Content uploaded by Dave Kendal
Author content
All content in this area was uploaded by Dave Kendal on Jun 17, 2017
Content may be subject to copyright.
Greenspacecontextandvegetationcomplexityshapepeople’spreferencesforurbanpublicparksand
residentialgardens
VirginiaHarris1,DaveKendal*1,2,AmyHahs2,CaraghGThrelfall1
1UniversityofMelbourne,DepartmentofEcosystemandForestSciences,500YarraBoulevard,Richmond
2AustralianResearchCentreforUrbanEcology,RoyalBotanicGardensVictoria,c/oSchoolofBiosciences,
TheUniversityofMelbourne,Parkville3010,Australia
*Correspondingauthordkendal@unimelb.edu.au
Abstract
Thepublic’slandscapepreferencesshapedecisionmakinganddrivetheecologicaloutcomesofurban
landscapes.Existingpreferencestudieshavemainlybeenconductedwithinasinglegreenspacecontext.Yet
differentkindsofgreenspaceafforddifferentkindsofusesandexperiencesanditislikelythisinfluences
preferences.Weinvestigatehowpeople’slandscapepreferencesareshapedbythegreenspacecontext
(publicparkvsprivateresidentialgardenlandscapes)andbyphysicalfeaturessuchasvegetationcomplexity.
ApostalquestionnairewassenttohouseholdsnearsevenurbanparksinMelbourne,Australia.Results
showedthatlandscapesweregroupedintofourcategoriesbasedonpatternsofpreferenceresponse.
Landscapeswithmoderatevegetationcomplexitywereplacedinseparatecategoriesdistinguishedbygreen
spacecontext(parksvsgardens),whileverysimpleandverycomplexlandscapeswereplacedindifferent
categoriesirrespectiveofgreenspacecontext.Surprisingly,densevegetationwashighlypreferredby
respondents.Asareasofdensevegetationalsoprovidecomplexhabitatsforwildlife,thishighlightsthe
possibilityofdevelopingpoliciesanddesigninglandscapesthatcanbenefitbothpeopleandnature.
Introduction
Urbanvegetationinpublicparksandprivategardensprovidesmanybenefitstopeople(Tzoulasetal.,2007).
Exposuretogreener,morenaturalenvironmentscanhavepositiveeffectsonpsychologicalwellbeing
(Kaplan,1995;vandenBergetal.,2003;Fulleretal.,2007;Lucketal.,2011),longevity(Takanoetal.,2002),
communityattachment(KimandKaplan,2004),physicalhealth(Pereiraetal.,2012)andmorbidity(Maaset
al.,2009).Atthesametime,urbangreenspacesalsoplayakeyroleintheecologicalfunctioningofcitiesby
maintainingbiodiversity(KurzandBaudains,2010;Thompsonetal.,2003)andprovidingecosystemgoods
andservices,suchasairandwaterpurification,pollinationandinsectregulation(BolundandHunhammar,
1999;Chiesura,2004).Humanperceptionofvegetationisimportantasitcanmediatesomeofthesepositive
effects(e.g.vandenBergetal.,2003),influenceperceivedappropriatenessofgreenspaceforcertain
activities(Bjerkeetal2006)andinfluencethelandmanagementdecisionspeoplemake(e.g.Kendaletal.,
2012a)thatshapeecologicalfunctioningandultimately,theongoingexistenceofthelandscape(Gobsteret
al.,2007).Understandingtherelationshipsbetweenthesocialandecologicalrolesofvegetationingreen
spacesisimportantforurbanplannersanddesignersseekingtoprovidespacesthatmeettheneedsofthe
communityandhavepositiveecologicaloutcomes.
Thecharacterofurbangreenspacesisdeterminedbytheactionsandinteractionsofnaturalandhuman
influences(CouncilofEurope,2000;Kendaletal.,2012b;Zube,1987).Whiletheeffectofnaturalinfluences
suchastemperatureandrainfallonvegetationarereasonablywellunderstood,socialinfluencesareless
wellstudied.Landscapepreferenceisthemeasurementofhowmuchpeople‘like’theappearanceofa
landscape(Scherer,2005).Itisausefulframeworkforinvestigatingthehumanrelationshipwiththesesocial
andecologicalsystems,andhasbeenwidelyusedtostudyrelationshipsinmanydifferentkindsofurban
landscapes.Forexample,peoplehavebeenshowntoconsistentlyprefernaturallandscapestobuilt
environments(KaplanandKaplan,1989),andpark‐likelandscapes(thatcontainspecificfeaturessuchas
scatteredtreeswithminimalunderstorey)areoftenhighlypreferred(Bjerkeetal.,2006;Kaltenbornand
Bjerke,2002;Ulrich,1993).Preferenceforlandscapehasbeenexplainedusingbothevolutionary(Kaplan
2
andKaplan,1989;Ulrich,1993),andculturaltheories(Nassaueretal.,2009;VandenBergandKoole,2006).
EvolutionarytheoriessuggestthatpreferenceisanevolvedresponsethataidedsurvivalintheAfrican
savannahlandscapeswhereearlyhominidslived(e.g.Appleton,1975).Culturaltheoriessuggestthat
preferenceislearnedandthatculturaltraditions(e.g.Gobster,1995),demographicvariables(suchasage,
education,gender),expertise(Hoffmanetal.,2012),personalexperienceorcognitivefactors(suchasvalues
andbeliefs)canallinfluenceanindividual’spreferencefordifferentlandscapes(Howley,2011;Lyons,1983;
VandenBergandKoole,2006;YangandBrown,1992;Yu,1995).Recently,thereisgrowingconsensusand
evidencethatsuggestsbothmodesoperateindetermininglandscapepreference(Bourassa,1990;Gobster,
1999;Kendaletal.,2012a).
Preferenceforlandscapesisdeterminedinpartbythecontextandspatialarrangementoffeatureswithina
landscape(Odeetal.,2009).Studiesexploringtheinfluenceofvegetationfeatureshaveshownthat
elementssuchastreesandshrubs(JimandChen,2006;KurzandBaudains,2010;Schroeder,1987),the
neatnessofvegetation(vandenBergandvanWinsum‐Westra,2010),andvegetationcharacteristicssuchas
colour(KaufmanandLohr,2004;Kendaletal.,2012a)andleaftexture(WilliamsandCary,2002)canshape
people’spreferences.Thediversityofvegetationpresentinalandscapecanbeaccuratelyperceivedby
people(Fulleretal.,2007;Qiuetal.,2013).However,thereisnotalinearrelationshipbetweendiversityand
preference,andmoderatelydiverselandscapescanbemorepreferredthanhighlydiverselandscapes(Qiuet
al.,2013).Preferencefordensevegetationalsovariesbyexpertise,forexample,landscapedesignerscan
havequitedifferentpreferencesthanthegeneralpublic(Hoffmanetal.,2012).Severalstudieshave
exploredpreferenceforlandscapefeatureswithinaspecificcontext.Inparks,featuressuchastrees,shrubs
andwaterarepreferred(Bjerkeetal.,2006;Schroeder,1987)whiledensevegetationnearpathsisnot
preferred(Jorgensenetal.,2002).Inresidential/allotmentgardens,featuressuchasthecomplexityand
structureofvegetationinfluencepreference,butthedirectionoftheeffectisinfluencedbyaspectsof
peoplespersonality(vandenBerg&vanWinsumWestra,2010)andtheirattitudestowardsnative
vegetation(KurzandBaudain,2010).
Togetherparksandgardenscontainmostofthevegetationthatexistswithinmanycities(Lorametal.,
2007).Urbanpublicparksareprimarilymanagedbygovernmentagencies,arelarger,andexistforpublic
use,whileresidentialgardensaretypicallysmaller,existprimarilyforprivateuseandaremoreintensively
managedbyindividualhouseholdersorcontractors(Goddardetal.,2010).Aspublicgreenspaces,urban
parksprovidespaceforresidentstoundertakephysicalactivitythroughactiverecreation(Groenewegenet
al.,2006;Leslieetal.,2010),andenhancecommunitycohesionandinteractionbetweenresidents
(ArnbergerandEder,2012;Homeetal.,2010;Westphal,2003).Incontrast,residentialgardensare
generallynotopentothepublic,andarerecognisedasplacesforcreativeexpression(DunnettandQasim,
2000;Kirkpatricketal.,2009),forimprovingpropertyandneighbourhoodappearance(LarsenandHarlan,
2006),forphysicalactivitythroughgardening(Bhatti,2006)andforrelaxation(Headetal.,2004).Cultural
preferencetheoriessuggestthatthesocialandphysicalcontextofthesedifferenttypesofgreenspacemay
leadtodifferencesinpeople’spreferences(CoolenandMeesters,2011;Gibson,1979;Hadavietal.,2015).
Thissuggeststhatitislikelythatpeople’spreferencesvarybetweenpublicparksandresidentialgardens,
andthesedifferencescouldhaveimportantimplicationsformanagementofdifferentkindsoflandscapes.A
fewstudieshavesuggestedthatlandscapetypecouldbeanimportantfactorinfluencingpreference(Bulut
andYilmaz,2008;SevenantandAntrop2009).Thishasbeenexploredinsomegeneralcontextssuchaswild
vsmorehumanlandscapes(deGrootandVandenBorn2003),vandenBergandKoole,2006),andithas
beenshownthatlandscapeelementsareperceiveddifferentlyindifferentlandscapecontexts(Filovaetal.,
2015).
Thisstudyinvestigateshowpeople’spreferencesvaryacrosspublicurbanparkandprivateresidential
gardengreenspaces,withvaryinglevelsofstructuralcomplexityofvegetation.Usingimagesofresidential
gardensandpublicparks,weinvestigatethehypothesisthatpeople’spreferenceswillberelatedtoboth
greenspacecontexts,andtophysicalfeatures(suchasthestructuralcomplexityofvegetation)withinthose
3
contexts.Specifically,weexplorewhetherpublicparksandprivategardensarepreferreddifferentlyby
people,andwhetherpreferenceforstructurallysimplevegetation(lawns)andmorestructurallycomplex
vegetation(trees,shrubsandflowers)isconsistentacrossparksandgardens.Understandingtherelationship
betweengreenspacecontexts,landscapepreference,andphysicalsite‐levelfeatureswillhelpurban
plannersanddesignerstobettermanagepotentiallyconflictingsocialandecologicalfunctionsandusesof
urbangreenspace.
Methods
Studylocation
Mosturbanlandscapepreferencestudieshavebeenconductedinthenorthernhemisphere,particularlyin
EuropeandNorthAmericaandmorestudiesarerequiredinotherplacestodevelopamorecomprehensive
understandingoflandscapepreferenceacrossdifferentsocialandphysicalenvironments.Herewe
conductedthestudyinMelbourne,thelargestcityandcapitalofthestateofVictoriainsouth‐eastern
Australia(37.7833°S,144.9667°E),withapopulationofalmost4millionpeople(AustralianBureauof
Statistics,2013).FourmajorbioregionsoccurwithinthegreaterMelbournearea,andthisstudyfocuseson
asinglebioregioninthecity’ssouth‐east.Thesouth‐easternregionofMelbourneincludesseveralolder,
establishedsuburbsaswellassuburbsthathavebeenrecentlydevelopedasMelbourne’spopulation
expands.AsseenthroughoutAustralia,themajorityofhousinginboththeolderandnewersuburbs
consistsoflowdensitydetachedhouseswithbothfrontandreargardens(Kellett,2011;Kirkpatricketal.,
2011).Gardensinsouth‐eastMelbourne’soldersuburbsareoftenlargerandmorelikelytocontaintrees,
whilethoseinnewersuburbsaregenerallysmaller,lessestablishedandincludeshrubsandgroundcovers
ratherthanlargerplantforms(GhoshandHead,2009).
Siteselectionandsurveymethodology
Sevenpublicneighbourhoodparksmanagedbytherelevantlocalgovernmentauthority,andfourteen
nearbyprivateresidentialgardens(twoperpark)managedbyresidentswerepurposivelychoseninsouth‐
easternMelbournetoensurerepresentationofarangeofgreenspacefeatures,includingavaryingamount
oftrees,understoreyshrubsandflowers,andlawn.Multiplephotographsweretakenwithineachpublic
parkandprivategardentocapturetherangeofformspresent.Thephotographsweretakeninovercast
conditionsandadjustedusingphotographicsoftware(GIMPDevelopmentTeam,2004)sothattheywere
consistentinbrightnessandcolour.Thefinalsetofimageswereselectedtoprovidearepresentative
sampleofthedifferentpublicparkandprivategardenlandscapesprevalentthroughoutsouth‐eastern
Melbourne,andthosephotographscontainingvehiclesandwaterbodieswererejectedduringthisselection
process,aspreviousstudieshaveshownthatthesearepowerfuldriversofpreference(Howley,2011;
Kaplan,2001;KaplanandKaplan,1989;YangandBrown,1992).
Aquantitativepostalquestionnairewassentto755randomlyselectedhouseholdsfromwithinthecensus
collectiondistrictscontainingthesevenparks(AustralianBureauofStatistics,2006).Respondentswere
givencolourphotographsrepresenting26differentfrontgardenandparklandscapes,andaskedhowmuch
theylikedeachlandscapeusinga7‐pointLikertscale(from‘Don’tlikeitatall’to‘Likeitverymuch’).This
methodhasbeenwidelyusedinlandscapepreferencestudies(e.g.DanielandMeitner,2001;Kaplanand
Kaplan,1989;Trentetal.,1987;Ulrich,1977).FourimagesperA4pagewereprovidedtoresidentstoallow
foreasierviewing,andtwodifferentrandomsequencesofphotographswereusedtoreduceeffects
associatedwithorderofpresentation(Bryman,2008).
Thephysicalattributesofeachlandscapeweremeasuredasrelativecoverofthreevegetationlayers(lawn,
shrubsandflowers,andtrees)andthepresenceofrecreationalfeatures.Quantificationofvegetationcover
wasachievedbyplacinga10x10gridovereachlandscapeimageandcountingthenumberofgridcells
representedbylawn,shrubsandflowers,andtrees(Tveit,2009).Treesweredefinedastaller,canopy
formingplantstypicallygrowinggreaterthanfivemetrestall;mownlawn;andshrubsandflowerswere
consideredtobeanyplantthatwasnotlawnoratree(Kirkpatricketal.,2011).
4
StatisticalAnalysis
SurveydatawasanalysedusingSPSSStatisticsv20.0(IBM,2012)andRv3.2.3(RDevelopmentCoreTeam,
2007).TherewasalowrateofmissingLikertscores(1.4%ofpreferenceratings)andthesedatawere
replacedwithrespondentmeans(TabachnickandFidell,2001).Overallrelationshipsbetweenpreference
andgreenspacetype(parksorgardens)andtheproportionoflawnandotherplants(shrubs,treesand
flowers)wasassessedusingpolynomialregressions.Preferenceandproportionofotherplantswere
normallydistributed,andwhilelawnwasnot,itwastransformedbyaddingpolynomialtermsandthe
assumptionsoftheregressionwereconfirmedbycheckingthedistributionoftheresiduals(Tabachnickand
Fidell,2001).Theinfluenceofdemographicfactors(gender,ageandeducationlevel)onpreferencewasalso
testingusingthemultiplecomparisonTukey’sHSDtest.
Followingthecategoryidentificationmethod(CIM)developedbyKaplan&Kaplan(1989)andwidelyusedin
landscapepreferenceresearch(e.g.Williams,2002),PrincipalComponentsAnalysis(PCA)withVarimax
rotation,conductedusingtheprincipalfunctioninthepsychpackageinR,wasusedtogrouplandscapes
withsimilarpreferencescores,bysummarisingpatternsofcorrelationsamongtheseobservedvariablesinto
asmallersetofcomponents.AsordinaldatafromLikertscaleswasusedintheanalysis,thePCAwas
performedonasimilaritymatrixgeneratedusinganon‐parametricSpearman’scorrelationonpreference
scoresforthe26landscapesinthestudy(TabachnickandFidell,2001).Ascreeplotwasgeneratedto
determinethenumberofcomponentstoextractforthePCA.AllphotographswereassignedtothePCA
componenttheyloadedmosthighlyon,andeachcomponentwasgivenadescriptivenamebasedonthe
similarityofthephotographsassignedtothatcomponent.Asthecomponentsweredistinguishedby
patternsofpreferencevariability,averagepreferencescoreswerecalculatedforeachcomponentbasedon
themeanLikertscoreacrossallrespondents,anddifferencesinrespondentpreferencebetween
componentswastestedusingTukey’sHSDtestasthepreferencedataweredistributednormally.
Inordertodeterminewhetherthestructuralcomplexityofvegetationinfluencedthecategorisationof
photosintothedifferentprincipalcomponents,therelationshipbetweenpreferenceandtheproportionof
simplevegetation(lawn)andmorestructurallycomplexvegetation(shrubs,flowersandtrees)was
examinedforeachcomponent.Thesevariablesdidnotmeettheassumptionsofnormalityorhomogeneous
variance,andhencenon‐parametricKruskal‐Wallisonewayanalysisofvariance(ANOVA)testswereused
andpairwisecomparisonswereemployedtotestthedifferenceinphysicalattributesbetweenanytwo
components(SokalandRolhf,1995).
Results
Intotal,21%(159of755)ofsurveyswerereturned,aresponseratewhichiscomparabletoorhigherthan
recentsimilarstudiesinthisgeographicregion(e.g.Ives&Kendal,2013,Shawetal.,2013).The
respondentswerebroadlyrepresentativeofsouth‐easternMelbourne,withover87%ofallrespondents
owningtheirownhouseandthemajorityofhouseholdsconsistingofeitheracouplewithchildren(29.6%),a
couplewithnochildren(28.9%)orasingleperson(23.3%).Abroadrangeofeducationlevels,
socioeconomicstatusandagegroupswererepresented(Table1).Incomparisontothepopulationaverage
forthecensuscollectiondistrictscoveredbythissurvey(CityofCasey,FrankstonCityCouncilandBayside
CityCouncil),therespondentswereslightlyolder,hadahigherfemalerepresentation,weremorelikelyown
theirhouseandlesslikelytospeakalanguageotherthanEnglishathome(Table1).
LandscapePreferenceComponents
ThePCAofpreferencescoresidentifiedfourcomponentsthatexplained56.6%ofthevariation(Table2,
supplementarymaterial).Imagesofgardensandparkswereseparatedinthefirsttwocomponents,and
combinedintheothercomponents.Component1(labelledEnglishLandscapes)accountedfor15.8%ofthe
variationandcontainedphotosofpublicparklandscapeswithscatteredtrees,lawnandcontainedallphotos
showingrecreationalfacilities.Component2(labelledSuburbanGardens)consistedofphotographsoflow
5
vegetation(shrubsandgroundcovers)andlawninprivateresidentialgardensandexplained14.9%ofthe
variation.PhotographsloadingonComponent3(labelledOpen)explained13.6%ofthevariationand
includedgrassylandscapesinbothpublicparksandprivategardenswithlittleothervegetation,whilethose
inComponent4(labelledDense)containedlandscapesinpublicparksandprivategardenswithahigher
proportionofshrubsandflowers,explaining12.2%ofthevariation(Table2).
Thereweresignificant(Tukey’sHSDp<0.05)differencesinpreferencescoresforthefourdifferent
components.EnglishLandscapeswerethemostpreferred(meanLikertscore5.3,Figure1),followedby
Denselandscapes(meanLikertscore4.5)andSuburbanGardens(meanLikertscore3.9).Openlandscapes
weretheleastpreferred(meanLikertscore3.2).Therewerenosignificantdifferencesinpreferencebased
ondemographicfactors(Tukey’sHSDp>0.05).
PhysicalFeatureswithinLandscapes
Significantdifferenceswerefoundintheproportionofstructurallysimple(lawn)andstructurallycomplex
(trees,shrubsandflowers)vegetationinthedifferentpreferencecomponents(Fig2).Non‐parametric
pairwisetestsindicatedthattherewassignificantlymorelawninOpenlandscapesthanDenselandscapes
(K=18.7,p<0.001,Figure2a).Likewise,therewassignificantlylessshrubandflowercoverinOpen
landscapescomparedtobothDenselandscapes(K=‐15.0,p<0.01,Figure2b)andSuburbanGarden
landscapes(K=15.8,p<0.001,Figure2b).EnglishLandscapescontainedsignificantlymoretreesthan
SuburbanGardens(K=14.9,p<0.001,Figure2c)andOpenorDenselandscapes.
Whileoveralllevelsofpreferencewerehigherinparks(p<0.001),therelationshipbetweenthestructural
complexityofvegetationandpreferencewassimilarinbothparksandgardens(Figure3,Table3).Theeffect
ofstructurallysimplevegetation(lawn)onpreferencewasbestexplainedbyincludingaquadraticterm;
preferenceincreasedslightlyuntiltheproportionoflawnreachedabout20%,afterwhichpreference
declinedsharply(p<0.001forthelinearandquadraticterms).Theeffectofstructurallycomplexvegetation
(e.g.trees,shrubsandflowers)onpreferencewasalsobestexplainedbyincludingaquadraticterm;
preferenceincreaseduntiltheproportionoftheseplantsreachedabout70%,afterwhichitremainedsteady
(p<0.001forthelinearandquadraticterms).Themostpreferredcomponent,EnglishLandscapes,contained
thehighestpercentageoftrees(46%)andsecondhighestproportionoflawn(31%)alongwithallthe
landscapesthatcontainedrecreationalinfrastructure.Bycontrast,theleastpreferredcomponent,Open
landscapes,containedthehighestpercentageoflawn(58%),norecreationalfacilitiesandfewtrees(15%).
Discussion
Overallwefoundthattherespondent’spreferenceswereinfluencedbybothgreenspacecontext(public
parkvsprivateresidentialgarden)andbythestructuralcomplexityofvegetation.Privateresidential
gardensandpublicparksweredistinctlyseparatedacrossthefirsttwooffourprincipalcomponents,and
overalllevelsofpreferencewerehigherforparks.However,amixofpublicparksandprivateresidential
gardensloadedontheremainingtwoprincipalcomponents,whichdifferedbyhavingveryhighorverylow
levelsofstructuralcomplexityinthevisiblevegetation.Thestructuralcomplexityofvegetationwashighly
associatedwithresidents’preferencesacrossalllandscapes,withnon‐lawnvegetationexertingagenerally
positiveinfluenceonpreferenceandlawnexertingagenerallynegativeinfluence.Toourknowledge,oursis
thefirststudytodemonstratethatdifferenttypesofurbangreenspacesplayaroleinshapingpreferences
forlandscapes,andthisknowledgecanbeusefullyappliedtothedesignandmanagementofvegetationin
urbangreenspaces.
Itisclearthatgreenspacecontextinfluencedparticipant’sresponsestopublicparkandprivateresidential
gardenlandscapestosomeextent,asthisdistinctioncomprisedthefirsttwocomponentsofthePCA
explaining30.7%ofthevariabilityinthedata(Table2).Whilethereweresomephysicaldifferencesbetween
landscapesinthetwogreenspacecontexts(e.g.parkshadmoretrees),thelandscapesloadingonthe
EnglishLandscapeandSuburbanGardencomponentswerenotstrictlybasedonphysicalfeatures(e.g.there
6
wereanumberofSuburbanGardenlandscapesthatalsocontainedbothtreesandlawn).Thisshowsthat
respondentspreferencesdistinguishedbetweenpublicparksandprivateresidentialgardenlandscapes.
Thesepatternsmaybeexplainedbythedifferentaffordancesavailableinthesegreenspacecontexts.
Affordancesaretherelationalpossibilitiesofferedbyalandscapeorfeatureswithinit,asperceivedbya
person(ormorecorrectly,ananimal:Gibson,1979).Forexample,apathaffordswalking,largetreesafford
shadeandfloweringshrubsaffordbeauty(andperhapsasenseofprideforthegardener).Todate,therole
ofaffordanceshasbeenlargelyabsentfromstudiesofurbangreenspacepreference(Hadavietal.,2015).
Differentaffordancesmaybedrivingpreferenceforpublicparksandprivategardens.Forexample,itiswell
understoodthatpathswithinparklandscapesaffordmovementthroughthelandscape,andtheprospectof
surveillanceleadstoenhancedfeelingsofpersonalsafetyandincreasedpreferenceinpublicspaces
(Jorgensenetal.,2002)thatmaynotbemirroredinprivategardens.Thepresenceofrecreational
infrastructuresuchasbarbeques,playequipment,shadestructuresandpathwaysinfluencesthe
affordancesavailabletopeopleintheselandscapes,andisalsolikelytoalsoinfluencepreferencefor
landscapeswithintheseparksasdemonstratedbyHadavietal.(2015).Childrenareakeydriverof
preferenceforurbanpubliclandscapes,wheretheneedsofchildrenfeaturehighlyinpreferenceresponses
inqualitativestudies(Burgessetal.,1988).Ourresultswereconsistentwiththeideathatpreferenceis
alignedwiththeseaffordances;scatteredtrees,recreationalinfrastructureandplayequipmentwereall
preferredelements(Hadavietal.,2015).
However,ingardens,itispossiblethatdifferentaffordancesaredrivingpreference.Gardensareusedfor
manythingsincludingpublicdisplay,foodproductionandentertainment(Kendaletal,2012a;Bhatti,2004;
Daniels&Kirkpatrick,2006).Gardenersdescribetheirgardensasnature,butalsoasaplaceforbeautyand
recreation(Dahmus&Nelson,2014).Inthefrontgardenlandscapesincludedinthisstudy,therewasno
evidenceofutilitarianactivitiessuchasfoodproductionandpreferenceresponsesweremorelikelytobe
relatedtoaestheticandecologicaluses.Ingardens,factorssuchasfloweringandfoliagecolourandtexture
(Kendaletal.,2012a),vegetationcomplexity(KurzandBaudains,2010)orcompliancewithsocialnorms
(Nassaueretal.,2009)maybebetterpredictorsofpreferencethanthepresenceofrecreational
infrastructureortheprospectofsurveillance.
Thegroupingofthetwoadditionallandscapes(principalcomponents3and4,Table2),suggeststhatin
somelandscapesthephysicalfeaturesofthelandscapesuchasthestructuralcomplexityofvegetationmay
bemoreimportantthangreenspacecontext.TheDenseandOpenlandscapeswerelesstypicalpublicparks
andprivategardens,anditispossiblethatthesewerejudgedmoreonobservablephysicalattributes.For
example,theDenselandscapesincludeddenseshrubsandtreesingardenbedsinparks,ordense(mostly
native)vegetationinfrontyards.PreferenceforDenselandscapesmaybeassociatedwithpeople’s
preferencefornativevegetation,orwhethertheyperceivethisvegetationtobebeneficialforurbanwildlife
(Kendaletal.,2012a;KurzandBaudains,2010).Somewhatsurprisingly,Denselandscapeswerehighly
preferred,almostasmuchastheEnglishLandscapecomponent(averagepreferencescores4.7and5.3
respectively),andtheDensecomponentcontainedthemostpreferredphotograph(Photo.1inTable2).This
findingisnotconsistentwithmanyotherstudiesthathavegenerallyfoundthatspatiallyrestrictedsettings
areperceivedlessfavourably(Ulrich,1993)althougharecentstudyfoundnosignificantdifferenceinthe
restorativeeffectsofgreenspacelandscapeswithvaryingdensityofvegetation(vandenBergetal.,2014).
Thereareseveralexplanationsforourfindings.Previousstudieshavelargelylookedatpreferencefordense
vegetationnearpathswheresafetybecomesamajorconcern,whileourstudyexploreddensevegetationin
gardenbedsandalongfencelinesawayfrompaths.Itmaybethatawayfrompaths,densevegetationcan
beapreferredfeatureoflandscapes.Anumberofstudieshaveshownthatpeoplewhoindicateahigher
concernfornaturearelikelytoprefermorebiodiverseandecologicallysustainablelandscapes(Kurzand
Baudains,2010;Zhengetal.,2011),andthismaybereflectedinpreferencefordensevegetationawayfrom
paths.Alternatively,thispreferencefordensevegetationmayalsoreflectthefavourableaesthetic
propertiestheyprovide,suchasscreeninglessinterestingareasofthelandscapealongboundaryfencesor
7
adjacentroads.Thissuggeststhatwithcarefuldesign(e.g.placedawayfrompathsandaccompaniedbycues
tocare),densenaturalisticvegetationcouldbeincorporatedintopublicparkswithoutcompromisingtheir
perceptionassafeareas.
Openlandscapesweretheleastpreferredofthefourlandscapecomponents,whichisconsistentwiththe
resultsofpreviousstudies(ArnbergerandEder,2012;Burgessetal.,1988).Inourstudy,lawnwas
negativelyrelatedtopreference,particularlyforthesuburbangardencomponent.Itispossiblethatvery
openlandscapesmaynotprovidetherangeofaffordancesorvegetationfeaturestoattractandcaterforthe
rangeofsocialactivitiesthatresident’sdesirewithintheirgreenspaces.Whileveryopenlandscapesprovide
someaffordancessuchasthepotentialtoplayballsports,itislikelythattherangeofaffordancesprovided
byveryopenlandscapesismorelimitedthanthosewithmorevegetationcomplexity.Wefoundthatpeople
preferlandscapeswhentheyhavemorevegetationcomplexityandtrees,aswellasfeaturesthatallowfora
broadermixofrecreationalactivitiestooccur.Thispreferenceisconsistentacrosspublicparksandprivate
residentialgardensinthisstudy,andechoesfindingsfromotherstudies(Bjerkeetal.2006).
PracticalImplications
Unsurprisingly,EnglishLandscapestylepublicparkswerethemostpreferredlandscapesinthisstudy.These
aremulti‐purposelandscapesandcanbeusedinmanywaysbythecommunitywhileprovidingafeelingof
safety.Butourresultsalsoprovidesomeimpetusforconsideringtheincorporationofdensepatchesof
vegetationintheplanningandmanagementofurbangreenspaces,includingformalEnglishLandscapestyle
publicparks.Previousstudieshavefoundindividualpsychologicalbenefitsfromurbangreenspaceincreases
asafunctionofperceivedplantandanimalspeciesrichness(Fulleretal.,2007;Lucketal.,2011),whichare
expectedtobegreaterinmoredenselyvegetatedlandscapes;positivecorrelationshavebeenfound
betweendenselyvegetatedlandscapes,andthespeciesrichnessofbirdsandactivityofbats(Threlfalletal.,
2016).Previousresearchfindingsonthenegativeeffectsofdensevegetationonsafetyandpreferencehave
beenincludedinlandscapedesignguidelines(e.g.HumeCityCouncil,2003).Ourstudyinsteadsuggeststhat
intherightcontext,densevegetationcanalsobepreferredbypeople.Inlightofthesefindings,theremay
beopportunitiesforlandscapedesignersandmanagerstodiversifythestructureoflandscapesbeing
providedinpublicparks.Areasofdensevegetationcouldbeincorporatedintopublicparksinorderto
increasebiodiversityvaluesandpublicpreference,withoutcompromisingsafety,preferenceorvalueas
multi‐usespaces.Wherefeelingsofsafetycanbemaintainedusinglongviewsprovidedbybroadareasof
lawnbetweenthetreesandshrubs(Jorgensenetal.,2002),anddensevegetationcanbeframedinits
presentationthroughcuestocare(Nassauer,1995)itshouldbepossibletoeffectivelydevelopmulti‐
functionallandscapesfrombothsocialandecologicalperspectives(Pauleitetal.,2011).
Therangeofpreferenceresultsobservedinthisstudyindicatethaturbangreenspacesneedtoperform
multiplerolesbyprovidingmulti‐functionallandscapestoaccommodatethemajorityofresidents,an
outcomesupportedbythequalitativeresearchofBurgess(1988).A‘onesizefitsall’approachtolanduse
policymaynevermeetthegeneralpublic’sdesires(Howley,2011)andthereisincreasingrecognitionofthe
needforadequatepublicparticipationindecisionsaroundurbangreenspaceplanningtoensurethe
developmentofeffectiveurbanplanningpolicies.Swanwick(2009)expandsontheideaofthesediverse
needsbyfocussinginsteadontheindividual,proposingthatapersonneedsaccesstoa‘portfolioofplaces’,
includingdifferenttypesoflandscapesatdifferenttimesandfordifferentpurposes.Thereforeeach
individual,aswellaseachcommunity,requirescontactwitharangeoflandscapesthatcancaterforeachof
thosedifferentlandscapeexperiences,fromwhichtheycanselectsitesthatappealtotheirindividual
preferences(Schroeder,1987).Anumberofstudieshaveshownthatgreenspaceisunequallydistributedin
cities(e.g.Kendal,2012c)andfutureresearchshouldexplorethepreferencesandneedsalonggradientsof
socialdisadvantage.
Theintensepressuresplacedonbothprivateandpublicurbangreenspaceascitiesgrowinsizemeanthat
theselandscapesareoftenatriskfromcontinuedurbandevelopment(Jim,2004),despitebeinghighly
8
valuedbysomesectorsofsocietyfornon‐marketvalues.Rapidurbanisationandincreasingurbansprawl
canresultinareductionofgreenspaceandincreasedrecreationpressureonthepublicspacethatremains
(ArnbergerandEder,2012).Thiscanhavenegativeeffectsonurbanecosystemsandonurbanresidents,
suchasisolatingpeoplefromexperiencesofnature(Miller,2005).Thedevelopmentofawider,more
diverserangeofurbangreenspacelandscapeswouldencourageawiderrangeofusers,andthereforemay
increaserecognitionofthesocio‐economicvalueofthesespacesandassistintheirretentionandprotection
fromfutureurbandevelopment.
Acknowledgements
ThisstudyformspartoftheAustralianResearchCouncilLinkageprojectLP110100686.Fundingwasalso
receivedfromtheFrankKeenanTrust.Theauthorswouldalsoliketothankthesurveyparticipants.DKand
AKHwouldalsoliketoacknowledgefinancialsupportfromTheBakerFoundation.DKisfundedbytheClean
AirandUrbanLandscapehuboftheNationalEnvironmentalScienceProgramme.Wewouldliketothank
threeanonymousreviewerswhosecommentshaveresultedinagreatlyimprovedmanuscript.
References
Appleton,J.(1975).TheExperienceofLandscape,Wiley,London.
Arnberger,A.,&Eder,R.(2012).Theinfluenceofgreenspaceoncommunityattachmentofurbanand
suburbanresidents,UrbanForestry&UrbanGreening11(1),41‐49.
AustralianBureauofStatistics(2006).VicMapAddress2006databaseAustralianBureauofStatistics,
CommonwealthofAustralia.
AustralianBureauofStatistics(2013).2011CensusQuickStats,AustralianBureauofStatistics,
CommonwealthofAustralia.
Bhatti,M.(2006).‘WhenI’minthegardenIcancreatemyownparadise’:Homesandgardensinlaterlife,
TheSociologicalReview54,318–341.
Bjerke,T.,Østdahl,T.,Thrane,C.,&Strumse,E.(2006).Vegetationdensityofurbanparksandperceived
appropriatenessforrecreation,UrbanForestry&UrbanGreening5(1),35‐44.
Bolund,P.,&Hunhammar,S.(1999).Ecosystemservicesinurbanareas,EcologicalEconomics29(2),293.
Bourassa,S.C.(1990).Aparadigmforlandscapeaesthetics,EnvironmentandBehavior22,787‐812.
Bryman,A.(2008),SocialResearchMethods,OxfordUniversityPress,Oxford.
Burgess,J.,Harrison,C.M.,&Limb,M.(1988),People,parksandtheurbangreen:Astudyofpopular
meaningsandvaluesforopenspacesinthecity,UrbanStudies25,455‐473.
Bulut,Z.,&Yilmaz,H.(2008).“DeterminationofLandscapeBeautiesThroughVisualQualityAssessment
Method:ACaseStudyforKemaliye(Erzincan/Turkey).”EnvironmentalMonitoringandAssessment
141,121‐129.
Chiesura,A.(2004).Theroleofurbanparksforthesustainablecity,LandscapeandUrbanPlanning68(1),
129‐138.
CouncilofEurope(2000).TheEuropeanLandscapeConvention,Strasbourg.
Coolen,H.&Meesters,J.(2011).Privateandpublicgreenspaces:meaningfulbutdifferentsettings.Journal
ofHousingandtheBuiltEnvironment27,49–67
Dahmus,M.E.,&Nelson,K.C.(2014).NatureDiscoursesintheResidentialYardinMinnesota,Landscape
andUrbanPlanning,125,183–87
Daniel,T.C.,&Meitner,M.M.(2001).Representationalvalidityoflandscapevisualizations:Theeffectsof
graphicalrealismonperceivedscenicbeautyofforestvistas,JournalofEnvironmentalPsychology
21(1),61‐72.
deGroot,W.T.,&VandenBorn,R.J.G.(2003).“VisionsofNatureandLandscapeTypePreferences:An
ExplorationintheNetherlands.”LandscapeandUrbanPlanning63(3),127‐138.
Dunnett,N.,&Qasim,M.(2000).Perceivedbenefitstohumanwell‐beingofurbangardens,Horttechnology
10,40‐45.
9
Filova,L.,Vojarb,J.,Svobodovab,K.,&Sklenicka,P.(2015).Theeffectoflandscapetypeandlandscape
elementsonpublicvisualpreferences:waystouseknowledgeinthecontextoflandscapeplanning,
JournalofEnviornnmentalProtectionandManagement(inpress).
Fuller,R.A.,Irvine,K.N.,Devine‐White,P.,Warren,P.H.,&Gaston,K.J.(2007).Psychologicalbenefitsof
greenspaceincreasewithbiodiversity,BiologyLetters3,390‐394.
Ghosh,S.,&Head,L.(2009).Retrofittingthesuburbangarden:Morphologiesandsomeelementsof
sustainabilitypotentialoftwoAustralianresidentialsuburbscompared,AustralianGeographer
40(3),319‐346.
GIMPDevelopmentTeam(2004).GIMP:TheGNUImageManipulationProgram.
Gobster,P.H.(1995).AldoLeopold’secologicalesthetic:Integratingestheticandbiodiversityvalues,Journal
ofForestry93(2),6‐10.
Gobster,P.H.(1999).Anecologicalaestheticforforestlandscapemanagement,LandscapeJournal18(1),54.
Goddard,M.A.,Dougill,A.J.,&Benton,T.G.(2010),Scalingupfromgardens:Biodiversityconservationin
urbanenvironments,TrendsinEcology&Evolution25(2),90‐98.
Groenewegen,P.P.,vandenBerg,A.E.,deVries,S.,Verheij,R.A.(2006).VitaminG:effectsofgreenspace
onhealth,well‐being,andsocialsafety,BMCPublicHealth6,149.
Hadavi,S.,Kaplan,R.,&Hunter,M.C.R.,(2015).Environmentalaffordances:Apracticalapproachfordesign
ofnearbyoutdoorsettingsinurbanresidentialareas.LandscapeandUrbanPlanning134,19–32.
Head,L.,Muir,P.&Hampel,E.(2004).Australianbackyardgardensandthejourneyofmigration,
GeographicalReview94(3),326.
Hofmann,M.,Westermann,J.R.,Kowarik,I.&vanderMeer,E.(2012).Perceptionsofparksandurban
derelictlandbylandscapeplannersandresidents.UrbanForestryUrbanGreening11,1–10
Home,R.,Bauer,N.,&Hunziker,M.,(2010).CulturalandBiologicalDeterminantsintheEvaluationofUrban
GreenSpaces,EnvironmentandBehavior42(4),494‐523.
Howley,P.(2011).Landscapeaesthetics:Assessingthegeneralpublics'preferencestowardsrural
landscapes,EcologicalEconomics72,161‐169.
HumeCityCouncil(2003).OpenSpaceandLandscapeguidelines,HumeCityCouncil.
IBM,(2012).SPSSStatisticsVersion20.0,Ireland.
Jim,C.Y.(2004).Green‐spacepreservationandallocationforsustainablegreeningofcompactcities,Cities
21(4),311‐320.
Jim,C.Y.,&Chen,W.(2006).PerceptionandattitudeofresidentstowardurbangreenspacesinGuangzhou
(China),EnvironmentalManagement38(3),338‐349.
Jorgensen,A.,Hitchmough,J.,&Calvert,T.(2002).Woodlandspacesandedges:theirimpactonperception
ofsafetyandpreference,LandscapeandUrbanPlanning60,135‐150.
Kaltenborn,B.P.,&Bjerke,T.,(2002).Associationsbetweenenvironmentalvalueorientationsandlandscape
preferences,LandscapeandUrbanPlanning59(1),1‐11.
Kaplan,S.(1995).TheRestorativeBenefitsofNature:TowardanIntegrativeFramework,Journalof
EnvironmentalPsychology,15,169–82
Kaplan,R.,(2001).Thenatureoftheviewfromhome:Psychologicalbenefits,Environment&Behavior33,
507‐542.
Kaplan,R.,&Kaplan,S.,(1989).TheExperienceofNature:APsychologicalPerspective,CambridgeUniversity
Press,Cambridge.
Kaufman,A.J.,&Lohr,V.I.,(2004).Doesplantcoloraffectemotionalandphysiologicalresponsesto
landscapes?,ActaHorticulturae639,229‐233.
Kellett,J.(2011).TheAustralianquarteracreblock:thedeathofadream?,TPR:TownPlanningReview
82(3),263‐284.
Kendal,D.,Williams,K.J.H.,&Williams,N.S.G.,(2012a).,Planttraitslinkpeople'splantpreferencestothe
compositionoftheirgardens,LandscapeandUrbanPlanning105,34‐42.
Kendal,D.,Williams,N.S.G.,&Williams,K.J.H.,2012b,Acultivatedenvironment:exploringtheglobal
distributionofplantsingardens,parksandstreetscapes,UrbanEcosystems15,637‐652.
10
Kendal,D.,Williams,N.S.G.,&Williams,K.J.H.(2012c).Driversofdiversityandtreecoveringardens,parks
andstreetscapesinanAustraliancity.UrbanForestry&UrbanGreening,11,257–265.
Kim,J.,&Kaplan,R.,(2004).Physicalandpsychologicalfactorsinsenseofcommunity‐newurbanist
KentlandsandnearbyOrchardVillage,EnvironmentandBehavior36,313‐340.
Kirkpatrick,J.,Daniels,G.,&Davison,A.,(2009).Anantipodeantestofspatialcontagioninfrontgarden
character,LandscapeandUrbanPlanning93(2),103‐110.
Kirkpatrick,J.,Daniels,G.,&Davison,A.,(2011).Temporalandspatialvariationingardenandstreettreesin
sixeasternAustraliancities,LandscapeandUrbanPlanning101(3),244‐252.
Kurz,T.,&Baudains,C.,(2010).Biodiversityinthefrontyard:Aninvestigationoflandscapepreferenceina
domesticurbancontext,EnvironmentandBehavior44(2),166‐196.
Larsen,L.,&Harlan,S.L.(2006),Desertdreamscapes:Residentiallandscapepreferenceandbehavior,
LandscapeandUrbanPlanning78(1–2),85‐100.
Leslie,E.,Sugiyama,T.,Ierodiaconou,D.,&Kremer,P.(2010).Perceivedandobjectivelymeasuredgreenness
ofneighbourhoods:Aretheymeasuringthesamething?,LandscapeandUrbanPlanning95,28‐33.
Loram,A.,Tratalos,J.,Warren,P.,&Gaston,K.,(2007).Urbandomesticgardens(X):theextent&structure
oftheresourceinfivemajorcities.,LandscapeEcology22(4),601‐15.
Luck,G.W.,Davidson,P.,Boxall,D.,&Smallbone,L.(2011).RelationsbetweenUrbanBirdandPlant
CommunitiesandHumanWell‐BeingandConnectiontoNature.ConservationBiology,25(4),816–
826.
Lyons,E.(1983).DemographicCorrelatesofLandscapePreference,EnvironmentandBehavior15(4),487‐
511.
Maas,J.,Verheij,R.A.,Vries,S.d.,Spreeuwenberg,P.,Schellevis,F.G.,&Groenewegen,P.P.(2009).
Morbidityisrelatedtoagreenlivingenvironment,JournalofEpidemiologyandCommunityHealth
(1979‐)63(12),967‐973.
Miller,J.R.(2005).Biodiversityconservationandtheextinctionofexperience,TrendsinEcology&
Evolution20(8),430‐434.
Moore,D.S.,&McCabe,G.P.,2003,Introductiontothepracticeofstatistics,W.H.FreemanandCompany,
NewYork.
Nassauer,J.I.(1995).Messyecosystems,orderlyframes.LandscapeJournal,14(2),161–170.
Nassauer,J.,Wang,Z.,&Dayrell,E.,(2009).Whatwilltheneighborsthink?Culturalnormsandecological
design,LandscapeandUrbanPlanning92,282‐292.
Pauleit,S.,Liu,L.,Ahern,J.,&Kazmierczak,A.(2011).Multifunctionalgreeninfrastructureplanningto
promoteecologicalservicesinthecity,inNiemelä,J.,Breuste,J.,Elmqvist,T.,Guntenspergen,G.,
James,P.,McIntyre,N.E.(eds)UrbanEcology:Patterns,processesandapplications.Oxford
UniversityPress,NewYork,pp.272‐285.
Pereira,G,Foster,S.,Martin,K.,Christian,H.,Boruff,B.J.,Knuiman,M.,&Giles‐Corti,B.,(2012).The
associationbetweenneighborhoodgreennessandcardiovasculardisease:anobservationalstudy,
BMCPublicHealth12,466.
Ode,Å.,Fry,G.,Tveit,M.S.,Messager,P.,&Miller,D.(2009).Indicatorsofperceivednaturalnessasdrivers
oflandscapepreference,JournalofEnvironmentalManagement90(1),375‐383.
Peschardt,K.K.,Stigsdotter,U.K.,Schipperrijn,J.,2014.IdentifyingFeaturesofPocketParksthatMayBe
RelatedtoHealthPromotingUse.LandscapeResearch1–16.
Qiu,L.,Lindberg,S.,&Nielsen,A.B.(2013).Isbiodiversityattractive?—On‐siteperceptionofrecreational
andbiodiversityvaluesinurbangreenspace.LandscapeandUrbanPlanning.119,136–146
RDevelopmentCoreTeam,(2007).R:Alanguageandenvironmentforstatisticalcomputing,RFoundation
forStatisticalComputing,Vienna,Austria.
Scherer,K.R.(2005).Whatareemotions?Andhowcantheybemeasured?,SocialScienceInformation44,
695‐729.
Schroeder,H.W.(1987).Dimensionsofvariationinurbanparkpreference:Apsychophysicalanalysis,
JournalofEnvironmentalPsychology7(2),123‐141.
11
Sevenant,M.,&Antrop,M.(2009).“CognitiveAttributesandAestheticPreferencesinAssessmentand
DifferentiationofLandscapes.”JournalofEnvironmentalManagement90,2889‐2899.
Shaw,A.,Miller,K.,&Wescott,G.(2013).WildlifeGardeningandConnectednesstoNature:Engagingthe
Unengaged,EnvironmentalValues,22,483–502
Sokal,R.R.,&Rolhf,F.J.(1995).Biometry:ThePrinciplesandPracticesofStatisticsinBiologicalResearch,
W.H.Freeman,NewYork.
Swanwick,C.,(2009).Society'sattitudestoandpreferencesforlandandlandscape,LandUsePolicy26,S62‐
S75.
Tabachnick,B.G.,Fidell,L.S.,(2001).UsingMultivariateStatistics,AllynandBacon,Boston.
Takano,T.,Nakamura,K.,Watanabe,M.,(2002).Urbanresidentialenvironmentsandseniorcitizens’
longevityinmega‐cityareas:theimportanceofwalk‐ablegreenspace,JournalofEpidemiologyand
CommunityHealth56,913‐6.
Thompson,K.,Austin,K.C.,Smith,R.M.,Warren,P.H.,Angold,P.G.,&Gaston,K.J.(2003),Urbandomestic
gardens(I):Puttingsmall‐scaleplantdiversityincontext,JournalofVegetationScience14,71‐78.
Threlfall,C.G.,Williams,N.S.G.,Hahs,A.K.,&Livesley,S.J.(2016).Approachestourbanvegetation
managementandtheimpactsonurbanbirdandbatassemblages.LandscapeandUrbanPlanning,
153,28–39.
Trent,R.B.,Neumann,E.,&Kvashny,A.(1987).Presentationmodeandquestionformatartifactsinvisual
assessmentresearch,LandscapeandUrbanPlanning14,225‐235.
Tveit,M.S.(2009).Indicatorsofvisualscaleaspredictorsoflandscapepreference;acomparisonbetween
groups,JournalofEnvironmentalManagement90,2882‐2888.
Tzoulas,K.,Korpela,K.,Venn,S.,Ylipelkonen,V.,Kazmierczak,A.,Niemela,J.,&James,P.(2007).Promoting
ecosystemandhumanhealthinurbanareasusinggreeninfrastructure:Aliteraturereview,
LandscapeandUrbanPlanning81,167‐178.
Ulrich,R.(1977).Visuallandscapepreference:Amodelandapplication,Man‐EnvironmentSystems7(5),279‐
293.
Ulrich,R.(1993).Biophilia,biophobiaandnaturallandscapes,in:TheBiophiliaHypothesis(S.Kellert,E.
Wilson,eds.),IslandPress,Washington,pp.73‐137.
vandenBerg,A.E.,Jorgensen,A.,&Wilson,E.R.(2014).Evaluatingrestorationinurbangreenspaces:Does
settingtypemakeadifference?Landscape&UrbanPlanning127,173–181
vandenBerg,A.E.,&Koole,S.L.(2006).NewwildernessintheNetherlands:aninvestigationofvisual
preferencesfornaturedevelopmentlandscapes,LandscapeandUrbanPlanning78(4),362‐372.
vandenBerg,A.E.,vanWinsum‐Westra,M.,(2010).Manicured,romantic,orwild?Therelationbetween
needforstructureandpreferencesforgardenstyles,UrbanForestry&UrbanGreening9(3),179‐
186.
Westphal,L.M.(2003).Urbangreeningandsocialbenefits:astudyofempowermentoutcomes,Journalof
Arboriculture29(3),137‐147.
Williams,K.J.H.,&Cary,J.(2002).Landscapepreferences,ecologicalqualityandbiodiversityprotection,
Environment&Behavior34(2),257‐274.
Williams,K.(2002).ExploringresidentpreferencesforstreettreesinMelbourne,Australia.Journalof
Arboriculture,28(4),161–170.
Yang,B.E.,Brown,T.J.,(1992).Acrossculturalcomparisonofpreferencesforlandscapestylesand
landscapeelements,EnvironmentandBehavior24(4),471‐507.
Yu,K.J.,(1995).Culturalvariationsinlandscapepreference‐comparisonsamongChinesesubgroupsand
westerndesignexperts,LandscapeandUrbanPlanning32(2),107‐126.
Zheng,B.,Zhang,Y.,Chen,J.,(2011).Preferencetohomelandscape:wildnessorneatness?,Landscapeand
UrbanPlanning99(1),1‐8.
Zube,E.H.,(1987).Perceivedlandusepatternsandlandscapevalues,LandscapeEcology1,37‐45.
12
Figures
Figure1:AverageLikertscore(±standarderror)perrespondentforphotographswithinlandscapeeach
preferencecomponent.Componentswithdifferentlettersaresignificantlydifferent(Tukey’sHSDp<0.05).
Figure2:Percentageofa)lawn,b)shrubsandflowersandc)trees(±standarderror)inphotographed
landscapesofeachpreferencecomponent.Valueswiththesameletterwerenotsignificantlydifferent
(KruskalWallisANOVAp<0.05).
13
Figure3:Therelationshipbetweena)lawnandb)treesandshrubs(includinglinearandquadraticterms)
andpreference,forbothparksandgardens.Noteregressionswereperformedusingrespondentleveldata,
butsymbolsindicatemeanpreferenceandproportioncoverforeachlandscape.Standarderrorbarsare
shown.
14
Tables
Table1:Demographicqualitiesofsurveyrespondentscomparedtothebroaderpopulationaverage
withintheLocalGovernmentAreascoveredbythesurvey.
DemographicQuality SurveyRespondents
AverageforthreeLocal
GovernmentAreas(Australian
BureauofStatistics,2013)
Owntheirhouse 87% 74%
Female 60% 51%
Age(median) Between45and54years 37years
Speaklanguageotherthan
Englishathome 11% 22%
IndexofSocioeconomic
AdvantageandDisadvantage
decile
2‐10 7
15
Table2:Sampleofthephotographswithineachpreferencecomponent.Thecomponentloadingscorefor
eachphotographisshowninbrackets.
EnglishLandscape SuburbanGarden Open Dense
Photo.12(0.75)
Photo.22(0.72)
Photo.7(0.86)
Photo.8(0.85)
Photo.3(0.74)
Photo.13(0.71)
Photo.16(0.68)
Photo.27(0.73)
Photo.10(0.70)
Photo.2(0.68)
Photo.21(0.68)
Photo.9(0.73)
Photo.15(0.68)
Photo.20(0.61)
Photo.19(0.65)
Photo.5(0.70)
Photo.17(0.66)
Photo.26(0.60)
Photo.25(0.57)
Table3:Regressionsofvegetationformonaveragepreferencescores.xisthepercentageofeachvegetation
form,Parkisabinaryvariablesetto1forparksand0forgardens.
Vegetationform Regressionequation
%Lawn
y=3.9+1.1*Park+0.025x‐0.0007x2,R2=0.18
Standarderrors:Intercept=0.06,Park=0.05,x=0.004,x2=0.00006
%Shrubs&Trees
y=1.9+1.0*Park+0.055x‐0.0003x2,R2=0.20
Standarderrors:Intercept=0.10,Park=0.05,x=0.004,x2=0.00005
16
Supplementarymaterial
Componentloadingmatrix
Thecomponentloadingsofeachlandscapeoneachprincipalcomponent,sortedinloadingorderoneach
component.
Landscape English
Landscape
Suburban
Garden Open Dense
Photo.12 0.75 0.17 0.11 ‐0.16
Photo.3 0.74 0.07 ‐0.15 0.04
Photo.10 0.70 0.11 0.18 ‐0.14
Photo.15 0.68 0.03 0.13 0.30
Photo.17 0.66 0.02 0.17 0.04
Photo.28 0.64 0.07 0.01 0.13
Photo.23 0.52 ‐0.17 0.27 0.26
Photo.22 ‐0.16 0.72 0.06 ‐0.05
Photo.13 0.22 0.71 ‐0.23 0.17
Photo.2 0.15 0.68 0.04 ‐0.03
Photo.20 0.10 0.61 0.27 ‐0.15
Photo.26 0.12 0.60 0.11 0.05
Photo.4 ‐0.14 0.56 0.07 0.22
Photo.6 0.02 0.55 0.02 0.05
Photo.11 0.36 0.52 ‐0.15 0.07
Photo.18 0.02 0.49 0.44 ‐0.04
Photo.7 0.18 ‐0.11 0.86 ‐0.05
Photo.21 ‐0.27 0.32 0.68 0.07
Photo.16 0.32 0.12 0.68 ‐0.16
Photo.19 0.12 ‐0.03 0.65 0.18
Photo.25 ‐0.12 0.18 0.57 0.27
Photo.8 ‐0.20 0.05 0.03 0.85
Photo.27 ‐0.14 0.13 0.11 0.73
Photo.9 0.25 0.03 ‐0.12 0.73
Photo.5 0.32 ‐0.11 ‐0.05 0.70
Photo.14 0.24 ‐0.01 0.17 0.48
Proportion
Var
0.16 0.15 0.14 0.12
Cumulative
Var
0.16 0.32 0.45 0.57
Eigenvalue
7.7 3.0 2.5 1.6