelateHolocene.International Journal of Wildland Fire, 19, 1040–1049. https://doi.org/10.1071/WF09108
Christensen, J. H., Kumar, K., &Aldrian, E. (2013). Climate phenomenaandtheirrelevanceforfutureregionalclimatechange,Chapter14.InT. F. Stocker,D.Qin,G.-K. Plattner, et al. (Eds.),The physical science basis: Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (pp. 1217–1308).Cambridge,UK:CambridgeUniversityPress.
Clear, J. L.,Molinari, C., & Bradshaw, R. H.W. (2014). Holocene fire inFennoscandiaandDenmark. International Journal of Wildland Fire,23,781–789.https://doi.org/10.1071/WF13188
Daniau, A. L., Bartlein, P. J., Harrison, S. P., Prentice, I. C., Brewer, S.,Friedlingstein,P.,…Zhang,Y. (2012).Predictabilityofbiomassburn-ing in response to climate changes.Global Biogeochemical Cycles,26,GB4007.
Edinborough, K. (2005). Evolution of bow-arrow technology. London,UK: University of London. Available from: http://discovery.ucl.ac.uk/1444653/1/U591962.pdf
Fyfe,R.M.,Beaulieu,J.L.,Binney,H.,Bradshaw,R.H.W.,Brewer,S.,Flao,A.,…Tonkov,S.(2009).TheEuropeanPollenDatabase:Pasteffortsandcurrentactivities.Vegetation History and Archaeobotany,18,417–424.https://doi.org/10.1007/s00334-009-0215-9
Giesecke, T., Basil Davis, B., Simon Brewer, S., Finsinger, W., Wolters,S., Blaauw, M., … Bradshaw, R. H. W. (2014). Towards mappingthe late Quaternary vegetation change of Europe. Vegetation History and Archaeobotany, 23, 75–86. https://doi.org/10.1007/s00334-012-0390-y
Goosse,H.,Brovkin,V.,Fichefet,T.,Haarsma,R.,Huybrechts,P.,Jongma,J.,…Weber, S. L. (2010).Description of the Earth systemmodel ofintermediate complexity LOVECLIM version 1.2.Geoscientific Model Development,3,603–633.https://doi.org/10.5194/gmd-3-603-2010
Granström,A.,&Niklasson,M.Granström&Niklasson (2008).Potentialsandlimitationsforhumancontroloverhistoricfireregimesinthebo-realforest.Philosophical Transactions of the Royal Society of London, se-ries B,363,2353–2358.
Heikkilä, M., & Seppä, H. (2003). A 11,000 yr palaeotempera-ture reconstruction from the southern boreal zone in Finland.Quaternary Science Reviews, 22, 541–554. https://doi.org/10.1016/S0277-3791(02)00189-0
Hörnberg,G.,Bohlin,E.,Hellberg,E.,Bergman,I.,Zackrisson,O.,Olofsson,A.,…Påsse,T.(2005).EffectsofMesolithichunter-gatherersonlocalvegetationinanon-uniformglacio-isostaticlandupliftarea,northernSweden. Vegetation History and Archaeobotany,15,13–26.
Jackson,S.T.,Betancourt,J.L.,Booth,R.K.,&Grayd,S.T.(2009).Ecologyand the ratchet of events:Climatevariability, nichedimensions, andspeciesdistributions.Proceedings of the National Academy of Sciences of the United States of America, 106, 19685–19692. https://doi.org/10.1073/pnas.0901644106
Lahtinen, M., Oinonen, M., Tallavaara, M., Walker, J. W., & Rowley-Conwy,P.(2017).TheadvanceofcultivationatitsnorthernEuropeanlimit: Process or event? The Holocene, 27, 427–438. https://doi.org/10.1177/0959683616660164
Lavento, M. (2001). Textile ceramics in Finland and on the Karelian isth-mus: Nine variations and fugue on a theme of C. F. Meinander, Suomen Muinaismuistoyhdistyksen aikakauskirja. Helsinki, FI: The FinnishAntiquarianSociety.
Lechterbeck,J.,Edinborough,K.,Kerig,T.,Fyfe,R.,Roberts,N.,&Shennan,S.(2014).IsNeolithiclandusecorrelatedwithdemography?Aneval-uation of pollen-derived land cover and radiocarbon-inferred demo-graphic change fromCentral Europe.The Holocene,24, 1297–1307.https://doi.org/10.1177/0959683614540952
Manning,K.,Colledge,S.,Crema,E.,Shennan,S.,&Timpson,A. (2016).The cultural evolution of Neolithic Europe. EUROEVOLDataset 1:Sites,phasesandradiocarbondata.Journal of Open Archaeology Data,5,2.
Marquer,L.,Gaillard,M.-J.,Sugita,S.,Poska,A.D.,Trondman,A.-K.,Mazier,F.,…Seppä,H.(2017).QuantifyingtheeffectsoflanduseandclimateonHolocenevegetation in Europe.Quaternary Science Reviews,171,20–37.https://doi.org/10.1016/j.quascirev.2017.07.001
Mazier, F., Gaillard, M.-J., Kunes, P., Sugita, S., Trondman, A.-K., &Broström, A. (2012). Testing the effect of site selection and pa-rameter setting on REVEALS-model estimates of plant abundanceusing the Czech Quaternary Palynological Database. Review of Palaeobotany and Palynology, 187, 38–49. https://doi.org/10.1016/j.revpalbo.2012.07.017
Miller,P.A.,Giesecke,T.,Hickler,T.,Bradshaw,R.H.W.,Smith,B.,Seppä,H., … Sykes, M. T. (2008). Exploring climatic and biotic controls onHolocenevegetation change in Fennoscandia. Journal of Ecology,96,247–259.https://doi.org/10.1111/j.1365-2745.2007.01342.x
Molinari, C., Lehsten, V., Bradshaw, R. H. W., Power, M. J., Harmand,P., Arneth, A., … Sykes, M. T. (2013). Exploring potential drivers ofEuropean biomass burning over the Holocene: A data-model anal-ysis. Global Ecology and Biogeography, 22, 1248–1260. https://doi.org/10.1111/geb.12090
Ohlson, M., Brown, K. J., Birks, H. J. B., Grytnes, J.-A., Hörnberg, G.,Niklasson,M.,…Bradshaw,R.H.W.(2011).InvasionofNorwaysprucediversifiesthefireregimeinborealEuropeanforests.Journal of Ecology,99,395–403.
Regnell, M., Gaillard, M.-J., Bartholin, T. S., & Karsten, P. (1995).Reconstruction of environment and history of plant use duringthe late Mesolithic (Ertebølle culture) at the inland settlement ofBökebergIII,southernSweden.Vegetation History and Archaeobotany,4,67–91.
Reitalu,T.,Seppä,H.,Sugita,S.,Kangur,M.,Koff,T.,Avel,E.,…Veski,S.(2013).Long-termdriversofforestcompositioninaboreonemoralregion: The relative importance of climate and human impact.Journal of Biogeography,40, 1524–1534. https://doi.org/10.1111/jbi.12092
Seppä, H.,Alenius, T., Bradshaw, R. H.W., Giesecke, T., Heikkilä,M., &Muukkonen,P.(2009).InvasionofNorwayspruce(Picea abies)andtheriseof theborealecosystem inFennoscandia. Journal of Ecology,97,629–640.https://doi.org/10.1111/j.1365-2745.2009.01505.x
Shennan, S., Downey, S. S., Timpson, A., Edinborough, K., Colledge, S.,Kerig,T., …Thomas,M.G. (2013). Regional population collapse fol-lowed initial agriculture booms in mid-Holocene Europe. Nature Communications,4,1–8.
Shennan, S., & Edinborough, K. (2007). Prehistoric population history:From the LateGlacial to the LateNeolithic inCentral andNorthernEurope.Journal of Archaeological Science,34,1339–1345.https://doi.org/10.1016/j.jas.2006.10.031
Sköld,E.,Lagerås,P.,&Berglund,B.E.(2010).Temporalculturallandscapedynamicsinamarginaluplandarea:Agriculturalexpansionsandcon-tractionsinferredfrompalynologicalevidenceatYttraBerg,southernSweden. Vegetation History Archaeobotany, 19, 121–136. https://doi.org/10.1007/s00334-009-0234-6
Sørensen, L., & Karg, S. (2012). The expansion of agrarian societies to-wardsthenorth:NewevidenceforagricultureduringtheMesolithic/Neolithictransition inSouthernScandinavia.Journal of Archaeological Science,51,98–114.
Sugita, S. (2007). Theory of quantitative reconstruction of veg-etation I: Pollen from large sites REVEALS regional vegeta-tion composition. The Holocene, 17, 229–241. https://doi.org/10.1177/0959683607075837
Taavitsainen, J.-P., Simola, H., & Grönlund, E. (1998). Cultivation his-tory beyond the periphery: Early agriculture in the north Europeanboreal forest. Journal of World Prehistory, 12, 199–253. https://doi.org/10.1023/A:1022398001684
Tallavaara,M.,Pesonen,P.,&Oinonen,M. (2010).Prehistoricpopulationhistory ineasternFennoscandia.Journal of Archaeological Science,37,251–260.https://doi.org/10.1016/j.jas.2009.09.035
Tallavaara,M.,&Seppä,H. (2012).Did themid-Holoceneenvironmentalchangescausetheboomandbustofhunter-gathererpopulationsizein eastern Fennoscandia? The Holocene, 222, 215–225. https://doi.org/10.1177/0959683611414937
Tikkanen, M., & Oksanen, J. (2002). Late Weichselian and HoloceneshoredisplacementhistoryoftheBalticSeainFinland.Fennia,180,9–20.
Timpson,A.,Colledge,S.,Crema,E.,Edinborough,K.,Kerig,T.,Manning,K.,…Shennan,S. (2014).Reconstructingregionalpopulationfluctua-tionsintheEuropeanNeolithicusingradiocarbondates:Anewcase-studyusinganimprovedmethod.Journal of Archaeological Science,52,549–557.
Trondman,A.K.,Gaillard,M.J.,Mazier,F.,Sugita,S.,Fyfe,R.,Nielson,A.B., … Wick, L. (2015). Pollen-based quantitative reconstructions ofHoloceneregionalvegetationcover (plant-functional typesand land-cover types) in Europe suitable for climatemodelling.Global Change Biology,21,676–697.https://doi.org/10.1111/gcb.12737
Trondman, A. K., Gaillard, M. J., Sugita, S., Björkman, L., Greisman, A.,Hultberg,T.,…Mazier,F. (2016).Arepollen records fromsmall sitesappropriate for REVEALS model-based quantitative reconstructionsof past regional vegetation? An empirical test in southern Sweden.Vegetation History and Archaeobotany, 25, 131–151. https://doi.org/10.1007/s00334-015-0536-9
Welinder,S.(2011).Earlyfarminghouseholds3900–800BC.InJ.Myrdal,&M.Morell(Eds.),The Agrarian history of Sweden: From 4000 BC to AD 2000.SE:NordicAcademicPress,Lund.
inBritain:Acomparisonofpollen-basedland-coverandarchaeological14Cdate-inferredpopulationchange.Journal of Archaeological Science,51,216–224.https://doi.org/10.1016/j.jas.2012.10.025
Woodward, F. I., Lomas,M.R.,&Kelly,C.K. (2004).Global climate andthe distribution of plant biomes. Philosophical Transactions Royal Society of London, series B,359,1465–1476.https://doi.org/10.1098/rstb.2004.1525
Zhang,Y.,Renssen,H.,&Seppä,H. (2016).Effectsofmelting icesheetsandorbitalforcingontheearlyHolocenewarmingintheextratropicalNorthernHemisphere.Climate of the Past,12,1119–1135.https://doi.org/10.5194/cp-12-1119-2016
Zhang,Y.,Renssen,H.,Seppä,H.,&Valdes,P.J. (2017b).Holocenetem-perature trends in the extratropical NorthernHemisphere based oninter-modelcomparisons.Journal of Quaternary Science.(inreview).
SUPPORTING INFORMATION
Additional Supporting Information may be found online in the supportinginformationtabforthisarticle.Appendix S1MetadataforpollenrecordsAppendix S2MetadataforcharcoalrecordsAppendix S3TheREVEALScorrectionsforthe23planttaxaAppendix S4VegetationcompositionandexplanatoryvariablesAppendix S5Plantfunctionaltypes(PFTs)Appendix S6Summedprobabilitydistributions(SPD)ofarchaeologicalradiocarbondatesAppendix S7 Variation partitioning results for winter and summertemperatureAppendix S8Variationpartitioningresultswithplantfunctionaltypes(PFTs)Appendix S9Resultsfrommovingwindowanalysis
How to cite this article:KuosmanenN,MarquerL,TallavaaraM,etal.Theroleofclimate,forestfiresandhumanpopulationsizeinHolocenevegetationdynamicsinFennoscandia.J Veg Sci. 2018;00:1–11. https://doi.org/10.1111/jvs.12601