Market access, population density, and socioeconomic development explain diversity and functional group biomass of coral reef fish assemblages

Market

access,

population

density,

and

socioeconomic

development

explain

diversity

and

functional

group

biomass

of

coral

reef

fish

assemblages

Tom

D.

Brewer

a,

*

,

Joshua

E.

Cinner

a

,

Rebecca

Fisher

b

,

Alison

Green

c

,

Shaun

K.

Wilson

d,e

a_{AustralianResearchCouncilCentreofExcellenceforCoralReefStudies,JamesCookUniversity,Townsville,QLD4811,Australia}

b

AustralianInstituteofMarineScience,UWAOceansInstitute(M004),35StirlingHwy,Crawley6009,Australia

c

TheNatureConservancy,51EdmonstoneStreet,Brisbane4101,Australia

d

MarineScienceProgram,DepartmentofEnvironmentandConservation,17DickPerryAve,Kensington6151,Australia

e

OceansInstitute,UniversityofWesternAustralia,Crawley6009,Australia

1. Introduction

There is overwhelming evidence that human activities are profoundly altering marine ecosystems on a global scale (e.g.

Hughes,1994;Pandolfietal.,2003).Ofparticularconcernarethe poorly understood, yet potentially disastrous environmental changesthathumanactivityiscausingtothefunctioningofcoral reefecosystemsuponwhichmillionsofpeopledepend(Moraetal., 2011). Ecosystem function is conceptually and analytically complex, requiring a diverse array of metrics to understand ecosystemresponsetohumanactivity.Highbiologicaldiversityis thoughttocontribute tomaintainingecosystem resilience (e.g.

McCann,2000;Cardinaleetal.,2006;Tilmanetal.,2006)through increased response diversity to perturbations and functional redundancy (Naeem, 1998; Chapin et al., 2000; Hooper et al., 2005; Maestre et al., 2012; but see Ives and Carpenter, 2007), assumingthat species respond to threats uniquely. The useof diversitymetricsassurrogatesforecosystemfunctionhowever, does not come without criticism. There is, for example, some evidencethatparticularspecies(Bellwoodetal.,2003,2006;Hoey

and Bellwood, 2009), and functionalgroups (e.g.Hughes et al., 2007) perform disproportionately important functional roles, actingasenergyconduitsthroughtrophiclevelsandmaintaining broaderecosystemprocesses.Therefore,itisimportanttoconsider measuresofbothdiversityandfunctionalgroupstounderstand howecosystemsmayrespondtohumanactivities.

Coral reef fish are perhaps the most diverse of vertebrate communities, are vital to ecosystem function, and provide significantgoodsandservicestopeople.Arangeoffactorshave beenidentifiedasimportantdriversofthediversity,biomass,and abundanceofreeffishfunctionalgroups.Atlargebiogeographic scales,distributionsofdiversityandfunctioncanbeexplainedby environmental factors, including available habitat, latitude– longitudegradients,themid-domaineffect,gyreinfluence,history of environmental stress, and larval subsidy from species-rich regions(BellwoodandHughes,2001;Connollyetal.,2003;Mora et al., 2003; Bellwood et al.,2005; Mora and Robertson, 2005; McClanahan et al., 2011). At local and national socio-political scales,variousenvironmentalandsocialfactorshavebeenusedto explainfishdiversityandbiomassoffunctionalgroups. Environ-mental factors include depth, exposure to prevailing weather, season,reefzone,coralcover,substraterugosity,habitat complex-ity,andlarvaldispersal(LuckhurstandLuckhurst,1978;Molles, 1978; Bell and Galzin, 1984; Roberts and Ormond, 1987; ARTICLE INFO

Articlehistory:

Received23September2011

Receivedinrevisedform24January2012 Accepted26January2012

Availableonline23February2012

Keywords:

Distaldriver Proximatedriver Fishingpressure Coralcover

Structuralequationmodel Trade

ABSTRACT

Thereisoverwhelmingevidencethatmanylocal-scalehumanactivities(e.g.fishing)haveadeleterious effect on coral reef fish assemblages. Our understanding of how broad social phenomena (e.g. socioeconomicdevelopment)affectthediversityandfunctionofcoralreeffishassemblageshowever,is still poor. Here, we use structural equation models to reveal how human population density, socioeconomicdevelopment,andmarketaccessaffectfishingpressureandcoralcoverto,inturn,explain thediversityandbiomassofkeyfunctionalgroupsofreeffishassemblageswithinSolomonIslands. Fishingpressureispredominantlydrivenbybothmarketaccessandlocalpopulationdensity,andhasa clearnegativeeffectonthediversityandfunctionofcoralreeffishes.Thestrongpositiveeffectofmarket accessonfishingpressuremakescleartheimportanceofunderstandingsocial-ecologicallinkagesinthe contextof increasinglyconnected societies.This studyhighlightstheneed toaddressbroad social phenomenaratherthanfocusingonproximatethreatssuchasfishingpressure,toensurethecontinued flowofcoralreefgoodsandservicesinthistimeofrapidglobalsocialandenvironmentalchange.

ß 2012ElsevierLtd.Allrightsreserved.

*Correspondingauthor.Tel.:+61433976561.

E-mailaddress:[email protected](T.D.Brewer).

ContentslistsavailableatSciVerseScienceDirect

Global

Environmental

Change

j ou rna l hom e pa ge : w w w. e l s e v i e r. c om/ l o ca t e / gl oe n v cha

FriedlanderandParrish,1998;Friedlanderetal.,2003;Gratwicke andSpeight,2005;Jonesetal.,2005;Grahametal.,2006)(Table1). Incontrasttothedepthofworkassessingenvironmentaldriversof fish diversity and function, assessments of the potentially important role that human activity might have in shaping ecologicalassemblageshavefocusedlargelyonhumanpopulation density(Jenningsetal.,1995;JenningsandPolunin,1996,1997; Bellwoodetal.,2003;Dulvyetal.,2004a,b;Mora,2008;Williams etal.,2008,2011;Stallings,2009)andfishingpressure(DeMartini etal., 2008;Wilsonetal., 2008). However, recent researchhas highlighted the potentially important role of factors such as market access and socioeconomic development in explaining functionalgroupdistributions(AswaniandSabetian,2009;Brewer etal.,2009;Cinneretal.,2009a;Stallings,2009).Whatisnotclear, however,ishowmarketaccessandsocioeconomicdevelopment affect fish species diversity, and whether market access and socioeconomicdevelopmenthaveaneffecton fishdiversityand functionbeyondwhatisexplainedbyhumanpopulationdensity. Thispaperaimstocontributetothisresearchgapbyexamining relationshipsbetweensocialdriversandthediversityandfunction ofreeffishcommunities.

Socialscientistsworkingonhuman–environmentinteractions oftendifferentiatebetweenproximate(e.g.fishingpressure)and distal(e.g.marketaccess,socioeconomicdevelopment,andhuman populationdensity)driversofenvironmentaldegradation( Forest-er and Machlis, 1996; Agrawal and Yadama, 1997; Geist and Lambin,2002;Krameretal.,2009;McKinneyetal.,2010).Inacoral reefcontext,thereisclearevidencethat,atthelocal-scale,people directly affect coral reef fish diversity and function through proximatedriverssuchasfishingpressureandhabitatdegradation (Wilsonetal.,2008).Whatisnotclear,however,istheroleofdistal drivers,inshapingtheseproximatedriversandultimatelycoral reeffishdiversityandfunction.Forexample,increased socioeco-nomic development does not directly affect fish diversity and function,butmightintensifylocalfishingpressurethroughgreater

accesstomoreefficientfishinggear,whichmight,inturn,decrease diversityand functionofcoralreeffish.Alternatively,increased socioeconomic development might reducedependence on local resources, or enable improved resource management practices, resulting inincreasedfish diversity andfunction (Cinneret al., 2009a).

Here,weexplorethelinkagesbetweenthreerecognizeddistal drivers (population density, socioeconomic development and marketaccess),twoproximatedrivers(fishingpressureandcoral cover),andarangeofmetricsoffishdiversityandfunction.Weuse structuralequationmodelstounderstandthesequentialeffectsof distal drivers on proximate drivers, and proximate drivers on diversityandfunctionmetricsacross25sitesinSolomonIslands. SolomonIslandsisalow-incomenationwithexceptionallyhigh fishspeciesdiversity(Greenetal.,2006a).Thevastmajorityofthe rapidlygrowingpopulationlivesalongthecoastlinethroughout themultitudeofislandswhichcomprisethenation.Reeffishand fishingaresignificanttoSolomonIslandslivelihoods,yetfuture demandforfishisexpectedtoexceedecologicalreplenishment (Bell et al., 2009). Increasing market access, and evidence of declining fish biomass withincreasing market proximity (ANU Enterprise, 2008), particularly for certain taxonomic groups (Breweretal.,2009),suggestthatfishingpressure,andassociated social and economic change, contribute to variability in the distributionofreeffishdiversityandfunction.Byassessingand comparingtheeffectofdistalandproximatedriversonreeffish assemblagesthisstudyquantifiestherelativeimportanceofthese socialdriversinadiverseecologicalsystem.

2. Methods

2.1. Siteselectionanddelineation

The25sitesinthisstudyspansevenofthenineprovincesin SolomonIslands(Fig.1).Wecollectedbothecologicalandsocial

Table1

Keyenvironmentalandhumanfactorsthatexplaininsitucoralreeffishdiversityandfunctionalgroupdistributionsatbiogeographicscalesrelevanttothisstudy.

Factor Explainsdiversitymeasuresa

Explainsfunctionalgroupbiomassb

Controlledforinthisstudy?

Environmental

Depth 1,2,3.c

1,3. Sd

Exposure 4. 11,12. S

Time 5. – S

Reefzone 1. 1,13. S

%Coralcover 6,7. 1,3,7. Me

Habitatcomplexity 4,8,9. 9,14. *f

Habitatrugosityg

1,4,10. 1,15. *

Human

Proximatedrivers

Fishconsumption – 14. M

Fishingpressure 16. 16,19,20. M

Distaldrivers

Populationdensity 17,18. 15,17,18,21–26. M

Affluence – 15,18.h

M

Urbandevelopment – 24,27. M

Marketaccess – 28. M

a_{Referencesrelatetoanymeasureofdiversity(e.g.richness,evenness).}

b_{Referencesrelatetoanyfunctionalgroupabundanceorbiomassmeasure(e.g.herbivorebiomass)} c

Supportingreferences:1.FriedlanderandParrish(1998),(2).RobertsandOrmond(1987),(3).O¨ hmanandRajasuriya(1998),(4).Friedlanderetal.(2003),(5).Molles (1978),6.BellandGalzin(1984),(7).Wilsonetal.(2006),(8).GratwickeandSpeight(2005),(9).Grahametal.(2006),(10).LuckhurstandLuckhurst(1978),(11).Fultonand Bellwood(2005),(12).Floeteretal.(2007),(13).Russ(2003),(14).Wilsonetal.(2008),(15).Cinneretal.(2009a),(16).Jenningsetal.(1995),(17).JenningsandPolunin (1997),(18).Stallings(2009),(19).DeMartinietal.(2008),(20).JenningsandPolunin(1996),(21).Bellwoodetal.(2003),(22).Dulvyetal.(2004a),(23).Dulvyetal.(2004b), (24).Mora(2008),(25).Williamsetal.(2011),(26).Williamsetal.(2008),(27).AswaniandSabetian(2009),(28).Breweretal.(2009).Referencesincludeonlythosethat show,statistically,theeffectofthefactorslisted,onmeasuresofcoralreeffishdiversityorfunction.

d_{Controlledduringsampling.} e

Controlledinmodel. f

Correlatedwithcoralcover(Grahametal.,2008). g

Alsomeasuredasnumberof‘holes’andholevolumeofreefsubstrate(FriedlanderandParrish,1998). h

[image:2.595.34.557.461.632.2]

dataforeachsite.Sitelocationswerepredeterminedbyamarine assessmentwhichsurveyedbothreeffishpopulationsandcoral substrate (Green et al., 2006a). Itwas necessary to define the spatialextentofeachsitetodeterminewhichhumanpopulations were associated with the coral reef at the location where the marineassessmentoccurred.Thespatialextentofeachsitewas elicitedfromindividualspossessinglocalknowledgeofresource usebypeopleresidinginadjacentvillages.Themarineboundary wasdefinedasthearea,withinthevicinityofwherethefishwere surveyed,which was likely tobe exploitedby people living in villagesontheadjacentcoastline(Breweretal.,2009).Alternative methodsexistforestimatingresourceuseboundaries,including frictionmappingusingthiessenpolygons(MullerandZeller,2002), ethnographic studies (Aswani, 1999), and participatory GIS mapping(Aswani,2011).However,thelarge-scalenatureofthis study inhibited the use of these more localized resource use mappingtechniques.Ourapproachmadeuseofthebestavailable information at the spatial scale of our study. The terrestrial boundaryofeachsitewassetat1kmfromthecoastline,extending alongthecoasttowherethemarineareaboundaryintersectedthe coast.Socialdataon villageswithintheaforementioned bound-aries wasprovidedby theSolomon Islandsstatisticsoffice and derived from a national census (1999), and a national village resourcesurvey(2007/2008).

2.2. Ecologicalresponsevariables

Fish species used in this study includedall fishessurveyed acrossthe25sites(ESM1).Fishweresurveyed,fromMaytoJune 2004,usingunderwatervisualcensusalongfive50mtransectsat eachsite,atadepthof8–10monreefslopeinlocationsexposedto prevailingweatherconditions(Greenetal.,2006b).Fourmetricsof speciesdiversity wereusedto testthe effectofproximate and distaldriversonfish diversity;(1) speciesrichness, (2) Pielou’s speciesevenness,(3)averagetaxonomicdistinctness(AvTD),and (4)variationintaxonomicdistinctness(VarTD).Speciesevenness warrantsinvestigationasitconsiders therelative abundanceof species and can have important ecosystem ramifications well beforespeciesbecomelocallyextinct(Chapinetal.,2000).AvTDis ameasureoftheaveragedistancebetweenallpairsofspeciesina taxonomic tree, which captures phenotypic differences and functional richness (Clarke and Warwick, 1999; Rogers et al., 1999).VarTDisthevarianceofthepathlengthsbetweeneverypair ofspeciesinataxonomictree,andrepresentstheunevennessof the taxonomic tree (Clarke and Warwick, 2001). Taxonomic hierarchylevels used tomeasure AvTD and VarTDwere Class,

Order,Family,GenusandSpecies.Pathlengthsbetweentaxonomic levelswereequallyweighted.

Totalbiomassestimateswerederivedfortwo keyfunctional groups: piscivores and herbivores.Piscivores wereclassified as fishesthat,basedongutcontentanalyses(FroeseandPauly,2011) predominantlyconsumefishes.Piscivorescaninhibitincreasein abundanceoflowertrophiclevelspeciesthroughpredatorprey interaction(Jenningsetal.,1995; Grahametal.,2003),and are particularlysensitivetofishingpressure(e.g.JenningsandPolunin, 1997;DeMartinietal.,2008;Sandinetal.,2008).Herbivoreswere classifiedasthosespeciesthatpredominatelyfeedonlargefleshy algae orthe epilithicalgal matrix (censu Wilsonand Bellwood, 1997).Thisincludesfishthatremovepartofthereefbyscrapingor excavating the substratum, and grazers that mainly ingest filamentous algae (censu Choat et al., 2002). Herbivores are thought to play a critical role in the resilience of coral reef ecosystems by preventing algal overgrowth that can smother corals(Mumby,2006;Hughesetal.,2007;GreenandBellwood, 2009).Piscivoreandherbivorespeciesweredividedintotargetand non-targetspeciestofurtherexploretheeffectoffishingonthese functional groups.Thelistoftargetand non-targetspecieswas constructedusingexpertopinionofSolomonIslandstargetspecies (Greenetal.,2006b),andcreelsurveydatafromadjacentPapua NewGuinea(Cinneretal.,2009b)(ESM1).

2.3. Proximatedrivers

Wemeasured twoproximatedriverspreviouslyshowntobe related to fish diversity and function: (1) coral cover, and (2) fishing pressure. Coral cover was measured along the same transectsusedinthefishsurvey.Substratetypewasclassifiedas coralornon-coralatthreepointsat2mintervalsalongeach50m transect totalling 375points ateach site (Hughes,2006). Coral cover,asapercentageoftotalsubstrate,rangedfrom13to68% (mean=44%)amongsites.Wemeasuredfishingpressureasthe number of people selling fishand numbers of fiberglassboats, eskies(insulatediceboxesusedforfishpreservation),spearguns and fishing nets per coral reef area at each site. We included numberofpeoplesellingfishasavariablebecauseofevidencethat themarket-basedfisheryadverselyaffectscoralreeffishstocksin SolomonIslands(AswaniandSabetian,2009;Breweretal.,2009). Coral reef area wasmeasured usingGoogle Earth Professional, followingthemethodsdiscussedinBreweretal.(2009).Fishing pressurevariableswereconsolidatedtoasinglevariable(

l

=3.3; varianceexplained=65.4%)usingPrincipalComponentsAnalysis. Allfishing pressurevariables loaded positivelyon theprincipal componentat0.65.

2.4. Distaldrivers

We developed metrics for three key distal drivers that we hypothesizedcouldstronglyinfluencereeffishassemblages:(1) humanpopulationdensity,(2) socioeconomicdevelopment,and (3)marketaccess.Humanpopulationdensitywasmeasuredasthe totalnumberofinhabitantspercoralreefareaateachsite.Human populationdensityrangedfrom11to1035(mean=159)people perkm2ofcoralreef.Basedoncalculatedpopulationdensitiesand conservative per capitafish consumption estimates (Bell et al., 2009),eightofthe25sitesexceededestimatedsustainableharvest levelsof5metrictonneskm2yr1

(Newtonetal.,2007). Socioeco-nomic development was measured as the sumof a set of un-weightedinfrastructuresandamenitiessimilartothoseusedin previousstudies(Cinneretal.,2009a;Pollnacetal.,2010); pre-school, kindergarten,primary school,highschool,clinic,wharf, tradestore,supermarket,postalservice,fueldepot,creditfacility andairportateachsite.Theinfrastructureandamenityvariables

[image:3.595.47.289.53.216.2]

displayedhighinternalconsistency(Cronbach’s

a

=0.823) allow-ingaggregationofvariablesintoasinglemeasureofsocioeconomic development.Theaggregatescoreforeachsitewasthendividedby the number of villages at each respective site, to control for infrastructureandamenityaccessibility(CinnerandMcClanahan, 2006).Marketaccesswasmeasuredastheun-weightedsumofthe shortestdistancefromeach ecologicalsamplinglocation tothe nearestlocalfishmarket,provincialcapital,andnationalcapital,all ofwhichhavefishmarkets,usingroadsandseaaspossibleroutes withinthesamedistancemeasure(Breweretal.,2009).

2.5. Modelconstruction

Weusedpartialleastsquaresregression,intheprogramWarp PLS,tobuild structural equationmodels (SEMs)of the general form: distal drivers!proximate drivers!ecological response.

Partialleastsquareswaschosenovercovariance-basedapproaches primarily because it suited our small sample size (Chin and Newstead,1999;Reinartzetal.,2009).Thedistalandproximate driversremainedconsistentacrossmodelswithonlytheecological response changing between models (Fig. 2a). All distal drivers (market access, socioeconomic development and population density)werelinkedtothetwoproximatedrivers(fishingpressure andcoralcover),exceptmarketaccesstocoralcoverastherewas no theoreticaljustification for this link.Bothproximate drivers werelinkedtotheecologicalresponsevariableinallmodels.This resultedinuniquemodelsfor eachecologicalresponse; species richness,Pielou’sevenness,AvTD,VarTD,totalherbivorebiomass, non-targetherbivorebiomass,totalpiscivorebiomass,and non-targetpiscivorebiomass.Thepartialleastsquaresmethodpartials out each analysis (e.g. the effect of population density and socioeconomic development on coral cover) from the overall

[image:4.595.77.513.237.685.2]

modelandistherefore,inthisstudy,theequivalenttosetsof non-linearregressions,exceptthatoverallmodelfitstatisticsarealso generated.Themodelswerebootstrapped,setat100 iterations. Weconstrainedallmodelstosecond-orderpolynomial relation-ships,therebyallowingsimple,non-linearrelationshipsbetween variables.WarpPLSsoftwarehasaninbuiltfunctionwherebythe relationshipbetweentwovariableswilldefaulttoasmallerorder polynomial if it is deemed linear (Kock, 2010). The output generatedincludedindividualstandardizedpathcoefficients(

b

), partialmodel fit scores (r2_{), overall model p values calculated}

through resampling estimations coupled with Bonferroni-like corrections(Kock,2010),andindividualexplanatoryandresponse

x,yplots(ESM2).Thetotaleffectofdistaldrivers(marketaccess, socioeconomicdevelopmentandpopulationdensity)were calcu-latedbymultiplyingthestandardizedcoefficients(

b

)withineach pathwaythensummingthesevaluesforpathwaysassociatedwith eachdriver.

3. Results

3.1. Effectsofproximatedriversonfishfunctionanddiversity

Fishing pressure had a clear negative effect on both fish diversity and the biomass of key functional groups of fish. Specifically, fishing pressure correlated negatively withspecies richness,andAvTD,andpositivelywithspeciesevenness(Fig.2b– d).Fishingpressurehowever,didnotnoticeablyaffectVarTD,with onlyasmalldecreaseinVarTDassociatedwithincreasedfishing pressure(Fig.2e).Also,fishingpressurehadaclearnegativeeffect on both allpiscivore and allherbivore biomass, yetnon-target biomassofthetwofunctionalgroupswasnegligiblyaffectedby fishingpressure (Fig. 2f–i).Coral covergenerally had a smaller effect on diversity and functional group metrics than fishing pressure(Fig.2b–i);Coralcoverwaspositivelyrelatedtorichness, AvTD,VarTDandnon-targetpiscivores.Coralcoverwasstrongly negatively correlated with species evenness and all herbivore biomass(Fig.2c,g).

3.2. Effectsofdistaldriversonfishfunctionanddiversity

Distal drivers explained much of the variance of fishing pressure (r2=0.73), particularly market access and population density(Fig.2a).Populationdensityandsocioeconomic develop-mentwere,however,comparativelypoordescriptorsofcoralcover (r2_{=0.24).Socioeconomicdevelopmenthadaweaknegativeeffect}

onfishingpressureandoncoralcover,thushadbothpositiveand negative effects on fish diversity and function, except for all herbivorebiomasswhichwaspositivelyaffectedbysocioeconomic development through both decreased fishing pressure and decreased coralcover. Population densityhad a negative effect onallherbivorebiomassthroughincreasedfishingpressure,anda positive, but weaker, effect on all herbivore biomass through decreased coral cover. Increased market access and population density, more than socioeconomic development, explained de-creaseddiversity and functionofcoral reeffish(Fig.2b–i).The strongindirecteffectofmarketaccess,ondiversityandfunction, was particularly noteworthy because the model specified that market access indirectly affected function and diversity only throughfishingpressure,ratherthanthroughbothfishingpressure andcoralcover(Fig.2a).

4. Discussion

This study explored how habitat and social factors explain spatialvariabilityinthediversityandfunctionalgroupbiomassof coralreef fishesat25 sitesacrossSolomon Islands.Our results

demonstratethatpopulationdensityandmarketaccessincrease fishing pressure, which is a major driver of fish diversity and functional group biomass. These distal social drivers have a negative effect on the biomass of piscivores and herbivores targetedby fishers.Moreovertherelative abundanceof species becomesmoreeven,whilstspeciesrichnessandAvTDdeclineas populationdensityincreasesandmarketsbecomemoreaccessible.

4.1. Explainingtheeffectsofproximatedriversonfishfunctionand diversity

Adeclineintaxonomicdistinctnessisoftenassociatedwitha declineinfunctionaldiversity(Rogersetal.,1999;Chapinetal., 2000;Nystro¨metal.,2000).Thisissupportedherebythefinding thatfishingpressurehadanegativeeffectonaveragetaxonomic distinctnessandthebiomassoftwoimportantfunctionalgroups, herbivoresandpiscivores.Conversely,therewasnegligibleeffect offishingonnon-targetspeciesfromthesefunctionalgroups.Thus, thedirecteffectoffishingislikelytobeconfinedlargelytothose species and functional groups that are targeted by fishers. A decrease in diversity, and increased evenness with increased fishingpressure,mightrelatetoremovalofrelativelyrarelarge bodiedpredators,whichareoftentargetedbyfishers(Paulyetal., 1998).

Target specieson coral reefstendtobelarge bodied(Dulvy etal.,2004b;Grahametal.,2005)andmanyofthesmallspecies have closeaffiliation withthereefbenthos(Mundayand Jones, 1998).Thisissupportedbythefindingthatcoralcoverandfishing pressurehadasimilareffectonnon-targetherbivoreandpiscivore biomass, compared to all herbivore and all piscivore biomass whichwaslargelyexplainedbyfishingpressurealone.For non-targetherbivorestherelationshipwithcoralcoverwasnegative, possiblybecausemanyofthesefisharedamselfishesthatmaintain territories covered with algae (Ceccarelli, 2007) (ESM 1). Con-versely,biomassofpiscivores,particularlynon-targetedspecies, tendedtoincreasewithcoralcover.Thismaybebecausemany smallerbodied non-targetpredators and theirprey take refuge among corals. Indeed a loss of coral and associated structural complexitycanleadtodeclinesinsmallbodiedpreyfishandtheir mediumsizedpredators(Grahametal.,2007).Functionally, non-target speciesare likely toperforma verydifferentrole tothe larger bodied species targeted by fishers. Fishing and habitat degradationmightthereforehavedifferentconsequencesforboth herbivoreandpiscivoreassemblagesandthefunctionalservices associatedwiththesegroups.

4.2. Explainingtheeffectsofdistaldriversonfishfunctionand diversity

Bydisaggregatingdistalandproximatedriversandmodeling fish assemblage response to different causal paths of human activity,wehaveshownthatsocialdriverscanhavebothpositive andnegativeeffectsonfishcommunitiesandtheirfunctionalrole inecosystems.Populationdensityhadbothpositiveandnegative effectson allherbivorebiomassthroughcoralcoverandfishing pressure,respectively.Thepositiveeffect,throughdecreasedcoral cover,mightbeexplainedbyincreasednutrientlevelsindirectly caused by high coastal population densities without access to sewagetreatmentfacilities.Resultantexcessnutrientshavebeen showntoincreasealgalgrowth(e.g.PastorokandBilyard,1985), andconsequentlyincreasefoodavailabilitytoherbivores.

increasedspeciesrichness, average taxonomicdistinctness, and functionalgroupbiomass.Ourresultsarebroadlyconsistentwith studiesconductedacrossfiveIndianOceancountriesthatfounda decreaseinfishingwithhigherlevelsofsocioeconomic develop-ment(Cinneretal.,2009a;CinnerandBodin,2010).Incomparison tothelargesocioeconomicdevelopmentspectruminthese multi-nationstudies,therelativelysmall effectsizeof socioeconomic developmenton fishingpressure presentedin this studymight reflectasmalldevelopmentgradientinSolomonIslands.

Themajorityofstudiesthathaveexploredtheeffectofhuman activityoncoralreeffishdiversityandfunctionhaveshownthat theseassemblage characteristicsareexplainedbyeitherfishing pressure (Jennings et al., 1995; Jennings and Polunin, 1996; DeMartinietal.,2008)orhumanpopulationdensity(Jenningsand Polunin,1997;Bellwoodetal.,2003;Dulvyetal.,2004a,b;Mora, 2008;Williamsetal.,2008,2011;Stallings,2009)(Table1).While itisclearthatlocalhumanpopulationdensityanddirectfishing effectsareimportantinexplainingecologicalgradients,wehave shownhere thattrade, measuredas distancetomarket,isalso important(Fig.2).IntheSolomonIslandstradelikelyaffectsfish diversityandfunctionthroughsmall-scalecommercialfishingto supplyurbanmarkets,whereaspopulation densitylikelyaffects diversityandfunctionthroughsemi-subsistencebasedfishingto supply local needs. Trade allows societies toacquire resources fromfurtherafield,externalizingenvironmentalfootprintsbeyond local human–environmentsystems (Arrow et al., 1995; Berkes et al., 2006; Shandra et al., 2009). Resource management and biodiversityconservationinitiativesmustrecognizethattradeand localpopulationpressurerepresentdifferentdriversofecological degradation, and should therefore consider applying different strategies to address their different effects on ecosystems. For example, strong governance of markets through sustainable harvesting certification, and market-specific gear and species restrictions,willbecomeincreasinglyimportantifcoralreeffish continuetobeareadilytradedcommodity(Berkesetal.,2006).

4.3. Futuremodelextensions

Expansionofthemodelsdevelopedinthispapertoother social-ecologicalcontextswouldhelptoprovideabetterunderstanding ofhowmarineecosystemswillrespondtokeysocialdynamics. However,threekeyadvancementsarenecessarytoimprovethe predictivecapacityofsuchmodels.First,analysisoftheindirect effects of distal drivers on theproportionate representation of multiplefunctionalgroups(includinghigherresolutionherbivore functional groups such as grazers, scrapers, and excavators) (Wilson and Bellwood, 1997) and species might lend further insightintotheroleofdistaldriversinshapingecosystemfunction (Wilsonetal.,2008).Second,coralcoverisonlyonemeasureof coral reef habitat and more detailed models including other environmental and habitat variables (e.g. Wilson et al., 2008), couldshedadditionallight ontherelativecontributionofdistal driversondiversityandfunction,particularlyforspeciesrichness and non-target assemblages of coral reef fish. Third, temporal assessmentswouldbevitaltounderstandthefeedbacksthatmight existinthissystem.

5. Conclusion

Managementmeasureswhichaddressproximatedrivers,such asfishingpressure,typicallyhavelocalizedeffectsondiversityand ecosystemfunction.Yet,theyarelimitedintheirabilitytoalleviate the effects of distal social drivers such as market access and socioeconomicdevelopment(Birkeland,2004).Therefore,whilst managing proximate threats represents an important (if not limited) management approach, and means of increasing local

resilience, governing reefs in a changing world will require becoming better acquainted withthe challenges, and potential solutions posed by broadersocial drivers suchas marketsand development.

Acknowledgments

WethanktheAustralianResearchCouncilCentreofExcellence forCoralReefStudiesforfundingthisresearch.ThankyoutoA. Hughes for providing substrate data. Thank you to theNature ConservancyandtheSolomonIslandsstatisticsoffice,particularly N.Gagaheforprovidingdata.ThankyoutoK.MoonandN.Graham forhelpfulcommentsonearlierversionsofthemanuscript.

AppendixA.Supplementarydata

Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,atdoi:10.1016/j.gloenvcha.2012.01.006.

References

Agrawal,A.,Yadama,G.,1997.Howdolocalinstitutionsmediate marketand populationpressuresonresources?ForestPanchayatsinKumaon,India. De-velopmentandChange28,435–465.

ANUEnterprise,2008.People’sSurvey2008.AustralianNationalUniversity. Arrow,K.,Bolin,B.,Costanza,R.,Dasgupta,P.,Folke,C.,Holling,C.S.,Jansson,B.-O.,

Levin, S.,Ma¨ler, K.-G.,Perrings, C.,Pimentel, D., 1995.Economicgrowth, carryingcapacity,andtheenvironment.EcologicalEconomics15,91–95. Aswani,S.,1999.Commonpropertymodelsofseatenure:acasestudyfromthe

RovianaandVonavonaLagoons,NewGeorgia,SolomonIslands.HumanEcology 27,417–453.

Aswani,S.,Sabetian,A.,2009.Implicationsofurbanizationforartisanalparrotfish fisheriesintheWesternSolomonIslands.ConservationBiology24,520–530. Aswani,S.,2011.Socioecologicalapproachesforcombiningecosystem-basedand customarymanagementinOceania.JournalofMarineBiology,doi:10.1155/ 2011/845385.

Bell,J.D.,Galzin,R.,1984.Influenceoflivecoralcoveroncoral-reeffish communi-ties.MarineEcologyProgressSeries15,265–274.

Bell,J.D.,Kronen,M.,Vunisea,A.,Nash,W.J.,Keeble,G.,Demmke,A.,Pontifex,S., Andre´foue¨t,S.,2009.PlanningtheuseoffishforfoodsecurityinthePacific. MarinePolicy33,64–76.

Bellwood,D.R.,Hoey,A.S.,Choat,J.H.,2003.Limitedfunctionalredundancyinhigh diversitysystems:resilienceandecosystemfunctiononcoralreefs.Ecology Letters6,281–285.

Bellwood,D.R.,Hughes,T.P.,2001.Regional-scaleassemblyrulesandbiodiversityof coralreefs.Science292,1532–1535.

Bellwood,D.R.,Hughes,T.P.,Connolly,S.R.,Tanner,J.,2005.Environmentaland geometricconstraintsonIndo-Pacificcoralreefbiodiversity.EcologyLetters8, 643–651.

Bellwood,D.R.,Hughes,T.P.,Hoey,A.S.,2006.Sleepingfunctionalgroupdrives coral-reefrecovery.CurrentBiology16,2434–2439.

Berkes,F.,Hughes,T.P.,Steneck,R.S.,Wilson,J.A.,Bellwood,D.R.,Crona,B.,Folke,C., Gunderson,L.H.,Leslie,H.M.,Norberg,J.,Nystro¨m,M.,Olsson,P.,O¨ sterblom,H., Scheffer, M., Worm, B., 2006. Globalization, roving bandits, and marine resources.Science311,1557–1558.

Birkeland,C.,2004.Ratchetingdownthecoralreefs.BioScience54,1021–1027. Brewer,T.D.,Cinner,J.E.,Green,A.,Pandolfi,J.M.,2009.Thresholdsandmultiple

scaleinteractionofenvironment,resourceuse,andmarketproximityonreef fisheryresourcesintheSolomonIslands.BiologicalConservation142,1797– 1807.

Cardinale,B.J.,Srivastava,D.S.,EmmettDuffy,J.,Wright,J.P.,Downing,A.L., San-karan,M.,Jouseau,C.,2006.Effectsofbiodiversityonthefunctioningoftrophic groupsandecosystems.Nature443,989–992.

Ceccarelli,D.,2007.Modificationofbenthiccommunitiesbyterritorialdamselfish: amulti-speciescomparison.CoralReefs26,853–866.

ChapinIII,F.S.,Zavaleta,E.S.,Eviner,V.T.,Naylor,R.L.,Vitousek,P.M.,Reynolds,H.L., Hooper,D.U.,Lavorel,S.,Sala,O.E.,Hobbie,S.E.,Mack,M.C.,Diaz,S.,2000. Consequencesofchangingbiodiversity.Nature405,234–242.

Chin,W.W.,Newstead,P.R.,1999.Structuralequationmodelinganalysiswithsmall samplesusingpartialleastsquares.In:Hoyle,R.H.(Ed.),StatisticalStrategies forSmallSampleResearch.Sage,ThousandOaks,CA.

Choat,J.H.,Clements,K.D.,Robbins,W.D.,2002.Thetrophicstatusofherbivorous fishesoncoralreefs.1:Dietaryanalyses.MarineBiology140,613–623. Cinner,J.E.,McClanahan,T.R.,2006.Socioeconomicfactorsthatleadtooverfishing

Cinner,J.E.,McClanahan,T.R.,Daw,T.M.,Graham,N.A.J.,Maina,J.,Wilson,S.K., Hughes,T.P.,2009a.Linkingsocialandecologicalsystemstosustaincoralreef fisheries.CurrentBiology19,206–212.

Cinner,J.E.,McClanahan,T.R.,Graham,N.A.J.,Pratchett,M.S.,Wilson,S.K.,Raina, J.-B.,2009b.Gear-basedfisheriesmanagementasapotentialadaptiveresponseto climatechangeandcoralmortality.JournalofAppliedEcology46,724–732. Cinner,J.E.,Bodin,O¨ .,2010.Livelihooddiversificationintropicalcoastal

communi-ties:anetwork-basedapproachtoanalyzing‘LivelihoodLandscapes’.PLoSONE 5,e11999.

Clarke,K.R.,Warwick,R.M.,1999.Thetaxonomicdistinctnessmeasureof biodiver-sity:weightingofsteplengthsbetweenhierarchicallevels.MarineEcology ProgressSeries184,21–29.

Clarke,K.R.,Warwick,R.M.,2001.Afurtherbiodiversityindexapplicabletospecies lists:variationintaxonomicdistinctness.MarineEcologyProgressSeries216, 265–278.

Connolly,S.R.,Bellwood,D.R.,Hughes,T.P.,2003.Indo-pacificbiodiversityofcoral reefs:deviationsfromamid-domainmodel.Ecology84,2178–2190. DeMartini,E.E.,Friedlander,A.M.,Sandin,S.A.,Sala,E.,2008.Differencesin

fish-assemblagestructurebetweenfishedandunfishedatollsinthenorthernLine Islands,centralPacific.MarineEcologyProgressSeries365,199–215. Dulvy,N.K.,Freckleton,R.P.,Polunin,N.V.C.,2004a.Coralreefcascadesandthe

indirecteffectsofpredatorremovalbyexploitation.EcologyLetters7,410–416. Dulvy,N.K.,Polunin,N.V.C.,Mill,A.C.,Graham,N.A.J.,2004b.Sizestructuralchange inlightlyexploitedcoralreeffishcommunities:evidenceforweakindirect effects.CanadianJournalofFisheriesandAquaticSciences61,466–475. Floeter,S.R.,Krohling,W.,Gasparini,J.L.,Ferreira,C.E.,Zalmon,I.R.,2007.Reeffish

communitystructureoncoastalislandsofthesoutheasternBrazil:the influ-enceofexposureandbenthiccover.EnvironmentalBiologyofFishes78,147– 160.

Forester,D.J.,Machlis,G.E.,1996.Modelinghumanfactorsthataffectthelossof biodiversity.ConservationBiology10,1253–1263.

Friedlander,A.M.,Brown,E.K.,Jokiel,P.L.,Smith,W.R.,Rodgers,K.S.,2003.Effectsof habitat,waveexposure,andmarineprotectedareastatusoncoralreeffish assemblagesintheHawaiianarchipelago.CoralReefs22,291–305. Friedlander,A.M.,Parrish,J.D.,1998.Habitatcharacteristicsaffectingfish

assem-blagesonaHawaiiancoralreef.JournalofExperimentalMarineBiologyand Ecology224,1–30.

Froese,R., Pauly,D.,2011. Fishbase. World Wide WebElectronic Publication (accessed07.2011)www.fishbase.org.

Fulton,C.J.,Bellwood,D.R.,2005.Wave-inducedwatermotionandthefunctional implicationsforcoralreeffishassemblages.LimnologyandOceanography50, 255–264.

Geist,H.J.,Lambin,E.F.,2002.Proximatecausesandunderlyingdrivingforcesof tropicaldeforestation.BioScience52,143–150.

Graham,N.A.J., Dulvy,N.K.,Jennings, S.,Polunin, N.V.C.,2005. Size-spectraas indicatorsoftheeffectsoffishingoncoralreeffishassemblages.CoralReefs 24,118–124.

Graham,N.A.J.,Evans,R.D.,Russ,G.R.,2003.Theeffectsofmarinereserveprotection onthetrophicrelationshipsofreeffishesontheGreatBarrierReef. Environ-mentalConservation30,200–208.

Graham,N.A.J., McClanahan, T.R.,MacNeil, M.A.,Wilson, S.K.,Polunin, N.V.C., Jennings,S.,Chabanet,P.,Clark,S.,Spalding,M.D.,Letourneur,Y.,Bigot,L., Galzin, R.,o¨hman,M.C., Garpe,K.C., Edwards,A.J.,Sheppard,C.R.C.,2008. Climatewarming,marineprotectedareasandtheocean-scaleintegrityofcoral reefecosystems.PLoSONE3,e3039.

Graham,N.A.J.,Wilson,S.K.,Jennings,S.,Polunin,N.V.C.,Bijoux,J.P.,Robinson,J., 2006.Dynamicfragilityofoceaniccoralreefecosystems.Proceedingsofthe NationalAcademyofSciences103,8425–8429.

Graham,N.A.J.,Wilson,S.K.,Jennings,S.,Polunin,N.V.C.,Robinson,J.A.N.,Bijoux,J.P., Daw,T.M.,2007.Lageffectsintheimpactsofmasscoralbleachingoncoralreef fish,fisheries,andecosystems.ConservationBiology21,1291–1300. Gratwicke,B.,Speight,M.R.,2005.Therelationshipbetweenfishspeciesrichness,

abundanceand habitatcomplexity ina rangeof shallowtropical marine habitats.JournalofFishBiology66,650–667.

Green,A.,Bellwood,D.R.,2009.MonitoringFunctionalGroupsofHerbivorousReef FishesasIndicatorsofCoralReefResilience–APracticalGuideforCoralReef ManagersintheAsiaPacificRegion.IUCN,Gland,Switzerland.

Green,A.,Lokani,P.,Atu,W.,Ramohia,P.,Thomas,P.,Almany,J.,2006a.Solomon IslandsMarineAssessment:TechnicalReportofSurveyConductedMay13to June17,2004.TNCPacificIslandCountriesReportNo.1/06..

Green,A.,Ramohia,P.,Ginigele,M.,Leve,T.,2006b.Fisheriesresources:coralreef fishes.In:Green,A.,Lokani,P.,Atu,W.,Ramohia,P.,Thomas,P.,Almany,J. (Eds.),SolomonIslandsMarineAssessment:TechnicalReportofSurvey Con-ductedMay13toJune17,2004.TNCPacificIslandCountriesReportNo.1/06. Hoey,A.S.,Bellwood,D.R.,2009.Limitedfunctionalredundancyinahighdiversity system:singlespeciesdominateskeyecologicalprocessoncoralreefs. Eco-systems12,1316–1328.

Hooper,D.U.,Chapin,F.S.,Ewel,J.J.,Hector,A.,Inchausti,P.,Lavorel,S.,Lawton,J.H., Lodge,D.M.,Loreau,M.,Naeem,S.,Schmid,B.,Seta¨la¨,H.,Symstad,A.J., Van-dermeer,J.,Wardle,D.A.,2005.Effectsofbiodiversityonecosystem function-ing:aconsensusofcurrentknowledge.EcologicalMonographs75,3–35. Hughes,A.,2006. BenthicCommunities,Solomon IslandsMarineAssessment:

TechnicalReportofSurveyConductedMay13toJune17,2004.TNCPacific IslandCountriesReportNo.1/06..

Hughes,T.P.,1994.Catastrophes,phaseshifts,andlarge-scaledegradationofa Caribbeancoralreef.Science265,1547–1551.

Hughes,T.P.,Rodrigues,M.J.,Bellwood,D.R.,Ceccarelli,D.,Hoegh-Guldberg,O., McCook,L.,Moltschaniwskyj,N.,Pratchett,M.S.,Steneck,R.S.,Willis,B.,2007. Phaseshifts,herbivory,andtheresilienceofcoralreefstoclimatechange. CurrentBiology17,360–365.

Ives,A.R.,Carpenter,S.R.,2007.Stabilityanddiversityofecosystems.Science317, 58–62.

Jennings,S.,Grandcourt,E.M.,Polunin,N.V.C.,1995.Theeffectsoffishingonthe diversity,biomassandtrophicstructureofSeychelles’reeffishcommunities. CoralReefs14,225–235.

Jennings,S.,Polunin,N.V.C.,1997.Impactsofpredatordepletionbyfishingonthe biomassanddiversityofnon-targetreeffishcommunities.CoralReefs16,71–82. Jennings,S.,Polunin,N.V.C.,1996.Effectsoffishingeffortandcatchrateuponthe structure andbiomassofFijian reeffishcommunities.JournalofApplied Ecology33,400–412.

Jones,G.P.,Planes,S.,Thorrold,S.R.,2005.Coralreeffishlarvaesettleclosetohome. CurrentBiology15,1314–1318.

Kock,N.,2010.WarpPLSß1.0UserManual.ScriptWarpSystems,Laredo,TX. Kramer,D.B.,Urquhart,G.,Schmitt,K.,2009.Globalizationandtheconnectionof

remotecommunities:areviewofhouseholdeffectsandtheirbiodiversity implications.EcologicalEconomics68,2897–2909.

Luckhurst,B.E.,Luckhurst,K.,1978.Analysisoftheinfluenceofsubstratevariables oncoralreeffishcommunities.MarineBiology49,317–323.

Maestre,F.T.,Quero,J.L.,Gotelli,N.J.,Escudero,A.,Ochoa,V.,Delgado-Baquerizo,M., Garcı´a-Go´mez,M.,Bowker,M.A.,Soliveres,S.,Escolar,C.,Garcı´a-Palacios,P., Berdugo,M.,Valencia,E.,Gozalo,B.,Gallardo,A.,Aguilera,L.,Arredondo,T., Blones,J.,Boeken,B.,Bran,D.,Conceic¸a˜o,A.A.,Cabrera,O.,Chaieb,M.,Derak,M., Eldridge,D.J.,Espinosa,C.I.,Florentino,A.,Gaita´n,J.,Gatica,M.G.,Ghiloufi,W., Go´mez-Gonza´lez,S.,Gutie´rrez,J.R.,Herna´ndez,R.M.,Huang,X., Huber-Sann-wald,E.,Jankju,M.,Miriti,M.,Monerris,J.,Mau,R.L.,Morici,E.,Naseri,K., Ospina,A.,Polo,V.,Prina,A.,Pucheta,E.,Ramı´rez-Collantes,D.A.,Roma˜o,R., Tighe,M.,Torres-Dı´az,C.,Val,J.,Veiga,J.P.,Wang,D.,Zaady,E.,2012.Plant speciesrichnessandecosystemmultifunctionalityinglobaldrylands.Science 335,214–218.

McCann,K.S.,2000.Thediversity-stabilitydebate.Nature405,228–233. McClanahan,T.R.,Maina,J.M.,Muthiga,N.A.,2011.Associationsbetweenclimate

stressandcoralreefdiversityinthewesternIndianOcean.GlobalChange Biology17,2023–2032.

McKinney,L.A.,Kick,E.L.,Fulkerson,G.M.,2010.Worldsystem,anthropogenic,and ecologicalthreatstobirdandmammalspecies:astructuralequationanalysisof biodiversityloss.OrganizationandEnvironment23,3–31.

MollesJr.,M.C.,1978.Fishspeciesdiversityonmodelandnaturalreefpatches: experimentalinsularbiogeography.EcologicalMonographs48,289–305. Mora, C.,2008. Aclear human footprintinthe coralreefs oftheCaribbean.

ProceedingsoftheRoyalSocietyB:BiologicalSciences275,767–773. Mora,C.,Aburto-Oropeza,O.,AyalaBocos,A.,Ayotte,P.M.,Banks,S.,Bauman,A.G.,

Beger,M.,Bessudo,S.,Booth,D.J.,Brokovich,E.,Brooks,A.,Chabanet,P.,Cinner, J.E.,Corte´s,J.,Cruz-Motta,J.J.,CupulMagan˜ a,A.,DeMartini,E.E.,Edgar,G.J., Feary,D.A.,Ferse,S.C.A.,Friedlander,A.M.,Gaston,K.J.,Gough,C.,Graham,N.A.J., Green,A.,Guzman,H.,Hardt,M.,Kulbicki,M.,Letourneur,Y.,Lo´pezPe´rez,A., Loreau,M.,Loya,Y.,Martinez,C.,Mascaren˜ as-Osorio,I.,Morove,T.,Nadon, M.-O.,Nakamura,Y.,Paredes,G.,Polunin,N.V.C.,Pratchett,M.S.,ReyesBonilla,H.c., Rivera,F.,Sala,E.,Sandin,S.A.,Soler,G.,Stuart-Smith,R.,Tessier,E.,Tittensor, D.P.,Tupper,M.,Usseglio,P.,Vigliola,L.,Wantiez,L.,Williams,I.,Wilson,S.K., Zapata,F.A.,2011.Globalhumanfootprintonthelinkagebetweenbiodiversity andecosystemfunctioninginreeffishes.PLoSBiology9,e1000606. Mora, C.,Chittaro,P.M.,Sale,P.F.,Kritzer,J.P.,Ludsin,S.A.,2003.Patternsand

processesinreeffishdiversity.Nature421,933–936.

Mora,C.,Robertson,D.R.,2005.Causesoflatitudinalgradientsinspeciesrichness:a testwithfishesofthetropicalEasternPacific.Ecology86,1771–1782. Muller,D.,Zeller,M.,2002.LandusedynamicsinthecentralhighlandsofVietnam:

aspatialmodelcombiningvillagesurveydatawithsatelliteimagery interpre-tation.AgriculturalEconomics27,333–354.

Mumby,P.J.,2006.Theimpactofexploitinggrazers(Scaridae)onthedynamicsof Caribbeancoralreefs.EcologicalApplications16,747–769.

Munday,P.L.,Jones,G.P.,1998.Theecologicalimplicationsofsmallbodysizeamong coral-reeffishes.OceanographyandMarineBiologyAnnualReview36,373–411. Naeem, S., 1998. Speciesredundancy and ecosystem reliability. Conservation

Biology12,39–45.

Newton,K.,Cote,I.M.,Pilling,G.M.,Jennings,S.,Dulvy,N.K.,2007.Currentand futuresustainabilityofislandcoralreeffisheries.CurrentBiology17,655–658. Nystro¨m,M.,Folke,C.,Moberg,F.,2000.Coralreefdisturbanceandresilienceina human-dominatedenvironment.TrendsinEcologyandEvolution15,413–417. O¨ hman,M.C.,Rajasuriya,A.,1998.Relationshipsbetweenhabitatstructureandfish communitiesoncoralandsandstonereefs.EnvironmentalBiologyofFishes53, 19–31.

Pandolfi,J.M.,Bradbury,R.H.,Sala,E., Hughes,T.P., Bjorndal,K.A.,Cooke,R.G., McArdle,D.,McClenachan,L.,Newman,M.J.H.,Paredes,G.,Warner,R.R., Jack-son, J.B.C.,2003. Globaltrajectoriesofthelong-termdeclineofcoralreef ecosystems.Science301,955–958.

Pastorok,R.A.,Bilyard,G.R.,1985.Effectsofsewagepollutiononcoral-reef com-munities.MarineEcologyProgressSeries21,175–189.

Pauly,D.,Christensen,V.,Dalsgaard,J.,Froese,R.,Torres,F.,1998.Fishingdown marinefoodwebs.Science279,860–863.

Reinartz,W.,Haenlein,M.,Henseler,J.,2009.Anempirical comparisonofthe efficacyofcovariance-basedandvariance-basedSEM. InternationalJournal ofResearchinMarketing26,332–344.

Roberts,C.M.,Ormond,R.F.G.,1987.Habitatcomplexityandcoralreeffishdiversity andabundanceonRedSeafringingreefs.MarineEcologyProgressSeries41,1–8. Rogers,S.I.,Clarke,K.R.,Reynolds,J.D.,1999.Thetaxonomicdistinctnessofcoastal bottom-dwellingfishcommunitiesoftheNorth-eastAtlantic.JournalofAnimal Ecology68,769–782.

Russ,G.R.,2003.Grazerbiomasscorrelatesmorestronglywithproductionthan withbiomassofalgalturfsonacoralreef.CoralReefs22,63–67.

Sandin,S.A.,Smith,J.E.,DeMartini,E.E.,Dinsdale,E.A.,Donner,S.D.,Friedlander, A.M.,Konotchick,T.,Malay,M.,Maragos,J.E.,Obura,D.,Pantos,O.,Paulay,G., Richie,M.,Rohwer,F.,Schroeder,R.E.,Walsh,S.,Jackson,J.B.C.,Knowlton,N., Sala,E.,2008.Baselinesanddegradationofcoralreefsinthenorthernline islands.PLoSONE3,e1548.

Shandra,J.M.,Leckband,C.,McKinney,L.A.,London,B.,2009.Ecologicallyunequal exchange,worldpolity,andbiodiversityloss.InternationalJournalof Compar-ativeSociology50,285–310.

Stallings,C.D.,2009.Fishery-independentdatarevealnegativeeffectofhuman populationdensityonCaribbeanpredatoryfishcommunities.PLoSONE4, e5333.

Tilman,D.,Reich,P.B.,Knops,J.M.H.,2006.Biodiversityandecosystemstabilityina decade-longgrasslandexperiment.Nature441,629–632.

Williams,I.D.,Richards,B.L.,Sandin,S.A.,Baum,J.K.,Schroeder,R.E.,Nadon,M.O., Zgliczynski, B.,Craig, P.,McIlwain,J.L.,Brainard,R.E.,2011. Differencesin reef fish assemblages between populated and remote reefs spanning multiplearchipelagosacrossthecentralandWesternPacific.JournalofMarine Biology14.

Williams,I.D.,Walsh,W.J.,Schroeder,R.E.,Friedlander,A.M.,Richards,B.L., Sta-moulis,K.A.,2008.AssessingtheimportanceoffishingimpactsonHawaiian coralreeffishassemblagesalongregional-scalehumanpopulationgradients. EnvironmentalConservation35,261–272.

Wilson,S.,Bellwood,D.R.,1997.Crypticdietarycomponentsofterritorial damsel-fishes(Pomacentridae,Labroidei).MarineEcologyProgressSeries153,299– 310.

Wilson,S.K.,Fisher,R.,Pratchett,M.S.,Graham,N.A.J.,Dulvy,N.K.,Turner,R.A., Cakacaka,A.,Polunin, N.V.C., Rushton,S.P.,2008. Exploitationandhabitat degradationasagentsofchangewithincoralreeffishcommunities.Global ChangeBiology14,2796–2809.