towards a comparable quantification of resilience johannes ingrisch论文.pdf

该【towards a comparable quantification of resilience johannes ingrisch论文 】是由【小舍儿】上传分享，文档一共【9】页，该文档可以免费在线阅读，需要了解更多关于【towards a comparable quantification of resilience johannes ingrisch论文 】的内容，可以使用淘豆网的站内搜索功能，选择自己适合的文档，以下文字是截取该文章内的部分文字，如需要获得完整电子版，请下载此文档到您的设备，方便您编辑和打印。?cationofResilienceJohannesIngrisch1andMichaelBahn1,*-?,thewayresilienceisquanti?edaffectsthedegreetowhichdifferenttrajectoriesofecosystemrecoveryfromdisturbancearerepresentedasPreviousresiliencemetricsdonotper-‘resilient’,parisonofdisturbanceresponsesacrossecosys-,‘resilience’(seeGlossary)todescribethemanifoldresponsesofecosystemstodisturbances(.,climateextremes)[1].Disturbancescancausetransientchangesinecosystempropertiesandprocesses(.,temporarychangesofproductivity[2–4])anizationofanecosystembycausingashifttoanalternativestablestate[5](.,fromforesttosavannah[6]).Accordingly,ple-?rstfocusesonthetransientimpactofdisturbanceandthesubsequentrecoveryofanecosystemandhasbeentermed‘engineeringresilience’(.,[80_TD$IF][7–9]).Thesecondviewconsidersresilienceastheabilityofanecosystemtowithstandashifttoanalternativestateinthefaceofdisturbance(‘ecologicalresilience’)[9,10].Theincreasingrecognitionoftheimplicationsofdisturbancesforecosystemsandsocietyhasponents[1,8,11–15],withpartlycontradictoryde?nitionsandterminologiesproposedalong[8,12]andbeyond[16–20],ithasalsobroughtforthamuch-neededexplorationofourunderstandingofthemechanismsunderlyingresilience[8,11,15,16,21,22].Theattempttoquantifyecosystemresponsestodisturbancehasledtothedevelopmentofalargenumberof‘indicators’or‘metrics’ofresilience,,UniversityofofcriticaltransitionsbetweenalternativestablestatesandhighlightedthatonlyafewoftheInnsbruck,,6020theoreticallysuggestedmetricshavebeenexploredin?eldsituations[23].Bycontrast,abroadInnsbruck,Austriarangeofmetricshasbeenappliedtoobservationalandexperimentalstudiestoquantify‘ecosystemrecoveryfromdisturbancewithinagivenstabilitydomain(.,engineeringresil-*Correspondence:ience’).However,parableMichael.******@().TrendsinEcology&Evolution,MonthYear,,:///.??:stateofanassessingwhichecosystems,andwhichecosystempropertiesandprocesses,,intensities,:,-disturbancestate(baseline-normalized).Perturbation:cumulativereductionQuantifyingResilience:pare?inecosystemstateorfunctionThemetricsoftheresilienceofecosystemresponsetodisturbancewithinastabilitydomainintegratedovertherecoverytimequantifythecapacitiestoresistandtorecoverfromdisturbance(Box1)(notethatinthis(seegrayareainFigure3FigureIAinal.[12]).ThemetricsBox1).Recovery:capacityofanecosystemappliedsofarintheliteraturecanbegroupedintothreemajortypesthatdifferfundamentallytoreturntoundisturbedecosystemponentofresiliencetheyaddress(Table1).Thesimplestcategorystateandfunctioningfollowinga(prisesmetricsdescribingtheimpactofadisturbanceonanecosystemstatedisturbance.(orprocessrate)inrelationtothebaseline(seeFigureIBinBox1).ThebaselineistypicallyRecoveryrate:rateatwhichanecosystemstateorfunctionrecoversconsideredthepre-,relativeprimarilyonrecovery;thatis,ontheamountbywhichthesystemhasrecovered(post-tothepre-disturbancestate(baselinenormalized),orrelativetothedisturbancechangeinstate).Recoveryfromdisturbancecanbeexpressedrelativetothedisturbanceimpact(impactbaseline(CategoryII)orrelativetothedisturbanceimpact(CategoryIII).Thereby,themetricsnormalized).fallingintoCategoryIIindicatebywhatpercentageofthepre-disturbancestatethesystemhasRecoverytime:timefromtheendrecovered,(Box1).Resilience:capacityofanecosystemtopersistandmaintainitsUnfortunately,variousstudieshaveappliedthesameterminologyacrossallcategoriesstateandfunctionsinthefaceof(Table1).Furthermore,?capacitytoreducetheimpactusedindifferentstudiesleadtoverydifferentquanticationsofanygivenresponsetrajectory(resistance)andthecapacityto(Figure1).Forexample,inCategoryIfullrecoveryfromdisturbanceimpliesconvergenceoftherecoverfromtheimpactofmetricto0orto1,dependingonhowauthorscalculateresilience(Table1andFigure1A,B).disturbance(recovery).Similarly,inCategoryIIIfullrecoveryhasalsobeendenotedbyeither0or1(Table1)andcanResistance:abilitytopersistduring?adisturbance,measureablethrougheventendtowardsinnity(Figure1F).?atingametrictowardsin?,forexample,biodiversityonecosystemresilienceThehighertheresistance,thesmallertodrought[24].:temporalparingresponsetrajectoriesrepresentingdifferentecosystems,differ-changeofecosystemstateorenttypesofdisturbances,,(tx)ordingtoCategoryII(Figure1);systemdcanbemoreorlessresilientthansystemcdependingonthemetricapplied(Figure1).Thisresultsfromthefactthattrajectorydhasahigherrateofbaseline-normalizedrecoverybutalowerrateofimpact-normalizedrecoverythantrajectoryc(Box1).Oursurveyofpreviouslyusedmetricsthereforeindicatesthatthechoiceofthemetric(.,thewayresilienceisquanti?ed)affectsthedegreetowhichdifferenttrajectoriesofecosystemrecoveryfromdisturbancearerepresentedas‘resilient’.Inconsequence,?uenceresilience[8,12].Thisholdssimilarlyforrecentlyadoptedarea-basedmetricsof2TrendsinEcology&Evolution,MonthYear,,?edbythechangeinecosystemstate;thatis,thedifferencebetweenthepre-disturbancestateandthestateatthetimeofpeakimpact(FigureIA).Therecoveryratecanbequanti?(FigureIA):thesmallertheimpactandthehighertherecoveryrate,thesmallertheperturbation(grayareainFigureIA),parethedisturbanceresponsesrelativetothepre-(baselinestate;FigureIB).,inresponsetodisturbancesystembsuffersalargerimpactthansystema(FigureID).However,asthepre-disturbancestateofsystembisalsolarger,thebaseline-normalizedimpactsonthetwosystemsarethesame(FigureIE).Similarly,whiletherecoveryrateforsystemcishigherthanthatforsystemainabsoluteterms(FigureID),itisidenticalwhenconsideringitshigherbaselinestate(FigureIE).Baselinenormalizationofecosystemstatesfollowingdisturbancehasbeenfrequentlyusedtoderivemetricsofresilience(seeTable1inmaintext).paretherecoveryofsystemsindependentofdisturbanceimpact,somestudieshavenormalizedtherecoveryofthesystemstatetotheimpact(FigureICandsee[25_TD$IF]Table1inmaintext),de?ningthesystemstateatpeakimpactas0%andthebaselinestateas100%.Formathematicalreasons(seethesupplementalinformationonline),theimpact-,inourexamplethetwosystemswiththehighest,andidentical,impact-normalizedrecoveryrates(systemsbandc)(FigureIF)exhibitthelowest,andidentical,recoverytimes(FigureID).(A)Absolute(B)Baseline-normalized(C)Impact-normalized%%510001004PerturbaonImpact3Impact6040base50impactbaseS2SRRbase2080(arbitraryunits)1RSystemstateSimpact001000Timet0titxtrt0titxtrt0titxtrRecoveryme(D)(E)%(F)%10c100100cb880bcddbda660abaseimpactS440Sa(arbitraryunits)SystemstateS220000TimeRecoveryme(t).(A)Absolutechangeinsystemstate(S)(R)hetime(t)ittakesforthesystemtorecoveritspre-disturbancestate(recoverytime).Theperturbation(grayarea)causedbythedisturbancecorrespondstothetotalreductioninecosystemstateintegratedovertherecoverytime.(B)Disturbanceresponseofasystemrelativetothepre-disturbancestate(=baseline,attimet0)ofthesystem(baselinenormalization).Boththeimpactandtherecoveryrateareexpressedasapercentageofthebaselinestate.(C)Systemrecoveryfromdisturbanceasapercentageofthedisturbanceimpact(impactnormalization).(D–F)Examplesforfourhypotheticalresponsetrajectoriesa–ddifferingintheirbaselines,impacts,andrecoveryrates,expressedasabsolutechangesinstate(D),asbaseline-normalizedchangesinstate(E),andasimpact-normalizedchangesinstateduringrecovery(F).TrendsinEcology&Evolution,MonthYear,,,ponentofRecoveryTheyAddressNo.[250_TD$IF]CategoryAttributeofrecoveryRefsEquationaReferredtoasNote#1INormalizedsystemstate[28,29,41]PxResilienceConvergesto1forfullrecoveryC0#2INormalizedsystemstate[42]PxResilienceConvergesto1forfullrecoveryCxà#3INormalizedsystemstate[43]CxPxx100%ResilienceEquationcitedincorrectlyin[32,44],Cxconvergesto0forfullrecoveryhià#4INormalizedsystemstate[30]à100CxPxResilienceResilienceattCxx#5IIBaseline-normalizedrecovery[8]PxàPiResilienceIndexvalueforfullrecoverydependsC0Cionresistanceà#6IIBaseline-normalizedrecovery[28]PxPiRelativeresilienceChi0àà#7IIBaseline-normalizedrecovery[30]àà100CiPit100CxPxRelativeresilienceCiCxà#8IIIImpact-normalizedrecovery[45]CxPxPowerofrecoveryLimitis0forfullrecoveryCiàPià#9IIIImpact-normalizedrecovery[46]C0PxResilienceC0àPià#10IIIImpact-normalizedrecovery[24]jPiC0jResilienceFullrecoverytendstoin?nity,PxàC0increasingwithhigherrecovery;symmetricjàj#11IIIImpact-normalizedrecovery[32]2CiPià1ResilienceRestrictedbetweenà1and1,fulleTjCiàPijtjCxàPxjrecovery=1,increasingwithhigherrecovery;symmetricaCandParetheecosystemstateorprocessrateoftheundisturbedcontrolsystemandadisturbedsystem,respectively,atagivenpointintimesincedisturbance(t=x)wherebyt=0andt=icorrespondtothetimeimmediatelybeforeandimmediatelyafterdisturbance,[13,25,26],whichprimarilyquantifyaspectsofperturbation(Box1).Also,recoverytime(.,[80_TD$IF][27]),albeitintuitiveandinformative,mendableasasinglemetricforresilienceasitdoesnotattributetherelativeimportanceofresistanceandrecoveryandmaybeassociatedwithvariablelevelsofperturbation(Figure2Bandrel