1054A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands . - PDF document

1054A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands Ithaslongbeenknownthatdrainageofpeatlandscausesirreversibleloweringofthesurface(subsidence)asaconsequenceofpeatshrinkageandbiologicaloxidation,withthelatterresultinginalossofcarbonstock.Inpeatlandar-easasdifferentastheFenlandsoftheUK,theNetherlands,VeniceLagooninItaly,theEvergladesandSacramentoDeltaintheUnitedStatesandLakeHulainIsrael,atotalsubsi-denceof200to600cmoccurredover40to130yr,bringingsurfacelevelsclosetoorbelowsealevel(Schothorst,1977;Hutchinson,1980;Stephensetal.,1984;HambrightandZo-hary,1998;Gambolatietal.,2003;DeverelandLeighton,2010).Inallofthesecases,peatoxidationisreportedtobethemaincauseofsubsidence.Inrecentyears,rapidlyin-creasingpeatcarbonlossesfromdrainedSEAsianpeatlandshavebeenfoundtocontributesubstantiallytoglobalgreen-housegasemissions(Hooijeretal.,2006,2010;Couwenbergetal.,2010;Murdiyarsoetal.,2010).EstimatesofnetcarbonlossesandresultantCO2emissionsfrompeatlanddrainedforagriculturerangefrom40tCO2ha�1yr�1(Mellingetal.,2005;Murdiyarsoetal.,2010;Herchoualc'handVerchot,2011),to&#x]TJ/;ང ; .96; Tf;&#x 9.4;1 0;&#x Td[;60tha�1yr�1atwatertabledepthsaround0.7m(applyingrelationsbetweenwatertabledepthandemissionasproposedbyDIDSarawak,2001;Hooijeretal.,2006,2010;Couwenbergetal.,2010),excludingforestbiomassandrelosses.TheuncertaintyintherateofcarbonemissionfromdrainedtropicalpeatlandiscausedpartlybytherelianceonmeasurementsofgaseousCO2emissionsthataredifculttoconductandinterpret.UnlessCO2emissionstudiesarecar-riedoutonalargescale(i.e.alargenumberofmeasure-mentsconductedoveralongtimeperiodatalargenum-berofmonitoringlocations)andoverarangeofenviron-mentalconditions(intermsofwatertable,vegetationcoverandtemperature),datauncertaintyisconsiderable(Couwen-bergetal.,2010;Murdiyarsoetal.,2010;Jauhiainenetal.,2012).Forexample,widelyquotedestimatesofCO2emis-sionfromoilpalmplantationsonpeathavebeenbasedonfewerthan50observations,includingreplicates,atsinglelo-cations(MurayamaandBakar,1996;Mellingetal.,2005).Moreover,fewstudieshaveestimatednetCO2emissionsre-sultingfrompeatoxidationalone,excludingrootrespiration.Also,gasuxmeasurementsdonotaccountforcarbonlossesindischargewater(DOCandPOC),thatleavethepeatlandindrainagewater(Alkhatibetal.,2007;Baumetal.,2007;Mooreetal.,2010).Recenteffortstocalculatethenetchangeinpeatcarbonstockfromthedifferencebetweenallesti-mateduxesintoandoutofthepeat,includingchangesinbiomass(Herchoualc'handVerchot,2011),havebeenincon-clusivebecauseofthelimiteddataavailableandthecumula-tiveuncertaintiesassociatedwitheachcomponent.Thereisaneed,therefore,forasimpleandreliableap-proachtodeterminingnetcarbonlossesfromdrainedtropicalpeatlands,especiallyinviewoftheurgentrequirementforlanduseplanningpoliciesthatreduceCO2emissionsfromSEAsianpeatlands,whichformasubstantialpartofglobalemissions(Hooijeretal.,2006,2010;Malhi,2010).Mea-surementsoflandsubsidence,incombinationwithdataonpeatcharacteristics,provideadirectapproachtocarbonlossassessmentthatisrelativelystraightforwardtoconductintheeldandtointerpret.Allimpactsonthepeatcarbonstockareintegratedovertimewithoutrequiringinstantaneousmea-surements,therebyprovidingamoreaccuratevalueforto-talcarbonlosseveniftheindividuallosscomponents(CO2,CH4,DOCandPOC)cannotbeseparatedusingthismethodalone.TheuseofthisapproachonSEAsianpeatlandshasbeenhamperedbyascarcityofreliablelong-termsubsidencedataandadequateinformationonbulkdensityandcarboncontentofthepeat(cf.reviewbyCouwenbergetal.,2010).Whendeterminingcarbonlossfromsubsidencedata,itisnecessarytoknowtheextenttowhichsubsidenceisthere-sultofpeatoxidationcomparedtophysicalvolumereduc-tion.Thefollowingsubsidencecomponentsneedtobesepa-rated: – Oxidation:decompositionofpeatintheaeratedzoneabovethewatertableowingtobreakdownoforganicmatter,resultingincarbonlossthroughreleaseofgaseousCO2totheatmosphere(Nelleretal.,1944;Jauhiainenetal.,2005,2008;Hiranoetal.,2009),andremovalasDOCandPOCindrainagewater(Alkhatibetal.,2007;Baumetal.,2007;Mooreetal.,2010).Thisprocess,actingalone,doesnotincreasebulkdensityofthepeatandcouldinfactdecreaseit. – Compactionandshrinkage:volumereductionofpeatintheaeratedzoneabovethewatertable.Compactionre-sultsfromthepressureappliedonthepeatsurfacebyheavyequipment;shrinkageoccursthroughcontractionoforganicbreswhendrying.Thesetwoprocessescanoftennotbeseparatedinpracticeandtheyareconsid-eredtogetheras“compaction”inthispaper.Bothpro-cessesleadtoanincreaseinpeatbulkdensity. – Consolidation:thecompressionofsaturatedpeatbelowthewatertableowingtolossofbuoyancyofthetoppeat,increasingstrainonthepeatbelow.Primarycon-solidationiscausedbylossofwaterfromporesinthepeat;itoccursrapidlywhengroundwaterisremovedfast,especiallywhereadensedrainagesystemisim-plementedinpeatofhighpermeability.Secondarycon-solidationisafunctionoftheresistanceofthesolidpeatmaterialitselftocompression;thisisaslowprocessthatmakesuponlyasmallfractionoftotalconsolidation(Berry,1983;MesriandAljouni,2007).Bothprocessesincreasepeatbulkdensity.Fireanderosionofpeatparticlesbywaterowbothremovepeatfromnearthesurface.Theirimpactonthebulkdensityoftheremainingpeatisunknown. Biogeosciences,9,1053– 1071 ,2012www.biogeosciences.net/9/1053/2012/ 1056A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands Table1.Summaryofmeasurementsitecharacteristics.Averagesareprovidedwithstandarddeviations. AcaciaplantationOilpalmplantationPlantationmeanDrainedforest26yrafterdrainage18yrafterdrainage6–18yrafterdrainage Sitelocation(Lat/Long)102.334/0.595103.601/�1.566–102.334/0.595Numberofmeasurementlocations1254216751Peatdepthm92.67.71.48.49.93.2Peatbulkdensityoftop1mgcm�30.0890.0180.0870.0180.088–Peatbulkdensity�1mdepth1gcm�30.0730.0150.0780.0070.075–Watertabledepthm0.70.20.730.230.710.330.16 1Measuredat1to2.3mdepth;affectedbyinitialconsolidation.2Naturalforeststripupto2kmfromplantationboundary;affectedbydrainage. Fig.1.CrosssectionalongtypicalstudytransectinSumatra,6yrafterdrainage,showingvariationinpeatdepth,averagewaterlevel,landuseandmonitoringlocationdensity. 2.4PeatcharacteristicsPeatthicknessandtype(bric,hemicorsapric)weredeter-minedatthetimeofpoleinstallationusinglocallyproducedaugersandvisualinterpretation.Bulkdensity(BD)wasdeterminedat22locationsintheAcaciaplantation,19thatweredrained4to7yrbeforeand3abouttwoyearsafterdrainage,andat10locationsintheoilpalmplantation.Peatsampleswerecollectedfromthesidesofpitsexcavatedinpeat,usingsharpenedsteelcylinderstoavoidpeatcompressionthatmayresultfromusingaverticalcorer.Thiswasdonequicklyafterpitconstruction,toavoiddeformationordryingofthepeat.Tofurtheravoidcom-pressionandtoensureinclusionofsmallerwoodremains,relativelylargecylindersof8cmdiameterand8cmlength(402cm3)wereused.Allpitswereatleast1mawayfromtreesorpalms,wherethepresenceoftreerootswasfoundtobeminimal.Waterwaspumpedfromdeeperpitstofacilitatesampling.IntheAcaciaplantation,pitsof1mdiameterwereupto1.2mdeepandthreereplicatesamplesweretakenatin-tervalsof0.15to0.3mstartingat0.075mbelowthesurface.Intheoilpalmplantation,pitswere2to2.5mdeepandsam-pleswerecollectedatintervalsof0.1m,commencing0.1mbelowthesurface(Fig.2).Atotalof1201peatsampleswereovendriedat105Cforupto96h(aslongasnecessarytoensurethatdryweightofsampleshadfullystabilized)tore-movemoisture,andweighedtocalculateBD.Largewoodremainscouldnotbecollectedinthecylin-ders,soadditionalcheckswerecarriedouttodetermineifun-dersamplingofsuchremainswouldaffecttheBDvalues.Atotalof20samplesofpartlydecomposedwoodtakenin3soilpitsfrompeatbelow1mdepthinoilpalmplantations,withanaveragevolume(SD)of326104cm3,weredriedandweighedfollowingthesameprotocolascylindersamples.Inaddition,thewoodcontentofthepeatinoilpalmplantationsiteswasassessedvisuallyon10cleanedpitsides,throughdetaileddescriptionsofpeatsurfacesof0.1by0.3mat0.1mdepthintervals,priortosampling.Ashcontentin223subsamplesfromAcaciaplantationsiteswasdeterminedbylossonignitioninamufefurnace. Biogeosciences,9,1053– 1071 ,2012www.biogeosciences.net/9/1053/2012/ 1060A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands Fig.4.Top:averagesubsidenceratesasmeasuredat14locationsinAcaciaplantations,overtherst9yrafterdrainage.Bottom:asmeasuredatalargernumberofdrainedpeatlandlocationsinSumatra(thisstudy),Malaysia(fromW¨ostenetal.,1997,basedonDIDMalaysia,1996),MildredIslandintheCaliforniaSacramentoDelta(DeverelandLeighton,2010)andFloridaEverglades.TheEvergladesrecordisaveragedfromthreerecordspresentedbyStephensandSpeir(1969);asthersttwoyearsaftercompletingthedrainagesystemin1912weremissingfromthesubsidencerecord,whichstartedin1914,weaddedasubsidenceof22.5cmyr�1forthoseyears,whichistheaveragesubsidencerateover1914and1915andthereforealmostcertainlyanunderestimateofactualinitialsubsidence.Alsoshownarelong-termcalculatedsubsidenceratesforSEAsia,applyingboththerelationdeterminedforFloridaEverglades(Stephensetal.,1984),assumingawaterdepthof0.7mandanaveragetemperatureof30C,andtherelationfoundforSEAsiainthispaper. Applying92%oxidationtotheaveragesubsidenceratemeasuredintheAcaciaplantation,6yronaverageafterdrainage,theresultingcarbonlossis68tCO2eqha�1yr�1(CO2equivalents,i.e.assuminginthiscalculationthatnocarbonislostasCH4,DOCorTOC).Fortheoilpalmsite,18yrafterdrainage,thisvalueis78tCO2eqha�1yr�1.Fortheseplantationsingeneral,6yrormoreafterdrainage,anaverageminimumlossof73tha�1yr�1maybeacceptedatanaveragewatertabledepthof0.71m(Table2).Whencalculatingtotalcumulativecarbonlossfromplan-tations,boththeveryhighlossintherst5yrandthelowerlossinthesubsequentperiodmustbeaccountedfor.Overa25yrperiod,theaverageannualcarbonlossthusbecomes90tCO2eqha�1yr�1forAcaciaplantationand109tha�1yr�1foroilpalmplantation,withanaverageof100tha�1yr�1forallplantations.Overa50yrperiod,thesevaluesbecome79tha�1yr�1and94tha�1yr�1re-spectively,withanaverageof86tha�1yr�1.3.7RelationshipsbetweensubsidencerateandwatertabledepthLinearcorrelationregressionsbetweensubsidencerateandaveragewatertabledepthweredeterminedseparatelyforAcaciaplantationanddrainednaturalforest.Therelation-shipbetweenwatertabledepthandsubsidenceinAcaciaplantation,6ormoreyearsafterdrainage,isasfollows(seeFig.5): SD1:5�4:98WDwhere: – Regression:N=125,F=33.38,p0:001,R2D0:21, – Intercept=1.50,SE=0.63,pD0:02, – Slope=�4.98,SE=0.86,p0:001, – S=annualsubsidenceofthepeatsurface(cmyr�1) – WD=averagewatertabledepthbelowthepeatsurface(�m;negative). Biogeosciences,9,1053– 1071 ,2012www.biogeosciences.net/9/1053/2012/ A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands1061 Fig.5.SubsidenceratesandwatertabledepthsasmeasuredinAcaciaplantations,oilpalmplantationsandadjacentforestinSumatra,Indonesia.Top:dataforindividualmonitoringlocations.MeasurementsinMalaysianoilpalmplantationsareshownforcomparison(fromDIDMalaysia,1996).ThelinearrelationsshownareforAcaciaplantations(excludingoilpalmoilplantations)andforest.Bottom:averagesfortheSumatraplantationdata,groupedby(sub-)transectsof5to9adjacentmonitoringlocations.LinearrelationsforFloridaEvergladesarealsoshown(adaptedfromStephensetal.,1984). Table2.Subsidenceratesandcarbonlossoverdifferenttimeperiods,asdeterminedfromsubsidenceandbulkdensitydata. AcaciaplantationsitesOilpalmplantationsitesPlantationaverage Totalsubsidenceinrst5yr(m)1.42––afterdrainageAveragesubsidenceandSD,�5yrcmyr�152.25.41.15.2afterdrainageCarbonloss0–5yrtCO2eqha�1yr�1178––afterdrainage(measured)Carbonloss0–18yrtCO2eqha�1yr�1–119–afterdrainage(measured)Carbonloss5–8yrtCO2eqha�1yr�168––afterdrainage(measured)Carbonloss18yrtCO2eqha�1yr�1–78–afterdrainage(measured)Carbonloss0–25yrtCO2eqha�1yr�190109100afterdrainage(calculated)Carbonloss0–50yrtCO2eqha�1yr�1799486afterdrainage(calculated) www.biogeosciences.net/9/1053/2012/Biogeosciences,9,1053– 1071 ,2012 1062A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands Fig.6.DevelopmentincarbonlossovertimeintheAcaciaandoilpalmplantationsstudied. Theanalysisofvarianceconrmsthatthelinearrelationshipbetweenpeatsubsidencerateandpeatwatertabledepthisstatisticallyvalidatthe0.05signicancelevel.Basedonthet-testoncoefcients,theinterceptisfoundtodiffersig-nicantly(pD0:02)fromzero.Therelationshipfordrainedforest,atwatertabledepthsof0–0.7m,is: SD0:41�6:04WDwhere: – Regression:N=51,F=26.43,p0:001,R2D0:35, – Intercept=0.41,SE=0.43,pD0:34, – Slope=�6.04,SE=1.17,p0:001.Thisanalysis,too,conrmsthatthereisastatisticallyvalidlinearrelationshipbetweenthepeatsubsidencerateandpeatwatertabledepth.Thet-testoncoefcientssuggeststhattheinterceptdoesnotdiffersignicantly(pD0:34)fromzero.Therefore,wemayrecommendasomewhatmodiedrela-tionwithaninterceptofzeroforuseinapplicationswhereawatertabledepthatthepeatsurface(i.e.ofzero)isassumedinemissioncalculationsforintactnaturalpeatlandforest: SD�7:06WDwhere: – Regression:N=51,F=197.12,p0:001,R2D0:80, – Slope=�7.06,SE=0.50,p0:001.Thecorrelationsfoundfortheseregressionrelationshipsin-cludinginterceptarenothigh,partlyduetodatalimitationsandpartlybecausefactorsotherthanwatertabledepthalsoinuencesubsidence(seeDiscussion),butalsoasaconse-quenceofsplittingthedatasetinto“Acaciaplantation”and“forest”subsampleswithlimitedwatertabledepthranges.Ifarelationisttedthroughthecombined“Acaciaplantation”and“forest”dataset,resultinginan“intermediate”relationthatmaybeappliedwherepeatlandlandcoverconditionsarenotclear,astrongerrelationshipisobtained: SD0:69�5:98WDwhere: – Regression:ND176,FD128:42,p0:001,R2D0:43, – Intercept=0.69,SE=0.34,p0:05, – Slope=�5.98,SE=0.53,p0:001,Thisanalysisagainconrmsavalidlinearrelationshipbe-tweenthepeatsubsidencerateandpeatwatertabledepth.Basedonthet-testoncoefcients,theinterceptisfoundtodiffersignicantly(p0:05)fromzero.3.8RelationshipsbetweencarbonlossandwatertabledepthTheCO2emissionsequivalenttopeatsubsidencelossescausedbydrainageinAcaciaplantationandnaturalforest(Fig.7)weredeterminedbyapplyinganoxidationpercent-ageof92%,acarboncontentof55%andabulkdensityof0.075gcm�3.Forplantations,theresultingrelationshipwithaveragewatertabledepthis: CLD21�69WD:Fordrainednaturalforest(usingthesubsidencerelationwithaninterceptthroughzero): CLD�98WD:Combined(fordeforestedunproductivepeatlands): CLD9�84WDwhere: CLDcarbonloss;intCO2eqha�1yr�1: Biogeosciences,9,1053– 1071 ,2012www.biogeosciences.net/9/1053/2012/ A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands1063 Fig.7.Comparisonoftherelationbetweencarbonloss(CO2eq)andwatertabledepthintropicalpeatlands,morethan5yrafterdrainage,asdeterminedinthisandotherstudies.TheFloridaEvergladesrelationiscalculatedfromdatainStephensandSpeir(1969),thatmayalsohavebeenusedtocalculatetherelationinW¨ostenandRitzema(2001).TherelationsbyHooijeretal.(2006,2010)andCouwenbergetal.(2010)werebasedonpartlydifferentsetsofliteraturesources.TherelationbyJauhiainenetal.(2012)isbasedondaytimeCO2uxmeasurementsinthesameAcaciaplantationasthecurrentstudy,excludingrootrespirationandcorrectedfordiurnaltemperatureuctuation. 4Discussion4.1Comparisonwithotherpublishedtropicalandsub-tropicalpeatsubsidenceratesTheaveragesubsidencerateof5cmyr�1foundforAcaciaandoilpalmplantations,morethan5yrafterinitialdrainage,isclosetomostliteraturevalues.SubsidenceratesreportedforaJohor(Malaysia)oilpalmplantation,between14and28yrafterdrainage,were4.6cmyr�1onaverageat17lo-cationsforwhichwatertabledepthdataarenotavailable(W¨ostenetal.,1997),and3.7cmyr�1at11otherlocationswithanaveragewatertabledepthof0.5m(DIDMalaysia,1996).Mohammedetal.(2009)report,onthebasisofeldmonitoringinoilpalmplantationsonpeatof3to4minthickness,thatsubsidencestabilizesat4.3cmyr�1after15yearsunderbestpracticemanagementwithaveragewa-terdepthsof0.4m.DIDSarawak(2001)reportedacon-stantaveragesubsidencerateof5cmyr�1inSarawakaf-tertheinitialtwoyearsfollowingdrainage,atawaterta-bledepthof0.6m,butalsoproposedthattherateofan-nualsubsidenceincreasedby1cmforevery10cmlower-ingofthewatertable,whichwouldresultin7cmyr�1sub-sidenceatawatertabledepthof0.7m.Andriesse(1988)suggestedastabilizationofsubsidenceatlong-termratesofupto6cmyr�1,basedonobservationsinanumberofloca-tionsinSEAsia.IntheEverglades,USA,anaveragesub-sidencerateof3cmyr�1wasreportedovermorethan50yraftertheinitialyear(StephensandSpeir,1969;Fig.4),butthiswasforadifferentpeattypeinasub-tropicalregionwithalowersurfacepeattemperatureof25C,comparedto30CinplantationsitesinIndonesia(Jauhiainenetal.,2012),andforhigherwatertablelevels.Applyinganequationthatre-latessubsidencetotemperatureandwatertabledepth,basedonlong-termcontrolledeldplotexperiments,Stephensetal.(1984)calculatedthattheEvergladespeatwouldhavesubsidedby8cmyr�1inthelongtermhaditbeeninfullytropicalconditionswithapeatsurfacetemperatureof30C.Inpeatlandwithaninitialorganiccontentofaround80%intheSacramentoDelta,California,subsidenceaftertheinitial5yrproceededataconstantrateof7.5cmyr�1forover50yr(DeverelandLeighton,2010;Fig.4).Thesestudiessupportourndingthatsubsidenceratesstabilizebetween4cmyr�1and5.5cmyr�1indrainedSEAsianpeatlands,ataveragewatertabledepthsaround0.7m,afteraninitialphaseofmorerapidsubsidence.Thevari-ationinreportedsubsidenceratesintropicalpeatseemstobesmallcomparedtotemperatepeats(Couwenbergetal.,2010).Thismayberelatedtotheeffectofsoiltemperature,whichismoreconstantintimeandspaceinthetropics(ataround30C)comparedtotemperateclimates,onpeatoxi-dation.ThestudiesinSEAsiamentionedaboveapplytodeepdepositsofbricpeatwithapre-drainageBDofaround0.07to0.1gcm�3andlowmineralcontent.Wherelowersubsidenceratesarereported,thisisusuallyforshallowerpeatwithhigherBDandmineralcontent.MurayamaandBakar(1996)reportedthatsubsidenceratesindrainedpeat-landsinPeninsularMalaysia,withbulkdensitiesbetween0.1and0.35gcm�3,were2to4cmyr�1aftertheinitialyear,andthatsubsidencedecreasedasBDincreased.Dradjadetal.(2003)reportedasubsidencerangeof2.4to5.3cmyr�1overa14yrperiod,inpeatyswampsoilaround2mthickwithamineralcontentashighas73to86%(i.e.notre-allypeat).DeverelandLeighton(2010)alsofoundastrongrelationshipbetweensoilorganiccontentandsubsidencerateintheSacramentoDelta,withsubsidenceratessome100yrafterinitialdrainagehavingdeclinedtolessthan1cmyr�1 www.biogeosciences.net/9/1053/2012/Biogeosciences,9,1053– 1071 ,2012 A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands1065 FortheFloridaEverglades,StephensandSpeir(1969)concludedthatsubsidencecontinuedataconstantrateforover50yraftertheinitialphase.ConstantsubsidenceisalsoobservedintheSacramentoDelta(DeverelandLeighton,2010;Fig.4).Whilstagradualreductiontoasubsidencerateof2cmyr�1,beyond28yrafterdrainage,wassuggestedbyW¨ostenetal.(1997),itshouldbenotedthatthisvaluewasbasedonanestimatedprojectionratherthanonlyonmea-surements.Subsidenceratespresentedinthisstudy,aswellasthosepresentedbyW¨ostenetal.(1997),areallaround5cmyr�1after6,18and14–28yrrespectively,suggestingaconstantsubsidencerateratherthanacleargradualdecrease.OnthebasisoftheevidenceavailableweconcludethatsubsidenceratesinAcaciaandoilpalmplantationsinSEAsia,morethan5yrafterdrainageatconstantwatertabledepths,arelikelytoremainconstantornearlyconstantataround5cmyr�1aslongasthereis“fresh”peatavailableforoxidation(Fig.4).Whenthepeatdepositisnearlyde-pletedandallremainingpeatiscompactedanddrained,thesubsidenceratewouldbeexpectedtodecline.Thisalsoap-plieswhen,asoxidationhasremovedtheupperlayers,lowerpeatlayersarebeingaccessedthatmayhavehigherBDandmineralcontent.Wethereforeemphasizethatthecurrentas-sessmentappliestobrictohemicpeatwithverylowmin-eralcontentandlowBD.4.4UnexplainedvariationinsubsidenceratesandwatertabledepthsSubsidenceratesdeterminedatindividualmonitoringloca-tionsshowconsiderablevariation,withvaluesvaryingfrom1.2to11.2cmyr�1(Fig.5).However,whenvaluesareaver-agedoversubgroupsof5to9adjacentmonitoringlocationsoverrelativelyshortdistancesalongtransects,thisrangenar-rowsto2.9–7.4cmyr�1(Table3,Fig.5).Thestandardde-viationinsubsidencerates,expressedasapercentageofthemeanvalue,isstillconsiderableat42%onaverageatthesubgrouplevel(Table3).Thisvariationinsubsidenceisnotmatchedbyvariationinwatertabledepthorpeatthick-ness,withstandarddeviationsof21%and9%ofthemean,respectively,overthesamesubgroups.Thissuggeststhatthevariationwithinthesegroups,betweenindividualmeasure-ments,isrelatedpartlytohighlylocalizedvariationsinphys-icalconditions,includingheterogeneityinthenear-surfacepeatandincanopycover,withthelatteraffectingpeatsur-facetemperature.Itfollowsthataccuratesubsidenceandwatertabledepthmeasurementrequiresalargenumberoflocationsmonitoredoverlongperiodstocovernotonlytheobviousvariationsinlandcoverandwatermanagement,butalsotheunknownrandomheterogeneities.Intheend,however,substantialun-explainedvariationinmeasuredsubsidenceislikelytoal-waysremainassomephysicalconditionsthatmayaffectcar-bonlossandsubsidencecannotorhardlybemeasured.Forexample,itisnotpossibletomeasurethevariationinpeatcharacteristics(BD,woodcontent)atthemicro-scalewith-outsamplingthepeatinapit,whichdestroysthemonitoringlocation.Thenatureofwatertabledepthmeasurementsmayalsoexplainpartofthevariation.The2-yearsubsidencerecordsusedinthisstudydidnotallcoverthesameperiod,althoughalloverlapbyatleast1yr,introducingdifferencesinrainfallandsoilmoistureregimeexperiencedbythedifferentloca-tions;moreoverdatagapsoccurredonsomeoftherecords,whichaffectedtheaveragewatertabledepthnumbersbutnotthesubsidencenumbers.4.5StudysitebulkdensitycomparedtoliteraturevaluesTheaverage“original”BDvalueof0.075gcm�3foundinpeatbelow1mdepthinthisstudy,translatingtoavalueof0.07gcm�3priortoconsolidation,isatthelowendofval-uesforSEAsianpeatspresentedbyPageetal.(2011),whondthataveragevaluesreportedby15individualstudies,inbothintactanddeforestedpeatlands,arebetween0.08and0.13gcm�3withanaverageof0.09gcm�3.Lower-rangevaluesreportedbyindividualstudiesarebelow0.06gcm�3inonly3outof15studies,andall15reportaverageval-uesabove0.07gcm�3.Indeforestedpeatlands,Pageetal.(2011)ndBDvaluesofnear-surfacepeattobehigherthanatgreaterdepth,butinforestedpeatlandsnoconsistentincreaseordecreasewithdepthwasidentied.Thislackofacleartrendwithdepth,atdepthsbetween0and4m,isalsoconrmedbypeatprolesinprimaryandsecondaryforestinKalimantan(Indonesia)(Kooletal.,2006;Ansharietal.,2010)(Fig.3).Thevalueof0.061gcm�3over1–2mdepthderivedfromdatausedbyAnsharietal.(2010)isthelowestaveragere-portedforanygroupofpeatprolesinundrainedpeatlandforestinSEAsia.Bycontrast,dataforintactforestpresentedbyKooletal.(2006)yieldanaverageof0.074gcm�3over1–4m,whichishigherthanthepre-consolidationvalueof0.07gcm�3appliedinthecurrentstudy.Onthebasisoftheaboveassessmentofliteraturevalues,andnotingthattheBDvaluesbelow1matthedifferentlo-cationsinthecurrentstudyareallverysimilar,weconcludethatourBDvaluesbelow1mareindeedrepresentativeofthepre-drainageconditions,somewhatincreasedbyprimaryconsolidation.Therefore,totalcompactionsincethestartofdrainage,aftertheinitial1yrconsolidationphase,maybeestimatedbycomparingthecurrentBDofpeatbelow1mdepthwiththataboveitasexplainedinSects.2.5and3.6.4.6Determiningthecarbonlossfromoxidation,asapercentageoftotalsubsidenceVeryfewstudieshaveseparatedtheoxidationandcom-pactioncomponentsofsubsidenceintropicalandsub-tropicalpeatlandsusingBDproles.Stephensand www.biogeosciences.net/9/1053/2012/Biogeosciences,9,1053– 1071 ,2012 1066A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands Table3.Summarystatisticsofwaterdepth,peatthicknessandsubsidenceratealongallplantationtransects,ascalculatedovergroupsof5to9adjacentmeasurementlocationseach.Mean,maximumandminimumvaluesarecalculatedfromaveragevaluesforindividuallocations.The“Riau”locationsareinAcaciaplantation,the“Jambi”locationsinoilpalm. (Sub-)transectNoofmon. Watertabledepth Peatthickness Subsidencecodepoints MeanMinMaxSDSD MeanMinMaxSDSD MeanMinMaxSDSD mmmm%mean mmmm%mean cmyr�1cmyr�1cmyr�1cmyr�1%mean ARiau6 �0.56�0.72�0.410.1120 �5.3�5.7�4.60.48 5.94.29.51.932BRiau6 �0.63�0.77�0.470.1016 �6.7�7.2�6.20.57 5.23.47.72.140CRiau6 �0.54�0.80�0.290.1121 �7.8�10.2�7.10.68 4.52.27.82.146DRiau5 �0.72�0.91�0.650.1014 �7.8�8.6�6.90.57 5.73.17.42.442ERiau6 �0.84�1.05�0.560.1011 �8.1�8.4�7.80.57 5.62.310.42.138FRiau6 �0.56�0.69�0.430.1119 �8.6�9.1�8.10.67 4.01.56.31.948GRiau5 �0.43�0.56�0.280.1126 �11.6�11.8�11.50.76 3.41.65.12.059HRiau5 �0.42�0.50�0.350.1125 �11.8�11.9�11.60.54 2.91.54.61.551IRiau5 �0.67�0.97�0.470.1117 �6.0�7.7�5.10.46 3.81.36.21.437KRiau5 �0.74�0.83�0.620.1115 �8.3�9.3�7.70.33 4.71.611.01.736LRiau5 �0.61�0.71�0.550.1219 �12.2�13.1�11.60.22 3.11.24.11.757MRiau5 �0.52�0.60�0.460.1324 �14.6�16.9�13.30.11 3.52.34.51.953NRiau5 �0.74�0.80�0.660.1014 �12.8�14.7�10.20.11 5.44.56.21.834ORiau9 �0.71�0.99�0.490.1217 �12.7�14.4�10.50.33 5.84.38.51.932PRiau5 �0.88�1.07�0.780.1315 �8.7�9.3�8.40.34 5.33.07.92.038QRiau5 �0.73�0.81�0.620.1318 �8.8�9.4�8.40.44 3.32.34.52.369RRiau6 �0.69�0.81�0.580.2131 �9.0�10.0�8.51.415 4.01.45.72.461SRiau5 �0.73�0.81�0.660.2635 �4.6�5.9�3.81.533 7.43.111.22.432TRiau7 �0.75�0.97�0.550.2634 �8.3�10.1�7.51.518 7.33.710.52.433URiau7 �0.93�1.26�0.650.2123 �8.6�9.0�8.01.315 6.44.78.42.234VRiau6 �1.08�1.19�0.970.2220 �8.0�8.0�8.01.113 5.94.39.82.2371Jambi9 �0.75�0.84�0.660.1925 �6.4�7.2�6.01.016 5.34.56.02.0382Jambi8 �1.06�1.14�1.000.1918 �6.5�8.1�6.01.016 4.93.56.52.0413Jambi9 �0.73�0.90�0.510.2231 �6.9�8.5�5.61.218 4.83.56.01.9404Jambi9 �0.73�0.77�0.690.1216 �9.2�10.7�8.80.77 6.15.08.01.9315Jambi7 �0.32�0.34�0.300.1133 �9.2�9.9�8.70.66 5.94.08.01.831 AllRiau125 �0.70�0.86�0.560.1420 �8.89�9.81�8.170.628 4.922.767.511.9943AllJambi42 �0.72�0.80�0.630.1725 �7.64�8.87�7.010.9113 5.404.106.901.9336All167 �0.71�0.85�0.570.1421 �8.66�9.63�7.950.679 5.013.017.401.9842 Speir(1969)calculatedthatoxidationaccountedfor78%ofsubsidenceintheEvergladespeatlandsoveraperiodofmorethan50yrsincedrainage;thiswasconrmedbyCO2uxmeasurementsatthesamesites(Neller,1944)andunderlaboratoryconditions(Volk,1973).Inareassessmentofthesamedata,Stephensetal.(1984)estimatedthattheincreaseinBDineldplotsreportedbyNeller(1944)explainedonly10to15%ofsubsidence,implyingthat85to90%couldbeattributedtocarbonloss.DeverelandRojstaczer(1996)andDeverelandLeighton(2010)foundthatoxidationaccountedfor68%ofsubsidenceinCalifornianpeatlandsmorethan70yrafterdrainage,basedonCO2uxmeasurementsandacarbonbalancemodel.Thisvalueapplies,however,topeatwithamineralcontentthatis20%higherthanintheEv-ergladesorSEAsia,whichisexpectedtoreducetherela-tivecontributionofoxidationtosubsidence.BasedonCO2uxmeasurements,MurayamaandBakar(1996)concludedthatoxidationcaused50%to70%ofsubsidenceatsitesinMalaysia;howeverthiswasalsoforshallowpeatwithhighmineralcontent.InpeatlandplantationsinJohor,Malaysiaacumulativevalueof61%wasreported(DIDMalaysia,1996;W¨ostenetal.,1997);however,theauthorsdonotexplainoverwhatperiodafterdrainagethisvalueapplies.Moreover,thesamestudyalsoreportscompletelossofthepeatlayer,i.e.100%oxidation,atuptoonethirdofthesubsidencemonitoringlocationswherepeatwasthinatthestartofmoni-toring.Couwenbergetal.(2010)assumedaminimumoxida-tionpercentageof40%,butthiswasbasedmostlyonstudiesintemperateclimateswhereasothers(Volk,1973;Stephensetal.,1984;Brady,1997)haveshownthatpeatoxidationincreasesathighertemperatures,causingadoublingofsub-sidencerateforevery10Cincrease.Assuminganaveragepeatsurfacetemperatureof10Cintemperatepeatlandsand30Cinthetropics,theoxidationratewouldbeexpectedtobefourtimeshigherinthelatterandmakeupafarlargerproportionofsubsidence.PeattemperatureanditspotentialimpactontropicalpeatoxidationisdiscussedinmoredetailinJauhiainenetal.(2012).Whiletheoxidationcontributiontopeatsubsidencein-creasesovertherstfewyearsafterdrainageasprimaryconsolidationandcompactiondiminish,thenetcarbonlossinfactdecreasesoverthisperiod,beforestabilizing.Itmaybethatanitepoolofthemostlabilecarboncom-poundsdecomposesrapidly,leavingonlyrecalcitrantcar-boncompoundsthataremoreresistanttodecomposition(Berg,2000).Inaddition,alowerwatertableinthisini-tial“dewateringphase”,appliedtorapidlyconsolidatethepeatsurfacebeforeplanting,mayfurtherincreaseoxida-tion.Asimilarndingofaninitial“spike”incarbonlosswasreportedfortheSacramentoDeltabyDevereland Biogeosciences,9,1053– 1071 ,2012www.biogeosciences.net/9/1053/2012/ A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands1067 Leighton(2010),whocalculatedthatemissionsreducedfrom154tCO2eqha�1yr�1afewyearsafterdrainageto55tha�1yr�180yrlater,atanaveragewatertabledepthof1m(Drexleretal.,2009).ThefactthattheBDprolesinourstudysitesat3–7and18yrafterdrainageareverysimilar(Fig.3)indicatesthatnotonlyprimaryconsolidationbutalsocompactionmayinfacthavebecomenegligibleaftertherst5yr,atwhichpointnearly100%ofsubsidenceappearstobecausedbyoxida-tion.Aftertheinitialyearofsubsidence,ratesofcompactionandoxidationmayachieveequilibrium.Compactioncon-tinuesasuncompactedpeatfromthesaturatedzoneenterstheunsaturated,oxidativezone.Howeverthiscompactionappearstobebalancedbyoxidationinanunsaturatedpeatproleofconstantthicknessandbulkdensitythatmovespro-gressivelydownwardsovertimeasthewatertableisloweredtomatchsurfacesubsidence.Thecombinedresultofthetwoprocessesisapeatbulkdensityprolethatisstableintime.Thelong-termoxidationcontributiontosubsidenceof92%inoilpalmplantations,18yrafterdrainage,isatthehighendofearlierestimates.Ourstudyconrms,how-ever,thatthecontributionofoxidationtopeatsubsidenceincreasesintimewhileconsolidationandcompactionaremajorcontributorsonlyintheinitialperiod.Itshouldbenotedthatmostpublishedpercentageoxidationvalues,in-cludingthe61%reportedbyDIDMalaysiaandW¨ostenetal.(1996,1997)andthe85–90%suggestedbyStephensetal.(1984),areaveragesofthecumulativeoxidationsincethestartofdrainage,andthereforesystematicallyunderestimatethepercentageoxidationaftertheinitialperiod.Thelongertheperiodafterdrainagethatisconsidered,thegreaterthecumulativecontributionofoxidationtosubsidencethatwillbefound.Forcalculationsoflong-termcarbonemissionswethereforerecommenduseofthegureof92%oxidationthatwendforoilpalmplantation18yrafterdrainage,ratherthanthe75%thatwefoundforAcaciaplantations6yrafterdrainage.4.7SensitivityassessmentThecalculationofthepercentageoxidationcontributiontosubsidence,andtheresultingcarbonloss,issensitivetothevalueusedfortheoriginalpre-drainageBD,correctedforconsolidationimmediatelyafterdrainage.InSect.4.5wehaveshownthatthepre-drainageBDvaluesaround0.075gcm�3usedinouranalysis,resultingfromavalueof0.07gcm�3allowingforprimaryconsolidationimmediatelyafterdrainage,areatthelowendofpublishedvalues.Wehavealsoshownthatthelowestaveragepre-drainagevaluereportedinanystudyis0.061gcm�3(usingdataofAnsharietal.,2010).Thiswouldincreasetoaround0.065gcm�3allowingforconsolidation.Usingthelattervalueinsteadofthe0.078gcm�3onaverageappliedtotheoilpalmplan-tationinthisstudywouldhaveyieldedanoxidationper-centageof77%at18yrafterdrainageinsteadof92%,suggestingthatcarbonlossmorethan5yrafterdrainagecouldbeattheverymost20%lowerthantheaveragevalueof73tCO2eqha�1yr�1proposedinthispaper,i.e.around60tCO2eqha�1yr�1.However,thislow-endgureisun-likelytoapply,assuchlowBDvaluesappeartobeexcep-tional.Forassessingthespikeincarbonlossintherstyearaf-terdrainage,theestimateofprimaryconsolidationintherstyearisalsoasensitiveparameter.Ifwewouldassumeallthe75cmofsubsidenceintherstyeariscausedbyprimaryconsolidationonly,ratherthanallowingfor19cmbeingcausedbyoxidationandcompaction(Sect.3.5),theaveragecarbonlossovertherst5yrbecomes132tCO2eqha�1yr�1ratherthan178tCO2eqha�1yr�1.Thiswouldalsodecreasethepercentageoxidationthatiscalculatedovertheperiod,fromthe75%nowcalculatedforAcaciaplantationover5yrafterdrainage(Sect.3.6)to69%,andfrom92%foroilpalmplantationover18yrto90%.Whilewedeemtheunderly-ingassumptionofanabsenceofoxidationintherstyrtobeunrealistic,thesevaluesof132tha�1yr�1emissionovertherst5yrandalong-termoxidationpercentageof90%maybeseenasthelowestpossibleestimatesonthebasisoftheevidenceavailable.Thevalueforcarboncontentusedhasaproportionaleffectonthecarbonlosscalculatedfromsubsidence.Assumingcarboncontentof50%or60%insteadof55%,whichcoverstherangereportedinliteratureforbricandhemicpeatwithlowmineralcontent,wouldreduceorincreasecarbonlossby10%.4.8ComparisonofcarbonlossinsubsidencewithCO2emissionmeasurementsGaseousCO2emissionsatthepeatsurfaceinthesameAcaciaplantationlandscapehavebeenmeasuredusingtheclosedchambertechniqueat144locations(Jauhiainenetal.,2012).MeasuresweretakentoexcluderootrespirationsotheresultsonlyrepresentCO2emissionsfrompeatoxida-tion.Aftercorrectionfordiurnaltemperatureuctuations,thesemeasurementsfromthesamepeatlandyieldavalueof80tha�1yr�1atanaveragewatertabledepthof0.8m,whichisveryclosetothe76tCO2eqha�1yr�1yieldedbythesubsidencemethodforthesamewatertabledepth.More-over,theslopeoftherelationshipbetweenwatertabledepthandCO2emissionpresentedbyJauhiainenetal.(2012)isnearlyidenticaltothatusingthesubsidencemethodpre-sentedinthispaper(Fig.7).Weconcludethattheresultsofthetwoindependentapproachesaremutuallysupportive.4.9ComparisonwithotherpublishedCO2emissionsfromtropicalpeatlandAtwatertabledepthsbetween0.5and1m,thataremostcommoninplantations,theemissionrelationsfoundforAcaciaplantationsanddrainedforestaresimilartothe www.biogeosciences.net/9/1053/2012/Biogeosciences,9,1053– 1071 ,2012 1068A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands Fig.8.TimeseriesofwatertabledepthasmeasuredatindividuallocationsinthestudiedAcaciaandoilpalmplantations,andinnearbynaturalforestat2kmfromtheAcaciaplantation,overa3-yearsperiod.Inplantations,therecordsnearestthelowerandupper10-percentileaveragewaterlevelswereselected. linearrelationshipreportedbyHooijeretal.(2006,2010)andCouwenbergetal.(2010),thatwerebasedonmeta-dataassessmentsofstudiescarriedoutindeforestedtrop-icalpeatlands(Fig.7).AsimilarCO2eqemissionvaluewasalsoobtainedbyDIDSarawak(2001)andW¨ostenandRitzema(2001)whoproposedthatevery1.0cmofsubsi-denceresultsinaCO2eqemissionof13.3tha�1yr�1,equat-ingtoatotalCO2eqemissionof66tha�1yr�1atthesubsi-dencerateof5cmyr�1reportedinthesamepublications.Weconcludethatthecarbonlossesfoundinthisstudy,morethan5yrafterdrainage,areinagreementwithmostearlierstudies,forwatertabledepthsthatarecommoninplantationsandfortheperiodbeyondtheinitialyearsafterdrainage.Atlesserwatertabledepths,thedifferencewithexistingrelationshipsincreases,suggestinghigheremissionsfromdrainedpeatlandsthanhavebeenassumedtodateandastrongerrelationshipwithvegetationcover,andperhapswithfertilizationandpeatdisturbanceaswell.Moreover,wefoundthatcarbonlossintheinitialyearishigherthaninsubsequentyears,resultinginconsiderablyhigherlong-termaverageemissionsthanhavebeenreportedtodate.4.10PredictingsubsidenceandcarbonlossunderdifferentwatermanagementregimesTheaveragewatertabledepthsencounteredinthisstudyaresimilarinbothAcaciaandoilpalmplantations,at0.7and0.73mrespectively,whichislessthanthosereportedinsomeearlierstudies(e.g.0.95minHooijeretal.,2006,2010)andclosetothetargetof0.7mspeciedfortheAcaciaplanta-tionsstudied(Hooijeretal.,2009)andof0.6mforoilpalmplantationsingeneral(DIDSarawak,2001).However,thisdoesnotsuggestthathighandwell-controlledwaterlevelsarethenorminsuchplantations.Thelimitedoptionsforef-fectivewaterlevelcontrolareillustratedbythewiderangeoflevelsencounteredinthestudy,with10-percentilevaluesforannualaveragesrangingbetween0.33and1.03mandforin-dividualmeasurementsbetween0and1.6m(Fig.8).Watertabledepthcanvarybyuptoametreoverafewkilometresineachtypeofplantation,andalsoovertimewithinthedryandwetseasons.Itshouldalsobenotedthatourmeasurementswereobtainedinarelatively“wet”yearwithhighrainfalleveninthedryseason,andinplantationsthatarerelativelywellmanagedcomparedtoothersintheregion.Waterta-bledepthvariationsinnormalyears,andinotherareas,arelikelytobegreater.Theimplicationoftherelativelylowslopeoftheregres-sionbetweenwatertabledepthandsubsidencefoundinthisstudy,isthatthebenetofraisingwatertablestoreducecar-bonemissionsinplantationsmaybesmallerthanearlieras-sumed.Evenifanaveragewatertabledepthof0.6mcouldbeachievedintheplantationsnowstudied,whichisnotguar-anteed,subsidencewouldstillbe4.5cmyr�1overthelongterm,andthecarbonloss63tCO2eqha�1yr�1morethan5yrafterdrainage.Thiswouldbeareductionofnomorethan20%relativetotheaverageemissionscurrentlyoccur-ring.Moreover,itshouldbenotedthatthisreductioninan-nualsubsidenceandemissionmerelymeanstheyarepost-ponedtoalaterdate,unlessnaturalconditionscouldbere-storedintheseplantations.5ConclusionsWeshow,foramuchlargernumberoflocationsthanallpre-viousstudiesinSEAsiaonthissubjectcombined,thatmea-surementsofsubsidenceandbulkdensitycanyieldaccuratesoilcarbonlossvaluesfortropicalpeatlands,ifthecontribu-tionsfromthedifferentprocessesofoxidation,compactionandconsolidationcontributingtosubsidenceareaccountedfor.Thisreducestheuncertaintyofcarbonlossestimatescomparedtoearlierpeatsubsidenceandgaseousemissionstudies.Thisstudyisalsothersttodeterminecarbonlossfromtropicalpeatlandfromsubsidenceinparallelwithdirect Biogeosciences,9,1053– 1071 ,2012www.biogeosciences.net/9/1053/2012/ 1070A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands Berry,P.L.:Applicationofconsolidationtheoryforpeattothedesignofareclamationschemebypreloading,Q.J.Eng.Geol.,16,103–112,1983. Brady,M.A.:Organicmatterdynamicsofcoastalpeatde-positsinSumatra,Indonesia,Ph.D.thesis,UniversityofBritishColumbia,Vancouver,258pp.,1997. Couwenberg,J.,Dommain,R.,andJoosten,H.:GreenhousegasuxesfromtropicalpeatlandsinsoutheastAsia,GlobalChangeBiol.,16,1715–1732,2010. Deverel,S.J.andRojstaczer,D.A.:SubsidenceofagriculturallandsintheSacramento-SanJoaquinDelta,California:Roleofaqueousandgaseouscarbonuxes,WaterResour.Res.,32,2359–2367,1996. Deverel,S.J.andLeighton,D.A:Historic,recent,andfuturesub-sidence,Sacramento-SanJoaquinDelta,California,USA,SanFranciscoEstuaryandWatershedScience,8,23pp.,2010. DIDMalaysia:WesternJahoreintegratedAgriculturalDevelop-mentProject,PeatSoilManagementStudy,DepartmentofIr-rigationandDrainage,KualaLumpurandLandandWaterRe-searchGroup(LAWOO),Wageningen,100pp.,1996. DIDSarawak:Watermanagementguidelinesforagriculturaldevel-opmentinlowlandpeatswampsofSarawak,ReportoftheDe-partmentofIrrigationandDrainage,Sarawak,Malaysia,78pp.,2001. Dradjad,M.,Soekodarmodjo,S.,Hidayat,M.S.,andNitisapto,M.:SubsidenceofpeatsoilsthetidalswamplandsofBarambai,SouthKalimantan,JurnalIlmuTanahdanLingkungan,4,32–40,2003. Drexler,J.Z.,DeFontaine,C.S.,andDeverel,S.J.:ThelegacyofwetlanddrainageontheremainingpeatintheSacramento-SanJoaquinDelta,JournaloftheSocietyofWetlandsScientists,29,372–386,2009. Driessen,P.M.andSoepraptohardjo,M.:SoilsforagriculturalexpansioninIndonesia,PublicationofSoilResearchInstitute,Bull.1,Bogor,Indonesia,41–63,1974. Ewing,J.M.andVepraskas,M.J.:Estimatingprimaryandsec-ondarysubsidenceinanorganicsoil15,20,and30yrafterdrainage,WETLANDS,26,119–130,2006. Gambolati,G.,Putti,M.,Teatini,P.,andStori,G.G.:SubsidenceduetopeatoxidationanditsimpactondrainageinfrastructureinafarmlandcatchmentsouthofVeniceLagoon,RMZ–MaterialsandGeoenvironment,50,125–128,2003. Hambright,K.D.andZohary,T.:LakesHulaandAgmon:destruc-tionandcreationofwetlandecosystemsinnorthernIsrael,Wetl.Ecol.Manag.,6,83–89,1998. Hergoualc'h,K.andVerchot,L.V.:Stocksanduxesofcar-bonassociatedwithlandusechangeinSoutheastAsiantropi-calpeatlands:Areview,GlobalBiogeochem.Cy.,25,GB2001, doi:10.1029/2009GB003718 ,2011. Hirano,T.,Jauhiainen,J.,Inoue,T.,andTakahashi,H.:Controlsonthecarbonbalanceoftropicalpeatlands,Ecosystems,12,873–887,2009. Hooijer,A.:Hydrologyoftropicalwetlandforests:recentresearchresultsfromSarawakpeatswamps,in:Forests-Water-PeopleintheHumidTropics,editedby:Bonell,M.andBruijnzeel,L.A.,CambridgeUniversityPress,447–461,2005. Hooijer,A.,Silvius,M.,W¨osten,H.,andPage,S.:PEAT-CO2As-sessmentofCO2emissionsfromdrainedpeatlandsinSEAsia,DelftHydraulicsreportQ3943,36pp.,2006. Hooijer,A.,Page,S.,andJauhiainen,J.:KamparPeninsulaSci-enceBasedManagementSupportProject,InterimSummaryRe-port2007–2008;rstndingsonhydrology,watermanagement,carbonemissionsandlandscapeecology,Deltaresreport,37pp.,2009. Hooijer,A.,Page,S.,Canadell,J.G.,Silvius,M.,Kwadijk,J.,W¨osten,H.,andJauhiainen,J.:CurrentandfutureCO2emis-sionsfromdrainedpeatlandsinSoutheastAsia,Biogeosciences,7,1505–1514, doi:10.5194/bg-7-1505-2010 ,2010. Hutchinson,J.N.:TherecordofpeatwastageintheEastAnglianFenlandsatHolmePost,1848–1978AD,J.Ecol.,68,229–49,1980. Jauhiainen,J.,Takahashi,H.,Heikkinen,J.E.P.,Martikainen,P.J.,andVasander,H.:Carbonuxesfromatropicalpeatswampforestoor,Glob.ChangeBiol.,11,1788–1797,2005. Jauhiainen,J.,Limin,S.,Silvennoinen,H.,andVasander,H.:Car-bondioxideandmethaneuxesindrainageaffectedtropicalpeatbeforeandafterhydrologicalrestoration,Ecology,89,3503–3514,2008. Jauhiainen,J.,Hooijer,A.,andPage,S.E.:Carbondioxideemis-sionsfromanAcaciaplantationonpeatlandinSumatra,Indone-sia,Biogeosciences,9,617–630,doi:10.5194/bg-9-617-2012,2012. Koh,L.P.,Miettinen,J.,Liew,S.C.,andGhazoul,J.:Remotelysensedevidenceoftropicalpeatlandconversiontooilpalm,Proc.Natl.Acad.Sci.USA,108,5127–5132,2011. Kool,D.M.,Buurman,P.,andHoekman,D.H.:OxidationandcompactionofacollapsedpeatdomeinCentralKalimantan,Geoderma,137,217–225,2006. Leifeld,J.,M¨uller,M.,andFuhrer,J.:Peatlandsubsidenceandcarbonlossfromdrainedtemperatefens,SoilUseManage.,27,170–176,2011. Malhi,Y.:Thecarbonbalanceoftropicalforestregions,1990–2005,EnvironmentalSustainability,2,237–244,2010. Melling,L.,Hatano,R.,andGoh,K.J.:SoilCO2uxfromthreeecosystemsintropicalpeatlandofSarawak,Malaysia,TellusB,57,1–11,2005. Mesri,G.andAljouni,M.:Engineeringpropertiesofbrouspeats,J.Geotech.Geoenviron.,133,850–866,2007. Miettinen,J.andLiew,S.C.:Degradationanddevelopmentofpeat-landsinPeninsularMalaysiaandintheislandsofSumatraandBorneosince1990,LandDegrad.Dev.,21,285–296,2010. Miettinen,J.,Hooijer,A.,Tollenaar,D.,Page,S.,Malins,C.ChenghuaShi,C.,andSooChinLiew,S.C.:Historicalanal-ysisandprojectionofoilpalmplantationexpansiononpeatlandinSEAsia,WhitePaperno.17oftheInternationalCouncilofCleanTransportation,Washington,2012. Mohammed,A.T.,Othman,H.,Darus,F.M.,Harun,M.H.,Zam-bri,M.P.,Bakar,I.A.,andW¨osten,H.:Bestmanagementprac-ticesonpeat:watermanagementinrelationtopeatsubsidenceandestimationofCO2emissioninSessang,Sarawak,Proceed-ingsofthePIPOC2009InternationalPalmOilCongress(Agri-culture,BiotechnologyandSustainability),2009. Moore,S.,Gauci,V.,Evans,C.D.,andPage,S.E.:FluvialorganiccarbonlossesfromaBorneanblackwaterriver,Biogeosciences,8,901–909, doi:10.5194/bg-8-901-2011 ,2011. Murayama,S.andBakar,Z.A.:Decompositionoftropicalpeatsoils,estimationofinsitudecompositionbymeasurementofCO2ux,JARQ-JPN.Agr.Res.Q.,30,153–158,1996. Biogeosciences,9,1053– 1071 ,2012www.biogeosciences.net/9/1053/2012/ A.Hooijeretal.:Subsidenceandcarbonlossindrainedtropicalpeatlands1071 Murdiyarso,D.,Hergoualch,K.,andVerchot,L.V.:Opportunitiesforreducinggreenhousegasemissionsintropicalpeatlands,P.Natl.Acad.Sci.USA,107,19655–19660,2010. Neller,J.R.:Oxidationlossoflowmoorpeatineldswithdifferentwatertables,SoilSci.,58,195–204,1944. Page,S.E.,Siegert,F.,Rieley,J.O.,Boehm,H.D.V.,Jaya,A.,andLimin,S.:TheamountofcarbonreleasedfrompeatandforestresinIndonesiaduring1997,Nature,420,61–65,2002. Page,S.E.,Rieley,J.O.,andBanks,C.J.:Globalandregionalimportanceofthetropicalpeatlandcarbonpool,GlobalChangeBiol.,17,798–818,2011. Schothorst,C.J.:SubsidenceoflowmoorpeatsoilsintheWesternNetherlands,Geoderma,17,265–291,1977. Stephens,J.C.andSpeir,W.H.:SubsidenceoforganicsoilsintheUSA,IAHS-AIHSPublication,89,523–534,1969. Stephens,J.C.,Allen,L.H.,andChen,E.:Organicsoilsubsi-dence,GeologicalSocietyofAmerica,ReviewsinEngineeringGeology,Vol.VI,107–122,1984. Suhardjo,H.andWidjaja-Adhi,I.P.G.:Chemicalcharacteristicsoftheupper30cmofpeatsoilsfromRiau,Sumatra(Indonesia),in:FinalReport,AgriculturalTechnicalAssistanceProgramme(In-donesia–TheNetherlands)1974–1977,74–92,LembagaPeneli-tianTanah,Bogor,Indonesia,1977. Vernimmen,R.R.E.,Hooijer,A.,Mamenun,Aldrian,E.,andvanDijk,A.I.J.M.:Evaluationandbiascorrectionofsatelliterain-falldatafordroughtmonitoringinIndonesia,Hydrol.EarthSyst.Sci.,16,133–146, doi:10.5194/hess-16-133-2012 ,2012. Volk,B.G.:Evergladeshistosolsubsidence–Part1:CO2evolutionasaffectedbysoiltype,temperatureandmoisture,ProceedingsSoilandCropScienceSocietyofFlorida,32,132–135,1973. W¨osten,J.H.M.,Ismail,A.B.,andvanWijk,A.L.M.:Peatsub-sidenceanditspracticalimplications:acasestudyinMalaysia,Geoderma,78,25–36,1997. W¨osten,J.H.M.andRitzema,H.P.:LandandwatermanagementoptionsforpeatlanddevelopmentinSarawak,Malaysia,Interna-tionalPeatJournal,11,59–66,2001. www.biogeosciences.net/9/1053/2012/Biogeosciences,9,1053– 1071 ,2012