1

BringingfishlifebacktoNoosa:restoringlostoysterreefhabitatsintheNoosa

Biosphere

FinalReport

Dr Ben Gilby, Prof. Thomas Schlacher,

Dr Andrew Olds, Prof. Rod Connolly,

Dr Christopher Henderson, Miss Cassandra Duncan, Mr Nicholas Ortodossi,

Mr Thomas Brook, Ms Felicity Hardcastle, and

Mr Ashley Rummell.

School of Science and Engineering University of the Sunshine Coast

QLD, 4556, Australia

2

TableofContents

Acknowledgementtoprojectpartnersandcontributors……..….…………………………………..…………..…….3

Summary………………………………………………………………………………………………………………….…..…………..…….4

Chapter1:IntroductionandbackgroundtotheNoosaRiverOysterReefRestorationTrail……………..9

Chapter2:Spatialrestorationecology:placingrestorationinalandscapecontext………....……..….…29

Chapter3:Maximisingthebenefitsofoysterreefrestorationforfinfishandtheirfisheries..…………51

Chapter4:Seascapecontextmodifieshowfishrespondtorestoredoysterreefstructures…….……79

Chapter5:Landscapecontextmodifiestherateanddistributionofpredationaroundhabitat

restorationsites…………………………………………………………………………………………………………………….….……99

Chapter6:Effectsofoysterreefrestorationextendacrosscoastalseascapes…………………………...…119

Chapter7:Identifyingrestorationhotspotstomaximiseoutcomesformultiplerestorationbenefits

………………………………………………………………………………………………………………………………………………….…..135

References…………………………….………………………………………………………………………………….…….…….…….160

Appendix:MonitoringoftheNoosaRiverOysterReefs;November2017-November2018:Reportto

NoosaCouncil,andQueenslandDepartmentofAgricultureandFisheries……….………..……………..…190

3

Acknowledgement to project partners and contributors TheauthorsacknowledgeNoosaCouncil,NoosaBiosphereReserveFoundation,TheThomas

Foundation,NoosaParksAssociationandtheUniversityoftheSunshineCoastforprovidingfunding

forthisresearchandforchampioningoysterreefrestorationeffortsintheNoosaRiver.Theauthors

particularlyacknowledgethegenerouscontributionsofChrisGillies(TheNatureConservancy,

Australia),CherieO’Sullivan(NoosaCouncil),MichaelGloster(NPA)andSimonWalker(Ecological

ServiceProfessionals)totherestorationeffortsinNoosa.

Theauthorsacknowledgethecontributionofourco-authorsCharlesPeterson(UniversityofNorth

Carolina),ChristineVoss(UniversityofNorthCarolina),MelanieBishop(MacquarieUniversity),

MichaelElliot(UniversityofHull)andJonathanGrabowski(NortheasternUniversity)towardsthe

conceptsandideasdiscussedinchapterthree.

Finally,theauthorsacknowledgetheeffortsofNicholasYabsley,HaydenBorland,SarahThackwray,

EllenBingham,JarrenCollins,TysonJonesandLucyGoodridgeGainesandcountlessotherstudent

volunteersfromUniversityoftheSunshineCoasttowardsfieldworkandtheprocessingofdata.

4

Summary

Ecologicalrestorationfacilitatestherecoveryofecosystemsandcanaugmentorre-establishanimal

populationsbothatrestorationsitesandacrosslandscapesmorebroadly.Withincreasing

recognitionofthedamagethathumanactivitieshaveonecosystems,ecologicalrestoration

activitiesareincreasinginbothprevalenceandscaleacrossmarine,freshwaterandterrestrial

ecosystemsglobally,andoftenhavespecificgoalsaroundenhancingbothlostordegraded

ecosystems,andtheanimalpopulationsthatthoseecosystemssupport.Despitethis,therehasbeen

relativelyfewstudiesthathavequantifiedtheeffectsofecologicalrestorationonmultipleattributes

ofrestoredecosystems,andthenusedthisinformationtooptimisethedesignandplacementof

subsequentefforts.

Oysterreefs,abiogeniccoastalecosystemprovidinghabitatforadiversityofestuarinespecies,are

significantlythreatenedbyhumanactivities.Globally,upto85%ofoysterreefshavebeenlost,and

someregionshaveexperiencedupto96%lossduetothecombinedeffectsofdisease,declining

waterquality,andoverharvesting.Thislossofstructurallycomplexoysterreefsfromcoastaland

estuarineecosystemshasbeenhypothesisedasasignificantcontributortowardsfishdiversityand

fisheriesdeclinesinmanycoastalsystems.Consequently,oysterreefrestorationhasincreasedin

prevalenceglobally.Whilsttheseeffortshaveincreasinglyfocusedonenhancingtheabundanceand

diversityoffishandfisheriesatrestorationsites,fewstudieshavequantifiedtheeffectsofoyster

restorationonfishinregionsoutsideofNorthAmerica,ontheecologicalfunctionsthatfishprovide

aroundrestoredreefs,orthedegreetowhichthebenefitsofoysterrestorationexpandacross

entireestuaries.Consequently,thissortofinformationhasrarelybeenusedtooptimiseoysterreef

restorationplansandmonitoringprograms.

TheNoosaRiveroysterreefrestorationtrialisthefirstshellfishrestorationprojectinstalledin

Queensland,Australia.Theprojectsoughttorestorethestructurallycomplexoysterreefsthatwere

onceabundantintheNoosaRiver,therebyenhancinghabitatavailabilityandcomplexityforfish

speciesofcommercialandrecreationalsignificance,andsubsequentlytheimportantecological

functionsthatthesespeciesperform.Weinstalledandsurveyed14oysterreefrestorationsitesin

theNoosaRiverestuary,thatwereplacedtoencompassdifferencesinseascapeconnectivityand

proximitytonearbymangroves,seagrasses,urbanstructuresandtheestuarymouth.Thesereefs

thereforeprovidetheoptimalstudydesigntotestfortheeffectsofspatialplacementanddesignof

oysterrestorationactionsonfish,fisheriesandecologicalfunctions.Thisreporthassixkeyaims.We

settheglobalcontextfortheoysterreefrestorationprojectinNoosaby(1)identifyingthedegreeto

5

whichlandscapecontextisconsideredintheselectionofrestorationsitesacrossallenvironmental

realmsglobally;and(2)identifyinggapsintheglobalunderstandingoftheeffectivenessofoyster

restorationfinfish,andsynthesisecurrentresearchtooptimisefutureoutcomes.Wethenusedthe

oysterrestorationsitesintheNoosaRiverestuaryto;(3)identifywhethertheseascapecontextof

restoredoysterreefs(especiallyseascapeconnectivitywithseagrassesandmangroves)modifiesthe

degreetowhichtheyaugmentfishabundanceanddiversity;(4)determinetheeffectsofoysterreef

restorationontherateanddistributionofthekeyecologicalfunctionofpredationatandaroundthe

restoredoysterreefs;(5)quantifyhowrestoredoysterreefsinfluencethedistributionand

abundanceoffishacrosstheestuary(i.e.beyondthereefsthemselves);and(6)usespatial

modellingtechniquestooptimisetheplacementofsubsequentoysterreefrestorationeffortsinthe

riverforoystergrowth,fish,fisheriesandecologicalfunctioningbyincorporatingallinformation

gatheredinchapters3,4and5,andour2018annualmonitoringreport.

Toidentifythedegreetowhichlandscapecontext(i.e.thesize,shape,andspatialarrangementof

ecosystems)isconsideredwhenselectingrestorationsitesglobally(seeChapter2),wereviewed

restorationprojectsfromtheglobalprimaryliterature.Fewerthanoneineightrestorationprojects

consideredlandscapecontextintheselectionofrestorationsites(11%of472projects).Thisfigure

wasremarkablysimilaracrossterrestrial(10%of243projects),marine(13%of89)andfreshwater

(13%of164)ecosystems.Ofthe54restorationprojectsthatconsideredlandscapecontextinsite

selection,justoverhalf(56%)reportedthatanimalpopulationswerelargerormorediversethanin

controlareas.Theseresultsindicatethatmoretightlyintegratingconceptsfromspatialecologyand

systematicconservationplanningintorestorationpractice,suchastheconsiderationofthe

landscapecontext,couldimprovethedesign,optimiseplacementandenhancetheecological

effectivenessofrestorationprojectsinallecosystems.

Toidentifygapsintheglobalunderstandingoftheeffectivenessofoysterrestorationforfinfishand

synthesisecurrentresearchtooptimisefutureoutcomes(seeChapter3),wereviewedtheglobal

literatureofoysterreefrestorationforfinfish.Globaldeclinesinoysterreefshaveresultedin

reducedhabitatheterogeneity,extent,andqualityforsomecoastalfinfish,potentiallyreducingfish

populationsandcatches.Although,itiswellestablishedthathabitatrestorationresultsinhigher

finfishbiomassanddiversitywhereoysterreefsreplacebaresandormudsubstrates,theprinciples

ofhabitatquality,ecologicalconnectivityandbroaderecosystemmanagementarepoorlyintegrated

withinoysterreefrestorationecology.Theinclusionofsuchprinciplescouldbeinstructivein

enhancingthebenefitsofprojectsonfishpopulationsthroughoutestuarineseascapesandincrease

6

stakeholderengagementandcost-effectiveness.Thisreviewpresentsaframeworkforprojects

seekingtorestorebothoysterreefhabitatandfinfishcommunities.

Toidentifywhethertheseascapecontextofindividualreefsmodifiesthedegreetowhichthey

augmentfishabundanceanddiversity(seeChapter4),weusedbaitedremoteunderwatervideo

stations(BRUVS)tosurveyfishassemblagesatthe14-oysterreefandatcontrolsites.Theseascape

contextofcoastalecosystemsplaysapivotalroleinshapingpatternsinfishrecruitment,abundance

anddiversity,however,thereisapaucityofinformationregardingthetrajectoriesinwhichthese

relationshipsfollowwhilststructuringtherecruitmentoffishassemblagestorestoredhabitats.Our

resultsfoundthatfishassemblagesatoysterreefsdifferedfromthoseatcontrolsites,withhigher

speciesrichness(1.4times)andmoreindividualsoftaxathatareharvestedbyfishers(1.8times).

Thepresenceorabsenceofseagrassnearbyaffectedtheabundanceofakeyharvestablefish

species(yellowfinbreamAcanthopagrusaustralis)onoysterreefs,butnotthecompositionoffish

assemblages,speciesrichnessorthetotalabundanceofharvestablefishesoverall.Ourfindings

highlighttheimportanceofconsideringspecies-specificpatternsinseascapeutilisationwhen

selectingrestorationsitesandsettingrestorationgoalsandsuggestthattheeffectsofrestorationon

fishassemblagesmightbeoptimisedbyfocussingeffortsinprimepositionsincoastalseascapes.

Todeterminetheeffectsofoysterreefrestorationontherateanddistributionofpredation(see

Chapter5),weusedsquidpopassays(driedsquidtetheredusingfishingline)tomeasurepredation

ratesandsurveypredatorsatsixrestoredoysterreefsandnearbycontrolsites.Ecologicalfunctions

areimportantinmaintaining,andenhancing,coastalecosystems.Restoringcoastalhabitats

enhancesthesefunctions,butfewstudieshaveexplicitlyquantifiedthedegreeandspatialextentof

thesemodifications.Ithasbeenwellestablishedthattherateanddistributionofecological

functionsismodifiedbyhowspeciesrespondtothecompositionoflandscapes,however,itisnot

clearwhererestorationsitesshouldbelocatedinheterogeneouslandscapestomaximiseoutcomes

forecosystemfunctions.Predationratesatrestoredoysterreefsweredoublethoseatcontrolsites.

Seascapecontextwasimportantinmodifyingthesepredationrates;consumptionnearreefswas

significantlylowerwhenreefswereclosetoseagrassandmangroves.Bycontrast,higherrateswere

observedonreefssurroundedbynon-vegetatedseafloor,farfromseagrassandmangroves.In

addition,thedistanceoverwhichpredationextendedintothesurroundingunvegetatedareaswas

greateronreefsfartherfromvegetation.Thesefindingshaveimportantimplicationsforplanning

restorationactionsbothinseaandonlandbecausetheynecessitatethatpractitionersunderstand

thebasicspatialpatternsthatarelikelytodrivetheabundanceanddistributionoffunctionally

importantspeciesacrossecosystems.

7

Toquantifyhowtherestoredoysterreefsmodifythedistributionandabundanceoffishacrossthe

estuary(i.e.beyondthereefsthemselves)(seeChapter6),wemonitoredfishatvideomonitoring

sitesina200mgridacrosstheentirelowerNoosaestuarybothbeforeandaftertheinstallationof

oysterreefs.Quantifyingtheeffectsthatrestorationhasonthespatialdistributionofanimalsboth

atrestorationsites,andacrosslandscapesisanimportantfocusforrestorationecologists.Asthe

scaleofrestorationincreasesacrossallenvironmentalrealmsglobally,thecapacityfortheeffectsof

restorationonanimalpopulations,andthereforesomeofthebenefitsofrestoration,toexpand

acrosslandscapesincreases.Wefoundthatfishassemblagesvariedsignificantlyacrosstemporal

scalesintheNoosaRiver,andthattherewasasignificantchangeintheassemblagecomposition

anddistributionoffishintheriverfollowingtheinstallationoftheoysterreefs.Theseeffects

proliferatedtoeffectsonspeciesrichnessandharvestedfishabundance,whichwerebothhigherat

greaterdistancesfromtherestoredreefsfollowingreefinstallation.Thesefindingsruncounterto

theeffectsfoundinotherstudiesinthisreportthatfoundasignificantlyhigherabundanceoffish

occurringatreefsandsignificantlyhigherratesofkeyecologicalfunctions.Changesin

environmentalconditionsacrosstheestuary(especiallyfreshwaterrunofffromcatchments)might

havehamperedthereefsinreachingtheirecologicalpotential.Theseresultshaveseveralkey

consequencesforthedesignsofsubsequentrestorationactionsintheNoosaRiverestuary.Firstly,

ensuringthattheeffectsofrestorationexpandsignificantlyacrosstheentireseascapeoftheNoosa

River(i.e.theyincreasetheoverallcarryingcapacityofthesystemmorebroadly)willrequirelarger

restorationsitesacrosstheentireNoosaseascapeto1)enhancetheaugmentationeffectsofthe

reefsand2)spreadbiomassacrosstheestuary.Further,amorethoroughconsiderationofthe

effectsofseascapeconnectivitywithdeeperchannels,andalong-termcommitmenttothorough

monitoringandevaluationoftheeffectsoftheoysterreefsacrosstheentireestuaryisrequired.

Tooptimisetheplacementofoysterreefrestorationeffortforoystergrowth,fish,fisheriesand

ecologicalfunctioning(seeChapter7),weuseddistributionmodelstoidentifyhotspotsintheNoosa

Riverestuarywheresubsequentrestorationeffortswouldmaximiseoverallbenefits.Oysterreefs

canbesuccessfullyrestoredthroughouttheentireNoosaRiverestuarybecauseoysterssuccessfully

settledandgrewrapidlyateverytrialsite.Restorationsitesalsohadhigherfishspeciesrichness,

harvestablefishabundance,andratesofkeyecologicalfunctions(predationandcarrionscavenging)

thanatnearbycontrolsites.However,thegrowthofoysterreefs(intermsofoysterdensityand

size)wasmostcorrelatedwiththeproximityofrestorationsitestourbanisedshorelines(i.e.sources

ofoysterlarvae)andtheestuarymouth(i.e.growthalongthesalinitygradient),andtheareaof

8

mangrovesaroundthesite(i.e.sourceofoysterlarvaeandjuvenileoysterpredators).Conversely,

theabundanceanddiversityoffish,andtheratesofpredationandscavengingatreefstendedtobe

mostrelatedtotheirisolationfromseagrassesandmangroves.Thisresultedindifferentmetrics

drivingthescaleanddistributionofrestorationbenefitsforfishesandfunctionsandmeantthatthe

areathatmaximisedoutcomesforallofoysterreefgrowth,fishes,andfunctionswassignificantly

smallerthanthetotalareathatcouldberestored.Hotspotswhereoysterrestorationwouldresultin

thegreatestpossibleoutcomesacrossalloftherestorationtargetsconsideredoccurrednearthe

estuarymouth,onthenorthernbankoftheestuaryapproximately2kmfromtheestuarymouth,

andinalargelake-likeinletinthefarwestoftheestuary(knownlocallyasLakeDoonella).Withthe

scaleofrestorationactionsincreasingacrosslandscapesglobally,usinginformationregardingthe

distributionoflikelyrestorationbenefitsinaquantitativemannercanhelptoreducethefinancial

andsocialcostsofrestoration,whilstmaximisingsocial,economic,andenvironmentaloutcomes.

Ourfindingshighlighttheimportanceoftwokeyconceptsinoptimizingtheoutcomesofrestoration

forbothecosystemsandtheanimalsthatinhabitthem;

1. Restorationpractitionersshouldmoreeffectivelyconsiderthepivotalroleoflandscape

contextandhabitatconnectivityontheeffectivenessofrestorationactionsforanimals,and

2. Restorationpractitionersshouldmorethoroughlyadopttheprinciplesofsystematic

conservationplanningintheplacementanddesignofrestorationactions;especiallyforthe

settingofquantitativegoalsandinthedesignofsubsequentmonitoringprograms.

OurfindingsshowthatecologicalrestorationofoysterreefstotheNoosaRiverresultedin;

1. Oysterreefstructuresthatattractedandgrewoystersthatwilleventuallydevelopintoself-

sustainingoysterreefs,

2. Anaugmentationoffishspeciesrichness(1.4times),harvestablefishabundance(1.8times)

andtheecologicalfunctionsofpredation(2times)andscavenging(1.5times)atthe

restoredreefsrelativetonearbycontrols,and

3. Someeffectoftheoysterrestorationeffortsonthedistributionofecologicalfunctionsand

theabundanceoffishacrosstheNoosaRiverestuarymorebroadly.

Inthissense,theNoosaRiveroysterrestorationprojecthassetthestandardgloballyforquantifying,

andsubsequentlyoptimisingrestorationactionsacrosslandscapesforthepurposesofaugmenting

fish,fisheries,fishbiodiversityandecologicalfunctioning.

9

Chapter1

IntroductionandBackgroundtotheNoosaRiver

OysterReefRestorationTrial

BenL.Gilby,AndrewD.Olds,CassandraK.Duncan,NicholasL.Ortodossi,

ChristopherJ.Henderson,andThomasA.Schlacher

Oysterreefsarevitalfishhabitats

Oysterreefsarebiogenicstructureswheresubsequentgenerationsoflarvaloysterssettle

onoldergenerations,andgrowprogressivelyoutwards.Occurringpredominantlyinthe

lowerintertidalthroughoutsouth-eastQueensland,oysterreefsprovidestructurally

complexandfood-richhabitatforadiversityoffishspecies,includingmanyspecieswhich

areofcommercialand/orrecreationalvalue(Diggles2013)(Figure1).

TheimportanceofoysterreefsasahabitattypeisrecognisedbytheConventionon

WetlandsofInternationalImportance(TheRamsarConvention–habitattype‘Ga’)(Kasoar

etal.2015).Oysterreefsprovideimportanthabitatforestuarinefishesthroughtwokey

mechanisms:1)byofferingstructurallycomplexstructureswithinestuariesthatfishuseas

habitatforresting,seekingrefugefrompredators,andspawning;and2)bydeliveringa

diversityoffoodresourcesforfishes,includingplanktonicprey,smallerfishes,andthe

oystersthemselves,whichoccurinhigherconcentrationinandaroundthereefs(Harding

andMann2001).Asaconsequence,just10m2

ofrestoredoysterreefshavebeenshownto

yieldanadditional2.6kgoffishbiomassperyearoverthelifetimeofthereefsrelativeto

sandysubstrates(Petersonetal.2003).

Globally,85%ofoysterreefshavebecomeextinctasaresultofoverharvesting,disease,and

poorwaterquality(Becketal.2011).Lostoysterreefsarecommonlyreplacedwithhabitats

thatprovidelessfoodorpoorerprotectionfrompredatorsforfish,suchasbaremudsand

sands(Grabowskietal.2012).Theseprofoundhabitatchangestypicallypropagatetosharp

declinesinfishdiversityandbiomass,severelyimpactingfisheries(Coenetal.1999).For

thesereasons,oysterreefsarebeingrestoredinmanylocationsworldwide,andoftenwith

theexplicitgoalofenhancingfishandfisheries(Gilbyetal.2018c).Itisimportanttonote,

therefore,thatsimplyhavingoystersinasystem(e.g.oystersgrowingonjetties,pylons,

artificialrockwalls)doesnotsubstitutethefunctional,ecologicalorhabitatrolesthat

naturaloysterreefshaveforestuarineecosystems.Urbanstructuresalsodonotprovidethe

sameseascapecomplexity(Petersonetal.2003).Therefore,activerestorationeffortsthat

restoreboththelostoysterreefbiogenicstructurestoestuaries,aswellasthestructural

11

complexityandfoodwebbenefitsderivedfromoysterreefsthatenhancesecondary

productionoffishisvital.

Figure1NaturaloysterreefstructureatToorbulPoint,Queenslandintheearly1900’s.Reefsconstitutea

structurallycomplex,biogenichabitatthatprovidesfeedingopportunities,sheltersitesandspawningareasfor

fishinestuaries.Oysterreefscanalsoslowerosionprocessesandcanimprovewaterqualityatalocalscale.

Positiveeffectsofreefsonfisharethoughttoresultprimarilyfromenhancedlarval

settlementandsurvival,andthemovementofadultfishtoreefs(Breitburgetal.1995;

Gilbyetal.2018c).Notwithstandingtheverylargefinancialinvestmentsinreefrestoration

globally,itissurprisingthatquantitativedataonfishenhancementsattributabletoreefs,

thestatedgoalofmanyinvestments,arelimitedessentiallytoverysmallareasatafew

locationsineasternUSA(Gilbyetal.2018c;Petersonetal.2003;zuErmgassenetal.2016).

Itissurprisingthatquantitativeinformationonthewaysrestoredoysterreefsmeettheir

statedrestorationobjectivestoenhancefishstocksatthescaleofwholeestuariesisnot

moreprevalent(Gilbyetal.2017a).

Fishhabitatsaretypicallyarrangedinaspatialmosaicofdifferenttypes(i.e.‘seascapes’)

(Nagelkerkenetal.2015;Oldsetal.2016;Yeageretal.2011).Intheseseascapes,thespatial

propertiesofhabitats(e.g.theirproximityandconnectivitytostructurallycomplexhabitats

12

suchasmangroves,seagrasses,marshes,andreefs)isimportantindeterminingtheirvalue

asfishhabitats(Bostrometal.2011;Oldsetal.2012b;Pittman2018).Inthecontextof

restoredoysterreefs,itis,therefore,plausiblethatthespatialpropertiesofareefmay

modifyany‘reefeffects’onfishassemblagesandecologicalfunctionsperformedbyfish

(Grabowskietal.2005;IrlandiandCrawford1997;MicheliandPeterson1999;Piersonand

Eggleston2014);whilstthisisaclearandattractivehypothesis,ithasneverbeentested

withempiricaldata.

HistoryofoysterreefsintheNoosaRiver

ThecommercialoysterindustryintheNoosaRivercommencedinthe1870’sor1880s,and

ceased,havingbecomeeconomicallyunviable,bythe1940’s(Thurstan2016).Oystering

locationswererecordedbylocalfisheriesofficialsthroughoutthisperiod(from1934;see

Figure2),whichoutlinetheextentoftheoysterleaseareasandtherebyindicatethemost

probablehistoricalrangeofoysterbedswithintheestuary.Theseoysterleaseareaswere

dredgedintensivelyforoystersintheearly1900’s,resultingintheremovalofliveoyster

bedsandtheunderlyingbedrock.Thisresultedinthelossoflargeareasofintertidaland

subtidalhardsubstratesthatarerequiredforoysterspattosettle,grow,andformreefs.

Theendpointofthisprocesswasthatoysterreefswerenolongerpresentasasubstantial

habitatinthelowerandmiddlereachesoftheestuary(Thurstan2016).

Thespeciesofoystersthatoccurinsouth-eastQueensland(primarilyrockoysters

Saccostreaspp.)formreefspredominantlyinintertidalhabitats,anduptoadepthof1

metrebelowthelowestastronomicaltides.Thisknowndepthrangecanbeusedtopredict

thelikelydistributionofoysters(assumingtheywerenotremovedbydredging)throughout

thesystembyusingbathymetrydatafortheNoosaRiver.Historically,thelikelyspatial

extentofoysterreefsintheNoosaRiverisestimatedtobeintherangeof41,530to

207,650m2

(ESP2016).Aminimumof8,241m2

ofoysterbedswasremovedperyearduring

thepeakoysterharvest.Approximately16,850m2

ofoldoysterbedsandoysterrubbleis

currentlyburiedbysedimentinthehistoricaloysterleaseareas.Oftheoriginalextentof

naturallyoccurringoysterreefsinNoosaRiver,approximately20,647m2

ofdegradedliving

oysterbeds(i.e.smallandisolatedpatches,oftenheavilysiltcoveredandcontainingmany

13

deadshells)andoysterrubbleremainscatteredthroughoutthesystem.Theseoysterbeds

havealowprofileandalowsurfacearea–volumeratio,areeasilycoveredbysilt

(especiallyfromwindsuspensionorseasonalrainevents),andthereforedonotprovidethe

sameopportunitiesforoysterrecruitmentasestablishedoysterreefsdo(ESP2016).

Additionally,thecurrentextentofdegradedoysterbedsandoysterrubbleintheNoosa

River(20,647m2

)isonlyasmallfractionofthetotalareaofoysterreefhabitats(41,530to

207,650m2

)thatarelikelytohavebeenpresentbeforecommercialharvestingcommenced

(ESP2016).Oysterspatare,however,stillpresent,andabundant,intheNoosaRiver,and

recruittosuitablehardsubstratewhenitispresent(TNCandESP2015).Thereisnolonger

suitablesubstrateforoysterlarvaetosettleonnaturaloysterreefsubstrate(i.e.biogenic

reefsubstrates)intheriver.Therefore,itisclearthemethodsofdredgingoystersfrom

thoseareashaveremovedsubstratesuitableforoysterrecruitment.Provisionof

appropriatesubstratetoallowoysterreefstoredevelopwithintheNoosaRiver,would

restoreadiversityofhabitattoimportantfisheriesspecies.

Asafishhabitat,oysterreefsarenowconsideredfunctionallyextirpatedwithinthesystem.

Thisdrastichabitatchangehasbeenaccompaniedbysignificantdeclinesinfishstocksover

thelastcentury(Thurstan2016).Theextent,composition,andqualityoffishhabitatsinthe

NoosaRiverhasthereforechangedsubstantiallysinceEuropeansettlement,withlikely

adverseflow-oneffectsonaquaticbiotaandresources,includingfishandfisheries

(Thurstanetal.2017).ThestatusoffishandfisheriesintheNoosaRiver,andthecondition

andtypeofhabitatssupportingthem,isofconsiderableconcerntothelocalcommunity.

Thisisthecatalystforthecurrentprojectaimedatrestoringoysterreefsasanessentialfish

habitatthathasbecomeextirpatedinthesystem.Thisprojectaimstopartlyredressthis

habitatlossbyrestoringoysterreefsatanumberoflocationsinthelowerreachesofthe

estuary,aimingtoenhancethequalityofhabitatforfishspeciesknowntobeclosely

associatedwithestuarineoysterreefs(e.g.yellowfinbream,tailor,mosesperch,grouper).

14

Figure2Ahand-drawnchartfrom1934,madebytheFisheriesInspectortohighlighttheareasintheNoosa

Riverclosedtonetfishing.TherednumbersindicatethelocationsofallknownNoosaoystersections.Source:

QueenslandStateArchives.

OysterreefrestorationintheNoosaRiver

TheprincipalobjectiveoftheNoosaRiveroysterreefrestorationprojectistorestore

naturaloysterreefhabitatintheNoosaEstuarytoenhancetheabundanceanddiversityof

fishthroughouttheestuary.Thisrequirestheadditionofsuitablehardsubstrateforoyster

spattosettle,andallowingnaturalrecruitmentandreef-formationprocessestorecoverto

areasthathistoricallysupportedoysterreefs.

Theoystersettlementsubstratecomprisesoysterreefrestorationunits.Theseunitsare

madefromthemostsuitablerecruitmentmaterial-oystershellsheldtogetherbyanatural

coirfibrebag.Theunitsareraisedabovethemuddyandsandymarinesubstrates,thereby

mimickingtheverticalreliefoftheoriginaloysterreefstofacilitatenaturalrecruitment

processes(Figure3,4).Weexpectedthatnaturalrecruitmentprocesseswouldcementthe

15

deadshelltogetherandformpartofthemosaicofhabitatswithintheestuaryincluding

bothsoftandhardstructuralhabitattypes,therebycreatingastructurallydiversemosaicof

habitatsthatispredictedtobebeneficialtoarangeoffishspecies,includingspeciesof

harvestedincommercialandrecreationalfisheries(e.g.yellowfinbream,estuarycod,tailor,

mangrovejack,andmosesperch).

TherestorationsitesthatwerechosenaslocationsforoysterreefrestorationintheNoosa

River(Figure5)becausethey:

1) areinreachesofthelowerNoosaRiverestuaryandinWeybaCreekfromwhich

oysterswereharvestedhistorically;

2) arewithinthedepthrangeknowntobesuitableforoysterreefstogrow;

3) areinlocationswhereviableoysterlarvaewererecordedduringrecentrecruitment

studies;and

4) haveanextentwherethefinalrestoredareawillnotexceedthehistoricalareal

extentofoysterreefsbeforecommercialharvestingtookplace.

Figure3Coirmeshbagfilledwithrecycledoystershells;theprinciplerestorationunit

16

Figure4ApprovedcoastalengineeringdrawingoftheoysterreefrestorationunitsinstalledintheNoosaRiver

17

Figure5MapofoysterrestorationsitesandmarinehabitatsintheNoosaRiver.

18

TheinstallationofoysterreefunitstotheNoosaRiverwasdoneduringaperiodofthree

days,from20to22November2017.Allworksandsurveyswereconductedbyateam

comprisingprofessionalsfromNoosaJettyBuilders,andmarinescientistsfromthe

UniversityoftheSunshineCoast.Alloysterreefunitsandsignswereinstalledexactly

accordingtospecificationssetoutinthepermitconditionsandanyinstructionsprovided

regardingprocedures.Allreefunitsarelocatedtoavoidanydamagetomarineplantsor

surroundingecosystem(Table1).Keyobservationsduringtheplacementofthereefunits

andassociatedsignageincludethefollowing:

• Nomarineplantsweredisturbedduringtheinstallationofthereefunitsorsigns.

Beforeeachreefunitwasplacedontheseafloor,wecarefullyexaminedthesiteto

ascertainthatnoseagrass,mangroveaerialroots,oranyothermarinevegetation

waspresent.Thisensuredthatnodirectplacementimpactsoccurredatanyofthe

sites(Table1).

• Alloperationswerelocalisedtotheimmediatesitewherereefunitsweresunkonto

theseabed.Thus,nohabitatoutsidetheRAAareawasimpactedinanyformbythe

reefunitsorthesignage.

• Polesholdingthesignswereinstalledusinganarrowwaterjet‘spear’.Useofthis

gearensuredthatonlyaverynarrow(<20cmdiameterinallcases)areaofthe

seabedwasdisturbed(Figure6).Signpostswerethendrivenfurtherintothe

substrate,toameandepthof2.7m(min1.8m,max3.5m;Table1),usingapost

driver(Figure6).

• Accordingtoengineeringspecifications,weusedmarinegradehardwoodstakes

(sourcedfromtheAustralianHardwoodStakeCompany)tosecurethereefunitson

theseafloor.Thestakeswerehammeredintotheseabedtoameandepthof0.9m

(min0.5m,max1.2m)-dependingonhardnessoftheseabed(Table1),atamean

angleof82.4º(min75º,max90º).

19

Figure6Theresultingverysmallfootprintofsignpostsfollowinginstallationbythewaterjet(left)anda

completedoysterreefinWeybaCreek(right).

20

Table1.Completedoyster-reefinstallationsintheNoosaRiver,listingtheirexactposition,thedateofplacement,characteristicsofpostsandstakesdrivenintotheseabedtosecuresignageandreefunits,andanypresenceand/ordamagetomarinevegetation.SiteNumber Position Installation

dateDepthtowhichsignpostisdrivenintotheseabed(m)

Stakestosecurereefunits Presenceof,ordamageto,marineplantsand/oradjacenthabitatsorecosystems.Latitude Longitude Depthof

stakeintoseabed(m)

Angleofstakerelativetoseabed(degrees)

1 reefunitsnotinstalled-seabednotcompactenough 2 -26.38333096 153.045835 20/11/17 2.8 1.2 79 None3 -26.38673804 153.043291 20/11/17 2.8 1.2 75 None4 -26.39367197 153.044599 20/11/17 2.9 1.1 84 None5 -26.394757 153.049933 20/11/17 2.4 0.5 82 None6 -26.39245399 153.051636 20/11/17 3.2 1.2 85 None7 -26.39351204 153.054941 20/11/17 3.3 1.2 77 None8 -26.39491903 153.061679 20/11/17 2.4 1.2 90 None9 -26.38954598 153.070696 21/11/17 1.8 0.5 86 None10 -26.38452999 153.071546 21/11/17 2.0 0.6 87 None11 -26.39817104 153.080595 21/11/17 2.3 0.5 76 None12 -26.40045603 153.077035 21/11/17 2.9 0.5 79 None13 -26.41099703 153.071263 22/11/17 3.5 1.2 87 None14 -26.41184301 153.0712 22/11/17 3.2 0.5 83 None15 reefunitsnotinstalled–presenceofseagrass 16 -26.43981904 153.062079 22/11/17 2.9 0.6 84 None

21

OysterreefgrowthandstabilitymonitoringfortheNoosaRiveroysterreefrestoration

project

Aspartofthedevelopmentapprovalforinstallingtheoysterreefs,thestakeholdersare

requiredtoreportyearly(inDecember)toQueenslandDepartmentofAgricultureand

Fisheriesontheprogressoftheoysterrestorationproject,especiallyrelatingto;

1. Oysterrestorationunitstability(i.e.oysterreefrestorationunitsremainwithinthe

allocatedresourceallocationareas);

2. Restorationofnaturalrecruitmentprocessesoverlongterm(i.e.spatrecruitment

ratesdemonstratethatthebiogenicmatrixwillbesufficienttoholdthestructureof

theoysterreefrestorationunitsinplacefollowingcompletedegradationofthe

supportingcoirmaterialduringtheestablishmentphaseandtofacilitateongoing

naturalrecruitmentprocesses);

3. EquitableCommunityImpacts(i.e.ensuringfaircommunityuseoftheriversystemis

notimpactedbyplacementoftheOysterReefRestorationUnitswithintheResource

AllocationAreas-tobemonitoredusingacounciloperatedcommunityfeedback

system);and

4. Nonegativeimpactonmarineplantsorshorelineerosion(i.e.ensuringtheoyster

restorationunitsdonotimpedenaturalmarineplantgrowthoracceleratecoastal

erosionprocesses).

Theseperformanceobjectivesarequantifiedusingestablishedmonitoringprotocolsfor

oysterreefrestoration,andfollowinternationalbestpracticeformonitoringrestoredoyster

habitats(Baggettetal.2014).Themonitoringprogramisdesignedtobeadaptive,with

annualreviewsagainsttheperformanceobjectivesfortheproject(SeeTable2fordetailed

monitoringrequirements).Notethatthesemonitoringobjectivesagreedtowiththe

DepartmentofAgricultureandFisheriesreferonlytothegrowth,stabilityandeffectsofthe

oysterreefunitsthemselves,andarenotrequiredtomonitoranyaspectofthefish

assemblagesrecruitingtothereefs.Thefishassemblagemonitoringundertakenaspartof

therestorationprojectisdetailedintheotherchaptersofthisreport.

Resultsofthetwo2018compulsorymonitoringeventsareprovidedasinAppendix1.Inthis

report,weshowthat;

1. OysterrestorationunitshavenotmovedfromtheRAAarea,norhavetheymoved

greaterthan1mwithintheRAAarea.

22

2. Therehasbeensignificantspatsettlementandoystergrowthatalloysterreef

restorationsites.Whilstthishasnotyetproliferatedtocoverageofinvertebrateson

theoutsideofthecoirbags,ortotheproperstabilisationofshellwithinthebags

(i.e.cementingofthebiogenicreefmatrix),theseoystergrowthresultsarea

positivesignforthelikelysuccessofthereefsinachievingtheseperformance

criteriainthenearfuture.

3. Whilstwedidnotidentifyanywilfuldamagetotheoysterrestorationunits,wedid

identifythatmarkerbuoyswereregularlyremovedfromtheoysterrestorationsites,

andtherehavebeenseveralinstancesofboatpropellerandanchorstrikesonthe

reefs.Thiswasrepairedwherepossible(i.e.correctiveactions),andafurther

educationcampaignisrequiredtoeducateriverusers.

4. Therehasbeennosignificantchangetothedistribution,compositionorqualityof

seagrassormangrovesaroundtheoysterreefsites.Similarly,therehasbeenno

shorelineerosionatoysterreefsites.

Consequently,twoofthefourmonitoringtargetsreportedonherehavebeenfullymet,the

thirdrelatingtooystergrowthistrackingveryfavourablywiththecriterialikelytobemetin

thecomingyears,andthefinalrelatingtocommunityusagemayrequireclosermonitoring.

Weidentifiedsomedamagetotheoysterreefunitsfromboatingactivities(principally

anchorandpropdamage)atmanysitesduringtheNovember2018monitoringevent.We

havesomeevidencetosuggestthattheuseofoysterreefsbyfishershasincreasedover

time,andthatsomeoftheecologicalbenefitsofthereefsmightthereforebebeingoffset

byfishing.Correctiveactionsweremadetothesereefsimmediatelyfollowingthe

monitoringperiodinNovember2018.Thethreereefsmoredistantfromthemainboating

activitiesinLakeWeybaandWeybaCreek(duetothecombinedeffectsoflimitsonaccess

tohireboats,anddifficultyofaccessduetoshallowwaters)werethemostintactoyster

reefs,withnodamagerecordedineither2018monitoringevents.Whilstthesereefsdidnot

necessarilyhavethehighestratesofsettlement(theexceptionbeingreefnumber13,which

hasthehighestaverageliveoysterdensity),theirsuccessmightbethemostguaranteedas

theyarelesslikelytobedamagedbythesortsofimpactsoccurringonreefsinthecentral

stretchoftheriver.

23

Removalofoysterrestorationunitsin2019

Despitethepositiveoutcomesfromthe2018monitoringreport,weidentifiedsubstantial

damagetotheoysterrestorationunitsfollowingtheChristmasandNewYearholidaysof

2018-2019.Themajorityofthisdamageisthoughttohavebeencausedbytheoyster

restorationunitsbeingstruckbyboatpropellersoranchors(Figure7).Thisresultedinthe

coirmeshbagsbeingsplitopen,withoystershellsbeingspilledthroughouttherestoration

area.Thereappearedtobeanincreaseinthetheftofthemarkerbuoysthatwereplacedon

thecornersoftherestorationareatowarnboatsofthereeflocationsovertheChristmas

holidayswhichmayhavecontributedtowardstheincreaseddamage.

Asaconsequenceofthisdamage,adecisionwasmadebytheNoosaBiosphereReserve

FoundationtoremovetenoysterreefrestorationsitesfromtheNoosaRiver(Figure8).The

removalwasconductedbytheUSCteamonthe13thand14thofFebruary2019.Oyster

restorationunitsweredisposedofimmediatelyattheCleanawayDoonanSolidWaste

Depot,aspertheconditionsoftheBeneficialReusePermitfortheoystershellsissuedby

QueenslandDepartmentofEnvironmentandHeritageProtection,andtheconditionsofthe

approvedDevelopmentApplication.

Figure7TwodamagedoysterrestorationunitsremovedfromtheNoosaRiverinFebruary2019.Theuniton

therightwaspositionedasthetopbaginthetriangularprismrestorationunits,andsustainedsubstantial

damagetothetopsurfacefromeitheraboatpropelleroranchor.Thissortofdamagetothetopbagsinthe

oysterrestorationunitswascommonfollowingthe2018-19Christmasholidays.

24

Table2NoosaoysterreefrestorationmonitoringscheduleasagreedtobyallstakeholdersinRAA2016CA0575,andincludingmark-upsfromDepartmentofAgriculture

andFisheriesdated6January2017.NotethatthesemonitoringobjectivesagreedtowiththeDepartmentofAgricultureandFisheriesreferonlytothegrowth,stability

andeffectsoftheoysterreefunitsthemselves,andarenotrequiredtomonitoranyaspectofthefishassemblagesrecruitingtothereefs.Thefishassemblagemonitoring

undertakenaspartoftherestorationprojectisdetailedbelow.

Performanceobjective Monitoringmethod Frequency Correctiveactionwhereperformanceobjectiveisnot

met

1.OysterRestorationUnitlocationstability

OysterrestorationunitsremainwithinthedesignatedResourceAllocationAreas.

VisuallyinspectthestabilityofoysterreefunitsandrecordthepreciseGPSposition(±cmscale),andsizeofeachunit(followinginternationalbestpractice:Baggettetal.2015).UseGISsoftwaretocontrasttheposition,footprint,sizeandareaofoysterreefrestorationunitsbetweenmonitoringeventsandassessanypotentialmovement.

Every6monthsforaminimumofthreeyearsAdditionalmonitoringwillbeconductedwithin2weeksofsubstantialrainfallevents(i.e.eventsthatexceed50-yearAverageRainfallIntervals).

Aprofessionalcoastalengineershallbeconsultedtosuggestremedialactionforanyoysterreefrestorationunitsthatcollapse,orshiftby>1m.AnyoysterreefrestorationunitsthatmoveoutsidethedesignatedResourceAllocationAreaswithin3yearswillberemoved.Aprofessionalecologistwillconfirmthecauseoftheshift.

2.NaturalRecruitmentProcessesOystersandothersessilebenthicinvertebratesrecruittoreefrestorationunitstoestablishabiogenicmatrix,whichbindsoystershellsinplace,priortodegradationofcoirmaterial.TheKeyPerformanceIndicatorsforthisare:1.)Oysterrecruitment:successfulrecruitmentofoysterspatinatleast1outofthe3years(i.e.33%ofthetime)postdeployment(followinginternationalbestpracticeof40%:Baggettetal.2015).1;2.)Coverofoystersandothersessilebenthicinvertebrates:anupwardtrendinthecoverofsessilebenthicinvertebrates

Quantifytherecruitmentofoystersandothersessilebenthicinvertebratestoreefrestorationunits,andmeasurechangesinthecoverofoystersandothersessilebenthicinvertebratesovertime(followinginternationalbestpractice:Baggettetal.2015).1

ThemonitoringmethodsforeachIndicatorsare:1.)Oysterrecruitment:markedoystershellswillbefastenedtotheoutsideofrestorationunits.Theseshellswillbeharvestedatregularintervals(15oystershellsperlocationoneachevent–i.e.5shellsperunit)andthedensityandsizeofrecruitsrecorded;2.)Coverofoystersandothersessilebenthicinvertebrates:photographsoftenquadrats(25cmx25cm),distributedinastratifiedrandomdesign,acrosseachoysterreefrestorationunitwillbetakenatregularintervalstoquantifythechangein

Every6monthsforaminimumofthreeyears,unlessotherwisedetailedwithincorrectiveactions

Wewillcontrast:1)therateofoysterrecruitment;2)thecoverofoystersandallothersessilebenthicinvertebrates;and3)thestabilityofthebiogenicmatrixamongoysterreefrestorationunitsandbetweenmonitoringevents.AnyrestorationunitsthatfailtomeettheKeyPerformanceIndicatorswithin3yearswillberemoved.Ifmajordamageoccurstothecoirmeshofanyoysterreefrestorationunitsbeforeastablebiogenicmatrixhasformed,andlooseoystershellsarebeinglostfromthestructure,itwillberepairedin-situbyhandweaving.Anysuchrepairswillbeconductedcarefully,andbyhand,toensurethatthereisnodamagetoestablishedareasofbiogenicmatrix.

25

growingonrestorationunits;and3.)Establishmentofstable

biogenicmatrix:structuralrigidityofoysterrestorationunits,denotingastablebiogenicmatrixafter3yearspostdeployment,whichissufficienttoholdoystershellsinplace.

coverofoystersandothersessilebenthicinvertebrates(Baggettetal.2015)1.3.)Establishmentofstablebiogenicmatrix:assessthestructuralintegrityofeachoysterreefrestorationunitandmonitorthedegradationofcoirmaterial.Structuralintegritywillbequantifiedbymeasuringtheproportionofoystershells(from10shellsthatareselectedatrandomateachlocation),whichcanberemovedeasilybyhandmanipulation.Therateofcoirdegradationwillbemeasuredbymonitoringchangestotheconditionofcoirmeshovertimefromthetenphoto-quadrats,whicharecollectedtwiceperyeartoassesschangesinthecoverofbenthicinvertebratesateachoysterreefrestorationunit.

3.CommunityuseandenjoymentofthedeclaredFishHabitatArea

OysterrestorationunitsdonotsignificantlyimpaircommunityuseandenjoymentofthedeclaredFishHabitatArea,particularlyfishingactivities.

Maintainrecordsofcommunityfeedback,evidenceofvandalism,andvesselstrikesonthetrialoysterreefrestorationunits.Recordswillbecomprehensiveandinclude,asaminimumset:1)thenumberandtypeofcomplaintsreceived(also,toallowtheDepartmenttogaugethesuccess,anypositivecommentsshouldalsobeprovided);2)thetype,nature,andseverityofanyactsofwillfulvandalism;and3)thetypeandseverityofanyvesselstrike.

Annualreportingofallcomplaints,casesofwillfulvandalism,andinstancesofvesselstrikereceivedforthethree-yeartrialperiod.

Complaints:within3monthsofreceivingeachcomplaint,aninvestigation(includinginterviewwiththecomplainantifpossible)willbeconductedtodeterminewhethercomplaintsaredirectlyrelatedtooysterreefrestorationunits.PotentialresponseactionswillbeprovidedtotheDepartmentofAgricultureandFisheriesintheannualreport.Ifcomplaintspersist,andgrievancesexceedthedirectbenefitofoysterreefrestoration,theparticularoysterreefrestorationunitwillberemovedassoonaspractical,andpriortothe4thanniversarypostdeployment.WillfulvandalismandvesselStrikes

Ifthereareconsistentand/orsignificantcasesofwillfulvandalismto,orinstancesofvesselsstriking,theoysterreefrestorationunits,thecauseoftheimpactwillbeidentifiedandusedtoguidethedeliveryofacommunityeducationcampaignaimedatreducingthesetypesofincidents.Iftheeducationcampaigndoesnotreducecasesofwillfulvandalismand/orinstancesofvesselstrike,theoysterreefrestorationunitsthatareresponsiblewillberemovedpriortothe4thanniversarypostdeployment.

26

4.Otherpotentialeffects:

Oysterreefrestorationunitsdonotcauseadeclineintheextentofmarineplantswithina50mradiusoftherestorationunits,andarenotattributedtoerosionofadjacentshorelinesorotherambientenvironmentalimpacts.

Maptheareaofmarineplanthabitats(seagrass,mangroves)intheimmediatevicinity(i.e.witha50mbuffer)ofeachoysterrestorationarea,usinghigh-resolutionGPS(cmscale)andfield-validatedaerialimagery(sourcedfromNearmap).Monitorchangesinthecompositioncoverageandconditionofseagrasswithin50mradiusaroundtheoysterreefs.Mapthelocationandconditionofestuarineshorelinesthatoccurintheimmediatevicinity(i.e.witha50mbuffer)ofeachoysterreef,usinghigh-resolutionGPS(cmscale)andfield-validatedaerialimagery(sourcedfromNearmap).

Annuallyforaminimumofthreeyears,oruntilanyfailedrestorationunitshavebeenremoved.

Wheremonitoringshowsthattherearesubstantial,andconsistent,lossesofmarineplants,orerosionofestuarineshorelines,intheimmediatevicinityofoysterreefrestorationunits,andthesechangescanunambiguouslybeattributedtooysterreefrestorationactivities,theproblemunitsthatareresponsibleforsuchimpactsshallberemoved(assoonaspracticalandpriortothefollowingannualreportinganniversary).

27

Figure8MapofoysterrestorationsitesremovedduringFebruary2018,andthosethatremain.

28

ObjectivesThebroaderNoosaRiveroysterrestorationprojectthereforehadsixkeycomponents.Eachofthese

componentsarereportedoninseparatechaptershere:

• Reviewofallrestorationliterature:Identifythedegreetowhichlandscapecontextwas

consideredtoguidetheselectionofrestorationsites(Chapter2);

• Reviewofoysterreefrestorationforfishliterature:Identifygapsintheglobal

understandingoftheeffectivenessofoysterrestorationforfinfish,andsynthesisecurrent

researchtooptimisefutureoutcomes(Chapter3);

• Seascapecontext:Identifywhethertheseascapecontextofindividualreefsmodifiedtheir

augmentationoftheabundanceanddiversityoffish(Chapter4);

• Ecologicalfunctioning:Determinetheeffectsofoysterreefrestorationontherateand

distributionofpredationinanestuarineseascape(Chapter5);

• Effectsofoysterreefsthroughoutseascapes:Quantifyhowtherestoredreefsinfluencethe

distributionandabundanceoffishacrosstheestuary(i.e.beyondthereefsthemselves)

(Chapter6);and,

• Spatialprioritisationmapping:Optimisetheplacementofoysterreefrestorationeffortfor

oystergrowth,fish,fisheriesandecologicalfunctioningusingstatisticalandspatial

modellingtechniquesbyincorporatingallinformationgatheredinchapters4,5and6,and

the2018annualmonitoringreport(Chapter7).

29

Chapter2

Spatialrestorationecology:placingrestorationina

landscapecontext

GilbyBL,OldsAD,ConnollyRM,HendersonCJ,SchlacherTA(2018)Spatial

RestorationEcology:PlacingRestorationinaLandscapeContext.BioScience

68:1007-1019

30

IntroductionEstablishingnaturereservesandrestoringecosystemsarecomplementaryapproachesin

conservation(Holletal.2003;Palmeretal.2016;Presseyetal.2007;Soule1985)thatrequire

significantfinancialinvestment(DeGrootetal.2013;Wilsonetal.2006).Maximisingthenetreturns

oftheseinvestments,intermsofbothbiodiversityandecosystemprotection,isthereforesensible

(Halpernetal.2013;Possinghametal.2015).Commonstrategiesforenhancingconservation

returnsaretoplacereservesinareasthatarethreatened,supporthighbiodiversity,orincurlower

socialandeconomiccosts(Halpernetal.2013),ortoprotectorrestoresitesthatmighthave

synergisticbenefitsforconservation(Moilanenetal.2011;Thomsonetal.2009).

Conservationandrestorationareas,bytheirverynature,arepositionedwithinheterogeneous

landscapes,whichcomprisemultipleecosystemsofdifferentsizesandshapes,andwithvarying

degreesofinter-connectedness(Forman1995;Oldsetal.2018c).Thelandscapecontextcan

stronglyinfluenceseveralkeybioticattributesofasite(Holletal.2003).Manyoftheseattributes

areofdirectconservationandrestorationconcern(e.g.populationdynamics,biodiversity,

productivity)(BunnandArthington2002;Wardetal.1999),acrossterrestrial(HanskiandSaccheri

2006),marine(MicheliandPeterson1999)andfreshwater(Wiens2002)realms.Animalsrespond

stronglytolandscapeconfigurationbecausemostmovethroughoutlandscapes,andsomeuse

differenthabitatsthroughouttheirlifecycles.Somesitesare,therefore,ofhigherecologicalvaluefor

animalsthanothersbecausetheydifferintermsoftheirfoodresources,refugevaluefromweather

orpredators,accessibilitytodispersalpathways,andnumerousotherecologicalpropertiesthathelp

toshapethefitnessofindividuals,demographicsofpopulations,andthecompositionof

assemblages(Pittmanetal.2011).

Globally,acrossmultipleecosystemsandrealms,ithasbeenestablishedthatplacingconservation

areasstrategicallywithinlandscapescanhavesynergisticbenefitsfortheirecologicaleffectsonboth

ecosystems,andtheanimalsthatinhabitthem(Oldsetal.2016;Rudnicketal.2012).Forexample,

manymarinefishmoveamongcoralreefs,seagrassmeadowsandmangroveforeststhroughout

theirlives,andbothfishabundanceanddiversityisoftengreatestinsidemarinereservesthat

conservetheseecosystemsandthepathwaysthatlinkthemacrosslandscapes(Edwardsetal.2009;

Mumby2006;Oldsetal.2016).Becausesomeofthesefishperformimportantecologicalfunctions

(e.g.herbivory),protectingconnectionsbetweencoralreefs,seagrasses,andmangrovescanalso

helptoimprovethespatialresilienceoftheseecosystemstodisturbance(e.g.Magrisetal.2014;

Oldsetal.2014).Furthermore,thebenefitsofplacingreservesintonetworkstomaximisethe

31

exchangeofindividuals,matterandenergyamongthemiswidelyrecognised(e.g.ArturoSánchez-

Azofeifaetal.2003;Harrisonetal.2012;Hermosoetal.2016).Reservesthatareplacedinareas

withgreaterconnectivityamongpopulations,amongecosystems,orwithotherreserves,therefore

oftenperformbetterthanreserveswithimpoverishedlandscapeconnections(Oldsetal.2016;

Ribeiroetal.2009;Stomsetal.2005).

Landscapecontextcanhavesimilarlypositiveecologicaleffectsontheperformanceofrestoration

projects(MetzgerandBrancalion2016).Restoredecosystemswithstrongconnectionstoother

habitatpatches,ofeitherthesameordifferentecosystemtypes,aremorelikelytobesettledby

animals,andreceivelargersubsidiesofmatterandenergyfromadjacentecosystems.Thislandscape

contextmodifiesthedistribution,abundanceanddiversityofanimalsandplantsacrosslandscapes

(Fahrig2001;Hodgsonetal.2011;LeesandPeres2008;Pottieretal.2009),andpotentially

improvestheecologicalfunctioningofrestoredhabitats(daSilvaetal.2015;JonesandDavidson

2016a).

Modernalgorithmsforreserveselectionexplicitlyrecognizetheimportanceofpositioningwithin

thebroaderlandscape(Hiltyetal.2006;Magrisetal.2016;Rudnicketal.2012;Weeks2017).These

algorithmshavedeveloped,now,tothepointthatsystem-specific,andspecies-specific,dataare

beingusedinassociationwithcommonlyusedmodellingtechniques(e.g.MARXAN,Zonationand

networkmodels),broadlyacrossregions,andwithinindividualsystems(Engelhardetal.2017;Kool

etal.2013;Weeks2017).Whereastheselectionoflocationsforreservesisoftenguidedbyspatial

conceptsfromlandscapeecology,includingconnectivityandlandscapecontext(Almanyetal.2009;

MargulesandPressey2000;Sarkaretal.2006),siteselectionforrestorationappearstoadoptthese

principleslessfrequently(Hodgsonetal.2011).However,spatialprioritisationisnotwidelyusedin

restoration(e.g.Adameetal.2015;Ikinetal.2016),so,therehavebeenrecentcallsfortighter

integrationofthespatialprinciplesfromlandscapeecologyandconservationbiology(Audinoetal.

2017;WiensandHobbs2015)tobothinformthedesignofrestorationprojects,andguidethe

selectionofrestorationsites(JonesandDavidson2016a;McAlpineetal.2016).Thisissurprising

becausetherehavelongbeencallsforthemorestrategicplacementofrestorationacross

landscapes,includinginsomeimportantglobalrestorationguidelinesandpolicies(Keenleysideetal.

2012;SER2008).

Thelandscapecontextofrestorationsitesislikelytobeoneofthemostimportantfactors

influencingtheoutcomesofrestorationinvestments,particularlyforprojectsthataimtoenhance

32

animalpopulations(Figure1)(MillerandHobbs2007;Moreno-Mateosetal.2012).Because

restorationsitesusuallyrequirecolonisationbyanimalpopulationsfromotherhabitatpatches,the

degreeofconnectivitybetweenrestorationsitesandotherpatchesofexistinghabitatiscritical

(Hodgsonetal.2011;Hodgsonetal.2009;Scottetal.2001).Restorationsitesmightbeplacedin

areasthatareconsideredappropriatefortheestablishmentandgrowthofhabitat-formingspecies,

butwhicharesoisolatedorpoorlyconnectedthatanimalpopulationsrespondtoalesserdegree;

therebyreducingtheoverallvaluesofthehabitatrestorationforthewholeecosystemorlandscape

(HoweandMartinez-Garza2014;JonesandDavidson2016a).Forexample,restoringtoenhance

connectivitybetweenhabitatfragmentscanhelpwithrestoringmetapopulationdynamics(Fischer

andLindenmayer2007;Montalvoetal.1997).Increasedconsiderationofmetapopulationstructure

andconnectivitycanservetoenhancetheavailablegeneticpoolofanimalpopulationsthereby

potentiallyincreasingtheirfitness(Baguetteetal.2013),reducingpopulationextinctionrisk(Reed

2004),andincreasingresiliencetoexogenousdisturbances(Etienne2004).Suchconsiderationsare

nowoftenconsideredintheplacementofmarinereserves,especiallywherefisheriesenhancement

isagoal(Gerberetal.2003;PuckettandEggleston2016).However,theregularitytowhich

landscapecontextmorebroadlyisconsideredinrestorationsiteselection,andhowthisaffects

inhabitinganimalpopulationshas,however,notbeenestablished(MetzgerandBrancalion2016).

Consequently,thelinesofevidencerequiredforpractitionerstoproperlyjustifytheintegrationof

principlesfromlandscapeecologyintothedesignofindividualrestorationprojectshavenotbeen

established.

33

Figure1Thelandscapecontextofrestorationsiteshassignificantconsequencesforthenumberandtypesof

animalswhichinhabitrestoredecosystemsonland(A),inthesea(B),andwherehydrologicconnectivity(C)is

vital,especiallyinfreshwaterecosystems,andwetlands.Forexample,isolatedpatchesofrestoredgrassland

(A.,left),mightperformlesseffectivelyforanimalsthanrestoredgrasslandsclosetonearbyalternatehabitats

(right).Inthesea(B),connectivitybetweenmultiplehabitatsisoftenanimportantconsiderationbecause

faunausemultiplehabitatsthroughouttheirlifecycle.Here,restoredoysterreefsintheNoosaRiver,Australia

intwoseascapecontexts;atadistancefromnearbymangroves(left),andveryclosetonearbymangrovesand

seagrasses(right).Hydrologicconnectivity(C)islimitedwhereweirsordamscanrestrictflowofpropagules

andanimals(left).Thesechallengescanbeovercomebyconsideringconnectivitybetweenwaterbodiesand

movementoffaunaandimplementingrestorationinterventionssuchaswaterwayremeanderingandfish

ways(right).ImagescourtesyC.Duncan,M.Lavin(CC-BY-SA2.0),N.Carson.

34

Inthisstudy,weidentifiedrestorationprojectsfromtheglobalprimaryliteratureandassessedthe

degreetowhichlandscapecontextwasusedasacriteriontohelpguidetheselectionofsitesfor

restorationwhentherestorationofassociatedanimalpopulationsisalsoaprinciplegoal,andunder

whichscenariosthisismostlikelytooccur.Wechosehabitatrestorationforanimalpopulationsas

thefocusbecauseanimalsmovethroughoutlandscape,andsospatialmetricsarelikelytobe

investigatedfirstfortheseoverotheraimsforrestoration.Weconductedliteraturesearchesusinga

structuredliteratureclassificationframework.Wereviewedtheliteratureusingmultipletermsused

forecologicalrestorationandasuiteofecosystem-andanimal-specificphrases.Wethen

categorisedprojectsintothosethatconsideredlandscapecontextinsiteselection,andthosethat

didnot,andfurthercategorisedprojectsintoenvironmentalrealms,ecosystems,andtheanimal

communitiestheyseektoenhance.Weconcludebyoutlininghowgreateruptakeoffundamental

principlesgleanedfromlandscapeecologycanimproverestorationsuccess.

MethodsLiteraturesearches

Restorationprojectswereidentifiedusingtargetedliteraturesearchesandastructuredliterature

classificationframework(Figure2).AllliteraturesearcheswereconductedusingtheISIWebof

KnowledgedatabaseinJuly2017.Theinitialsearchwasfortheterm-(("habitatrestoration"or

"ecologicalrestoration"or"ecosystemrestoration")and(faunaorinhabit*oranimal*orbiodivers*

orwildlifeorfish)),resultingin2183articlesreturnedand333articlesdownloadedforpotential

inclusion.Basedonthecompositionofthisinitialshortlist,followupsearchesincludedcomponents

oftheinitialsearch,aswellas:wetland*,seagrass,oyster*,fish*,mangrove*,forest*,grass*,

insect*,invert*,bird*, mammal*,reptil*,stream*,river*,lake*,andcoral*.Allreviewarticlesand

meta-analysesidentifiedbyliteraturesearchesweredownloadedandscannedbythefirstauthorfor

additionalarticlesthatcouldbeincluded.Thismethodofliteraturesearchingwasnotdesignedtobe

exhaustive(i.e.systematic,comprehensive)inidentifyingallrestorationprojectsintheliterature

thatalsosoughttoenhanceanimals.Itwas,however,designedtogivethebestpossible

representationofrestorationprojectsacrossmultipleenvironmentalrealmsandecosystems,and

withintheconstraintsofoursearchterms.

35

Figure2Literatureclassificationframeworkanddecisionflowchartusedtoidentifyrestorationprojectsfrom

theglobalprimaryliterature.

Inclusioncriteriaanddataextracted

Wefollowedasetclassificationframeworkforallstudiesidentifiedbyliteraturesearches(Figure2).

Restorationprojectswereclassifiedasalsotargetingtherestorationofanimalsinhabitinganimals

withinthemif:(1)thestudytestedforeffectsoftherestorationeffortforanimals,or(2)ifthe

articlestatesexplicitlythatagoaloftherestorationeffortwasalsotoincreasetheabundance

and/ordiversityofanimals.

Weconcludedthatarestorationprojectconsideredthelandscapecontextofrestorationsitesifthe

projectspecificallychosesitesorsystemsoverotherareasbecauseoftheirpositionwithinthe

landscape,especiallyiftherestorationproject:(1)aimedtoreconnecthabitatfragments;(2)

consideredthepositionofsitesrelativetopatchesofeithersimilar,ordifferent,habitattypes;or(3)

incorporatedanaspectorunderstandingofanimalmovement,dispersalcapacityorconnectivity

intositeselection.Wedidnotconsiderastudytohaveconsideredlandscapecontextifsiteswere

simplychosen:(1)onthebasisofexistingimpacts(e.g.restorationofoilspillsites,erosion

mitigation);(2)torevegetateminedland;(3)using‘habitatsuitability’modelsforgrowthofhabitat

Doesthehabitatrestorationprojectconsiderlandscape

contextinsiteselection?

Ifyes-Howwerethesitesselected,were

landscapeecologyprinciplestestedforinthestudy,andwhat

effectdidthishavefortheinhabitingfauna?

No- articleexcluded

Doesthearticledescribeanactivehabitatrestoration

effort?

No- articleexcluded

Yes

Ifno-Howwerethesites

selected?

Yes

Doesthehabitatrestorationprojectalso

seektoenhanceinhabitingfauna?

36

formingspecieswhenthepositionofsitesrelativetootherhabitatpatcheswasnotalsoconsidered;

or(4)ifsiteswererandomlyselectedusinganumbergenerator,GIS,orsimilar.

Restorationprojectsweregroupedaccordingtocountry,regionandhabitatrestored,andwepooled

studiesthatreportedonthesamerestorationprojecttoavoidoverlapincountingprojects.Wealso

identifiedthetypesofanimals(inbroadgroups;e.g.birds,mammals,fish,insects,orany

combinationofthese)thateachrestorationprojectsoughttoenhance,basedonstatedoutcomes.

Forprojectsthatdidincorporatelandscapecontextintotheirsiteselection,weidentifiedwhether

therestorationeffortresultedinthestatedaimsbeingachievedforthatspecies(i.e.‘positive’

effectsonfauna).

Studieswereclassifiedintoenvironmentalrealms(terrestrial,marine,orfreshwater)accordingto

theirlocationandtheanimalspeciesbeingrestored.Forexample,wetlandscanbeeither

freshwater,marineorterrestrial;arestoredwetlandinsaltwaterseekingtorestorehabitatforfish

isclassifiedasmarine,whereasawetlandinsaltwaterenhancingbirdsorinsectsisclassifiedas

terrestrial.

ResultsDistributionandthematicfocusofidentifiedrestorationprojects

Thefinaldatabaseincluded472restorationprojectsdescribedinEnglishlanguagejournalsthat

soughttoenhanceanimalpopulationsfromrestorationefforts(Forthefulllistofincludedarticles

contacttheUSCteam).Weidentifiedrestorationprojectsin34countries;mostprojectswerefrom

theUSA(n=212,45%),followedbyAustralia(n=44,9%),Finland(n=20,4%),Sweden(n=14,3%),

andtheUnitedKingdom(n=14,3%).Themostcommonecosystemsrestoredwereforests(n=134;

28%),followedbystreams(n=87;18%),andgrasslands(n=67;14%).Overall,37%ofstudiesstated

explicitlythatthelandscapecontextofrestorationsiteswasimportantindeterminingthe

effectivenessofrestorationforanimals,irrespectiveofwhetherornottheyactuallyimplemented

landscapeconceptsintositeselection.

Terrestrialprojectscomprisedjustoverhalfofourdataset(n=243,52%)(Figure3).Forests(n=36,

15%ofterrestrialprojects)andgrasslands(n=34,14%)intheUSA,andforestsinAustralia(n=30,

12%)weretheecosystemsmostoftentargetedforrestoration(Figure4).Terrestrialrestoration

projectsmostoftensoughttoenhancepopulationsofbirds(n=107,44%),insects(n=53,22%)and

mammals(n=23;9%).Freshwaterprojectscomprisedroughlyonethirdofourdataset(n=151,

37

32%)(Figure3).StreamsandriversintheUSA(n=47,31%offreshwaterprojects)andFinland(n=

11,7%)weretheecosystemsmostoftentargetedforrestoration(Figure5).Freshwaterrestoration

projectsmostoftensoughttoenhancepopulationsoffish(n=89,59%)andmacroinvertebrates

(n=60,40%).Marineprojectscomprisedroughlyonesixthofourdataset(n=78,17%)(Figure3).

Oysterreefs(n=30,44%ofmarineprojects)andsaltmarshes(n=10,13%)intheUSAwerethe

ecosystemsmostoftentargetedforrestoration(Figure6).Marinerestorationprojectsmostoften

soughttoenhancepopulationsoffish(n=63,81%),nektoniccrustaceans(n=21,27%)or

macroinvertebrateinfauna(n=20,26%).

Effectsoflandscapecontextinrestoration

Acrossallrealms,11%ofprojects(54of472)includedlandscapecontextasacriterioninthe

selectionofrestorationsites(Table1,Figure3).Theintegrationoflandscapecontextinrestoration

projectswasremarkablysimilaracrossrealms:terrestrial(10%of243projects),marine(13%of89)

andfreshwater(13%of164).Interrestrialecosystems,projectsthatusedlandscapecontextmost

oftenwerethoseinpeatlands(100%ofpeatlandprojects,butonlyonewasidentified),wetlands

(14%ofwetlandprojects)andforests(10%)(Figure4B),andthosethattargetedamphibians(29%)

andmammals(13%).Freshwaterprojectsthatconsideredlandscapecontextmostoftenwerethose

inrivers(40%)andstreams(7%)(Figure5B),andthosetargetingamphibians(40%)and

macroinvertebrateinfauna(13%).Inmarinesystems,theconsiderationoflandscapecontextwas

mostprevalentinmarinewetlands(30%)andinseagrass(29%)(Figure6B),andwhenrestoration

targetedbirds(33%)ormacroinvertebrateinfauna(15%).

38

Table1.Summaryofrestorationprojectsinterrestrial,freshwaterandmarinerealmsfor(A)allrestorationstudies;(B)restorationstudiesthatconsideredlandscape

contextinthedesignphase;and(C)restorationstudiesthatdidnotconsiderlandscapecontextinthedesignphase.

Terrestrial Freshwater Marine Allrealms

%ofprojects #ofprojects %ofprojects #ofprojects %ofprojects #ofprojects %ofprojects #ofprojects

A.Allrestorationstudies 51 243 32 151 17 78 100 472

B.Landscapecontextconsideredindesign 10 25 13 19 13 10 11 54

Positiveeffects 7 18 4 6 8 6 6 30

Noeffectsreported 4 9 7 11 5 4 5 24

Landscapeecologyconceptused

Connectivitywithpatchesofsimilarhabitat 8 20 8 12 6 5 8 37

Connectivitywithotherhabitats 1 1 1 1 1 1 1 3

Hydrology 1 2 2 3 4 3 2 8

Other 1 2 2 3 1 1 1 6

C.LandscapecontextNOTconsideredindesign 46 218 28 132 14 68 88 417

Reasonforexclusion

Locatedonaminesite 7 15 0 0 0 0 4 15

Minimiseerosioneffects 0 0 3 4 0 0 1 4

Randomplacement 5 10 2 2 0 0 3 12

Indiscernible 81 177 86 113 85 58 83 348

Other 7 16 10 13 15 10 9 39

39

Figure3Summaryofincludedstudies.Numbersinwhitecirclesinthecentreofeachchartdenotethe

numberofstudiesineachcategory.A)thetotalnumberofstudiesincludedinthedatabase,divided

byenvironmentalrealm.B)Theproportionofstudieswhichdid,anddidnot,considerthespatial

contextofrestorationsitesduringecosystemrestorationprojects.C)Theproportionofstudiesthat

didconsiderlandscapecontextduringtherestorationdesignprocessthateitherdid,ordidnot,

reportthedirectionofoutcomesforanimals.

Positiveeffects Noeffectsreported

B.Landscapecontextconsideredindesign

C.LandscapecontextNOTconsideredindesign

Terrestr

ialFres

hwater

Marine

Positiveeffects Noeffectsreported Positiveeffects Noeffectsreported

Terrestrial

Freshwater

Marine

Positive effects Positive

effects

Positive effects

Effects notreported Effects not

reportedEffects not

reported

Terrestrial Marine Freshwater

B.Landscapecontextconsideredindesign

C.LandscapecontextNOTconsideredindesign

B.Landscapecontextconsideredindesign

C.LandscapecontextNOTconsideredindesign

Context NOT considered in design

Context NOT considered in design

Context NOT considered in design

A.Numberofstudies

B.Studiesconsideringlandscapecontext

C.Effectsofconsideringlandscapecontextforanimals

Context IS considered in design

Context IS considered in design

Context IS considered in design

472

243 151 78

25 19 10

40

Figure4(A)Globaldistributionofterrestrialrestorationprojectsthataimedtorestoreanimal

populations,assemblages,ordiversity.(B)Ofthe243terrestrialrestorationprojects,only25(10%)

consideredlandscapecontextintheirdesignphase(Table1).Theintegrationoflandscapeattributes

did,however,differamongterrestrialecosystems.Numbersabovebarsindicatethenumberof

projectsthatconsideredlandscapecontextinsiteselection.

0

10

20

30

40

50

60

70

80

90

100

Coastal Dunes Forest Grassland Marsh Peatlands Wetland

Perc

enta

ge o

f stu

dies

Habitats

Consider spatial ecology Do not consider spatial ecology

Coastal dunes (n=1)

Marsh (n=2)

Wetland (n=37)Forest (n=134)

Peatlands (n=1)

1 study

3 studies9 studies

Grassland (n=67)

A.

B.1

13 5600

41

Figure5(A)Globaldistributionoffreshwaterrestorationprojectsthataimedtorestoreanimal

populations,assemblages,ordiversity.(B)Ofthe151freshwaterrestorationprojects,only19(13%)

consideredlandscapecontextintheirdesignphase(Table1).Theintegrationoflandscapeattributes

did,however,differamongfreshwaterecosystems.Numbersabovebarsindicatethenumberof

projectsthatconsideredlandscapecontextinsiteselection.

0

10

20

30

40

50

60

70

80

90

100

Lake River Spring Stream Wetland

Perc

enta

ge o

f stu

dies

Habitats

Consider spatial ecology Do not consider spatial ecology

Lake (n=6)

Stream (n=87)

Spring (n=1)River (n=35)

Wetland (n=22)

1 study

3 studies9 studies

A.

B.

8 5

60 0

42

Figure6(A)Globaldistributionofmarinerestorationprojectsthataimedtorestoreanimal

populations,assemblages,ordiversity.(B)Ofthe78marinerestorationprojects,only10(13%)

consideredlandscapecontextintheirdesignphase(Table1).Theintegrationoflandscapeattributes

did,however,differamongmarineecosystems.Numbersabovebarsindicatethenumberofprojects

thatconsideredlandscapecontextinsiteselection.

0

10

20

30

40

50

60

70

80

90

100

Coral reef Boulder fields Macroalgae Mangroves Saltmarsh Oysters Estuary Seagrass Wetland

Perc

enta

ge o

f stu

dies

Habitats

Consider spatial ecology Do not consider spatial ecology

Coral reefs (n=3)Mangroves (n=9)Estuary (n=1)

Boulder fields (n=1)Saltmarsh (n=12)Seagrass (n=7)

Kelp (n=1)

Oysters (n=34)Wetland (n=10)

1 study

3 studies9 studies

A.

B.

1 1 3

32

0 0 0 0

43

Therewasgoodevidencetosuggestthatincorporatinglandscapecontextintotheselection

ofrestorationsitesresultedinpositiveoutcomesforanimals,with56%ofprojects(30of54

projects)thatdidconsiderlandscapecontextreportingpositiveoutcomesforanimals(Table

1,Figure3).Theremaining44%(24of54)ofprojectsdidnotexplicitlytestwhether

landscapecontextaffectedanimalsdirectly.Projectsthatreportedpositiveeffectsof

landscapecontextonanimalsweremorecommononland(72%of25projects)andinthe

sea(60%of10),thaninfreshwater(32%of19).

Mostincludedstudiesfocussedontheecologicaleffectsofconnectivity(74%of54projects),

eitherbetweenpatchesofthesamehabitat,orwithotherhabitats.Projectsthatreported

positiveeffectsofconnectivityonanimalswerecommonacrossterrestrial(84%of25

projects),marine(60%of10),andfreshwater(68%of19)landscapes.Themajorityofthese

(69%of54projects)consideredconnectivitybetweenpatchesofsimilarhabitats,andfewer

consideredtheeffectsofconnectivitywithadjacentalternatehabitats(3%of54projects,

andoneineachenvironmentalrealm;Table1).Thepotentialecologicaleffectsofother

landscapeconceptshave,however,rarelybeentestedwithempiricaldata(26%of54

projects),withmostresearchlimitedtotheeffectsofhydrologyinaquaticecosystems

(Table1).

Inmostprojects(89%of472projects),landscapecontextwasnotlistedasacriterioninthe

selectionofrestorationsites(Table1,Figure3).Whererestorationstudiesdidnotconsider

thespatialattributesoflandscapes,projectsweremostoftenatminesites(15projects,6%

ofterrestrialprojects),locatedtominimisetheeffectsoferosiononfreshwaterecosystems

(4projects,3%offreshwaterprojects),orplacedrandomlyinlandscapes(12projects,4%of

terrestrialprojectsand1%offreshwaterprojects).Formoststudies,however,wewerenot

abletoidentifyhowrestorationsiteswereselectedfromthedescriptionsprovided(74%of

472projects).Thiswasacommontrendacrossterrestrial(177,73%ofallterrestrial

projects),marine(58,75%ofallmarineprojects)andfreshwater(113,75%ofallfreshwater

projects)realms.

Thefirstrestorationstudiesthatexplicitlyconsideredlandscapecontextwhenselecting

restorationsiteswerepublishedin1996forfreshwaterecosystems,2001forterrestrial

ecosystems,and2004formarineecosystems(Figure7A).Theproportionofstudiesthat

consideredlandscapecontexthasremainedhighlyvariablewithnocleartrendbetween

44

yearsoverthepastthreedecades(Figure7A).Bycontrast,therehasbeenasharpincrease

intheintegrationofspatialconceptsfromlandscapeecologyintothewiderfieldsofbiology

andecologyduringthisperiod(Figure7B).

Figure7Summaryofpublishedstudiesthathaveconsideredlandscapecontextin:(A)restoration

(separatedforterrestrial,marineandfreshwaterrealms);and(B)acrossthefieldsofbiologyand

ecology.DataextractedfromISIWebofKnowledgethekeywordsrestoration("habitatrestoration"

or"ecologicalrestoration"or"ecosystemrestoration"),“biology”or“ecology”,and“landscape

ecology”or“spatialecology”.2017ispartyearonly.

0

10

20

30

40

50

60

70

80

90

PErc

enta

ge o

f stu

dies

Year

Terrestrial Marine Freshwater

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

Perc

enta

ge o

f stu

dies

Year

A.

B.

45

Discussion

Theecologicaloutcomesofrestorationprojectscan,inmanycases,beimprovedbyplacing

sitesatlocationsinlandscapestomaximisetherecruitmentofindividualstonewlycreated

habitats(JonesandDavidson2016a).Whereasitiswidelyappreciatedthatthelandscape

contextofrestoredhabitatscanshapethesuccessofrestorationprojects,theprinciplesof

landscapeecologyremainrarelyconsideredinrestorationdecisions.Inthisstudy,weshow

thatamarkeddiscrepancyexistsbetweenthestatedimportanceoflandscapecontextfor

restoration(37%ofstudiesreviewed),andtheextenttowhichthespatialpropertiesof

landscapesareconsideredinrestorationprojects(12%ofstudiesreviewed).Furthermore,

wefoundnoevidencethattheconsiderationoflandscapecontextinrestorationhas

increasedoverthepastthreedecades.Thisissurprisingasspatialconceptshavebeenmore

widelyadoptedinconservationduringthesameperiod,andthereareimportantglobal

restorationpolicydocumentswhichadvocateforitsinclusion(Keenleysideetal.2012;SER

2008).Thisresultcouldeventuatebecauseofunderreportinginthedescriptionof

restorationsites,restrictingouruseofonlytheprimary.However,peer-reviewedliterature

offerssomesafeguardagainstsubstantialunder-reportingofthedesignprocess.

Wefoundnoclearevidenceastowhytheuptakeofspatialecologicalprinciplesvaried

acrossdifferenthabitats;likelybecausetheuptakewasconsistentlylowacrossallhabitats.

Theexceptionswerehabitatswherefewrestorationprojectswereidentified,buthadvery

highuptakeoflandscapecontextprinciples(e.g.peatlands);however,littlecanbegleaned

fromthesefewexamplesfromhabitatswithfewerrestorationprojects.Therewerealsono

cleartrendsinhabitatswhichhavebeenstudiedextensivelyintermsoflandscapecontext.

Forexample,thestudyoflandscapecontextoncoralreefsishighlyadvanced(Pittmanand

Olds2015),butnocoralrestorationprojectshaveyetconsidereditinrestorationsite

selection.

Weidentifiedseveralkeytypesofspatialmetricsthatareconsideredintheselectionof

restorationsites.Predominantly,studiesassessedconnectivitywithsimilarpatchesof

habitats(e.g.Angelierietal.2016;Derheetal.2016);unsurprisinggiventhattheroleof

restorationisoftentoenhanceorreconnectthreatenedhabitats,soconsideringnearby

patchesofthishabitatisusuallyimportant.Conceptssuchashydrologicconnectivityand

connectivitywithalternativehabitats(i.e.thoseotherthanthehabitatbeingrestored)were,

howeverconsideredanorderofmagnitudelessthanconnectivitywithsimilarhabitat

46

patches.Thisrelativelylownumberofstudiesfocusingonhydrologicconnectivityis

surprisingbecausereconnectingwaterbodieshydrologicallyisanimportantfocusin

wetlands,andaquaticecosystemsglobally(JacksonandPringle2010;Kondolfetal.2006).

Similarly,conceptsregardingthespreadofpropagulesbywindorothermechanisms

receivedlittleattentioninthestudiesweidentified.Thismeansthatthereareseveralkey

conceptsandmetricswithinspatialecologythathaveyettobeproperlystudiedin

restoration,butthatarelikelytobehelpfulinplacingrestorationefforts.Broadeningthe

scopeofthetypesofconsiderationsandmetricsusedtoplacerestorationshouldbeakey

focusofrestorationresearchersandpractitionersalike.

Mostprojectsthatincorporatedconceptsfromlandscapeecologyintotheirdesignandsite

selectionreportedpositiveoutcomesforanimals,albeitwithsomevariationbetween

terrestrial(72%positiveoutcomes),marine(60%)andfreshwater(31%)realms.For

example,infreshwaterecosystems,wetlandrestorationsforamphibiansreportedmarked

increasesinadultbreedingpopulationswithinthreeyearsofrestorationofpondsthatwere

highlyconnectedtoeachother(Petrankaetal.2003);aneffectthatisinconsistentforother

similarfreshwaterrestorationprojectswherecontextwasnotconsidered(Shulseetal.

2012).Thesesortsofconsiderationsofadult,breedingmetapopulationswererare,butare

likelyinstructivefortheenhancementofanimalpopulationsacrosslandscapes(McAlpineet

al.2016;Montalvoetal.1997).Somecounterintuitiveresultswereuncoveredforoyster

reefs.Higherconnectivity(inthiscase,simplyproximity)withadjacenthabitatsisusually

viewedasbeneficialforcoastalmarineorganisms(Oldsetal.2018c).However,restored

reefsinNorthAmericacontainedhigherfishabundancewhenfurtherfromexistingmarshes

becausetheyprovidednewcomplexhabitatsonpreviouslylowcomplexitymuddyareas

(Grabowskietal.2005).Conversely,higherconnectivitybetweenextantreefsandrestored

reefswasviewedasapositiveinfluenceoninhabitingfaunainotherstudies(Gregalisetal.

2009).Althoughnostudieshaveassessedhowprojectsorsitesthatdidincorporatespatial

contextintotheirdesign,versusthosethatdidn’t(e.g.randomly,orforsomeother

ecologicalreasoning)performedrelativetoeachother,thesefindingssuggestthatthe

spatialpropertiesoflandscapesmighthavebroad,andlargelyunrecognised,effectsonthe

successofecologicalrestorationprojects.

Ecologicalrestorationisconductedformanypurposes,andnotalltypesofrestorationare

designedtobenefitanimals.Forexample,manyrestorationprojectsseektorestorewhole

47

ecosystems(e.g.wholeoflakerestoration),toreverseaparticulartypeofimpact(e.g.mine

siterehabilitation,oilspillremediation)(e.g.BradyandNoske2010),ortolimitthe

ecologicaleffectsofongoingdisturbances,suchaserosion,sedimentation,nutrient

enrichmentorpollution(e.g.restoringoysterreefstoreduceshorelineretreat,orforeststo

reducegullyerosion)(e.g.Piazzaetal.2005).Underthesecircumstances,restorationsites

mightbeplacedinareasthatareconsideredappropriatefortheestablishmentofhabitat-

formingspecies,butwhichmightbeisolatedfromotherremnantpatchesofthesame

habitatthatserveassourcesfortherecruitmentofanimals.Becausethelocationsofthese

typesofprojectsareoftenfixed,itmightnotalwaysbepossibleforrestorationdecisionsto

beplacedinalandscapecontext.Alternatively,restorationsitesmightbeplacedinareasto

limitthepotentialforconflictwithotherhumanuses(e.g.fishing,farming,recreationaluse,

transport),orlegislation(e.g.mooringareas,otherformsofconservation)(Presseyand

Bottrill2008),andthesemightnotbeoptimalfortherestorationofecosystems,ortheir

inhabitinganimals(Fahrig2001;Pottieretal.2009).Bothapproachesmightresultin

restorationsitesbeingunintentionallyrestrictedtolocationsthatprovidepoorhabitat

valuesforanimals(cf.‘residual’conservationareas)(PresseyandBottrill2008;Presseyetal.

2000),andwhichthereforehavethepotentialtoyieldsuboptimalrestorationreturnsfor

animals.

Thereareseveralkeyresearchareasthatshouldbepromotedtoassistinbetterprioritising

restorationacrosslandscapes.Weraiseinthisreviewapaucityofrestorationprojects

incorporatingsystematicconservationplanningregimensintheselectionofrestoration

sites.Studiesshouldbeconductedtodeterminethevalidityofusingsystematic

conservationpractices(MargulesandPressey2000)forrestorationacrossenvironmental

realms(researchquestion1;Table2).Despitethisreviewuncoveringadiversityof

ecosystemsandanimalsthathavebeenthefocusofrestoration,thereareseveral

ecosystems(researchquestion2;Table2)andanimalgroups(researchquestion3;Table2)

thatremainunderrepresentedwithintheliterature.Forexample,veryfewstudieshave

assessedthecapacityforcoastaldunestoberestoredforanimals,especiallybirds,despite

beachecosystemsbeingregularlyrestoredforthesepurposes(Masloetal.2012;Masloet

al.2011).Todate,mostworkontheeffectsoflandscapecontextontheoutcomesof

restorationhavefocusedspecificallyonhowthemetricofconnectivity(especiallyproximity

andisolation)betweenpatchesofsimilarhabitataffectstheabundanceofanimals.

Restorationsitesarealwayspositionedwithinheterogeneouslandscapesofmultiplehabitat

48

types,sofurtherresearchshouldbeconductedonhowconnectivitywithnearbypatchesof

differenthabitattypesaffectsthesuccessofrestorationmorebroadly(researchquestion4;

Table2),aswellastheconditionofanimalpopulationsthemselves,wherethisisakey

restorationgoal(researchquestion5;Table2).Similarly,thereareseverallandscape

metrics,beyondsimpledistance-basedconnectivitymetrics(especiallyEuclideandistance),

thatmightassistindevelopingmorerobustandrepresentativemodelsforincorporating

landscapesecologyintorestoration.Landscapeecologistshavedevelopedseveral

multivariateormultimetricvariablesthatcanbeusedtodescribethecomplexityof

landscapesforanimals(e.g.McGarigaletal.2012).Theefficacyofthesemetricsin

restorationshouldbefurtherinvestigatedfordifferentecosystemsandtargetanimals

(researchquestion6;Table2).Severalauthorshavediscussedthevalidityoflandscape

metrics,andconceptsacrossenvironmentalrealms(Pittmanetal.2018).Forexample,there

issomeevidencetosuggestthatthebestmetricstodescribespatialpatternsinanimal

abundancemightdifferbetweenthelandandtheseaduetothemovementofwaterbodies

inaquaticecosystems(Pittmanetal.2018).Establishingthevalidityofthesemetricsacross

environmentalrealmswillassistingeneralisingrestorationplanningregimensacrossrealms

(researchquestion7;Table2).Collectively,answeringthesepriorityresearchquestionswill

assistinestablishingmoreeffectiveregimensforsystematiclandscaperestorationacrossall

environmentalrealms,andprovidetheevidencethatmanagersneedtosupporttheir

decisionmaking-processes.

49

Table2.Listofpriorityresearchquestionsforintegratinglandscapeprinciplesintorestoration

ecology.

PriorityResearchQuestions

1. Aretheprinciplesofsystematicconservationplanning(MargulesandPressey2000)directly

translatabletoallrestorationprojects,andifnot,whichprinciplesneedtobetailoredspecificallyfor

restoration?

2. Forwhichhabitatsdoeslandscapecontextmattermost,andwhichmetrics,andwhichconnections,

shouldbeprioritisedforindividualhabitattypes?

3. Forwhichanimalgroupsdoeslandscapecontextmattermost,andwhichmetrics,andwhich

connections,shouldbeprioritisedforindividualanimalgroups?

4. Towhatdegreedoesconsideringconnectivitywithotherhabitats(i.e.inter-habitatconnectivity)

affecttheoutcomesofrestorationprojects(e.g.Unsworthetal.2008)?Todate,mostresearchand

focusintheliteraturehasbeenonconnectivitybetweenpatchesofthesamehabitat(i.e.intra-

habitatconnectivity).

5. Towhatdegreecanincorporatingmetapopulationdynamicsandconnectivityintorestoration

planningenhanceanimalpopulationfitness,persistence,andresilience?

6. Aretheremorethorough,orbetter,metricsthatcanbeusedtooptimisethespatialplacementof

restoration(e.g.McGarigaletal.2012)?Todate,mostworkonthespatialmetricsthatinfluence

restorationhasfocusedonconnectivity,especiallyproximity(Euclidean)betweensimilarhabitat

patches.

7. Isthereconsistencyintheefficacyofspatialmetricsacrossenvironmentalrealms?(e.g.Pittmanetal.

2018)

50

Systematicconservationplanninghasmadegreatprogressinusingadiversityoflandscape

characteristicsandsophisticatedalgorithmstoguidethedesignofprotectedareasand

reservenetworks(MargulesandPressey2000;Moilanenetal.2009).Theprinciplesof

landscapeecology,andthetechniquesofconservationplanning,mightbeusefulinthe

designofrestorationareas,buthavenotbeenwidelyappliedtointroduceacriticalspatial

elementtorestorationplanning(Hodgsonetal.2011).Whenselectingsitesforanetworkof

reserves,theconservationplanningprocessstartswithabroad,landscape-scale

perspective.Ecosystems,habitatsandlocationsareselectedforprotectiontomaximise

conservationbenefitsacrosstheentirelandscape,andsomepotentialsitesarethen

eliminatedbecauseofotherconsideration(e.g.extractiveindustries,tourism)(Wattsetal.

2009).Bycontrast,restorationprojectsoftenstartwithanarrowerperspectivefocusingon

thespecificecosystemorhabitattoberestored.Sitesarethenselectedbasedontheir

suitabilitytosupporttheparticularecosystemofinterest,withlittleconsiderationgivento

theattributesoflandscapesbeyondtherestorationsite.Incorporatingthelessonslearned

fromsystematicconservationplanning(i.e.goalsetting,data-basedfeedbacks,and

improvements)andtheprinciplesoflandscapeecology(i.e.theplacementofsitesin

heterogeneousland-andseascapes)intorestorationshould,therefore,leadtosignificant

improvementsinthedesign,placementandecologicaleffectivenessofrestorationprojects.

Weadvocateforalandscape-scaleapproachtorestoration,andsuggestthatspatial

restorationecologyshouldstartwiththeidentificationofecosystemsandhabitatsthatare

inthegreatestneedofrestoration(basedonthebesthistoricalinformationavailable).Sites

forrestorationshouldthenbeselectedfromallsuitablelocationswithinthelandscapeof

interesttomaximisetheirpotentialecologicalbenefitsfortheecosystemsthemselves,the

animalsandfoodwebsthesesupport,andtheecologicalfunctionsandecosystemservices

theyprovide.Thisspatialapproachtorestorationecologyshould,therefore,helptobroaden

boththescopeandperceivedecologicalbenefitsofmanyrestorationprojects,andmight

alsoimprovereturnsoninvestmentacrossrestoredlandscapes.

51

Chapter3

Maximisingthebenefitsofoysterreefrestoration

forfinfishandtheirfisheries

GilbyBL,OldsAD,PetersonCH,ConnollyRM,VossCM,BishopMJ,Elliott

M,GrabowskiJH,OrtodossiNL,SchlacherTA(2018)Maximizingthe

benefitsofoysterreefrestorationforfinfishandtheirfisheries.Fishand

Fisheries19:931-947

52

Introduction

Globally,85%ofoysterreefshavebeenlost,mainlyduetooverharvesting,disease,and

poorwaterquality(Becketal.2011).Lostoysterreefsareoftenreplacedwithecosystems

thatprovidelowerhabitatvaluesforfish(i.e.lessfood,orpoorerprotectionfrom

predators),suchasbaresediments(Grabowskietal.2012).Thesechangescanbeassociated

withreductionsinfishdiversity,biomassandabundance,andwithdeclinesinthelandings

ofrecreationalandcommercialfishes(Coenetal.1999;Petersonetal.2003).Asa

consequence,therestorationofoysterreefsasfishhabitats,andthereforetoenhancefish

and/orfisheriesisoften,butnotalways,akeyaimofoysterrestorationprojects(Coenand

Luckenbach2000;zuErmgassenetal.2016).

Ecologicalrestorationofoysterreefsforfinfishandtheirfisheriesisanimportant

componentofmanycoastalmanagementandenhancementschemes(Baggettetal.2015;

Creightonetal.2015;HumphriesandLaPeyre2015).Oysterreefrestorationencompasses

bothcategoriesofecoengineering;TypeA,therestorationofhabitatsthusallowingthe

desiredspeciestocoloniseorexpand,andTypeB,whichinvolvesthedirectincreaseofa

species,suchasthroughrestockingorreplanting(Elliottetal.2016).Globally,46studies

detailoysterrestorationprojectswhichseektoenhancefinfishand/ortheirfisheriesaround

reefs(asidentifiedfromtheISIWebofKnowledge;Figure1,TableS1).Restoringoyster

reefscanaugmentfishbiomassbyupto260g/m2ofrestoredreefperyear(Petersonetal.

2003).Theseeffectsofrestorationonfishpopulationsresultfromenhancedlarval

settlementandsurvival(Breitburgetal.1995),andtheimmigrationofadultfishtoreefs,

wheretheyfeedandtakeshelter(Gittmanetal.2016;Harwelletal.2011).Byenhancing

fishbiomass,restoredoysterreefscanconveyeconomicbenefitsofuptoUS$4123ha-1yr-1

forlocalcommercialfisheries(Grabowskietal.2012).Whilerestoredoysterreefsare

relativelycommoninNorthAmerica(87%ofstudieshavebeenconductedontheAtlanticor

GulfcoastsoftheUSA;Figure1),oysterreefrestorationprojectsfocusingonfish

enhancementareonlyjustgaininginterestandmomentuminAustralia(Gilliesetal.2015b),

Europe(Farinas-FrancoandRoberts2014)andAsia(Quanetal.2009).Whilstenhancement

offishpopulationsisjustoneofthemultipleecosystemservicesprovidedbyrestoredoyster

reefs,andisnotalwaysthekeydriverofrestoration(e.g.intertidaloysterreefrestoration

maybetostabiliseshorelinesandprotectthemagainsterosion)(Grabowskietal.2012),

win-winscenarios(fortheecologyandtheeconomy)maybecreatedif,irrespectiveofthe

keygoal,reefsarealsodesignedtoenhancefish.Byexpandingthegoalsofoysterreef

53

restorationtoincludefishandfisheries,wemight,therefore,alsoenhancetheeconomic,

socialandculturalvaluesassociatedwithrestorationefforts,andmaximisestakeholder

engagement,bothlocallyandglobally(LaPeyreetal.2012).

Figure1Globaldistributionsof(A)studiesassessingtheeffectsofoysterreefrestorationforfish(as

identifiedfromISIWebofKnowledgesearchfor(oyster*andfishandrestor*)),and(B)theglobal

distributionofkeyhabitatswithwhichoysterreefshaveafunctionallinkage.Informationonthe

extentandgeographicdistributionofecosystemtypessourcedfromtheUnitedNationsEnvironment

ProgrammeWorldConservationMonitoringCentre(UNEP-WCMC).

54

Whileitiswellestablishedthatrestoringoysterreefscanaugmentfishbiomassand

enhancefinfishfisheries,thepublishedliteratureonoysterreefrestorationpoorly

integratesseveralimportantecologicalconceptsthatshapefishpopulationsincoastal

waters.Forexample,itiswidelyacceptedthattheextentofkeycoastalnurseryhabitats,

suchasseagrass,marshes,andmangroves(Nagelkerkenetal.2015),andthedegreeto

whichthesehabitatsareconnected(Oldsetal.2016;Pittmanetal.2011),aresignificant

determinantsoffishassemblagesincoastalseascapes.Therefore,futureoysterreef

restorationprojectscanbuildonthefirstgenerationofoysterreefprojects,whichhave

demonstratedthatreefsaugmenttheproductivityofanumberoffinfishspecies,by

adoptingalandscape-scaleapproachthataccountsforhowreefsinteractwithother

structuredhabitats(Bostrometal.2011).Advancesinourunderstandingofhowmanaging

catchmentwatersupply,sedimentandnutrientrunoffaffectscoastalecosystems(Gorman

etal.2009;Kleinetal.2012),andmoreeffectivefisheriesrestrictions(Gilbyetal.2017b)

canalsoimprovetheecologicalconditionofcoastalhabitatsandmodifythecompositionof

fishassemblages.Anincreasingliteratureonecohydrologywithecoengineeringprinciples

areprovidingcase-studiesofsuccessandfailure(Elliottetal.2016).However,theextentto

whichtheseinterventionsaffecttheoutcomesofoysterreefrestorationprojectsforfishis

unclear.Thecapacityofoysterrestorationprojectstopromotefishbiomassandenhance

fisheriescan,therefore,beimprovedbybetterincorporatingmodernconceptsinfish

ecologyandfisheriesmanagementintothedesignofreefprojects.

Inthispaper,wereviewtherelevantliteraturetobetterintegratethefieldsoffishhabitat

research,seascapeecology,andcoastalmanagementintooysterreefrestorationprojects.

First,weoutlinethreeimportantprerequisitesthatmustbeestablishedforoysterreef

restorationprojectsthatseektoenhancefishpopulationsandfisheries(Figure2).Wethen

introducefourkeyconcepts(Figure3)thatwillimproverestorationoutcomesforfishand

fisheries:(1)viewoysterreefsasfishhabitats;(2)recognisethatoysterreefsarepartofa

widerseascapethatincludesotherfishhabitats;(3)considertheimpactofother

managementinterventions(e.g.fishingrestrictions,catchmentrunoffreductions),and(4)

monitortheeffectsofrestorationforbothoystersandfishacrosstheentireseascape,and

implementchangestorestorationplanswherenecessary.Theoverarchingintentistofine-

tunethedesign,placement,andmanagementofrestoredoysterreefstominimisetheir

economiccostsandmaximisetheirecologicalbenefitsforbothoysterreefsandfinfish

fisheries.

55

Figure2Therearethreeimportantprerequisitesforoysterreefrestorationprojectsthatseekto

enhancefinfishandtheirfisheries:(A)evidenceforthehistoricaloccurrenceofoysterreefsinthe

targetregion;(B)whetherthedetrimentaleffectsofhumanactivitiesthatcausedthelossofoyster

reefshavebeencontrolledandmodernconditionsaresuitableforreefrestoration;and(C)iflost

reefswerereplacedbyhabitats(e.g.baresediments)thatsupportsignificantlyfewerfishthanoyster

reefs.ImagescourtesyUSFishandWildlifeService,Kaensu(Flickr)(CCBY2.0),andD.Schwen(CCBY

3.0).

A.Historicaloysteroccurrence

B.Causesofreefdeclinearrested,modernconditionsappropriate

C.Newnon-reefhabitatssupport fewerfish

PriortoRestoration FollowingRestoration

56

Figure3Keyconceptstoimproverestorationoutcomesforfishandfisheries.(A)Oysterreefsmustbe

healthyandcomplexbiogenichabitatsthatarepositionedwithinthephysicochemicalnicheofboth

oystersandfish.(B)Oysterreefrestorationprojectsshouldbepositionedtoimproveconnectivity

(betweenreefsandamongreefsandotherhabitats),promotefishrecruitmentandenhancethe

nurseryfunctionofcoastalseascapesforfish.(C)Otherimpactingprocesses(e.g.fishing,

sedimentation,eutrophication)mustbemanagedtolimittheireffectsontheperformanceof

restoredoysterreefs.(D)Managementdecisionsmustbeadaptable,respondingtokeycriteriafrom

thoroughmonitoringprogramsthatarespecificallydesignedtoascertainthehealthanddevelopment

ofthereefs,anddetecttheeffectsofoysterreefsforfishbothonreefs,andinthesurrounding

seascape.

Prerequisitesforrestoringoysterreefsforfinfishandtheirfisheries

Addingoysterreefstosystemswheretheydidnotoccurhistorically(i.e.atanecosystemor

embayment-widescale)canbeviewedastheartificialmodificationofcoastalseascapes.

Theabsenceofoysterreefsfromwhereareaswheretheyaretoberestoredmightbe

becausetheareaishaseitherunsuitablesubstratumorwaterquality(includingturbidity

andsalinity)ortherenotbeinganyspatsupply.Inthecaseofadegradedsystem,e.g.where

historicalpopulationsarenowabsent,failingtoremedythehumanactivitiesthatledtothe

degradation,orloss,ofoysterreefsorpreventtheirrecovery(e.g.over-harvesting,poor

waterquality)willseverelylimitthesuccessofanyrestorationproject.Insomeinstances,

oysterreefsmayhavebeenreplacedbyotherstructurallycomplexecosystems(e.g.

seagrass,kelps,ormangroves)thatalsoprovideimportanthabitatforfish;restoringoyster

reefsintheselocationsmightnotresultinoverallnetimprovementsinfishorfisheries

(Grabowskietal.2005).Consequently,therearethreeimportantconsiderations(Figure2)

foroysterreefrestorationprojectsthatseektoenhancefishandfisheries:

A)evidenceforthehistoricaloccurrenceofoysterreefsinthetargetregion;

C.Fisheries&CatchmentManagementFisheriesmanacrossagementCatchmentmanagement

B.TheSeascapeContextConnectivitywithotherhabitats

ConnectivitywithotheroysterreefsEnhancing fishrecruitmentandnurseryvalue

A.OysterReefsasFishHabitatEnvironmentalvariables

Reefproperties

D.MonitoringandAdaptabilityOystersandReefs

Fishacrosstheseascape

57

B)causesofoysterreefdeclinemustbereversed,andmoderndayconditionsare

suitableforoystergrowth;and,

C)newnon-reefhabitatssupportfewerfishthanoysterreefs.

Keyconceptstoimproverestorationoutcomesforfishandfisheries

1.Viewoysterreefsasfishhabitats

Theenvironmentalconditionsunderwhichoysterreefrestorationismostsuccessfulare

welldocumented(e.g.Baggettetal.2015).Manyoftheconditionsthataffecttheoutcomes

ofoysterproductionalsoinfluencefishabundanceanddiversity,therebyfunctionallylinking

oysterreefhabitatswithfishviawaterqualityandotherattributesoftheenvironment.We

will,therefore,examinehowthefactorsthatcontrolestablishmentofrestoredoysterreefs

mightalsoaffectthefishassemblagesthatcolonisethesereefs.

Environmentalvariables

Bothoystersandindividualfishspecieseachhaveoptimal,andoftendifferent,

physicochemicalenvelopesinwhichtheyprefertolive(e.g.SolanandWhiteley2016).For

example,oysterrestorationsitesmustbelocatedwithinthephysiologicaltolerancesofthe

mainreef-buildingoysterand/ormusselspecies,withrespecttotemperature,salinity,

dissolvedoxygen,andturbidity(Baggettetal.2015).Catchmentrun-off,tides,waves,and

currentsmodifytheseenvironmentalvariablesoveravarietyoftemporalscales(Rodriguez

etal.2014),therebyemphasisingtheneedforagoodunderstandingofecohydrology

principles(WolanskiandElliott2015).Therefore,considerationoftherangeandthreshold

(tolerances)valuesoftheseabioticfactorsismoreimportantthantheirmeanvaluesin

determiningtheplacementofoysterreefs.Inaddition,foraself-sustainingbed,thesitehas

tobewithinasuitablehydrographicregimetogetthespatdeliveredtothearea,i.e.

hydrographicconcentration.Selectingsitesthatmatchtheoptimalenvironmentaltolerance

envelopesforbothoystersandfishshouldthereforebeconsideredakeygoalwhenseeking

torestorefishassemblagesaroundrestoredoysterreefs.

Mostestuarinefishspecieshaveevolvedtocopewithvariationinthephysicochemical

propertiesofcoastalwaters(ElliottandQuintino2007),butareneverthelessalso

susceptibletoextremewatertemperatures(MarshallandElliott1998),salinitylevels

(BachmanandRand2008;MarshallandElliott1998),andlowdissolvedoxygen

concentrations(Stevensetal.2006)thatarebeyondtheirphysiologicallimits.Tominimise

58

effectsofvariableandespeciallypoorwaterqualityonbothoystersandfish,oysterreef

restorationprojectscantargetlocationsthatprovidehighlyoxygenatedwaterswithgood

waterflow(Lenihanetal.1999;Lenihanetal.2001).Theappropriatewaterqualityhastobe

atthereefsite,aswellasthesiteofthesourceoysterorfishpopulations.

Highsedimentation,especiallyfromfinesiltsandclays,detrimentallyaffectsoysterspat

settlement,andlowersthebodycondition,reproductiveoutput,andgrowthofoysters

(Kimbroetal.2014;Lenihan1999;Tamburrietal.2008).Highturbiditylevelscanalsobe

harmfultosomeestuarinefishspecies(BenfieldandMinello1996),particularlyvisually-

orientingpredators(LuntandSmee2015).Whilstoysterreefscanhelptoreduceturbidityat

localscales,thiseffectmaytakeseveralyearstodevelopandreliesonthepersistenceof

adultoysters(LaPeyreetal.2014b;NewellandKoch2004).Positioningoysterreefsatsites

withinestuariesthatregularlyexperienceveryhighturbiditymay,therefore,limitboththe

growthofoysterreefsandtherateatwhichtheyarecolonisedbyfish.Manyestuaries,

especiallymacrotidalones,havestrongerosion-depositioncycleswhichneedincorporating

intotheassessmentofrisktothebeds.

Reefproperties

Aprimefunctionofrestoredoysterreefsisoftentocreateorenhancefishhabitat.The

biologicalandstructuralpropertiesofrestoredoysterreefsareoftenimportantfactorsin

determiningthequalityofreefhabitattofish,includingprovidingfoodandshelterfrom

predation(i.e.reefsize,verticalrelief,watercover,andstructuralcomplexity).Thereef

couldproducetheappropriatehabitateveniftheoysterswerenolongeralive.

Oysterreefsareactivelyrestoredeitherbyreplacinghardsubstratathathavebeenlost,

therebyallowingnewoysterspattosettleandgrow(e.g.TypeAecoengineering),and/orby

reintroducinglivingoysters(Baggettetal.2015;LaPeyreetal.2014a),(e.g.TypeB

ecoengineering)(Elliottetal.2016).Bothmethodsrequiresuccessfullarvalsettlement,

whichcanresultfrommatureoystersonrestoredreefsorasoysterspatbroughtfromother

locations(Lipciusetal.2008).Reef-associatedfishesoftenfeeddirectlyonoysterspat,adult

oystersandotherinvertebrates(e.g.polychaetes,amphipods,sponges,mussels)thatgrow

onoysterreefs,orusereefsashabitats(JohnsonandSmee2014;LehnertandAllen2002).

Therefore,productivereefscomposedofhealthyoysters,andotherinvertebrates,arelikely

tobemorebeneficialforfish(Petersonetal.2003).

59

Subtidalreefsthatextendhighabovetheseabedandoutofpotentiallowoxygen

concentrationboundarylayers,resultinimprovedoystersettlement,growth,andsurvival

onthereefcrests(Lenihanetal.1999).Tallerreefsmightalsobemoreresilienttothe

potentialimpactsofsedimentation,sea-levelrise,parasitesanddiseases(Lenihanetal.

1999;Rodriguezetal.2014).However,aswithmarinemusselstheincreaseinbivalve

individualsizetomaketheorganismsprotrudeabovethebedboundarylayercouldmake

themsusceptibletobeginpulledoffthebedbythestrongercurrents.Oysterreefswith

greaterverticalreliefprovidemorecalmwaterintheirlee(i.e.currentshadow),whichis

favouredbysomefishspecies(Breitburgetal.1995;Lenihan1999),andcanbepositively

relatedtofishbiomass(GratwickeandSpeight2005b).

Placingreefsinintertidallocationsresultsintheirdryingatlowtide,thusrestrictingthetime

forfishtousethearea.Becausefishneedwater,thehabitatvalueofintertidalreefsforfish

maybelowerthanthatofsubtidalreefs(e.g.LehnertandAllen2002).Whiletheseintertidal

reefs,likemanyotherintertidalhabitats(e.g.mangroves,intertidalflats),mightproviderich

feedingopportunitiesduringhightide,thestructureofthefishassemblagethatutilises

themoftendependsonthecompositionofthesurroundingseascapethatfishuseduring

lowtide(Oldsetal.2012a;Pittmanetal.2004).Thisindicatestheimportanceofthe

knowledgeofthebiologicalandhydrographicconnectivitybetweenareasandhence

feeding,breedingorrefugiamigrations.Thedispersalratesoflarvae(ofoysters)andpost-

larvae(offishes)combinedwithtidaldiscursiondistanceswilldictatethedeliveryofspat

andrecruitstotherestoredareas.Thisconceptissimilartotheeffectsoftidesonthe

habitatfunctionsofmangroves;mangroveforeststhatfalldryatlowtideareoftenpoorer

fishhabitatsthatthosethataresubmergedpermanently(e.g.Bakeretal.2015).Where

projectsnecessitateintertidalreefs(e.g.duetooysterspeciesbiology,disease

considerations,ortoensuresafenavigationbyboats),itmaybeimportantthattheyare

positionedclosertosubtidalhabitatsnearbywherefishescanseekrefugeduringlowtides,

suchasonseagrassorotherreefs(Oldsetal.2012b;Petersonetal.2003).

Intermsofreefandprojectextent,arestorationprojectseekingtoenhanceseascape

heterogeneityforfishcanaimtorestorethegreatestextentofoysterreefspossiblewithin

financialandtimelimitations,uptoanyestablishedhistoricextentsofoystersinthetarget

estuary.Thesizeofarestorationsitehastwocomplementaryfacets:1)theareaofseabed

60

coveredbylivingandnon-livingoystershells,and2)thetotalaerialdimensionsofthe

projectfootprint,includinganynon-reefareasbetweenoysterreefs(Baggettetal.2015).

Habitatcomplexityandextentwilldictatetheabilityoftherestoredareatogainorregainit

functioning;forexample,Wootton(1992)emphasisedthatfishdiversityincreasedwith

habitatheterogeneity,thensizeandthenproductivity.Habitatsthatarestructurally

complex(e.g.reefs,mangroves,seagrass)typicallyharbourhigherfishbiomassanddiversity

thanhabitatsthataresimpler(e.g.sandandmudflats)(Grabowskietal.2005;Gratwicke

andSpeight2005a;Shermanetal.2002).Thearchitecturalcomplexityofoysterreefscanbe

measuredintheformofrugosity(roughness)(scale:cmtotensofm),andintheformof

newspatialheterogeneitythatthereefsaddtocoastalseascapes(scale:mtokm).Atthe

scaleofindividualoysterreefs,theprovisionofholesandcrevicesofvariablesizesthatcan

beinhabitedbyarangeoffishes,andusedasrefuges,feedinggrounds,andspawningsites,

whichwillhelptopromotefishdiversityandbiomass(GratwickeandSpeight2005a).Atthe

scaleofoysterrestorationprojects,reefsthataredesignedtoprovideadiversityofhabitat

structures(i.e.,withhighrugosityandverticalrelief)acrossthesitearelikelytocontain

moreandahigherdiversityoffish(Bozecetal.2015)becausetheyprovidemorefeeding

opportunitiesandbettersanctuariesfrompredators,especiallyforjuvenilefish(Petersonet

al.2003).Theseconceptsarewidelyacceptedforartificialreefs(Shermanetal.2002;

WilsonandElliott2009),butrequirefurtherinvestigationforoysterreefs(Table1).

61

Table1Listofresearchquestionsforfishandfisheriesassociatedwithrestoredoysterreefs.

ResearchField PriorityResearchQuestions

A.Oysterreefsasfishhabitats

1.Foodforfish:Underwhatscenariosareoysters,andthelarvaeofothersessileinvertebrate(e.g.ascidian,polychaete,sponge),mostlikelytosettle,grow,andprovidehigh-qualityfoodforfish?(e.g.Grabowskietal.2005)

2.Protectionfrompredators:Whatattributesofoysterreefs(i.e.area,height,architecture)providefishwiththebestprotectionfrompredators?(e.g.Shermanetal.2002)

B.Oysterreefsaspartofwiderseascapes

3.Connectivity:Howdoestheseascapecontextofoysterreefsaffectecologicalprocesses(predation,nutrientturn-over),fishlarvalsettlement,andhabitatvalueforadultandjuvenilefishes(e.g.Bostrometal.2011;Grabowskietal.2005),arethesepatternsconsistentacrossdifferenttypesofseascapes,andwhichoysterreefdesignsmaximisethesemetrics?

4.Fishmovement:Towhichalternatehabitatsshouldoysterreefsbeconnectedtomaximisereefvaluetofishes,andoverwhatscalesdotheseconnectivityeffectsoccur?(e.g.Nagelkerkenetal.2015)

C.Fisheriesandcatchmentmanagement

5.Peopleandoysterreefs:Whataretheeffectsoffishersandfishingontherestorationoffishbiomass?(e.g.Powersetal.2009)Whatarethesocialandculturalvaluesofrestoredoysterreefs,andhowcanthesebemaximised?(e.g.Kingsley-Smithetal.2015;Venturellietal.2017)

D.Monitoringandadaptability

6.Indicatorspeciesandprocesses:Whatarethebestecologicalindicatorsofrestorationsuccessforfinfish,andwhichsuiteofindicatorsaremostappropriateformonitoringeffectsonecosystemconditionandfunction?(e.g.Valesinietal.2017)

62

2.Recognisethatoysterreefsarepartofawiderseascapethatincludesotherfishhabitats

Thesuccessfulrestorationrequiresagoodknowledgeofecologicalprinciples.Firstlythatthe

physico-chemicalenvironmentwillsetupthefundamentalsubstratumorwatercolumn

nicheswhichthengetoccupiedbyorganismstogivecommunitystructure(theso-called

environment-biologyrelationships)(GrayandElliott2009).Followingthis,biology-biology

relationshipscreatetheecologicalfunctioningfromthestructure,andthenthebiotastart

modifyingtheenvironment(biology-environmentrelationships)especiallyinthecaseof

ecosystemengineerssuchasreef-formingspecies.

Seascapeecologytransfersecologicalprinciplesandconceptsfromlandscapeecologyto

marinesystems(Pittmanetal.2011).Atthecoreofseascapeecologyistherecognitionthat

theecologicalfunctionsofecosystemsarecontingentonthetype,conditionandspatial

arrangementofotherecosystemstructuresacrossentireseascapes(Bostrometal.2011;

Gustafson1998).Themovementofmatterandorganismsacrossseascapesfunctionallylinks

ecosystems,andthisecologicalconnectivityshapesspeciesdistributions,foodweb

structure,andecosystemfunction(Oldsetal.2016;Pittmanetal.2011).Therecoveryoffish

assemblagesonrestoredoysterreefsreliesoncolonisationfromelsewhereintheseascape;

therefore,seascapepositioningisavitalconsiderationwhenrestoringoysterreefsforfish.

Connectivitywithotherhabitats

Incoastalseascapes,manyfishmovedaily,ortidally,betweenmarshes,mangroves,

seagrassesandnaturalreefsatscalesofmetrestohundredsofmetres(Bostrometal.2011;

Grober-Dunsmoreetal.2009;Oldsetal.2018c;Potteretal.2015).Ecosystemsthatare

betterconnected(i.e.,closertogether,orlinkedbycurrents),therefore,usuallyharbour

morefishthanthosethatareisolated(Nagelkerkenetal.2015;Oldsetal.2018c)(Figure4).

Thelevelofconnectivitybetweenoysterreefsandotherecosystemsis,therefore,an

importantconsiderationinthedesignofoysterrestorationprojects;particularlyfor

intertidalreefsthatdryonebbtidesandforcefishtomoveintosubtidalhabitats(Grabowski

etal.2005;Petersonetal.2003).Despitethis,connectivityisparticularlydifficultto

quantify.TheUKMarineConservationZoneconceptaimedforacoherentandconnectedset

ofsitesanddefinedsitestobeconnectedasdictatedbylarvaltimeinthewatercolumn

(Robertsetal.2010).However,suchaconnectivityrule-of-thumbthendependsonthetidal

excursion,tidaloscillationsandthepresenceofoceanicfrontswhichcanbeabarrierto

movement(Greenetal.2014;Oldsetal.2018c).

63

Asrestorationprojectsareincreasinglybeingconductedinhighlymodifiedseascapes,the

degreetowhichbuiltinfrastructuremayserveasabarriertofishmovements(Bishopetal.

2017),therebyreducingthebenefitsofrestorationprojectsforfinfish,mayalsobea

significantconsideration.Connectionswithmarshes,seagrasses,andmangrovesarelikelyto

bethemostimportantforoysterreefs(Figure1),dependingontheseascapeinwhichthey

areimbedded(Bostrometal.2011;Gainetal.2017)andonthelocalfishcommunity

requirementsforfeeding,spawningandrefugiamigrations.Globally,studiesonhabitat

connectivitywithoysterreefsareentirelyrestrictedtomarsh-dominatedseascapes(Figure

1),withnostudiesconductedinsubtropicalseascapes,especiallyaroundmangroves,

despitethecurrentorhistoricalpresenceofoysterreefsinmanyoftheseareas.While

seagrassesoccurinbothmarshandmangrove-dominatedseascapes,fewstudieshave

explicitlyassessedtheeffectsofseagrassconnectivityforfishonoysterreefs(TableS1).

Determiningtheimportanceoftheseconnectionswithalternatehabitats,andthedistances

overtheyfunctionindifferentseascapecompositions(e.g.mangrove-versusmarsh-

dominatedseascapes),shouldthereforebeapriorityforresearch(researchpriorities3and

4,Table1).

Whilstitcanbegeneralisedbroadlyacrosscoastalecosystems,thathigherconnectivitywith

alternatehabitatsispositiveforfishassemblages(Oldsetal.2018c),therearesome

contrastingresultswithintheoysterreefliterature.Forexample,studiesonoysterreefsin

NorthCarolina,USA,concludedthatrestoredreefsdirectlyadjacent(<10m)toexisting

vegetatedhabitatsdidnotaugmentfishabundancetothesamedegreeasmoreisolated

reefs(e.g.onmudflats)(Geraldietal.2009;Grabowskietal.2005).Conversely,arecent

studyinTexas,USA,indicatedthatreefsneartoseagrasshadhigherabundanceof

macrofaunathanmorepoorlyconnectedreefs(Gainetal.2017).Theconsistencyofthese

effectswithinmangrove-orseagrass-dominatedseascapesthereforeremainsunclear.

Consequently,furtherstudieswhichseektodeterminetheoptimaldistancefortheisolation

areanimportantrequirementforoptimisingfuturerestorationprojects(researchpriorities

3and4,Table1).

Thescaleofpatchconnectivityeffectsbetweenhabitatsisusuallybetween100and1000m

inmostcoastalseascapes(Bostrometal.2011),andisdictatedby:1)fishmobility;2)the

typeofmigrationbeingundertaken(e.g.feeding,reproductive);3)thecompositionof

64

seascapes;and4)hydrology(Edwardsetal.2009;Nagelkerken2009;Oldsetal.2012b).For

example,fishmovesmallerdistancesamonghabitatstofeedthantheydoduring

ontogeneticmigrations.Feedingmigrationsintoseagrassmeadowsareoftenshorterthan

similarforaysintohabitatswithmoreverticalrelief(e.g.mangroves,reefs),andtidal

migrationsamonghabitatsareshorterinmicrotidalsystemsthaninareasthatexperience

largertidalrangeswhosetidalexcursioncanbeusedasatransportmechanism(Grober-

Dunsmoreetal.2009;Oldsetal.2018c).System-specificinformationonthelocationand

conditionofotherfishhabitats,andthescaleoverwhichfishmovementslinkecosystemsin

focalseascapesisthereforevital(Gilbyetal.2018b;Nagelkerkenetal.2015)(Table1).

Connectivitywithotheroysterreefs

Habitatsthatareclosetootherpatchesofthesametypeofhabitatoftensupporthigherfish

diversity,abundance,andbiomassthanisolatedpatches(GustafsonandGardner1996;

Soonsetal.2005).Effectsofthistypeofhabitatconnectivityhavebeenreportedwidelyin

seagrass,marsh,andcoralreefecosystems,butarerarelytestedforoysterreefs(Bostrom

etal.2011).Wheretheyhavebeentested,connectivityhasshownhighlyvariableeffects

(Grabowskietal.2005;Gregalisetal.2009).Whilstthe‘optimal’distancethatmaximises

fishmovementbetweenoysterreefsisunknown(Table1),distancesarelikelytobesystem-

specific,andscaleonthedispersalcapacityofspecieswithinindividualsystems.From

publishedliteratureonbothoystersandfish,wecansurmisethatrestoredoysterreefs

shouldbesufficientlytoexistingreefstoensurethattheyreceiveagoodsupplyofboth

oysterlarvaeandfish(Gregalisetal.2008;Steppeetal.2016),butalsosufficientlyfarapart

toprovideadditionalreefnodesinthenetworkofoysterreefsthatarelinkedbyfish

movement(GustafsonandGardner1996;Soonsetal.2005)(Figure4).

Thespatialseparationofrestoredoysterreefsshouldbeinformedbythemigrationpatterns

offishspeciesthataretargetsforrestoration.Fishshouldbeabletomoveeasilyamong

oysterreefstoaccessmultiplerestoredreefsinthefocalseascape.Previousstudieshave

suggestedthatmultiplesmallerreefsmightprovidesimilarhabitatvaluesforfishassingle

largerreefs(Harwelletal.2011).Consequently,severalsmallerrestoredreefsthatarewell

connectedtoeachother(withinthe100to1000mrange)acrossaseascape,thereforeare

morelikelytoeffectiveatenhancingfishpopulationsacrossentireestuaries(Table1;Figure

4).Thehydrologyofestuariesandcoastalwatersalsoshapesthecompositionoffish

assemblagesbyregulatingthelikelihoodofjuvenilesettlement(HannanandWilliams1998)

65

andtheprobabilityofvisitationbyadults(ConnollyandHindell2006;Hendersonetal.

2017a).Fishemploytidalexcursioncurrentstotraverselargedistancesbetweeninshore

habitats(e.g.feedingareas,juvenilenurseryhabitats)andoffshorehabitats(e.g.spawning

areas,adulthabitats),anduseotherstructurallycomplexecosystemsassteppingstones

(i.e.,feedingandrestingareas)duringthesemigrations(Engelhardetal.2017;Nagelkerken

etal.2015)(Figure4).Byselectivelyrestoringoysterreefsinlocationstoaddbothspatial

andstructuralheterogeneitytocoastalseascapes,restorationprojectsmightalsoprovide

additionalhabitatsforfishtouseassteppingstonesduringtheseontogeneticmigrations

(Mullaney1991;zuErmgassenetal.2016)(Figure4).Hencethisalsoensuresconnectivity

whichisthecentralmechanismforcreatingcoherenceamongstrestoredandprotected

environments(D’Agostinietal.2015;Planesetal.2009).

Enhancingfishrecruitmentandnurseryvalue

Oysterreefsareimportantsitesforfishspawning(TolleyandVolety2005),attractfish

larvae(Breitburgetal.1995),andfunctionasnurseryareasformanyfishspecies(Coenetal.

1998;Petersonetal.2003;zuErmgassenetal.2016)(Figure4).Theextenttowhichoyster

reefsfunctionasnurseryhabitatsforjuvenilefishesisdeterminedbythreeinter-related

factors:(1)thelikelihoodthatfishlarvaesettleonreefs;(2)theabundance,growthand

survival,ofjuveniles;and(3)thelevelofsuccessthatjuvenilefishhaveinmigratingfrom

oysterreefstotheiradulthabitatshabitats(sensuBecketal.2001).

Fishlarvaeoftenenterestuariesthroughpassagestotheopensea(e.g.estuarymouths,

coastalbays,surfbars)(Blaber2008),andsomeestuarinefishesspawnoversurfbarsin

thesepassagesinmanyregionsoftheworld(Oldsetal.2017;Sheavesetal.1999).Placing

oysterreefsnearthemouthsofestuariesmight,therefore,enhancethelikelihoodofreefs

beingusedasspawningsites,andalsopromotethelikelihoodoflarvalsettlement(Pichleret

al.2017).Larvalrecruitmentmightalsobeenhancedbycreatingmorecomplex,andtaller

oysterreefs,whichcreateeddiesinwhichlarvaeaccumulate(Breitburgetal.1995).

However,theplacementofreefsnearertoestuarymouthsmightincreasetheirvulnerability

tobeingcoveredbymovingsediments,andoysterdiseaseinsomepartsoftheworld

(associatedwithhighersalinity;Lenihanetal.1999).Despitethis,reefbedsatthemouthsof

estuariesmaynotbeself-sustainingifthelarvaegettransportedawayandthereareno

seedingpopulationswithintheinterconnectedhydrographicsystems(WolanskiandElliott

2015).

66

Thegrowthandsurvivalofjuvenilefishdependsontheavailabilityofqualityfood,and

protectionfrompredation(Blaber2008).Thisisturnreliesonthepresence,andcreationor

lossofhabitats,whichagainareinfluencedbyhabitatchangethroughrestorationor

anthropogenicpressures(Amorimetal.2017).Asmanycoastalfishrequiremultiplehabitats

throughouttheirlives,especiallyduringearlyontogeneticmovements,itisthequalityof

bothindividualhabitatsandthesurroundingseascapethatenhancesnurseryvalueforlarval

andpost-settlementfishes(Nagelkerkenetal.2015).Oncefishhaverecruitedtooyster

reefs,orintothesurroundingseascape,thearea’svalueasanurseryisdeterminedbyfood

availability,predationpressure,competitiveinteractionsforfoodandspacewith

cohabitants,andtheavailabilityofalternativeforagingandrefugehabitatsintheseascape

(Gittmanetal.2016;Pittmanetal.2007).Oysterreefsthatarerestoredinappropriate

locationscanmodifyeachofthesefeaturesbyprovidingfeedingandsheltering

opportunities,whichservetoreducecompetitionandpredator-inducedmortalityforfishon

reefsandinadjacenthabitats,andmightthereforeenhancethenurseryfunctionofcoastal

seascapes(Figure4).

3.Considertheimpactoffisheriesandcatchmentmanagement

Anthropogenicstressorssuchasrunofffromalteredcatchments(Gilbyetal.2015;Lerberg

etal.2000)andoverharvestingbothinthecatchmentandatsea(Paulyetal.2003;Paulyet

al.2005)havesubstantialconsequencesformarineecosystemconditionandresilience.In

thecontextofmanagingfishstocksassociatedwithrestoredoysterreefs,managersneedto

consider:

(1)howtomanagefishstocksusingcatchrestrictions,includingthedesignationor

expansionofno-takereserves;and

(2)howtomanageotherpotentialimpactsfromthecatchment,andwiderseascapes,in

whichreefsarelocated.

Thislevelofmanagementthereforehastorespondtoawholesuiteofpressures,both

exogenicunmanagedandendogenicones(Elliott2011).Theexogenicpressures,which

emanatefromoutsidethemanagementareaandinwhichmanagementsometimesonlybe

abletorespondtotheconsequencesandnotthecauses,includeclimatechangeeffects,

runoff,andthelossofbreedingfishpopulationsawayfromthesite.Thesecanbebothout

atseaandalsoelsewhereinthecatchment(Elliottetal.2017).Theendogenicmanaged

pressuresincludethoseimpactsinanareasuchashabitatlossandpollutingdischarges.

67

Fisheriesmanagement

Theeffectivemanagementoffisheriesincoastalecosystemsreliesonmaintainingahigh

biomassoflarge,maturebreedingfish.Themostcommonmanagementinterventionis

catchrestrictionsintheformofeitherreserves(Edgaretal.2014)orbyimplementingsize

andbaglimits(BartholomewandBohnsack2005).Notwithstandingsomeuncertainty

regardingthesurvivalratesofreleasedindividuals,andthecriticaleffectsofthedegreeof

enforcement(Guidettietal.2008),andfishingeffortdisplacedbyreservedeclaration

(Halpernetal.2004;Lédéeetal.2012),theconsensusisthatcatchrestrictions(i.e.limitson

thenumbersbysizeclassoffishcaughtbyanglers)havepositiveeffectsontheabundance

ofharvestedspeciesinthegreatmajorityofcases(Edgaretal.2014;Tetzlaffetal.2013).

Strategicallyplacedreserves,whicharewhollyno-take,well-designed,andwell-policed,

increasetheabundanceandbiomassofharvestedspecies(Edgaretal.2014),restoretrophic

relationshipsandfoodwebs,resultinginhabitatimprovements(GilbyandStevens2014),

andinsomecases,enhancesurroundingfisheries(Halpernetal.2009).Wheretheyare

designedtoformacoherentnetwork,asisrequiredbyanincreasingamountoflegislation

globally(e.g.theUKMarineandCoastalAccessAct),thenallthesiteshavetobeconsidered

asafunctionalunit.Reserveeffectivenessisenhancedbyprotectingmultiplefishhabitats

andthespatialconnections,andcorridors,betweenthesehabitats(Oldsetal.2016).

Consequently,wesuggestthatitwouldbeprudenttoestablishrestoredoysterreefsinno-

takemarinereserves,andanyimportantconnectioncorridorswithadjacenthabitats(Olds

etal.2012a).Wesuggestthatoysterreefsthatarerestoredinoptimalpositionsin

heterogeneousseascapes(Bostrometal.2011;MicheliandPeterson1999)andalso

protectedinno-takemarinereserves,wouldlikelyfunctionbetterforbothoystersandfish,

thanrestoredreefsthatareopentofishing(Oldsetal.2016).Conversely,reefsopentonon-

sustainableoysterharvestingwilllikelybequicklydegraded(Kirby2004;Rothschildetal.

1994),therebylikelyalsoreducinganyvalueforassociatedfishcommunities.However,the

degreetowhichfishbiomassonrestoredoysterreefsisaugmentedbyplacementwithin

marinereserveshasnotbeentested(Nevinsetal.2014)(Table1).

Catchmentmanagement

Humanpressuresonestuarineandcoastalecosystemsarediverse(e.g.pollution,fishing,

habitatdestruction)andoccurthroughouttheadjacentcatchmentandmarineareas(Elliott

68

etal.2014;Lotzeetal.2006).Thissuiteofexogenicandendogenicanthropogenicstressors

andtheirlargespatialfootprintnecessitatesabroad‘land-to-sea’frameworkformanaging

potentialchangesinwaterquality,fishpopulations,andhabitats(CicchettiandGreening

2011;Gilbyetal.2016).Oysterreefrestorationcanbecompromisedbyexcessivenutrient

andsedimentinputs(Wallesetal.2016),associatedeutrophicsymptomsincludinghypoxia

(Becketal.2011),andlossofconnectivitywithfunctionallylinkedhabitats(e.g.seagrasses,

mangroves,coastalseas;Nagelkerkenetal.2015;Oldsetal.2017;Whitfield2017)through

eitherhabitatdegradation.Fishaccesstoreefsand/orestuariesmightbelimitedbyblocking

offishpassagethroughcoastaldefences,sandbankdevelopmentduetolowriverflowfrom

increasedabstractionupstream,andurbanbarriersandinfrastructures(Bishopetal.2017).

Theoutcomesofoysterreefrestorationwill,therefore,bemaximisedonlywhen

managementminimisesthedeleteriouseffectsofotherimpactingprocessesthataffectthe

focalestuaryanditscatchment.Theexpansionofoysterreefhabitatcan,however,augment

otherestuarinehabitatsthroughtheintroductionofproductionecosystemservicesleading

tosocietalgoodsandbenefits(Elliottetal.2017;TurnerandSchaafsma2015).These

include:(1)improvingwaterqualitybyreducingexcessnitrogen(PiehlerandSmyth2011;

Smythetal.2015)andfilteringparticulates,whichincreasesthelevelofsunlightreaching

theseabed(Walletal.2008);(2)thefertilizationofbenthichabitatsfrompseudofaeces

(PetersonandHeck1999).Theseservicesfacilitatemarshandseagrasshabitats,whichin

turnincreasefishproduction(Whitfield2017)andagaininturnleadtoincreasedsocietal

goodsandbenefitssuchascommercialfishyields,recreationbenefitsorcoastaldefences

(TurnerandSchaafsma2015).

4.Monitorreefsandfishacrosstheseascapeformanagement,andimplementchanges

whererequired

Restorationprojectsshouldhaveexplicitgoals,executedbybestpracticesthatcanbe

adaptedbasedonresultsfromongoingmonitoringandnewresearch(MargulesandPressey

2000;WiensandHobbs2015).Indeed,failedorineffectiverestorationisoftenduetopoor

orill-definedobjectives(Elliottetal.2016).Hence,managementofoysterreefrestoration

projectsrequiresthereviewofexistingmanagementinterventions,andtherefinementof

anypracticesthatareineffective(seethelessonslearnedinElliottetal.2016).Thus,oyster

reefrestorationprojectsshouldcontinuallymeasurehoweffectiveactionsareinmeeting

restorationgoals(WiensandHobbs2015).Whilemonitoringprotocolsandmetricsfor

restoredoysterreefsareestablishedforthereefsthemselves(seeBaggettetal.2015for

69

specificdetails),andbasicmonitoringprotocolshavebeendetailedforfinfish(Baggettetal.

2014),generalmetricsforassessinghowreefsaffectthequalityofsurroundingfishand

fisheries,beyondtherestorationsite(i.e.ataseascapescale)havenotbeenestablished.

Ingeneral,therearetwoalternativestodeterminingwhethertherestoredsiteisperforming

asdesired:oneisthatacomparisonwiththerestoredsiteandanothercontrolsite(or

anothercontroltime)couldbeused,oralternativelytheenvironmentalmanagersneedto

clearlyindicatewhatisdesiredforarestoredsiteandthenallofthemonitoringistocheck

deviationfromthatobjective;bothofthesearewhollydependentonclearobjectivesbeing

setfortherestoredsiteanditsdependentpopulationsandspecies.

Thechoiceofmonitoringmetricswilllargelybedeterminedbythegoalsandobjectivesof

specificprojects(McDonaldetal.2016).Thereare,however,severalminimumrequirements

thatshouldbemettoenableproperestimationsoftheeffectsofoysterreefsonfish.Ata

minimum,allmonitoringshouldencompasscountsofentirefishassemblagesatmultiple

timepointsandcontrolsitesbothbeforeandafterrestoration.SuchaBACI-PS(Before-

After-Control-ImpactPairedSeries)designisneededtodisentangletheeffectsoftidal,

seasonalorannualvariationonfishassemblages(Underwood1994),andtomeasure

secondaryproduction(i.e.theaccumulationoffishbiomassovertime).Itisalsodesirableto

monitornotonlyrestorationandcontrolsites,butalsoreferencessitesorremnanthabitats

ofthetypewhichtherestorationeffortisaspiringtorecreate(Graysonetal.1999;

McDonaldetal.2016).Thephysico-chemicalenvironmentneedstobemonitoredaswellas

theecologicalstructureandfunctioningotherwisechangesinthelattercannotbeexplained.

Ideally,multiplecontrolestuaries,withnooysterreefrestoration(againfollowingBACI-PS),

shouldalsobemonitoredduringtheperiodofreefestablishmentandfishrecruitment,so

thattheecologicalbenefitsofoysterrestorationcanbeproperlyseparatedfromanyother

regionalchangesthatmightalsobeaffectingoysterreefsandfishassemblages(Underwood

1994).Projectsrunningforanumberofyearswillbenefitgreatlyfrommeasuringthe

recruitmentandsizedistributionsoffocalspecies(Shinetal.2005).

Asfishmoveamongecosystemsincoastalseascapes(e.g.fromoysterreefstoother

habitats)(Nagelkerkenetal.2015),thepotentialfisheriesbenefitsofrestoredoysterreefs

willnotberestrictedtothenear-fieldfootprintofindividualrestorationprojectsbutalso

requireconsiderationoffar-fieldeffects.Toestablishtheextentofanysuchfisheries

70

benefitsitwill,therefore,bebeneficialtomeasurepotentialchangestofishassemblages

andfishcatchesatawiderseascapescale(i.e.uptokmfromrestoredoysterreefs).In

essence,themonitoringhastocoverallcomponentpairs:environment-oysters,

environment-fishes,oysters-fishes,oysters-predators,andfishes-predators.Thisgivesboth

thestructuralandfunctionalmeasurements.

Inadditiontomonitoringfish,thereareseveralmetricsthatmightbemonitored,andthat

areusuallycorrelatedorassociatedwiththeabundanceoffish,thatmanagerscanuseto

demonstratetheeffectivenessofoysterreefsforfish,andfortheecosystemmorebroadly.

Severalecologicalprocesses,suchasscavenging(Webley2008),predation,andnutrient

sequestrationorturnover(Kelloggetal.2013),areintimatelylinkedwiththeabundanceof

fishesincoastalecosystems,andsoareincreasinglyusedasmeasuresofecosystemhealth

(HavstadandHerrick2003).Theuseofindicatorsofecosystemconditionandfunctioning

(e.g.indicatorspecies,umbrellaspecies)isincreasingforestuariesglobally(e,g.Gilbyetal.

2017a;Montagnaetal.2008;Valesinietal.2017).However,furtherstudiesarerequiredto

identifypotentialindicatorspeciesthatmightbeusefulforoysterreefs;especiallythose

thatmightbeusedtocomparepatternsacrossbiogeographicregions(Table1).Asrestored

oysterreefsaccumulatefishbiomass,theymightalsoalterthespatialdistributionoffishing

effortincoastalseascapes(i.e.,fishingcouldeasilybecomeconcentratedoversuccessful

oysterreefs).Alternatively,dependingonthegearused,thereefmaydiscouragebottom

trawlingtherebyactingasade-factono-takezone.Hence,monitoringpotentialchangesin

thedistributionoffishingeffort(e.g.mappinganglersandcommercialfishinginrelationsto

reefs)is,therefore,necessarytoinvestigatehowthecombinedeffectsofrestorationand

fishingalterfishassemblagesonoysterreefs.

Ithasbeenrecommendedthatthesettlementandgrowthofoystersonrestoredreefs

shouldbemonitoredforupto6yearspostinstallation(Baggettetal.2015).Fish

assemblageswillcontinuetochangethroughthisperiod(i.e.,asoysterreefsbecome

established),andcouldtakeupwardsof10yearstodevelop(zuErmgassenetal.2016).

MonitoringrecoverytowardsthegenericstandardsprovidedbyMcDonaldetal.(2016)

wouldthereforereduceambiguityaroundgoalsandsuccess.

Thesecriteriashouldofcoursebeconsideredasidealoroptimalsamplingregimes.

Gatheringofevenbasicabundanceanddiversitydataatindividualsitesshouldbe

71

consideredanimportantgoalforalloysterreefrestorationprojects.Similarly,cost-effective

methodsofmeasuringhabitatsizeand/orqualityusing,forexample,LIDARordrone

techniques,willdeliverrapidinformationonthesupportingecosystemservices.

Discussion

Oysterreefrestorationiscostly,sorestorationeffortsshouldseekwin-winscenarios(e.g.

forecologyandeconomy),whereoysterreefrestorationachievesmultiplebenefits(e.g.

shorelinestabilisation,andenhancementoffisheriesproductivity).Restoringoysterreefs

augmentsfishbiomassrelativetobaresubstrata,andintegratingseveralkeyconceptsfrom

estuarinefishecologyintothedesignandmonitoringofrestorationprojectswillhelp

maximisetheirreturnoninvestment.Placingcomplexreefstructuresinstrategiclocations

withinheterogeneousestuarineseascapesmightenhanceestuarinehabitatdiversityand

promotetheperformanceofrestoredoysterreefsforfishandfisheries.Thephysical

installationofoysterreefsshouldnotbeviewedasthefinaloutcomeofrestoration

programs.Restoredreefsmustbemanaged,togetherwithotherimpactingprocessesthat

mightthreatenrestorationsuccess(e.g.fishing,sedimentation,eutrophication),and

monitoredovertimetomaximisetheaccumulationoffishbiomass,andthebenefitsof

restorationforfisheries(MargulesandPressey2000;WiensandHobbs2015).Oysterreef

restorationprojectsthataccountforthereefdesignsorplacementsthatmightserveto

maximisetheutilityofseascapesforfishwillleadtogreaterfishdiversity,abundance,and

harvestablefishbiomassthroughoutcoastalecosystems.Asshowninthisreview,thereare

severalimportantresearchquestionsthatneedaddressing(Table1),however,thereare

alsoseveralimportantquestionsformanagerstoaskbasedonexistingliteraturewhen

aimingcreatesuccessfulandsustainablereefsforfinfishandtheirfisheries(Table2).By

understandingtheneedforappropriatemanagementmeasuresandtargets(suchasthose

inTable2)whichhavetorespondtothelargeuncertaintiesintheecologicalfunctioning,

thenthelikelihoodofsuccessfulandsustainablereefscanbeenhanced.

Inmanyinstances,therearelogisticandlegislativechallengestooptimisingthedesign,

positioning,managementandmonitoringofreefsaccordingtothesecriteria.Forexample,

thelocationsinwhichoysterreefscanberestoredmightbelimitedbyregulations

concerningshipping,recreationalactivities,orhabitatconservation(e.g.legislative

protectionofmarineplants).Theymaybelimitedbyapoorknowledgeofhistorical

evidenceofoystersortheunknownreasonsforthedeclineofthepreviousstocks.Funding

72

agenciesmightalsoprefersimpleoysterreefdesignsforeaseofinstallation,tolimitcosts,

ortomaximisetheiraccessibility(Kingsley-Smithetal.2015).Inmanyinstances,theactual

designofoysterrestorationprojectswillbeacompromiseacrossthesemultipleconstraints,

anddependonspecificprojectgoals.Forexample,theenhancementoffisheriesmaybe

secondarytoshorelinestabilisationundersomescenarios.Notwithstandingthese

challenges,ifthegoalofoysterreefrestorationistoenhancefishpopulationsandbenefit

fisheriesitisimperativetooptimisethehabitatvalues,seascapecontext,andongoing

monitoringandmanagementofrestoredreefs.

73

Table2Questionsformanagementwhererestorationseekstorestorebothoysterreefsand

surroundingfinfishandtheirfisheries

Broadfields Specificquestionsformanagement

PrerequisitesDidoysterreefshistoricallyoccurinthearea?

- whatwastheirextent/distribution?- whichspeciesoccurred?

Whatisthetimelineandreasonsofdecline/extirpation?

- diseases/pathogens/parasites?- fishing/harvesting?- waterquality(e.g.suspendedsolids)?- hasthisreasonbeenarrested?

Aretherenowfewerfish? - evidenceofdeclinedbiodiversityand/orcatches?- canthisbetiedtothelossofoysterreefs?

A-OysterreefsasfishhabitatsDotherighthabitatsandphysico-chemicalconditionsexistforoysters?

- istheretheappropriateelevation/depth?- isthehydrographicconnectivitymaintainedbothupstreamandat

sea?- isthesalinity,turbidity,primary(plankton)productivity,suitablefor

settlement,survivalandlong-termgrowth?Cantheenvironmentalvariablesthataffectfishandoystersbematched?

- whatenvironmentalenvelopesdoesthetargetoysterspeciesprefer?- whatenvironmentalenvelopesdothetargetfishspeciesprefer?- wherearetheareasintheestuaryinwhichthesevaluesoverlap?

Reefstructuressuitableforfish?

- howcanthegrowthofinvertebratesandsmallfishbeencouragedtoenhancefeedingopportunitiesforfish?

- howdoesthereeffunctiontoimproveprotectionfrompredation?B-Theseascapecontext

Istheareaavailableandsuitableforreefplacements?

- restrictionsduetoshipping,boatingorrecreationalusages?- restrictionsonoystertranslocation(e.g.Biosecurityexclusions)?

Linkagestootherhabitattypes? - whichalternatehabitats(e.g.seagrass,marsh,mangroves)aremostimportantintermsoflinkageswithineachsystem?

- overwhichscalesdotheselinkagesoccur?- cantheplacementofreefsbeoptimizedunderthiscontextto

improvecarryingcapacity?Linkagestootheroysterreefs? - aretheresourcesavailableforoysterlarvaefromotherreefs?

- canfishusetherestoredreefsasanetwork?- overwhichscalesdotheselinkagesoccur?- cantheplacementofreefsbeoptimizedunderthiscontextto

improvecarryingcapacity?C-Fisheriesandcatchmentmanagement

Areexistingfisheriesmanagementsufficient?

- doexistingbaglimitsormarinereservesservetoenhancethebreedingbiomassoffisharoundreefs?

- ifno,can,andhowmightthisbechangedtodoso?Areexistingcatchmentmanagementplanssufficient?

- arethereexogenousthreatstothesurvivalorgrowthofoystersand/orfish?

- whatarethey,wheredotheyexist(atthesite,upstream,orinthecatchment?),andtowhatdegreedotheyinfluenceoutcomesforfishandoysters(i.e.inwhichordershouldyoutacklethem?)

- whatapproachesshouldbetakentominimiseornegatetheireffects?D-MonitoringandadaptabilityArethereappropriaterestorationgoals?

- whatarethespecific,quantifiablegoalsofrestoration?- aretheseachievablewithinthelifetimeoftheproject?- howwillmonitoringaddresswhetherthesegoalsaremet?- howwilllessonsfrommonitoringbefedbackintothemanagementof

thereefsorbroaderestuary?Howwillthereefsandfishbemonitored?

- whichmetrics,whatmethods?- howdoesthisrelatetothevalueofthereefhabitatforfish(e.g.food

and/orprotectionfrompredators?)?- fishbeyondthereefsite?(i.e.ataseascapescale)- howdoesthisrelatetothebroaderprojectgoals?

74

Therestorationofhabitatsandbiodiversityisvaluedbysociety,andcanconveysignificant

psychologicalbenefitstousers(Fulleretal.2007;ReyBenayasetal.2009).Successful

restorationprojectsthatimprovetheconditionofhabitats,orenhancefishpopulations,also

providevaluableecosystemservicesandsocietalgoodsandbenefits(e.g.foodprovision,

bankreinforcement,biodiversityenhancement)andare,therefore,anassettolocalpeople.

Restoringlosthabitatsshouldbeviewedasasignificantachievement,irrespectiveofthe

goalsoftheprojectorwhetherfisheriesareenhancedbytherestorationefforts.Bybetter

integratingthegoalofsupplementingfishandfisheries,withtheobjectivesofoysterreef

restoration,wemightthereforeincreasestakeholderengagement(LaPeyreetal.2012),and

helptoensurethatrestoredoysterreefsfunctionoptimallywithinsocio-ecologicalsystems.

Inessence,theaimwillbetoimprovetheecologicalstructureandfunctioningandnot

merelyachieveanexerciseofmorevalue‘totheecologiststhantheecology’(Elliottetal.,

2016).

Furtherdevelopmentoftheseconceptsrequiresongoinginvestigationoftheeffectsof

oysterreefrestorationonfishassemblages.Designingoysterreefstructuresthatare

attractiveasbothsourcesoffoodandrefugesfrompredationmustbeapriorityforoyster

reefrestorationprojectswhererestoringfishandfisheriesisalsoagoal(researchpriorities1

and2,Table1).Whilemanystudieshavedemonstratedthatreefsserveasnurseryandadult

finfishhabitats,andaugmentfishproductionlocally(10sto100sm),thenextcriticalstepis

todeterminewhetherthesebenefitsconveytofishpopulations(e.g.Petersonetal.2003),

ecologicalfunctions(e.g.RodneyandPaynter2006),ornurseryvalues(Nagelkerkenetal.

2015)atlargerspatialscales(kmto10sofkm)beyondthereefsites.Furthermore,while

studieshavedemonstratedthatbetter-connectedreefsharbourmoreanimals(e.g.Gainet

al.2017),thedegreetowhichthesemetricsoffishandfish-associatedecologicalfunctions

areenhancedacrossdifferenttypesofseascapes,whethertheseeffectsareconsistent,and

whichoysterreefdesignsoptimizetheeffectsisunclear(researchpriorities3and4,Table

1).Properlymanagingfishbiomasson,andaround,restoredoysterreefsrequiresaclearer

understandingofhowpeopleinteractwithreefs,andhowtheseinteractionsmightmodify

theresponsesofrecoveringfishcommunities(researchpriority5,Table1).Tohelpoptimise

futurerestorationprojectsforfinfish,wemustidentifysuitableindicatorsthatcanbeused

tomeasurerestorationsuccessforfinfishspecifically(researchpriority6,Table1).Thereis

nowanacceptedlistofattributesrequiredbysuitableindicators(e.g.Elliott2011)toensure

75

thatnotonlyaretheyoperationalbutthatmanagerswillknowwhentheyhavebeen

reached.Hencethecentralfunctionofmanagementbeingthatmeasures(suchas

restoration)willbeseentoachievedthedesiredaims.

Restoringoysterreefscanhavesignificant,oftenpositive,effectsforfishandfisheries.The

managementandresearchrecommendationspresentedhereareabasicsetthatcanbe

expanded,refined,andadaptedbyindividualprojectstobestmatchgoalsandobjectives,

andcanbeeasilyintegratedintomostprojects.Weemphasisethattheeffectivenessof

oysterreefrestorationprojectsforfishandfisheriescanbeimprovedbyoptimisingthe

habitatvaluesandseascapecontextofindividualreefs,bymanagingotherimpacting

processes,andthroughadaptivemonitoringwithappropriateindicatorsofrestoration

performance.

76

SupplementaryMaterialsTableS1Listofstudiesthatinvestigatetheeffectsofrestoredoysterreefsforfishglobally.ThesestudiesweresourcedusinganISIWebofKnowledgesearchoftheterm(oyster*andfishandrestor*).Authors Year Title Journal Volume Pages RegionAbbott,R.R.andR.Obernolte

2012 AcousticallytaggedfishutilizationofanartificialreefconstructedfornativeOlympiaoysterrestoration

JournalofShellfishResearch

31 257-258

California

Arve,J. 1960 Preliminaryreportonattractingfishbyoyster-shellplantingsinChincoteagueBay,Maryland

ChesapeakeScience

1 58-65 Maryland

Boswell,K.M.,M.P.Wilson,P.S.D.MacRae,C.A.WilsonandJ.H.Cowan

2010 SeasonalestimatesoffishbiomassandlengthdistributionsusingacousticsandtraditionalnetstoidentifyestuarinehabitatpreferencesinBaratariaBay,Louisiana

MarineandCoastalFisheries

2 83-97 Louisiana

Broekhuizen,N.,C.J.Lundquist,M.G.HadfieldandS.N.Brown

2011 Dispersalofoyster(Ostreachilensis)larvaeinTasmanBayinferredbyusingaverifiedparticletrackingmodelthatincorporateslarvalbehaviour

JournalofShellfishResearch

30 643-658

NewZealand

Brown,L.A.,J.N.Furlong,K.M.BrownandM.K.LaPeyre

2014 OysterreefrestorationinthenorthernGulfofMexico:effectofartificialsubstrateandageonnektonandbenthicmacroinvertebrateassemblageUse

RestorationEcology

22 214-222

Mississippiandothers

Dillon,K.S.,M.S.PetersonandC.A.May

2015 FunctionalequivalenceofconstructedandnaturalintertidaleasternoysterreefhabitatsinanorthernGulfofMexicoestuary

MarineEcologyProgressSeries

528 187-203

Mississippi

Farinas-Franco,J.M.andD.Roberts

2014 Earlyfaunalsuccessionalpatternsinartificialreefsusedforrestorationofimpactedbiogenichabitats

Hydrobiologia 727 75-94 Ireland

George,L.M.,K.DeSantiago,T.A.PalmerandJ.B.Pollack

2015 Oysterreefrestoration:effectofalternativesubstratesonoysterrecruitmentandnektonhabitatuse

JournalofCoastalConservation

19 13-22 Texas

Geraldi,N.R.,S.P.Powers,K.L.HeckandJ.Cebrian

2009 Canhabitatrestorationberedundant?Responseofmobilefishesandcrustaceanstooysterreefrestorationinmarshtidalcreeks

MarineEcologyProgressSeries

389 171-180

Alabama

Gittman,R.K.,C.H.Peterson,C.A.Currin,F.J.Fodrie,M.F.PiehlerandJ.F.Bruno

2016 Livingshorelinescanenhancethenurseryroleofthreatenedestuarinehabitats

EcologicalApplications

26 249-263

NorthCarolina

Grabowski,J.H.,A.R.Hughes,D.L.KimbroandM.A.Dolan

2005 Howhabitatsettinginfluencesrestoredoysterreefcommunities

Ecology 86 1926-1935

NorthCarolina

Gregalis,K.C.,M.W.JohnsonandS.P.Powers

2009 Restoredoysterreeflocationanddesignaffectresponsesofresidentandtransientfish,crab,andshellfishspeciesinMobileBay,Alabama

TransactionsoftheAmericanFisheriesSociety

138 314-327

Alabama

Harding,J.M.andR.Mann

1999 Fishspeciesrichnessinrelationtorestoredoysterreefs,PiankatankRiver,Virginia

BulletinofMarineScience

65 289-299

Virginia

Harding,J.M.andR.Mann

2000 EstimatesofnakedGoby(Gobiosomabosc),stripedblenny(Chasmodesbosquianus)andEasternoyster(Crassostreavirginica)larvalproductionaroundarestoredChesapeakeBayoysterreef

BulletinofMarineScience

66 29-45 Virginia

Harding,J.M.andR.Mann

2001 Oysterreefsasfishhabitat:Opportunisticuseofrestoredreefsbytransientfishes

JournalofShellfishResearch

20 951-959

Virginia

Harwell,H.D.,M.H.PoseyandT.D.Alphin

2011 Landscapeaspectsofoysterreefs:Effectsoffragmentationonhabitatutilization

JournalofExperimentalMarineBiologyandEcology

409 30-41 NorthCarolina

Humphries,A.T.andM.K.LaPeyre

2015 Oysterreefrestorationsupportsincreasednektonbiomassandpotentialcommercialfisheryvalue

Peerj 3 e1111 Louisiana

77

Humphries,A.T.,M.K.LaPeyre,M.E.KimballandL.P.Rozas

2011 Testingtheeffectofhabitatstructureandcomplexityonnektonassemblagesusingexperimentaloysterreefs

JournalofExperimentalMarineBiologyandEcology

409 172-179

Louisiana

Kellogg,M.L.,J.C.Cornwell,M.S.OwensandK.T.Paynter

2013 Nitrogenremovalandnutrientsequestrationofarestoredoysterreef

MarineEcologyProgressSeries

480 1-19 Maryland

Kingsley-Smith,P.R.,R.E.Joyce,S.A.Arnott,W.A.Roumillat,C.J.McDonoughandM.J.M.Reichert

2012 Habitatuseofintertidaleasternoyster(Crassostreavirginica)reefsbynektoninSouthCarolinaEstuaries

JournalofShellfishResearch

31 1009-1021

SouthCarolina

LaPeyre,M.K.,A.T.Humphries,S.M.CasasandJ.F.LaPeyre

2014 Temporalvariationindevelopmentofecosystemservicesfromoysterreefrestoration

EcologicalEngineering

63 34-44 Louisiana

Lehnert,R.L.andD.M.Allen

2002 NektonuseofsubtidaloystershellhabitatinasoutheasternUSestuary

Estuaries 25 1015-1024

SouthCarolina

Lenihan,H.S. 1999 Physical-biologicalcouplingonoysterreefs:Howhabitatstructureinfluencesindividualperformance

EcologicalMonographs

69 251-275

NorthCarolina

Lenihan,H.S.,C.H.Peterson,J.E.Byers,J.H.Grabowski,G.W.ThayerandD.R.Colby

2001 Cascadingofhabitatdegradation:Oysterreefsinvadedbyrefugeefishesescapingstress

EcologicalApplications

11 764-782

NorthCarolina

Lunt,J.,J.ReustleandD.L.Smee

2017 Waveenergyandflowreducetheabundanceandsizeofbenthicspeciesonoysterreefs

MarineEcologyProgressSeries

569 25-36 Texas

MacRae,P.S.D.andJ.H.CowanJr.

2010 Habitatpreferencesofspottedseatrout,Cynoscionnebulosus,incoastalLouisiana:Asteptowardsinformingspatialmanagementinestuarineecosystems

TheOpenFishScienceJournal

3 154-163

Louisiana

Marenghi,F.,G.Ozbay,P.ErblandandK.Rossi-Snook

2010 Acomparisonofthehabitatvalueofsub-tidalandfloatingoyster(Crassostreavirginica)aquaculturegearwithacreatedreefinDelaware'sInlandBays,USA

AquacultureInternational

18 69-81 Delaware

Nevins,J.A.,J.B.PollackandG.W.Stunz

2014 CharacterizingthenektonuseofthelargestunfishedoysterreefintheUnitedStatescomparedwithadjacentestuarinehabitats

JournalofShellfishResearch

33 227-238

Texas

Peters,J.R.,L.A.YeagerandC.A.Layman

2015 Comparisonoffishassemblagesinrestoredandnaturalmangrovehabitatsalonganurbanshoreline

BulletinofMarineScience

91 125-139

EastFlorida

Quan,W.M.,L.Zheng,B.J.LiandC.G.An

2013 Habitatvaluesforartificialoyster(Crassostreaariakensis)reefscomparedwithnaturalshallow-waterhabitatsinChangjiangRiverestuary

ChineseJournalofOceanologyandLimnology

31 957-969

China

Quan,W.-m.,J.-x.Zhu,Y.Ni,L.-y.ShiandY.-q.Chen

2009 Faunalutilizationofconstructedintertidaloyster(Crassostrearivularis)reefintheYangtzeRiverestuary,China

EcologicalEngineering

35 1466-1475

China

Rezek,R.J.,B.Lebreton,E.B.Roark,T.A.PalmerandJ.B.Pollack

2017 Howdoesarestoredoysterreefdevelop?Anassessmentbasedonstableisotopesandcommunitymetrics

MarineBiology 164 54 Texas

Robillard,M.M.R.,G.W.StunzandJ.Simons

2010 Relativevalueofdeepsubtidaloysterreefstootherestuarinehabitattypesusinganovelsamplingmethod

JournalofShellfishResearch

29 291-302

Texas

Rodney,W.S.andK.T.Paynter

2006 Comparisonsofmacrofaunalassemblagesonrestoredandnon-restoredoysterreefsinmesohalineregionsofChesapeakeBayinMaryland

JournalofExperimentalMarineBiologyandEcology

335 39-51 Maryland

Scyphers,S.B.,S.P.PowersandK.L.Heck,Jr.

2015 Ecologicalvalueofsubmergedbreakwatersforhabitatenhancementonaresidentialscale

EnvironmentalManagement

55 383-391

Alabama

Scyphers,S.B.,S.P.Powers,K.L.Heck,Jr.andD.Byron

2011 Oysterreefsasnaturalbreakwatersmitigateshorelinelossandfacilitatefisheries

PlosOne 6 e22396 Alabama

Stunz,G.W.,T.J.MinelloandL.P.Rozas

2010 RelativevalueofoysterreefashabitatforestuarinenektoninGalvestonBay,Texas

MarineEcologyProgressSeries

406 147-159

Texas

78

Taylor,J.andD.Bushek 2008 IntertidaloysterreefscanpersistandfunctioninatemperateNorthAmericanAtlanticestuary

MarineEcologyProgressSeries

361 301-306

NorthAtlantic

Tolley,S.G.andA.K.Volety

2005 Theroleofoystersinhabitatuseofoysterreefsbyresidentfishesanddecapodcrustaceans

JournalofShellfishResearch

24 1007-1012

WestFlorida

Tolley,S.G.,A.K.VoletyandM.Savarese

2005 InfluenceofsalinityonthehabitatuseofoysterreefsinthreesouthwestFloridaestuaries

JournalofShellfishResearch

24 127-137

WestFlorida

Ulanowicz,R.E.andJ.H.Tuttle

1992 ThetrophicconsequencesofoysterstockrehabilitationinChesapeakeBay

Estuaries 15 298-306

Marylandandothers

Walters,K.andL.D.Coen

2006 Acomparisonofstatisticalapproachestoanalyzingcommunityconvergencebetweennaturalandconstructedoysterreefs

JournalofExperimentalMarineBiologyandEcology

330 81-95 SouthCarolina

Wenner,E.,H.R.BeattyandL.Coen

1996 Amethodforquantitativelysamplingnektononintertidaloysterreefs

JournalofShellfishResearch

15 769-775

SouthCarolina

Wilber,D.H.,N.H.HadleyandD.G.Clarke

2012 ResidentcrabassociationswithsedimentationonrestoredintertidaloysterreefsinSouthCarolinaandtheimplicationsforsecondaryconsumers

NorthAmericanJournalofFisheriesManagement

32 838-847

SouthCarolina

Workman,I.,A.Shah,D.FosterandB.Hataway

2002 Habitatpreferencesandsitefidelityofjuvenileredsnapper(Lutjanuscampechanus)

ICESJournalofMarineScience

59 S43-S50

Mississippi

79

Chapter4

Seascapecontextmodifieshowfishrespondto

restoredoysterreefstructures

GilbyBL,OldsAD,HendersonCJ,OrtodossiNL,ConnollyRM,SchlacherTA

(2019)Seascapecontextmodifieshowfishrespondtorestoredoysterreef

structures.ICESJournalofMarineScience.InPress(doi:10.1093/icesjms/

fsz019).

80

Introduction

Habitatsthatprovidehighstructuralcomplexityandverticalreliefincoastalseascapesareoften

hotspotsforfishbiodiversity(Whitfield2017).Fishcongregatearoundstructurallycomplexhabitats

forprotectionfrompredation,accesstoalternatefoodsources,ortoshelterfromcurrents(Brooket

al.2018;Lenihan1999;MicheliandPeterson1999).However,thesehabitats,thatincluderocky

outcrops,logsnags,mangrovesandseagrasses,areamongthefirsttoberemovedwhenhumans

modifyaquaticecosystemsforshipping,shorelinestabilisation,urbandevelopment,orfishing

activities(Halpernetal.2008).Consequently,therestorationofstructurallycomplexhabitatsisa

techniquethathasbeenusedwidelyinaquaticecosystemstoaugmentoraggregatefishbiomass,

richnessorfisheriesvaluesfollowingdisturbancefromhumanactivities(Beckeretal.2018;Miller

2002;Ronietal.2002;Sunetal.2017;zuErmgassenetal.2016).However,somerestorationefforts

donotfullyconsiderhowtheplacementofsitescouldmaximisetheabundanceordiversityoffish

aroundthem,meaningthatthefullbenefitsofrestorationmightnotbeachieved(Gilbyetal.2018a;

JonesandDavidson2016a).

Theseascapecontextofhabitatsrelativetootherecosystems,togetherwithvariationinthesize

andqualityofhabitatpatches,structurethecompositionoffishassemblagesacrosscoastal

seascapes(Grober-Dunsmoreetal.2009;Pittman2018).Forexample,theproximityofmangrove,

logsnagsandrockyoutcropstonearbyseagrassshapesthecompositionoffishassemblagesacross

subtropicalestuarineseascapes(Gilbyetal.2018b).Similarly,theproximityofcoralreefsand

mangroveforestsmodifiestheabundanceanddiversityoffishintropicalreefseascapesacrossthe

PacificOcean(Oldsetal.2013)andCaribbeanSea(Mumbyetal.2004).Consequently,connectivity

betweenecosystemsisnowanimportantconsiderationinmarineconservationplanning(Hidalgoet

al.2016;Weeks2017).Intemperateseascapes,marshfragmentationinfluencesboththe

abundanceanddiversityoffishinnearbysubtidalchannels,andtheeffectsofthishabitat

transformationonfishandfisheriesarefurthercompoundedbytheimpactsofurbanisationandsea

levelrise(Rudershausenetal.2018;TorioandChmura2015).Inadditiontotheseeffectsonfish

assemblages,theseascapecontextofecosystemscanshapethedistributionofecologicalfunctions

incoastalseascapes(Martinetal.2018;Oldsetal.2018a),andmodifythebenefitsofrestoration

andconservationforbothbiodiversityandecosystemfunctioning(Grabowskietal.2005;Henderson

etal.2017b;Oldsetal.2016).Withrestorationprojectsnowestablishedformanyecosystems

globally,informationregardingthegrowthandhealthoftheserestoredecosystemsisoftenreadily

available,meaningthatwearenowbetterpoisedtodeterminehowtheplacementofrestoration

sitesinheterogeneousseascapesmightoptimisetheirperformance(Gilbyetal.2018a;Rudnicket

81

al.2012).Asaresult,wemightenhancetherecruitmentoffishtonewhabitats(eitherrestoredor

artificial)byoptimisinghabitatdetectability,accessibilityandlocationincoastalseascapes

(HuntingtonandLirman2012).

Oysterreefsarestructurallycomplexcoastalecosystems,whichprovidehighvaluehabitatforfish

assemblagesbecausetheyprovideabundantfoodandprotectionfrompredation,especiallyfor

juvenileandsmallerfish(Petersonetal.2003;zuErmgassenetal.2016).Oysterreefsare,however,

alsothreatened,withanestimated85%lostglobally(Becketal.2011),andupto96%lostinsome

regions(Diggles2013).Consequently,oysterreefrestorationhasbecomeawidespread

managementresponseglobally(Gilliesetal.2015b).Whilstthereareseveralpotentialbenefitsto

restoringoysterreefs,includingimprovedwaterquality,reducedsedimentationandincreased

nutrientsequestration(Coenetal.2007;Gilliesetal.2015a),manyprojectsexplicitlyseektorestore

reefstoenhancethevalueofcoastalseascapesforfishandfisheries(CoenandLuckenbach2000;zu

Ermgassenetal.2016).Previousresearchhasestablishedthattheseascapecontextofrestored

oysterreefscanaffectboththeabundanceanddiversityoffishinsurroundingseascapes(Grabowski

etal.2005).Forexample,connectionswithnearbymarshesstructuretheabundanceanddiversityof

macroinvertebrateassemblagesonoysterreefs,andtheratesatkeyecologicalfunctions,suchas

predation,areperformed(MicheliandPeterson1999).Thelong-termeffectsoftheplacementof

restoredreefsindifferentcontextscanoftenbespeciesspecific(Ziegleretal.2018).Onoccasion,

however,fishabundanceanddiversityisgreateronrestoredoysterreefsthatareisolatedfrom

otherecosystems,possiblybecausetheseisolatedreefsprovidenew,andstructurallycomplex,

habitatinlocationsthatwerepreviouslylowcomplexity,unvegetatedsoftsediments(Grabowskiet

al.2005;Ziegleretal.2018),andthatreefsplacednearotherbiogenichabitats(e.g.marshes)might

notbeaseffectiveasisolatedreefsforenhancingfishandcrustaceanabundance(Geraldietal.

2009).Theunderstandingsbeginningtoemergefromthesestudiesonseascapeeffectsforfishon

oysterreefeffectshavebeenpredominantlybuiltfromanarrowgeographicrange.Forexample

moststudiesoftheeffectsofseascapecontextonthefishassemblagesofnaturalorrestoredoyster

reefshavebeenconductedintemperatemarsh-dominatedseascapes,anditisnotclearwhether

similareffectsoccurinsubtropicalortropicalseascapeswheremangrovesdominate(Gilbyetal.

2018c).Effectsofrestoredoysterreefsinmangrove-dominatedsystems,especiallyinmeso-or

macrotidalareas,mightdiffertoeffectsinmarsh-dominatedseascapesduetodifferencesin

mangroveaccessibility,foodavailability,orprotectionfrompredators(e.g.Sheavesetal.2016).

82

WerestoredoysterreefsintheNoosaRiverinQueensland,easternAustralia,asystemwhereoyster

reefshadbeenfunctionallyextirpatedoveracenturyago,andusedtheseasamodelsystemtotest

whethertheseascapecontextofindividualreefsmodifiedtheiraugmentationoftheabundanceand

diversityoffish.Weanticipatedthatfishassemblageswoulddifferbetweenoysterreefsandnearby

unvegetatedlocations,whichweresurveyedascontrolsites,withgreaterspeciesrichnessand

abundanceofharvestablefishexpectedonoysterreefs.However,theoysterreefsalsodifferedin

termsoftheirpositionsrelativetoseagrassmeadows,andvariationinthedegreeofseagrass

connectivitycanstructurethecompositionoffishassemblagesacrossestuarineseascapes(Gilbyet

al.2018b).Consequently,wehypothesisedthattheeffectivenessofrestoringoysterreefsforfish

woulddependontheirproximitytoadjacentseagrassmeadows,butthatthedirectionthatthis

relationshipaffectsfishassemblageswouldbedependentonspecies-specificpatternsofseagrass

utilisation.

Materialandmethods

Studysystem

TheNoosaRiverisasubtropicalestuaryineasternAustralia(~24°S)(Figure1).Itsupportsa

heterogeneousmixofunvegetatedsandysubstrate,mangroves(mostlyAvicenniamarina)and

seagrassmeadows(mostlyZosteramuelleriwithleaflengths30-40cm),whichcontributestothe

diversityandabundanceoffishesinthisseascape(Gilbyetal.2018b).Mangrovesare,however,the

dominantcomponentoftheseascapeintermsofarea,withover200timesmoreaerialextentof

mangrovesthanseagrassthroughouttheNoosaRiver.Oysterreefswereonceabundantinthe

estuary,butdisappearedfromthesystemintheearlytomid1900s,likelyduetothecombined

effectsofoverharvesting,diseaseanddecliningwaterquality(Thurstan2016).

Oysterreefrestorationcommencedat14sitesintheNoosaRiverestuaryinNovember2017,andso

formpartofanactiverestorationeffortintheNoosaregion.Theprinciplegoaloftheoysterreef

restorationeffortistorestorestructurallycomplexhabitatsandtoenhanceseascapecomplexityfor

fish,includingforspeciesofcommercialandrecreationalsignificance.Thestructuresthatwere

installedtorestoreoysterreefswerecomprisedofthree1mlongby30cmdiameterbagsof2.5cm

gaugecoconutmeshfilledwithoystershellsstackedinatriangularprism.Eachreefsiteincluded

threeofthesestructures,whichwerearrangedinatriangle(with5msides)andpositionedatthe

depthoflowestastronomicaltide,sothatreefemergedatlowtideapproximatelytwiceannuallyin

thismesotidalestuary(tidalrange~1-2m).Naturallyoccurringoysterlarvae,whichremainpresent

83

insufficientnumbersintherivertoallowrestoration,settleonthereefstructures,andareexpected

tocementthemtogethertoformfunctioningoysterreefsafterapproximatelythreeyears.

Figure1Locationofrestoredoysterreefs,controlsitesandhabitatsintheNoosaRiver,easternAustralia.

84

Oysterreefsiteswerepositionedevenlyalongthelowerestuarywithinthehistoricrangeofoyster

reefsintheriver(Thurstan2016)(Figure1).Allreefstructureswereplacedwithin100mofadjacent

mangroves.Therefore,effectsfrommangrovesonfishcongregatingaroundoysterreefswere

consistentacrossallsites(Gilbyetal.2018b;Oldsetal.2012b).Bycontrast,reefstructuresdiffered

intermsoftheirconnectivitywithpersistentseagrassbeds(i.e.presentoverthelasttwoyears)

(Gilbyetal.2018b).Seagrassmeadowsinfluencethecompositionoffishassemblagesonrockyreefs

inestuariesacrosstheregionforupto500m;thedistancethatbestreflectsthedailyhomeranges

ofmanyfishspeciesthatinhabitoysterreefsinthearea(Gilbyetal.2018b;Oldsetal.2012b).

Consequently,oysterreefsweregroupedbasedonthepresenceorabsenceofseagrassnearby(i.e.

within500m).Weselected14controlsitesspreadthroughouttheNoosaRiverinthesamerangeas

oysterreefs,alwaysonunvegetatedsedimentsatleast200mfromoysterreefsandothercontrol

sitestoensurespatialindependence,andwiththesamenumberofseagrasssitespresentorabsent

nearbyasforoysterreefsites(Figure1).

Fishsurveys

Weusedbaitedremoteunderwatervideostations(BRUVS)tosurveyfishonoysterreefstructures

andatcontrolsites.BRUVSconsistedofa5kgweightwitha1mlengthofPVCpipetoattachbaits

atafixeddistanceof50cmfromaGoProcamerarecordinginhighdefinition(1080p).Baitswere

500gofcrushedpilchardsSardinopssagaxplacedina20x30cmmeshbagwith5mmopenings.

BRUVSweredeployedforaperiodof1hourateachsiteduringeachsurveyevent.Becausesalinity

isaprincipledeterminantforthedistributionoffishinestuaries,salinitywasmeasuredateachsite

BRUVSdeployment.AllBRUVSweredeployedinthecentreofoysterreefsites.Surveyswere

conductedduringtheday(0900-1600hrs)andwithintwohoursofhightidetominimisepotential

confoundingeffectsfromtidalanddielvariation.Wesurveyedfishassemblagesimmediately

followinginstallationofoysterreefs(time0),andthenat6weekintervalsforthreeadditional

events(i.e.n=4eventsintotal).Wesurveythesereefsinthefirst6monthsofinstallationwhenthe

valueofindividualreefsunitsforfisharesimilar,andhavenotyetbeensignificantlymodifiedby

differentialgrowthinpotentialfooditemsonthereefsthemselves(Gilbyetal.2018c).Further,as

thisrestorationeffortisconsideredanactiveanddevelopingrestorationeffort,andtherehasbeen

littlesettlementofinvertebratesamongsttheoysterreefunitstothispoint,wecanassumethatany

responsesfoundinthisstudyareduemostlytotheadditionofstructurallycomplexhabitattothe

estuary,asopposedtoanysignificantprey-orfood-availabilityeffects.Reefswerepurposefully

positionedinareasintheestuarywithintermediatetidalflowratesto1)reducescouringaroundthe

reefsbyhightidalflows,but2)stillprovideampleoysterlarvalrecruitmenttoreefs.Consequently,

85

thereisunlikelytobeanysignificantdifferencesinodourplumesfromBRUVSthatwouldbias

results.

FishassemblageswerequantifiedfromvideofootageusingthestandardMaxNmetric(seeGilbyet

al.2017a).Toaccountforvariablevisibility,fishwereonlycountediftheyswamthroughthefieldof

viewwithin1mofthecamera(asdeterminedbytheabove-describedlengthofPVCpipe).Visibility

wassufficienttousethis1mfieldofviewforallsurveys,hencetherewasnoeffectofwatercolumn

turbidityorvisibilityontheeffectivenessofoursurveys.Wecalculatedspeciesrichness(i.e.total

numberofspecies)andnumberofharvestablefish(fishedstatusaccordingtoFishBase,andequals

thesumofMaxNvaluesofallspeciesconsidered‘harvested’withintheregion)(FroeseandPauly

2018)foreachBRUVSdeployment.

Statisticalanalyses

Weusedpermutationalmultivariateanalysisofvariance(PERMANOVA)calculatedonsquareroot

transformedBrayCurtisdissimilaritytotestfordifferencesinthecompositionoffishassemblages

betweenoysterreefsandcontrolsites(fixedfactor,2levels),withvariationinthepresenceor

absenceofseagrassnearby(fixedfactor,2levels;seagrasspresentorseagrassabsentnearby),and

theinteractionbetweenthesefactors,whilstaccountingforsamplingperiod(covariable,4levels;0,

6,12,and18weeks).PERMANOVAresultswerevisualisedusingcanonicalanalysisofprinciple

components(CAP).Inusing‘samplingperiod’,ourfocuswasnottoascertainwhether‘timesince

restoration’resultedinanythresholdeffectsforfishassemblagesonoysterreefstructures(i.e.reef

assemblagesreaching‘maturity’),asthisislikelytotakemanyyearstodevelop(zuErmgassenetal.

2016),butrathertoaccountforpotentialdifferencesinenvironmentalconditionsbetweensampling

periods.PERMANOVAwasfollowedbyDufrene-Legendreindicatorspeciesanalyses(Dufreneand

Legendre1997)inthelabdsvpackageoftheRstatisticalframework(RCoreTeam2018)to

determinethefishspeciescontributingmosttodifferencesbetweenoysterreefsandcontrolsites.

Wethenusedgeneralisedlinearmodels(GLMs)inRtotestforcorrelationsbetweentheabundance

ofspeciesidentifiedfromindicatorspeciesanalyses,alongwithspeciesrichnessandtheabundance

ofharvestablespecies(i.e.thesumofMaxNofspeciesidentifiedasharvestedbyFroeseandPauly

2018).ThefactorsandmodelstructureofGLMswereidenticaltothoseusedinPERMANOVA

analyses.

86

Results

Fishassemblages

Werecorded42fishspeciesacrossallsurveys,with34speciesoccurringonoysterreefs,and31

speciesatcontrolsites(FigureS1).Ofthespeciesoccurringonreefs,12speciesoccurredexclusively

atreefs(TableS1);eightofthesearetargetedbylocalfisheries.Conversely,eightspeciesoccurred

onlyatcontrolsites,ofwhichthreespeciesaretargetedbylocalfisheries.Oysterreefssupported,

onaverage,1.4timesmorefishspecies(3.6+/-0.3se)thancontrolsites(2.6+/-0.2se)acrossall

samplingperiods(Figure2A).Similarly,oysterreefssupported1.8timesmoreharvestablefish(6.5

+/-0.7se)thancontrolsites(3.6+/-0.5se)acrossallsamplingperiods(Figure2B).Fishassemblages

onreefsweredominatednumericallybyestuaryperchlet(Ambassismarianus;Ambassidae),

yellowfinbream(Acanthopagrusaustralis;Sparidae),tarwhine(Rhabdosargussarba;Sparidae)and

southernherring(Herklotsichthyscastelnaui;Clupeidae),whichcomprised43%,27%,4%and3%of

totalfishabundancerespectively.Bycontrast,fishassemblagesatcontrolsitesweredominated

numericallybyestuaryperchlet,yellowfinbream,weepingtoadfish(Torquigenerpleurogramma;

Tetraodontidae)andstripedtrumpeter(Pelatesquadrilineatus;Terapontidae),whichcomprised

69%,27%,4%and3%oftotalfishabundancerespectively.

Effectsofseascapepositiononreeffish

Fishassemblagesdifferedbetweenoysterreefsandcontrolsites,andbetweensiteswhereseagrass

waspresentandabsentnearby(Figure3),butthesetwofactorsdidnotinteract(Table1).The

compositionoffishassemblagesalsodifferedamongsamplingperiods,withbothfewerfishspecies

andfewerharvestablefishesbeingrecordedonthefinalsamplingevent(Figure2A,B).Sampling

periodwassignificantlycorrelatedwithsalinity;salinitywasverylowacrosstheentireestuaryon

finalsamplingevent,followingheavyrainfallinthecatchment(Figure2C).Therefore,salinityvalues

werenotincludedinsubsequentanalyses.Whilstthespeciesrichnessofassemblagesandthe

abundanceofharvestablefishwaslowerduringthisfinalsampling,valueswerestillsignificantly

higheronoysterreefsthanatcontrolsitesforthreeofthefoursamplingperiods(Figure2).

Differencesinassemblagecomposition,speciesrichnessandfishabundancebetweenoysterreefs

andcontrolsiteswereprimarilydrivenbyvariationintheabundanceofyellowfinbream,moses

perch(Lutjanusrusselli;Lutjanidae),andsouthernherring,whichwereallmoreabundantand

occurredmoreoftenonreefsthanatcontrolsites(Figure3,Table2,TableS2).

Thepresenceorabsenceofseagrassnearoysterreefsdidnotaffectspeciesrichnessormodifythe

abundanceofharvestablefishesonoysterreefs(Figure4A,B,Table3,TableS2).Despitetherebeing

87

nointeractionbetweentreatmentandseagrasspresenceattheassemblagelevel,yellowfinbream

weremoreabundantatoysterreefsthancontrolsites,butonlywhenoysterreefswerenear

seagrass(Figure4C,Table3).Mosesperchweremoreabundantonoysterreefsthatatcontrolsites,

butwerealsocommonatbothoysterreefsandcontrolsitesnearseagrass(Figure4D,Table3).

Southernherringweremoreabundantonoysterreefsthanatcontrolsites,andwerenotaffected

bythepresenceorabsenceofseagrassnearby(Figure4E,Table3).

Figure2Average(+/-SE)(A)speciesrichness,(B)harvestablefishabundanceand(C)salinityvaluesforallsites

betweencontrolsandoysterreefsacrossthefoursamplingperiods.

0

1

2

3

4

5

6

7

0 weeks 6 weeks 12 weeks 18 weeks

ControlsOyster reefs

0

2

4

6

8

10

12

0 weeks 6 weeks 12 weeks 18 weeks

ControlsOyster reefs

0

5

10

15

20

25

30

35

0 weeks 6 weeks 12 weeks 18 weeks

Abun

danc

eR

ichn

ess

Salin

ity (p

pt)

C. Salinity

A. Species richness

B. Harvestable fish abundance

Weeks since installation

88

Figure3Canonicalanalysisofprinciplecomponentsvisualisingfishassemblagesaroundrestoredoysterreefs

intheNoosaRiver,Australiafordifferencesacrosstreatment,andwithseagrasspresentorabsentnearby.

VectoroverlaysarePearsoncorrelationsofpotentialindicatorspeciesfromDufrene-Legendreindicator

speciesanalysis.

Yellowfin bream

Southern herring Moses perch

Oyster reefs, seagrass presentOyster reefs, seagrass absentControls, seagrass presentControls seagrass absent

89

Figure4Average(+/-SE)of(A)speciesrichnessand(B)harvestablefishabundance,andtheabundanceof

significantindicatorspecies(C-E)atrestoredoysterreefsandcontrolsites,andwithseagrasspresentor

absentnearby(i.e.within500m).PlotsreflecttheresultsofGLManalyses(Table3).Starsabovecolumns

indicatesignificantdifferencesbetweenlevelsofthatfactorusingTukeypairwisecomparisons.Letteringon

panelCindicatessignificantdifferencescalculatedusingTukeypairwisecomparisonsontheinteraction

betweentreatmentandseagrasspresence.

0

0.2

0.4

0.6

0.8

Seagrass absent

Seagrass present

0

0.2

0.4

0.6

0.8

Control Oyster reef

01234567

Control Oyster reef Control Oyster reef

Seagrass absent Seagrass present

Abu

ndan

ceA

bund

ance

C. Yellowfin bream D. Moses perch

E. Southern herring

A. Species richness B. Harvestable fish abundanceR

ichn

ess

A. A.A.

B.

0

1

2

3

4

5

Seagrass absent

Seagrass present

0

1

2

3

4

5

Control Oyster reef012345678

Control Oyster reef

0

0.1

0.2

0.3

0.4

0.5

Control Oyster reef

Abu

ndan

ce

012345678

Seagrass absent

Seagrass present

0

0.1

0.2

0.3

0.4

0.5

Seagrass absent

Seagrass present

* * *

* *

*

90

Table1Permutationalmultivariateanalysisofvarianceoutputtestingassemblageleveldifferencesbetween

treatment(fixedfactor,2levels;oysterreefsorcontrolsites)andseagrasspresencenearby(fixedfactor,2

levels;seagrassabsentandseagrasspresentnearby),correctedforsamplingperiod(covariate).Pvaluesin

boldsignificantata=0.05.

Source df MS Pseudo-F P

Samplingperiod 1 10900 3.4 0.001

Treatment(Tr) 1 7771.4 2.5 0.009

Seagrasspresence(Se) 1 6963.7 2.2 0.018

TrxSe 1 2527.6 0.8 0.627

Res 107 3164.4

Total 111

Table2Indicatorspeciesanalysistestingforthespeciesthatcontributemosttowardsdifferencesin

assemblagesbetweenoysterreefsandcontrolsites.‘Group’indicatestheleveloffactor‘Treatment’forwhich

thatspeciesisasignificantindicator.

Species Group IndicatorValue PYellowfinbream Oysterreefs 0.56 0.001Mosesperch Oysterreefs 0.21 0.013Southernherring Oysterreefs 0.17 0.015

91

Table3Generalisedlinearmodelresultstestingforcorrelationsbetweenkeyattributesoffishassemblages

andsignificantindicatorspecies(seeTable1)withtreatment(fixedfactor,2levels;oysterreefsorcontrol

sites)andseagrasspresencenearby(fixedfactor,2levels;seagrassabsentandseagrasspresentnearby),

correctedforsamplingperiod(covariate).Pvaluesinboldsignificantata=0.05.

Generalfishassemblageindicators

Source d.f. X2 P

Speciesrichness

Samplingperiod 1 20.61 <0.001

Treatment(Tr) 1 9.92 <0.001

Seagrasspresence(Se) 1 0.1 0.74

TrxSe 1 0.02 0.87

Harvestablefishabundance

Samplingperiod 1 83.03 <0.001

Treatment(Tr) 1 44.22 <0.001

Seagrasspresence(Se) 1 9.87 <0.001

TrxSe 1 0.01 0.91

Indicatorsofoysterreefassemblages

Source d.f. X2 P

Yellowfinbream

Samplingperiod 1 75.71 <0.001

Treatment(Tr) 1 31.74 <0.001

Seagrasspresence(Se) 1 4.53 0.03

TrxSe 1 3.67 0.05

Mosesperch

Samplingperiod 1 0.11 0.74

Treatment(Tr) 1 8.19 <0.01

Seagrasspresence(Se) 1 4.73 0.03

TrxSe 1 1.93 0.16

Southernherring

Samplingperiod 1 0.27 0.61

Treatment(Tr) 1 9.93 <0.01

Seagrasspresence(Se) 1 0.05 0.81

TrxSe 1 1.46 0.22

92

Discussion

Habitatsareoftenrestoredincoastalseascapestoenhancefishabundanceanddiversity,orto

improvethefisheriesvaluesofestuaries(CoenandLuckenbach2000;GrabowskiandPeterson2007;

Hecketal.2003).Restorationsitesthatarestrategicallyplacedwithinheterogeneousestuarine

seascapesmightbemoreeffectiveforthesepurposes,butthetrajectoriesoftheserelationships,

andconsistenciesinpatternsacrossdifferenttypesofseascapesrequiresfurtherinvestigation(Gilby

etal.2018c).Inthisstudy,wefoundthatthefishassemblagesofrestoredoysterreefsweremore

diverseandcontainedahigherabundanceofharvestablefishthancontrolsites.Ourresultsalso

showthattheeffectsofrestorationforsomespecies,includingtaxathatwereatargetforoyster

reefrestoration,canbemodifiedbytheseascapecontextinwhichhabitatsarerestored.The

presenceorabsenceofseagrassnearbyaffectedtheabundanceofakeyharvestablefishspecies

(i.e.yellowfinbream)onoysterreefs,butnotthecompositionoffishassemblages,fishspecies

richnessorthetotalabundanceofharvestablefishes.Whilstthesefindingsareatthisstagelikely

onlyresponsesoffishtotheoysterreefstructuresthemselves,asopposedtoanybenefitsof

significantoystergrowth,thesefindingssuggestthattheecologicalbenefitsofrestorationforboth

fishassemblages,andspeciesthatareharvestedbyfishers,mightbeimprovedbyoptimizingthe

seascapecontextinwhichrestorationtakesplace.

Oysterreefsareagloballythreatenedecosystem(Becketal.2011)providingsignificantvalueto

surroundingfishassemblagesandassociatedfisheries(Grabowskietal.2012;HumphriesandLa

Peyre2015).Consequently,therestorationoffishbiomassand/orbiodiversityisoftenakeygoalof

oysterreefrestorationprojects(Gilbyetal.2018c;GrabowskiandPeterson2007;Petersonetal.

2003).Inthisstudywereportpositiveeffectsofoysterreefrestorationonbothfishabundanceand

diversity,evenearlyinthegrowthoftheserestoredreefs.Theseresultsconcurwiththefindingsof

severalstudies(e.g.Grabowskietal.2005;HardingandMann1999;LaPeyreetal.2014b)andtwo

seminalreviewpapers(Petersonetal.2003;zuErmgassenetal.2016),whichconcludethatoyster

reefrestorationcanaugmentfishdiversityandabundanceincoastalseascapes.Moststudieshave,

however,beenconductedinmarshandseagrassdominatedseascapesinthenorthernhemisphere

(Gilbyetal.2018c).Ourresultsshow,forthefirsttime,thatoysterreefrestorationcanalsobenefit

fishassemblagesinmangrove-dominatedseascapes.Thisisanimportantfindingbecausethescale

ofoysterreefrestorationisincreasingglobally,asisthewillingnessofcountriestoinvestinoyster

restoration(Gilliesetal.2015b),andweshowthattherestorationofoysterreefscanhavepositive

effectsonfishabundanceanddiversityinmultiplecoastalseascapes,evenatearlystagesofreef

development.Therestoredoysterreefsthatwestudiedwererelativelysmallandyoung,butour

93

resultsstilldemonstratethepotentialforthissortofseascape-scalerestorationtoenhancefish

assemblagesacrossestuaries.

Therestorationofoysterreefshadstrong,consistentandpositiveeffectsonbothfishspecies

richnessandthetotalabundanceofharvestablefishes.Theresponsesofindividualspecieswere,

however,morenuancedanddependedontheseascapecontextinwhichoysterreefswereplaced.

Thepresenceofseagrassnearbytosurveysitesmodifiedtheabundanceofakeyspeciesonoyster

reefsthataretargetedinlocalfisheries(i.e.yellowfinbream),andanotherharvestedspecies(i.e.

mosesperch)morebroadlyacrosstheentireseascape(Oldsetal.2018b;Webleyetal.2015).This

findingsuggeststhatecologicalbenefitsofoysterreefrestorationcandependonboththeseascape

contextinwhichoysterreefsareplaced.Placingoysterreefsinlocationswithdifferentseascape

characteristicsmight,therefore,improvethewidercapacityofrestorationprojectstoachieve

multiple,ordifferent,fisheriesenhancementgoals(sensuGilbyetal.2018c).Forexample,reefscan

beplacedinmultiplecontextstoachievegoalsforbotherosioncontrolandfisheriesbenefits,or

placedindifferentcontextstobenefitmultiplespeciesoffish.Thisisastrategicapproachtohabitat

restoration,whichwouldrequireexplicitgoals,andcarefulconsiderationofhowrestoration

performance(includingtothelevelofindividualfishspecies)mightbeshapedbythespatialfeatures

ofseascapes(Ehrenfeld2000;Gilbyetal.2018a;Guerreroetal.2017;Hallettetal.2013).

Theresponsesofyellowfinbreamandmosesperchtooysterreefrestorationandthepresenceor

absenceofseagrassnearbyindicatethenuancedresponsesofindividualspeciestorestorationin

differentcontexts.Inthissense,theeffectsfromseagrassoperatedinasymmetricalways,which

correspondedtodifferencesinthebiologyandecologyofthesespecies.Oysterreefsthatwere

closetoseagrassweremoreeffectiveforaugmentingtheabundanceofyellowfinbream,whichare

generalistzoobenthivoresthatcongregatearoundstructurallycomplexhabitats,includingoyster

reefsandseagrassmeadows,tofeedonoysterspat,otherepibenthicinvertebratesandfish(Brook

etal.2018;Oldsetal.2018a).Yellowfinbreamrecruittoseagrassmeadowsasjuvenilesandmove

tootherstructurallycomplexhabitatsasadultswheretheyaretargetedincommercialand

recreationalfisheries(Oldsetal.2012b;Webleyetal.2015),andarethereforeanimportantspecies

forwhichoysterrestorationintheregionseekstoenhance.Oysterreefswereeffectiveinenhancing

theabundanceofmosesperchrelativetocontrols,butthesefishwerealsocommonatallsitesnear

seagrass.Mosespercharegeneralistpiscivores,whichresideinstructurallycomplexmangrove

forests,oysterreefsandseagrassmeadowsinthesubtropicalestuariesofeasternAustralia(Oldset

al.2012b,Martinetal.2018).Theyrecruitintoestuariesasjuvenilesandmovetooffshorereefsas

94

adultsweretheyareimportanttargetsforfisheriesintheregion(Webleyetal.2015).Southern

herringareseasonalvisitorstotheestuariesofregion,andaggregatearoundhighreliefhabitats,

includingreefsandartificialstructures,toshelterfrompredatorsandfeedonarichsupplyof

plankton(WalthamandConnolly2013).Theyarealsoimportantpreyforlargerpredatoryfishes

(includingbothyellowfinbreamandmosesperch)inthestudyarea(MillerandSkilleter2006;Olds

etal.2018b).Thesefindingsdemonstratethattherestorationofoysterreefstructurescanimprove

boththehabitatandnurseryvaluesofestuarineseascapes,withfunctionaleffectsonplanktivores,

zoobenthivoresandpiscivoresthataresuggestiveofwiderbenefitsforcoastalfood-webs.

Determiningwhethertheseeffectscontinueasthereefsmature,andtheavailabilityofdifferent

sortsofpreychanges(i.e.oystersandotherinvertebrateswillincreaseindensity,andprovideother

feedingopportunities),isanimportantnextstepinquantifyingtheeffectsofseascapecontexton

thesereefs.

Theoverallqualityofestuarineecosystemsforfish,especiallyasfishnurseries,iscontingentonthe

presenceofamultitudeofhabitatswithinaheterogeneousseascape(Nagelkerkenetal.2015;

Whitfield2017).Whilsttheeffectsofseagrassnearoysterreefswerenotalwayspositiveinthis

study,thepresenceofseagrassinestuariescanpromotefishabundanceandrichness,andthe

fisheriesvaluesofestuariesintheregion(Gilbyetal.2018b;Pittmanetal.2004;Skilleteretal.

2017).Theeffectsofseagrassonthefishassemblagesofadjacentstructuralcomplexhabitatsoften

occuroverascaleof500m(Gilbyetal.2018b;Oldsetal.2012b).Theresultsofthisstudy,however,

suggestseagrasscanexertbothpositiveandneutraleffectsonthefishspeciescongregatingaround

oysterreefs.Thespatialscaleoverwhichseagrassinfluencesfishassemblagesonoysterreefswill,

therefore,beanimportantconsiderationforfurtherresearch.Theoysterreefswestudiedwere,

however,onlyrathersmallcomponentsofthebroaderseascapeinwhichtheywereplaced.Itwill,

therefore,beimportanttoascertainwhether,andhow,theseascapeeffectswereportscalewith

changesintherelativesizeofrestoredoysterreefs.Therearemultipleotheroysterreefrestoration

projectsinAustralia,varyingintermsofboththesizeoysterreefsbeingrestored(frommetresto

hectares)andtheproximityofseagrassbeds(AustralianShellfishReefRestorationNetwork2018),

whichcouldbeusedtotestthishypothesis.

Inthisstudy,wefoundthatrestoredoysterreefsintheNoosaRiverestuarycontainfish

assemblagesthataremorediverse,andcontainmoreharvestablefishthanatnearbycontrolsites.

Ourresultsindicatedspecies-specificeffectsofseascapepositioningonfishassemblagesthatmust

beconsideredwhendecidingonthelocationoffuturerestorationefforts.Here,reefscanbeplaced

95

inspecificcontextstotargettheenhancementindividualfishspecies,orinmultiplecontextsto

benefittwoormorespeciesthathavedifferenthabitatrequirements.Whilstthepatternswe

observedinthisstudyaresimplyanearlyindicationofthesuccessoftheoysterrestorationefforts

forfishinthisestuary,theyarelikelyaresponseonlytotheadditionofcomplexstructureto

previouslyunvegetatedsediments,andsoitisimportanttocontinuetotrackthesepatternsinfish

assemblagesasthereefsmatureandgrow.Inanycase,ourresultsareapositiveindicationofthe

potentialforrestoredoysterreefstoaugmentfishandfisheriesvaluesintheNoosaRiverand

beyond.If,giventheyoungageofthesereefs,fishsurveyedinthisstudywereonlyrespondingto

theactualstructureofthereefsthemselves,asopposedtoanystrongbenefitsassociatedwithfood

provisionfromthereefsasthereefsgrow,thenwecouldhypothesisethatthesepatternsin

augmentationofrichnessandharvestablefishwillsimplyincreasefurtherovertime.This,however,

willrequirefurtherinvestigation.Ourresultshaveconsequencesfortheplacementofstructurally

complexhabitatsinestuariesmorebroadly.Bystrategicallyplacingstructuresincoastalecosystems

toenhancethespecificcomponentsoffishassemblagesthatweseektoaugmentorcentralise,we

mightmoreefficientlyreachourconservation,restoration,orfisheries-relatedgoals.

96

SupplementaryMaterials

FigureS1Speciesaccumulationcurvesforrestoredoysterreefstructuresandnearbycontrolsitesinthe

NoosaRiver,Australia.

0

5

10

15

20

25

30

35

40

0 10 20 30 40 50 60

Spec

ies

coun

t

Samples

Oyster reefs Controls

97

TableS1ListofspeciesidentifiedatrestoredoysterreefstructuresandnearbycontrolsitesintheNoosa

River,Australia,andtheirfrequencyofoccurrence(FOC)atthesetwotypesofsites(outofatotalof56

possiblesurveyevents).Speciesinboldidentifiedexclusivelyonoysterreefs.‘Harvested?’indicatesspecies

classifiedastargetedbylocalrecreationalandcommercialfisheries.

Species Harvested? FOCatoysterreefs FOCatcontrolsAcanthopagrusaustralis(Yellowfinbream) Yes 49 35Ambassismarianus(Estuaryperchlet) 7 10Anguillareinhardtii(Longfineel) 5 0Aptychotremarostrata(Easternshovelnoseray) 0 2Arothronmanilensis(Narrowlinedpuffer) 7 8Atulemate(Yellowtailscad) Yes 0 1Carangoideschrysophrys(LongnoseTrevally) Yes 1 0Caranxignobilis(GiantTrevally) Yes 0 1Congerverreauxi(Congereel) 1 0Epinepheluscoiides(Estuarycod) Yes 2 0Favonigobiusexquisitus(Easternsandgoby) 6 5Gerresfilamentosus(Filamentousbiddy) 0 1Gerressubfasciatus(Silverbiddy) 12 6Glaucostegustypus(Giantshovelnoseray) 1 0Gymnothorapseudothyrsoideus(HighfinMoray) 0 1Hemitrygonfluviorum(Estuaryray) 1 5Herklotsichthyscastelnaui(Southernherring) 10 2Himanturauarnak(Reticulatewhipray) 7 8Hyporhamphusquoyi(Longtailgarfish) Yes 0 1Leiognathusequulus(Ponyfish) 2 0Lutjanusargentimaculatus(Mangrovejack) Yes 1 0Lutjanusfulviflamma(Blackspotsnapper) Yes 2 0Lutjanusrussellii(Mosesperch) Yes 16 4Marilynapleurosticta(CommonToadfish) 2 1Monodactylusargenteus(Butterbream) 12 4Mugilsp.(Mullet) Yes 3 1Pastinachusater(CowtailStingray) 1 2Pelatesquadrilineatus(Fourlinedtrumpeter) 6 7Platycephalusfuscus(Duskyflathead) Yes 1 2Pomadsyskaakan(Barredjavelin) 2 1Pomatomussaltatrix(Tailor) Yes 0 1Pseudolabrusguentheri(Gunther'swrasse) 1 0Pseudorhombusjenynsii(Smalltoothflounder) 1 0Rhabdosargussarba(Tarwhine) Yes 14 6Scatophagusargus(SpottedScat) 3 2Selenotocamultifasciata(Stripedscat) 1 0Siganusfuscescens(Rabbitfish) Yes 2 2Sillagosp.(Whiting) Yes 15 11Sphyraenaobtusata(Stripedbarracuda) 1 0Strongylurakrefftii(Longtom) 0 2Teraponjarbua(Crescentgrunter) 3 1Torquigenerpleurogramma(Weepingtoadfish) 3 8

98

TableS2Listofstudysites,andtheaveragevalues(withstandarderror(SE))andfrequencyofoccurrence(wherepossible;FOC)forspeciesrichness,harvestedfish

abundanceandsignificantindicatorspeciesforthefactortreatment.

Site Seagrasspresence/absence

Speciesrichness Harvestablefishabundance Yellowfinbream Mosesperch SouthernherringAverage SE Average SE FOC Average SE FOC Average SE FOC Average SE

OysterReef2 Present 2.8 0.9 3.0 1.1 3 2.9 0.5 1 1.0 0 1 1.1 0OysterReef3 Present 4.5 1.7 8.5 3.5 3 9.0 2.3 3 1.7 0.3 2 3.0 1.4OysterReef4 Absent 5.3 1.4 10.3 2.3 3 6.0 1.3 2 1.0 0 1 12.0 0OysterReef5 Absent 4.3 1.4 6.1 0.9 4 4.3 1.1 2 1.5 0.4 0 0.0 0OysterReef6 Present 3.0 0.8 2.8 0.6 4 1.0 0 3 1.3 0.3 1 1.0 0OysterReef7 Present 4.3 1.7 7.8 2.7 4 6.5 2.2 0 0.0 0 1 1.0 0OysterReef8 Absent 4.0 0.4 6.5 1.6 4 4.7 0.8 0 0.0 0 1 1.1 0OysterReef9 Absent 3.0 1.1 8.7 3.2 3 7.3 1.8 0 0.0 0 1 1.1 0OysterReef10 Absent 5.3 1.4 8.5 5 4 5.8 3 1 1.0 0 0 0.0 0OysterReef11 Absent 4.0 1.6 6.5 3 3 6.7 1.6 1 1.0 0 0 0.0 0OysterReef12 Present 4.0 0.7 6.8 3.8 4 5.0 3.7 1 2.0 0 0 0.0 0OysterReef13 Present 2.3 0.3 5.5 2.1 4 4.3 1.7 0 0.0 0 1 1.0 0OysterReef14 Present 3.5 0.9 7.0 2.8 4 6.0 2.2 1 1.0 0 1 1.1 0OysterReef16 Absent 1.0 0.4 3.7 1.9 2 6.2 0.7 1 1.0 0 0 0.0 0Control2 Present 1.8 0.9 1.5 1 1 1 0 0 0 0 1 1 0Control3 Present 4 1.4 6.3 2.2 3 4.3 1.5 2 2 0.7 0 0 0Control4 Absent 2.5 1 3 2.3 1 4 0 1 1 0 0 0 0Control5 Absent 2.5 0.6 4.3 1 4 3 0.7 0 0 0 0 0 0Control6 Present 1.5 0.6 0.8 0.3 2 1 0 0 0 0 0 0 0Control7 Present 2.3 0.9 3 2 2 4 1.4 0 0 0 0 0 0Control8 Absent 2.8 1.2 3.5 1.3 3 3.7 0.8 0 0 0 0 0 0Control9 Absent 3.3 0.8 6.8 1.8 4 6 1.5 0 0 0 0 0 0Control10 Absent 3.8 1 3.5 1 2 4 0.7 0 0 0 0 0 0Control11 Absent 1.8 0.3 0.8 0.5 2 1.5 0.4 0 0 0 0 0 0Control12 Present 2.3 0.5 2.5 1.6 2 4.5 1.8 0 0 0 0 0 0Control13 Present 2.5 0.6 4.5 1.3 4 3.5 1 0 0 0 0 0 0Control14 Present 5 0.7 7.3 1.8 4 5.5 1.6 1 2 0 1 1 0Control16 Absent 0.5 0.3 3 3 1 12 0 0 0 0 0 0 0

99

Chapter5

Landscapecontextmodifiestherateanddistributionof

predationaroundhabitatrestorationsites

DuncanC,GilbyB,OldsA,ConnollyR,OrtodossiN,HendersonC,SchlacherT

(2019)Landscapecontextmodifiestherateanddistributionofpredation

aroundhabitatrestorationsites.BiologicalConservation.Inpress.

100

Introduction

Themaintenanceofecosystemconditioniscontingentuponthepreservationofecologicalfunctions

thatenableecosystemstoresistorrecoverfromdisturbance(Deckeretal.2017;Risser1995).The

distributionofmanyecologicalfunctionsinlandscapescorrelateswiththepresenceorabundanceof

functionallyimportantspecies(BroseandHillebrand2016).Thesefunctionallyimportantspeciesare

underthreatfromhumanactivitiesinmanysettings(Vitouseketal.1997).Forexample,habitatloss

anddegradationhasresultedinthelossoffunctionallyimportantspecies(e.g.herbivoresand

predators)inmarine(Waycottetal.2009),freshwater(Quesnelleetal.2013)andterrestrial

ecosystems(Kormannetal.2016).Thiscanhaveknock-oneffectsfortheratesanddistributionsof

keyecologicalfunctionsinbothdisturbedhabitatpatches(Valiente-Banuetetal.2015),andin

surroundinglandscapes(Tylianakisetal.2010).Rehabilitatingorrestoringdegradedecosystemsis

anincreasinglyimportantmanagementinterventioninallmodifiedlandscapes(AertsandHonnay

2011;Bouleyetal.2018;Cosentinoetal.2014).Whilsthabitatrestorationhasinmanysettingsbeen

showntoincreasetheratesofkeyecologicalfunctions(Fraineretal.2018),itremainsuncommon

forrestorationprojectstoexplicitlytargettherestorationofmobileanimalsthatperformimportant

ecologicalfunctions(Gilbyetal.2018a).

Thepositionofrestorationsitesinlandscapesplaysapivotalroleinshapingtheassemblagesof

animalswhichcoloniserestoredhabitats,andtheratesofecologicalfunctionsthatanimalsprovide

(Belletal.1997;Gilbyetal.2018c;JonesandDavidson2016a;Laszloetal.2018).Restoringhabitats

atsiteswithhighconnectivitytonearbyecosystems,whichprovidealternativehabitatsorsource

populationsforanimals,canenhancerecruitmentintorestoredhabitats(PullingerandJohnson

2010;Volketal.2018;zuErmgassenetal.2016).Forexample,restoringcorridorsbetweenforest

patchesincreasesfaunalabundancebyfacilitatingspeciesmovementandsettlement(Leesand

Peres2008;Tewksburyetal.2002).Similarly,restoringhabitatpatchesinlocationswithconnections

tomanyhabitatpatchesofdifferenttypes,mightservetoenhancetheabundanceanddiversityof

animalsthatusemultiplehabitatsduringtheirlives(MicheliandPeterson1999;Nagelkerkenetal.

2015).Whilsttheprinciplesoflandscapeecologyareregularlysuggestedasimportant

considerationsinrestorationplans,theyarerarelyimplementedwhenselectingpossiblesitesfor

restorationactivities,withonly12%ofrestorationsitesgloballyhavingbeenplacedstrategicallyin

landscapestoenhancepossibleeffectsonanimals(Gilbyetal.2018a).Consequently,empiricaldata

thatcanbeusedtotestthefunctionaleffectivenessofrestorationindifferentlandscapecontextsis

limited.Moststudiesthathaveexaminedpossiblelandscapeeffectsofhabitatrestorationhave

focusedonchangesinanimalabundance,however,theabundanceofanimalsdoesnotalways

101

correlatewiththefunctionstheyperform(Bullocketal.2011;GamfeldtandRoger2017).

Quantifyingtheeffectsofrestorationindifferentlandscapecontextsanddeterminingwhether

thesechangesinspeciesabundanceproliferatetodifferencesinkeyecologicalfunctionsis,

therefore,pivotalforoptimisingthedesignandplacementofrestorationefforts(Gilbyetal.2018a).

Humanshavefundamentallytransformedmanycoastalseascapes(i.e.marinelandscapes)viathe

combinedeffectsofurbanisation,poorwaterquality,dredgingandfishing,andthesechangeshave

resultedinthelossordegradationofmanymarineecosystems(Halpernetal.2008).Consequently,

therestorationofcoastalecosystemshasbecomeanimportantfocusinmarinespatialplanning

(Barbieretal.2011),andenhancingtheabundanceofanimals(especiallyfishesandlarge

crustaceans)andecosystemfunctioningisaprimaryobjectiveformanyrestorationprojects

(Baggettetal.2015;zuErmgassenetal.2016).Oysterreefsareahighlythreatenedbutrestorable

ecosystem(Becketal.2011),consequentlyoysterrestorationprojectsarenowexpandingrapidlyin

numberglobally(Allewayetal.2015).Whilstoysterreefsarerestoredformultiplepurposes(e.g.

shorelinestabilisation,waterquality,returnoflosthabitats),oysterreefsprovideimportanthabitats

formanycoastalfishspeciesandareoftenrestoredtoaugmentfishabundanceanddiversity

(Baggettetal.2015).Oysterrestorationcanhavepositiveeffectsonfishassemblagesoverwhat

werepreviouslyunstructuredsediments(Grabowskietal.2005;HardingandMann1999;Peterson

etal.2003),however,thepossiblebenefitsofoysterrestorationforecologicalfunctionshaverarely

beentestedwithempiricaldata(Gilbyetal.2018c;Smythetal.2015).Thelandscapecontextof

oysterreefscanmodifythecompositionoffishassemblages,bothoverreefsandinsurrounding

temperate(Grabowskietal.2005;MicheliandPeterson1999),andsubtropical(Gilbyetal.2018b)

seascapes,butthereisnodatatodescribewhethertheseeffectsalsomodifythespatialdistribution

ofecologicalfunctions(Gilbyetal.2018c).

Restorationprojectsoftenseektoenhancetheconditionofecosystemsandthediversityor

abundanceofanimalsthatusetheseecosystemsashabitat(JonesandDavidson2016a;Middendorp

etal.2016).Manyrestorationeffortsalsoaimtopromoteecologicalfunctions,butthepotential

functionaleffectsofrestorationarerarelymeasuredormonitored.Thisstudyquantifiedtheeffects

ofoysterreefrestorationontherateanddistributionofpredationinanestuarineseascape.

Predationisanimportantecologicalfunctionthathelpstomaintaincommunitystructureinall

ecosystems(Estesetal.2010;RippleandBeschta2012;RitchieandJohnson2009).Quantifyingrates

ofpredationaroundhabitatrestorationprojectsisimportantbecausepredationissignificantly,and

quickly,modifiedbytherapidcolonisationofpredatorstorestoredcoastalecosystems(Harding

102

1999;MicheliandPeterson1999;Petersonetal.2003)andpredatorsaresensitivetoecosystem

changesastheyrelyonpreyavailabilitytosurviveandreproduce,andsoaregoodindicatorspecies

forthispurpose(Gilby2017;González-TokmanandMartínez-Garza2015).Weaimedtodetermine:

1)thedegreetowhichoysterreefrestorationenhancesthefunctionofpredationatrestoration

sites;2)thedistanceoverwhichpredationextendsintotheseascapesurroundingrestoredoyster

reefs,3)howtheseascapecontextofrestoredoysterreefsmodifiestheireffectsonecological

functions,and4)theidentityofthespeciesperformingthefunction.Wesurveyedratesofpredation

atsixrestoredoysterreefs,andintheseascapesurroundingeachreef,whichdifferedintermsof

theirproximitytonearbyseagrassmeadowsandmangroveforests.Wehypothesisedthatoyster

reefrestorationwouldenhancepredationratesbothonreefsandinthesurroundingseascape

(relativetonearbycontrolsites)andexpectedthatthesefunctionaleffectsofrestorationwould

dependonthespatialcontextofoysterreefsrelativetootherhabitats(e.g.seagrass,mangroves)

thatprovidehigh-reliefandstructurallycomplexhabitatsforfish.

MethodsStudysystem

ThisstudywasconductedintheNoosaRiver;asubtropicalestuary(~24°S)ontheeastcoastof

Australia.TheNoosaRiverseascapeiscomprisedofmangroveforestsandseagrassmeadows,

interspersedamongstamatrixofunvegetatedsandysubstrates(Figure1).Oysterreefswere

historicallyabundantintheNoosaRiver,butbecamefunctionallyextinctintheearly1900’s

(Thurstan2015).OysterreefswererestoredintheNoosaRiverinNovember2017,withaprinciple

aimtorestorestructurallycomplexhabitats(i.e.relativeto‘lowcomplexity’unvegetatedmudsor

sands)andtoenhanceseascapecomplexityforfish.Reefswereconstructedusingcoconut-fibre

meshbags(1mlongx30cmdiameterwitha2cmaperture)filledwithrecycledwholeSydneyrock

oyster(Saccostreaglomerata)shell.Eachoysterreefsiteiscomprisedofnineoysterreefbags

stackedinthreepilesofthreeat5mdistances,forminganequilateraltriangleandpositioned

intertidallytoabuttheleveloflowestastronomicaltide,andweresitedwithinthehistoricalrange

ofoystersinthisseascape(Thurstan2015).Thesereefsprovideoysterlarvae,whichoccurnaturally

inthissystem(TheNatureConservancy&EcologicalServiceProfessionals2015),aplacetosettle

andgrow.Overtime,oysterswillgrowandcementoystershellstogethertoformfullyfunctioning

oysterreefs.

Inthisstudy,surveysandexperimentswereconductedatsixoysterreefrestorationsites,which

werechosentorepresenttherangeofseascapecontexts(especiallywithrespecttotheareasof

103

nearbymangroveforestsandseagrassmeadows)availablewithinthissystem(Figure1)(Gilbyetal.

2017b).Thesesitesalsorepresentagradientofsalinityandlightpenetrationintheriver.These

metrics,therefore,werethekeyenvironmentalmetricstestedinthisstudy(Table1).Thesemetrics

havebeenshowninpreviousstudiesintheregiontobeimportantdriversofthedistributionand

diversityofestuarinefish(Gilbyetal.2018b).Sixcontrolsiteswereselectedonthebasisthatthey

hadthesamesuiteofseascapecontextsasreefsites,butwereatleast200mfromeachoysterreef

site(Figure1).

Figure1Distributionofrestoredoysterreefs,controlsites,andothermarinehabitatsintheNoosaestuary,

easternAustralia(spatialdatasources:QLDDepartmentofEnvironmentandHeritageProtectionandGilbyet

al.(2018b)).

104

Table1Environmentalmetricsincludedinstatisticalmodels,theirdefinitionsandsources(whereapplicable).

Variable Definition SourceDistancefromreef Pre-establisheddistances(i.e.camera

positions)fromeachsite(i.e.0,1,2,5,10,15,

20,25m)inmetresfromthecentreofthe

restoredreef.

-

Seascapecontext

Seagrassarea

Theareacoveredbyseagrassinmetres

withina500mdistancebuffer.Thiswas

selectedbasedonthelikelyhomerangesof

fishintheseestuarinesystems(Oldsetal.

2012b).

(Gilbyetal.2018b;Olds

etal.2012b)

Mangrovearea

Theareacoveredbymangrovesinmetres

withina500mdistancebuffer,selectedbased

onthelikelyhomerangesoffishinthese

estuarinesystems(Oldsetal.2012b).

(Oldsetal.2012b)

GIS,habitatlayersfrom

(Oldsetal.2012b;

Queensland

Government2016)

WaterqualitySalinity Thesalinity(psu)ofthewaterateachsite,

quantifiedusingarefractometeratthetime

ofdeploymentofeverycamera.

-

Secchidepth(turbidity) Themeasureofwatercolumnlight

penetration(m)ateachsiteusingastandard

30cmdiametersecchidiskatthetimeof

deploymentofeverycamera.

-

105

StudyDesign

Thefundamentalpremiseofourstudydesignwastoquantifythefunctionofpredation,andthe

speciesperformingthefunctionofpredationatreefssitesandnearbycontrols,andwithincreasing

distanceawayfromthesesites.Todothis,wequantifiedpredatoryfishassemblagesandpredation

usingeightunderwatercameraunitsatpre-establisheddistances(i.e.camerapositions)fromeach

site(i.e.0,1,2,5,10,15,20,25m)(FigureS1).Thesedistanceswereselectedtomirrorthescaleof

functionaleffectsfromsimilar-sizedartificialreefsinothercoastalecosystems(e.g.Hendersonetal.

2017;Jelbartetal.2007;Laymanetal.2016;Skilleteretal.2017).Thefirstdeployment(i.e.at0m)

waslocatedinthecentreofeachoysterreefandcontrollocation.Allotherdeploymentswere

placedatrandomanglesseawardfromtheoysterreefsite,attheappropriatedistancefromthis

centrepoint(FigureS1),ndwerealwaysplacedonunvegetatedsandsormuds(i.e.notinseagrass

meadowsormangroveforests).Surveysandpredationexperimentswereconductedonfour

occasions:immediatelyafterreefswereinstalled,andat2,4and6monthspostinstallation(i.e.

surveyperiods).Assurveyswereconductedinthefirst6monthspostinstallation,individualreefs

unitswereinasimilardevelopmentalstageandwerenotyetsignificantlyalteredbythesettlement

andgrowthofoysters,orotherinvertebrates.Theeffectsthatwereport,therefore,representa

responseoffishtotheadditionalstructurallycomplexhabitatsthatreefsprovide,ratherthanany

effectofvariablefoodavailabilityamongreefs.Insurveyingoursitesacrossmultiplesurveyperiods,

ourintentwasnottoquantifyanyeffectsof‘timesincerestoration’astheseoysterreefswilltake

manyyearstodevelopandgrow.Rather,wesoughttoaccountfordifferentenvironmental

conditionsbetweensamplingperiods.

FishSurveysandPredationExperiments

Weused‘squidpops’attachedtocameraunitstoquantifyratespredationandidentifypredators

aroundoursites.Squidpopsarenowastandardmethodforindexingrelativepredationratesof

marinemeso-predators,andhavebeenusedextensivelyforthispurposeincoastalseascapes(Duffy

etal.2015;RodemannandBrandl2017).Squidpopsconsistofasingle1cm2pieceofdriedsquid

mantletetheredtoa20cmlongbamboostakeusinga10cmlengthoffishingline.Cameraunits

werecomprisedofa5kgweightwithaGoProcamerarecordinginhighdefinition(1080p).The

squidpopstakewasthenfastenedtothecameraunitusinga15mmgaugePVCarmatadistanceof

45cmfromthecamera,sothatthesquidpopwasvisibleatalltimes(FigureS2).Atotalof384

squidpopdeploymentswereconductedduringthisstudy.Allcameradeploymentsweremadeon

unvegetatedsubstrate(i.e.notinseagrassmeadowsormangroveforests)andwereconducted2

hourseithersideofhightidetomaximisewatervisibilityandaccessibilitytooysterreefsites(Gilby

106

etal.2017a).Eachcameradeploymentwas1h;deploymenttimeswereselectedbasedonapilot

studyintheNoosain2017.Squidpopswereconsumedbypredatorswhenthesquidpiecehadbeen

entirelyremovedbyafish.Theidentityoffishpredatorswasdeterminedbyviewingvideofootage

fromeachdeployment.

DataAnalyses

Ouranalyticalapproachcomprisedthreekeysteps.First,weusedalogisticregressioninR(R-Core-

Team2017)todeterminehowtheeffectoftreatment(fixedfactor,twolevels;oysterreefsand

controlsites)andsurveyperiod(fixedfactor,fourlevels;event1,2,3,4)influencedpredationrates

acrossalldeployments(abinomialresponsevariable,1=squidpopconsumed,0=squidpopnot

consumed).Secondly,weusedgeneralisedadditivemodels(GAMs)inthemgcvpackage(Wood

2018)toexaminecorrelationsbetweenpredationratesandourenvironmentalmetricsofinterest

(Table1).WecalculatedmodelswithallpossiblecombinationsoffactorsusingMuMInpackage

(Barton2018),andselectedthebestfitmodelusingAikaike’sInformationCriterion(AIC).Relative

factorimportancewascalculatedforeachvariablebytakingthesumofweightedAICvaluesforall

modelsinwhichthatfactorwasincluded(Burnham2002).Finally,weusedgeneralisedlinear

models(GLMs)toquantifyhowthevariablesfromthebestfitmodelinteractedwitheachother.

Here,weanalysedinteractionsbetweenpairsofvariablesfromthebestfitmodelseparately(as

opposedtocalculatingallcomparisons,includingthreeorfourwayinteractionsinthesamemodel)

toavoidoverinterpretingtheseinteractioneffectsgiventhenumberofreefsites(n=6)sampled.

Results

Habitatrestorationenhancesecologicalfunction

Wemeasuredpredationtobe212%higheratoysterreefs(n=106events)thanatcontrolsites

(n=50events).Sixspeciesconsumedsquidpops(Figure2A):yellowfinbream(Sparidae;

Acanthopagrusaustralis),narrow-linedpuffer(Tetraodontidae;Arothronmanilensis),butterbream

(Monodactylidae;Monodactylusargenteus),commonponyfish(Leiognathidae;Leiognathus

equulus),mudcrab(Portunidae;Scyllaserrata)andyellowfintripodfish(Triacanthidae;Tripodichthys

angustifrons).Predationwasdominatedbyyellowfinbream,whichconsumed88%ofall

deployments,followedbynarrow-linedpuffer(7%)andbutterbream(2%)(Figure2b).The

likelihoodofpredationwassignificantlyhigheratoysterreefsthanatcontrolsites(x2=34.5,P<0.001)

(Figure2c).Thelikelihoodofpredationalsoincreasedsignificantlywithsurveyperiod(x2=34,

P<0.001)(Figure2d).However,thesefactorsdidnotinteractsignificantly.Thehighratesof

predationatrestoredoysterreefsmirroredthedistributionandfeedingactionsofyellowfinbream,

107

whichweremoreabundantatoysterreefsthanatcontrollocations(Figure2b).Bycontrast,the

diversityofpredatorswasgreateratcontrolsites(n=5species),thanatrestoredoysterreefs(n=3

species).Giventhestrong,andconsistenteffectsofoysterreefsonpredationrates,allsubsequent

analysesconsideredreefsitesandcontrolsitesseparately.

Figure2(a)SixpredatorsconsumedsquidpopsintheNoosaRiver,Australia.(b)Piechartsrepresentthe

proportionofsquidpopsconsumedbyeachpredatoratrestoredoysterreefsandcontrolsitesduringeach

surveyperiod(colourscorrespondtothoseusedinpanela).Numbersinsidepiechartsrepresentthetotal

numberofpredationeventsforthattreatmentbysamplingperiodcombination(outof48totaldeployments).

(c)Probabilityofpredationatrestoredoysterreefsandcontrolsiteswithinonehourfromdeployment.(d)

Probabilityofpredationacrosssurveyperiods(1,2,3,4)withinonehourfromdeployment.Imagesbythe

authorsandB.Sarangi(CC1.0).

108

Table2Bestfitmodelonprobabilitiesofpredationaroundoysterreefsitesfromgeneralisedadditivemodel

outputwithimportancevalues.Thebestfitmodelhadanr2valueof0.26.

Valuesinboldaresignificantat

p=0.05.

Table3Summaryofgeneralisedlinearmodelstestingforrelationshipsbetweentheprobabilityofpredation

andthreepredictors:Distancefromreef(D),Seagrassarea(S)andMangrovearea(M).Valuesinboldare

significantatp=0.05).

Source Variableimportance ChiSq P

Distancefromreef 0.99 10.99 <0.001

Mangrovearea 0.95 23.82 <0.001

Seagrassarea 0.57 4.46 0.035

Source df x2 P

Seagrass

Distancefromreef(D) 1 9.12 0.003

Seagrassarea(S) 1 23.23 <0.001

DxS 1 2.55 0.11

Mangroves

Distancefromreef(D) 1 9.12 0.003

Mangrovearea(M) 1 24.59 <0.001

DxM 1 0.67 0.41

109

Seascapecontextshapestheeffectsofhabitatrestorationonecologicalfunction

Thelikelihoodofpredationatoysterreefswasbestexplainedbyacombinationofthedistanceof

surveysitestotheoysterreefs,andtheareaofmangrovesandseagrassintheseascapesurrounding

theoysterreefs(Table2,Figure3).Noothervariableswereincludedinthebestfitmodel(Table3).

Theprobabilityofpredationdecreasedwithincreasingdistancefromrestoredoysterreefs(Figure

3A),andwasalsoloweradjacenttoreefsthatwereborderedbyalargerareaofmangrovesor

seagrasses(Table2,Figure3B,C).Atcontrolsites,thelikelihoodofpredationwasbestexplainedby

acombinationofsalinityandareaofseagrassandmangroves(TableS1).Thebestfitmodelfor

controlsitesdidnotincludedistancefromreefs,therebyconfirmingthattheeffectsfoundatreefs

wereduetotherestorationofreefs,andnotanyeffectsof‘baitattraction’.Thepresenceof

mangrovesorseagrassesintheseascapesurroundingrestoredoysterreefsdidnotmodifytherate

atwhichtheprobabilityofpredationdeclinedwithdistancefromindividualreefs(i.e.therewasno

interactionbetweendistancefromreefandareaofadjoininghabitat)(Table3,Figure4).Predation

rateswerehigheratreefswithoutnearbymangrovesorseagrassesandthisremainedconsistent

withincreasingdistancefromtheoysterreef(Figure4).

110

Figure3Generalisedadditivemodel(GAM)outputsillustratingrelationshipsbetweentheprobabilityof

predationand(a)distancefromreef,(b)mangrovearea,and(c)seagrassarea.Greyshadedpolygons

represent95%confidenceintervals.

A.

B.

C.

111

Figure4Generalisedlinearmodel(GLM)outputsillustratingrelationshipsbetweentheprobabilityof

predationand(a)seagrassarea,(b)mangrovearea,categorisedintolowandhighbynaturalsplitsinthedata

asdistancefromreefincreases.Greyshadedpolygonsrepresent95%confidenceintervals.Notethatthe

analysesunderlyingthesegraphswererunontwocontinuousvariables(distancefromreef,andareaof

habitatwithin500mofthereefs),butareavaluesareshownhereascategoriesforeaseofillustration,with

categoriesbasedonnaturalsplitsinthevaluesofhabitatextent(formangroveslow=397489.12m2-

459372.51m2,high=463206.33m

2-549056.19m

2;forseagrass,low=seagrassabsentwithin500m,

high=seagrasspresentwithin500m).

B.

A.

Distance from reef (m)

Distance from reef (m)

Prob

abili

ty o

f pre

datio

nPr

obab

ility

of p

reda

tion

High seagrass extentLow seagrass extent

High mangrove extentLow mangrove extent

112

Discussion

Animportantgoalofmanyhabitatrestorationprojectsistore-establishanimalpopulationsandthe

ecologicalfunctionstheyperform(CoenandLuckenbach2000;Kaiser-Bunburyetal.2017;Miller

2002).Thepotentialeffectsofhabitatrestorationonecologicalfunctionsare,however,rarely

quantified.Ourresultsshowthatoysterreefrestorationenhancestheecologicalfunctionof

predationonpreviouslyunstructuredsubstratesandindicatethatthespatialcontextofrestored

ecosystemscanshapetheirfunctionaleffectsincoastalseascapes.Inthissense,highseascape

complexity,characterisedinthisstudybyextensiveareasofnearbyseagrassesormangroves,

reducestherateofpredationrelativetositeswithlowerextentsofmangrovesorseagrassesnearby.

Similarly,wefoundthattheseeffectsremainedconsistentwithincreasingdistanceintothenearby

lowcomplexityunvegetatedareasandthattherewasnoeffectofnearbyseascapeelementsonthe

rateofdeclineofpredationrateswithincreasingdistance.Wesuggestthattheseeffectsarelikely

becausethevalueoftherestoredoysterreefsforfishiscontingentonthequalitycontrastbetween

thereefandthesurroundinghabitatpatches.Here,wehypothesisethatthefishareusingthereefs

asacentralpointoftheirhomerangeandperformingfeedingexcursionsradiallyfromthereefat

relativelyfixeddistances.Wherethereareexistinghigh-qualityhabitatsneartorestoredoyster

reefs,fisharelesslikelytomovetoanewstructuredhabitat,therebyreducingtheaugmentation

effectofrestorationefforts(Gilbyetal.2019).Thesefindingssuggestthatplacingrestorationsitesin

strategicallyselectedlocationswithinheterogeneouslandscapescanmaximisetheratesof

ecologicalfunctionatrestorationsites,andthedistanceoverwhichtheyextendawayfromthe

restorationsite.

Thecolonisationoffunctionallyimportantspeciestorestorationsitesiscontingentonthebenefits

suppliedbyrestoredhabitatsrelativetoalternativehabitats(JonesandDavidson2016a;zu

Ermgassenetal.2016).Furthermore,restoringecosystemstoenhancethediversityofanimals

performingthefunctions(i.e.functionalredundancy)canincreasethecapacityofecosystemto

withstandthelossofindividualspecies(MicheliandHalpern2005).Inthisstudy,predationwas

significantlyhigheratoysterrestorationsitesthanatnearbyunvegetatedcontrolsites,irrespective

oftheirlandscapecontext.Thispositiveeffectofhabitatrestorationonecologicalfunctionsislikely

duetofishcongregatingaroundstructurallycomplexhabitatsforforaging,refuge,spawningand

dispersalacrossdifferentlifestages(BlaberandBlaber1980;Whitfield2017).Previousstudieshave

shownthatthiscanextendtothefunctionsthatcongregatingspeciesprovide(Laymanetal.2013;

Laymanetal.2014;Oldsetal.2018a),thoughthisisrarelyquantifiedforrestorationactions(Gilby

etal.2018a).Despitethesepositiveeffectsofrestoration,wefoundrelativelylowfunctional

113

redundancyinthecompositionofpredatorassemblagesonreefs.Previousstudiesonboththe

effectsofrestorationonratesofpredationand/ortheavailabilityofpreyresources(Micheliand

Peterson1999;Ziegleretal.2018)haveillustratedsimilarpatternsinregardstothevalueof

restoredoysterreefsforthesurvivalofvariousmacroinvertebratespecies(i.e.crustaceans,bivalves

andgastropods)acrossdifferentlifestages.Predationonreefswasdominatedbyyellowfinbream,a

voraciousgeneralistconsumerinthisestuary(Oldsetal.2018a;Pollock1982),whichaggregates

aroundavarietyofstructurallycomplexhabitats,includingoysterreefsandartificialstructures

(Gilbyetal.2018b).Yellowfinbreammightcompetitivelyexcludeotherpotentialpredatorsfrom

restoredoysterreefsbecausetheyareabundantandaggressivepredatorsinthissystem,aneffect

thatwouldbeexacerbatedonreefsthatareatearlysuccessionalstages,andwhichmightprovide

onlyalimitedpoolofresourcestobepartitionedbymobileconsumers.Giventhatthediversityof

predatorswashighestatunstructuredcontrolsites,andisalsolikelytoincreaseatoysterreefsover

time,itwillbeimportanttodeterminehowthesespatialandtemporalchangescombinetoshape

functionalredundancyinthecompositionofpredatorassemblagesacrossthiscoastalseascape.

Conventionally,highseascapeheterogeneityisthoughttoenhanceecologicalfunctioningincoastal

habitatpatchesduetoincreasedconnectivitybetweenadjacenthabitats(MicheliandPeterson

1999;Pottieretal.2009),includinginsystemswithinourstudyregion(Gilbyetal.2018b;Oldsetal.

2012b).Weshow,however,thatthisisnotalwaysthecaseforrestoration.Previousstudieshave

identifiedgreaterfishabundanceanddiversityatoysterreefrestorationsitesmoreisolatedfrom

nearbymarshesbecausetheyprovidenew,structurallycomplexhabitattopreviouslylow-

complexitymuddysubstrata(Geraldietal.2009;Grabowskietal.2005).Ourfindingsshowthat

theseeffectsofrestorationonfishassemblagescanalsoextendtokeyecologicalfunctions.Where

existingcomplexhabitatsareavailable,fishmaybelesslikelytomigratetonewlyrestored

ecosystemsnearbybecausetheymayprovidelessfood,haveahigherriskofpredation,orare

energeticallymoreexpensivetoreachorinhabit(GrabowskiandPowers2004;Ziegleretal.2018).

Bycontrast,fishwillalsomovesomedistanceoverlowcomplexityhabitattocongregatearound

restoredstructures(Gregalisetal.2009;IrlandiandCrawford1997).Inthissense,thevalueofa

restoredhabitatpatchinacoastalseascapeiscontingentuponthecontrastbetweenthevalueof

therestoredhabitatandthevalueofthehabitatsimmediatelysurroundingtherestorationsite.In

addition,itislikelythattheoverallvalueoftherestoredoysterreefsforfishisalsoshapedbythe

combinedeffectsof1)thelevelofconnectednessthattherestoredhabitatshavewithalternate

habitats,2)themovementcapacityofthefishwithinthesystemandthescalesoverwhichthey

move,and3)therelativepredationriskfeltbythefishmakingmovementstothenewreefs.

114

However,therestoredoysterreefsinthisstudywererelativelysmallstructuralcomponentsinthe

overallseascape,particularlyincomparisontotheextensivenearbyremnantseagrassand

mangrovehabitats.Theeffectsoflandscapecontextonverylargerestorationsites(i.e.10sofm2)

mightbedifferenttothosefoundherebecausetheeffectofreefsdrawingfishawayfromnearby

naturallyoccurringstructurallycomplexhabitatsmightbegreaterwithalargerrestorationfootprint.

Seekingconsistenciesintheseeffectsonrestorationfordifferentrestoredreefdesigns,inmultiple

settings,andacrossmultiplefunctionsisthereforeanimportantresearchgap.

Habitatrestorationcanaffectthedistributionofanimalsandthefunctionstheyprovidebeyondthe

footprintoftheecosystemsthathavebeenrestored(Gilbyetal.2018c).Ourresultsindicated

significantdeclinesinpredationwithincreasingdistancefromrestoredreefsandshowthatthe

trajectoryofthiseffectwasconsistentirrespectiveofthecomplexityoftheseascapearoundeach

reef.However,wedidnotidentifyadistanceatwhichtheeffectsofthereefonthefunction

plateaued.Thisissurprisingbecausetheeffectsofartificialcinderblockreefs,whichareofasimilar

sizetoouroysterreefs,havebeenshowntoextendforonly15mintothesurroundingseagrass-

dominatedseascape(Laymanetal.2016).Previousstudiesonreefsinthisregionhaveshownthat

thescaleofseascapescanmodifytheeffectsofconnectivityonfishassemblages.Inthissense,the

effectswefindheremightnotconfertoseascapesofalargerscales,andtheremightbeagiven

seascapescale(likelymanyhundredsofmeters)overwhichthresholdsoccurontheeffectsof

seascapeconnectivity(Martinetal.2015;Oldsetal.2012b).Themainpredatoryfishinoursystem

mighthavelargerhomerangesthanthoseinotherstudies(Pollock1982),andthelackofextensive

structurallycomplexseagrassesaroundourreefsmightresultinfishbeinglesstightlyassociated

withthesereefsthantheywereinotherstudies.Alternatively,theseeffectsmightbeduetohigh

levelsofnaturalpredationacrossallecosystemsintheseascapewestudied(Baueretal.2010;Foam

etal.2005).Thisisunlikely,sincepredationrateswereloweronreefssurroundedbylowcomplexity

seascapesthanonreefsthatwereborderedbycomplexhabitats,andthesetrajectoriesdidnot

interceptatthefurthestdistancesurveyed.Itisthereforemorelikely,thatthedistancesfromreefs

thatwesurveyedweresimplynotfarenoughtodetecttheseeffects,andsofurtherstudyisneeded

tounderstandthedistanceoverwhichthefunctionaleffectsofrestoredoysterreefsextendinto

surroundingseascapes.Whilstwefoundsignificantlyhigherratesofpredationwithincreasingtime

(i.e.surveyperiod),itisdifficulttoconcludethattheseeffectswereduetotherestorationefforts

because1)wedidnothaveasignificantinteractionbetweensurveyperiodandtreatment,and2)

thereareseveralseasonalandenvironmentalconsiderationsthatarelikelytooverwhelmthese

‘time’effects.

115

Inthisstudy,weshowthattheaugmentationofecologicalfunctionsatrestorationsitesis

contingentupontheirpositioninheterogeneouslandscapes,andhowfunctionallyimportantspecies

respondtotheselandscapepatterns.Inaddition,weshowthattherestorationofalosthabitattoa

lowcomplexity,unvegetatedareaincoastalecosystemscanresultinkeyecologicalfunctionshaving

afootprintthatextendssignificantlybeyondtherestorationsiteitself.Ifthespeciesweidentifiedas

predatorsinthisstudyweresimplyrespondingtothepositiveeffectsoftherestoredstructureonly

(i.e.reefsunitsonlyandnotnecessarilytoanyfood-itembenefitsthatmightbegainedfromafully

grown,maturereefs),thenitmightbehypothesisedthattheseeffectswouldonlystrengthenwith

time.Theseeffects,however,requirefurthertesting.Ourfindingshaveimportantimplicationsfor

planningrestorationactionsbothinseaandonlandbecausetheynecessitatethatpractitioners

understandthebasicspatialpatternsthatarelikelytodrivetheabundanceanddistributionof

functionallyimportantspeciesacrossecosystems.Theseresultsalsosignaltheimportanceof

quantifyingsystem-specificresponsestorestoration,becausethepatternsthatwefoundwent

againsttheconventionalwisdomregardinglandscapepatternsforourstudysystem.Giventhe

paucityofinformationabouttheeffectsofhabitatrestorationonecologicalfunctions,determining

whethertheseeffectsfoundinthisstudyareconsistentacrossfunctions,ecosystemsand

environmentalrealms,isimportanttooptimisefuturerestorationefforts.

116

SupplementaryMaterials

FigureS1Conceptualdiagramillustratingcamerapositionsat0,1,2,5,10,15,20,25mdistancesfroma

reefsite.Thefirstcameraunitislocatedwithinthecentreofthereef(i.e.0m),theremaining(i.e.1,2,5,

10,15,20,25m)arepositionedatrandomanglesseawardfromthefirstcamera.

Shoreline

= Squid Pop Camera 0m

1m2m

5m

10m

15mAngles Random

20m

25m

Lowest astronomical tide

117

FigureS2Squidpopassaydesign.Squidpopsconsistofasingle1cm2pieceofdriedsquidmantletethered

toa20cmlongbamboostakeusinga10cmlengthoffishingline.Cameraunitswerecomprisedofa5kg

weightwithaGoProcamerarecordinginhighdefinition(1080p).Thesquidpopstakewasthenfastened

tothecameraunitusinga15mmgaugePVCarmatadistanceof45cmfromthecamera,sothatthe

squidpopwasvisibleatalltimesandforeasewhendeployingfromaboat.

118

TableS1Bestfitmodelonprobabilitiesofpredationatcontrolsitesfromgeneralisedadditivemodel

outputwithimportancevalues.Thebestfitmodelhadanr2valueof0.26.

Valuesinboldaresignificantat

p=0.05.

Source Variable

importance

ChiSq P

Salinity 0.99 24.11 <0.001

Mangrovearea 0.95 23.93 <0.001

Seagrassarea 0.57 25.95 <0.001

Secchidepth 0.67 5.3 0.09

119

Chapter6

Effectsofoysterreefrestorationextendacrosscoastal

seascapes

OrtodossiNL,GilbyBL,SchlacherTA,HendersonCJ,OldsAD(2019)Effectsof

oysterreefrestorationextendacrosscoastalseascapes.

120

Introduction

Landscapefragmentationandhabitatdegradationfromanthropogenicdisturbancesresultsina

simplificationofthestructuralcomplexityofhabitats,andthesechangeshaveledtounprecedented

lossesinbiodiversityandecosystemservices,globally(Butchartetal.2010;Montoyaetal.2012;

Sannigrahietal.2018).Thesenegativeconsequencesfromhabitatlossandmodificationhave

motivatedsignificantinterestintherestorationofdegradedecosystemsonland,andinthesea

(Lokeetal.2015;Montoyaetal.2012).Restoredecosystemsincreaselandscapediversity,and

habitatcomplexity,andcanenhancebiodiversity,re-establishecosystemfunctioningandsafeguard

ecosystemservices(Lokeetal.2015;MicheliandPeterson1999).Thesepositiveeffectsof

restoration,particularlyenhancementofanimalabundanceanddiversity,havebeenreported

widely,andfrommostrestoredecosystems(Bullocketal.2011;ReyBenayasetal.2009).Thereis,

however,someuncertaintyastowhetherthehighabundanceanddiversityoffaunaresultsfrom

themovementofanimalsintorestoredhabitatsfromnearbylocations(i.e.attraction)orfromthe

increasedcarryingcapacitythatrestoredhabitatsconfertolandscapes,andprovidenew

opportunitiesforthecolonisationofadditionalindividuals(i.e.supplementation).Theseattraction

versussupplementationhypotheseshavebeendebatedforsometime,particularlyinrelationtothe

habitatvaluesofartificialreefs(Brickhilletal.2005;PickeringandWhitmarsh1997),buttheconcept

hasnotbeensufficientlyexaminedwhenconsideringthepotentialvaluesofrestorationprojectson

land,orinthesea.

Theextentandconditionofmanycoastalandestuarineecosystemshavebeenreducedduetothe

cumulativeimpactsofmultipleanthropogenicstressors,withdetrimentaleffectstobiodiversity,

food-webs,ecologicalfunctioningandecologicalresilience(Elliottetal.2007;Hughesetal.2005;

WormandDuffy2003).Manyrestorationeffortshave,therefore,beenemployedincoastaland

estuarineseascapestocombatdecliningbiodiversityandpromoteecosystemsservices(Coenetal.

2007;Elliottetal.2007;Hughesetal.2005).Positiveeffectsofhabitatrestorationhavebeen

reportedwidelyformanymarineorganisms,butespeciallyforfishes,fromarangeofcoastal

ecosystems,includingseagrassmeadows(BlandonandzuErmgassen2014),mangroveforests

(Bosireetal.2008),saltmarshes(Raposa2002)andoysterreefs(Gilbyetal.2019).Itis,however,

notclearwhethertheseeffectsofrestorationresultfromthemovementofmobilemarine

organismsintorestoredhabitatsfromnearbylocations(i.e.attraction),orfromtherecruitmentand

colonisationofnewindividuals(i.e.supplementation)(Brickhilletal.2005;PickeringandWhitmarsh

1997).Itisimportanttodistinguishbetweenthesetwoprocessesbecauseincreasingthe

biomass/carryingcapacityofcoastalandestuarinesystems,particularlyforfishandcrustaceans,is

121

importantforsustainingthesubstantialrecreationalandcommercialfisheriesthatarederivedfrom

theseecosystemsaroundtheworld(Barbieretal.2011;Tayloretal.2017).

Oysterreefsarestructurallycomplexcoastalecosystems,andarewidelyregardedasoneofthe

mostvaluableandproductivecoastalfishhabitats(Coenetal.2007;GrabowskiandPeterson2007).

Oysterreefshave,however,experiencedsignificantdeclinesinthelastcentury,with>85%ofall

oysterreefsbeingreportedasfunctionallyextinctduetothecombinedeffectsofoverharvesting,

disease,anddecliningwaterquality,globally(Becketal.2011;Coenetal.2007;Grabowskietal.

2012).Lostoysterreefsarecommonlyreplacedwithhabitatsthatprovidelessfoodorpoorer

protectionfrompredatorsforfish,suchasbaremudsandsands(Gilbyetal.2018c;Grabowskietal.

2012;Grabowskietal.2005).Theseprofoundhabitatchangesareoftenirreversiblewithout

intervention,andaretypicallyassociatedwithdeclinesinfishdiversity,abundanceandbiomass,and

severeimpactstolocalrecreationalandcommercialfisheries(Coenetal.1999;Gilbyetal.2018c).

Forthesereasons,animportantgoalofmanyoysterreefrestorationprojectsistheenhancementof

fish,andfisheriesintheareassurroundingrestoredreefs(Baggettetal.2015;HumphriesandLa

Peyre2015).Previousresearchhasestablishedthatoysterreefsenhancefishabundanceand

diversityinareaswherereefsarerestoredonpreviouslybaresediments(Geraldietal.2009;

Petersonetal.2003;PiersonandEggleston2014;zuErmgassenetal.2016).Thesefindingswere,

however,limitedtotheimmediatevicinityofrestoredoysterreefs(i.e.mto10sofm)intemperate

marsh-dominatedseascapes(Grabowskietal.2005;MicheliandPeterson1999).Consequently,itis

notclearwhethersimilareffectsoccurinsubtropicalortropicalseascapeswhere

mangroves/seagrassdominate,andiftheseeffectscanbeextendedtolargerscales(i.e.kmto10’s

km)toaffectthebiomass/carryingcapacityofentireestuarineseascapes.

Inthisstudy,weuserestoredinter-tidaloysterreefsinasubtropicalestuaryinsouth-east

Queensland,Australiatotestwhethertherestorationofoysterreefssupplementthefish

communitybyprovidingadditionalstructurallycomplexhabitatsandenhancethebiomass/carrying

capacityoftheestuary,orsimplyredistributefishthroughoutthesystem.Therestoredoysterreefs

inthissystemreplacelowcomplexitysandyandmuddysubstrates,andareplacedacrossthe

estuaryinfundamentallydifferentseascapepositionsintermsoftheirproximitytootherfish

habitats(i.e.mangroves,seagrass)andintheirlevelofconnectivitywiththeopensea.This

restorationprojectis,therefore,anidealtestsystemforquantifyingwhether,andhow,restored

oysterreefsaugmentboththeabundanceanddiversityoffish,andwhethertheinstallationof

oysterreefsredistributetheirabundancethroughouttheestuary.Giventhepotentialforoyster

122

reefstoenhancefishbiomass(Geraldietal.2009;Petersonetal.2003;zuErmgassenetal.2016),

wehypothesisethattherestoredoysterreefswillaugmentfishabundanceanddiversitythroughout

thesystem(i.e.largespatialscales~kmto10’skms)andre-distributefishclosertorestoredoyster

reefs.

MethodsStudyseascape

Wesampledfishwithremoteunderwatervideostations(RUVS)withintheNoosaRiverestuary,in

Queensland,Australia(26°22'S153°04'E)(Figure1).TheNoosaRiverestuaryiscomprisedofa

heterogeneousmixofnaturalestuarinehabitatsincludingmangroves,seagrass,sandflats,rock

bars,andwoodydebris,andincludesamoderatelyurbanisedshoreline(≤50%oftheriverhas

urbanshorelines).Variationinthecompositionandarrangementofhabitatsinthisseascapehas

beenlinkedtochangesintheabundance,diversityanddistributionoffishintheestuary(Gilbyetal.

2018b).NaturaloysterreefswereonceabundantthroughouttheNoosaRiverestuary,butdeclined

precipitouslythroughouttheearly1900’sduetoanthropogenicactivities(especiallyoverharvesting,

decliningwaterquality,andassociatedrickofdisease),andarenowfunctionallyextinct(Thurstan

2016).Thetransitionofthesecomplexoysterreefstoun-vegetatedsandysubstrateshas

contributedsignificantlytoreductionsinfishdiversityandfishcatchesthroughouttheestuary

(Thurstan2016).OysterreefswererestoredtotheNoosaRiverestuaryinNovember2017,withthe

primaryobjectiveofenhancingfishabundanceanddiversitybyreplacingthelosthabitatvalues,and

heterogeneity,intheseascape.

Figure1.Locationofrestorationsites(orange)andremoteunderwatervideostations(RUVS)deployments

(blue)intheNoosaRiverestuary,Australia.

123

Oysterreefdesign

Oysterreefswereconstructedofbiodegradablecoconutmeshbags(1mlongx30cmdiameter)

filledwithrecycledoystershellsobtainedfromlocalsuppliers.Naturallyoccurringoysterlarvae

settleandgrowonthisrestoredstructure,andovertimesolidifytherecycledoystershellstogether,

formingfunctioningoysterreefsoncethecoconutmeshcasingdegradesaway.Eachoysterreefis

comprisedof9oyster-shellfilledbagsstackedinthreepilesofthreeintriangularprisms,withone

sidepositionedatlowestastronomicaltide(LAT),andwitheachpileseparatedby5m.Atotalof14

sitesthroughouttheestuarywererestoredwiththeseoysterreefs.Restorationsiteswerechosen

basedonthreekeycriteria:1)positionedwithinthehistoricalrangeofnaturaloysterreefsinthe

estuary;2)positionedtorepresentthefullspectrumofseascapecontextsavailablethroughoutthe

estuary;and3)positionedtoprovidesteppingstonesforfishmigrationsthroughouttheestuary.

Therestoredoysterreefsinthissystem,therefore,differintermsoftheirproximitytootherfish

habitats(i.e.mangroves,seagrass)andintheirlevelofconnectivitywiththeopensea(asourceof

fishandoysterlarvae);makingthemanidealnaturallaboratorytotesthowoysterreefrestoration

initiativescanmodifyfishabundanceanddiversity,andpotentiallyenhancefisheriesoverlarge

spatialscales.

Fishsurveys

Fishassemblagesweresampledusingremoteunderwatervideostations(RUVS).RUVSconsistedof

aGoProcamera(Hero5recordingin1080phighdefinition)mountedtoa3kgcustom-built

deploymentdevice.RUVSareincreasinglyusedforthistypeofstudyastheydonotbiasthe

structureoffishassemblagesviatheuseofbaitattractants(Gilbyetal.2018b).Surveyswere

conductedatthreemonthintervalsbetweenFebruary2017andAugust2018,andincludedthree

samplingeventsbeforetheinstallationoftheoysterreefs,andthreesamplingeventsafter

installation.Allsurveyswereconductedduringdaylight(07:00-16:00h),andwithin2hrseitherside

ofhightide,tomaximisevisibilityandaccessibilitytositesintheshallowreachesoftheestuary.We

deployedRUVSat100sites,whichwereselectedbyusingQuantumGIStooverlayagridwith200m

intervalsovertheentiremarinereachoftheNoosaRiverestuary(i.e.fromthemouthtothemost

upstreamreachescoveringallareaswhereoysterreefrestorationoccurred).EachRUVSdeployment

sampledfishfor30minutes,givingatotalvideosamplingtimeof50hrspersamplingeventand300

hrsoverallforthestudy.Fishabundanceanddiversitywerequantifiedfromvideosusingthe

standardMaxNstatistic;themaximumnumberofindividualsofaspeciesobservedinanysingle

videoframe(Ortodossietal.2018).Wealsoquantifiedthenumberofspecieswhichwereharvested

124

bybothrecreationalandcommercialfishersusinginformationfromFishBase(FroeseandPauly

2018).

Dataanalysis

Weusedpermutationalmultivariateanalysisofvariance(PERMANOVA)inPRIMERE(Andersonetal.

2008)todeterminedifferencesintheassemblagesoffishintheNoosaRiverbetweenpre-andpost-

installationevents(fixedfactor,twolevels),andsurveymonths(fixedfactor,threelevels;February,

MayandJuly/August).PERMANOVAresultswerevisualisedusingnon-metricmulti-dimensional

scaling(nMDS)ordinations.Fishspeciesthatdrovethesedifferenceswereidentifiedusingindicator

speciesanalysis(INDVAL)(DufreneandLegendre1997)inthelabdsvpackage(Roberts2013)ofthe

Rstatisticalframework(RCoreTeam2019).Significantindicatorspecieswereplottedoverthe

nMDSordinationusingPearsonvectoroverlays.

Wefurtherinterrogatedthesepatternsbytestingfortheinteractiveeffectsofthedistanceofeach

surveysitetothenearestoysterreefrestorationsite(distancetoreef),pre-andpost-installation

events,andsurveymonthsonspeciesrichnessandharvestablefishabundanceacrosstheestuary.

Thesedifferenceswerevisualisedusing‘HeatMaps’createdinQGIS(QGISDevelopmentTeam

2019).

ResultsFishassemblages

FishassemblagesintheNoosaRivervariedsignificantlybetweenpre-andpost-installationevents,

andwithsurveymonth(Table1).Pairwisetestsindicatedthatallpairwisecomparisonsofthe

pre/postinstallationandsurveymonthinteractiontermdifferedsignificantly(Figure2).Sixspecies

wereidentifiedinINDVALanalysisasbeingtheprincipledriversofthesepatterns;yellowfinbream

Acanthopagrusaustralis,sandwhitingSillagocilliata,seamulletMugilcephalus,spottedscat

Scatophagusargus,southernherringHerklotsichthyscastelnaui,andnarrow-linedpufferArothron

manilensis.YellowfinbreamandsandwhitingweresignificantindictorsofthePre(July/August)

event,whereaseachoftheotherspeciesweresignificantindicatorsPre(February)event.We

identifiedasignificantreductioninaveragespeciesrichnessandharvestablefishabundance

betweenthethreepreinstallationsurveysandthethreepostinstallationsurveys(Figure3).

125

Table1.Permutationalmultivariateanalysisofvarianceresults(PERMANOVA)testingfortheeffectsof

pre/postinstallationandsurveymonthontheassemblageoffishatallsitessurveyedintheNoosaRiver.

Pairwisetestsindicatethatallcombinationsintheinteractiontermdifferedsignificanttoeachother.

Source df SS MS Pseudo-F P

Pre/PostInstallation 1 14734 14734 4.1099 0.003

Surveymonth 2 79795 39897 11.129 0.001

Pre/PostInstallationxSurveyMonth 2 51417 25708 7.1709 0.001

Residual 602 2.16E+06 3585.1

Total 607 2.30E+06

Figure2.Non-metricmultidimensionalscalingordination(nMDS)ofcentroidsforpairsofpre/postinstallation

andsurveymonthwithPearsonvectoroverlaysofsignificantspeciesfromINDVALanalyses.

Pre (Feb)

Pre (May)

Pre (July/Aug)Post (Feb)

Post (May)Post (July/Aug)

Sand whiting

Yellowfin bream

Sea mullet

Spotted scatSouthern herring

Narrow-lined puffer

Pre (Feb)Pre (May)Pre (July/Aug)Post (Feb)Post (May)Post (July/Aug)

126

Effectsofproximitytorestoredoysterreefs

Fishdiversityandabundancewereconcentratedinthesameplacebothbefore,andafter,

installation(Figure4);inthelowerestuary,thatusuallyhashighsalinity,anabundanceofstructure

habitats(inparticularlogsnags,mangrovesandurbanstructures,andispositionedatthemeeting

pointoftwodifferentestuaries(WeybaCreek,andthemainNoosaRiver).Speciesrichnessinthe

NoosaRiverestuarywasmodifiedbythedistanceofmonitoringsitestoreefs,andtheinteraction

betweenpre/postinstallationandsurveymonth(Table2,Figures4,5).Here,speciesrichnesswas

highestatmonitoringsitesfarthestfromoysterreefs(Figure4,5A),andwasvariable,withno

consistentpatterning,betweensurveymonthsandsurveyeventspreandpostinstallation(Figure

4B).TheabundanceofharvestablefishintheNoosaRiverestuarywasmodifiedbytheinteractions

betweenpre/postinstallationanddistancetoreef,pre/postinstallationandsurveymonth,and

surveymonthanddistancetoreef(Table2,Figure4,6).Here,theabundanceofharvestablefish

increasedconsistentlywithdistancefromrestoredreefs,butincreasedtoagreaterdegreepost

installation(Figure6A).Similarly,tothepatternsthroughtimewithspeciesrichness,harvestablefish

abundancevariedconsiderablythroughtimeintheestuary,withnoconsistentpatterns(Figure6B).

Table2.Resultsofgeneralisedlinearmodelanalysestestingfortheeffectsofdistancetooysterreef,pre/post

installationandsurveymonthonA)speciesrichnessandB)harvestablefishabundance.

Source Df X2 P

SpeciesRichness

Pre/postinstallation 1 24.8 <0.001

Distancetoreef 1 8.1 <0.001

Surveymonth 2 33.9 <0.001

Pre/postinstallationxDistancetoreef 1 1.7 0.189

Pre/postinstallationxSurveymonth 2 85.4 <0.001

SurveymonthxDistancetoreef 2 1.0 0.612

Pre/postinstallationxDistancetoreefXSurveymonth 2 0.9 0.625

Harvestablefishabundance

Pre/postinstallation 1 116.1 <0.001

Distancetoreef 1 19.8 <0.001

Surveymonth 2 149.5 <0.001

Pre/postinstallationxDistancetoreef 1 13.4 <0.001

Pre/postinstallationxSurveymonth 2 258.1 <0.001

SurveymonthxDistancetoreef 2 45.1 <0.001

Pre/postinstallationxDistancetoreefXSurveymonth 2 4.0 0.137

127

Figure3Average(+/-SE)speciesrichnessandharvestablefishabundanceacrossallsitesforthethree

monitoringeventspriortoreefinstallation,andthethreeeventsafterreefinstallation.

01234567

Species richness Harvestable fish abundance

Average

Pre Installation Post installation

128

Figure4Heatmapsofbeforeandafterreefinstallationforspeciesrichnessandharvestablefishabundance.

129

Figure5EffectsofA)distancetoreefandB)theinteractionbetweenpre/postinstallationbysurveymonthon

speciesrichness.

2.0

2.5

3.0

3.5

0 500 1000Distance to reef (m)

Pred

icte

d sp

ecie

s ric

hnes

s

0

0.5

1

1.5

2

2.5

3

3.5

4

Feb May July/Aug Feb May July/Aug

Pre Post

Spec

ies

richn

ess

A. Distance to reef

B. Pre/post installation x survey month

130

Figure6EffectsofA)theinteractionbetweenpre/postinstallationanddistancetoreefandB)theinteraction

betweenpre/postinstallationbysurveymonthonharvestablefishabundance.

3

4

5

6

7

0 500 1000Distance to reef (m)

Pred

icte

d ha

rves

ted

fish

abun

danc

ePeriod Post Pre

0

2

4

6

8

10

Feb May July/Aug Feb May July/Aug

Pre Post

Harv

esta

ble

fish

abun

danc

e

A. Pre/post installation x distance to reef

B. Pre/post installation x survey month

131

DiscussionQuantifyingtheeffectsthatrestorationhasonthespatialdistributionofanimalsbothatrestoration

sites,andacrosslandscapesisanimportantfocusforrestorationecologists(Gilbyetal.2018a;Jones

andDavidson2016).Asthescaleofrestorationincreasesacrossallenvironmentalrealmsglobally,

thecapacityfortheeffectsofrestorationonanimalpopulationstoexpandacrosslandscapesalso

increases.Quantifyingthedegreetowhichcurrentrestorationactionsmodifytheseeffects,andthe

attributesofrestorationprojectsthatmightmaximizetheproliferationofrestorationeffectsmore

broadly,isthereforevitalinoptimizingfuturerestorationplans(Guerreroetal.2017;Haggeretal.

2017).Inthisstudy,wefoundthatfishassemblagesvariedsignificantlyacrosstemporalscalesinthe

NoosaRiver,andthattherewasasignificantchangeintheassemblagecompositionanddistribution

offishintheriverfollowingtheinstallationoftheoysterreefs.Here,allsignificantindicatorspecies

wereinhigherabundanceandmoreprevalentpriortotheinstallationoftherestoredoysterreefs.

Theseeffectsproliferatedtoeffectsonspeciesrichnessandharvestedfishabundance,whichwere

bothhigheratgreaterdistancesfromtherestoredreefs.Similarly,weidentifiedhigherspecies

richnessandharvestablefishabundanceduringthemonitoringeventspriortotheinstallationofthe

oysterreefsthanafter.Thesefindingsarecountertotheeffectsfoundinotherstudies,including

otherstudiesofthesereefsthatfoundasignificantlyhigherabundanceoffishoccurringatreefsand

significantlyhigherratesofkeyecologicalfunctions.Thesestudiesidentifiedasignificanteffectof

freshwaterrunofffromcatchmentsonthediversityandabundanceoffishintheNoosaRiver

estuary(Gilbyetal.2019),andthesecondyearofourmonitoringprogramherehadsignificantly

higherrainfallinthebroaderNoosacatchmentthanthefirstyearofsampling(Bureauof

Meteorology2019).Thissuggeststhatadditionalenvironmentalvariableswillneedtobeaccounted

forinsubsequentanalysestoproperlyquantifytheeffectsofthereefsonfishassemblages

throughouttheestuary.Consequently,giventhefindingsofotherstudiesonthesereefs(Duncan

2018;Gilbyetal.2019),andthefactthattheeffectsofoysterreefsonfishcantakeseveralyearsto

develop(thisstudywascompletedduringtheyearfollowingoysterreefinstallation)(zuErmgassen

etal.2016),wecannotconcludethattheeffectsoftheinstallationofthereefshasbeenanoverall

declineinfishabundanceanddiversityintheriver,asanalysesthataccountforvariablerunoffwill

berequired.Similarly,wecannotconclusivelydeterminewhethertherehasbeenanoveralladditive

ortake-backeffectofthereefsonthebroaderfishassemblagethroughouttheNoosaRiver.More

complexspatialanalyses,andpotentiallymoresurveyeventswillberequiredtobetterdisentangle

theseeffects.Irrespectiveofthesepatterns,theresultsfoundherewillprovideinvaluablebaseline

informationregardingtheeffectsofoysterreefsonfishandfisheriesthroughouttheNoosaRiver,

especiallywithplanstoexpandtherestorationeffortsinNoosamorebroadlynowunderway.

132

Ithasbeenwellestablishedgloballythatrestoringoysterreefstopreviouslyunvegetated,sandyor

muddysubstratesaugmentstheabundanceanddiversityoffish(Petersonetal.2003;zuErmgassen

etal.2016).IntheNoosaRiverestuary,restoredoysterreefscontained1.4timesthenumberof

speciesand1.8timestheabundanceofharvestablefishthanatnearbycontrolsites(Gilbyetal.

2019).Similarly,studiesquantifyingthedistributionofkeyecologicalfunctionsaroundtherestored

reefsfoundopposingeffectstotheresultsfoundhere(Duncan2018).Thefunctionofpredation,an

importantfunctionthatmaintainsfoodwebstructureandensurestheexchangeofnutrientsand

energythroughfoodwebs,declinedwithincreasingdistancefromtheserestoredoysterreefs

(Duncan2018).Ourfindingsinthisstudywerethereforecountertoourpredictionsoftheeffectsof

restorationonthedistributionoffish.Nostudies,however,haveattemptedtoquantifytheeffects

ofoysterrestorationacrossseascapes(i.e.awayfromrestorationsites)tothescaleattemptedhere.

Thereareseveralpotentialexplanationsforthepatternswefoundinthisstudy.Firstly,itislikely

thatchangesinenvironmentalconditionswithintheestuarybetweenthe2017and2018sampling

eventshadasubstantialeffectonthecompositionoffishwithintheestuary(Gilbyetal.2017a;

Gilbyetal.2019).Secondly,theoysterreefsinstalledinNoosaarerelativelysmall;werestored14

oysterreefsof(approximately25m2)foratotalrestorationfootprintofapproximately350m2across

theentireestuary.Thismeansthatthesereefsremainarelativelysmallcomponentofthebroader

seascape,sotheireffectswererelativelylowacrosstheseascapedespiteourfindingofsignificant

benefitsatthereefsitesthemselves(Duncan2018;Gilbyetal.2019).AreasintheNoosaRiverstudy

areathatarefarthestfromoysterreefs,arealsothosenearthemouth,andwhereWeybaCreek

joinsthemainsystem.Thisiswherefishabundanceanddiversitywasaggregatedbothbeforeand

afterthereefswereinstalled.Thestrongereffectsofdistancefromreefsonfishassemblagespost-

installationlikelyreflectstotheaggregationofmorefishfrommorespeciesintheseareasafterthe

reefswentin.Thisisnotlikelyrelatedtothereefsthemselves,butratherpossiblyreflects

movementdownstreamtospawn,orfollowingsuccessiverainfallevents.Increasingthesizeof

restorationsitesinsubsequentiterationsoftherestorationeffortsinNoosaisthereforeakey

recommendation,andwilllikelyresultinagreater,andmoreimmediatelydetectable,effectonfish

andfisheriesacrosstheentireestuary.Globalanalysesoftheeffectsofrestoredoysterreefsonfish

indicatethatitcantakeupto15yearsfortheeffectsofrestorationonfishassemblagestofully

develop(zuErmgassenetal.2016).Thesetimeframesreflectthetimeframesidentifiedforthe

developmentofeffectsformarinereserves(Edgaretal.2014).Theshort,oneyearperiodsurveyed

heremaynothaveprovidedsufficienttimefortheeffectsofthereefstoevenbedetectedproperly

133

acrosstheNoosaRiverestuary,letalonefullymature,especiallyinthebackgroundoftheincreased

rainfallexperiencedinthesecondyearofsurveys(BureauofMeteorology2019).Consequently,

long-termmonitoringofrestorationeffortsisencouragedsothatthefullbenefitscanbeproperly

quantifiedinthecontextofannualandseasonalvariationsintemperature,rainfallandfishing

pressure(Underwood1991).Maintainingsuchlong-term,consistentmonitoringhasbeena

significantchallengeforrestorationinmanysystems,butservestobetterinformstakeholdersofthe

benefitsthatrestorationbrings,increasesthecapacitytooptimizesubsequentrestorationactions,

andprovideslong-termevidenceofrestorationbenefits(Baggettetal.2015;Elliottetal.2016;Gilby

etal.2018c).

Thepositioningofoysterreefsonintertidalbanksmayhavecontributedtowardsthehigher

assemblagespeciesrichnessandabundanceofharvestablefishfurtherawayfromrestoredoyster

reefs.AlloftheoysterreefsrestoredintheNoosaRiverestuarywereplacedintertidally,andthe

vastmajoritywereplacedonthenorthernbankoftheestuary.Thismeansthatthedeeperchannels

whichfishusetomigrateinandoutoftheestuary,andseekshelterduringlowtideswerelocated

somedistancefromtherestorationsites.Broadly,connectivitywiththesedeeperchannelshas

showntobeasignificantfactorthatchangesthedistributionoffishacrosstheestuarineseascapes

ofNoosaandtheSunshineCoast(Brooketal.2018;Gilbyetal.2017b).Consequently,wehavetwo

keyrecommendationsfromthisfinding.Firstly,subsequentanalysesshouldaccountforthedistance

ofthesedeeperchannelsfromthemonitoringsites,andfortheactualdepthofeachmonitoring

sites.Similarly,subsequentrestorationeffortsintheNoosaRivershouldbetteraccountfor

connectivitywiththesedeeperwaterchannels,astheseareimportantmovementpathwaysfor

manyoftheharvestablefishspeciesforwhichtheNoosarestorationactionsseektoenhance(Davis

etal.2014;Davisetal.2017).

Theeffectsofconservationinterventionsonthebroaderdistributionofanimalsthroughout

landscapeshasbeenasignificantfocusovermanyyears.Here,studieshaveprincipallyfocusedon

theeffectsofnaturereservesonthedistributionofanimalsaroundreserves.Thesetypesofstudies

havefocusedonthis‘spillover’effectintwoways;positivelythroughthespilloverofharvestable

biomassandlarvaetofishedareasfrommarinereservesasfishassemblageswithinmarinereserves

mature(McClanahanandMangi2000;Stobartetal.2009),andthenegativeeffectsofanimal

interactionsandpathogensspillingovertoareasofhumanhabituationfromnationalparksasthe

carrycapacityoftheparkisreached(Kilpatricketal.2009;RippleandBeschta2007).Expanding

researchintotheeffectsofrestorationonthese‘spillover’effectstosurrounding,unrestoredsites

134

willenhanceourcapacitytooptimizerestorationinseveralkeywaysanalogoustothoseconducted

forreserves;wecanoptimizeourrestorationdesignsinsituationswherethespilloverofbiomassis

favourable(likethecasestudyhere),andlimitconflictswhensuchspilloverislessfavourable.

Despitetheequivocalresultsfoundhereregardingtheeffectsofoysterreefinstallationonthe

assemblagesoffish,anddistancetorestorationsitesfollowingrestoration,severalkeyconclusions

canstillbedrawnforfuturerestorationeffortsinNoosa.Principally,ensuringthattheeffectsof

restorationexpandsignificantlyacrosstheentireseascapeoftheNoosaRiver(i.e.theyincreasethe

overallcarryingcapacityofthesystemmorebroadly)willrequirelargerrestorationsitesacrossthe

entireNoosaseascapeto;1)enhancetheaugmentationeffectsofthereefs,and2)spreadbiomass

acrosstheestuary.Further,betterconsiderationofconnectivityeffectswithdeeperchannels,anda

long-termcommitmenttothoroughmonitoringandevaluationoftheeffectsoftheoysterreefs

acrosstheentireestuaryisrequired.

135

Chapter7

Identifyingrestorationhotspotstomaximiseoutcomes

formultiplerestorationbenefits

GilbyB,OldsA,DuncanC,OrtodossiN,HendersonC,SchlacherT(2019)

Identifyingrestorationhotspotstomaximiseoutcomesformultiplerestoration

benefits.(underreview)

136

Introduction Theprevalenceandsizeofecologicalrestorationprojectsisincreasingacrossmarine,freshwater

andterrestrialecosystemsglobally(Middendorpetal.2016;Paiceetal.2016).Ecologicalrestoration

principallyseekstore-establishorrestorelostordegradedecosystemsandtheirfunctioning

(Balagueretal.2014;Haggeretal.2017;Stanturfetal.2014).Restorationprojectscanalsohave

importantgoalsaroundenhancingorre-establishingpopulationsofanimals(Gilbyetal.2018a;

JonesandDavidson2016b),andtheprovisionofecosystemservices(DameandLibes1993;Smyth

etal.2015).Usingquantitativeinformationtooptimiserestorationplansmayresultingreater

ecological,economicandsocialbenefitsderivedfromrestorationforthesamefinancialcosts,and

canpotentiallyminimisesocialcostsanddisruptions(Bullocketal.2011).Identifyingrestoration

actionsthatcreatehotspotsformultiplerestorationbenefits(i.e.animalpopulations,ecological

functionsand/orecosystemsservices,inadditiontothegrowthoflostordegradedecosystems)is

thereforeanimportantgoalforecologistsandrestorationpractitioners(SchulzandSchroder2017).

Theprincipleofmanagingorimprovingecosystemsformultiplebenefitshasbeenlong-recognised

(HectorandBagchi2007;Pengetal.2019),butonlyrarelyappliedtothedesignofrestoration

projectsinaformalmanner(CrossmanandBryan2009;Gilbyetal.2018c).Forexample,thereis

increasingrecognitionthatmanagingecosystemstoenhancebiodiversity,andthereforeratesand

redundancyofmultipleecologicalfunctions(e.g.recruitment,predation,herbivoryandscavenging),

canincreasetheresilienceofecosystemstoarangeofpotentialdisturbances(Cardinaleetal.2006;

GamfeldtandRoger2017).Restorationliteraturehasincreasinglyrecognisedtheimportanceof

designingandplacingrestorationeffortstomaximisebiodiversityoutcomes,increaseratesofkey

ecologicalfunctions,andprovideecosystemservices(likecarbonsequestrationandharvestable

species)(Stanturfetal.2014;Wortleyetal.2013)bothatrestorationsites,andthroughout

landscapesmorebroadly(Brudvig2011;Holletal.2003).Suchimportant‘secondary’benefitsof

ecologicalrestoration(i.e.beyondsimplytherecoveryoflostordegradedecosystems)have

graduallysteeredrestorationtargetsawayfromtherestorationofecosystemstohistoricalbaselines

orreferencesites,andtowardsmorecomplexgoalsfocusedontheprovisionofecologicalservices

andtherestorationofself-sustainingecosystems,inadditiontoreversingbiodiversitylosses(Bullock

etal.2011;ReyBenayasetal.2009).However,quantifyingthefeaturesofrestorationsitesthat

maximisemultiplebenefitsinastructuredmanner,andthensettingquantitativerestorationgoals

tooptimiserestorationplans,remainsrare(Guerreroetal.2017;Tobonetal.2017;Wiensand

Hobbs2015).

137

Thelandscapecontextofecosystems(theirsize,shapeandspatialarrangement)modifiesthe

conditionandcompositionofhabitat-formingspecies,theirvalueashabitatsforanimals,andthe

ratesanddistributionoftheecologicalfunctionsthoseanimalsprovide(Massoletal.2011;Pittman

etal.2011).Thelandscapecontextofrestorationsitescanthereforeaffectplantandanimal

recruitment,speciespersistence,ratesofkeyecologicalfunctions,andtheprovisionofecosystem

services(Grabowskietal.2005;SchultzandCrone2005;Simenstadetal.2006).Theseeffectscan

haveknock-onconsequencesfortheconditionofrestoredecosystems,thetimeinwhichrestored

ecosystemstaketobecomeself-sustaining,andthelikelihoodofachievingrestorationtargets

(Brudvigetal.2009;Rudnicketal.2012;Tambosietal.2014).Despitetherecognisedimportanceof

landscapecontextinoutcomesforrestorationactions,onlyapproximately11%ofrestorationsites

havebeenselectedwithproperconsiderationoflandscapeeffects(Gilbyetal.2018a).Identifying

restorationsiteswithinheterogeneouslandscapesthatfulfilboththebiologicalrequirementsof

habitatformingspecies,andthatalsomaximiserestorationoutcomesformultiplesecondary

restorationtargets(i.e.animalsandmultifunctionality)ischallenging,andreliantupon

understandingthespatialdriversofallrestorationbenefits(SchulzandSchroder2017).After

decadesofresearchfocusingonthewaystomaximisethegrowthofhabitatformingspeciesin

restorationprojects,andwithlandscapescalerestorationplansbeingimplementedinmany

settings,researcherscannowfocusmorestronglyonoptimisingthespatialplacementofrestoration

actionsinlandscapestoensureoutcomesformultiplebenefits(Gilbyetal.2018c;Jonesand

Davidson2016a;Rudnicketal.2012;Wortleyetal.2013).

Coastalecosystemsareidealstudysystemstoquantifytheeffectsofrestorationactionsformultiple

benefits.Coastalecosystemsarehighlythreatenedbyhumanactivities(Halpernetal.2008),butare

alsovitalhabitatsandnurseriesformanycoastalanimalspecies(Whitfield2017),meaningthat

thereisastrongdemandforactivecoastalrestoration(Bayraktarovetal.2016).Forexample,an

estimated85%ofnativeoysterreefshavebeenlostgloballyduetooverharvesting,decliningwater

qualityanddisease(Becketal.2011).Thereplacementofstructurallycomplexoysterreefswithlow

complexitysandyandmuddysubstratesarelikelycontributorstowardsbothfisheriesdeclines

(Petersonetal.2003;zuErmgassenetal.2016)andreductionsinecologicalfunctioninginmany

coastalecosystems(Grabowskietal.2005;MicheliandPeterson1999).Therefore,oysterreef

restorationisoftenusedtore-establishlosthabitats,aswellastoenhancepopulationsofanimals

andtheecologicalfunctionstheyprovide(Gilbyetal.2019;Grabowskietal.2005;Turneretal.

1999).Thereisamountingbodyofevidencetosuggestthattheeffectsofrestoredoysterreefson

fishassemblagesismodifiedsignificantlybytheirlandscapecontext(Duncan2018;Gilbyetal.2019;

138

Grabowskietal.2005).However,oystersalsooftenhavenarrowenvironmentalenvelopesinwhich

theycanrecruit,growandthrive(e.g.tide,salinity,turbidityandtemperatureranges)(Baggettetal.

2015;Lenihanetal.1996;Tamburrietal.2008).Thismeansthatrestorationplansmayneedto

incorporatesystem-specificinformationonthespatialdistributionofoysterreefgrowth,

assemblagesoffishatrestorationsites,andratesofkeyecologicalfunctionstomaximisethescale

ofoverallbenefits(Gilbyetal.2018c).

Withincreasinginvestmentsbeingmadegloballyonecologicalrestoration,andthescaleof

restorationprojectsincreasingconcomitantly,identifyingrestorationsitesinlandscapesthatresult

inhotspotsformultiplerestorationbenefitsisanincreasingresearchfocus(Zhangetal.2018).This

requirestheselectionofsiteswherethephysicalrestorationeffortwillbesuccessful,aswellas

providingheightenedsecondarybenefitsthatpeoplecovet.Inthisstudy,wesurveythegrowthof

oystersandthedevelopmentofsecondarybenefitsat13restoredoysterreefsintheNoosaRiver

EstuaryinQueensland,Australia.Oysterreefswerehistoricallyabundantintheestuary,butwere

extirpatedfromthesystemintheearly1900’sduetothecombinedeffectsofoverharvesting,

sedimentation,decliningwaterquality,anddisease(Thurstan2016).Thislossofcomplex,biogenic

habitatsfromthesystemhasbeensuggestedascontributingtoreducedfishcatchesandecosystem

conditionintheestuaryoverthepastcentury.Enhancingfishabundanceanddiversity(especiallyof

harvestablespecies)isthereforeaprinciplegoalofthecurrentrestorationproject.Previousstudies

attherestoredreefsinNoosashowed1.4timesgreaterspeciesrichness,1.8timesmorefishthat

aretargetedbycommercialandrecreationalfishers,anddoublingofpredationratesthanatnearby

controlsites(Duncan2018;Gilbyetal.2019).Thebasicpremiseofthisstudywastoidentify

locationswhereplacingadditionalrestoredoysterreefswouldresultinthegreatestnetbenefits.

Wesurveyedsixpotentialrestorationbenefits(i.e.restorationbenefits)ateachoftherestoredreefs

intheNoosaRiverestuarythatcouldbebroadlygroupedintothreecategories.Firstly,wequantified

thedensityandsizeofoystersgrowingontherestorationunits,therebyprovidinginformation

regardingthegrowthoftheprimaryhabitatformingspecies.Secondly,wequantifiedthediversityof

fishassemblagesandtheabundanceofharvestedfishspeciesatreefs,therebyproviding

informationonthefishassemblagescongregatingaroundtherestoredreefs.Finally,wequantified

ratesofpredationandcarrionscavenging,therebyprovidinginformationonthekeyecological

functionsthatsupportoysterreefswithinthisestuary.Wethenusedregressionanalysesand

distributionmodelstoidentifyrestorationhotspotsthataremostlikelytomaximiserestoration

outcomesforallsixvariables.Thisstudyrepresentsoneofthefirstattemptstousethoroughspatial

139

analysesofmultipleaspectsofrestorationeffortstoidentifyrestorationhotspotsincoastal

ecosystemsformultiplerestorationbenefits.

MaterialandMethodsStudysystem

Oysterreefswererestoredto13sitesintheNoosaRiverestuaryinsouthernQueensland,Australia,

inNovember2017(Figure1).Oysterreefrestorationunitscomprisedthree1mlongby30cm

diametercoirmeshbags(2.5cmaperture)filledwithrecycledoystershellsstackedinatriangular

prism.Threeoftheseoysterreefunitswereplacedinatrianglewith5msidesandonesideatthe

positionoflowestastronomicaltideateachrestorationsite.Theseoysterreefstructuresare

designedtoallowoysterlarvae,whichstilloccurinnumbersgreatenoughtoallownatural

recruitmentandgrowth,tosettleamongsttheoystershellsandcementthemtogetherovertime.

TheNoosaRiverestuarysupportsaheterogeneousmixofunvegetatedsandysubstrates,mangrove

forests(mostlyAvicenniamarina),seagrassbeds(mostlyZosteramuelleri)andhardened(urbanised)

shorelines(Figure1).The13oysterreefsitesrepresentthediversityoflandscapecontexts

(especiallywithrespecttoproximitytonearbyseagrassesandmangroves)thatexistcurrentlywithin

theestuary.

Figure1MapoftheNoosaRiverestuary,oysterrestorationsitesandcurrent-daydistributionsofseagrass

bedsandmangroveforests.Sitenumbersprovidedforoysterrestorationsites.

140

Fieldsurveys

Wesurveyedsixattributesofoysterrestorationsites,henceforth‘restorationbenefits’,thatcould

bebroadlygroupedintothreecategories;oysterspatsettlementandgrowth,fishassemblage

richnessandharvestablefishabundance,andecologicalfunctioning.

Oysterspatsettlementandgrowthsurveys

WequantifiedthenumberandsizeofSydneyrockoystersSaccostreaglomerata

(i.e.thekeyreef-formingoysterspeciesinthissystem)settlingontheoysterrestorationunitsusing

twotechniques.Firstly,sixshellswereaffixedtotheoutsideofthecoirbagsusingfishinglineand

wereusedtoquantifythenumberandsizeofjuvenileoysterssettlingontheoutsideoftheoyster

units(i.e.the‘necklace’method).Secondly,sixshellswererandomlyremovedfromeachoysterreef

restorationunittoquantifythenumberandsizeofjuvenileoysterssettlingwithinthematrixofthe

oysterrestorationunititself.Shellswerecollectedfromtheoysterrestorationunitsatperiodsofsix

months(i.e.May2018)andoneyear(i.e.November2018)aftertheoysterreefswereinstalled,and

wereplacedimmediatelyoniceandthenfrozenpriortoquantificationinthelaboratory.Oyster

densitywasestimatedbycountingthenumberofSydneyrockoystersidentifiedonallsurfacesof

collectedshells,andusingatransparent,flexibleplasticsheetwitha1cmgridoverlaytoestimate

thetotalsurfacearea(tothenearestcm2)ofeachshellremovedfromthefield.Quantificationof

oystersizefollowedpreviousoystergrowthquantificationmethods,andinvolvedmeasuringthe

heightofeachoyster(i.e.thedistancefromtheumbotothedistalmarginoftheshell)usingVernier

calliperstothenearest1/10mm(Baggettetal.2015).

Fishassemblagerichnessandharvestablefishabundancesurveys

Wesurveyedfishassemblagesatoysterreefrestorationsitesusingbaitedremoteunderwatervideo

stations(BRUVS)atsix-weekintervalsforfoureventsfollowingtheinstallationofthereefs.BRUVS

werecomprisedofaGoProcamerarecordinginhigh-definition(1080)mountedtoa5kgweight,

withabaitbagsecured50cmfromthecamerausinga1mlongbaitarmmadeof15mmgaugePVC

pipe.BRUVSwerebaitedwith~500gofpilchards(Sardinopssagax)ina20x30cmmeshbaitbag

with0.5cm2aperture;thestandardbaitandquantityforcoastalfishsurveysusingBRUVS(e.g.

Hendersonetal.2017a).BRUVSweredeployedatthecentreofoysterreefrestorationsiteforone

hour.Surveyswereconductedtwohourseithersideofhightidetomaximisefishaccessibilitytothe

reefsandwatervisibility(Gilbyetal.2017a).VideofootagewasanalysedbyquantifyingMaxN;the

maximumnumberofeachspeciesidentifiedinanyoneframewithineachvideo.Toaccountfor

variablewaterclarity,onlyfishseenswimmingbetweenthecameraandtheendofthebaitarm(i.e.

141

1mfromthecamera)werecounted.Nosurveyswereconductedwheretheendofthebaitarm

couldnotbeseenbyobservers,sowatercolumnturbiditydidnotinfluenceourresults.Species

richness(i.e.thetotalnumberofspecies)andtheabundanceofharvestedfishspecies(thesumof

MaxNvaluesforallharvestedspeciesidentifiedinFishBase)(FroeseandPauly2018)werethen

calculatedfromtheresultingmatrixofMaxNvalues.

Scavengingratesurveys

Carrionconsumption(i.e.scavenging)isanimportantecologicalfunctionincoastalecosystems.

Manyfishandinvertebratespeciesscavengeanimalcarcasses,andtheiractionsrecyclenutrients

andenergy,andpreventthebuildupofpotentiallyharmfulcarrion(Oldsetal.2018;Wilsonand

Wolkovich2011)

ScavengingrateswerequantifiedatrestoredoysterreefsduringBRUVSsurveysbydeployingthree

pilchardcarcassesofknownweighttothesurfaceofeachBRUVSbaitbag(Oldsetal.2018).After

the1hrdeployment,pilchardcarcasseswerere-weighedtoestablishthechangeinweightduring

thedeployment,withscavengingrateastheproportionofbaitconsumedduringtheonehour

deployment.Controldeployments(wherescavengingdidnotoccurduringdeployments)showno

significantchangeinweightovertheonehourdeploymentperiod(Oldsetal.2018).

Predationsurveys

Predationisanimportantecologicalfunctionthathelpstomaintainfoodwebstructureandthe

exchangeofenergyinallecosystems(Estesetal.2010;RippleandBeschta2012;Ritchieand

Johnson2009).Quantifyingratesofpredationaroundhabitatrestorationprojectsisimportant

becausepredationissignificantly,andquickly,modifiedbytherapidcolonisationofpredatorsto

restoredcoastalecosystems,andpredationaroundreefsisagoodindicatorthatfoodwebsare

developinginassociationwiththerestoredstructures(Harding1999;MicheliandPeterson1999;

Petersonetal.2003).

Weusedsquidpopstoquantifyrelativepredationaroundoursitesat8-weekintervalsacrossfour

eventsfollowingtheinstallationofreefs,withthefirstsurveyoccurringintheweekfollowingoyster

reefinstallation.Squidpopsareastandardmethodusedincoastalseascapestoindextherelative

predationratesofmeso-predators(Duffyetal.2015).Squidpopsaredesignedtomimicasmall

invertebrateorfishswimminginthewatercolumnandaremadeofa1cm2pieceofsquidmantle

connectedtoabamboostakewitha10cmpieceoffishingline.Eightsquidpopsweredeployed

142

aroundeachoysterreef(atleast1mfromeachother),duringeachsurveyperiod.Squidpopswere

deployedonthesamevideoarraysdescribedabovefortheBRUVS(inplaceofthepilchardsandbait

bag,andneveronthesamedayasBRUVS),onthePVCpipearmatadistanceof45cmfromthe

cameraGoProcamera,toensurethesquidpopsremainedvisible.Wedeployedsquidpopsarraysfor

onehour(basedontheresultsofapilotstudypriortothecommencementofsurveys),duringa

periodtwohourseithersideofhightide(toalignwithperiodsthatwealsosampledfish

assemblages),tomaximiseaccessibilitytotheoysterreefs,andtomaximisevisibilityonthevideos.

Squidpopsareclassifiedasconsumedwhenananimalremovestheentirepieceofsquidfromthe

array.Cameraswereusedtoconfirmthatpredationeventswereduetoconsumptionbyanimals,

andnottheresultofanyotherrandomevent(e.g.lossofsquidmantleduetocurrent/tidesetc).A

subsetof6sites(reefsiiithroughviii)waschosenforpredationsurveysthatrepresentedthefull

suiteofcontextsinwhichreefswererestored.

StatisticalAnalysis

Environmentalvariables

Environmentalvariablescouldbebroadlygroupedintotwocategories(Table1).Firstly,we

accountedfordifferencesinexperimentaldesign(foroystergrowthmetricsonly;i.e.oystergrowth

method)andthetimingofsurveys(i.e.event).Secondly,weincludedfourvariablesrelatingtothe

landscapecontextoftheoysterreefrestorationsitesrelativetonearbyseagrassmeadows,

mangroveforests,hardenedshorelines(i.e.urbanisedshorelines),andtheestuarymouth.These

variableswerechosenbecausewehadspecifichypothesesforthewaysinwhichtheywouldmodify

thebenefitsofrestorationthatwerebasedontheresultsofpreviousstudiesintheregion(Table1).

Identifyingspatialpatternsinbenefitsofoysterreefrestoration

Wetestedforcorrelationsbetweenthesixrestorationbenefits(i.e.thedependentvariablesin

thesemodels)andenvironmentalvariablesusinggeneralisedadditivemodelsinthemgcvpackage

(Wood2017)ofR(RCoreTeam2018).ModeloverfittingwasminimisedbyfittingGAMswithfour

polynomiallinesorfewer(i.e.k<4),andbyrunningallpossiblecombinationsoffourvariablesor

fewer(BurnhamandAnderson2002).BestfitmodelswerethosewiththelowestAikaike

InformationCriterion(AIC)value.

Identifyingoysterhotspots

Weusedthebestfitmodelsforrestorationbenefitstocreatedistributionmodels(usingthepredict

functioninmgcv)forthewholeNoosaRiverestuaryat5mpixelresolution(chosentoreflectthesize

143

ofcurrentrestorationsites).AlldistributionmodelsweremappedinQGIS(QGISDevelopmentTeam

2019).Whereeithersamplingeventoroystergrowthmethodwaspresentinthebestfitmodel,we

quantifiedthedistributionmodelusingmeanvaluesforalllevelswithinthatfactor.

Valueswereextractedfromeachpixelforeachofthesixdistributionmodels.Valuesforeach

restorationbenefitwerestandardisedfrom0(theminimumvalueforeachtarget)and1(the

maximumvalueforeachtarget).Oysterreefrestorationhotspotswereidentifiedusingtwoseparate

oysterreefrestorationindices;thefirstbytakingasumofthesestandardisedvalues(resultinginan

indexrangingfrom0to6),andthesecond,takinganaverageofthesestandardisedvalues(resulting

inanindexrangingfrom0to1).Thedistributionoftheserestorationindiceswasthenalsomapped

inQGIS.

144

Table1Listofenvironmentalmetricsincludedinanalyses,theirdefinitions,hypothesesunderlyingtheirinclusioninmodels,andsources.

Metric Definitionandjustification Hypothesis-oystergrowth Hypothesis-fishassemblagesandecologicalfunctions

Datasource

Experimentaldesign Oystergrowthmethod Includedinoystergrowthanalysesonly.Twomethods-1)

shellsremoveddirectlyfromtheoysterrestorationunits,and2)markedshellsaffixedtotheoutsideoftheoysterrestorationunits.

- Notincluded -

Event Surveyperiodintimesincetheinstallationoftheoysterreefs.Patternsmaychangeovertimeorbetweenseasons.

- - -

Landscapecontext Seagrass Thepresenceorabsenceofseagrasswithin500mofeach

oysterreefsite.Reefsweredeliberatelyplacedintheestuarytobeeitherwithin500mofseagrasses,orbeyond500mofseagrasses;afactorshowninpreviousstudiestobeimportantinstructuringfishassemblagesinthisregion(Gilbyetal.2018b;Gilbyetal.2019).

Thepresenceofnearbyseagrasswillmodifythenumberandsizeofoysterswillbeloweratsitesneartoseagrassbyprovidingsubsidiesandmodifyingoysterpredationrisk.

Thepresenceofnearbyseagrasswillmodifyhabitatconnectivity,andthereforethenumberandtypeoffishcongregatingatoysterreefs.

Gilbyetal.(2018b)

Mangroves Thearea(inm2)ofmangroveswithin500mofeachoysterreefsite.Alloysterreefswereplacedwithin100mofmangroves,meaningthattheeffectsofproximitytomangroveswereconsistentacrosssitesinthisseascape(Gilbyetal.2018b;Gilbyetal.2017a).Therewere,however,variationsintheareaofmangroveswithin500mofthesites.

Theextentofnearbymangroveswillmodifythenumberandsizeofoystersbyprovidingsubsidiesandmodifyingoysterpredationrisk.

Theextentofnearbymangroveswillmodifyhabitatconnectivity,andthereforethenumberandtypeoffishcongregatingatoysterreefs.

QueenslandGovernment(2015b)

Distancetohardenedshorelines

Thedistance(inm)fromeachoysterreeftothenearesthardenedshoreline.Hardenedshorelinesareimportantsourcesofoysterspatandhomesforfish,andtheireffectsproliferateinpredictabledistancesintosurroundingseascapes(Brooketal.2018;Gilbyetal.2018b).

Hardenedshorelinesaresourcesforoysterlarvae,sooysterdensitywillbehighernearertohardenedshorelines,butsizemightbelowerduetohigherpredationrisk.

Theproximityofsitestohardenedshorelinesmodifieshabitatconnectivity,andthereforethenumberandtypeoffishcongregatingatoysterreefs.

QueenslandGovernment(2015a)

Distancetoestuarymouth

Thedistance(inm)fromeachoysterreeftothecentreoftheestuarymouth.Proximitytotheestuarymouthmodifieslikelihoodofrecruitmentandmigrationtodifferentsites(Gilbyetal.2018b;Hendersonetal.2017a).

Oystergrowthandabundancewillbegreaternearertotheestuarymouthduetohighersalinitylevels.

Fishdiversityandtheabundanceoftargetedfishspeciestendstobelowerwithincreasingdistancefromtheestuarymouth

QGISDevelopmentTeam(2019)

145

ResultsPatternsinthebenefitsofoysterreefrestoration

Weidentifiedextensiveoystergrowthontheoysterrestorationunits,withbetween145.7and946

oystersgrowingonaveragepersquaremetreofshell,andtoanaveragesizeofbetween12.4and

17.4mm(Table2).Thedensityofoystersgrowingontheoysterrestorationunitswashighestonthe

‘necklace’methodontheoutsideoftherestorationunits,andatrestorationsiteswithan

intermediateareaofmangrovesnearby,atdistancesapproximately3000mand6000mfromthe

estuarymouth,andatsitesnearertohardenedshorelines(Table3,FigureS1,Figure2A).Oysters

grewlargerwithincreasingtimesincerestoration,atrestorationsiteswithalowerareaof

mangrovesnearby,andatsitesnearertotheestuarymouthandfurtherfromhardenedshorelines

(Table3,FigureS1,Figure2B).

Diverseandabundantfishassemblagesrecruitedtotherestoredoysterreefrestorationunits,with

between2.3and5.3species,and2.5and10.3harvestablefishidentifiedonaverageateachoyster

restorationsitepersurvey(Table2).Previousstudieshaveshownthatthisequatesto1.4times

greaterspeciesrichnessand1.8timesmoreharvestablefishatrestorationsitesthanatnearby

controlsites(Gilbyetal.2019).Thespeciesrichnessoffishassemblesandtheabundanceof

harvestablefishcongregatingatoysterrestorationsiteswasvariableovertime,butalwayshigherat

siteswithoutseagrassnearby(Table3,FigureS2,Figure2C,D).Theabundanceofharvestablefish

wasalsohigheratoysterrestorationsiteswithalowerextentofmangrovesnearby(Table3,Figure

S2,Figure2D).

Predationratesrangedbetween13and81%likelihoodofpredation,andscavengingratesbetween

19and100%consumptiononaverage(Table2).Previousstudieshaveshownthatthisequatestoa

doublingofpredationratesatrestorationsitesthanatnearbycontrolsites(Duncan2018).

Scavengingrateswere32%higheratoysterrestorationsitesthanatnearbycontrolsites(FigureS3).

Ratesofpredationwerevariablethroughtimebutwerehighestatoysterrestorationsiteswithout

seagrassnearbyandwithlessthan50haofmangroveswithin500m(Table3,FigureS4,Figure2E).

Ratesofscavengingwashigheratoysterrestorationsitesprogressivelyfurtherfromtheestuary

mouthandatsiteswithlessthan45haofmangroveswithin500m,andlowestatoysterrestoration

sitesgreaterthan400mfromhardenedshorelines(Table3,FigureS4,Figure2F).

146

Table2Averagevalues(withstandarddeviation)foreachrestorationtargetateachofthe13oysterreefrestorationsites(perFigure1).

Site Oysterdensity(oysters/m2) Oystersize(mm) Fishspeciesrichness Harvestablefishabundance Predation(probability) Scavenging(proportionconsumed)

Av. StDev Av. StDev Av. StDev Av. StDev Av. StDev Av. StDev

i 149.4 249.6 17.4 8.1 2.8 1.9 3.0 2.1 - - 0.19 0.30

ii 161.8 225.8 13.0 4.8 4.5 3.4 8.5 7 0.13 0.34 0.50 0.37

iii 145.7 216.3 13.0 5.3 5.3 2.8 10.3 4.6 0.81 0.40 0.98 0.04

iv 722.6 665.9 14.0 6.4 4.3 2.9 6.1 1.9 0.56 0.50 0.97 0.03

v 271.7 345.1 15.5 5.6 3 1.6 2.5 1 0.38 0.49 0.24 0.21

vi 252.6 342.5 15.6 4.6 4.3 3.3 7.8 5.4 0.66 0.48 0.99 0.03

vii 946.0 888.1 14.0 5.2 4 0.8 6.5 3.1 0.78 0.42 1 0

viii 312.8 536.7 12.6 6.9 3 2.2 8.7 6.4 - - 0.62 0.45

ix 282.0 472.0 15.9 6.1 5.3 2.8 8.5 9.9 - - 0.88 0.23

x 317.9 403.1 12.7 6.3 4 3.3 6.5 5.9 - - 0.98 0.04

xi 424.9 467.2 14.9 5.4 4 1.4 6.8 7.7 - - 0.70 0.48

xii 627.9 493.3 13.4 5.7 2.3 0.5 5.5 4.1 - - 0.82 0.35

xiii 351.8 482.4 12.4 5.8 3.5 1.7 7.0 5.6 - - 0.40 0.45

147

Table3Resultsofgeneralisedadditivemodelandmulti-modelinferenceanalyses.Valuesprovidedforenvironmentalvariablesarevariableimportancevalues,withvalues

nearertooneindicatinggreatercontributiontowardstop-rankedmodels.Boldvaluesindicatevariablesincludedinbestfitmodelforeachrestorationtarget.R2and

weightvaluesprovidedfortheoverallbestfitmodel.‘Weight’istheweightedAICvalueforthebestfitmodel.‘Secondarymodels’indicatesthenumberofmodelswhich

explainedasimilaramountofvariancetothebest-fitmodel(i.e.thosewithin2AICcofthebestfitmodel),withvariablesinparenthesesbeingthoseincludedinadditional

secondarymodels.Wherenovariablesareprovided,secondarymodelscontainedonlycombinationsofthoseincludedinthebestfitmodel.

Environmentalvariables ModelattributesRestorationtarget Method Event Seagrass Mangroves Hardened

shorelinesDistancetoestuarymouth

R2 Weight Secondarymodels

Oysterdensity 1 0.33 0.07 1 0.42 1 0.23 0.42 2(Event)

Oystersize 0.18 1 0.22 0.38 0.99 0.56 0.14 0.21 3

Fishspeciesrichness - 0.96 0.36 0.28 0.31 0.27 0.17 0.13 5(Mouth,hardened

shorelines,mangroves)

Harvestablefish

abundance

- 0.99 0.6 0.48 0.28 0.31 0.2 0.21 5(Mouth,hardened

shorelines)

Predation - 0.95 0.49 0.84 0.45 0.37 0.25 0.19 7(Mouth,hardened

shorelines)

Scavenging - 0.2 0.41 0.85 0.53 0.71 0.4 0.17 4(Mouth,seagrass)

148

Figure2DistributionmodelsforA)oysterdensity,B)oystersize,C)fishspeciesrichness,D)harvestablefish

abundance,E)predationandF)scavengingcalculatedusingthebestfitgeneralisedadditivemodelsforeach

restorationbenefit(describedinTable2).

149

Hotspotsforrestorationeffectiveness

Becausesettlementandgrowthofoysterswasgreatenoughtoallowsuccessfulrestorationatall

sites,wedidnotneedtofirstconsiderwhererestorationcouldbesuccessfulintheNoosaEstuary.

Therefore,wesimplypooledthestandardisedvaluesforoystergrowthmetricswiththeother

restorationbenefitstoidentifyrestorationhotspots.

Whilstthesumandaverageapproachestocalculatingtherestorationindexdifferslightly

conceptually(i.e.totalrestorationbenefitsatsitesversusaveragerestorationbenefits),wefound

fewdifferencesintheoutcomeswhenusingthesumversusaverageapproachtocreatetheoyster

restorationindices(Figure3).Hotspotswhereoysterrestorationwouldresultinthegreatest

possibleoutcomesacrossalloftherestorationtargetsconsidered(i.e.restorationindexcategory6

infigures3AandB)occurredneartheestuarymouth,onthenorthernbankoftheestuary

approximately2kmfromtheestuarymouth,andinalargelake-likeinletinthefarwestofthe

estuary(knownlocallyasLakeDoonella).Thesehotspotswereinsimilarpositionsforbothsum

(Figure3A)andaverageapproaches(Figure3B).However,thetotalareaofcategory6restoration

indexvalues(i.e.highrestorationbenefits)wouldbereducedsomewhatifrestorationoccursonlyin

intertidalareas(i.e.thosesimilartothelocationsusedforrestorationsitesinthisstudy)(Figure3C).

Wefoundnolocationsthatwereclassifiedasthehighestrestorationindexcategoryforall

restorationbenefits.Patternsintherelativevaluesofeachoftherestorationbenefitsacrossthe

differentlevelsofrestorationindiceswereinconsistent(Table3).Here,therewasnoclearsequence

inincreasingvaluesforeachrestorationbenefitacrossincreasingrestorationindexcategories,

irrespectiveofwhichapproachwasusedtoquantifytheindex(Table4).Onlypredationwashighest

atcategory6thanallothercategories,whilstfishspeciesrichnesswasequalhighestwithatleast

oneotherrestorationcategory(Table4).Infact,often,thevaluesforseveraloftherestoration

benefitsaremuchloweratthehighercategories,andforscavenging,thevalueisactuallyatits

lowest.Theapproachesusedtocalculatetherestorationindexsimplymeansthatlocationswith

highscoreshavethegreatestoverallpositiveeffectsacrossthewholesuiteofmetricsquantified

anddoesnotnecessarilymeanthatallrestorationbenefitsaremaximised.

150

Table4Listofoysterrestorationindexcategories(fromFigure3),theaveragevaluesforeachrestorationtargetforcategory1(i.e.thelowestrestorationindexcategory),

andthepercentagedifferenceinaveragevaluesfromthecategoryoneaverageforeachofthesubsequentfourcategoriesforA)theapproachoftakingthesumof

standardisedrestorationtargetvalues,andB)takingtheaverageofstandardisedrestorationtargetvalues.

Oysterrestorationindexcategory Oysterdensity Oystersize Fishspeciesrichness Harvestablefishabundance Predation Scavenging

A.Sumapproach

Category1(minimum) 6.8oysters/m2 2.7mm 1.4species 2.2fish 0.2probability 0.49prop.consumed

2 11.6 1.4 0.0 -1.0 33.1 15.13 34.1 3.1 0.0 -6.2 214.6 16.34 -81.8 6.0 13.5 21.0 68.9 0.05 -18.6 2.5 14.0 20.9 104.4 6.16 2.8 6.4 14.0 16.6 253.5 -40.2

B.AverageApproach

Category1(minimum) 7.1oysters/m2 2.7mm 1.4species 2.2fish 0.2probability 0.6prop.consumed2 23.7 2.3 0.0 -4.6 176.3 -23.23 33.5 3.7 0.0 -7.8 298.3 8.94 -86.1 5.5 14.0 21.8 79.3 -18.05 -47.3 4.8 14.0 20.7 117.8 -24.46 -8.9 1.4 14.0 19.9 150.5 -10.3

151

Figure3Mapofthefinaloutputsoftherestorationprioritisationprocessillustratinghotspotsforoyster

restorationbenefitsintheNoosaRiver,Australiausingsummed(A)andaveraging(B)approaches,and(C)the

intertidalareaswithintheNoosaRiver.

152

DiscussionEcologicalrestorationprojectsareincreasinginprevalenceandsizeglobally(Brudvig2011;Palmer

etal.2014).Whilstthereisalonghistoryofsuccessfullyrestoringlostordegradedecosystems,

thereremainslittleevidencetosuggestthatrestorationprojectsaredesignedtoexplicitlymaximize

outcomesformultiplerestorationbenefitsinasystematicway(Gilbyetal.2018a).Inthisstudy,

oysterreefrestorationprovedsuccessfulfortheestablishmentofgrowingoysterreefsthatwill

eventuallycementtogetherandformself-sustainingoysterreefs(i.e.oysterdensityandsize).These

oysterreefsalsoattractedadiversityandabundanceoffish(i.e.fishspeciesrichnessand

harvestablefishabundance)thatperformseveralimportantecologicalfunctionsthatarealsogood

indicatorsofecosystemandfoodwebcondition(i.e.predationandscavenging).Therefore,properly

managedoysterrestorationunitscouldberestoredtoanyintertidalareaoftheNoosaRiverestuary,

andthiswouldresultinthesettlementandgrowthofoysters,andsomeadditionalbenefitsfor

ecologicalfunctioningandfishassemblagesmorebroadlyincomparisontonearbysandyormuddy

substrates.However,wealsofoundthatthecumulativebenefitsoffutureoysterreefrestoration

effortscanbemaximisedbymorecloselyconsideringthespatialplacementofrestorationsites.In

thissense,eachoftherestorationbenefitswetestedvariedsignificantlywithatleastonespatial

attributeofthebroaderlandscape.Thisresultedintheextentofareaswherethecombinedbenefits

ofoysterreefrestorationwerehighestbeingsignificantlynarrowerthantheareainwhichintertidal

oysterreefscouldtheoreticallygrow.Theapproachusedinthisstudyisdeliberatelyconceptually

andcomputationallysimpleandservestodemonstratehowdecisionscanbemadeacrossbroader

areasindefensible,butrelativelystraightforwardways.Usingsuchapproachestooptimisethe

placementofrestorationeffortsacrosslandscapewillhelptomaximisetheoverallbenefitsof

restoration,andpotentiallyservetoreducesocialandeconomiccosts.

Giventheinconsistentpatternsinrestorationindexscoreswiththelevelsofdifferentrestoration

benefits,thereareseveralpotentialpathwaysforselectingoptimalrestorationsites.Iffundingis

limited,restorationcouldbeplacedonlyinthecategory6restorationindexareasbecausethese

areasmaximisethetotaloutcomes.However,ifabroaderapproachispossibleduetogreater

fundingoradesiretospreadeffort,thenrestorationplanscouldselectsitesacrosstheestuarythat

fulfilspecificcriteriarelatingtotherestorationbenefits.Forexample,ifweassumethatthemost

importantaspectoftherestorationeffortisthesuccessfulgrowthofthereefsthemselves,then

restorationplanswouldstartwithacombinedmapofoystersettlementandgrowth(i.e.figure2A

andB),andthensubsetfromtheretoplaceseveralreefsindifferentareasthata)fulfilthisinitial

‘oystergrowth’criteria’andb)alsoinvitemorediversefishcommunitiesor,separately,ahigher

153

abundanceofharvestablefishorenhancedratesofecologicalfunctions.Thissecondapproach

amountstotheCAR(comprehensive,adequate,representative)principleofspatialconservation

planning(CommonwealthofAustralia1997).Theapproachcomprehensivelyrestoresoysterreefsto

areasthatprovideallofthedifferentbenefitsforwhichreefsprovideandallowsthemtobe

restoredinoptimallocationsforthatpurpose(asopposedtoonlyinareasthat,onaverage,satisfy

allrestorationtargetstosomedegree;perFigure3).Itisadequatetotheextentthatthereare

severalrepresentativesamplesrestoredtoeachofthoselocationsasacontingencyintheeventthat

oneisdestroyedorimpacted.Finally,itisrepresentativeintherestorationoffeaturesofoyster

reefsthatservetobothmaintaintheconditionofcoastalsystemsandprovidefortheecosystem

servicesforwhichpeoplecovet.

Systematicallyselectingpotentialrestorationlocationsbyusingquantitativeoutcomesfromexisting

restorationeffortshasseveralkeyadvantages.Firstly,sitescanbeselectedwithmoreconfidence

thatthefullsuiteofecological,socialandeconomicbenefitssoughtfromrestorationwilleventuate

followingtheinitialexpense(inbothfinancialandtimeoutlays)oftherestorationefforts(Gilbyetal.

2018a;Wilsonetal.2011).Similarly,usingquantitativeinformationtoprioritiseactionsmakes

decisionshighlydefensible,isrobusttochallenges,andcanbeusedtoadvocateforadditional

restoration(MargulesandPressey2000).Usingactualvaluesforeachoftherestorationbenefits

alsoallowsforquantitativegoalstobesetmoreeasily;anongoingchallengeforrestorationinmany

settings(PetersonandLipcius2003;RondininiandChiozza2010;Thompson2011).Further,this

approachhasthepotentialtoreducesocialconflictsbyreducingthetotalfootprintofrestoration

sitesandallowingforthestrategicavoidanceofareaswhererestorationisnotpossible(e.g.

roadways,harbors,areasofhumanhabitation),whilststillensuringmaximumderivedrestoration

benefits.Suchapproacheswillbecrucialasthescaleandnumberofrestorationprojectsincreases

globally.Spatialprioritisationalgorithms(e.g.Marxan,Zonation)thatbetterconsidersocial(e.g.

avoidanceofareasofexistingusage),andeconomic(e.g.costofrestoringinonepartoftheestuary,

versusanother)costswouldlikelyenhancetheoperationalisationoftheseconcepts(Wattsetal.

2009).Theseconsiderationswould,however,likelyservetofurtherreducethenumberofcandidate

sites,meaningthatincorporatinganexistingunderstandingofdesignsandrestorationsitingthat

likelymaximisemultiplebenefitsbecomesincreasinglyimportant.

Wefoundextensiverecruitmentandgrowthofoystersontheoysterrestorationunitswithinthe

NoosaRiverestuaryinallcontexts.Thisfindingsuggeststhatoysterreefrestorationwouldbe

successfulinanyintertidalareainthesystem,aslongasthosereefsareproperlymanagedfrom

154

externalinfluenceslikeboatstrike,harvesting,orsignificantsedimentation.Mangrovesprovide

sourcesofoysterspatthatgrowonproproots,butarealsohomesforpotentialoysterpredators

(Sheavesetal.2015).Therefore,restorationsiteswithintermediateareasofmangrovesnearby

likelycontainedthegreatestdensityofoystersbecausetheyhadgoodaccesstosettlingspatbut

wereseparatedspatiallyfrompotentialpredators.Mangrovesare,however,alsogoodsourcesof

organicmaterialforthegrowingoysters(Dittmaretal.2006),meaningthatlargeroysterswere

foundatsiteswithgreaterextentsofmangrovesnearby.Similarly,urbanstructuresareboth

sourcesofoysterlarvaeandpotentialoysterpredatorssuchasfishandlargecrustaceans(Brooket

al.2018).Oysterrestorationsitesnearertourbanshorelinesthereforecontainahigherdensityof

oysters,buttheseoystersaremorelikelytobeconsumedbyfishthatcongregatearoundurban

shorelinesbeforetheyreachasizerefuge.Effectsofproximitytotheestuarymoutharemore

complex,andlikelyrelatedtooystersurvivalwithvariablesalinityfurtherupstream(e.g.Lenihanet

al.1999),theprevalenceofpotentialoysterpredators(e.g.Soniatetal.2004),andperhapsabrasion

frommobilesandsnearertotheestuarymouth.Previousstudieshaveshownthatoystersgrow

largerandmorequicklyinmoresalinewaters,potentiallyduetolongerperiodsoftimewhere

oysterscanopentofeed(oystersclosewhenwateristoofresh)(DoveandO'Connor2007).These

findingsprovideimportantinsightsintotheareaswhereoysterrestorationwouldbesuccessfulin

theestuariesofthisregion.Ifrestoringthelostoysterhabitatwastheonlygoaloffuturerestoration

efforts,thenpractitionerscouldsimplyusethedistributionmodelsforoysterdensityandsize(i.e.

Figure2A,B)toidentifythemostsuccessfulrestorationsites.

Enhancingtheabundanceand/ordiversityoffishassemblagesisanimportantgoalofmanyoyster

reefrestorationprojectsglobally(Gilbyetal.2018c;zuErmgassenetal.2016).Wefoundthatthe

effectsofrestorationontherichnessoffishassemblages,theabundanceofharvestablefish

congregatingaroundoysterreefs,andtheecologicalfunctionsthesefishperformwerevariableover

time;thiswaslikelyduetotheeffectsofvariablefreshwaterrunoffinthesystem(Gilbyetal.2019).

Thereisanincreasingbodyofevidencetosuggestthatrestoringoysterreefsatsitesisolatedfrom

existingstructurallycomplexecosystemsresultsingreateraugmentationoffishspeciesrichnessand

abundance,andthiscanoftencascadetotheecologicalfunctionsthosefishprovide(Duncan2018;

Gilbyetal.2019;Grabowskietal.2005).Thisislikelybecausemoreisolatedrestorationactions

providenew,andstructurallycomplex,habitatinlocationsthatwerepreviouslylowcomplexity,

unvegetatedsoftsediments(Gilbyetal.2019).Thefindingsofthisstudyprovidefurtherevidenceto

supporttheseconclusionsregardingthesitingofoysterrestorationeffortsforfishandtheecological

functiontheyperform.Thesefindingsalsoprovidefurtherevidencetosupporttheuseofoysterreef

155

restorationasatargetedrecoverymechanismforrestoringmultipleaspectsofcoastalfish

assemblages,butalsohighlighttheimportanceofconsideringprojectdesignandsitinginorderto

maximisethesederivedbenefits.

Landscape-scalerestorationeffortsnecessitatemorethoroughmethodsandtheoriesforthe

planningofrestorationactions(Gilbyetal.2018a;Wilsonetal.2011).Asfinancialcostsand

likelihoodofsocialdisruptionsincreases,ensuringthatbenefitsderivedfromrestorationare

maximisedperunitcostwillbevital.Theprinciplesusedheretoselectpotentialhotspotsfor

restorationoutcomesaresimilarconceptuallytotheplacementofreservesinheterogenous

landscapes.Reserveselectionprocessesusemapsofthecurrentdistributionofhabitats,and

increasinglyconsiderhabitatconnectivity(Oldsetal.2016;Weeks2017),ecologicalfunctionsand

resilience(Oldsetal.2012c),andbothnaturalandhumandisturbances(Huijbersetal.2015),and

thenusequantitativeconservationgoalstoallocatereservelocation.Theseprincipleshave,

however,rarelybeenconsideredfortheplacementofrestorationsites(Gilbyetal.2018a).

Identifyingpotentialrestorationsitesusingquantitativegoalsbasedonourbestunderstandingof

locationswhererestorationwouldbemostsuccessful,butalsomaximisepotentialadditional

outcomes,shouldbecomethestandardapproachforplanningrestorationactionsacross

landscapes.

156

Supplementarymaterials

FigureS1Barplotsandgeneralisedadditivemodelpartialplotsfortheeffectsofenvironmentalvariablesontherestorationtargetsofoysterdensity(A),andoystersize

(B).Variablesinredboxeswereincludedinthebest-fitmodel.

Oyst

er d

ensit

y (o

yste

rs/m

2 )

Mangrove area (m2) Distance to estuary mouth (m) Distance to hardened shoreline (m)

Mod

elle

d ef

fect

SeagrassPresentAbsent

MethodBag Necklace

750

0

150

300

450

600

500

400

300

200

100

0

EventMay-18 Nov-18

Oyst

er si

ze (m

m)

0

5

10

15

20

Mod

elle

d ef

fect

Mangrove area (m2) Distance to estuary mouth (m) Distance to hardened shoreline (m)SeagrassMethodBag Necklace

EventMay-18 Nov-18

500

400

300

200

100

0

0

5

10

15

20

0

5

10

15

20

PresentAbsent

A. Oyster density

B. Oyster size

Method Event Seagrass Mangroves D. to estuary mouth D. to hardened shoreline

157

FigureS2Barplotsandgeneralisedadditivemodelpartialplotsfortheeffectsofenvironmentalvariablesontherestorationtargetsoffishspeciesrichness(A),and

harvestablefishabundance(B).Variablesinredboxeswereincludedinthebest-fitmodel.

Event1 2 3 4

Fish

spec

ies r

ichne

ss

0

2

4

6

8

0

2

4

6

SeagrassPresentAbsent

Harv

esta

ble

fish

abun

danc

e

Event1 2 3 4

0

3

6

9

12

SeagrassPresentAbsent

0

2

4

6

8

10

Mod

elle

d ef

fect

Mangrove area (m2)

Mangrove area (m2) Distance to estuary mouth (m) Distance to hardened shoreline (m)

Distance to hardened shoreline (m)Distance to estuary mouth (m)

Mod

elle

d ef

fect

A. Fish species richness

B. Harvestable fish abundance

Event Seagrass Mangroves D. to estuary mouth D. to hardened shoreline

158

FigureS3Average(+/-standarderror)scavengingratesatoysterrestorationsitesandnearbycontrolsites.

Controlsiteswereselectedtobeatleast200mfromanoysterreefsite,andsothatthesameseascape

contextswerecoveredbybothcontrolandrestorationsites(i.e.samenumberneartoseagrass,andwith

similarareasofmangrovesnearby).Notethatthesetwovaluesarehighlystatisticallydifferent(t-test,

P<0.001).

0

0.2

0.4

0.6

0.8

1

Control Oyster reef

Prop

ortio

n co

nsum

ed

159

FigureS4Barplotsandgeneralisedadditivemodelpartialplotsfortheeffectsofenvironmentalvariablesontherestorationtargetsofpredation(A),andscavenging(B).

Variablesinredboxeswereincludedinthebest-fitmodel.

A. Predation

B. Scavenging

Event Seagrass Mangroves D. to estuary mouth D. to hardened shoreline

Event1 2 3 4

Prob

abili

ty o

f pre

datio

n

0

0.4

0.6

0.8

1

0.2

Prop

ortio

n ca

rrio

n co

nsum

ed

Event1 2 3 4

SeagrassPresentAbsent

0

0.4

0.6

0.8

1

0.2

SeagrassPresentAbsent

0

0.25

0.5

0.75

1

Mangrove area (m2) Distance to estuary mouth (m) Distance to hardened shoreline (m)

Mod

elle

d ef

fect

Mangrove area (m2) Distance to estuary mouth (m) Distance to hardened shoreline (m)M

odel

led

effe

ct0

0.25

0.5

0.75

1

ReferencesAdameMF,HermosoV,PerhansK,LovelockCE,Herrera-SilveiraJA(2015)Selectingcost-effective

areasforrestorationofecosystemservicesConservationBiology29:493-502

doi:10.1111/cobi.12391

AertsR,HonnayO(2011)Forestrestoration,biodiversityandecosystemfunctioningBMCecology

11:29

AllewayH,CookP,CrawfordC,CreightonC,DigglesB,FordJ,GilliesC,HamerP,LebraultE,LePortA

(2015)Shellfishreefhabitats:asynopsistounderpintherepairandconservationof

Australia’senvironmental,socialandeconomicallyimportantbaysandestuariesReportto

theNationalEnvironmentalScienceProgramme,MarineBiodiversityHubCentreforTropical

WaterandAquaticEcosystemResearch(TropWATER)Publication,JamesCookUniversity,

Townsville

AlmanyGR,ConnollySR,HeathDD,HoganJD,JonesGP,McCookLJ,MillsM,PresseyRL,Williamson

DH(2009)Connectivity,biodiversityconservationandthedesignofmarinereserve

networksforcoralreefsCoralReefs28:339-351doi:10.1007/s00338-009-0484-x

AmorimE,RamosS,ElliottM,FrancoA,BordaloAA(2017)Habitatlossandgain:influenceon

habitatattractivenessforestuarinefishcommunitiesEstuarineCoastalandShelfScience

197:244-257

AndersonMJ,GorleyRN,ClarkeKR(2008)PERMANOVA+forPrimer:Guidetosoftwareand

statisticalmethods.PRIMER-ELtd,Plymouth,UnitedKingdom

AngelieriCCS,Adams-HoskingC,FerrazK,deSouzaMP,McAlpineCA(2016)UsingSpecies

DistributionModelstoPredictPotentialLandscapeRestorationEffectsonPuma

ConservationPLoSOne11doi:10.1371/journal.pone.0145232

ArturoSánchez-AzofeifaG,DailyGC,PfaffASP,BuschC(2003)IntegrityandisolationofCostaRica's

nationalparksandbiologicalreserves:examiningthedynamicsofland-coverchange

BiologicalConservation109:123-135doi:https://doi.org/10.1016/S0006-3207(02)00145-3

AudinoLD,MurphySJ,ZambaldiL,LouzadaJ,ComitaLS(2017)Driversofcommunityassemblyin

tropicalforestrestorationsites:roleoflocalenvironment,landscapeandspaceEcological

Applications:n/a-n/adoi:10.1002/eap.1562

AustralianShellfishReefRestorationNetwork(2018)AustralianShellfishReefRestorationNetwork.

https://www.shellfishrestoration.org.au/.AccessedJanuary2018

BachmanPM,RandGM(2008)EffectsofsalinityonnativeestuarinefishspeciesinSouthFlorida

Ecotoxicology17:591-597doi:10.1007/s10646-008-0244-7

161

BaggettLP,PowersSP,BrumbaughR,CoenLD,DeAngelisB,GreeneJ,HancockB,MorlockS(2014)

Oysterhabitatrestorationmonitoringandassessmenthandbook.TheNatureConservancy,

Arlington,VA,USA

BaggettLP,PowersSP,BrumbaughRD,CoenLD,DeAngelisBM,GreeneJK,HancockBT,MorlockSM,

AllenBL,BreitburgDL,BushekD,GrabowskiJH,GrizzleRE,GrosholzED,LaPeyreMK,

LuckenbachMW,McGrawKA,PiehlerMF,WestbySR,zuErmgassenPSE(2015)Guidelines

forevaluatingperformanceofoysterhabitatrestorationRestorationEcology23:737-745

doi:10.1111/rec.12262

BaguetteM,BlanchetS,LegrandD,StevensVM,TurlureC(2013)Individualdispersal,landscape

connectivityandecologicalnetworksBiologicalReviews88:310-326doi:10.1111/brv.12000

BakerR,SheavesM,JohnstonR(2015)Geographicvariationinmangrovefloodingandaccessibility

forfishesandnektoniccrustaceansHydrobiologia762:1-14doi:10.1007/s10750-015-2329-7

BalaguerL,EscuderoA,Martin-DuqueJF,MolaI,AronsonJ(2014)Thehistoricalreferencein

restorationecology:Re-definingacornerstoneconceptBiolConserv176:12-20

doi:10.1016/j.biocon.2014.05.007

BarbierEB,HackerSD,KennedyC,KochEW,StierAC,SillimanBR(2011)Thevalueofestuarineand

coastalecosystemservicesEcologicalmonographs81:169-193

BartholomewA,BohnsackJA(2005)Areviewofcatch-and-releaseanglingmortalitywith

implicationsforno-takereservesReviewsinFishBiologyandFisheries15:129-154

doi:10.1007/s11160-005-2175-1

BartonK(2018)Multi-ModelInferenceRpackageversion1421,https://cranr-

projectorg/web/packages/MuMIn/MuMInpdf

BauerS,EnsBJ,KlaassenM(2010)ManyroutesleadtoRome:potentialcausesforthemulti-route

migrationsystemofRedKnots,CalidriscanutusislandicaEcology91:1822-1831

BayraktarovE,SaundersMI,AbdullahS,MillsM,BeherJ,PossinghamHP,MumbyPJ,LovelockCE

(2016)ThecostandfeasibilityofmarinecoastalrestorationEcolAppl26:1055-1074

doi:10.1890/15-1077

BeckMW,BrumbaughRD,AiroldiL,CarranzaA,CoenLD,CrawfordC,DefeoO,EdgarGJ,HancockB,

KayMC,LenihanHS,LuckenbachMW,ToropovaCL,ZhangG,GuoX(2011)Oysterreefsat

riskandrecommendationsforconservation,restoration,andmanagementBioscience

61:107-116doi:10.1525/bio.2011.61.2.5

BeckMW,HeckKL,AbleKW,ChildersDL,EgglestonDB,GillandersBM,HalpernB,HaysCG,Hoshino

K,MinelloTJ,OrthRJ,SheridanPF,WeinsteinMR(2001)Theidentification,conservation,

162

andmanagementofestuarineandmarinenurseriesforfishandinvertebratesBioscience

51:633-641doi:10.1641/0006-3568(2001)051[0633:ticamo]2.0.co;2

BeckerA,TaylorMD,FolppH,LowryMB(2018)Managingthedevelopmentofartificialreefsystems:

TheneedforquantitativegoalsFishFish19:740-752doi:10.1111/faf.12288

BellSS,FonsecaMS,MottenLB(1997)LinkingrestorationandlandscapeecologyRestoration

ecology5:318-323

BenfieldMC,MinelloTJ(1996)Relativeeffectsofturbidityandlightintensityonreactivedistance

andfeedingofanestuarinefishEnvironmentalBiologyofFishes46:211-216

doi:10.1007/bf00005223

BishopMJ,Mayer-PintoM,AiroldiL,FirthLB,MorrisRL,LokeLHL,HawkinsSJ,NaylorLA,Coleman

RA,CheeSY,DaffornKA(2017)Effectsofoceansprawlonecologicalconnectivity:impacts

andsolutionsJournalofExperimentalMarineBiologyandEcology492:7-30

doi:10.1016/j.jembe.2017.01.021

BlaberSJ(2008)Tropicalestuarinefishes:ecology,explorationandconservation.JohnWiley&Sons,

BlaberS,BlaberT(1980)Factorsaffectingthedistributionofjuvenileestuarineandinshorefish

Journaloffishbiology17:143-162

BlandonA,zuErmgassenPSE(2014)Quantitativeestimateofcommercialfishenhancementby

seagrasshabitatinsouthernAustraliaEstuarine,CoastalandShelfScience141:1-8

doi:10.1016/j.ecss.2014.01.009

BosireJO,Dahdouh-GuebasF,WaltonM,CronaBI,LewisRR,FieldC,KairoJG,KoedamN(2008)

Functionalityofrestoredmangroves:AreviewAquatBot89:251-259

doi:10.1016/j.aquabot.2008.03.010

BostromC,PittmanSJ,SimenstadC,KneibRT(2011)Seascapeecologyofcoastalbiogenichabitats:

advances,gaps,andchallengesMarineEcologyProgressSeries427:191-217

doi:10.3354/meps09051

BouleyP,PoulosM,BrancoR,CarterNH(2018)Post-warrecoveryoftheAfricanlioninresponseto

large-scaleecosystemrestorationBiolConserv227:233-242

doi:10.1016/j.biocon.2018.08.024

BozecYM,Alvarez-FilipL,MumbyPJ(2015)Thedynamicsofarchitecturalcomplexityoncoralreefs

underclimatechangeGlobalChangeBiology21:223-235doi:10.1111/gcb.12698

BradyCJ,NoskeRA(2010)SuccessioninBirdandPlantCommunitiesovera24-YearChronosequence

ofMineRehabilitationintheAustralianMonsoonTropicsRestorationEcology18:855-864

doi:10.1111/j.1526-100X.2008.00511.x

163

BreitburgDL,PalmerMA,LoherT(1995)Larvaldistributionsandthespatialpatternsofsettlement

ofanoysterreeffish:responsestoflowandstructureMarineEcologyProgressSeries

125:45-60doi:10.3354/meps125045

BrickhillMJ,LeeSY,ConnollyRM(2005)Fishesassociatedwithartificialreefs:attributingchangesto

attractionorproductionusingnovelapproachesJFishBiol67:53-71

BrookTW,GilbyBL,OldsA,ConnollyRM,HendersonCJ,SchlacherTA(2018)Theeffectsofshoreline

armouringonestuarinefisharecontingentuponthebroaderurbanisationcontextMarEcol

ProgSerInPress

BroseU,HillebrandH(2016)Biodiversityandecosystemfunctioningindynamiclandscapes

PhilosophicaltransactionsoftheRoyalSocietyofLondonSeriesB,Biologicalsciences371

BrudvigLA(2011)Therestorationofbiodiversity:wherehasresearchbeenandwheredoesitneed

togo?AmJBot98:549-558doi:10.3732/ajb.1000285

BrudvigLA,DamschenEI,TewksburyJJ,HaddadNM,LeveyDJ(2009)Landscapeconnectivity

promotesplantbiodiversityspilloverintonon-targethabitatsProcNatlAcadSciUSA

106:9328-9332doi:10.1073/pnas.0809658106

BullockJM,AronsonJ,NewtonAC,PywellRF,Rey-BenayasJM(2011)Restorationofecosystem

servicesandbiodiversity:conflictsandopportunitiesTrendsinEcologyandEvolution

26:541-549doi:10.1016/j.tree.2011.06.011

BunnSE,ArthingtonAH(2002)Basicprinciplesandecologicalconsequencesofalteredflowregimes

foraquaticbiodiversityEnvironmentalmanagement30:492-507doi:10.1007/s00267-002-

2737-0

BureauofMeteorology(2019)Monthlyrainfallrecords.BureauofMeteorology,Canberra,Australia

BurnhamKP(2002)Informationandlikelihoodtheory:abasisformodelselectionandinference

Modelselectionandmultimodelinference:apracticalinformation-theoreticapproach:49-97

BurnhamKP,AndersonDR(2002)ModelSelectionandMultimodelInference.Springer,NewYork

ButchartSH,WalpoleM,CollenB,vanStrienA,ScharlemannJP,AlmondRE,BaillieJE,BomhardB,

BrownC,BrunoJ,CarpenterKE,CarrGM,ChansonJ,CheneryAM,CsirkeJ,DavidsonNC,

DentenerF,FosterM,GalliA,GallowayJN,GenovesiP,GregoryRD,HockingsM,KaposV,

LamarqueJF,LeveringtonF,LohJ,McGeochMA,McRaeL,MinasyanA,HernandezMorcillo

M,OldfieldTE,PaulyD,QuaderS,RevengaC,SauerJR,SkolnikB,SpearD,Stanwell-SmithD,

StuartSN,SymesA,TierneyM,TyrrellTD,VieJC,WatsonR(2010)Globalbiodiversity:

indicatorsofrecentdeclinesScience328:1164-1168doi:10.1126/science.1187512

CardinaleBJ,SrivastavaDS,DuffyJE,WrightJP,DowningAL,SankaranM,JouseauC(2006)Effectsof

biodiversityonthefunctioningoftrophicgroupsandecosystemsNature443:989

164

CicchettiG,GreeningH(2011)EstuarineBiotopeMosaicsandHabitatManagementGoals:An

ApplicationinTampaBay,FL,USAEstuariesandCoasts34:1278-1292doi:10.1007/s12237-

011-9408-4

CoenLD,BrumbaughDR,BushekD,GrizzleR,LuckenbachMW,PoseyMH,PowersSP,TolleySG

(2007)EcosystemservicesrelatedtooysterrestorationMarEcolProgSer341:303-307

CoenLD,GiottaRE,LuckenbachMW,BreitburgDL(1999)Oysterreeffunction,enhancement,and

restoration:Habitatdevelopmentandutilizationbycommercially-andecologically-

importantspeciesJournalofShellfishResearch18:712

CoenLD,LuckenbachMW(2000)Developingsuccesscriteriaandgoalsforevaluatingoysterreef

restoration:Ecologicalfunctionorresourceexploitation?EcologicalEngineering15:323-343

doi:10.1016/S0925-8574(00)00084-7

CoenLD,LuckenbachMW,BreitburgDL(1998)Theroleofoysterreefsasessentialfishhabitat:A

reviewofcurrentknowledgeandsomenewperspectives.In:BenakaLR(ed)FishHabitat:

EssentialFishHabitatandRehabilitation,vol22.AmericanFisheriesSocietySymposium.pp

438-454

CommonwealthofAustralia(1997)Nationallyagreedcriteriafortheestablishmentofa

comprehensive,adequateandrepresentativereservesystemforforestsinAustralia.

AustralianGovernmentPublishingService,Canberra,Australia.

ConnollyRM,HindellJS(2006)Reviewofnektonpatternsandecologicalprocessesinseagrass

landscapesEstuarineCoastalandShelfScience68:433-444doi:10.1016/j.ecss.2006.01.023

CosentinoBJ,SchooleyRL,BestelmeyerBT,KellyJF,CoffmanJM(2014)Constraintsandtimelagsfor

recoveryofakeystonespecies(Dipodomysspectabilis)afterlandscaperestoration

LandscapeEcology29:665-675doi:10.1007/s10980-014-0003-5

CreightonC,BoonPI,BrookesJD,SheavesM(2015)RepairingAustralia'sestuariesforimproved

fisheriesproduction-whatbenefits,atwhatcost?MarineandFreshwaterResearch66:493-

507doi:10.1071/mf14041

CrossmanND,BryanBA(2009)Identifyingcost-effectivehotspotsforrestoringnaturalcapitaland

enhancinglandscapemultifunctionalityEcolEcon68:654-668

doi:10.1016/j.ecolecon.2008.05.003

D’AgostiniA,GherardiDFM,PezziLP(2015)Connectivityormarineprotectedareasanditsrelation

withtotalkineticenergyPLoSOne10:e0139601

DameR,LibesS(1993)OysterreefsandnutrientretentionintidalcreeksJournalofExperimental

MarineBiologyandEcology171:251-258

165

daSilvaFR,MontoyaD,FurtadoR,MemmottJ,PizoMA,RodriguesRR(2015)Therestorationof

tropicalseeddispersalnetworksRestorationEcology23:852-860doi:10.1111/rec.12244

DavisB,BakerR,SheavesM(2014)Seascapeandmetacommunityprocessesregulatefish

assemblagestructureincoastalwetlandsMarEcolProgSer500:187-202

doi:10.3354/meps10680

DavisJP,PittKA,OldsAD,HarborneAR,ConnollyRM(2017)Seagrasscorridorsandtidalstate

modifyhowfishusehabitatsonintertidalcoralreefflatsMarEcolProgSer581:135-147

doi:10.3354/meps12311

DeckerE,LinkeS,HermosoV,GeistJ(2017)Incorporatingecologicalfunctionsinconservation

decisionmakingEcologyandEvolution7:8273-8281

DeGrootRS,BlignautJ,vanderPloegS,AronsonJ,ElmqvistT,FarleyJ(2013)BenefitsofInvestingin

EcosystemRestorationConservationBiology27:1286-1293doi:10.1111/cobi.12158

DerheMA,MurphyH,MonteithG,MenendezR(2016)Measuringthesuccessofreforestationfor

restoringbiodiversityandecosystemfunctioningJournalofAppliedEcology53:1714-1724

doi:10.1111/1365-2664.12728

DigglesBK(2013)Historicalepidemiologyindicateswaterqualitydeclinedriveslossofoyster

(Saccostreaglomerata)reefsinMoretonBay,AustraliaNZJMarFreshwaterRes47:561-581

doi:10.1080/00288330.2013.781511

DittmarT,HertkornN,KattnerG,LaraRJ(2006)Mangroves,amajorsourceofdissolvedorganic

carbontotheoceansGlobalBiogeochemicalCycles20:n/a-n/adoi:10.1029/2005gb002570

DoveMC,O'ConnorWA(2007)SalinityandtemperaturetoleranceofSydneyrockoystersSaccostrea

glomerataduringearlyontogenyJShellfishRes26:939-947doi:10.2983/0730-

8000(2007)26[939:Sattos]2.0.Co;2

DuffyJE,ZieglerSL,CampbellJE,BippusPM,LefcheckJS(2015)Squidpops:asimpletoolto

crowdsourceaglobalmapofmarinepredationintensityPLoSOne10:e0142994

DufreneM,LegendreP(1997)Speciesassemblagesandindicatorspecies:theneedforaflexible

asymmetricalapproachEcolMonogr67:345-366

DuncanCK(2018)Theeffectofhabitatrestorationontheecologicalfunctioningofcoastal

ecosystems.UniversityoftheSunshineCoast

EdgarGJ,Stuart-SmithRD,WillisTJ,KininmonthS,BakerSC,BanksS,BarrettNS,BecerroMA,

BernardAT,BerkhoutJ,BuxtonCD,CampbellSJ,CooperAT,DaveyM,EdgarSC,ForsterraG,

GalvanDE,IrigoyenAJ,KushnerDJ,MouraR,ParnellPE,ShearsNT,SolerG,StrainEM,

ThomsonRJ(2014)Globalconservationoutcomesdependonmarineprotectedareaswith

fivekeyfeaturesNature506:216-220doi:10.1038/nature13022

166

EdwardsHJ,ElliottIA,PresseyRL,MumbyPJ(2009)Incorporatingontogeneticdispersal,ecological

processesandconservationzoningintoreservedesignBiologicalConservation143:457–470

EhrenfeldJG(2000)DefiningtheLimitsofRestoration:TheNeedforRealisticGoalsRestorEcol8:2-9

doi:doi:10.1046/j.1526-100x.2000.80002.x

ElliottM(2011)Marinescienceandmanagementmeanstacklingexogenicunmanagedpressures

andendogenicmanagedpressures–anumberedguideMarinePollutionBulletin62:651-655

ElliottM,BurdonD,AtkinsJP,BorjaA,CormierR,deJongeVN,TurnerRK(2017)"AndDPSIRbegat

DAPSI(W)R(M)!"-AunifyingframeworkformarineenvironmentalmanagementMarine

PollutionBulletin118:27-40doi:10.1016/j.marpolbul.2017.03.049

ElliottM,BurdonD,HemingwayKL,ApitzSE(2007)Estuarine,coastalandmarineecosystem

restoration:Confusingmanagementandscience-ArevisionofconceptsEstuarineCoastal

ShelfSci74:349-366doi:10.1016/j.ecss.2007.05.034

ElliottM,CuttsND,TronoA(2014)Atypologyofmarineandestuarinehazardsandrisksasvectors

ofchange:areviewforvulnerablecoastsandtheirmanagementOcean&Coastal

Management93:88-99

ElliottM,ManderL,MazikK,SimenstadC,ValesiniF,WhitfieldA,WolanskiE(2016)Ecoengineering

withEcohydrology:SuccessesandfailuresinestuarinerestorationEstuarine,Coastaland

ShelfScience176:12-35doi:10.1016/j.ecss.2016.04.003

ElliottM,QuintinoV(2007)Theestuarinequalityparadox,environmentalhomeostasisandthe

difficultyofdetectinganthropogenicstressinnaturallystressedareasMarinePollution

Bulletin54:640-645doi:10.1016/j.marpolbul.2007.02.003

EngelhardSL,HuijbersCM,Stewart-KosterB,OldsAD,SchlacherTA,ConnollyRM(2017)Prioritising

seascapeconnectivityinconservationusingnetworkanalysisJournalofAppliedEcology

54:1130-1141doi:10.1111/1365-2664.12824

ESP(2016)AssessmentofOysterHabitatinNoosaRiverEstuary.EcologicalServiceProfessionals,

Brisbane,Australia

EstesJA,TinkerMT,BodkinJL(2010)Usingecologicalfunctiontodeveloprecoverycriteriafor

depletedspecies:seaottersandkelpforestsintheAleutianarchipelagoConservation

Biology24:852-860

EtienneRS(2004)Onoptimalchoicesinincreaseofpatchareaandreductionofinterpatchdistance

formetapopulationpersistenceEcologicalModelling179:77-90

doi:10.1016/j.ecolmodel.2004.05.003

FahrigL(2001)Howmuchhabitatisenough?BiologicalConservation100:65-74doi:10.1016/s0006-

3207(00)00208-1

167

Farinas-FrancoJM,RobertsD(2014)Earlyfaunalsuccessionalpatternsinartificialreefsusedfor

restorationofimpactedbiogenichabitatsHydrobiologia727:75-94doi:10.1007/s10750-013-

1788-y

FischerJ,LindenmayerDB(2007)Landscapemodificationandhabitatfragmentation:asynthesis

GlobalEcologyandBiogeography16:265-280doi:10.1111/j.1466-8238.2007.00287

FoamPE,MirzaRS,ChiversDP,BrownGE(2005)Juvenileconvictcichlids(Archocentrus

nigrofasciatus)allocateforagingandantipredatorbehaviourinresponsetotemporal

variationinpredationriskBehaviour142:129-144

FormanRTT(1995)Landmosaics:theecologyoflandscapesandregions.CambridgeUniversity

Press,Cambridge

FrainerA,PolviLE,JanssonR,McKieBG(2018)Enhancedecosystemfunctioningfollowingstream

restoration:therolesofhabitatheterogeneityandinvertebratespeciestraitsJournalof

AppliedEcology55:377-385

FroeseR,PaulyD(2018)Fishbase.

FullerRA,IrvineKN,Devine-WrightP,WarrenPH,GastonKJ(2007)Psychologicalbenefitsof

greenspaceincreasewithbiodiversityBiologyletters3:390-394

GainIE,BrewtonRA,RobillardMMR,JohnsonKD,SmeeDL,StunzGW(2017)Macrofaunausing

intertidaloysterreefvariesinrelationtopositionwithintheestuarinehabitatmosaic

MarineBiology164doi:10.1007/s00227-016-3033-5

GamfeldtL,RogerF(2017)Revisitingthebiodiversity–ecosystemmultifunctionalityrelationship

Natureecology&evolution1:0168

GeraldiNR,PowersSP,HeckKL,CebrianJ(2009)Canhabitatrestorationberedundant?Responseof

mobilefishesandcrustaceanstooysterreefrestorationinmarshtidalcreeksMarEcol-Prog

Ser389:171-180doi:10.3354/meps08224

GerberLR,BotsfordLW,HastingsA,PossinghamHP,GainesSD,PalumbiSR,AndelmanS(2003)

Populationmodelsformarinereservedesign:Aretrospectiveandprospectivesynthesis

EcologicalApplications13:S47-S64

GilbyBL,MaxwellPS,TibbettsIR,StevensT(2015)Bottom-upfactorsforalgalproductivityoutweigh

no-fishingmarineprotectedareaeffectsinamarginalcoralreefsystemEcosystems

18:1056-1069

GilbyBL,OldsAD,ConnollyRM,HendersonCJ,SchlacherTA(2018a)Spatialrestorationecology:

placingrestorationinalandscapecontextBioscience68:1007-1019

168

GilbyBL,OldsAD,ConnollyRM,MaxwellPS,HendersonCJ,SchlacherTA(2018b)Seagrassmeadows

shapefishassemblagesacrossestuarineseascapesMarineEcologyProgressSeries588:179-

189

GilbyBL,OldsAD,ConnollyRM,StevensT,HendersonCJ,MaxwellPS,TibbettsIR,SchoemanDS,

RissikD,SchlacherTA(2016)Optimisingland-seamanagementforinshorecoralreefsPLoS

One11:e0164934doi:10.1371/journal.pone.0164934

GilbyBL,OldsAD,ConnollyRM,YabsleyNA,MaxwellPS,TibbettsIR,SchoemanDS,SchlacherTA

(2017a)Umbrellascanworkunderwater:usingthreatenedspeciesasindicatorand

managementsurrogatescanimprovecoastalconservationEstuarineCoastalandShelf

Science199:132-140

GilbyBL,OldsAD,HendersonCJ,OrtodossiNL,ConnollyRM,SchlacherTA(2019)Seacapecontext

modifeshowfishrespondtorestoredoysterreefstructuresICESJournalofMarineScience

InPress

GilbyBL,OldsAD,PetersonCH,ConnollyRM,VossCM,BishopMJ,ElliottM,GrabowskiJH,Ortodossi

NL,SchlacherTA(2018c)Maximizingthebenefitsofoysterreefrestorationforfinfishand

theirfisheriesFishandFisheries19:931-947

GilbyBL,OldsAD,YabsleyNA,MaxwellPS,ConnollyRM,SchlacherTA(2017b)Enhancingthe

performanceofmarinereservesinestuaries:justaddwaterBiologicalConservation210:1-7

GilbyBL,StevensT(2014)Meta-analysisindicateshabitat-specificalterationstoprimaryproducer

andherbivorecommunitiesinmarineprotectedareasGlobalEcologyandConservation

2:289-299doi:10.1016/j.gecco.2014.10.005

GilliesCL,CreightonC,McLeodIM(2015a)Shellfishreefhabitats:asynopsistounderpintherepair

andconservationofAustralia’senvironmentally,sociallyandeconomicallyimportantbays

andestuaries.ReporttotheNationalEnvironmentalScienceProgramme,Marine

BiodiversityHub.CentreforTropicalWaterandAquaticEcosystemResearch(TropWATER)

Publication,JamesCookUniversity,Townsville,Australia

GilliesCL,FitzsimonsJA,BraniganS,HaleL,HancockB,CreightonC,AllewayH,BishopMJ,BrownS,

ChamberlainD,ClevelandB,CrawfordC,CrawfordM,DigglesB,FordJR,HamerP,HartA,

JohnstonE,McDonaldT,McLeodI,PinnerB,RussellK,WinstanleyR(2015b)Scaling-up

marinerestorationeffortsinAustraliaEcologicalManagement&Restoration16:84-85

doi:10.1111/emr.12159

GittmanRK,PetersonCH,CurrinCA,FodrieFJ,PiehlerMF,BrunoJF(2016)Livingshorelinescan

enhancethenurseryroleofthreatenedestuarinehabitatsEcologicalApplications26:249-

263doi:10.1890/14-0716

169

González-TokmanD,Martínez-GarzaC(2015)EffectofEcologicalRestorationonBodyConditionofa

PredatorPLoSOne10:e0133551doi:10.1371/journal.pone.0133551

GormanD,RussellBD,ConnellSD(2009)Land-to-seaconnectivity:linkinghuman-derivedterrestrial

subsidiestosubtidalhabitatchangeonopenrockycoastsEcologicalApplications19:1114-

1126doi:10.1890/08-0831.1

GrabowskiJH,BrumbaughRD,ConradRF,KeelerAG,OpaluchJJ,PetersonCH,PiehlerMF,Powers

SP,SmythAR(2012)Economicvaluationofecosystemservicesprovidedbyoysterreefs

Bioscience62:900-909doi:10.1525/bio.2012.62.10.10

GrabowskiJH,HughesAR,KimbroDL,DolanMA(2005)Howhabitatsettinginfluencesrestored

oysterreefcommunitiesEcology86:1926-1935doi:10.1890/04-0690

GrabowskiJH,PetersonCH(2007)Restoringoysterreefstorecoverecosystemservices.In:

CuddingtonK,ByersJE,WilsonWG,HastingsA(eds)EcosystemEngineers.ElsevierAcademic

Press,Burlington,MA,pp281–298

GrabowskiJH,PowersSP(2004)HabitatcomplexitymitigatestrophictransferonoysterreefsMar

Ecol-ProgSer277:291-295

GratwickeB,SpeightMR(2005a)EffectsofhabitatcomplexityonCaribbeanmarinefish

assemblagesMarineEcologyProgressSeries292:301-310doi:10.3354/meps292301

GratwickeB,SpeightMR(2005b)Therelationshipbetweenfishspeciesrichness,abundanceand

habitatcomplexityinarangeofshallowtropicalmarinehabitatsJFishBiol66:650-667

doi:10.1111/j.1095-8649.2005.00629.x

GrayJS,ElliottM(2009)EcologyofMarineSediments:sciencetomanagement.OxfordUniversity

Press,Oxford

GraysonJE,ChapmanMG,UnderwoodAJ(1999)Theassessmentofrestorationofhabitatinurban

wetlandsLandscapeandUrbanPlanning43:227-236

GreenAL,FernandesL,AlmanyG,AbesamisR,McLeodE,AlinoPM,WhiteAT,SalmR,TanzerJ,

PresseyRL(2014)DesigningMarineReservesforFisheriesManagement,Biodiversity

Conservation,andClimateChangeAdaptationCoastalManagement42:143-159

doi:10.1080/08920753.2014.877763

GregalisKC,JohnsonMW,PowersSP(2009)RestoredOysterReefLocationandDesignAffect

ResponsesofResidentandTransientFish,Crab,andShellfishSpeciesinMobileBay,

AlabamaTransactionsoftheAmericanFisheriesSociety138:314-327doi:10.1577/T08-041.1

GregalisKC,PowersSP,HeckJrKL(2008)Restorationofoysterreefsalongabio-physicalgradientin

MobileBay,AlabamaJournalofShellfishResearch27:1163-1169doi:10.2983/0730-8000-

27.5.1163

170

Grober-DunsmoreR,PittmanSJ,CaldowC,KendallMS,FrazerTK(2009)Alandscapeecology

approachforthestudyofecologicalconnectivityacrosstropicalmarineseascapes.

EcologicalConnectivityamongTropicalCoastalEcosystems.doi:10.1007/978-90-481-2406-

0_14

GuerreroAM,ShooL,IaconaG,StandishRJ,CatterallCP,RumpffL,deBieK,WhiteZ,MatzekV,

WilsonKA(2017)Usingstructureddecision-makingtosetrestorationobjectiveswhen

multiplevaluesandpreferencesexistRestorEcol25:858-865doi:10.1111/rec.12591

GuidettiP,MilazzoM,BussottiS,MolinariA,MurenuM,PaisA,SpanòN,BalzanoR,AgardyT,Boero

F,CarradaG,Cattaneo-ViettiR,CauA,ChemelloR,GrecoS,ManganaroA,Notarbartolodi

SciaraG,RussoGF,TunesiL(2008)Italianmarinereserveeffectiveness:Doesenforcement

matter?BiologicalConservation141:699-709doi:10.1016/j.biocon.2007.12.013

GustafsonEJ(1998)Quantifyinglandscapespatialpattern:Whatisthestateoftheart?Ecosystems

1:143-156doi:10.1007/s100219900011

GustafsonEJ,GardnerRH(1996)Theeffectoflandscapeheterogeneityontheprobabilityofpatch

colonizationEcology77:94-107

HaggerV,DwyerJ,WilsonK(2017)Whatmotivatesecologicalrestoration?RestorEcol25:832-843

doi:10.1111/rec.12503

HallettLM,DiverS,EitzelMV,OlsonJJ,RamageBS,SardinasH,Statman-WeilZ,SudingKN(2013)Do

WePracticeWhatWePreach?GoalSettingforEcologicalRestorationRestorEcol21:312-

319doi:10.1111/rec.12007

HalpernBS,GainesSD,WarnerRR(2004)Confoundingeffectsoftheexportofproductionandthe

displacementoffishingeffortfrommarinereservesEcologicalApplications14:1248-1256

doi:10.1890/03-5136

HalpernBS,KleinCJ,BrownCJ,BegerM,GranthamHS,MangubhaiS,RuckelshausM,TullochVJ,

WattsM,WhiteC,PossinghamHP(2013)Achievingthetriplebottomlineinthefaceof

inherenttrade-offsamongsocialequity,economicreturnandconservationProceedingsof

theNationalAcademyofSciences110:6229-6234

HalpernBS,LesterSE,KellnerJB(2009)Spilloverfrommarinereservesandthereplenishmentof

fishedstocksEnvironmentalConservation36:268-276

HalpernBS,WalbridgeS,SelkoeKA,KappelCV,MicheliF,D'AgrosaC,BrunoJF,CaseyKS,EbertC,

FoxHE,FujitaR,HeinemannD,LenihanHS,MadinEM,PerryMT,SeligER,SpaldingM,

SteneckR,WatsonR(2008)AglobalmapofhumanimpactonmarineecosystemsScience

319:948-952doi:10.1126/science.1149345

171

HannanJC,WilliamsRJ(1998)Recruitmentofjuvenilemarinefishestoseagrasshabitatina

temperateAustralianestuaryEstuaries21:29–51

HanskiI,SaccheriI(2006)Molecular-levelvariationaffectspopulationgrowthinabutterfly

metapopulationPlosBiology4:719-726doi:10.1371/journal.pbio.0040129

HardingJM(1999)SelectivefeedingbehavioroflarvalnakedgobiesGobiosomaboscandblennies

ChasmodesbosquianusandHypsoblenniushentzi:preferencesforbivalveveligersMarEcol-

ProgSer179:145-153

HardingJM,MannR(1999)Fishspeciesrichnessinrelationtorestoredoysterreefs,Piankatank

River,VirginiaBullMarSci65:289-299

HardingJM,MannR(2001)Oysterreefsasfishhabitat:Opportunisticuseofrestoredreefsby

transientfishesJournalofShellfishResearch20:951-959

HarrisonHB,WilliamsonDH,EvansRD,AlmanyGR,ThorroldSR,RussGR,FeldheimKA,van

HerwerdenL,PlanesS,SrinivasanM,BerumenML,JonesGP(2012)Larvalexportfrom

marinereservesandtherecruitmentbenefitforfishandfisheriesCurrentBiology22:1023-

1028doi:10.1016/j.cub.2012.04.008

HarwellHD,PoseyMH,AlphinTD(2011)Landscapeaspectsofoysterreefs:Effectsoffragmentation

onhabitatutilizationJournalofExperimentalMarineBiologyandEcology409:30-41

doi:10.1016/j.jembe.2011.07.036

HavstadKM,HerrickJE(2003)Long-termecologicalmonitoringAridLandResearchand

Management17:389-400doi:10.1080/15324980390229300

HeckKL,HaysG,OrthRJ(2003)Criticalevaluationofthenurseryrolehypothesisforseagrass

meadowsMarEcolProgSer253:123-136doi:10.3354/meps253123

HectorA,BagchiR(2007)BiodiversityandecosystemmultifunctionalityNature448:188-U186

doi:10.1038/nature05947

HendersonCJ,OldsAD,LeeSY,GilbyBL,MaxwellPS,ConnollyRM,StevensT(2017a)Marine

reservesandseascapecontextshapefishassemblagesinseagrassecosystemsMarine

EcologyProgressSeries566:135-144

HendersonCJ,StevensT,GilbyBL,LeeSY(2017b)Spatialconservationoflargemobile

elasmobranchsrequiresanunderstandingofspatio-temporalseascapeutilisationICESJMar

Sci72:553-561

HermosoV,AbellR,LinkeS,BoonP(2016)Theroleofprotectedareasforfreshwaterbiodiversity

conservation:challengesandopportunitiesinarapidlychangingworldAquatic

Conservation-MarineandFreshwaterEcosystems26:3-11doi:10.1002/aqc.2681

172

HidalgoM,SecorDH,BrowmanHI(2016)Observingandmanagingseascapes:linkingsynoptic

oceanography,ecologicalprocesses,andgeospatialmodellingICESJMarSci73:1825-1830

doi:10.1093/icesjms/fsw079

HiltyJA,LidickerWZ,MerenlenderAM(2006)CorridorEcology:Thescienceandpracticeoflinking

landscapesforbiodiversityconservation.IslandPress,Washington,DC,USA

HodgsonJA,ThomasCD,CinderbyS,CambridgeH,EvansP,HillJK(2011)Habitatre-creation

strategiesforpromotingadaptationofspeciestoclimatechangeConservationLetters

4:289-297doi:10.1111/j.1755-263X.2011.00177.x

HodgsonJA,ThomasCD,WintleBA,MoilanenA(2009)Climatechange,connectivityand

conservationdecisionmaking:backtobasicsJournalofAppliedEcology46:964-969

doi:10.1111/j.1365-2664.2009.01695.x

HollKD,CroneEE,SchultzCB(2003)Landscaperestoration:movingfromgeneralitiesto

methodologiesBioscience53:491-502

HoweHF,Martinez-GarzaC(2014)RestorationasexperimentBotanicalSciences92:459-468

HughesTP,BellwoodDR,FolkeC,SteneckRS,WilsonJ(2005)Newparadigmsforsupportingthe

resilienceofmarineecosystemsTrendsEcolEvol20:380-386

HuijbersCM,ConnollyRM,PittKA,SchoemanDS,SchlacherTA,BurfeindDD,SteeleC,OldsAD,

MaxwellPS,BabcockRC,RissikD(2015)Conservationbenefitsofmarinereservesare

undiminishednearcoastalriversandcitiesConservationLetters8:312-319

doi:10.1111/conl.12128

HumphriesAT,LaPeyreMK(2015)Oysterreefrestorationsupportsincreasednektonbiomassand

potentialcommercialfisheryvaluePeerj3doi:10.7717/peerj.1111

HuntingtonBE,LirmanD(2012)Species-arearelationshipsincoralcommunities:evaluating

mechanismsforacommonlyobservedpatternCoralReefs31:929-938doi:10.1007/s00338-

012-0917-9

IkinK,TullochA,GibbonsP,AnsellD,SeddonJ,LindenmayerD(2016)Evaluatingcomplementary

networksofrestorationplantingsforlandscape-scaleoccurrenceoftemporallydynamic

speciesConservationBiology30:1027-1037doi:10.1111/cobi.12730

IrlandiE,CrawfordM(1997)Habitatlinkages:theeffectofintertidalsaltmarshesandadjacent

subtidalhabitatsonabundance,movement,andgrowthofanestuarinefishOecologia

110:222-230

JacksonCR,PringleCM(2010)Ecologicalbenefitsofreducedhydrologicconnectivityinintensively

developedlandscapesBioscience60:37-46doi:10.1525/bio.2010.60.1.8

173

JelbartJE,RossPM,ConnollyRM(2007)Fishassemblagesinseagrassbedsareinfluencedbythe

proximityofmangroveforestsMarineBiology150:993-1002

JohnsonKD,SmeeDL(2014)Predatorsinfluencethetidaldistributionofoysters(Crassostrea

virginica)MarineBiology161:1557-1564doi:10.1007/s00227-014-2440-8

JonesME,DavidsonN(2016a)Applyingananimal-centricapproachtoimproveecological

restorationRestorationEcology24:836-842doi:10.1111/rec.12447

JonesME,DavidsonN(2016b)Applyingananimal-centricapproachtoimproveecological

restorationRestorationEcology24:836-842

Kaiser-BunburyCN,MougalJ,WhittingtonAE,ValentinT,GabrielR,OlesenJM,BlüthgenN(2017)

EcosystemrestorationstrengthenspollinationnetworkresilienceandfunctionNature

542:223

KasoarT,zuErmgassenPSE,CarranzaA,HancockB,SpaldingM(2015)Newopportunitiesfor

conservationofathreatenedbiogenichabitat:aworldwideassessmentofknowledgeon

bivalve-reefrepresentationinmarineandcoastalRamsarSitesMarFreshwRes66

doi:10.1071/mf14306

KeenleysideKA,DudleyN,CairnsS,HallCM,StoltonS(2012)Ecologicalrestorationforprotected

areas;principles,guidelinesandbestpractices.IUCN,Gland,Switzerland

KelloggML,CornwellJC,OwensMS,PaynterKT(2013)Denitrificationandnutrientassimilationona

restoredoysterreefMarineEcologyProgressSeries480:1-19

KilpatrickAM,GillinCM,DaszakP(2009)Wildlife–livestockconflict:theriskofpathogen

transmissionfrombisontocattleoutsideYellowstoneNationalParkJApplEcol46:476-485

KimbroDL,ByersJE,GrabowskiJH,HughesAR,PiehlerMF(2014)Thebiogeographyoftrophic

cascadesonUSoysterreefsEcolLett17:845-854

Kingsley-SmithP,StoneB,KepplerB,LefflerJ(2015)Expandinglivingshorelinesthrough

stakeholder-drivensiteselectionsforintertidaloysterreefbuildingintheAceBasinNerr,

SouthCarolina,USAJournalofShellfishResearch34:647-648

KirbyMX(2004)Fishingdownthecoast:Historicalexpansionandcollapseofoysterfisheriesalong

continentalmarginsProceedingsoftheNationalAcademyofScience101:13096-13099

KleinCJ,JupiterSD,SeligER,WattsME,HalpernBS,KamalM,RoelfsemaC,PossinghamHP(2012)

ForestconservationdelivershighlyvariablecoralreefconservationoutcomesEcological

Applications22:1246-1256

KondolfGM,BoultonAJ,O'DanielS,PooleGC,RahelFJ,StanleyEH,WohlE,BångA,CarlstromJ,

CristoniC(2006)Process-basedecologicalriverrestoration:visualizingthree-dimensional

connectivityanddynamicvectorstorecoverlostlinkagesEcologyandsociety11

174

KoolJT,MoilanenA,TremlEA(2013)Populationconnectivity:recentadvancesandnewperspectives

LandscapeEcology28:165–185

KormannU,ScherberC,TscharntkeT,KleinN,LarbigM,ValenteJJ,HadleyAS,BettsMG(2016)

Corridorsrestoreanimal-mediatedpollinationinfragmentedtropicalforestlandscapesProc

RSocB283:20152347

LaPeyreM,FurlongJ,BrownLA,PiazzaBP,BrownK(2014a)Oysterreefrestorationinthenorthern

GulfofMexico:Extent,methodsandoutcomesOcean&CoastalManagement89:20-28

doi:10.1016/j.ocecoaman.2013.12.002

LaPeyreMK,HumphriesAT,CasasSM,LaPeyreJF(2014b)Temporalvariationindevelopmentof

ecosystemservicesfromoysterreefrestorationEcologicalEngineering63:34-44

doi:10.1016/j.ecoleng.2013.12.001

LaPeyreMK,NixA,LabordeL,PiazzaBP(2012)Gaugingstate-levelandusergroupviewsofoyster

reefrestorationactivitiesinthenorthernGulfofMexicoOcean&CoastalManagement

67:1-8doi:10.1016/j.ocecoaman.2012.06.001

LaszloE,GyorgyKD,ZoltanB,BenceK,CsabaN,JanosKP,CsabaT(2018)Habitatheterogeneityasa

keytohighconservationvalueinforest-grasslandmosaicsBiolConserv226:72-80

doi:10.1016/j.biocon.2018.07.029

LaymanCA,AllgeierJE,MontañaCG(2016)Mechanisticevidenceofenhancedproductionon

artificialreefs:AcasestudyinaBahamianseagrassecosystemEcologicalEngineering

95:574-579doi:https://doi.org/10.1016/j.ecoleng.2016.06.109

LaymanCA,AllgeierJE,YeagerLA,StonerEW(2013)Thresholdsofecosystemresponsetonutrient

enrichmentfromfishaggregationsEcology94:530-536

LaymanCA,JudZR,ArcherSK,RieraD(2014)Provisionofecosystemservicesbyhuman-made

structuresinahighlyimpactedestuaryEnvironmentalResearchLetters9:044009

LédéeEJI,SuttonSG,TobinRC,DeFreitasDM(2012)Responsesandadaptationstrategiesof

commercialandcharterfisherstozoningchangesintheGreatBarrierReefMarinePark

MarinePolicy36:226-234doi:http://dx.doi.org/10.1016/j.marpol.2011.05.009

LeesAC,PeresCA(2008)Conservationvalueofremnantriparianforestcorridorsofvaryingquality

forAmazonianbirdsandmammalsConservationBiology22:439-449doi:10.1111/j.1523-

1739.2007.00870.x

LehnertRL,AllenDM(2002)NektonuseofsubtidaloystershellhabitatinasoutheasternUSestuary

Estuaries25:1015-1024doi:10.1007/bf02691348

175

LenihanHS(1999)Physical-biologicalcouplingonoysterreefs:Howhabitatstructureinfluences

individualperformanceEcologicalMonographs69:251-275doi:10.1890/0012-

9615(1999)069[0251:pbcoor]2.0.co;2

LenihanHS,MicheliF,SheltonSW,PetersonCH(1999)Theinfluenceofmultipleenvironmental

stressorsonsusceptibilitytoparasites:Anexperimentaldeterminationwithoysters

LimnologyandOceanography44:910-924

LenihanHS,PetersonCH,AllenJM(1996)Doesflowspeedalsohaveadirecteffectongrowthof

activesuspension-feeders:AnexperimentaltestonoystersLimnolOceanogr41:1359-1366

LenihanHS,PetersonCH,ByersJE,GrabowskiJH,ThayerGW,ColbyDR(2001)Cascadingofhabitat

degradation:OysterreefsinvadedbyrefugeefishesescapingstressEcologicalApplications

11:764-782doi:10.2307/3061115

LerbergSB,HollandAF,SangerDM(2000)Responsesoftidalcreekmacrobenthiccommunitiesto

theeffectsofwatersheddevelopmentEstuaries23:838-853doi:10.2307/1353001

LipciusRN,EgglestonDB,SchreiberSJ,SeitzRD,ShenJ,SissonM,StockhausenWT,WangHV(2008)

Importanceofmetapopulationconnectivitytorestockingandrestorationofmarinespecies

ReviewsinFisheriesScience16:101–110

LokeLH,LadleRJ,BoumaTJ,ToddPA(2015)Creatingcomplexhabitatsforrestorationand

reconciliationEcolEng77:307-313

LotzeHK,LenihanHS,BourqueBJ,BradburyRH,CookeRG,KayMC,KidwellSM,KirbyMX,Peterson

CH,JacksonJBC(2006)Depletion,degradation,andrecoverypotentialofestuariesand

coastalseasScience312:1806-1809doi:10.1126/science.1128035

LuntJ,SmeeDL(2015)Turbidityinterfereswithforagingsuccessofvisualbutnotchemosensory

predatorsPeerj3:e1212doi:10.7717/peerj.1212

MagrisRA,PresseyRL,WeeksR,BanNC(2014)Integratingconnectivityandclimatechangeinto

marineconservationplanningBiologicalConservation170:207-221

doi:10.1016/j.biocon.2013.12.032

MagrisRA,TremlEA,PresseyRL,WeeksR(2016)Integratingmultiplespeciesconnectivityand

habitatqualityintoconservationplanningforcoralreefsEcography39:649-664

doi:10.1111/ecog.01507

MargulesCR,PresseyRL(2000)SystematicconservationplanningNature405:243-253

doi:10.1038/35012251

MarshallS,ElliottM(1998)EnvironmentalinfluencesonthefishassemblageoftheHumberestuary,

UKEstuarineCoastalandShelfScience46:175-184doi:10.1006/ecss.1997.0268

176

MartinTSH,OldsAD,OlaldeABH,BerkströmC,GilbyBL,SchlacherTA,ButlerIR,YabsleyNA,Zann

M,ConnollyRM(2018)Habitatproximityexertsopposingeffectsonkeyecologicalfunctions

LandscEcol33:1273-1286doi:10.1007/s10980-018-0680-6

MartinTSH,OldsAD,PittKA,JohnstonAB,ButlerIR,MaxwellPS,ConnollyRM(2015)Effective

protectionoffishoninshorecoralreefsdependsonthescaleofmangrove-reefconnectivity

MarEcol-ProgSer527:157-165

MasloB,BurgerJ,HandelSN(2012)Modelingforagingbehaviorofpipingploverstoevaluatehabitat

restorationsuccessJournalofWildlifeManagement76:181-188doi:10.1002/jwmg.210

MasloB,HandelSN,PoverT(2011)RestoringBeachesforAtlanticCoastPipingPlovers(Charadrius

melodus):AClassificationandRegressionTreeAnalysisofNest-SiteSelectionRestoration

Ecology19:194-203doi:10.1111/j.1526-100X.2010.00709.x

MassolF,GravelD,MouquetN,CadotteMW,FukamiT,LeiboldMA(2011)Linkingcommunityand

ecosystemdynamicsthroughspatialecologyEcolLett14:313-323doi:10.1111/j.1461-

0248.2011.01588.x

McAlpineC,CatterallCP,MacNallyR,LindenmayerD,ReidJL,HollKD,BennettAF,RuntingRK,

WilsonK,HobbsRJ,SeabrookL,CunninghamS,MoilanenA,MaronM,ShooL,LuntI,VeskP,

RumpffL,MartinTG,ThomsonJ,PossinghamH(2016)Integratingplant-andanimal-based

perspectivesformoreeffectiverestorationofbiodiversityFrontiersinEcologyandthe

Environment14:37-45doi:10.1002/16-0108.1

McClanahanTR,MangiS(2000)Spilloverofexploitablefishesfromamarineparkanditseffecton

theadjacientfisheryEcolAppl10:1792-1805

McDonaldT,JonsonJ,DixonKW(2016)Nationalstandardsforthepracticeofecologicalrestoration

inAustraliaRestorationEcology24:S4-S32.

McGarigalK,CushmanSA,EneE(2012)FRAGSTATSv4:SpatialPatternAnalysisProgramfor

CategoricalandContinuousMaps.Computersoftwareprogramproducedbytheauthorsat

theUniversityofMassachusetts,Amherst.

http://www.umass.edu/landeco/research/fragstats/fragstats.html.AccessedMay2017

2017

MetzgerJP,BrancalionPHS(2016)LandscapeEcologyandRestorationProcesses.In:PalmerMA,

FalkDA,ZedlerJB(eds)FoundationsofRestorationEcology.SecondEdition.IslandPress,

Washington,DC

MicheliF,HalpernBS(2005)LowfunctionalredundancyincoastalmarineassemblagesEcology

Letters8:391-400

177

MicheliF,PetersonCH(1999)Estuarinevegetatedhabitatsascorridorsforpredatormovements

ConservationBiology13:869-881doi:10.1046/j.1523-1739.1999.98233.x

MiddendorpRS,PerezAJ,MolinaA,LambinEF(2016)Thepotentialtorestorenativewoodyplant

richnessandcompositioninareforestinglandscape:amodelingapproachintheEcuadorian

AndesLandscapeEcology31:1581-1599doi:10.1007/s10980-016-0340-7

MillerMW(2002)UsingecologicalprocessestoadvanceartificialreefgoalsIcesJournalofMarine

Science59:S27-S31doi:10.1006/jmsc.2001.1162

MillerJR,HobbsRJ(2007)Habitatrestoration-Doweknowwhatwe'redoing?RestorationEcology

15:382-390doi:10.1111/j.1526-100X.2007.00234.x

MillerSJ,SkilleterGA(2006)Temporalvariationinhabitatusebynektoninasubtropicalestuarine

systemJExpMarBiolEcol337:82-95doi:10.1016/j.jembe.2006.06.010

MoilanenA,LeathwickJR,QuinnJM(2011)Spatialprioritizationofconservationmanagement

ConservationLetters4:383-393doi:10.1111/j.1755-263X.2011.00190.x

MoilanenA,WilsonKA,PossinghamHP(2009)Spatialconservationprioritization.OxfordUniversity

Press,Oxford

MontagnaPA,EstevezED,PalmerTA,FlanneryMS(2008)Meta-analysisoftherelationshipbetween

salinityandmolluscsintidalriverestuariesofsouthwestFlorida,USAAmerican

MalacologicalBulletin24:101-115

MontalvoAM,WilliamsSL,RiceKJ,BuchmannSL,CoryC,HandelSN,NabhanGP,PrimackR,

RobichauxRH(1997)Restorationbiology:ApopulationbiologyperspectiveRestoration

Ecology5:277-290doi:10.1046/j.1526-100X.1997.00542.x

MontoyaD,RogersL,MemmottJ(2012)Emergingperspectivesintherestorationofbiodiversity-

basedecosystemservicesTrendsEcolEvol27:666-672

Moreno-MateosD,PowerME,CominFA,YocktengR(2012)Structuralandfunctionallossin

restoredwetlandecosystemsPLoSBiol10:e1001247doi:10.1371/journal.pbio.1001247

MullaneyMD(1991)Trophicontogeny,age,growth,andontogenyofthefeedingapparatusin

Mycteropercamicrolepis(Pisces:Serranldae).CollegeofCharleston

MumbyPJ(2006)Connectivityofreeffishbetweenmangrovesandcoralreefs:Algorithmsforthe

designofmarinereservesatseascapescalesBiologicalConservation128:215-222

doi:10.1016/j.biocon.2005.09.042

MumbyPJ,EdwardsAJ,Arias-GonzalezJE,LindemanKC,BlackwellPG,GallA,GorczynskaMI,

HarborneAR,PescodCL,RenkenH,WabnitzCCC,LlewellynG(2004)Mangrovesenhance

thebiomassofcoralreeffishcommunitiesintheCaribbeanNature427:533-536

doi:10.1038/nature02286

178

NagelkerkenI(2009)Evaluationofnurseryfunctionofmangrovesandseagrassbedsfortropical

decapodsandreeffishes:patternsandunderlyingmechanisms.In:NagelkerkenI(ed)

Ecologicalconnectivityamongtropicalcoastalecosystems.Springer,Heidelberg,pp

357−399

NagelkerkenI,SheavesM,BakerR,ConnollyRM(2015)Theseascapenursery:anovelspatial

approachtoidentifyandmanagenurseriesforcoastalmarinefaunaFishandFisheries

16:362-371doi:10.1111/faf.12057

NevinsJA,PollackJB,StunzGW(2014)Characterizingthenektonuseofthelargestunfishedoyster

reefintheUnitedStatescomparedwithadjacentestuarinehabitatsJournalofShellfish

Research33:227-238doi:10.2983/035.033.0122

NewellRIE,KochEW(2004)Modelingseagrassdensityanddistributioninresponsetochangesin

turbiditystemmingfrombivalvefiltrationandseagrasssedimentstabilizationEstuaries

27:793-806

OrtodossiN,OldsAD,GilbyBL,ConnollyRM,SchlacherTA(2018)Seascapeconnectivityshapes

conservationperformanceonexposedcoastlinesConservBiol:Inpress

OldsAD,AlbertS,MaxwellPS,PittKA,ConnollyRM(2013)Mangrove-reefconnectivitypromotes

theeffectivenessofmarinereservesacrossthewesternPacificGlobEcolBiogeogr22:1040-

1049doi:10.1111/geb.12072

OldsAD,ConnollyRM,PittKA,MaxwellPS(2012a)Habitatconnectivityimprovesreserve

performanceConservationLetters5:56-63doi:10.1111/j.1755-263X.2011.00204.x

OldsAD,ConnollyRM,PittKA,MaxwellPS(2012b)Primacyofseascapeconnectivityeffectsin

structuringcoralreeffishassemblagesMarEcol-ProgSer462:191-203

doi:10.3354/meps09849

OldsAD,ConnollyRM,PittKA,PittmanSJ,MaxwellPS,HuijbersCM,MooreBR,AlbertS,RissikD,

BabcockRC,SchlacherTA(2016)Quantifyingtheconservationvalueofseascape

connectivity:aglobalsynthesisGlobalEcologyandBiogeography25:3-15

doi:10.1111/geb.12388

OldsAD,FrohloffBA,GilbyBL,ConnollyRM,YabsleyNA,MaxwellPS,SchlacherTA(2018a)

UrbanisationsupplementsecosystemfunctioningindisturbedestuariesEcography41:2104-

2113

OldsAD,GilbyBL,ConnollyRM,TibbettsIR,HendersonCJ,StevensTF,ThackwraySK,SchlacherTA

(2018b)FishesofMoretonBay:ecology,humanimpactsandconservation.In:TibbettsIR,

RothlisbergPC,NeilDT,HomburgTA,BrewerDT,ArthingtonAH(eds)MoretonBay

179

QuandamookaandCatchment:past,present&future.UniversityofQueenslandPress,

Brisbane,

OldsAD,NagelkerkenI,HuijbersCM,GilbyBL,PittmanSJ,SchlacherTA(2018c)Connectivityin

CoastalSeascapes.In:PittmanSJ(ed)SeascapeEcology.JohnWileyandSons,Oxford.

OldsAD,PittKA,MaxwellPS,BabcockRC,RissikD,ConnollyRM(2014)Marinereserveshelpcoastal

systemscopewithextremeweatherGlobalChangeBiology20:3050-3058

OldsAD,PittKA,MaxwellPS,ConnollyRM(2012c)Synergisticeffectsofreservesandconnectivityon

ecologicalresilienceJApplEcol49:1195-1203

OldsAD,Vargas-FonsecaE,ConnollyRM,GilbyBL,HuijbersCM,HyndesG,LaymanCA,WhitfieldAK,

SchlacherTA(2017)Theecologyoffishinthesurfzonesofoceanbeaches:aglobalreview

Fishandfisheries

PaiceRL,ChambersJM,RobsonBJ(2016)Outcomesofsubmergedmacrophyterestorationina

shallowimpounded,eutrophicriverHydrobiologia778:179-192

PalmerMA,FalkDA,ZedlerJB(2016)FoundationsofRestorationEcology.SecondEdition.Island

Press,Washington,DC

PalmerMA,HondulaKL,KochBJ(2014)Ecologicalrestorationofstreamsandrivers:shifting

strategiesandshiftinggoals.In:FutuymaDJ(ed)AnnualReviewofEcology,Evolution,and

Systematics,Vol45,vol45.AnnualReviewofEcologyEvolutionandSystematics.pp247-+.

doi:10.1146/annurev-ecolsys-120213-091935

PaulyD,AlderJ,BennettE,ChristensenV,TyedmersP,WatsonR(2003)Thefutureforfisheries

Science302:1359-1361doi:10.1126/science.1088667

PaulyD,WatsonR,AlderJ(2005)Globaltrendsinworldfisheries:impactsonmarineecosystems

andfoodsecurityPhilosophicalTransactionsoftheRoyalSocietyB-BiologicalSciences

360:5-12doi:10.1098/rstb.2004.1574

PiersonKJ,EgglestonDB(2014)Responseofestuarinefishtolarge-scaleoysterreefrestoration

TransAmFishSoc143:273-288doi:10.1080/00028487.2013.847863

PengJ,HuXX,QiuSJ,HuYN,MeersmansJ,LiuYX(2019)Multifunctionallandscapesidentification

andassociateddevelopmentzoninginmountainousareaSciTotalEnviron660:765-775

doi:10.1016/j.scitotenv.2019.01.023

PetersonBJ,HeckJKL(1999)Thepotentialforsuspensionfeedingbivalvestoincreaseseagrass

productivityJournalofExperimentalMarineBiologyandEcology240:37-52

PetersonCH,GrabowskiJH,PowersSP(2003)Estimatedenhancementoffishproductionresulting

fromrestoringoysterreefhabitat:quantitativevaluationMarineEcologyProgressSeries

264:249-264doi:10.3354/meps264249

180

PetersonCH,LipciusRN(2003)Conceptualprogresstowardspredictingquantitativeecosystem

benefitsofecologicalrestorationsMarEcolProgSer264:297-307

PetrankaJW,MurraySS,KennedyCA(2003)Responsesofamphibianstorestorationofasouthern

Appalachianwetland:Perturbationsconfoundpost-restorationassessmentWetlands

23:278-290doi:10.1672/7-20

PiazzaBP,BanksPD,LaPeyreMK(2005)Thepotentialforcreatedoystershellreefsasasustainable

shorelineprotectionstrategyinLouisianaRestorationEcology13:499-506

doi:10.1111/j.1526-100X.2005.00062.x

PickeringH,WhitmarshD(1997)Artificialreefsandfisheriesexploitation:areviewofthe‘attraction

versusproduction’debate,theinfluenceofdesignanditssignificanceforpolicyFisheries

Research31:39-59

PiersonKJ,EgglestonDB(2014)Responseofestuarinefishtolarge-scaleoysterreefrestoration

TransAmFishSoc143:273-288doi:10.1080/00028487.2013.847863

PittmanSJ(2018)SeascapeEcology.WileyBlackwell,Oxford

PittmanSJ,CaldowC,HileSD,MonacoME(2007)Usingseascapetypestoexplainthespatial

patternsoffishinthemangrovesofSWPuertoRicoMarineEcologyProgressSeries

348:273-284doi:10.3354/meps07052

PittmanSJ,KneibRT,SimenstadCA(2011)PracticingcoastalseascapeecologyMarineEcology

ProgressSeries427:187-190doi:10.3354/meps09139

PittmanSJ,McAlpineCA,PittmanKM(2004)Linkingfishandprawnstotheirenvironment:a

hierarchicallandscapeapproachMarineEcologyProgressSeries283:233-254

PittmanSJ,OldsAD(2015)Seascapeecologyoffishesoncoralreefs.In:MoraC(ed)Ecologyof

fishesoncoralreefs.

PittmanSJ,WiensJA,WuJ,UrbanDL(2018)Landscapeecologists'perspectivesonseascape

ecology.In:PittmanSJ(ed)SeascapeEcology.JohnWileyandSons,Oxford.

PichlerHA,GrayCA,BroadhurstMK,SpachHL,NagelkerkenI(2017)Seasonalandenvironmental

influencesonrecruitmentpatternsandhabitatusageamongresidentandtransientfishesin

aWorldHeritageSitesubtropicalestuaryJFishBiol90:396-416doi:10.1111/jfb.13191

PiehlerMF,SmythAR(2011)Impactsofecosystemengineersonestuarinenitrogencycling

Ecosphere2:art12

PlanesS,JonesGP,ThorroldSR(2009)Larvaldispersalconnectsfishpopulationsinanetworkof

marineprotectedareasProceedingsoftheNationalAcademyofScience106:5693-5697

181

PollockB(1982)Movementsandmigrationsofyellowfinbream,Acanthopagrusaustralis(Günther),

inMoretonBay,QueenslandasdeterminedbytagrecoveriesJournalofFishBiology20:245-

252

PossinghamHP,BodeM,KleinCJ(2015)Optimalconservationoutcomesrequirebothrestoration

andprotectionPlosBiology13doi:10.1371/journal.pbio.1002052

PotterIC,TweedleyJR,ElliottM,WhitfieldAK(2015)Thewaysinwhichfishuseestuaries:a

refinementandexpansionoftheguildapproachFishandFisheries16:230-239

doi:10.1111/faf.12050

PottierJ,BédécarratsA,MarrsRH(2009)Analysingthespatialheterogeneityofemergentgroupsto

assessecologicalrestorationJournalofappliedecology46:1248-1257doi:10.1111/j.1365-

2664.2009.01714.x

PowersSP,PetersonCH,GrabowskiJH,LenihanHS(2009)Successofconstructedoysterreefsinno-

harvestsanctuaries:implicationsforrestorationMarineEcologyProgressSeries389:159-170

doi:10.3354/meps08164

PresseyRL,BottrillMC(2008)Opportunism,threats,andtheevolutionofsystematicconservation

planningConservationBiology22:1340-1345doi:10.1111/j.1523-1739.2008.01032.x

PresseyRL,CabezaM,WattsME,CowlingRM,WilsonKA(2007)Conservationplanninginachanging

worldTrendsinEcology&Evolution22:583-592doi:10.1016/j.tree.2007.10.001

PresseyRL,HagerTC,RyanKM,SchwarzJ,WallS,FerrierS,CreaserPM(2000)Usingabioticdatafor

conservationassessmentsoverextensiveregions:quantitativemethodsappliedacrossNew

SouthWales,AustraliaBiologicalConservation96:55–82

PuckettBJ,EgglestonDB(2016)Metapopulationdynamicsguidemarinereservedesign:importance

ofconnectivity,demographics,andstockenhancementEcosphere7doi:10.1002/ecs2.1322

PullingerMG,JohnsonCJ(2010)Maintainingorrestoringconnectivityofmodifiedlandscapes:

evaluatingtheleast-costpathmodelwithmultiplesourcesofecologicalinformation

LandscapeEcology25:1547-1560doi:10.1007/s10980-010-9526-6

QGISDevelopmentTeam(2019)QGISGeographicInformationSystem.OpenSourceGeospatial

Foundation.http://qgis.osgeo.org.

QuanW-m,ZhuJ-x,NiY,ShiL-y,ChenY-q(2009)Faunalutilizationofconstructedintertidaloyster

(Crassostrearivularis)reefintheYangtzeRiverestuary,ChinaEcologicalEngineering

35:1466-1475doi:10.1016/j.ecoleng.2009.06.001

QueenslandGovernment(2015a)Queenslandlandusemappingprogram.QueenslandGovernment,

Brisbane,Australia

182

QueenslandGovernment(2015b)Regionalecosystemmapping-mangroves.Queensland

Government,Brisbane,Australia

QueenslandGovernment(2016)Regionalecosystemmapping.QueenslandGovernment,Brisbane,

Australia

QuesnellePE,FahrigL,LindsayKE(2013)Effectsofhabitatloss,habitatconfigurationandmatrix

compositionondecliningwetlandspeciesBiologicalConservation160:200-208

RCoreTeam(2017)R:ALanguageandEnvironmentforStatisticalComputingRFoundationfor

StatisticalComputing,https://wwwR-projectorg/

RCoreTeam(2018)R:Alanguageandenvironmentforstatisticalcomputingvolhttp://www.R-

project.org/.Vienna,Austria

RCoreTeam(2019)R:Alanguageandenvironmentforstatisticalcomputingvolhttp://www.R-

project.org/.Vienna,Austria

RaposaK(2002)EarlyresponsesoffishesandcrustaceanstorestorationofatidallyrestrictedNew

EnglandsaltmarshRestorEcol10:665-676

ReedDH(2004)ExtinctionriskinfragmentedhabitatsAnimalConservation7:181-191

doi:10.1017/s1367943004001313

ReyBenayasJM,NewtonAC,DiazA,BullockJM(2009)Enhancementofbiodiversityandecosystem

servicesbyecologicalrestoration:ameta-analysisScience325:1121-1124

doi:10.1126/science.1172460

RibeiroMC,MetzgerJP,MartensenAC,PonzoniFJ,HirotaMM(2009)TheBrazilianAtlanticForest:

Howmuchisleft,andhowistheremainingforestdistributed?Implicationsforconservation

BiologicalConservation142:1141-1153doi:10.1016/j.biocon.2009.02.021

RippleWJ,BeschtaRL(2007)RestoringYellowstone'saspenwithwolvesBiolConserv138:514-519

doi:10.1016/j.biocon.2007.05.006

RippleWJ,BeschtaRL(2012)TrophiccascadesinYellowstone:thefirst15yearsafterwolf

reintroductionBiologicalConservation145:205-213

RisserPG(1995)BiodiversityandecosystemfunctionConservationBiology9:742-746

RitchieEG,JohnsonCN(2009)Predatorinteractions,mesopredatorreleaseandbiodiversity

conservationEcologyletters12:982-998

RobertsCM,HawkinsJP,FletcherJ,HandsS,RaabK,WardS(2010)Guidanceonthesizeandspacing

ofMarineProtectedAreasinEngland.UniversityofYork,York

RobertsDW(2013)labdsv:Ordinationandmultivariateanalysisforecology.http://CRAN.R-

project.org/package=mgcv.

183

RodemannJR,BrandlSJ(2017)Consumptionpressureincoastalmarineenvironmentsdecreases

withlatitudeandinartificialvs.naturalhabitatsMarEcolProgSer574:167-179

doi:10.3354/meps12170

RodneyWS,PaynterKT(2006)Comparisonsofmacrofaunalassemblagesonrestoredandnon-

restoredoysterreefsinmesohalineregionsofChesapeakeBayinMarylandJournalof

ExperimentalMarineBiologyandEcology335:39-51doi:10.1016/j.jembe.2006.02.017

RodriguezAB,FodrieFJ,RidgeJT,LindquistNL,TheuerkaufEJ,ColemanSE,GrabowskiJH,Brodeur

MC,GittmanRK,KellerDA,KenworthyMD(2014)Oysterreefscanoutpacesea-levelrise

NatureClimateChange4:493-497doi:10.1038/nclimate2216

RondininiC,ChiozzaF(2010)Quantitativemethodsfordefiningpercentageareatargetsforhabitat

typesinconservationplanningBiolConserv143:1646-1653

doi:10.1016/j.biocon.2010.03.037

RoniP,BeechieTJ,BilbyRE,LeonettiFE,PollockMM,PessGR(2002)AReviewofStream

RestorationTechniquesandaHierarchicalStrategyforPrioritizingRestorationinPacific

NorthwestWatershedsNAmJFishManage22:1-20doi:10.1577/1548-

8675(2002)022<0001:AROSRT>2.0.CO;2

RothschildBJ,AultJS,GoulletquerP,HeralM(1994)DeclineoftheChesapeakeBayoyster

population:AcenturyofhabitatdestructionandoverfishingMarineEcologyProgressSeries

111:29-39

RudershausenPJ,MerrellJH,BuckelJA(2018)Fragmentationofhabitataffectscommunitiesand

movementofnektoninsaltmarshtidalcreeksMarEcolProgSer586:57-72

doi:10.3354/meps12385

RudnickD,RyanSJ,BeierP,CushmanSA,DieffenbachF,EppsC,GerberLR,HartterJ,JennessJS,

KintschJ(2012)Theroleoflandscapeconnectivityinplanningandimplementing

conservationandrestorationprioritiesIssuesinEcology

SannigrahiS,BhattS,RahmatS,PaulSK,SenS(2018)Estimatingglobalecosystemservicevaluesand

itsresponsetolandsurfacedynamicsduring1995–2015JEnvironManage223:115-131

SarkarS,PresseyRL,FaithDP,MargulesCR,FullerT,StomsDM,MoffettA,WilsonKA,WilliamsKJ,

WilliamsPH,AndelmanS(2006)Biodiversityconservationplanningtools:Presentstatusand

challengesforthefutureAnnualReviewofEnvironmentandResources31:123-159

doi:10.1146/annurev.energy.31.042606.085844

SchultzCB,CroneEE(2005)Patchsizeandconnectivitythresholdsforbutterflyhabitatrestoration

ConservBiol19:887-896doi:10.1111/j.1523-1739.2005.00462.x

184

SchulzJJ,SchroderB(2017)IdentifyingsuitablemultifunctionalrestorationareasforForest

LandscapeRestorationinCentralChileEcosphere8:e01644

ScottTA,WehtjeW,WehtjeM(2001)Theneedforstrategicplanninginpassiverestorationof

wildlifepopulationsRestorationEcology9:262-271doi:10.1046/j.1526-

100x.2001.009003262.x

SER(2008)Ecologicalrestorationasatoolforreversingecosystemfragmentation.Societyfor

EcologicalRestorationInternational,Tucson,Arizona

SheavesM,BakerR,NagelkerkenI,ConnollyRM(2015)Truevalueofestuarineandcoastalnurseries

forfish:incorporatingcomplexityanddynamicsEstuarCoasts38:401-414

doi:10.1007/s12237-014-9846-x

SheavesM,JohnstonR,BakerR(2016)Useofmangrovesbyfish:newinsightsfromin-forestvideosMarEcolProgSer549:167-182doi:10.3354/meps11690

SheavesMJ,MolonyBW,TobinAJ(1999)Spawningmigrationsandlocalmovementsofatropical

sparidfishMarineBiology133:123-128doi:10.1007/s002270050450

ShermanRL,GillianDS,SpielerRE(2002)Artificialreefdesign:voidspace,complexity,and

attractantsIcesJournalofMarineScience59:S196-S200doi:10.1006/jmsc.2001.1163

ShinY,RochetM,JenningsS,FieldJ,GislasonH(2005)Usingsize-basedindicatorstoevaluatethe

ecosystemeffectsoffishingICESJournalofMarineScience62:384-396

doi:10.1016/j.icesjms.2005.01.004

ShulseCD,SemlitschRD,TrauthKM,GardnerJE(2012)Testingwetlandfeaturestoincrease

amphibianreproductivesuccessandspeciesrichnessformitigationandrestoration

EcologicalApplications22:1675-1688

SimenstadC,ReedD,FordM(2006)Whenisrestorationnot?EcolEng26:27-39

doi:10.1016/j.ecoleng.2005.09.007

SkilleterGA,LoneraganNR,OldsA,ZharikovY,CameronB(2017)Connectivitybetweenseagrassand

mangrovesinfluencesnektonassemblagesusingnearshorehabitatsMarEcolProgSer

573:25-43doi:10.3354/meps12159

SmythAR,PiehlerMF,GrabowskiJH(2015)Habitatcontextinfluencesnitrogenremovalbyrestored

oysterreefsJournalofAppliedEcology52:716-725doi:10.1111/1365-2664.12435

SolanM,WhiteleyN(2016)Stressorsinthemarineenvironment:physiologicalandecological

responses:societalimplications.OxfordUniversityPress,Oxford

SoniatTM,FinelliCM,RuizJT(2004)Verticalstructureandpredatorrefugemediateoysterreef

developmentandcommunitydynamicsJExpMarBiolEcol310:163-182

doi:10.1016/j.jembe.2004.04.007

185

SoonsMB,MesselinkJH,JongejansE,HeilGW(2005)Habitatfragmentationreducesgrassland

connectivityforbothshort-distanceandlong-distancewind-dispersedforbsJournalof

Ecology93:1214-1225

SouleME(1985)Whatisconservationbiology?Bioscience35:727-734doi:10.2307/1310054

StanturfJA,PalikBJ,DumroeseRK(2014)Contemporaryforestrestoration:Areviewemphasizing

functionForEcolManage331:292-323doi:10.1016/j.foreco.2014.07.029

SteppeCN,FredrikssonDW,WallendorfL,NikolovM,MayerR(2016)DirectsettingofCrassostrea

virginicalarvaeinatidaltributary:applicationsforshellfishrestorationandaquaculture

MarineEcologyProgressSeries546:97-112doi:10.3354/meps11604

StevensPW,BlewettDA,CaseyJP(2006)Short-termeffectsofalowdissolvedoxygeneventon

estuarinefishassemblagesfollowingthepassageofHurricaneCharleyEstuariesandCoasts

29:997-1003

StobartB,WarwickR,GonzálezC,MallolS,DíazD,ReñonesO,GoñiR(2009)Long-termandspillover

effectsofamarineprotectedareaonanexploitedfishcommunityMarineEcology-Progress

Series384:47-60doi:10.3354/meps08007

StomsDM,DavisFW,AndelmanSJ,CarrMH,GainesSD,HalpernBS,HoenickeR,LeibowitzSG,

LeydeckerA,MadinEMP,TallisH,WarnerRR(2005)Integratedcoastalreserveplanning:

makingtheland-seaconnectionFrontiersinEcologyandtheEnvironment3:429-436

doi:10.2307/3868659

SunP,LiuXZ,TangYL,ChengWZ,SunRL,WangXX,WanR,HeinoM(2017)Thebio-economic

effectsofartificialreefs:mixedevidencefromShandong,ChinaICESJMarSci74:2239-2248

doi:10.1093/icesjms/fsx058

TambosiLR,MartensenAC,RibeiroMC,MetzgerJP(2014)Aframeworktooptimizebiodiversity

restorationeffortsbasedonhabitatamountandlandscapeconnectivityRestorEcol22:169-

177doi:10.1111/rec.12049

TamburriMN,LuckenbachMW,BreitburgDL,BonniwellSM(2008)SettlementofCrassostrea

ariakensislarvae:effectsofsubstrate,biofilms,sedimentandadultchemicalcuesJournalof

ShellfishResearch27:601608

TaylorMD,ChickRC,LorenzenK,AgnaltAL,LeberKM,BlankenshipHL,VanderHaegenG,Loneragan

NR(2017)FisheriesenhancementandrestorationinachangingworldFisheriesResearch

186:407-412

TetzlaffJC,PineWE,III,AllenMS,AhrensRNM(2013)Effectivenessofsizelimitsandbaglimitsfor

managingrecreationalfisheries:acasestudyoftheGulfofMexicorecreationalgagfishery

BulletinofMarineScience89:483-502doi:10.5343/bms.2012.1025

186

TewksburyJJ,LeveyDJ,HaddadNM,SargentS,OrrockJL,WeldonA,DanielsonBJ,BrinkerhoffJ,

DamschenEI,TownsendP(2002)Corridorsaffectplants,animals,andtheirinteractionsin

fragmentedlandscapesProceedingsofthenationalacademyofsciences99:12923-12926

TheNatureConservancy&EcologicalServiceProfessionals(2015)RestorationofNoosaEstuary:An

AssessmentofOysterRecruitment.

ThompsonBA(2011)PlanningforImplementation:Landscape-LevelRestorationPlanninginan

AgriculturalSettingRestorationEcology19:5-13doi:10.1111/j.1526-100X.2010.00666.x

ThomsonJR,MoilanenAJ,VeskPA,BennettAF,MacNallyR(2009)Whereandwhentorevegetate:

aquantitativemethodforschedulinglandscapereconstructionEcologicalApplications

19:817-828doi:10.1890/08-0915.1

ThurstanRH(2016)HistoricalecologyoftheNoosaestuaryfisheries.UniversityofQueensland,

Brisbane,Australia

ThurstanRH,GameE,PandolfiJM(2017)Popularmediarecordsrevealmulti-decadaltrendsin

recreationalfishingcatchratesPLoSONE12:e0182345doi:10.1371/journal.pone.0182345

TNC,ESP(2015)RestorationofNoosaEstuary:AnAssessmentofOysterRecruitment.ReportbyThe

NatureConservancyandEcologicalServiceProfessionals.TheNatureConservancyand

EcologicalServiceProfessionals,Brisbane,Australia

TobonW,Urquiza-HaasT,KoleffP,SchroterM,Ortega-AlvarezR,CampoJ,Lindig-CisnerosR,

SarukhanJ,BonnA(2017)RestorationplanningtoguideAichitargetsinamegadiverse

countryConservBiol31:1086-1097doi:10.1111/cobi.12918

TolleySG,VoletyAK(2005)Theroleofoystersinhabitatuseofoysterreefsbyresidentfishesand

decapodcrustaceansJournalofShellfishResearch24:1007-1012

TorioDD,ChmuraGL(2015)ImpactsofSeaLevelRiseonMarshasFishHabitatEstuarCoasts

38:1288-1303doi:10.1007/s12237-013-9740-y

TurnerRK,SchaafsmaM(2015)Coastalzonesecosystemservices:fromsciencetovaluesand

decisionmaking.Switzerland.

TurnerSJ,ThrushS,HewittJ,CummingsV,FunnellG(1999)Fishingimpactsandthedegradationor

lossofhabitatstructureFisheriesManagementandEcology6:401-420

TylianakisJM,LalibertéE,NielsenA,BascompteJ(2010)Conservationofspeciesinteraction

networksBiologicalconservation143:2270-2279

UnderwoodAJ(1991)BeyondBACI-Experimentaldesignsfordetectingenvironmentalimpactson

temporalvariationsinnaturalpopulationsAustJMarFreshwaterRes42:569-587

UnderwoodAJ(1994)OnbeyondBACI-Samplingdesignsthatmightreliablydetectenvironmental

disturbancesEcologicalApplications4:3-15doi:10.2307/1942110

187

UnsworthRKF,DeLeónPS,GarrardSL,JompaJ,SmithDJ,BellJJ(2008)HighconnectivityofIndo-

PacificseagrassfishassemblageswithmangroveandcoralreefhabitatsMarineEcology

ProgressSeries353:213-224doi:10.3354/meps07199

ValesiniFJ,CottinghamA,HallettCS,ClarkeKR(2017)Interdecadalchangesinthecommunity,

populationandindividuallevelsofthefishfaunaofanextensivelymodifiedestuaryJFish

Biol90:1734-1767doi:10.1111/jfb.13263

Valiente-BanuetA,AizenMA,AlcántaraJM,ArroyoJ,CocucciA,GalettiM,GarcíaMB,GarcíaD,

GómezJM,JordanoP(2015)Beyondspeciesloss:theextinctionofecologicalinteractionsin

achangingworldFunctionalEcology29:299-307

VenturelliPA,HyderK,SkovC(2017)Anglerappsasasourceofrecreationalfisheriesdata:

opportunities,challengesandproposedstandardsFishandFisheries18:578-595

doi:10.1111/faf.12189

VitousekPM,MooneyHA,LubchencoJ,MelilloJM(1997)HumandominationofEarth'secosystems

Science277:494-499

VolkXK,GattringerJP,OtteA,Harvolk-SchoningS(2018)Connectivityanalysisasatoolforassessing

restorationsuccessLandscapeEcology33:371-387doi:10.1007/s10980-018-0611-6

WallCC,PetersonBJ,GoblerCJ(2008)FacilitationofseagrassZosteramarinaproductivityby

suspension-feedingbivalvesMarineEcologyProgressSeries357:165-174

doi:10.3354/meps07289

WallesB,TroostK,vandenEndeD,NieuwhofS,SmaalAC,YsebaertT(2016)Fromartificial

structurestoself-sustainingoysterreefsJournalofSeaResearch108:1-9

doi:10.1016/j.seares.2015.11.007

WalthamNJ,ConnollyRM(2013)Artificialtidallakes:Builtforhumans,homeforfishEcolEng

60:414-420doi:10.1016/j.ecoleng.2013.09.035

WardJV,TocknerK,SchiemerF(1999)Biodiversityoffloodplainriverecosystems:Ecotonesand

connectivityRegulatedRivers-Research&Management15:125-139doi:10.1002/(sici)1099-

1646(199901/06)15:1/3<125::aid-rrr523>3.0.co;2-e

WattsME,BallIR,StewartRS,KleinCJ,WilsonK,SteinbackC,LourivalR,KircherL,PossinghamHP

(2009)MarxanwithZones:Softwareforoptimalconservationbasedland-andsea-use

zoningEnvironmentalModelling&Software24:1513-1521

doi:10.1016/j.envsoft.2009.06.005

WaycottM,DuarteCM,CarruthersTJ,OrthRJ,DennisonWC,OlyarnikS,CalladineA,Fourqurean

JW,HeckKL,HughesAR(2009)Acceleratinglossofseagrassesacrosstheglobethreatens

coastalecosystemsProceedingsoftheNationalAcademyofSciences106:12377-12381

188

WebleyJAC(2008)Theecologyofthemudcrab(Scyllaserrata):theircolonisationofestuariesand

roleasscavengersinecosystemprocesses.GriffithUniversity

WebleyJ,McInnesK,TeixeiraD,LawsonA,QuinnR(2015)StatewideRecreationalFishingSurvey

2013-14.QueenslandGovernment,Brisbane,Australia

WeeksR(2017)IncorporatingseascapeconnectivityinconservationprioritisationPLoSOne

12:e0182396doi:10.1371/journal.pone.0182396

WhitfieldAK(2017)Theroleofseagrassmeadows,mangroveforests,saltmarshesandreedbedsas

nurseryareasandfoodsourcesforfishesinestuariesReviewsinFishBiologyandFisheries

27:75-110doi:10.1007/s11160-016-9454-x

WiensJA(2002)Riverinelandscapes:takinglandscapeecologyintothewaterFreshwaterBiology

47:501–515

WiensJA,HobbsRJ(2015)IntegratingconservationandrestorationinachangingworldBioscience

65:302-312doi:10.1093/biosci/biu235

WilsonEE,WolkovichEM(2011)Scavenging:howcarnivoresandcarrionstructurecommunities

TrendsEcolEvol26:129–135

WilsonJC,ElliottM(2009)ThepotentialforhabitatcreationproducedbyoffshorewindfarmsWind

Energy12:203-212

WilsonKA,LulowM,BurgerJ,FangYC,AndersenC,OlsonD,O'ConnellM,McBrideMF(2011)

Optimalrestoration:accountingforspace,timeanduncertaintyJApplEcol48:715-725

doi:10.1111/j.1365-2664.2011.01975.x

WilsonKA,McBrideMF,BodeM,PossinghamHP(2006)Prioritizingglobalconservationefforts

Nature440:337-340doi:10.1038/nature04366

WolanskiE,ElliottM(2015)EstuarineEcohydrology:anintroduction.Elsevier,Amsterdam

WoodS(2017)mgcv:MixedGAMComputationVehiclewithGCV/AIC/REMLsmoothnessestimation.

Rpackageversion1.8-24.http://CRAN.R-project.org/package=mgcv.

WoodS(2018)MixedGAMComputationVehiclewithAutomaticSmoothnessEstimationRpackage

version18-24,https://cranr-projectorg/web/packages/mgcv/mgcvpdf

WoottonRJ(1992)FishEcology.Springer,Amsterdam

WormB,DuffyJE(2003)Biodiversity,productivityandstabilityinrealfoodwebsTrendsEcolEvol

18:628-632

WortleyL,HeroJM,HowesM(2013)Evaluatingecologicalrestorationsuccess:areviewofthe

literatureRestorEcol21:537-543doi:10.1111/rec.12028

YeagerLA,LaymanCA,AllgeierJE(2011)Effectsofhabitatheterogeneityatmultiplespatialscales

onfishcommunityassemblyOecologia167:157–168

189

ZhangY,CioffiW,CopeR,DaleoP,HeywoodE,HoytC,SmithC,SillimanB(2018)Aglobalsynthesis

revealsgapsincoastalhabitatrestorationresearchSustainability10:1040

ZieglerSL,GrabowskiJH,BaillieCJ,FodrieFJ(2018)Effectsoflandscapesettingonoysterreef

structureandfunctionlargelypersistmorethanadecadepost-restorationRestorEcol

26:933-942doi:10.1111/rec.12651

zuErmgassenPSE,GrabowskiJH,GairJR,PowersSP(2016)Quantifyingfishandmobileinvertebrate

productionfromathreatenednurseryhabitatJournalofAppliedEcology53:596-606

doi:10.1111/1365-2664.12576

190

Appendix A- Monitoring of the Noosa River Oyster Reefs; November 2017 - November 2018

Report to Noosa Council, and Queensland Department of Agriculture and Fisheries

Dr Ben Gilby Prof Thomas Schlacher,

Dr Andrew Olds, Mr Nicholas Ortodossi,

Mr Ashley Rummell, Dr Christopher Henderson,

Mr Thomas Brook, Miss Cassandra Duncan,

Ms Felicity Hardcastle.

December 2018

191

Table of Contents Executive summary 3 Introduction and context 4 Noosa River oyster reef restoration monitoring schedule 9 Monitoring criteria 1- Oyster restoration unit location stability 12 Monitoring criteria 2- Natural recruitment processes 21 Monitoring criteria 3- Community use and enjoyment of the declared fish habitat area 25 Monitoring criteria 4- Other potential effects 28 Conclusions 29 References 29 Appendices 30

192

Executive summary

1) Background & Rationale The principle objective of the Noosa Oyster Reef Restoration Project is to restore ecologically-functioning oyster reefs to the Noosa River Estuary. Reef units were installed in November 2017 to begin the restoration. An annual monitoring is part of the approval (OPW17/0016) that needs to report on: a) reef unit stability; b) oyster spat recruitment processes; c) lack of negative community usage; and d) impact on marine plants and shoreline erosion. This report addresses these requirements, reporting on the findings of two mandated monitoring events (May 2018, Nov. 2018) done by USC. The methods used were precisely those prescribed in the approved monitoring schedule, with no modifications.

2) Stability and Position

The reef units can be considered stable. The mean recorded change in position over 12 months was 0.16 m (+/- 0.12 m SD). All oyster reef restoration units moved less than 0.65 m over one year. No reef unit moved from the designated resource allocation areas at any site. No reef unit was physically displaced from its anchor points in the sediment or removed otherwise.

3) Recruitment

Substantial spat fall and oyster growth was found in both monitoring events. In May 2018, an average spat fall of 387.5 oysters/m2 (+/- 537.7 oysters/m2 SD) was recorded, and in November 2018 it was 306.1 oysters/m2 (+/- 575.1 oysters/m2 SD). There was a significant increase in the size of newly recruited oysters on the reefs. Here oysters increased in size from 14.3 mm (+/- 6.5 mm SD) in May 2018, to 19.1 mm (+/- 8.9 mm SD) in November 2018. The favourable spat fall and oyster growth are an encouraging bellwether for the reefs to develop into biologically stable systems over the coming years.

4) Interaction with Public There were signs of damage to a few reef units most likely caused by boat propellers and anchors, evident during the November 2018 monitoring event. We repaired these units where possible. Limited data suggest that the use of the oyster reefs by recreational fishers possibly appears to have increased over time. We suggest that closer monitoring of this aspect may be useful in coming years.

5) Marine Plants & Shoreline Erosion

Whilst there was some change to the edges of seagrass beds, there were no significant changes to the distribution, cover, or species composition of existing seagrass beds or mangroves forests within 50 m of each reefs. We found no shoreline erosion in the vicinity of the reef units. At two sites in Lake Weyba and Weyba Creek dugong grass Halophila ovalis was found near reef sites where it was not recorded at the time of reef placement.

6) Conclusions

Overall, the results of the 2018 monitoring of the Noosa oyster reefs are encouraging and positive. Two of the targets have been fully met (reef unit stability, no impact on plants and shorelines). The target of oyster and other invertebrate growth is tracking very favourably. The observation of reef unit damage from anchoring needs consideration of closer monitoring and possibly targeted education / information may be considered for this issue. The next scheduled monitoring event is mandated for May 2019.

193

Introduction and context Oyster reef restoration in the Noosa River; The principal objective of the Noosa River Oyster Reef Restoration project is to restore natural oyster reef habitat in the Noosa Estuary. This requires the addition of suitable hard substrate for oyster spat to settle, and allowing natural recruitment and reef-formation processes to recover to areas that historically supported oyster reefs. The oyster settlement substrate comprises oyster reef restoration units. These units are made from the most suitable recruitment material - oyster shells held together by a natural coir fiber bag. The units are raised above the muddy and sandy marine substrates, thereby mimicking the vertical relief of the original oyster reefs to facilitate natural recruitment processes (Figure 1, 2). We expected that natural recruitment processes would cement the dead shell together and form part of the mosaic of habitats within the estuary including both soft and hard structural habitat types, thereby creating a structurally diverse mosaic of habitats that is predicted to be beneficial to a range of fish species, including species of harvested in commercial and recreational fisheries (e.g. yellowfin bream, estuary cod, tailor, mangrove jack, and moses perch). The restoration sites that were chosen as locations for oyster reef restoration in the Noosa River (Figure 3) because they:

5) are in reaches of the lower Noosa River estuary and in Weyba Creek from which oysters were harvested historically;

6) are within the depth range known to be suitable for oyster reefs to grow; 7) are in locations where viable oyster larvae were recorded during recent recruitment

studies; and 8) have an extent where the final restored area will not exceed the historical areal

extent of oyster reefs before commercial harvesting took place.

Figure 1 Coir mesh bag filled with recycled oyster shells; the principle restoration unit

194

Figure 2 Approved coastal engineering drawing of the oyster reef restoration units installed in the Noosa River

195

Figure 3 Map of oyster restoration sites and marine habitats in the Noosa River

196

Oyster reef unit installations The installation of oyster reef units to the Noosa River was done during a period of three days, from 20 to 22 November 2017 (Figure 4). All works and surveys were conducted by a team comprising professionals of Noosa Jetty Builders, and marine scientists from the University of the Sunshine Coast. All oyster reef units and signs were installed exactly according to specifications set out in the permit conditions and any instructions provided regarding procedures (Table 1). All reef units are located to avoid any damage to marine plants or surrounding ecosystem (Table 1, Figure 4). Key observations during the placement of the reef units and associated signage include the following:

• No marine plants were disturbed during the installation of the reef units or signs. Before each reef unit was placed on the seafloor, we carefully examined the site to ascertain that no seagrass, mangrove aerial roots, or any other marine vegetation was present. This ensured that no direct placement impacts occurred at any of the sites (Table 1).

• All operations were localised to the immediate site where reef units were sunk onto the seabed. Thus, no habitat outside the RAA area was impacted in any form by the reef units or the signage.

• Poles holding the signs were installed using a narrow water jet ‘spear’. Use of this gear ensured that only a very narrow (< 20 cm diameter in all cases) area of the seabed was disturbed (Figure 4). Sign posts were then driven further into the substrate, to a mean depth of 2.7 m (min 1.8 m, max 3.5 m; Table 1), using a post driver (Figure 4).

• According to engineering specifications, we used marine grade hardwood stakes (sourced from the Australian Hardwood Stake Company) to secure the reef units on the seafloor. The stakes were hammered into the seabed to a mean depth of 0.9 m (min 0.5 m, max 1.2 m) - depending on hardness of the seabed (Table 1), at a mean angle of 82.4º (min 75 º, max 90 º).

• Pre- and post-installation photographs and videos of all sites are available electronically (OneDrive folder). Full access to the all electronically stored imagery has been provided to Noosa Council. In addition, we are happy to grant access to whoever may require it (please email: bgilby@usc.edu.au).

Figure 4 The resulting very small footprint of sign posts following installation by the water jet (left) and a completed oyster reef in Weyba Creek (right).

197

Table 1. Completed oyster-reef installations in the Noosa River, listing their exact position, the date of placement, characteristics of posts and stakes driven into the seabed to secure signage and reef units, and any presence and/or damage to marine vegetation.

Site Number

Position Installation date

Depth to which sign post is driven into the seabed (m)

Stakes to secure reef units Presence of, or damage to, marine plants and/or adjacent habitats or ecosystems.

Latitude Longitude Depth of stake into seabed (m)

Angle of stake relative to seabed (degrees)

1 reef units not installed - seabed not compact enough 2 -26.38333096 153.045835 20/11/17 2.8 1.2 79 None 3 -26.38673804 153.043291 20/11/17 2.8 1.2 75 None 4 -26.39367197 153.044599 20/11/17 2.9 1.1 84 None 5 -26.394757 153.049933 20/11/17 2.4 0.5 82 None 6 -26.39245399 153.051636 20/11/17 3.2 1.2 85 None 7 -26.39351204 153.054941 20/11/17 3.3 1.2 77 None 8 -26.39491903 153.061679 20/11/17 2.4 1.2 90 None 9 -26.38954598 153.070696 21/11/17 1.8 0.5 86 None 10 -26.38452999 153.071546 21/11/17 2.0 0.6 87 None 11 -26.39817104 153.080595 21/11/17 2.3 0.5 76 None 12 -26.40045603 153.077035 21/11/17 2.9 0.5 79 None 13 -26.41099703 153.071263 22/11/17 3.5 1.2 87 None 14 -26.41184301 153.0712 22/11/17 3.2 0.5 83 None 15 reef units not installed – presence of seagrass 16 -26.43981904 153.062079 22/11/17 2.9 0.6 84 None

198

Noosa River oyster reef restoration monitoring schedule As part of the development approval for installing the oyster reefs, the stakeholders report yearly (in December) to Queensland Department of Agriculture and Fisheries on the progress of the oyster restoration project, especially relating to;

5. Oyster restoration unit stability (i.e. oyster reef restoration units remain within the allocated resource allocation areas);

6. Restoration of natural recruitment processes over long term (i.e. spat recruitment rates demonstrate that the biogenic matrix will be sufficient to hold the structure of the oyster reef restoration units in place following complete degradation of the supporting coir material during the establishment phase and to facilitate ongoing natural recruitment processes);

7. Equitable Community Impacts (i.e. ensuring fair community use of the river system is not impacted by placement of the Oyster Reef Restoration Units within the Resource Allocation Areas - to be monitored using a council operated community feedback system); and

8. No negative impact on marine plants or shoreline erosion (i.e. ensuring the oyster restoration units do not impede natural marine plant growth or accelerate coastal erosion processes).

These performance objectives are quantified using established monitoring protocols for oyster reef restoration, and follow international best practice for monitoring restored oyster habitats (Baggett et al. 2014). The monitoring program is designed to be adaptive, with annual reviews against the performance objectives for the project (See Table 2 for detailed monitoring requirements). Two monitoring events occurred in 2018; one 6 months post installation in May 2018, and the other 12 months post installation in November 2018. The results of these monitoring events are covered in this report. There have also been detailed surveys of the fish and crustacean assemblages associating with the reefs throughout 2018. Although the details of these surveys are not covered in this report, they are available from USC upon request.

199

Table 2 Noosa oyster reef restoration monitoring schedule as agreed to by all stakeholders in RAA 2016CA0575, and including mark-ups from Department of Agriculture and Fisheries dated 6 January 2017.

Performance objective Monitoring method Frequency Corrective action where performance objective is not met

1. Oyster Restoration Unit location stability Oyster restoration units remain within the designated Resource Allocation Areas.

Visually inspect the stability of oyster reef units and record the precise GPS position (± cm scale), and size of each unit (following international best practice: Baggett et al. 2015). Use GIS software to contrast the position, footprint, size and area of oyster reef restoration units between monitoring events and assess any potential movement.

Every 6 months for a minimum of three years Additional monitoring will be conducted within 2 weeks of substantial rainfall events (i.e. events that exceed 50-year Average Rainfall Intervals).

A professional coastal engineer shall be consulted to suggest remedial action for any oyster reef restoration units that collapse, or shift by > 1 m. Any oyster reef restoration units that move outside the designated Resource Allocation Areas within 3 years will be removed. A professional ecologist will confirm the cause of the shift.

2. Natural Recruitment Processes Oysters and other sessile benthic invertebrates recruit to reef restoration units to establish a biogenic matrix, which binds oyster shells in place, prior to degradation of coir material. The Key Performance Indicators for this are: 1.) Oyster recruitment: successful recruitment of oyster spat in at least 1 out of the 3 years (i.e. 33% of the time) post deployment (following international best practice of 40%: Baggett et al. 2015).1; 2.) Cover of oysters and other sessile benthic invertebrates: an upward trend in the cover of sessile benthic invertebrates growing on restoration units; and 3.) Establishment of stable biogenic matrix: structural rigidity of oyster restoration units, denoting a stable biogenic matrix after 3 years post deployment, which is sufficient to hold oyster shells in place.

Quantify the recruitment of oysters and other sessile benthic invertebrates to reef restoration units, and measure changes in the cover of oysters and other sessile benthic invertebrates over time (following international best practice: Baggett et al. 2015).1

The monitoring methods for each Indicators are: 1.) Oyster recruitment: marked oyster shells will be fastened to the outside of restoration units. These shells will be harvested at regular intervals (15 oyster shells per location on each event – i.e. 5 shells per unit) and the density and size of recruits recorded; 2.) Cover of oysters and other sessile benthic invertebrates: photographs of ten quadrats (25 cm x 25cm), distributed in a stratified random design, across each oyster reef restoration unit will be taken at regular intervals to quantify the change in cover of oysters and other sessile benthic invertebrates (Baggett et al. 2015)1. 3.) Establishment of stable biogenic matrix: assess the structural integrity of each oyster reef restoration unit and monitor the degradation of coir material. Structural integrity will be quantified by measuring the proportion of oyster shells (from 10 shells that are selected at random at each location), which can be removed easily by hand

Every 6 months for a minimum of three years, unless otherwise detailed within corrective actions

We will contrast: 1) the rate of oyster recruitment; 2) the cover of oysters and all other sessile benthic invertebrates; and 3) the stability of the biogenic matrix among oyster reef restoration units and between monitoring events. Any restoration units that fail to meet the Key Performance Indicators within 3 years will be removed. If major damage occurs to the coir mesh of any oyster reef restoration units before a stable biogenic matrix has formed, and loose oyster shells are being lost from the structure, it will be repaired in-situ by hand weaving. Any such repairs will be conducted carefully, and by hand, to ensure that there is no damage to established areas of biogenic matrix.

200

manipulation. The rate of coir degradation will be measured by monitoring changes to the condition of coir mesh over time from the ten photo-quadrats, which are collected twice per year to assess changes in the cover of benthic invertebrates at each oyster reef restoration unit.

3. Community use and enjoyment of the declared Fish Habitat Area Oyster restoration units do not significantly impair community use and enjoyment of the declared Fish Habitat Area, particularly fishing activities.

Maintain records of community feedback, evidence of vandalism, and vessel strikes on the trial oyster reef restoration units. Records will be comprehensive and include, as a minimum set: 1) the number and type of complaints received (also, to allow the Department to gauge the success, any positive comments should also be provided); 2) the type, nature, and severity of any acts of willful vandalism; and 3) the type and severity of any vessel strike.

Annual reporting of all complaints, cases of willful vandalism, and instances of vessel strike received for the three-year trial period.

Complaints: within 3 months of receiving each complaint, an investigation (including interview with the complainant if possible) will be conducted to determine whether complaints are directly related to oyster reef restoration units. Potential response actions will be provided to the Department of Agriculture and Fisheries in the annual report. If complaints persist, and grievances exceed the direct benefit of oyster reef restoration, the particular oyster reef restoration unit will be removed as soon as practical, and prior to the 4th anniversary post deployment. Willful vandalism and vessel Strikes If there are consistent and / or significant cases of willful vandalism to, or instances of vessels striking, the oyster reef restoration units, the cause of the impact will be identified and used to guide the delivery of a community education campaign aimed at reducing these types of incidents. If the education campaign does not reduce cases of willful vandalism and / or instances of vessel strike, the oyster reef restoration units that are responsible will be removed prior to the 4th anniversary post deployment.

4. Other potential effects: Oyster reef restoration units do not cause a decline in the extent of marine plants within a 50 m radius of the restoration units, and are not attributed to erosion of adjacent shorelines or other ambient environmental impacts.

Map the area of marine plant habitats (seagrass, mangroves) in the immediate vicinity (i.e. with a 50 m buffer) of each oyster restoration area, using high-resolution GPS (cm scale) and field-validated aerial imagery (sourced from Nearmap). Monitor changes in the composition coverage and condition of seagrass within 50 m radius around the oyster reefs. Map the location and condition of estuarine shorelines that occur in the immediate vicinity (i.e. with a 50 m buffer) of each oyster reef, using high-resolution GPS (cm scale) and field-validated aerial imagery (sourced from Nearmap).

Annually for a minimum of three years, or until any failed restoration units have been removed.

Where monitoring shows that there are substantial, and consistent, losses of marine plants, or erosion of estuarine shorelines, in the immediate vicinity of oyster reef restoration units, and these changes can unambiguously be attributed to oyster reef restoration activities, the problem units that are responsible for such impacts shall be removed (as soon as practical and prior to the following annual reporting anniversary).

201

Monitoring criteria 1- Oyster restoration unit location stability Performance objective Oyster restoration units remain within the designated Resource Allocation Areas. Monitoring method and frequency Visually inspect the stability of oyster reef units and record the precise GPS position (± cm scale), and size of each unit (following international best practice: Baggett et al. 2015). Use GIS software to contrast the position, footprint, size and area of oyster reef restoration units between monitoring events and assess any potential movement. Frequency: Every 6 months for a minimum of three years. Additional monitoring will be conducted within 2 weeks of substantial rainfall events (i.e. events that exceed 50-year Average Rainfall Intervals). Was the performance objective met? No units moved outside of the RAA area, and all repairable reefs were repaired. Therefore, all aspects of this performance objective were met, and no further action is necessary. Supporting Evidence No oyster restoration units moved beyond the RAA areas, and no units moved more than 1 m during either monitoring event (Table 3). Maps of each of the sites (Figures 5 to 18) are provided below to support this lack of movement of bags. High-resolution maps and csv files of the locations of the reefs at all three survey points (following installation, and at 6 and 12 month monitoring events) are available from USC upon request.

202

Table 3 List of reef site sites and reef units, their positions as given by the coastal surveyor, and the distance these units had moved from the initial installations (as recorded by the coastal surveyor) to the 6 and 12 month monitoring events. Distances measured are from the center points of oyster restoration units.

Site Surveyor ID Easting (GDA94)

Northing (GDA94)

Distance moved 6 months (m)

Distance moved 12 months (m)

2 2.1 504557.693 7081870.54 0.01 0.12 2.2 504562.401 7081871.525 0.09 0.04 2.3 504561.797 7081866.981 0.11 0.06

3 3.1 504297.232 7081482.809 0.05 0.17 3.2 504300.191 7081487.149 0.21 0.08 3.3 504302.058 7081482.224 0.30 0.03

4 4.1 504447.682 7080720.227 0.39 0.09 4.2 504451.864 7080722.81 0.33 0.28 4.3 504452.517 7080718.084 0.13 0.17

5 5.1 504977.921 7080596.146 0.20 0.02 5.2 504980.268 7080598.985 0.08 0.11 5.3 504983.03 7080596.309 0.28 0.37

6 6.1 505140.806 7080848.337 0.19 0.09 6.2 505147.768 7080846.296 0.20 0.42 6.3 505142.59 7080842.496 0.20 0.14

7 7.1 505475.222 7080728.716 0.02 0.21 7.2 505481.921 7080724.223 0.06 0.07 7.3 505476.882 7080722.755 0.20 0.20

8 8.1 506148.168 7080578.329 0.44 0.11 8.2 506151.843 7080583.145 0.12 0.16 8.3 506154.632 7080579.701 0.57 0.65

9 9.1 507047.3 7081174.147 0.24 0.22 9.2 507052.487 7081172.727 0.46 0.12 9.3 507048.33 7081169.39 0.03 0.15

10 10.1 507131.275 7081731.424 0.10 0.21 10.2 507136.871 7081731.471 0.18 0.11 10.3 507135.843 7081726.658 0.27 0.10

11 11.1 508035.278 7080213.498 0.03 0.17 11.2 508040.293 7080216.237 0.15 0.14 11.3 508038.955 7080212.182 0.23 0.07

12 12.1 507678.59 7079963.401 0.17 0.07 12.2 507684.058 7079964.39 0.18 0.11 12.3 507683.416 7079959.64 0.06 0.10

13 13.1 507103.328 7078792.463 0.20 0.12 3.2 507107.243 7078796.008 0.16 0.08 13.3 507107.661 7078790.417 0.06 0.15

14 14.1 507090.833 7078704.893 0.18 0.25 14.2 507094.457 7078706.761 0.20 0.39 14.3 507094.824 7078701.819 0.12 0.13

16 16.1 506199.719 7075595.579 0.21 0.07 16.2 506204.963 7075595.325 0.18 0.04 16.3 506200.595 7075590.153 0.16 0.19 Mean 0.18 0.16 StDev 0.12 0.12

203

Figure 5 Map of (A) mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 2 in the Noosa River.

Figure 6 Map of (A) aerial extents of seagrass and mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 3 in the Noosa River.

204

Figure 7 Map of (A) mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 4 in the Noosa River.

Figure 8 Map of (A) aerial extents of seagrass and mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 5 in the Noosa River.

205

Figure 9 Map of (A) aerial extents of seagrass and mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 6 in the Noosa River.

Figure 10 Map of (A) aerial extents of seagrass and mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 7 in the Noosa River.

206

Figure 11 Map of (A) mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 8 in the Noosa River.

207

Figure 12 Map of (A) mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 9 in the Noosa River.

208

Figure 13 Map of (A) mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 10 in the Noosa River.

209

Figure 14 Map of (A) mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 11 in the Noosa River.

210

Figure 15 Map of (A) aerial extents of seagrass and mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 12 in the Noosa River.

Figure 16 Map of (A) aerial extents of seagrass and mangrove edges and (B) oyster reef

211

unit positions (at installation, and at 6 and 12 months post installation) at site 13 in the Noosa River.

212

Figure 17 Map of (A) aerial extents of seagrass and mangrove edges and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 14 in the Noosa River.

213

Figure 18 Map of (A) vegetation monitoring area (note there was no seagrass or mangroves at this site) and (B) oyster reef unit positions (at installation, and at 6 and 12 months post installation) at site 16 in the Noosa River.

214

Monitoring criteria 2- Natural recruitment processes Performance objective: Oysters and other sessile benthic invertebrates recruit to reef restoration units to establish a biogenic matrix, which binds oyster shells in place, prior to degradation of coir material. The Key Performance Indicators for this are:

1. Oyster recruitment: successful recruitment of oyster spat in at least 1 out of the 3 years (i.e. 33% of the time) post deployment (following international best practice of 40%: Baggett et al. 2015);

2. Cover of oysters and other sessile benthic invertebrates: an upward trend in the cover of sessile benthic invertebrates growing on restoration units; and

3. Establishment of stable biogenic matrix: structural rigidity of oyster restoration units, denoting a stable biogenic matrix after 3 years post deployment, which is sufficient to hold oyster shells in place.

Monitoring method: Quantify the recruitment of oysters and other sessile benthic invertebrates to reef restoration units, and measure changes in the cover of oysters and other sessile benthic invertebrates over time (following international best practice)(Baggett et al., 2015). The monitoring methods for each Indicators are:

1. Oyster recruitment: marked oyster shells will be fastened to the outside of restoration units. These shells will be harvested at regular intervals (15 oyster shells per location on each event – i.e. 5 shells per unit) and the density and size of recruits recorded;

2. Cover of oysters and other sessile benthic invertebrates: photographs of ten quadrats (25 cm x 25cm), distributed in a stratified random design, across each oyster reef restoration unit will be taken at regular intervals to quantify the change in cover of oysters and other sessile benthic invertebrates (Baggett et al., 2015).

3. Establishment of stable biogenic matrix: assess the structural integrity of each oyster reef restoration unit and monitor the degradation of coir material. Structural integrity will be quantified by measuring the proportion of oyster shells (from 10 shells that are selected at random at each location), which can be removed easily by hand manipulation. The rate of coir degradation will be measured by monitoring changes to the condition of coir mesh over time from the ten photo-quadrats, which are collected twice per year to assess changes in the cover of benthic invertebrates at each oyster reef restoration unit.

Frequency: Every 6 months for a minimum of three years, unless otherwise detailed within corrective actions Was the performance objective met? We identified significant oyster spat settlement and growth at all oyster reefs. Whilst this has not yet proliferated to cover of oysters on the outside of the restoration units, or to fully cementing shells within the units together, these positive spat fall and growth results indicate that these performance criteria are likely to be met in the coming years. Therefore, no further action is necessary at this early stage of the project. Supporting evidence Oyster recruitment We collected shells from two sources;

1. 5 shells collected from within the coir bags at each oyster restoration unit (henceforth ‘bags’), and

2. 5 shells collected from ‘oyster necklaces’ which were strings of drilled oyster shells tied together with fishing line and then affixed to the top of the oyster restoration units during installation (henceforth ‘necklaces’).

215

All live oysters growing on the shells were identified to species, counted, and their height measured, and the density of oysters calculated as number of oysters per square meter of shell (as recommended by Baggett et al., 2015). We identified significant oyster settlement during both monitoring events on both necklaces and in the oyster restoration unit bags (Figure 19, Figure S1, Table 4, Table 5). Overall, we identified an average of 387.5 spat/m2 in May 2018, 306.1 oysters/m2 in November 2018, and 349.9 oysters/m2 across both monitoring events (Table 4). Whilst we identified some changes in the densities of oysters at specific restoration sites (Figure S1, Table 5), with some sites showing increased oyster density between sampling events, and others lower oyster density, the overall pattern was for a reduction in oyster density between monitoring events 1 and 2 (Table 4). This change was, however, offset by a significant change in the size-frequency distribution of oyster shells on the reefs (Figure 20, Figure S2). Here oysters increased in size from 14.3 mm (+/- 6.5 mm SD) in May 2018, to 19.1 mm (+/- 8.9 mm SD) in November 2018. Consequently, it is likely that intermediate sized oysters (6-15 mm in length) in May 2018 grew to the larger sizes (i.e. <30 mm) by November 2018, and that the smaller oysters (<6mm) either grew to an intermediate size, or did not survive the winter. Consequently, despite a reduction in overall average density, this was offset by a larger average oyster size during the second monitoring event.

Figure 19 Example images of juvenile oysters growing on shells collected from within the oyster reef restoration units. Table 4 List of oyster quantification methods, and the average densities (with standard deviations) of oysters during both monitoring events, and for all monitoring events in 2018. Monitoring event 1 Monitoring event 2 Overall (both events)

Average (spat/m2) StDev Average (spat/m2) StDev Average (spat/m2) StDev

Necklaces 543.7 643.5 493.9 513.4 525.8 599.3 Bags 272.9 413.3 230.9 438.2 251.9 425.9 All 385.7 537.7 306.1 475.1 349.9 511.7

216

Table 5 Summary of average oyster density (with standard deviation) across all oyster restoration sites. Note that all sites recorded significant spat settlement. Necklaces were unrecoverable from three sites in November 2018 sue to anchor damage.

May 2018 Nov 2018 Bags Necklace Bags Necklace Site Avg (spat/m2) StDev Avg (spat/m2) StDev Avg (spat/m2) StDev Avg (spat/m2) StDev 2 114.7 153.4 159.9 413.5 135.7 153.3 212.5 316.4 3 131.5 132.8 217.9 219.4 37.9 82.4 320.2 364.6 4 49.6 75.1 229.0 274.9 72.4 92.7 317.6 305.0 5 544.8 307.4 1399.2 961.6 473.1 315.7 550.5 601.5 6 165.8 181.3 475.0 562.5 282.8 323.9 No shells recoverable 7 191.1 279.7 664.4 363.3 94.5 195.3 No shells recoverable 8 810.0 944.4 1229.5 527.7 1101.2 984.6 0.0 0.0 9 300.5 374.0 611.7 861.9 57.8 172.3 672.9 278.2 10 145.7 156.9 842.1 722.7 82.1 131.5 No shells recoverable 11 279.8 321.7 409.9 377.3 178.6 277.4 666.1 764.4 12 340.5 308.6 825.5 548.3 86.5 126.6 758.9 511.0 13 589.8 527.2 551.3 316.8 475.4 518.8 1056.5 425.7 14 142.0 178.3 428.7 675.4 117.2 164.2 793.7 468.3 16 15.5 60.0 0 0 37.8 79.2 45.5 150.8

Figure 20 Size frequency distribution of oyster shells across all samples taken in May and November 2018. Cover of oysters and other sessile invertebrates At this stage, we were not able to identify any significant settlement of invertebrates on the outside of the coir bags. This is unsurprising as significant growth of invertebrates on the outside of the coir mesh bags was not expected within the first year, and therefore not a compulsory success criteria for the first year in the monitoring program. We anticipate, however, that given the positive indications of spat recruitment and oyster growth, that it is likely that this target will be achieved fully in the coming years. We were, however, able to use the quadrat samples (see figure 21) to show that there has been no significant degradation of the coir mesh, that it is still strong, and that the gauge of the rope has not been narrowed by any natural decomposition as yet. Similarly, we not that there is minimal macroalgal coverage on the bags (<1% in all cases), meaning that there is no significant fouling occurring that would limit the settlement of oysters either on, or into, the oyster restoration units into the future. Establishment of stable biogenic matrix The development of the reefs is in the early stages, therefore 100% of randomly selected oyster shells were still easily moved during both monitoring events. This is unsurprising as the development of stable biogenic matric was not expected within the first year. We

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0-2

2-4

4-6

6-10

10-15

15-20

20-25

25-30

30-40

40-50

>50

0-2

2-4

4-6

6-10

10-15

15-20

20-25

25-30

30-40

40-50

>50

May2018 November2018

Prop

ortio

nofoysters

SizeRange

217

anticipate, however, that given the positive indications of spat recruitment and oyster growth, that this performance criteria will likely will be achieved in the coming years.

Figure 21 Example images of quadrat photos taken on oyster reef restoration units, here at reefs 13 (top row) and 16 (bottom row). We did not identify any significant growth of invertebrates on the outside of the oyster reef units, but there was also no significant degradation of the coir mesh noted.

218

Monitoring criteria 3- Community use and enjoyment of the declared fish habitat area Performance objective: Oyster restoration units do not significantly impair community use and enjoyment of the declared Fish Habitat Area, particularly fishing activities. Monitoring method: Maintain records of community feedback, evidence of vandalism, and vessel strikes on the trial oyster reef restoration units. Records will be comprehensive and include, as a minimum set:

1. the number and type of complaints received (also, to allow the Department to gauge the success, any positive comments should also be provided);

2. the type, nature, and severity of any acts of wilful vandalism; and 3. the type and severity of any vessel strike.

Frequency: Annual reporting of all complaints, cases of willful vandalism, and instances of vessel strike received for the three-year trial period. Was the performance objective met? Noosa Council holds the data / information on any complaints from the public. We did not identify any willful vandalism of the reefs, but we did identify several instances of boat propeller and anchor strikes in the November 2018 monitoring event. Supporting Evidence Noosa Council is the designated contact for comments regarding the oyster reef restoration project in the Noosa River, and of how this project affects the use and enjoyment of the declared fish habitat area by the community. This is clearly stated on the signage at each oyster reef restoration site (Figure 22). We identified some damage to oyster reefs during the monitoring events (summarized in Table 6). During the monitoring event in May 2018, only one reef required repairs. In November 2018, most reefs required repair, and some coir bags were unrepairable. It is likely that the majority of damage to the oyster reefs is caused by being struck by boat propellers (especially at sites where bags were torn), and/or poorly placed or dragging boat anchors (especially at sites where bags were torn and/or moved from their position on the reef). During the November 2018 monitoring event, four of the reefs were also badly fouled by fishing line and/or abandoned crab pots. This fishing gear was removed from the reefs and disposed of. Combined, this damage resulted oyster spat necklaces being impossible to retrieve from some reefs (see monitoring criteria 2, Table 5). Members of the USC team have conducted several surveys on the fish fauna associated with the oyster reefs in the Noosa River, beyond the monitoring events detailed here. In association with these surveys, team members have taken note of the number of reefs where fishing activities have been occurring (Table 7). Whilst this is simply a snapshot view of the types of activities taking place around the reefs, it is indicative of the potential damage occurring to the reefs. The key observation is a general increase in the number of fishing boat at the reefs with time since installations. This is likely due to 1) the general public becoming increasingly familiar with the locations and benefits of the reefs, 2) media reports of the positive results regarding fish assemblages coming to light, and 3) the timing of September school holidays. Whilst we did not identify any acts of willful vandalism of the oyster reef restoration units during our monitoring events, many marker buoys (placed on the corner of the reef units; a compulsory condition of the development approval) were missing during monitoring events. These were immediately replaced by the USC team. Some buoys were, however, removed from sites the night after being reinstalled, thereby indicating very regular removal of reef

219

hardware by members of the public at some sites. This lack of buoys on the corner of the oyster reefs is likely a contributor to some of the reefs being hit by boat propellers etc. No site signage was damaged or modified during the past year, so no maintenance or replacement has been necessary (Figure 22).

Figure 22 Oyster reef restoration signage being installed at an oyster reef restoration site in the Noosa River, with text on the sign indicating the contact details at Noosa Council for comments on the broader oyster restoration project.

220

Table 6 List of recorded damage and repairs to oyster reefs during both monitoring events Site Damage May 2018 Repairs conducted Damage November 2018 Repairs conducted 2 None NA All top bags torn. Buoy

missing. Repairs to bags completed where possible. Replaced buoy.

3 None NA All top bags torn. Buoy missing.

Repairs to bags completed where possible. Replaced buoy.

4 None NA Bottom right, and top right bag torn. Middle reef unit covered by significant sand. Right hand reef unit and nearby buoy tangled with crab pot.

Repairs to bags completed where possible. Crab pot removed.

5 Reef struck by boat which broke mooring, one buoy missing

Reef repaired and put back in place. Replaced buoy.

All top bags torn. Buoy missing.

Repairs to bags completed where possible.

6 None NA Middle top bag ripped. Repaired bag. 7 None NA Middle top bag ripped. Repaired bag. 8 Right hand bag moved and

torn, 1 buoy missing Repaired and put back on reef. Replaced buoy.

Right hand top bag torn. Buoy missing.

Repaired bag, replaced buoy.

9 Right hand bag moved and torn, 2 buoys missing.

Repaired and put back on reef. Replaced buoys.

Left hand top bag torn, right hand reef significantly covered by sand, middle top bag ripped.

Repaired bags.

10 One buoy missing. Replaced buoy. Top middle bag destroyed, significant covering by sand. Buoy missing.

Repaired bag, replaced buoy.

11 Two buoys missing. Replaced buoys. Top middle bag torn. Repaired bag. 12 None NA Right top bag torn. Repaired bag. 13 None NA None NA 14 None NA None NA 16 None NA None NA

Table 7 List of days USC team members have surveyed the reefs, and the number of reefs upon which fishing activities have been noted during these surveys.

Date of field work Number of days Sites visited Activity being completed Sites with boats present

25-26 November 2017 2 3-8 Ecological function study 0 27-30 November 2017 4 All Fish surveys 0 29-30 December 2017 2 All Fish surveys 2 29-30 January 2018 2 3-8 Ecological function study 0 15-16 February 2018 2 All Fish surveys 1 26-27 March 2018 2 3-8 Ecological function study 2 29-30 March 2018 2 All Fish surveys 1 31 May, 1 June 2018 2 3-8 Ecological function study 0 15-18 May 2018 4 All Fish surveys 2 20-22 May 2018 3 All 6 month monitoring 3 27-28 June 2018 2 All Fish surveys 4 16-17 August 2018 2 All Fish surveys 2 1-2 October 2018 2 All Fish surveys 2 19-20 November 2018 2 All Fish surveys 3 26-27 November 2018 2 All 12 month monitoring 1

221

Monitoring criteria 4- Other potential effects Performance objective Oyster reef restoration units do not cause a decline in the extent of marine plants within a 50 m radius of the restoration units, and are not attributed to erosion of adjacent shorelines or other ambient environmental impacts. Monitoring method Map the area of marine plant habitats (seagrass, mangroves) in the immediate vicinity (i.e. with a 50 m buffer) of each oyster restoration area, using high-resolution GPS (cm scale) and field-validated aerial imagery (sourced from Nearmap) (NearMap, 2018). Monitor changes in the composition coverage and condition of seagrass within 50 m radius around the oyster reefs. Map the location and condition of estuarine shorelines that occur in the immediate vicinity (i.e. with a 50 m buffer) of each oyster reef, using high-resolution GPS (cm scale) and field-validated aerial imagery (sourced from Nearmap). Frequency Annually for a minimum of three years, or until any failed restoration units have been removed. Was the performance objective met? All aspects of this performance objective were met, so no further action is necessary. Supporting evidence Mangroves and seagrasses were successfully mapped at each site using the described methods (Figures 8-21). We identified no significant change in the distribution of mangroves or seagrasses within the 50 m monitoring area (Table 8; t test, P>0.9), or any erosion to nearby shorelines. Whilst the footprint of seagrass was lower at some sites, there was no loss of seagrass attributable to the installation of the reefs (i.e. significant decline of seagrass away from the reefs, especially within the RAA areas). Photographs of the mangrove fringe and shoreline within the 50 m monitoring area both before the installation, and during the 12 month monitoring event are available from USC upon request. Table 8 Extent of seagrass and length of mangrove edge at within 50m of each oyster restoration site at installation and 12 months post installation.

Reef Mangroves at installation (m)

Mangroves at 12 months (m)

Seagrass at installation (m2)

Seagrass at 12 months (m2)

2 113.9 119.4 0 0 3 92.0 87.9 1995.9 1443.6 4 135.6 133.8 0 0 5 162.3 175.2 0 0 6 39.4 43.6 1372.1 1590.0 7 43.1 48.2 1480.8 1913.5 8 108.8 112.0 0 0 9 180.8 181.8 0 0 10 110.7 114.4 0 0 11 138.5 137.5 0 0 12 114.4 116.9 151.9 499.0 13 81.7 82.3 2105.0 1707.5 14 134.5 136.2 854.6 765.5 16 0 0 0 0

222

No change to seagrass composition or density was identified at any restoration sites. All sites were dense (>70% cover) and long (>20 cm) eelgrass (Zostera capricorni) (except for sites 13 and 16- see below). During the 12 month monitoring event, we identified small (leaf height <1 cm), low coverage (<5%) growth of dugong grass Halophila ovalis inside the RAA area at site 13, and some small patches within the rocky shore adjacent to site 16. This seagrass was not present at the sites during installation (Table 1). The growth of seagrass at these sites is likely due to lower runoff in Lake Weyba and Weyba Creek over the past yeah, and especially in the past 6 months, and the stabilizing and baffling effects of the reefs on surrounding sediments (especially at site 13). Further monitoring will be required on these growing seagrasses. Conclusions In this report, we show that;

5. Oyster restoration units have not moved from the RAA area, nor have they moved greater than 1 m within the RAA area.

6. There has been significant spat settlement and oyster growth at all oyster reef restoration sites. Whilst this has not yet proliferated to coverage of invertebrates on the outside of the coir bags, or to the proper stabilisation of shell within the bags (i.e. cementing of the biogenic reef matrix), these oyster growth results are a positive sign for the likely success of the reefs in achieving these performance criteria in the near future.

7. Whilst we did not identify any wilful damage to the oyster restoration units, we did identify that marker buoys were regularly removed from the oyster restoration sites, and there have been several instances of boat propeller and anchor strikes on the reefs. This was repaired where possible.

8. There has been no significant change to the distribution, composition or quality of seagrass or mangroves around the oyster reef sites. Similarly, there has been no shoreline erosion at oyster reef sites.

Consequently, two of the four monitoring targets reported on here have been fully met, the third relating to oyster and invertebrate growth is tracking very favorably with the criteria likely to be met in the coming years, and the final relating to community usage may require closer monitoring. We identified some damage to the oyster reef units from boating activities (principally anchor and prop damage) at many sites during the November 2018 monitoring event. We have some evidence to suggest that the use of oyster reefs by fishers has increased over time, and that some of the ecological benefits of the reefs might therefore be being offset by fishing. The three reefs more distant from the main boating activities in Lake Weyba and Weyba Creek (due to the combined effects of limits on access to hire boats, and difficulty of access due to shallow waters) are the most intact oyster reefs, with no damage recorded in either 2018 monitoring events. Whilst these reefs do not necessarily have the highest rates of settlement (the exception being reef number 13, which has the highest average live oyster density), their success might be the most guaranteed as they are less likely to be damaged by the sorts of impacts occurring on reefs in the central stretch of the river. References BAGGETT, L. P., POWERS, S. P., BRUMBAUGH, R. D., COEN, L. D., DEANGELIS, B. M.,

GREENE, J. K., HANCOCK, B. T., MORLOCK, S. M., ALLEN, B. L., BREITBURG,

D. L., BUSHEK, D., GRABOWSKI, J. H., GRIZZLE, R. E., GROSHOLZ, E. D., LA

PEYRE, M. K., LUCKENBACH, M. W., MCGRAW, K. A., PIEHLER, M. F., WESTBY,

223

S. R. & ZU ERMGASSEN, P. S. E. 2015. Guidelines for evaluating performance of

oyster habitat restoration. Restoration Ecology, 23, 737-745.

NEARMAP. 2018. NearMap photomaps [Online]. Available: http://www.nearmap.com.au/

[Accessed November 2018].

224

Appendices

Figure S1 Map of oyster reef restoration sites with average (+/- SD) values for oyster density during May and November 2018 monitoring events in the oyster reef bags (B) and on the oyster necklaces (N).

225

Figure S2 Map of oyster reef restoration sites with average (+/- SD) values for oyster length during May and November 2018 monitoring events in the oyster reef bags (B) and on the oyster necklaces (N).