Response of temperate grasslands at different altitude to simulated summer drought different - PDF document

2526A.K.GilgenandN.Buchmann:Responseoftemperategrasslandstodrought Table1.DescriptionofthethreeexperimentalgrasslandsitesinSwitzerland. ChamauFr¨ueb¨uelAlpWeissenstein PoliticalcommunityH¨unenberg,ZGWalchwil,ZGBerg¨un,GRCoordinatesWGS84471203700N,82403800E47605700N,83201600E463406000N,94702600EElevation393ma.s.l.982ma.s.l.1978ma.s.l.Growingseasonlengthmid-ApriltoOctoberMaytomid-Octobermid-JunetoSeptemberSoilTypecambisolagleysolahumoussandyloambC-horizonstartsat(approx.)120cma�120cma30cmSand/silt/clayfraction(approx.)a35/40/25%30/40/20%NAAnnualprecipitationsumc1179mm1632mm918mmMeanannualtemperaturec9.8C7.7C2.3CVegetationtype(dominantspecies)articialgrass-legumemixturepermanentmanagedpasturepermanentalpinepasture(Trifoliumrepens,Loliumperenne,(Dactylisglomerata,Alopecurus(Festucarubra,Phleumrha´eticum,Poapratensis)pratensis,Taraxacumofcinale)dTrifoliumrepens)eNumberofspecies(approx.)2535around20Managementtypeofusagesilage;sheepgrazinginautumnsilage/hay;cattlegrazinginautumncattleandhorsegrazingintensityintensiveintermediateextensivenumberofcutsperyear61–20typeoffertiliserliquidmanuresolidmanurenone aRoth(2006)bSch¨arer(2003)cdatafromZeeman(2008)adaptedwithdatafromMeteoSwissdSautier(2007)eKeller(2006) Bollingeretal.,1991;StampiandZeiter,2004)and(2)eldexperimentsmanipulatingrainfallamounts.Suchma-nipulationstudieshavebeencarriedoutinaridorsemi-aridregionswheredroughtisoccurringfrequently(e.g.Sternbergetal.,1999;GrecoandCavagnaro,2003;K¨ochyandWilson,2004;Englishetal.,2005;Schwinningetal.,2005;Heisler-Whiteetal.,2008;Sherryetal.,2008)butalsointemperategrasslandswheredroughtisnotasevereproblemtoday(e.g.Grimeetal.,2000;Morecroftetal.,2004;Kahmenetal.,2005;Mikkelsenetal.,2008).AnumberofstudiesfocusedontheeffectsofchangesinprecipitationtimingandamountsonC4dominatedgrasslands(e.g.Knappetal.,2001;Fayetal.,2002,2003;Nippertetal.,2009).Inmostofthesema-nipulationstudies,productivityofcommongrasslandspecieswasgreatlyreduced,carbon(C)allocationtobelow-groundpartsincreasedanddeeprootedspeciesweremoredroughtresistant,supportingstudiescarriedoutwithsinglespeciesundercontrolledconditions.However,detailedinformationonthedroughtresponseofC3dominatedtemperategrass-landsinEuropeisstillrare.Thisstudythereforeaimedtoassesstheresponsestosimu-latedextremesummerdroughtofthreetemperategrasslandsatdifferentaltitudesinSwitzerland.Weexcludedprecipi-tationwithtransparentrainsheltersandfocusedonthefol-lowingquestions:(1)Howdoescommunityabove-groundbiomassproductivityofthesethreegrasslandsrespondtosimulatedsummerdrought?(2)Iscommunitybelow-groundbiomassproductivityaffectedbysimulatedsummerdroughtaswell?(3)Dodifferentplantfunctionaltypesdifferintheirresponsetothedroughttreatment?Wehypothesisedthat(1)thesimulatedsummerdroughtdecreasescommunityabove-groundproductivitybecausewaterisoneofthekeyresourcesforplantgrowth,(2)rootbiomassproductivityincreaseswithdroughtbecauseofincreasedforagingforwater,(3)differentplantfunctionaltypesdifferintheirresponsetosimulatedsummerdroughtduetotheirdifferentrootingpatterns.2Materialsandmethods2.1ExperimentalsitesandsetupThestudywasconductedatthreedifferenttemperategrass-landsitesacrossSwitzerland,representingSwissgrasslandsystemsattherespectivealtitudes(Table1).Theexperi-mentwasestablishedbytheendofJune2005atthetwolowersites,ChamauandFr¨ueb¨uel,andinJuly2006atthealpinesite,AlpWeissenstein,andwascontinueduntil2007,thusduringthreeandtwogrowingseasons,respectively.Ateachsite,weinstalledveportablerainsheltersandexcludedrainfallinspring/summertosimulateapronounceddrought.In2006,twoadditionalrainsheltersandcontrolplotswereinstalledatChamauandoneeachatFr¨ueb¨uel.Thetunnel-shapedrainsheltershadanareaof33.5mandwereabout2.1mhighatthehighestpoint.Duringtheentiredrought Biogeosciences,6,2525– 2539 ,2009www.biogeosciences.net/6/2525/2009/ A.K.GilgenandN.Buchmann:Responseoftemperategrasslandstodrought2527 Table2.Datesofrainexclusionperiodandamountsofprecipitationexcludedduringthethreeyearsofexperimentandaveragelong-termannualprecipitationsumatallsites.Thefractionofannualprecipitationisgiveninbrackets. 200520062007 TreatmentAmountTreatmentAmountTreatmentAmountAnnualperiod[mm]period[mm]period[mm]precipitation[mm]a Chamau24June–19Sep491(42%)31May–17Aug271(24%)2May–10July403(33%)1159Fr¨ueb¨uel23June–19Sep763(52%)31May–17Aug388(24%)7May–20July589(33%)1534AlpWeissenstein––6July–24Aug248(29%)20June–23Aug311(32%)975 adatafromMeteoSwissstationsCham(1972–2000,forChamau),Zugerberg(1972–2008,forFr¨ueb¨uel)andPreda(1961–1974,forAlpWeissenstein) treatmentperiod,thesteelframeswerecoveredwithtrans-parentplasticfoils(200m;Gew¨achshausfolieUV5,fo-litecAgrarfolien-VertriebsGmbH,Westerburg,Germany).Toensuresufcientaircirculation,therainshelterswerenotcloseddowntothegroundandthemainopeningoftheshel-terswasorientedtowardsthemainwinddirection.Basedonregionalclimatemodelprojections(e.g.Freietal.,2006),wedeterminedthatreducingMaytoAugustprecipitationby45%comparedtotodaywouldsimulateanextremesummerdroughtinthefuture.Aperiodofeighttotenweeksbegin-ninginMayatChamauandFr¨ueb¨uelandsixtoeightweeksatAlpWeissensteinwasrequiredtoachievethistargetre-duction.Duetotheveryvariableweatherconditions,theamountsofprecipitationexcludedvariedoverthethreeyearsoftreatmentbutrelativeamountsatthethreesiteswerecom-parablewithinyears(Table2).Weestablishedacoreareaof12minthemiddleoftherainshelterstoexcludeanydirectraininputonplots.Allrainfallremovedbytheshel-terswasleadawayfromtherespectiveplots.Controlplots(equalnumberofreplicatesasfortheshelters)locatednexttotheshelteredplotsreceivednaturalrainfallamounts.Thesameplotswereusedthroughoutthethreeyearsoftheexper-iment.Duringthethreeyearsoftheexperiment,nofertiliserwasappliedtotheplotsandnograzingwasallowedontheplots.Inautumn2006,theweedspeciesRumexobtusifoliusL.hadtoberemovedmanuallyfromtheexperimentalplotsatChamauduetoSwissregulations(Gilgenetal.,2010),overseedingtookplaceinearlyMarch2007tore-establishaclosedvegetationcover.2.2MicrometeorologicalmeasurementsMicroclimaticvariableswerecontinuouslymonitoredclosetothecentreoftwodroughtandcontrolplotspersite,start-inginautumn2005atFr¨ueb¨uel,spring2006atChamauandsummer2006atAlpWeissenstein.Soiltemperature(Pre-cisionICTemperatureTransducerAD592AN,AnalogDe-vices,Norwood,MA,USA),soilmoistureatthreesoildepth(5,15and30cm)using20cmlongECH2Oprobes(EC-20,DecagonDevices,Inc.,Pullman,WA,USA)andsoilheatux(HFP01,HukseuxThermalSensorsB.V.,Delft,Netherlands)closetothesoilsurfacewererecorded.Wealsomonitoredairtemperatureat60and160cm(usingthesamesensorsasforsoiltemperaturebutwithventilation)andpho-tosyntheticallyactiveradiation(PARLITE,Kipp&ZonenB.V.,Delft,Netherlands).Theairtemperaturesensorsweredisconnectedinwinterandspring2006/2007duetotechnicalmodicationsinthesetup.Aftermid-October2007,theairtemperaturesensorswerenotventilatedanylongertosavebatterylifetime.Measurementsweremadeevery10swhiletenminuteaverageswereloggedwithaCR10Xdatalogger(CampbellScienticInc.,Logan,UT,USA).Inaddition,weassessedsoilmoisturebytakingadditionalmeasurementsofgravimetricsoilwatercontentin2007.Wethereforesampled15cmdeepsoilcoresfromthecentreoftheplotsanddividedthemintothree5cmpieces.Sampleswerestoredintightlysealedplasticbagsandfreshweightwasmeasuredimmediatelyafterreturningfromtheeld.Thesoilcoresweredriedtoweightconstancyat100C(somedays)anddryweightwasmeasuredafterwards.Gravi-metricsoilwatercontentwasthencalculatedasthediffer-enceofthefreshandthedryweight(weightofthewater)dividedbythedryweight(weightofthesoil).2.3Above-groundproductivityAbove-groundbiomasswasharvestedatthecuttingdatesofthesurroundingfarm,i.e.sixtimesperyearatChamau(threetimesin2005),twotimesperyearatFr¨ueb¨uel(oncein2005)andonceattheendofthegrowingseason(endofSeptember)atAlpWeissenstein.Biomasswascollectedusing2050cmframesthatwererandomlyplacedontheplotsin2005andinstalledatxedlocationsstartingspring2006.Cuttingheightofthevegetationwasapproximately7cmabovethesoilaccordingtothecommonmanagementpracticeonthefarms.Twosamplesperplotweretakenandthenpooledfortheanalyses(representing0.2m2).Biomasswasstoredinplasticbagsat4Cforamaximumofoneweek www.biogeosciences.net/6/2525/2009/Biogeosciences,6,2525– 2539 ,2009 2528A.K.GilgenandN.Buchmann:Responseoftemperategrasslandstodrought untilitcouldbeseparatedintospecies(plantfunctionaltypesforAlpWeissenstein)anddriedat60Cuntilweightcon-stancy.Drymatterwasthendeterminedforeachspeciessample,andthesumofallsamplesfromthesameplotswasusedasanestimateforcommunityabove-groundbiomassproductivity.Eachspecieswasassignedaplantfunctionaltype(grass,forb(i.e.non-leguminousforb)orlegume).Thedeadbiomass(oftenalsoreferredtoasnecromass)wascon-sidereditsownplantfunctionaltypeandwasneverseparatedintospecies(i.e.biomassoftheotherthreefunctionaltypeswasalivebydenition).Itwashoweverincludedincommu-nityabove-groundproductivityestimates(oftenalsocalledphytomass).2.4LAIandvegetationheightDuringthegrowingseasons2006and2007(only2007atAlpWeissenstein),leafareaindex(LAI)wasmeasuredapprox-imately7cmabovesoil(cuttingheight)usinganLAI-2000PlantCanopyAnalyzer(LI-CORBiosciences,Lincoln,NE,USA).Fivemeasurementsdistributedovertheplotwereav-eragedtorepresenttheplotLAI.A270viewcapwasusedtoreduceplotssizerequiredformeasurements,i.e.only90ofthesensorviewwereusedformeasurements.VegetationheightwasestimatedatthesamedatesasLAIusingastickandafallingstyrofoamplate(0.50.5m,1cmthick).Twotofourmeasurementswereaveragedforeachplot.2.5RootbiomassproductivityRootbiomassproductivityatthecommunityleveloverthecompletegrowingseasonwasdeterminedatChamauandFr¨ueb¨uelin2007andovernearlytwogrowingseasons(2006and2007)atAlpWeissensteinusingingrowthcores(4.4cmdiameter,30cmlength).Coreswerepositionedinanangleofapproximately45.Allingrowthcoreswerelledwithroot-free,sieved(2mm)soilfromthecorrespondingsite.In-growthcoreswereinstalledfrom13March2007to1Decem-ber2007atChamau(263days),from15December2006to1December2007atFr¨ueb¨uel(351days),andfrom14July2006to25September2007atAlpWeissenstein(438days).OneingrowthcoreperplotwasburiedatAlpWeissensteinwhiletwocoresperplotwereusedatthetwoothersites.Af-terremovalfromthesoil,coreswerestoredat4Cuntilfur-theranalysis.Theingrownrootswerewashedfromthesoilcoresinthelaboratory,rootsweredriedtoweightconstancyat60Candthedryweightwasdetermined.Wheretwoin-growthcoreshadbeeninstalledandrecovered,rootbiomassoftwocoreswaspooledtocalculaterootproductivityoftheplot.Insomecases,partofthesoilwaslostwhenremovingtheingrowthcoresfromthesoil.Therefore,theactuallengthofthesoilcorewasusedtocalculatetheamountofrootsperunitsoildepth.Astheingrowthcoresremainedintheeldfordifferenttimeperiodsatthethreesites,therootweightwasdividedbythenumberofdaysthatrootshadbeenal-lowedtogrowintothecoresandthenmultipliedby365torepresenttherootmassperm2andyear.2.6CarbonisotopeandnitrogenconcentrationmeasurementsTodeterminecarbonisotoperatios(13C)andnitrogencon-centrations,themostabundantspeciesweresampledatChamau(AgrostisstoloniferaL.,AlopecuruspratensisL.,DactylisglomerataL.,LoliummultiorumLAM.,PhleumpratenseL.AGG.,PoapratensisL.AGG.,PoatrivialisL.S.L.,RumexobtusifoliusL.,andTrifoliumrepensL.)andFr¨ueb¨uel(AgrostiscapillarisL.,A.stolonifera,A.praten-sis,AnthoxanthumodoratumL.,P.pratense,P.pratensis,P.trivialis,RumexacetosaL.,R.obtusifolius,andT.repens),whilethefourplantfunctionaltypeswereanalysedatAlpWeissenstein.BiomasswasgroundtoanepowderandanalysedusingaFlashEA1112Serieselementalanalyser(ThermoItaly,formerCEInstruments,Rhodano,Italy)cou-pledtoaFinniganMATDeltaplusXPisotoperatiomassspec-trometer(FinniganMAT,Bremen,Germany)viaa6-portvalve(Brooksetal.,2003),aConFloIII(Werneretal.,1999)andanadditionalNaon-trapbackedbyaconven-tionalMg(ClO4/2-trapfollowedbya4-portvalve(Werner,2003)betweenreductiontubeandGCcolumn.Post-runoff-linecalculations(blank,offsetandpossiblydriftcorrections)wereperformedtoassignthenal-valuesontheV-PDBandAIR-N2scalesaccordingtoWernerandBrand(2001).Thelong-timeprecisionforthelab'squalitycontrolstandardtyrosine(2.5years)was0.05‰for13C.PrecisionforNconcentrationswas0.05%.Forstatisticalanalyses,isotopeandnitrogendatawereweightedbybiomasstocalculateavegetationandfunctionalgroupmeanforeachplot.2.7StatisticalanalysisDuetotherathersmallnumberofreplicates(n=5–7),allplotswereincludedinthestatisticalanalysesofbiomass,vegetationheight,13CandNconcentrations,i.e.noout-lierswereexcluded.LAIvalueswereexcludedonlyifthestandarderrorofthevemeasurementsperplotwashigherthanathirdofthemeanLAIoftheplotorifMTA(meantiltangle)wasbelow30%(criteriausedfor2007sincethisinformationwasnotavailablefor2006).StatisticalanalyseswereperformedusingR2.5.0(RDe-velopmentCoreTeam,2007).Communityabove-groundandfunctionaltypebiomass,LAI,vegetationheightandiso-topesignaturesweregenerallyallanalysedwiththesameANOVAmodel.Inarststep,thefulldatasetswereanalysed(allsitesoverallyears),usinganANOVAmodelconsider-ingsite,harvestdate,treatment(controlvs.droughttreat-ment)andallinteractions,includingthetripleinteraction.Inasecondstep,datasetswereanalysedforeachyearsep-aratelyusingthesamemodel.Inathirdstep,thedatasetwasanalysedforeachsiteseparatelyandnallyforeachharvest Biogeosciences,6,2525– 2539 ,2009www.biogeosciences.net/6/2525/2009/ A.K.GilgenandN.Buchmann:Responseoftemperategrasslandstodrought2531 andlegumeswerenotsignicantlyaffectedinanyofthethreeyears.Communityabove-groundregrowth(i.e.biomassgrownbetweencuttingdates)acrossallsitesandallthreeyearsoftheexperiment(Fig.3,Table4,lowerpart)wasaffectedbysiteandharvestdate(P0.001)aswellasbythedroughttreatment(P=0.003).Asignicantsitetreatmentinter-action(P=0.001)indicateddifferinggrasslandresponsestodroughtatthethreesites,asalreadyobservedfortheannualcommunityabove-groundproductivities.Includingyearintheanalysisdidnotchangesignicancelevelsandwasthusomitted.Whenanalysingthethreeyearsseparately,siteandharvestdateexplainedalargefractionofthevariationinproductivityinallthreeyears(2005–2007),whilethetreat-menteffectwasinsignicantin2005,itbecamemarginallysignicantin2006(P=0.06)andstayedsignicantin2007(P=0.01).In2007,thedroughttreatmentshowedhighlysig-nicanteffectsonabove-groundproductivitiesatChamau(P=0.001)andAlpWeissenstein(P0.001),butstillnoef-fectsatFr¨ueb¨uel(P=0.5;Fig.3).TheseannualpatternscouldalsobeseenintheresponseofthesingleharveststhatwereunaffectedbythedroughttreatmentatFr¨ueb¨uelandinthetworstyearsatChamau(exceptforthefthharvestin2006),butsignicantlyreducedatChamauinthethirdyear(rst,thirdandfourthharvest)andalsostronglyreducedatAlpWeissenstein(Fig.3).Thedatetreatmentinteractiontermwasneversignicant,indicatingthatthedirectionofthedroughtresponsesoftherespectivegrasslandswasstableovertime(Table4,lowerpart).FocusingonthedifferentPFTsclearlyshowedthatgrassesresembledthedroughtresponseofthecommunityabove-groundproductivitybest(independentofRumex).Thedroughttreatmentclearlydecreasedgrassregrowthabove-ground(P=0.001).Althoughthisresponsedifferedamongsites(droughtstronglydecreasedgrassproductivitiesatChamauandAlpWeissensteinbutnoeffectwasobservedatFr¨ueb¨uel;sitetreatmentinteraction:P0.001),there-sponseofgrassestodroughtwasstableoverthegrowingseason(harvestdatetreatmentinteraction:notsigni-cant).Furthermore,similarpatternsforcommunityandgrassabove-groundproductivitywerealsofoundwhenanalysingtheyearsseparately(exceptsiteeffectin2005):grasspro-ductivitydifferedamongthethreesitesandgrassregrowthchangedduringthegrowingseason(siteandharvestdateef-fects:P0.001).Forbsreactedtodroughtwithdecreasedregrowth(overall:P=0.007),particularlyin2006.Unliketheirannualproductivities,forbregrowthdifferedbetweensitesandalsoduringtheseason(siteandharvestdateef-fects:P0.001),butthedirectionofthedroughtresponsewasunaffectedbysiteandharvestdate(overallsitetreat-mentandharvestdatetreatmentinteractions:P&#x]TJ/;ན ; .96; Tf;&#x 15.;Ց ;� Td;&#x[000;0.6).Furthermore,legumesweregenerallypositivelyaffectedbythedroughttreatment(overall:P=0.005),particularlyintheyear2007,butthiswasdrivenbythepositivedroughtre-sponseatChamau(datanotshown). Fig.2.Relationshipsofannualprecipitationsums(a)andofex-cludedamountsofprecipitation(b)withtheaveragechangeinan-nualcommunityabove-groundbiomassproductivitiesinresponsetothedroughttreatment.R2andPvaluesfortheregressionofannualabove-groundbiomassexcludingRumexobtusifoliusatChamauaregiveningrey(opensymbols,dashedline). 3.3LAIandvegetationheightLAImeasurementsduring2006and2007(Fig.4)repre-sentedabove-groundbiomassregrowthpatternsinhighertemporalresolutionthancommunityabove-groundproduc-tivityandvariedsignicantlyduringthegrowingseasons(P0.001).Similartoproductivitymeasurements,nosig-nicantdroughteffectonLAIwasfoundatChamauin2006(exceptveryearlyintheseason),probablyconfoundedbyRumexabundance.Incontrast,thedroughtresponseofLAIin2007wasstrongerthanthatofabove-groundproductivity,withdroughtloweringLAIvaluessignicantly(P0.001).AtFr¨ueb¨uel,LAIwasnegativelyaffectedbydroughtin www.biogeosciences.net/6/2525/2009/Biogeosciences,6,2525– 2539 ,2009 2532A.K.GilgenandN.Buchmann:Responseoftemperategrasslandstodrought Table4.ResultsoftheANOVAmodelsforannualsumsofabove-groundbiomassandabove-groundharvestyieldsofcommunity(comm.=sumofdeadandalivebiomass)andplantfunctionaltype(grass,forb(excludingRumexobtusifoliusatChamau),legume;onlyalivebiomass)aswellasdeadabove-groundbiomass.Mainfactorsaresite(Chamau,Fr¨ueb¨uelorAlpWeissenstein),harvestdateandtreatment(droughtvs.control).SignicantPvalues(P0.05)aregiveninbold,marginallysignicantvalues(0.1P�0.05)initalics;n=5–6. Allyears200520062007 Comm.GrassForbLeg.DeadComm.GrassForbLeg.DeadComm.GrassForbLeg.DeadComm.GrassForbLeg.Dead Annualabove-groundbiomasssumsSite0.0010.0010.20.0010.0010.0020.0010.010.070.40.0010.0010.30.0010.0010.0010.0010.20.0060.001Treatment0.10.080.060.21.00.70.70.30.20.70.10.20.11.00.050.050.030.60.20.4SiteTreatment0.20.050.70.10.60.60.10.20.040.40.30.40.70.90.80.050.0040.40.20.2Above-groundharvestyieldsSite0.0010.0010.0010.0010.0010.020.50.0010.70.0010.0010.0010.0010.0030.0010.0010.0010.0010.0020.001Harvestdate0.0010.0010.0010.050.0010.010.0040.90.50.0010.0010.0010.0010.020.0030.0010.0010.80.20.05Treatment0.0030.0010.0070.0050.80.60.50.20.10.70.060.050.010.40.010.010.0020.50.010.3SiteHarvestdate0.20.30.0010.090.001–––––––––––––––SiteTreatment0.0010.0010.70.030.010.60.20.0020.060.40.060.070.30.50.10.0060.0050.20.10.001HarvestdateTreatment0.60.60.60.70.021.00.90.90.20.60.30.50.80.70.30.20.20.90.61.0 Fig.3.Effectofsummerdroughtoncommunityabove-groundbiomassproduction(i.e.deadandaliveplantmaterial)atChamau(excludingR.obtusifoliusbiomass;a–c),Fr¨ueb¨uel(d–f)andAlpWeissenstein(g–h)duringtheexperiment.Barsrepresentbiomassproductionperharvest(i.e.biomassregrownbetweencuttingdates).Meansandstandarderrorsaregiven(n=5–7).Periodsofdroughttreatmentareshadedingrey.0.05P�0.01,0.01P�0.001. Biogeosciences,6,2525– 2539 ,2009www.biogeosciences.net/6/2525/2009/ A.K.GilgenandN.Buchmann:Responseoftemperategrasslandstodrought2533 Fig.4.Effectofsummerdroughtoncommunityleafareaindex(LAI)atChamau(a,b),Fr¨ueb¨uel(c,d)andAlpWeissenstein(e)dur-ingtheexperiment.Meansandstandarderrorsaregiven(n=3–7).Periodsofdroughttreatmentareshadedingrey.()0.1P�0.05,0.05P�0.01,0.01P�0.001,P0.001. bothyears(P0.001)althoughtherewasnosucheffectoncommunityabove-groundbiomass.AtAlpWeissenstein,droughtreducedLAIvalues(P=0.02),butdespitethelargedecreaseinbiomassunderdroughtconditionsthistrendwasnotsignicantatthedifferentmeasurementdates.Nore-lationshipofLAIbeforethecutwithharvestedcommunityabove-groundbiomasswasfoundatanyofthethreesites(R20.1,P&#x]TJ/;ན ; .96; Tf;&#x 15.;Ւ ;� Td;&#x[000;0.2butP0.001forChamau).Vegetationheight(datanotshown)developedsimilartoLAIbutwasnotasstronglyaffectedbythedroughttreat-mentasLAI(exceptforAlpWeissensteinwherevegetationheightwassignicantlyreducedbydrought).Vegetationheightbeforethecutswasnotrelatedtocommunityabove-groundbiomassatChamauandAlpWeissenstein,incontrasttoFr¨ueb¨uel(R2=0.4,P0.001).3.4RootbiomassRootbiomassproductivitydifferedsignicantlyamongsites(P=0.01),withproductivitiesbeingalmosttwiceashighatAlpWeissensteincomparedtoFr¨ueb¨uelandChamau(Fig.5).However,below-groundproductivitywasnotaf-fectedbythedroughttreatment(P=0.9).Inaddition,thesitetreatmentinteractiontermwasnotsignicant(P=0.9),in-dicatingthattheresponseofbelow-groundproductivitytodroughtwassimilaracrossallthreesites.3.5CarbonisotopeandnitrogenconcentrationmeasurementsBulkabove-groundcarbonisotoperatios(13C)rangedbe-tween�31.3‰and�26.7‰forthegrasslandvegetation(Fig.6),withsignicantdifferencesamongsites(P0.001). www.biogeosciences.net/6/2525/2009/Biogeosciences,6,2525– 2539 ,2009 2538A.K.GilgenandN.Buchmann:Responseoftemperategrasslandstodrought Kalapos,T.,vandenBoogaard,R.,andLambers,H.:Effectofsoildryingongrowth,biomassallocationandleafgasexchangeoftwoannualgrassspecies,PlantSoil,185,137–149,1996. Karsten,H.D.andMacAdam,J.W.:Effectofdroughtongrowth,carbohydrates,andsoilwaterusebyperennialryegrass,tallfes-cue,andwhiteclover,CropSci.,41,156–166,2001. Keller,P.:VegetationskundlicheUntersuchungenundErtragderGr¨unland¨achenimGebietderETH-ForschungsstationAlpWeissensteinGR,DiplomaThesis,UniversityofZurich,Switzerland,119pp.,2006. Knapp,A.K.,Briggs,J.M.andKoelliker,J.K.:Frequencyandextentofwaterlimitationtoprimaryproductioninamesictem-perategrassland,Ecosystems,4,19–28,2001. Knapp,A.K.andSmith,M.D.:Variationamongbiomesintempo-raldynamicsofabovegroundprimaryproduction,Science,291,481–484,2001. K¨ochy,M.andWilson,S.D.:Semiaridgrasslandresponsestoshort-termvariationinwateravailability,PlantEcol.,174,197–203,2004. Lambers,H.,Chapin,F.S.,andPons,T.L.:Plantphysiologicalecology,Springer,NewYork,NY,USA,1998. Leuzinger,S.,Zotz,G.,Asshoff,R.,andK¨orner,C.:ResponsesofdeciduousforesttreestoseveredroughtinCentralEurope,TreePhysiol.,25,641–650,2005. Marquard,E.,Weigelt,A.,Temperton,V.M.,Roscher,C.,Schu-macher,J.,Buchmann,N.,Fischer,M.,Weisser,W.W.,andSchmid,B.:Plantspeciesrichnessandfunctionalcompositiondriveoveryieldingina6-yeargrasslandexperiment,Ecology,inpress,2009. Mikkelsen,T.N.,Beier,C.,Jonasson,S.,Holmstrup,M.,Schmidt,I.K.,Ambus,P.,Pilegaard,K.,Michelsen,A.,Albert,K.,An-dresen,L.C.,Arndal,M.F.,Bruun,N.,Christensen,S.,Dan-baek,S.,Gundersen,P.,Jorgensen,P.,Linden,L.G.,Kongstad,J.,Maraldo,K.,Prieme,A.,Riis-Nielsen,T.,Ro-Poulsen,H.,Stevnbak,K.,Selsted,M.B.,Sorensen,P.,Larsen,K.S.,Carter,M.S.,Ibrom,A.,Martinussen,T.,Miglietta,F.,andSverdrup,H.:Experimentaldesignofmultifactorclimatechangeexperi-mentswithelevatedCO2,warminganddrought:theCLIMAITEproject,Funct.Ecol.,22,185–195,2008. Molyneux,D.E.andDavies,W.J.:Rootingpatternandwaterrela-tionsofthreepasturegrassesgrowingindryingsoil,Oecologia,58,220–224,1983. Monti,A.,Barbanti,L.,andVenturi,G.:Photosynthesisonindivid-ualleavesofsugarbeet(Betavulgaris)duringtheontogenyatvariablewaterregimes,Ann.Appl.Biol.,151,155–165,2007. Morecroft,M.D.,Masters,G.J.,Brown,V.K.,Clarke,I.P.,Taylor,M.E.,andWhitehouse,A.T.:Changingprecipitationpatternsalterplantcommunitydynamicsandsuccessioninanex-arablegrassland,Funct.Ecol.,18,648–655,2004. Nilsen,E.T.andOrcutt,D.M.:Thephysiologyofplantsunderstress,JohnWiley&Sons,Inc.,NewYork,USA,1996. Nippert,J.B.,Knapp,A.K.,andBriggs,J.M.:Intra-annualrainfallvariabilityandgrasslandproductivity:canthepastpredictthefuture?,PlantEcol.,184,65–74,2006. Nippert,J.B.,Fay,P.A.,Carlisle,J.D.,Knapp,A.K.,andSmith,M.D.:Ecophysiologicalresponsesoftwodominantgrassestoalteredtemperatureandprecipitationregimes,ActaOecol.,35,400–408,2009. Paruelo,J.M.,Lauenroth,W.K.,Burke,I.C.andSala,O.E.:Grasslandprecipitation-useefciencyvariesacrossaresourcegradient,Ecosystems,2,64–68,1999. Penuelas,J.,Prieto,P.,Beier,C.,Cesaraccio,C.,deAngelis,P.,deDato,G.,Emmett,B.A.,Estiarte,M.,Garadnai,J.,Gorissen,A.,Lang,E.K.,Kroel-Dulay,G.,Llorens,L.,Pellizzaro,G.,Riis-Nielsen,T.,Schmidt,I.K.,Sirca,C.,Sowerby,A.,Spano,D.,andTietema,A.:Responseofplantspeciesrichnessandprimaryproductivityinshrublandsalonganorth-southgradientinEuropetosevenyearsofexperimentalwarminganddrought:reductionsinprimaryproductivityintheheatanddroughtyearof2003,GlobalChangeBiol.,13,2563–2581,2007. RDevelopmentCoreTeam:R:Alanguageandenvironmentforsta-tisticalcomputing,2.5.0ed.,RFoundationforStatisticalCom-puting,Vienna,Austria,2007. Roth,K.:BodenkartierungundGIS-basierteKohlenstofn-venturvonGraslandb¨oden:UntersuchungenandenETH-ForschungsstationenChamauundFr¨ueb¨uel(ZG,Schweiz),DiplomaThesis,UniversityofZurich,Switzerland,132pp.,2006. Saab,I.N.,Sharp,R.E.,Pritchard,J.,andVoetberg,G.S.:In-creasedendogenousabscisicacidmaintainsprimaryrootgrowthandinhibitsshootgrowthinmaizeseedlingsatlowwaterpoten-tials,PlantPhysiol.,93,1329–1336,1990. Sautier,S.:ZusammensetzungundProduktivit¨atderVegetationimGebietderETHZ-ForschungsstationFr¨ueb¨uel(ZG),DiplomaThesis,UniversityofZurich,Switzerland,123pp.,2007. Sch¨arer,M.:KurzberichtderFallstudieAlpWeissenstein.N¨ahrstoffanalyseninAlpweidenSommer2002,TermPaper,ETHZurich,Switzerland,44pp.,2003. Schwinning,S.,Starr,B.I.andEhleringer,J.R.:Summerandwin-terdroughtinacolddesertecosystem(ColoradoPlateau)partII:effectsonplantcarbonassimilationandgrowth,J.Arid.Envi-ron.,61,61–78,2005. Seibt,U.,Rajabi,A.,Grifths,H.,andBerry,J.A.:Carbonisotopesandwateruseefciency:senseandsensitivity,Oecologia,155,441–454,2008. Sherry,R.A.,Weng,E.,ArnoneIII,J.A.,Johnson,D.W.,Schimel,D.,Verburg,P.S.,Wallace,L.L.,andLuo,Y.Q.:Laggedef-fectsofexperimentalwarminganddoubledprecipitationonan-nualandseasonalabovegroundbiomassproductioninatallgrassprairie,GlobalChangeBiol.,14,2923–2936,2008. Signarbieux,C.andFeller,U.:Effectsofanextendeddroughtpe-riodongrasslandsatvariousaltitudesinSwitzerland:Aeldstudy,in:Photosynthesis.EnergyfromtheSun.14thInterna-tionalCongressonPhotosynthesis,editedby:Allen,J.F.,Gantt,E.,Golbeck,J.H.,andOsmond,B.,Springer,Dordrecht,Nether-lands,1371–1374,2008. Signarbieux,C.:GrasslandsinachangingclimateinSwitzer-land:photosyntheticperformanceandwaterrelationsofselectedspeciesunderarticialdrought,PhDThesis,UniversityofBern,Switzerland,127pp.,2009. Signarbieux,C.andFeller,U.:Effectsofanextendeddroughtpe-riodonphysiologicalpropertiesofgrasslandsspeciesinSwitzer-land,inpreparation,2009. Stampi,A.andZeiter,M.:Plantregenerationdirectschangesingrasslandcompositionafterextremedrought:a13-yearstudyinsouthernSwitzerland,J.Ecol.,92,568–576,2004. Biogeosciences,6,2525– 2539 ,2009www.biogeosciences.net/6/2525/2009/