Assessing the feasibility of freshwater mussel restoration in ...

Virginia Commonwealth University Virginia Commonwealth University

VCU Scholars Compass VCU Scholars Compass

Theses and Dissertations Graduate School

2020

Assessing the feasibility of freshwater mussel restoration in Assessing the feasibility of freshwater mussel restoration in

urban streams. urban streams.

John-Reid Ryan

Follow this and additional works at: https://scholarscompass.vcu.edu/etd

Part of the Terrestrial and Aquatic Ecology Commons

© The Author

Downloaded from Downloaded from https://scholarscompass.vcu.edu/etd/6405

This Thesis is brought to you for free and open access by the Graduate School at VCU Scholars Compass. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of VCU Scholars Compass. For more information, please contact [email protected].

P a g e |1

©John-ReidRyan2020AllRights

Reserved

P a g e |2

Assessingthefeasibilityoffreshwatermusselrestorationinurbanstreams.

AthesissubmittedinpartialfulfillmentoftherequirementsfortheDegreeofMasterofScienceinEnvironmentalStudiesatVirginiaCommonwealthUniversity

ByJohn-ReidRyan,B.S.EnvironmentalStudiesVirginiaCommonwealthUniversityCenterforEnvironmentalStudies,VirginiaCommonwealthUniversity,

[email protected]

ThesisAdvisor:PaulBukaveckasPh.D.Professor,VCUCenterforEnvironmentalStudies

VirginiaCommonwealthUniversityRichmond,Virginia

July,2020

P a g e |3

Acknowledgements

Iwouldliketothank my advisorPaulBukaveckasforhisextensivepatienceandadvicethat

movedtheprojectintherightdirection,aswellasfundingthisresearch. RachelMair,Amy

Maynard,BryceMaynard,BenDavisfromHarrisonLakeFishHatcheryforsupplyingAlewife

Floaterandadditionalguidance. TheRiceRiversResearchGrantorganizationforprovidedme

withfundingandsupportduring2019withIamverythankfulfor.MacLee,SpencerTassone,

andRachelHendersonfromtheVCUAnalyticalLabgreatlycontributedtothisprojectby

processingwaterqualitysamplesanddata. Additionally,JoeWoodandotherCBFstaffassisted

intheinitialliteraturereviewandmusselconferencethatgreatlyhelpedmegetthisproject

focusedandmovingearly.Dr.Vonesh,AlbertByers,ElizabethEdmondson,SuzanneKirk,Todd

Janeski,CaseyJohnsonandtheirinvolvementonNOAABivalveprojectatVCUhelpedmerefine

mypresentationofthisresearch.Also,FamilyandFriendswhoIdraggedouttohelpmewith

samplingsitesandGrandparentsforloaningmemetalstakestosecurecages,Icouldnothave

donethiswithoutyou.

P a g e |4

TableofContents

Abstract……………………………………………………………………………………………….5

Introduction…………………………………………………………………………………………6-9

Methods………………………………………………………………………………………………9-12

Results…………………………………………………………………………………………………12-15

Discussion……………………………………………………………………………………………15-17

Conclusion…………………………………………………………………………………………..17-18

References…………………………………………………………………………………………..19-22

Tables…………………………………………………………………………………………………..23-25

Figures………………………………………………………………………………………………….26-33

P a g e |5

Abstract

Themainobjectiveofthisstudywastodeterminewhetherintroducedfreshwater

mussels(Alewifefloater,Utterbackianaimplicata)cansurviveandgrowinurbanstreamsinthe

JamesRiverwatershed.AsecondaryobjectivewastoassessdifferencesinU.implicatasurvival

andgrowthofinthecontextofdifferingwaterqualityandfoodresourceconditionsamong

threeurbansitesandthreeruralsites.Resultsfromthisstudyshowlargedifferencesingrowth

andsurvivorshipofmusselsacrosssites.Highersurvivorshipwasobservedamongmussels

stockedintoruralstreams(35%and44%)incomparisontourbanstreams(3%,6%and14%).

Highmortalityinurbanstreamswaslargelyduetowashoutandburialofmussels.These

findingssuggestthatthe“flashy”hydrologytypicallyassociatedwithurbanstreamsisa

significantimpedimenttosuccessfulintroductionatthesesites.Highgrowthrateswere

observedinoneoftheruralstreams(HerringCreek:57mg/d),whereasgrowthrateswereless

than15mg/datallothersites.Foodresourcemetricsshowedstatisticallysignificant

differencesamongsiteswithhighervaluesofTSS,particledensity,organicmattercontentand

chlorophyll-acontentatruralsitesrelativetourbansites.Thesefindingssuggestthatrural

siteshadmorefavorablefoodresourcesthanruralstreams,thoughwedidnotfindthatfood

metricswereasignificantpredictorofvariationingrowthratesamongsites.Wedidnotfind

thatwaterqualitymetrics(temperature,dissolvedoxygen)wereasignificantpredictorof

variationinmusselgrowthrates.Overall,thesefindingssuggestthathydrologicconditionsin

urbanstreamsposeasignificantchallengetothesuccessfulreintroductiononnativemussels.

P a g e |6

Introduction

NorthAmericahasthelargestfreshwatermusseldiversityintheworld(Williamsetal.,

1993).Inadditiontobeingadiverseanduniquefauna,musselsprovideimportantecosystem

servicesbyremovingparticulatematterfromthewatercolumn(Vaughn,2018).Nutrients

containedinparticulatemattermaybestoredinmusselshellsandtissues,orexcretedas‘bio

deposits’(Vaughn,2018).Thetransferofnutrientsfromsuspendedparticulatematterto

benthicdepositsmakesmusselsanimportantlinkwithinnitrogenandphosphorouscyclesasit

increasesthelikelihoodthatnutrientswillbesequestered(throughin-streamburial)orlostvia

denitrification(Hoelleinetal.,2017).Priorworkonurbanriversshowedthatfreshwater

musselsstimulatemicrobialactivityanddenitrificationthroughwasteproduction.Hoelleinet

al.(2017)showedthatnitrogenuptakeanddenitrificationratesinsedimentalonewerearound

2%,whileratesinsedimentswithmusselswere8-12%.Ecosystemservicesprovidedby

freshwatermusselsmaybebeneficialtomitigatinganthropogenicpollutionofChesapeakeBay.

Humanactivitiessuchasagricultureandurbandevelopmentincreasesedimentandnutrient

transportviatributarystreamstransportingsedimenttotheBay(Eshleman&Sabo,2016).

Thesenon-pointsources,coupledwithpointsourceinputssuchaswastewatertreatment

plantshavedegradedwaterqualitywithintheBayanditstributaries(Eshleman&Sabo,2016).

ReducingsedimentandnutrientloadstotheBaytoimprovewaterclaritywilldependon

implementationofbestmanagementpracticesinuplandareasandtributarystreams

(McConnell,2017).

MusselpopulationshavedeclinedinmanywatershedsthroughouttheUnitedStates

overthepast50years(Williamsetal.,1993).TheNatureConservancyestimatesthat55%of

musselspeciesinNorthAmericahaveprogressedtoextinctionorimperiledstatus(Williamset

al.,1993).Declinesinmusselpopulationsarepartlyduetohabitatdegradationassociatedwith

landusechange(urbanizationandagriculture).Urbanstreamsareoftendevoidofmussel

populations,andinunrestoredurbanstreamstheirbiodiversitycanbe47%lessthanreference

streams(Smucker&Detenbeck,2014).Recentadvancesintheabilitytopropagatemussels

providesanopportunitytorestorepopulations,howeverlittleisknownregardingin-stream

conditionsthatwouldinfluencethesuccessofrestoration(e.g.foodandwaterquality

P a g e |7

conditions).Generally,musselsrequirewell-oxygenatedflowingwaterwithsuitablesubstrate

conditions(amixtureofsand,gravelandsilt;NRCS,2007).Foodquantityandqualityislikelyto

bedependentontheabundanceofsuspendedparticulatematteranditscomposition,

includingcontributionsfromsuspendedalgae(JeagerandCherry,1994).Thepresenceof

impoundments,suchasstormwaterretentionponds,mayincreasefoodquantityandquality

astheytrapinorganicparticulatematter,andmaypromotegrowthofphytoplankton(Winter

andDuthie1998).Furtherresearchisneededtodocumentthesuccessofmusselintroduction

effortsindiversestreamhabitatconditionstobetterinformmanagementefforts.

Conditionsfoundinurbanstreamsmaypresentspecialchallengestomusselrestoration

(Walshetal.,2016).Theseinclude“flashy”streamflowconditionsduetorapidrunofffrom

impervioussurfaces(Nedeauetal.,2003;Walshetal.,2016).Highdischargeeventsalterstream

morphologyduetoincreasedbankerosionwhichcausesunstablesedimentsandburialof

mussels(Walshetal.,2016).Urbanstreamsarealsosubjecttoscouringeventswherethe

streambottomisremoved(Walshetal.,2016).Thisremovessubstratesincludingleaflitterand

organicdepositsthatmakeupmusselhabitat(Walshetal.,2016).Theabsenceofmusselsin

urbanstreamsmayalsobefromyearsofpoorwaterqualityconditions(lowdissolvedoxygen,

toxicpollutants)thatwereprevalentpriortopassageoftheCleanWaterAct.

TheChesapeakeBayhassufferedfromeutrophicationandsedimentpollutionfora

numberofdecades.Inordertoproperlymanagethisissue,thereisaneedforrestoration

practicesthatreducesedimentandnutrientloads.Thisisaccomplishedbyimplementingbest

managementpractices(BMPs).Improvementsinstreamconditionhavebeenbroughtaboutby

avarietyofmanagementpracticesthatseektoreduceurbanrunoff(e.g.,viastormwater

retention),improvewaterquality(e.g.,bypreventingCSOevents),and,insomecasesby

undertakingstreamrestorationprojects,whichreshapethestreamchanneltoreduceerosion

andwithstand“flashy”hydrology(NRCS,2007).Typically,streamrestorationprojectsfocuson

thegeomorphologyofthechannelanddonotconsiderthepotentialforbiologicalrestoration

asameanstoimproveecosystemservices(NRCS,2007).Incontrast,biologicalrestorationis

usedwithintheBayitselftoachievewaterqualitytargets.Forexample,theChesapeakeOyster

BMP(Cornwelletal.,2016)wasestablishedonthebasisthatoystersfilterparticulatematter

P a g e |8

fromthewatercolumnandincreasenitrogenremovalthroughenhanceddenitrification

(Cornwelletal.,2016).ResearchfromthePartnershipforDelawareEstuary(PDE)hasshown

thattheclearanceandfiltrationratesoffreshwatermusselsrivalthatofoysters.Kreegeretal.

(2017)reportedthatmass-specificclearanceratesforfreshwatermusselsrangedfrom0.5to

3.4lh-1g-1,whileEasternOystershadaclearancerateof6to6.4lh-1g-1.Thesamestudy

suggestedthatmusselbedshaveahigherclearanceratethanoysterbeds,duetohigher

populationdensity.Thecriticalecosystemservicesthatoysterscontributetotheenvironment

arepotentiallythesameecosystemservicesthatfreshwatermusselscouldprovideinupstream

areas.Thesefindingssuggestthatstockingfreshwatermusselsintributarystreamsmaybea

usefulmeansforreducingnutrientandsedimentinputstotheBay.

MusselrestorationeffortshaveoccurredthroughouttheAtlanticslope,thoughthese

typicallyfocusonspeciesthatareapriorityforconservation,andinhighqualityhabitats.In

additiontostateandfederalfacilities,someenvironmentalgroupshavealsobegunmussel

restorationactivitiestoaugmentthepopulationofAlewifeFloater,Utterbackiaimplicata,

mussels(DelawareEstuary,2016).TheNationalStrategyforConservingFreshwaterMussels

enumeratesseveralgoalsforpropagatingmusselsandunderstandingwhatfactorsdegradethe

population(Haag&Williams,2014).However,thestrategyprovideslittleguidanceastowhere

therestorationsshouldoccur(Haag&Williams,2014).Manystockingprogramsfocusonlyon

augmentingpopulation,andonlyinareaswhereothermembersofthesamespeciescanbe

found.Asaresult,propagationprogramsleaveoutstreamswheremusselsareorwere

historicallyabsent,andwhichmaybenefitfrommusselrestorationtoimprovelocaland

downstreamwaterquality.Dataareneededtoassesstheviabilityofstockingmusselsinto

impairedstreamssuchasthosefoundinurbanenvironmentsandtobetterunderstandthe

factorswhichaffecttheirperformance(e.g.,survivorship,growthrate).

ObjectivesandHypothesis

Mythesisprojectfocusedonthequestion:Canfreshwatermusselsberestoredinurban

streams?Toaddressthisquestion,A.implicatawerestockedinthreeurbanstreamsinthe

metroareaofRichmond,Virginia.Asacontrolforcomparison,A.implicatawerestockedintwo

ruralstreamsandahatcherypond.Allofthestreamsusedinthisstudyaretributariesofthe

P a g e |9

JamesRiver.Oneoftheruralstreams(HerringCreek)isalsothesourceofwaterforthefish

hatcherypond.Datawerecollectedtoassessmusselsurvivorshipandgrowth(massand

length)at1-2monthintervals.Theprimaryobjectivewastodeterminewhetherfreshwater

musselscansurviveandgrowinurbanstreams.Asecondaryobjectivewastoassessinter-

streamdifferencesinsurvivorshipandgrowthinrelationtoin-streamhabitatconditionssuch

aswaterquality,foodquantity,andfoodquality.Waterqualitymetricsofinterestincluded

watertemperature,pHanddissolvedoxygen.Inurbanstreams,lowpHandoxygenconditions

mayarisefromchronicorepisodicinputsofwastewater(e.g.,CSOevents),whichmayhave

detrimentaleffectsonmusselgrowthandsurvivorship.Temperaturewouldgenerallybe

expectedtohaveapositiveeffectongrowthrates,thoughhightemperatureconditions(e.g.,in

theabsenceofaripariancanopy)maybedetrimental.MetricsoffoodquantitysuchasTotal

SuspendedSolids(TSS)orparticledensity,maybepositivelyrelatedtogrowthandsurvivorship,

thoughinurbanstreams,bedandbankerosionmaysuspendedmaterialsoflowfoodquality

(e.g.,sand,siltandclay).Therefore,measuresoffoodquality(organicmatterandchlorophyll-a

content)werealsousedtoassesstheirrelationshiptogrowthrates.

Methods

JuvenileA.implicatamussels(meanlength42mmSTDdependentonsite)wereplaced

incagesinstalledineachofthreeurbanstreamsandtwonon-urbanstreams.AlewifeFloater

musselswerechosenduetotheiravailabilityfromHarrisonLakeNationalFishHatchery

(CharlesCity,VA),becausetheyarenativetotheregionandhavebeenusedinprior

restorationprojects(Kreegeretal.,2017).Themusselswereindividuallytaggedandmonitored

onamonthlybasisforsurvivorshipandgrowth.Waterquality,substratecharacteristicsand

foodqualityweredocumentedforeachstreamtointerpretdifferencesinsurvivorshipand

growth.Thesamedatawerecollectedfromcontrolsites:arearingpondattheHarrisonLake

FishHatchery,andtwonon-urbanstreams(KimagesCreek,HerringCreek).TheHatcherypond

waschosenasacontrolsiteasithasbeenusedtosuccessfullyrearmusselsforlocalstocking

efforts.HerringCreekandKimagesCreekarelocatednearbyandwerechosenbasedontheir

rurallocationandaccess.KimagesCreekislocatedattheVCURiceRiversCenterandhasa

long-termrecordofbi-monthlywaterqualitymonitoring.

P a g e |10

TheRichmondurbanstreamsselectedforthisstudywerepartofanetworkof7sites

currentlymonitoredbyVCU.Bi-monthlydataarecollectedtomeasuredischargeandwater

quality(temperature,pH,conductivity,turbidity,dissolvedoxygenandTSS).Individualstreams

werechosenbasedontheiraccessibilityandpriordatacharacterizinghydrology,

geomorphologyandwaterquality.Habitatconditions(hydrology)andfoodavailability

(quantityandqualityofparticulatematter)wereexpectedtodifferamongthestudysites.

BroadRockCreek,ReedyCreekandGilliesCreekdrainpredominantlyurbanareaswithahigh

proportionofimpervioussurfacesandlowproportionofforestedareasintheirwatershed

(Table1).AtBroadRock,thepresenceofanupstreamimpoundmentwasexpectedtoprovide

morestableflowconditions,andpotentiallyimprovefoodqualitybytrappingsedimentand

allowingforphytoplanktonproductionthatcouldhavepossiblyimitatedimpoundmentsat

HerringCreek.Itwasexpectedfromlandusedatathatruralstreamswouldharborimproved

musselsurvivalandgrowthrates,buttheriseinimpervioussurfacesatKimagescreekcould

showvariationinmusselsurvivorshipandgrowthbetweenruralsites.

In-StreamDeploymentofMussels

JuvenileA.implicatamusselswerestockedatallsitesinApril-May2019withstocking

ratesrangingfrom106to129persite.Supplementalstockings(20-60mussels)werecarried

outatsomesitesfollowinglossofindividualsduetowashout.Atotalof880musselswere

usedinthisstudy.Sixenclosures(Figure1)containing~19individualsperenclosurewere

placedateachsite.Streamenclosureswereconstructedfromplasticcrates(approx.25x25x

25cm)withplasticwiremesh(1cm)addedalongthesides(2layers),bottom(3layers)andtop

(1layer).Twometalpoleswithflattopscommonlyusedtosecuremulchbarrierswereplaced

atthefronttwocornerstosecuretheenclosuretothestreambottom.Musselenclosuresat

thefishhatcherysiteconsistedoffloatingbaskets(Pattersonetal.2018).Threeenclosures

containing40musselsperbasketwereplacedintoapondattheHarrisonLakeNationalFish

Hatchery.AllmusselsweregivenauniqueIDtag(number1to1,300)ontheirshellbelowthe

umbousingaZinglaserengraver.

P a g e |11

DataCollection

Survivorshipandgrowthweremonitoredat1-2monthintervalsfromthetimeof

stocking(April-May2019)totheconclusionoftheexperiment(December2019).Survivorship

wasdeterminedfromthenumberofalivemusselsinallenclosuresatasiteforagivencensus

date.Separatetallieswererecordedfordeadmussels(recoveryofemptyshells)vs.lost

mussels(arisingfromlossofcagesormissingmusselswithinrecoveredcages).Duetolossof

enclosures,andinsomecases,thelownumberofsurvivingindividualsperenclosure,wedid

notderiveseparateestimatesofsurvivalforeachenclosure.Length(mm)andweight(wet

mass;g)weremeasuredforallindividuals.Growthrates(mm/dandmg/d)werederivedfor

eachmonitoringintervalbasedonrepeatmeasurementsoftaggedindividuals.Inthelater

stagesoftheexperiment,sometagsbecameunreadable.Werecordedlengthandweightfor

theseunknowns,andderivedpopulation-basedestimatesofgrowthratebasedonthechange

inmeanlengthandmassduringthemonitoringinterval.Waterqualitydata(temperature,pH,

conductivity,turbidityanddissolvedoxygen)weremeasuredinconjunctionwithmonitoringof

musselsusingaYSIProDSSsonde.Watersampleswerecollectedtoassessfoodquantityand

qualitybasedontotalsuspendedsolids(TSS),organicmatter(OM)content,andchlorophyll-a

(CHLa)content.SampleanalysisfollowedprotocolsdevelopedfortheVCUEnvironmental

AnalysisLab,astate-accreditedwaterqualitytestingfacility.SamplesforTSSandCHLawere

filteredthroughaGF/Aglassfiberfilters(0.5-μmnominalporesize).FiltersforCHLaanalyses

wereextractedfor18hinbufferedacetoneandanalyzedonaTurnerDesignTD-700

Fluorometer.FiltersforTSSweredriedat60oCfor48handanalyzedusingaPerkin-ElmerCHN

Analyzertodeterminetheorganicmattercontent,expressedasparticulateorganicC(POC).

Particlesizeanddensityweremeasuredbetween2.16and60µmusingaCoulterCounter

Multisizer4e(BeckmanCoulter,Pasadena,California).SampleswerepreservedwithLugol’s

iodinesolutionandrefrigerated.Samplesweredilutedwithelectrolytesolutionstartingwith

5mlofsampleto5mlofelectrolyteandrepeatedthreetimesforarangeofconcentrations(2x,

4x,8x,and16x).TheCoulterCountermeasuresallparticleswithinthespecifiedsizerange

inclusiveofcells(bacteria,phytoplankton)andnon-livingparticulates(e.g.,silt,clay,etc.)with

resultsreportedasnumberofparticlesperunitvolume.

P a g e |12

StatisticalAnalysis

Waterqualitydataandmetricsoffoodquantityandqualitywereanalyzedusingaone-

wayANOVAtodeterminewhetherdifferencesamongsiteswerestatisticallysignificant.

Statisticalanalysiswaslimitedtothe4sitesforwhichbi-monthlydatawereavailable;Reedy

Creek,GilliesCreek,BroadRockCreekandKimagesCreek.

ResultsSurvivorshipofmusselsvariedamongthe6sites(Figure2).Highestsurvivorshipwas

observedatthefishhatcherypondwhere88%ofthestockedmusselssurvivedthroughtheend

oftheexperiment(May-November).Onlyasmallnumber(N=3)werelostduetomortality

(presenceofdeadmussels),whileotherswereunaccountedforanddesignatedaslost.

Survivorshipattheruralsites(KimagesandHerringCreeks)was35%and44%,respectively.At

thesesites,relativelyfewmusselswerelostduetomortality(Kimages=21individuals;Herring

=13individuals),butalargernumberofindividuals(40-60musselspersite)werelostdueto

washoutofcagesfollowingaJulystormevent.Lowestsurvivorshipwasobservedamongthe

threeurbansites.AtReedyCreek,only4musselssurvivedthroughmid-August(3%

survivorship),whileatBroadRockandGilliesCreek,thenumberofsurvivingmusselswas12

and20individuals(6%and14%survivorship,respectively).Attheurbansites,themajorityof

thedecline(84%)wasduetolossofcageseitherthroughwashoutorburial,thoughhigher

ratesofmortalitywerealsoobserved(Reedy=42individuals,Gillies=28individuals).

Length-basedgrowthratesvariedseasonallyandamongsites(Table2;Figure3).

Highestindividual-basedgrowthrates(0.23±0.06mm/d)weremeasuredinHerringCreek

duringJulytoSeptember.Averagegrowthratesatthissitewere0.093mm/dacrossallcensus

periods.Bycomparison,growthratesatthehatcherypondandKimagesCreekwere0.047

mm/dand0.044mm/d,respectively.ThehighgrowthratesatHerringCreekcorrespondtoan

increaseinlengthof31%(from45.4±0.3mmto59.5±0.8mm)overthe5-monthperiod

(May-October).MusselsstockedatthehatcherypondandKimagesCreekincreasedinlengthby

10%and9%,(respectively)overthesameperiod.Thelownumberofsurvivingindividuals

limitedthenumberofdatesforwhichgrowthratescouldbecalculatedattheurbansites.

Averagegrowthrateswereloweramongtheurbanstreamsincomparisontothenon-urban

sites:0.025mm/d(Gillies),0.021mm/d(BroadRock)and0.015mm/d(Reedy).

P a g e |13

Highestmass-basedgrowthrateswereobservedinHerringCreekwheremeangrowth

rateswere57mg/dandpeakvalueswere94±9mg/dfromJulytoSeptember(Table3;Figure

3).HighmeangrowthrateswerealsoobservedduringSeptember-Octoberatthissite(48±14

mg/d)andatthehatcherypond(40±2mg/d)duringMay-July.Amongothersites,average

growthratesrangedfrom1mg/datBroadRockto15mg/datthehatcherypondandReedy

Creek.Negativegrowthrateswererecordedinsomecensusperiods.

Growthratesestimatedusingthelargerdatasetofallmeasuredindividuals(including

thosethatcouldnotbeidentifiedbytagnumber)generallyshowedgoodagreementwiththose

basedonrepeatedmeasurementsoftaggedindividuals(Figure4).Mass-basedgrowthrates,

whetherderivedfromrepeatmeasurementsofindividualsorpopulationmeanvaluesshoweda

highdegreeofcorrespondence(R2=0.96,p<0.001).Length-basedestimatesderivedbythe

twomethodsshowedaweakeragreement(R2=0.52,p<0.001).Increasesinlengthshoweda

strongcorrespondencetoincreasesinmassforboththeindividual-based(R2=0.65,p<0.001)

andpopulation-basedassessments(R2=0.72,p<0.001).Slopesderivedfromthetwodatasets

werenotstatisticallydifferent(0.0017±0.0003and0.0016±0.0002mm/mg,respectively).

Waterqualityconditionsvariedseasonallyandamongsites(Table4;Figure5).

Statisticalanalysisofthesedatafocusedonthe4sitesforwhichbi-weeklydatawereavailable

(N=17measurementsduringMay-December).Watertemperaturesweresimilaramongthe3

urbanandonenon-urban(Kimages)streams(meanrange=16.4to18.0C;p=0.8).Peakwater

temperaturesduringtheperiodofstudywere~22Catthesesites.Higherwatertemperatures

wereobservedinthehatcherypond,whichexceeded25CduringJune-September(peak=32

C).Specificconductancewassignificantlydifferentamongthe4sites(p<0.001)withlowest

valuesatKimagesCreek(mean=104±11uS/cm)andhighervaluesamongurbanstreams

(meanrange=144to197uS/cm).SimilarconductivityvalueswereobservedinHerringCreek

(mean=150±41uS/cm),whereasthehatcherypondexhibitedlowvaluesrelativetothe

streams(mean=38±5uS/cm).Dissolvedoxygenconcentrationsdifferedsignificantlyamong

sites(p<0.001)withlowestvaluesobservedinKimagesCreek(mean=6.7±0.3mg/L;

saturation=70±3%).Dissolvedoxygenvalueswerehigheramongthethreeurbanstreams

(rangeofmeans=9.0to9.6mg/L;saturation=91-100%).Lowoxygenconditionswerealso

P a g e |14

observedinHerringCreek(mean=5.8±1.4mg/L;saturation=59±12%)andononeoccasion

inthehatcherypond(June=2.2mg/L;27%saturation).Highestdischargewasobservedin

HerringCreek(mean=233±114L/s).Amongtheurbansites,GilliesCreekexhibitedthe

highestaveragedischarge(mean=74±4L/s)andhighestpeakvalues>100L/sinJuneand

October).AveragedischargewassimilaramongBroadRock,ReedyandKimagesCreeks(range

ofmeans=14to39L/s).

Foodquantityandqualityvariedbetweensites(Figure6).TSSconcentrations(mg/L)

weresignificantlydifferentamongthe4sitesforwhichbi-monthlydatawereavailable(p<

0.001).HighestconcentrationsweremeasuredinKimagesCreek(mean=16.2±5.2mg/L).TSS

concentrationswereloweramongthethreeurbanstreams(rangeofmeans=1.7to2.2mg/L).

RelativelyfewmeasurementsofTSSwereobtainedfromthehatcherypondandHerringCreek

(N=3and4,respectively),buttheseshowedsomewhatelevatedlevelsrelativetotheurban

streams(mean=8.7and3.8mg/L,respectively).

VariationinCHLaconcentrationsamongsitesgenerallytrackeddifferencesinTSS.

Highestconcentrationswereobservedatthehatcherypond(mean=5.65µg/L±1.16).Average

concentrationsatHerringandKimagesCreekwere2.68±0.63and2.46µg/L±0.58,

respectively).UrbansiteshadlowerCHLaconcentrationsrelativetoruralsiteswithmean

valuesrangingfrom0.76±0.12µg/L(BroadRock)to1.91±0.29µg/L(Reedy).Differencesin

CHLaamongsiteswerestatisticallysignificant(p<0.001).POCconcentrationsgenerally

followedTSSandCHLashowinglowerconcentrationsattheurbansites.KimagesCreekandthe

hatcherypondexhibitedthehighestPOCconcentrations(means=1.16±0.22and1.08±0.11

mg/L,respectively),followedbyHerringCreek(0.66±0.08mg/L).Urbanstreamshadtwo-fold

lowerPOCconcentrationscomparedtotheruralsiteswithmeanvaluesrangingfrom0.22±.03

(Gillies)to0.22±0.04mg/L(BroadRock)to0.31±0.03mg/L(Reedy).Differencesinturbidity

amongsitesmirroredvariationsinTSS,CHLa,andPOC.KimagesCreekandthehatcherypond

hadthehighestaverageconcentrationsofallsites(14.1±3.8and10.2±1.4NTU,respectively).

TherewasnoconsistentseasonalpatterninCHLa,TSS,Turbidity,orPOCconcentrationsamong

thesitesforwhichbi-monthlydatawereavailable.

P a g e |15

Particledensityandparticlesizevariedbetweensites.Higherparticledensitieswere

observedamongtheruralsitesrelativetotheurbansites.Thehatcherypondexhibitedthe

highestparticledensity(mean=158,000±40,000#/ml),thoughthisestimatewasbasedon

only4measurements.Amongthe4sitesforwhichbi-monthlydatawerecollected,Kimages

Creekexhibitedthehighestparticledensity(mean=91,400±3,000#/ml).Averageparticle

densitiesinHerringCreekwere53,1000±4,000#/ml.Urbansiteshadlessthanhalftheparticle

densityofruralsites(Gillies=31,700±3000,BroadRock=29,400±7600,Reedy34,900±7500

#/ml).Differencesamongthe4siteswherebi-monthlydatawereavailableweremarginally

significant(p=0.052).MedianparticlesizewashighestatBroadRock(mean=3.27µm)but

wasotherwisesimilaramongsites(range=2.89to2.98µm;p=0.006).

Seasonalandinter-streamvariationinmusselgrowthrateswasanalyzedinrelationto

waterqualityandfoodresourcemetrics(Figure7).Waterqualitymetricsgenerallydidnot

showastatisticallysignificantrelationshipwithmusselgrowthrates,withtheexceptionof

dissolvedoxygen.Growthratesweregenerallysimilaramongsiteswhereoxygensaturation

wasgreaterthan60%,buttwoofthesiteswherehighgrowthratesweremeasured(Herring

Creekandhatcherypond)exhibitedlowdissolvedoxygen.Thisresultedinaweak(R2=0.25

butstatisticallysignificant(p=0.029)negativerelationshipbetweengrowthanddissolved

oxygen.Foodmetricsdidnotshowstatisticallysignificantrelationshipswithmusselgrowth

rateswiththeexceptionofamarginallysignificant(p=0.07)positiverelationshipwithCHLa(R2

=0.23).

DiscussionDifferencesinsurvivorshipamongurbanandruralsitesreflectedexpectationsbasedon

theurbanstreamsyndrome.Specifically,issuesof“flashy”hydrologyandsedimentationwere

evidentaturbansites(Walshetal.,2016).Multiplecagelossesoccurredateachoftheurban

sites(e.g.,3atGillies,4atBroadRockand6atReedyCreeks).Mostofthesecagelosses

occurredfromlateMaythroughJune.Thoughwelacksite-specificdischargedata,itislikely

thathighrunofffromimpervioussurfacesresultedinlargeincreasesinstreamdischarge

followingrainfallevents,whichresultedinenclosuresbeingsweptaway.Oneofthestudysites

(ReedyCreek)hasa3-yearrecordofwaterleveldata(2015-2018),whichwasusedin

combinationwithtwice-monthlymeasurementsofdischargetodevelopastage-discharge

P a g e |16

relationship(P.Bukaveckas,unpubl.data).Atbaseflow,waterlevelwas10-20cmand

dischargelessthan10L/s.Duringevents,waterlevelexceeded2manddischargeexceeded

3000L/s.Thelargerangeofvariationindepthanddischargehighlightthedifficultiesin

establishingresidentpopulationsofmussels.Sedimentdepositionfrombankerosionwasalso

evidentateachofthesesites,resultinginsiltandsandloadsincagesthatmayhavesuffocated

mussels.PriorstudiesinurbansystemsusingtheAsianClam(Corbicula)alsoreportedhigh

ratesof(Nobles&Zhang,2015),whereasenclosuresinnon-urbansystemsreportedhigher

ratesofsurvivorship(e.g.,60-90%;Haagetal.,2019),whichwerecomparabletothose

observedatourruralsites.Mortalityeffectsmeasuredinthisstudyreflectinparttheuseof

enclosures,whichareneededtorecoverindividualsformeasurementsofgrowthand

survivorship.Lossofcagesisascribedtomortality,thoughitispossiblethatmusselsstocked

directlyintostreamsmaybebetterabletoavoidwashout,ortofindsuitablehabitatin

downstreamareas.Somelossofmusselsmayhavealsooccurredduetoescapefromcages,as

indicatedbytheabsenceofmusselsandlackofdeadshells,butthiswasminorcomponentof

theoverallmortalityincomparisontowashoutlosses.

AlewifeFloatermusselswereselectedforthisstudyinpartduetotheirexpectedhigh

growthrates.Observedgrowthratesbasedonlengthrangedfrom0.04to0.23mm/d,

exceedingthosereportedforothermussels(e.g.,Lampsiliscardium=0.001-0.020mm/d;

Ohlman&Pegg,2020).HighestgrowthrateswereobservedatHerringCreekandthehatchery

pond.ForcomparisonAlewifefloatermusselshaveafastgrowthrateinhatcherysystems

capableof0.219mm/d(Kreegeretal.,2018).Itispossiblethatgrowthratecouldhavebeen

limitedbyaconcentratedpopulationsizeof40perfloatingbasketasopposedto20percage.It

ispossiblethattheconcentrationof40perbasketwasatathresholdwheremusselswere

competingwitheachotherforfeedingandphysicallylimitedduetolackofspacewithinthe

basket.Ifthisweretrue,itmightwouldexplainwhythegrowthratedroppedfrom0.10to0.05

mm/d.GrowthratesintheurbanstreamsandKimagesCreekweregenerallybelowHerring

Creekandthehatcherypond,withonlyafewpeaksreachingequalgrowth.Inter-site

differencesinmass-basedgrowthratesgenerallyfollowedsurvivalrates,showingthaturban

streamshadlowergrowthratesandhighermortalitycomparedtoruralstreams.While

P a g e |17

KimagesCreekdoesnothavealltheattributesoftheurbansitessuchasflashyhydrologyora

watershedwithhighamountsofimpervioussurfaces,thegrowthrateandsurvivorshipof

musselsismorecomparabletourbansitesthanruralsites.

Trendsinfoodqualityandquantityindicatethatdifferencesbetweensitesmaypartly

accountfordifferencesinsurvivalandgrowth.Theresponseoffreshwatermusselstoincreased

algalorsuspendedmatterfluxwithwatervelocityiscomplexandvariesbyspeciesdueto

morphology(Mistry&Ackerman,2015.;Tuttle-raycraft&Ackerman,2019).High

concentrationsofsuspendedparticulatematterareindicativeofpotentiallygreaterfood

resources,thoughmuchofthisparticulatemattermaybeoflowquality(e.g.,siltandclay).

Studieshaveshownthattheclearancerate(numberofparticlesfiltered)ofmusselsdecreases

withhighlevelsofTSS(Tuttle-raycraft&Ackerman,2019).Itisthereforeimportanttoalso

considermetricsoffoodqualitysuchasorganicmattercontentandchlorophyllacontent.

RuralsitesexhibitedhigherTSS,POC,chlorophyll-aandparticledensityrelativetourbansites.

Morefavorablefoodresourceconditionsmayaccountforhighergrowthratesobservedat

someofthenon-urbansites(HerringCreek,Hatcherypond)thoughwedidnotfindsignificant

relationshipsforthese,orwaterqualityparameters,inpredictingvariationingrowthrates.

Conclusion

Priorstudieshavedocumentedtheeffectsofurbanizationonstreamhydrologyand

geomorphology.Typically,theseareassociatedwithhighrunofffromimpervioussurfaces

duringstormevents,whichleadtobedandbankerosion.Theseconditionsareoften

associatedwithimpairmentofstreamfunction(e.g.,reducedsedimentandnutrientretention;

lowbiodiversity).Resultsfromthisstudyfurtherhighlightthechallengestorestoringurban

streamsasexemplifiedbythehighrateofwashoutandburialofintroducedmussels.

Restorationeffortswithinthecatchmentareneededtoreduceurbanrunoff,whichmaythen

allowforsuccessfulre-introductionofmusselsandassociatedimprovementsinstream

ecosystemservices.Datafromthisstudyshowthatwhileintroductionofmusselstourban

streamshadlimitedsuccess,growthandsurvivorshipinnearbyruralstreamswasindicativeof

greaterpotentialforsuccess.Thestockingoffreshwatermusselsintothese,andpotentially,

restoredurbanstreams,maybeausefulapproachtomitigatingnutrientandsediment

P a g e |18

transporttoChesapeakeBay.Furtherstudiesareneededtobetterunderstandwaterquality

andfoodconditionsthatareconducivetosuccessfulestablishmentofmussels,andhowstream

restorationeffortsmaybegearedtoaccommodatingmusselintroduction.

P a g e |19

References

Cornwell,J.C.,Rose,J.,Kellogg,M.L.,Luckenbach,M.W.,Bricker,S.B.,Paynter,K.T.,…

Hudson,K.(2016).PanelrecommendationsontheoysterBMPnutrientandsuspended

sedimentreductioneffectivenessdeterminationdecisionframeworkandnitrogenand

phosphorusassimilationinoystertissuereductioneffectivenessforoysteraquaculture

practices.OysterBMPExpertPanelFirstIncrementalReport.

Eshleman,K.N.,&Sabo,R.D.(2016).Decliningnitrate-NyieldsintheUpperPotomacRiver

Basin:WhatisreallydrivingprogressundertheChesapeakeBayrestoration?Atmospheric

Environment,146,280–289.https://doi.org/10.1016/j.atmosenv.2016.07.004

Estuary,D.,Estuary,N.,December,P.,&No,P.D.E.R.(2016).FreshwaterMussel

ReintroductionResearchProject2016 :AdvancementsinFreshwaterMusselRestoration

atLongwoodGardensFreshwaterMusselReintroductionResearchProject2016 :

AdvancementsinFreshwaterMusselRestorationatLongwoodGardens,(16).

Haag,W.R.,Culp,J.J.,Mcgregor,M.A.,Bringolf,R.,&Stoeckel,J.A.(2019).Growthand

survivalofjuvenilefreshwatermusselsinstreams :Implicationsforunderstanding

enigmaticmusseldeclines.FreshwaterScience38:753–770.

https://doi.org/10.1086/705919.

Haag,W.R.,&Williams,J.D.(2014).Biodiversityonthebrink:Anassessmentofconservation

strategiesforNorthAmericanfreshwatermussels.Hydrobiologia,735(1),45–60.

https://doi.org/10.1007/s10750-013-1524-7

Hoellein,T.J.,Zarnoch,C.B.,Bruesewitz,D.A.,&DeMartini,J.(2017).Contributionsof

freshwatermussels(Unionidae)tonutrientcyclinginanurbanriver:filtration,recycling,

P a g e |20

storage,andremoval.Biogeochemistry,135(3),307–324.https://doi.org/10.1007/s10533-

017-0376-z

Kreeger,D.A.,Gatenby,C.M.,&Bergstrom,P.W.(2017).Comparativeevaluationofnative

bivalvespeciesrestorationforwaterqualityimprovementintheChesapeakeBayand

othermid-Atlanticwatersheds.PartnershipfortheDelawareEstuaryReport,(17-05),96.

Kreeger,D.A.,Gatenby,C.M.,Bergstrom,P.W.,Service,W.,Great,L.,Fish,L.,…Road,C.

(2018).Restorationpotentialofseveralnativespeciesofbivalvemolluscsforwaterquality

improvementinMid-Atlanticwatersheds.JournalofShellfishResearch37(5),1121–1157.

https://doi.org/10.2983/035.037.0524

Lucas,R.(2019).CharacterizingwaterqualityandhydrologicpropertiesofUrbanStreamsin

CentralVirginia.,VCURichmond,VA.

McConnell,K.A.(2017).LimitsofAmericanFarmBureauFederationv.EPAandtheClean

WaterAct’sTMDLProvisionintheMississippiRiverBasin.EcologyLawQuarterly,44(2),

469–501.https://doi.org/10.15779/Z38M61BP5N

Nedeau,E.J.,Merritt,R.W.,&Kaufman,M.G.(2003).Theeffectofanindustrialeffluentonan

urbanstreambenthiccommunity:Waterqualityvs.habitatquality.Environmental

Pollution,123(1),1–13.https://doi.org/10.1016/S0269-7491(02)00363-9

Nobles,T.,&Zhang,Y.(2015).Survival,growthandconditionoffreshwatermussels:Effectsof

municipalwastewatereffluent.PLoSONE,10(6),1–20.

https://doi.org/10.1371/journal.pone.0128488

NRCS.(n.d.).Chapter1:Introduction:EcologicalandPhysicalConsiderationsforStream

Projects.NationalEngineeringHandbook,654.

P a g e |21

Ohlman,L.M.,&Pegg,M.A.(2020).Handlingeffectsonsurvivalandgrowthofplain

pocketbookLampsiliscardium(Rafinesque,1820)freshwatermussels.Hydrobiologia,

847(2),457–467.https://doi.org/10.1007/s10750-019-04106-y

Patterson,M.A.,Mair,R.A.,Eckert,N.L.,Gatenby,C.M.,Brady,T.,Jones,J.W.,Simmons,B.

R.,Devers,J.L..(2018)Freshwatermusselpropagationforrestoration.Cambridge

UniversityPress,334pp.ISBN-13:978-1108445313.

Smucker,N.J.,&Detenbeck,N.E.(2014).Meta-analysisoflostecosystemattributesinurban

streamsandtheeffectivenessofout-of-channelmanagementpractices.Restoration

Ecology,22(6),741–748.https://doi.org/10.1111/rec.12134

Tuttle-Raycraft,S.,&Ackerman,J.D.(2019).Livingthehighturbiditylife :Theeffectsoftotal

suspendedsolids,flow,andgillmorphologyonmusselfeeding,Limnology&

Oceanography64:2526–2537.https://doi.org/10.1002/lno.11202

Vaughn,C.C.(2018).Ecosystemservicesprovidedbyfreshwatermussels.Hydrobiologia,

810(1),15–27.https://doi.org/10.1007/s10750-017-3139-x

Walsh,C.J.,Roy,A.H.,Feminella,J.W.,Cottingham,P.D.,Groffman,P.M.,Ii,R.P.M.,&Ii,R.

A.P.M.O.(2016).Theurbanstreamsyndrome :currentknowledgeandthesearchfora

cure.JournaloftheNorthAmericanBenthologicalSociety,24(3),706–723.

https://doi.org/10.1016/B978-0-12-801231-4.00027-6

Williams,J.D.,Warren,Jr.,M.L.,Cummings,K.S.,Harris,J.L.,&Neves,R.J.(1993).

ConservationStatusofFreshwaterMusselsoftheUnitedStatesandCanada.Fisheries,

18(9),6–22.https://doi.org/10.1577/1548-8446(1993)018<0006

Winter,J.G.,&Duthie,H.C.(1998).Effectsofurbanizationonwaterquality,periphytonand

P a g e |22

invertebratecommunitiesinasouthernOntariostream.CanadianWaterResources

Journal23:245-257.https://doi.org/10.4296/cwrj2303245

Yeager,M.M.,Cherry,D.S.,&Neves,D.J.(1994).Feedingandburrowingbehaviorsofjuvenile

RainbowMussels,Villosairis(Bivalvia :Unionidae).JournaloftheNorthAmerican

BenthologicalSociety,13:217–222.

P a g e |23

Table1.Location,watershedarea,substratecomposition,andsurroundinglandusefrom

NOAAlandusedatabaseandgraduatestudentthesis(Lucas,2019)(NOAA,2010)offive

streamsselectedformusselintroduction.

P a g e |24

Table2.Growthratesofmussels(lengthinmm/d)stockedintourbanandnon-urbanstreams.

Individualgrowthratesarebasedonrepeatmeasurementsofthesameindividual;population

growthratesarederivedfromtheaveragelengthofallindividuals.Stockingdatesforeachsite

areshowninthefirstcolumnandcensusdatesinthesecondcolumn.

Site Date Mean SE N Mean NHatchery 7/15/2019 0.072 0.003 116 0.071 119

5/15/2019 8/27/2019 0.022 0.006 99 0.007 11210/2/2019 0.040 0.027 92 -0.002 10411/2/2019 0.052 0.027 86 0.015 100

Kimages 7/24/2019 0.031 0.003 45 0.044 655/18/2019 9/1/2019 0.071 0.043 39 0.014 45

9/30/2019 0.006 0.006 31 0.007 4710/29/2019 0.096 0.041 23 -0.005 4011/21/2019 0.014 0.040 17 -0.006 33

Herring 7/14/2019 0.052 0.006 17 0.052 375/17/2019 9/7/2019 0.225 0.061 14 0.137 52

10/22/2019 0.082 0.015 18 0.075 62Gillies 8/3/2019 0.013 NA 1 0.02 29

5/27/2019 9/15/2019 0.005 0.001 21 0.009 2311/2/2019 0.057 0.051 17 -0.011 21

Broad Rock 7/1/2019 0.028 0.023 4 -0.031 45/20/2019 8/15/2019 0.034 0.029 13 0.034 13

9/17/2019 0.001 0.004 11 0.001 11Reedy

5/22/2019 7/9/2019 0.015 0.027 4 0.027 107/7/2019 8/12/2019 0.064 0.050 2 0.076 4

Individual GR (mm/d) Population GR (mm/d)

P a g e |25

Table3.Growthratesofmussels(massinmg/d)stockedintourbanandnon-urbanstreams.

Individualgrowthratesarebasedonrepeatmeasurementsofthesameindividual;population

growthratesarederivedfromtheaveragelengthofallindividuals.Stockingdatesforeachsite

areshowninthefirstcolumn.

Site Date Mean SE N Mean NHatchery 7/15/2019 40.1 1.6 116 39.3 119

5/15/2019 8/27/2019 10.5 1.4 99 4.3 11210/2/2019 1.2 3.3 92 -1.1 10411/2/2019 5.7 2.6 86 1.3 100

Kimages 7/24/2019 10.5 1.6 45 16.0 655/18/2019 9/1/2019 13.8 1.9 39 14.2 45

9/30/2019 6.6 4.9 31 3.2 4710/29/2019 -1.1 6.2 23 -10.7 4011/21/2019 -2.5 4.3 17 -5.5 33

Herring 7/14/2019 29.0 4.9 17 31.5 375/17/2019 9/7/2019 94.2 9.1 14 77.0 52

10/22/2019 48.0 13.9 18 39.6 62Gillies 8/3/2019 9.5 NA 1 9.4 29

5/27/2019 9/15/2019 10.6 1.9 21 8.4 2311/2/2019 -3.7 1.3 17 -8.3 21

Broad Rock 7/1/2019 7.5 2.3 2 0.0 85/20/2019 8/15/2019 -4.6 5.3 13 -4.6 13

9/17/2019 0.9 1.4 11 0.9 11Reedy

5/22/2019 7/9/2019 11.3 8.4 4 10.2 107/7/2019 8/12/2019 15.8 1.3 2 0.0 4

Individual GR (mg/d) Population GR (mg/d)

P a g e |26

Table4.MeanvaluesofwaterqualityvariablesrecordedateachsiteduringApriltoDecember2019.

Site Temp pH Conductivity DO DO DischargeC µS/cm % mg/L m3/s

Herring 18.7 6.9 150.3 59.4 5.8 0.233Pond 23.2 7.3 38.4 81.4 7.3 N/AKimages 16.4 7.1 103.7 69.5 6.7 0.021BroadRock 17.9 7.5 144.0 99.8 9.6 0.039Reedy 17.2 7.4 196.8 90.9 9 0.014Gillies 17.8 7.2 157.6 94.1 9.1 0.072

P a g e |27

Figure1.Designofinstreammusselenclosures

P a g e |28

Figure2.StockingnumberandfateofAlewifeFloatermusselsinthreeurbanstreams(Reedy,

BroadRock,andGillies),2non-urbanstreams(Herring,Kimages)andafishhatcheryrearing

pond.Thebarsrepresentthenumberofmussels,bluerepresentingstockedmussels,orange

representingnumberofmusselsalive,greyrepresentingnumberofmusselslost,andyellow

representingnumberofmussels’dead.

P a g e |29

Figure3.Lengthandmassbasedgrowthrates(withstandarderror)ofAlewifeFloatermussels

inurbanandruralstreamsaswellasthehatcherypond.Datashownareaveragevaluesbased

onrepeatmeasurementsofindividuallytaggedmussels.

-20

0

20

40

60

80

100

120

Jul Aug Oct Nov Jul Aug Sep Oct Nov Jul Sep Oct

GR(m

g/d)

GR-mass

-20

0

20

40

60

80

100

120

Aug Sep Nov Jul Aug Sep Jul Aug

GR(m

g/d)

GR-mass

-0.05

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

Jul Aug Oct Nov Jul Aug Sep Oct Nov Jul Sep Oct

GR(m

m/d)

GR-length

Non-UrbanSites UrbanSites

-0.05

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

Aug Sep Nov Jul Aug Sep Jul Aug

GR(m

m/d)

GR-length

JulAugOctNovJulAugSepOctNovJulSepOctHatcheryKimagesHerring

AugSepNovJulAugSepJulAugGilliesBroadRockReedy

P a g e |30

Figure4.Comparisonsofindividualandpopulationbasedgrowthratesasmassandlength(upperpanels).Comparisonsofmassandlengthbasedgrowthratesderivedfromrepeatedmeasurementsofindividualsandpopulationmeans(lowerpanels).Allregressionsp<0.001.

R²=0.9607

-20

0

20

40

60

80

100

-20 0 20 40 60 80 100

Popu

latio

nGR

(mg/d)

IndividualGR(mg/d)

Indvs.PopGR(mass)

R²=0.5213

-0.05

0.00

0.05

0.10

0.15

0.00 0.05 0.10 0.15 0.20 0.25

Popu

latio

nGR

(mm/d)

IndividualGR(mm/d)

Indvs.PopGR(length)

R²=0.6495

0.00

0.05

0.10

0.15

0.20

0.25

-20 0 20 40 60 80 100

Individu

alGR(m

m/d)

IndividualGR(mg/d)

GR-massvs.GR-length(Ind)

R²=0.8744

-0.05

0.00

0.05

0.10

0.15

-20 0 20 40 60 80 100

Popu

latio

nGR

(mm/d)

PopulationGR(mg/d)

GR-massvs.GR-length(Pop)

P a g e |31

Figure5.Waterqualityconditionsatstudysitesduringtheperiodwhenmusselsweredeployed

fromJune-December2019.

P a g e |32

Figure6.Foodquantityandqualitymetricsforurbanstreams(Gil=Gillies,BR=BroadRock,

Ree=Reedy)and3non-urbansites(Kim=Kimages,Her=Herring,Pon=hatcherypond).Data

shownaremedians(darkblackline),25%and75%quartiles(boxes),95%confidenceintervals

(bars)andoutliers(circles).

P a g e |33

Figure7.Waterqualityandfoodmetricsaspredictorsofmusselmassgrowthrates.Regression

linesdenotestatisticallysignificantrelationships.

-20

0

20

40

60

80

100

0 10 20 30 40

GR(m

g/d)

WaterTempC

-20

0

20

40

60

80

100

0 20 40 60 80 100 120

GR(m

g/d)

DisO2(%)

-20

0

20

40

60

80

100

0 50000 100000 150000 200000 250000

GR(m

g/d)

ParticleDensity(#/ml)

-20

0

20

40

60

80

100

0 5 10 15 20 25

GR(m

g/d)

TSS(mg/L)

-20

0

20

40

60

80

100

0 2 4 6 8 10

GR(m

g/d)

CHLa(µg/L)

R2 = 0.23p=0.07

-20

0

20

40

60

80

100

0.0 0.5 1.0 1.5 2.0

GR(m

g/d)

POC(mg/L)

R2 = 0.25p=0.03