Validation of the California Rapid Assessment Method ...

23
1 Validation of the California Rapid Assessment Method Depressional Wetland Module: Final Report Central Coast Wetlands Group at Moss Landing Marine Labs Authors: Cara Clark Sarah Stoner-Duncan Sarah Pearce Kevin O’Connor Kamille Hammerstrom Ross Clark December, 2017

Transcript of Validation of the California Rapid Assessment Method ...

Page 1: Validation of the California Rapid Assessment Method ...

1

ValidationoftheCaliforniaRapidAssessmentMethod

DepressionalWetlandModule:FinalReport

CentralCoastWetlandsGroupatMossLandingMarineLabs

Authors:CaraClark

SarahStoner-DuncanSarahPearce

KevinO’ConnorKamilleHammerstrom

RossClark

December,2017

Page 2: Validation of the California Rapid Assessment Method ...

2

TableofContents

EXECUTIVESUMMARY 5

INTRODUCTION 6

CRAMDEVELOPMENTPROCESS 7

METHODS 8

IDENTIFYTHEGRADIENTOFSTRESS 8SELECTLEVEL3DATA 8IDENTIFYMETRICSFROMLEVEL3DATA 8CONCEPTUALMODELS 9IDENTIFYFIELDSITES 10CONDUCTFIELDASSESSMENTS 11PROCESSLEVEL3DATA 12ANALYZECORRELATIONSBETWEENCRAMANDLEVEL3DATA 13

RESULTS 12

DISCUSSION 19

CONCLUSIONS 21

REFERENCES 22

Page 3: Validation of the California Rapid Assessment Method ...

3

ListofFiguresFigure1.Mapofsitesselectedforsampling..............................................................................................10

Figure2.FieldteamconductingCRAMassessmentandcollectinginvertebratesamplesatMendocinoCollegepond.......................................................................................................................................11

Figure3.SamplingregimeforLevel3protocols........................................................................................11

Figure4.FieldteamsiftinginvertebratesampleatAshCreekinpreparationfordeliverytothelabforanalysis...............................................................................................................................................12

Figure5.HistogramofCRAMIndexscores(n=45)...................................................................................13

Figure6.HistogramsshowingthedistributionofdataineachCRAMAttribute.......................................14

Figure7.HistogramsshowingthedistributionofdatafortheBMIandalgaeIBIs....................................14

Figure8.CorrelationplotofCRAMIndexscorevs.BMIIBI.......................................................................16

Figure9.CorrelationplotofCRAMIndexscorevs.AlgaeIBI.....................................................................16

Figure10.CorrelationplotofCRAMIndexscorevs.SpecificConductance...............................................17

Figure11.CorrelationplotsoftheBMIIBIversusCRAMAttributescores…………………………………………….18

Figure12.CorrelationplotsoftheAlgaeIBIversusCRAMAttributescores………………………………………….19

ListofTablesTable1.CRAMAttributesandmetricswithsummariesofeachmetric......................................................5

Table2.PredictedrelationshipbetweenCRAMandLevel3metrics..........................................................9

Table3.Spearman’srankcorrelations(ρ)andfdr-correctedp-valuesforallCRAMandCRAMmetriccomparisonstoindependentvariables.P-valuessignificantattheα=0.05levelareprintedinboldfont.....................................................................................................................................................15

Page 4: Validation of the California Rapid Assessment Method ...

4

ExecutiveSummaryDepressionalwetlandsareabundantandwidespreadinCalifornia.Stateagenciesandwetlandscientistshavedeterminedthatthereisaneedtocharacterizetheirconditioninordertoassessimpacts,establishprotectionstrategies,andinitiaterestorationefforts.TheWetlandandRiparianAreaMonitoringPlan(WRAMP)wascreatedasaframeworkformonitoringandassessmenttoachievethis.TheCaliforniaRapidAssessmentMethod(CRAM)ispartofthisframeworkandisusedtorapidlycharacterizetheoverallhealthofwetlands.

TheCRAMdevelopmentprocessincludesprescribedstepstocompletion,includingvalidationbyconfirmingcorrelationswithmoreintensiveassessmentmeasures.TheCRAMvalidationprocessfortheDepressionalCRAMwetlandmoduleisdescribedinthisreport.

TheintensivemeasuresofconditionusedtovalidateDepressionalCRAMwereabenthicmacroinvertebrateindexofbioticintegrity(IBI),analgaeIBI,andseveralwaterchemistryparameters.

TheoverallCRAMIndexscoreandindividualCRAMAttributescoresweresignificantlycorrelatedwiththemacroinvertebrateIBI,thealgaeIBI,andseveralofthewaterchemistryparameters.

TheDepressionalCRAMmoduleprovidesameaningful,repeatable,andaccurateassessmentofwetlandconditionacrossthestateofCalifornia.

Page 5: Validation of the California Rapid Assessment Method ...

5

IntroductionDepressionalwetlandsareabundantandwidespreadinCalifornia.Freshwaterwetlandscomprise60%ofallwetlandsinthestate(NRA2010).Manyofthesewetlandsareisolatedgeographicallyandvulnerabletodevelopmentpressure(Brownetal.2016).Stateagenciesandwetlandscientistshavedeterminedthatthereisaneedtocharacterizetheirconditioninordertoassessimpacts,establishprotectionstrategies,andinitiaterestorationefforts.

TheCaliforniaWaterQualityMonitoringCouncil(Council)wasconvenedin2007underamandatefromlegislation,CASenateBill1070,tocoordinateandintegratewaterqualityandrelatedecosystemmonitoring,assessment,andreporting(mywaterquality.ca.gov2017).TheCaliforniaWetlandMonitoringWorkgroup(CWMW)wasestablishedasasub-groupoftheCounciltobuildtoolsforwetlandmonitoring(CWMW2013).TheCWMWoverseestheimplementationoftheWetlandandRiparianAreaMonitoringPlan(WRAMP).TheWRAMPisacoordinatedmonitoringandassessmentstrategythatisstructuredundertheUSEPA’sthreelevelframeworkforaquaticsystemassessment.Theframeworkcategorizeswetlandmonitoringunder:Level1,GISmappingandinventoryofaquaticresources;Level2,field-basedrapidassessmentsofwetlandcondition;andLevel3,moreintensivemeasuresofspecificfunctions,suchaswaterqualityorspeciessampling.

Californiahasthehighestlossofhistoricalwetlandsinthelower48statesat91%reductionoverthe20thcentury(Dahl1990).Thestatehasprioritizedtheassessmentandprotectionofanyremainingwetlands,whicharestillverydiverse.Rapidassessmentofwetlandsallowsforcost-effectiveandrepeatablecharacterizationofwetlandhealthatscalesfromlocaltowatershedtoregionorstatewide.TheCaliforniaRapidAssessment(CRAM)wasdevelopedtosupportthesemonitoringneeds.CRAMprovidesanoverallIndexscore(rangingfrom25to100)thatindicatesthegeneralhealthofawetlandanditscapacitytoperformimportantfunctionsandservices.TheIndexscoreisanaverageoffourmain“Attributes”ofcondition.EachAttributeiscomposedoftwotofivemetricsandsubmetrics(Table1).Theassessmentofeachmetricorsubmetricisbasedonvisualindicatorssurveyedduringafieldvisitoflessthanhalfaday.

Table1.CRAMAttributesandmetricswithsummariesofeachmetric

Attributes Metrics MetricSummaryBufferandLandscapeContext

AquaticAreaAbundance

Measuresextentofwetlandswithin500m

PercentwithBuffer

Percentofareasurroundedbyatleast5mofbufferlandcover

BufferWidth Averageof8bufferwidthmeasurementsupto250mBufferCondition Vegetationquality(nativevs.non-native),degreeofsoil

disturbance,andimpactofhumanvisitationHydrology WaterSource Anthropogenicinfluenceonwatersources(extractionsor

inputs)withinlocalwatershedupto2km

Page 6: Validation of the California Rapid Assessment Method ...

6

Hydroperiod DirectanthropogenicinputsordiversionsHydrologicConnectivity

Accesstoadjacentslopeswithoutlevees,roadgrades,orotherobstructions

PhysicalStructure StructuralPatchRichness

Numberofhabitatstructurespresentfromalistofpotentialpatchtypesfordepressionalwetlands

TopographicComplexity

Complexityofmicro-andmacro-topographicfeatures

BioticStructure NumberofPlantLayers

Numberofplantheightclassesthatcoveratleast5%ofthearea

NumberofCo-dominantSpecies

Totalnumberoflivingplantspeciesthatcompriseatleast10%ofanyplantlayer

PercentInvasiveSpecies

Thepercentofthetotalnumberofco-dominantspeciesthatarelistedbyCal-IPCasinvasive

HorizontalInterspersion

Thecomplexityofplantzones(speciesassemblagesormono-specificstands)

VerticalBioticStructure

Twomethodoptions:1)wetlandswithwoodyvegetationscoreoverlapoflayers;2)wetlandswithemergentmarshplainsscoreentrainmentandvegetationcanopy

CRAMDevelopmentProcessTherearesixstepstoCRAMdevelopment,asdescribedinSutulaetal.(2006)andoutlinedontheCRAMwebsite(http://www.cramwetlands.org/about).Thesestepsinclude:

1. Definitionphase2. Basicdesignphase3. Verificationphase4. Validationphase5. Moduleproductionphase6. Ambientsurveyphase

Previouswork,fundedbytheUSFWS(Agreement#M11AF00103),accomplishedphasesonethroughthree.Verificationwascompletedin2012and2013withastatewidesurveyacrossagradientofhydroperiod(ClarkandO’Connor2014).Thisstudywasasolidfoundationtolaunchthecurrentproject,asitverifiedthatthedepressionalmodulewaseffectivelydifferentiatingbetween“good”,“fair”,and“poor”sitesascategorizedaprioribythedevelopmentteam.ThefieldbookforDepressionalCRAMwasrevisedtoimproveitsperformanceandutility.

ThisprojectaimedtocompletethevalidationphaseofDepressionalCRAMdevelopment.Validationhasbeendefinedas“theprocessofdocumentingrelationshipsbetweenCRAMresultsandindependentmeasuresofconditioninordertoestablishCRAM’sdefensibilityasameaningfulandrepeatablemeasureofwetlandcondition”(Steinetal.2009).ValidationoftheDepressionalmodulewillestablishitsscientificcredibilityandfurtheritsuseinlocal,state,andfederalprograms.

Page 7: Validation of the California Rapid Assessment Method ...

7

MethodsValidationoftheDepressionalCRAMmodulefollowedthesystematicprocessdescribedbySteinetal.(2009),whichprescribedseveralstepstovalidation:

1. Identifythegradientofstress2. IdentifyappropriateLevel3datatovalidatetheCRAMmodule3. IdentifymetricsthatwillbecalculatedfromthedetailedLevel3data4. Createconceptualmodelsoftheexpectedrelationshipbetweenthedetaileddataand

CRAMscores5. IdentifyfieldsiteswhereLevel3dataareavailableorpossibletocollect6. ConductCRAMassessmentsatthesitesidentified7. AnalyzecorrelationsbetweenCRAMscoresandLevel3metrics

IdentifytheGradientofStressDepressionalwetlandscanbeimpactedbysurroundinglanduse.Landscapeconditionscanbeaneffectivepredictorofwetlandhealth(Rothetal.1996,MicacchionandGara2008).Adjacentandupstreamlandcoveraffectswetlandsthroughmanyprocesses,includingpollutedrunoff,habitatloss,andalterationofhydrologicdynamics.Whendepressionalwetlandsaresurroundedbynaturalopenspacetheyaremuchmorelikelytosupportflora,fauna,andimportantwetlandfunctions.Conversely,whentheyareclosetodevelopedareassuchasurbanoragriculturallandcovers,theyaremorelikelytohavereducedfunctionanddiversity.Thisstudyselectedarangeofsitesalongagradientofdevelopmentpressure,includingsomesitesinopenspacepreservesorparks,andothersincitiesandagriculturalareas.

SelectLevel3DataThisprojectbenefitedfrompreviousworkondepressionalwetlands.LundeandResh(2012)developedamacroinvertebrateindexofbioticintegrity(IBI)fordepressionalwetlandsintheSanFranciscoBayArea.Brownetal.(2016)expandedthisworkinsouthernCaliforniatoincorporatemultipleindicatorsofwetlandhealth,includingmacroinvertebrates,algae,waterquality,andhabitatassessment.OurcolleaguesattheSouthernCaliforniaCoastalWaterResearchProject(SCCWRP)andtheBayAreaRegionalWaterQualityControlBoard(RB2)graciouslysharedtheirdatawithus.TheLevel3toolswereinplacebuthadonlybeenusedinparticularregionsofthestate.Wewereabletousethestandardprotocolsdevelopedforsamplingmacroinvertebrates,algae,andwaterquality(Fetscheratal.2015)toexpandthegeographicrangeofthevalidationsites.

IdentifyMetricsfromLevel3DataThemacroinvertebrateIBIisacombinationofseveralmetricsbasedontheassemblageofspeciesfoundineachsample.Theseincludescraperrichness(afunctionalgroup);percentEphemeroptera,Odonata,andTrichoptera(EOT);EOTrichness;Oligochaetarichness;andseveralothers.ThefinalIBIscoreisonascaleof1-100andresultsfromcombiningalloftheselectedmetrics.WeusedtheoverallIBIscoreasthecomparisonmetric.

Page 8: Validation of the California Rapid Assessment Method ...

8

ThediatomalgaeassemblagesarescoredusingtheD18index,whichcombinesseveralmetricsincluding:proportionsedimenttolerant;proportionlowPindicators;proportionNheterotrophs,proportionrequiring>50%DOsaturation;andproportionhalobiontic(Fetscheretal.2014).

Standardwaterchemistryandwaterqualityparameterswereselectedforanalysisaswell.Theseincludedturbidity,temperature,pH,specificconductance,salinity,dissolvedoxygen(DO),andalkalinity.

ConceptualModelsTheexpectedrelationshipbetweenCRAMIndexandAttributescoresandLevel3datawerepredictedaprioriforeachLevel3indicator.BoththemacroinvertebrateanddiatomalgaeIBIsarescoredfromlowtohigh,withahigherscoreindicatingbetterqualitywetlands.SiteswithhigherIBIscoresprovidemoreintactfunctionsandhavefewerstressors.Therefore,apositiverelationshipisexpectedbetweenCRAMIndexandAttributescoresandinvertebrateandalgalIBIs.

ThewaterqualityparametersaregenerallymorevariableanditwasdifficulttopredictrelationshipswithconditionindicessuchasCRAM.Waterchemistrycanbeaffectedbymanyfactors,whichmaynotnecessarilyberelatedtothelocalconditionoralocaldisturbance.Waterchemistryintegratesalloftheupstreamlanduses,andmaybeaffectedmorebyimpactsupstreamthannearbyconditions.However,somewaterqualityindicatorsmayberelatedtotheconditionoftheadjacentaquaticsystem.Specificconductanceisageneralmeasureofwaterqualityandcanbeincreasedbyrisinglevelsofsaltsorotherinorganiccompounds,oftencausedbyhumandisturbance(EPA2017).Therefore,asspecificconductance(salinity)increases,CRAMscoresareexpectedtodecrease.Humanactionssuchaswaterextractionforirrigationoradditionsofwastewaterrunofftowetlandscanalsoaffectconductance.Increasesinsalinitycanbestressfulformanyaquaticbiota(EPA2017).WepredictedthathighersalinitylevelswouldcorrelatewithlowerCRAMscores.OtherwaterchemistrymeasureswerenotpredictedtocorrelatewithCRAMscores.

Table2.PredictedrelationshipsbetweenCRAMandLevel3metrics

CRAMIndexScore

BufferandLandscape

Hydrology Physical Biotic

BMIIBI + + + + +AlgaeIBI + + + + +Turbidity - - - Temperature - - - -pH + SpecificConductance - Salinity - DissolvedOxygen(mg/L)

Page 9: Validation of the California Rapid Assessment Method ...

9

DissolvedOxygen(%) Alkalinity

IdentifyFieldsitesTheprojectbenefitedfromdatapreviouslycollectedinSouthernCaliforniaandtheSanFranciscoBayArea.However,forfullvalidationassessmentsiteswereneededacrossthestate.Weselectedsitesinseveralregions,includingtheSierraNevada,theCentralValley,theModocregion,andtheNorthCoast.Siteswereselectedtorepresentthegradientofhumandisturbanceandstress.Siteaccesspermissionwasalsoaconsiderationinsiteselection.Mostsiteswereonpublicland,whileseveralwerelocatedwithinprivatepreservesoronschoolcampuses.Weincludedasub-setofsitesfromthepreviouslysampledareasinSouthernCaliforniaandtheBayArea.Tominimizeregionalbias,only15siteswereselectedfromeachoftheseareasforsynthesiswith15sitessampledinotherregionsofthestate.Atotalof45siteswereincludedintheanalysis(Figure1).

Page 10: Validation of the California Rapid Assessment Method ...

10

Figure1.Mapofsitesselectedforsampling

Page 11: Validation of the California Rapid Assessment Method ...

11

ConductFieldAssessmentsFieldassessmentswereconductedusingtheDepressionalWetlandCRAMmodule(version6.1)atfifteennewsitesforthisprojectduringspringandsummer2014(Figure2).CRAMhadpreviouslybeenconductedatthesitesinSouthernCaliforniaandtheBayAreausingthesameversionoftheprotocol.AllassessmentsfollowedthequalityassuranceproceduresoutlinedintheCRAMQAPlan(CWMW2016)andtheQAPPforthisproject(CCWG2014).

Ateachsite,thesamesuiteofLevel3indicatorswascollected.Thisincludedsamplingmacroinvertebrates,algae,

andwaterchemistryaccordingtostandardprotocols(Fetscheretal.2015).Samplingwasdistributedaroundeachwetlandalongten“nodes”whereindividualsampleswerecollected(Figure3).Thenodeswereevenlyspacedaroundthewetlandbymeasuringthecircumferenceofthepondanddividingthatlengthintotensegments.Eachofthesamplingnodeshasthreeparalleltransects:oneforwaterchemistry,oneformacroinvertebratesampling,andoneforalgaesampling.

Figure2.FieldteamconductingCRAMassessmentandcollectinginvertebratesamplesatMendocinoCollegepond

Figure3.SamplingregimeforLevel3protocols

Page 12: Validation of the California Rapid Assessment Method ...

12

ProcessLevel3dataMacroinvertebratesamples(Figure4)wereprocessedbytheAquaticBioassessmentLaboratoryandusedthestandardtaxonomiceffortnamingconventionoftheSouthwestAssociationofFreshwaterInvertebrateTaxonomists.AlgaesampleswereprocessedbyEcoAnalysts,Inc.accordingtoSurfaceWaterAmbientMonitoringProgram(SWAMP)standards.

MacroinvertebrateIBIscoreswerecalculatedaccordingtoLundeandResh(2012).TherawmacroinvertebratedatawererunthroughanRscripttocalculatetheIBIwithgracioushelpfromJeffBrownattheSouthernCaliforniaCoastalWaterResearchProject(SCCWRP).ThediatomalgaeIBIwascalculatedusingSCCWRP’sonlinetool(SCCWRP2016)withtroubleshootingassistancefromBettyFetscher.ThediatomalgaeIBIwasdevelopedforSouthernCaliforniastreams(Fetscheretal.2014)buthasbeenusedindepressionalwetlandsaswell(Brownetal.2016).

AnalyzeCorrelationsBetweenCRAMandLevel3DataSpearmanrankcorrelationswereconductedfortheoverallCRAMIndexscoreandeachoftheCRAMAttributesagainsttheBMIIBI,algaeIBI,andallwaterqualityparameters.Percentdissolvedoxygenwasnotusedduetoerrorsinthedata.Thenon-parametricSpearmanrankcorrelationwasusedbecauseitdoesnotrequireanassumptionofbivariatenormality(Dodge2010).ToaccountforthelargenumberofcorrelationsandtocontrolforType-Ierror,p-valueswerecorrectedusingthefalsediscoveryrate(fdr,BenjaminiandHochberg1995).EachmetricwithinCRAMwastreatedasindependentofeachother,socorrectedp-valueswerecalculatedusingfdrseparatelyforeachCRAMmetricandindependentmeasures(BMIIBI,IBID18,turbidity,watertemperature,etc.).AllcalculationswereconductedusingSAS9.3software(SASInstituteInc.2011).

ResultsAneffectiverapidassessmentmethodmustberesponsivetoarangeofconditionsandbesensitivetohumandisturbance(Sutulaetal.2006,Steinetal.2009).TheCRAMIndexscoreisacompositeofthefourAttributescoresandrepresentstheoverallecologicalconditionofthewetland.TheCRAMtoolgeneratesaminimumvalueof25andamaximumvalueof100.TheCRAMIndexscorescollectedforthisprojectrangedfrom31to96,withamedianscoreof60(Figure5).Wedeterminedthatthescoresarenotbiasedtowardshighorlowvalues(skewness

Figure4.FieldteamsiftinginvertebratesampleatAshCreekinpreparationfordeliverytothelabforanalysis

Page 13: Validation of the California Rapid Assessment Method ...

13

=-0.04).ThebroadrangeofscoresconfirmstheresponsivenessofthedepressionalCRAMmodule.

CRAMIndexScore

Figure5.HistogramofCRAMIndexscores(n=45)

AnextensiverangeofscoresweremeasuredforeachCRAMAttribute:BufferandLandscapeContext25-93,Hydrology33-100,PhysicalStructure25-100,andBioticStructure25-97(Figure6).WedeterminedthateachAttributeisresponsivetovaryingconditionsinandaroundthewetlandofinterest.

Page 14: Validation of the California Rapid Assessment Method ...

14

CRAMAttributeScore

Figure6.HistogramsshowingthedistributionofdataineachCRAMAttribute

IBIScore

Figure7.HistogramsshowingthedistributionofdatafortheBMIandalgaeIBIs

Page 15: Validation of the California Rapid Assessment Method ...

15

TheLevel3indicatorswerefoundtohaveasimilarwiderangeofIBIscoresasfoundwithCRAM(Figure7).

TheoverallCRAMIndexscoreandeachAttributescoreweretestedforsignificantcorrelationswithLevel3data,includingtheBMIIBI,algaeIBI,andwaterchemistryparameters.

Table2liststheresultsofallanalyseswithsignificantcorrelations’p-valuesshowninboldfont(significantwhencomparedtoα=0.05).Interestingly,whenawaterqualityparameterwassignificantlycorrelated,ittendedtobesignificantacross3CRAMAttributes:BufferandLandscapeContext,Hydrology,andBioticStructure.BoththeBMIIBIandtheIBID18correlatedsignificantlywiththeCRAMIndexscoreaswellasmostAttributes.Waterqualityparametersofturbidity,pH,anddissolvedoxygendidnotsignificantlycorrelatewithanyAttributes,withtheexceptionofpH,whichcorrelatedwiththeBioticStructure.CRAMPhysicalStructuredidnotcorrelatesignificantlywithanyparameters,althoughitwasmarginallysignificantlycorrelatedwiththeAlgaeIBI(p=0.1).

Table3.Spearman’srankcorrelations(ρ)andfdr-correctedp-valuesforallCRAMandCRAMmetriccomparisonstoindependentvariables.P-valuessignificantattheα=0.05levelareprintedinboldfont.

OverallCRAM

CRAMBuffer&LandscapeContext

CRAMHydrology

CRAMBioticStructure

CRAMPhysicalStructure

BMIIBI ρ=0.42p=0.01

ρ=0.34p=0.03

ρ=0.47p=0.005

ρ=0.29p=0.07

ρ=0.11p=0.50

AlgaeIBI ρ=0.49p=0.001

ρ=0.36p=0.02

ρ=0.32p=0.04

ρ=0.50p=0.001

ρ=0.24p=0.10

Turbidity(NTU) ρ=-0.19p=0.63

ρ=-0.04p=0.83

ρ=-0.05p=0.83

ρ=-0.30p=0.40

ρ=-0.15p=0.63

WaterTemp(oC)

ρ=0.23p=0.52

ρ=0.05p=0.88

ρ=0.13p=0.85

ρ=0.03p=0.88

ρ=0.37p=0.20

pH ρ=-0.28p=0.10

ρ=-0.20p=0.10

ρ=-0.21p=0.24

ρ=-0.42p=0.01

ρ=-0.16p=0.30

SpecificConductance(μS/cm)

ρ=-0.38p=0.02

ρ=-0.36p=0.02

ρ=-0.33p=0.04

ρ=-0.39p=0.02

ρ=-0.09p=0.56

DO(mg/L) ρ=-0.07p=0.85

ρ=-0.13p=0.85

ρ=-0.06p=0.85

ρ=-0.13p=0.85

ρ=0.01p=0.96

Alkalinity(CaCO3average)

ρ=-0.32p=0.05

ρ=-0.34p=0.05

ρ=-0.34p=0.05

ρ=-0.25p=0.12

ρ=-0.07p=0.65

ThesignificantcorrelationsbetweentheCRAMIndexscoreandthelevel3indicatorsareofparticularinterest.TheCRAMIndexscorewassignificantlycorrelatedwithbothoftheselected

Page 16: Validation of the California Rapid Assessment Method ...

16

independentassessmentprotocols(BMIIBIandtheAlgaeIBI)aswellasexpectedwaterqualityparameters(specificconductance)(Figures8-11).

Figure9.CorrelationplotofCRAMIndexscorevs.AlgaeIBI

Figure8.CorrelationplotofCRAMIndexscorevs.BMIIBI

Page 17: Validation of the California Rapid Assessment Method ...

17

Figure10.CorrelationplotofCRAMIndexscorevs.SpecificConductance

TheindividualCRAMAttributeswerealsotestedforcorrelationwiththeIBIs.TheBMIIBIwassignificantlycorrelatedwiththeBufferandLandscapeContextAttributeandtheHydrologyAttribute(Figure12).ItwasnotsignificantlycorrelatedwithPhysicalStructureorBioticStructure.TheAlgaeIBIwassignificantlycorrelatedwiththeBufferandLandscapeContext,Hydrology,andBioticStructureAttributes,butnotthePhysicalStructureAttribute(Figure13).

Page 18: Validation of the California Rapid Assessment Method ...

18

CRAMAttributeScoreFigure11.CorrelationplotsoftheBMIIBIversusCRAMAttributescores

Page 19: Validation of the California Rapid Assessment Method ...

19

CRAMAttributeScoreFigure12.CorrelationplotsoftheAlgaeIBIversusCRAMAttributescores

DiscussionThegoalofthisprojectwastovalidatetheCRAMmoduleforDepressionalwetlands.ToensurethattheCRAMmethodmeetsestablishedCRAMdevelopmentguidelines(Steinetal.2009),theCRAMValidationteamsetouttoconfirmthataCRAMmodulefordepressionalsystemsmetasetofkeycriteria.AvalidatedCRAMmoduleshouldgeneratescoreswhichappropriately

Page 20: Validation of the California Rapid Assessment Method ...

20

representafullrangeofwetlandconditionsfoundwithinthestate.Thetoolshouldalsoberepeatableandcorrelatewithothertrophicorfunctionspecificindicatorsofcondition.

ThesiteselectionprocessensuredthatsampledwetlandsrepresentedthefullrangeofclimaticandecologicalconditionfoundinCalifornia.BypartneringwithwetlandscientiststhroughoutthestatewithextensiveexperienceinCaliforniadepressionalwetlands,wehavedevelopedatoolthatcanbeusedsuccessfullybymostCaliforniawetlandpractitioners.WecreatedaconceptualmodelfromwhichwepredictedandtestedrelationshipsbetweenCRAMscoresandvariousLevel3indicatorsofcondition.

AtleastoneLevel-3metriccorrelatedsignificantly,andintheexpecteddirection,foreachdepressionalAttribute,withtheexceptionofPhysicalHabitat,althoughitsrelationshiptoAlgaeIBIscoreswasnearlysignificant(ρ=0.24,p=0.10).OuranalysisfoundthatsiteCRAMIndexscoresweresignificantlycorrelatedwiththebenthicinvertebrateindex(BMIIBI),thealgaeindex(AlgaeIBI),andseveralwaterqualityparameters(specificconductanceandalkalinity).TheCRAMAttributesBufferandLandscapeContextandHydrologyweresignificantlycorrelatedwiththesameLevel3parameters.Sincethealgaeandmacroinvertebrateindexesweredevelopedtorespondtodisturbanceandstressinthewetland,CRAMattributesthatevaluatesimilarfunctionsandareasofconditionshouldreflectasimilargradientofimpacts.BoththeoverallIndexscoreandseveralpredictedCRAMAttributescoreswerecorrelatedwiththeIBIscores.TheBufferandLandscapeContextAttributemeasuresanthropogenicimpactsfromsurroundinglanduse.Similarly,themacroinvertebratesandalgaeindexesaresensitivetothosesameimpacts.TheHydrologyAttributeevaluatesthesourcesofwaterandpotentialcontamination,theartificialmanipulationofwaterflow,andtheconnectiontoadjacenttransitionalhabitat.Thesefactorssimilarlyaffectthecompositionofthealgaeandmacroinvertebratecommunities.

ThewaterqualityparametersthatweremoststronglycorrelatedwithCRAMscoreswerespecificconductanceandalkalinity.Inaddition,pHwascorrelatedwiththeBioticStructureAttribute.Intheabsenceofmarinesources,higherlevelsofconductivityoftenindicateanincreaseinhumandisturbance,whichresultsinrunoffhighinsaltsanddissolvedsolids(EPA2017).ThesignificantcorrelationbetweenCRAMandtheseparametersshowsthatCRAMisabletodetectadjacentlanduseswhichleadtoimpairedwaterquality.

InvalidatingthisCRAMmodule,thegoalwastohavebroadcorrelationwithmultipleL3metricsthatrepresentarangeofecologicalfunctionsandservices.However,wedidnotexpectthosecorrelationstobetight,withhighSpearmans’Rhovalues,asthiswouldnegatetheneedfordevelopinganewmethodofassessment.CRAMismeanttomeasuremultiplepotentialwetlandfunctions,notanysinglefunction,asrepresentedbytheL3data.

Anumberofthewaterchemistrymeasurementsdidn’tcorrelatewithCRAM,includingturbidity,temperature,anddissolvedoxygen,whilepHonlycorrelatedwithoneCRAMAttribute(BioticStructure).WeexpectedanegativecorrelationbetweenCRAMAttributesand

Page 21: Validation of the California Rapid Assessment Method ...

21

turbidityandtemperaturemeasurements,becausebothindicateuplanddisturbances.TurbidityandtemperaturewereexpectedtocorrelatenegativelywithCRAMIndexscores,asareductioninripariancovercanleadtohighertemperatures,increasederosion(turbidity)andlowerCRAMIndexscores.SimilartofindingsbyCRAMvalidationeffortsforothermodules,otherfactorsappeartobeaffectingthequalityofwaterflowingintothesampledwetlands,suchasspecificupstreamlanduses,localclimateandgeography(Sutulaetal.2006).

MostofthesiteswithhighwatertemperatureswereintheCentralValley,buthadarangeofCRAMIndexscoresreflectingsiteconditions.Otherwaterqualityparametersincludingdissolvedoxygenalsoseemstobeinfluencedbyfactorsthatdon’trelatetotheoverallconditionofthewetland.Dissolvedoxygencanvarywidelyatasinglewetlandinresponsetodynamicfactorssuchasdailyrespirationfluctuations,soitmaynotbethebestindicatorofwetlandcondition.ThepHofasystemismoreinfluencedbylocalgeologyandrainfallthanoverallwetlandhealth.

ConclusionsThisworkwaspresentedtotheLevel2/RapidAssessmentCommitteeoftheCWMWinJuly,2017,andtheiradvicecontributedtofurtheranalyses.TheLevel2/RapidAssessmentCommitteeapprovedthevalidationoftheDepressionalCRAMmoduleattheOctober,2017meeting.

TheDepressionalCRAMmoduleisnowvalidatedandmeetsthegoalsdefinedbytheLevel2Committee.OuranalysisshowsthatthereisasignificantcorrelationbetweenCRAMIndexandAttributescoresandLevel3intensivemeasuresofconditionandfunction.Therefore,weconcludethattheDepressionalCRAMmoduleprovidesameaningful,repeatable,andaccurateassessmentofwetlandconditionacrossthestateofCalifornia.

Page 22: Validation of the California Rapid Assessment Method ...

22

ReferencesBenjaminiY,YHochberg.1995.Controllingthefalsediscoveryrate:apracticalandpowerfulapproachtomultipletesting.JRStatSocB57:289-300.

Brown,J.S.,E.D.Stein,C.Solek,A.E.Fetscher.2016.AssessmentoftheConditionofSouthernCaliforniaDepressionalWetlands.CostaMesa,CA.SCCWRPTechnicalReport921.

Clark,C.andK.O’Connor.2014.VerificationoftheDepressionalCRAMWetlandModule,CIAPTask3SummaryReport.MossLanding,CA.https://ccwg.mlml.calstate.edu/sites/default/files/documents/finalreport-depressionalCRAM.pdf(accessed11/20/17).

[CCWG]CentralCoastWetlandsGroup.2014.QualityAssuranceProjectPlanforSupportforL2CommitteePriorityToolDevelopment:ValidationofThreeCRAMModulesTask3:ValidationoftheDepressionalWetlandCRAMModule.FundingNumber:CD-99T05801-0

[CWMW]CaliforniaWetlandsMonitoringWorkgroup.2016.DataQualityAssurancePlan:CaliforniaRapidAssessmentforWetlands.Sacramento,CA.http://www.cramwetlands.org/sites/default/files/CRAM%20data%20QA%20plan%20v7-2016.9.19.pdf(accessed11/30/2017)

[CWMW]CaliforniaWetlandMonitoringWorkgroup.2013.CaliforniaRapidAssessmentMethod(CRAM)forWetlandsUser’sManual,Version6.1pp.67.http://www.cramwetlands.org/sites/default/files/2013-04-22_CRAM_manual_6.1%20all.pdf

Dahl,T.(1990)WetlandslossesintheUnitedStates1780’sto1980's.U.S.DepartmentoftheInterior,FishandWildlifeService,Washington,D.C.13pp.

Dodge,Y.2010.TheConciseEncyclopediaofStatistics.NewYork(NY):Springer.pp502-505.

[EPA]EnvironmentalProtectionAgency.2017.NationalAquaticResourceSurveys.https://www.epa.gov/national-aquatic-resource-surveys.(accessed11/14/17).

Lunde,K.B.,V.H.Resh.2012.Developmentandvalidationofamacroinvertebrateindexofbioticintegrity(IBI)forassessingurbanimpactstoNorthernCaliforniafreshwaterwetlands.EnvironmentalMonitoringandAssessment,v.184,pp3653-3674.

Mywaterquality.ca.gov.2017.CaliforniaWaterQualityMonitoringCouncil.http://www.mywaterquality.ca.gov/monitoring_council/index2.html.(accessed11/20/17).

[NRA]NaturalResourcesAgency.2010.StateoftheState’sWetlands:TenYearsofChallengesandProgress.Sacramento,CA.http://www.resources.ca.gov/docs/SOSW_report_with_cover_memo_10182010.pdf(accessed11/20/17).

Page 23: Validation of the California Rapid Assessment Method ...

23

SASInstituteInc.2011.BaseSAS®9.3ProceduresGuide.Cary,NC:SASInstituteInc.

[SCCWRP]SouthernCaliforniaCoastalWaterResearchProject.2016.AlgaeIBIAssessmentTool.http://www.sccwrp.org/Data/DataTools/algaeIBI.aspx.(accessed2/24/15).

Sutula,M.A.,E.D.Stein,J.N.Collins,A.E.FetscherandR.Clark.2006.Apracticalguideforthedevelopmentofawetlandassessmentmethod:TheCaliforniaexperience.JournaloftheAmericanWaterResourcesAssociation42:157-175.

SteinE.D.,A.E.Fetscher,R.P.Clark,A.Wiskind,J.L.Grenier,M.Sutula,J.N.Collins,C.Grosso.2009.Validationofawetlandrapidassessmentmethod:UseofEPA’slevel1-2-3frameworkformethodtestingandrefinement.Wetlands29:648-665.