Cities 42 (2015) 130–141

Contents lists available at ScienceDirect

Cities

journal homepage: www.elsevier .com/locate /c i t ies

Community attachment and resident attitude toward old masonry wallsand associated trees in urban Hong Kong

http://dx.doi.org/10.1016/j.cities.2014.09.0060264-2751/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Room 3.07, Building G31, Griffith University, GoldCoast Campus, QLD 4222, Australia. Tel.: +61 7 55527419; fax: +61 7 55528244.

E-mail addresses: alex.lo@griffith.edu.au (A.Y. Lo), hragjcy@hkucc.hku.hk(C.Y. Jim).

Alex Y. Lo a,⇑, C.Y. Jim b

a Urban Research Program, Griffith School of Environment, Griffith University, Gold Coast, QLD 4222, Australiab Department of Geography, University of Hong Kong, Pokfulam Road, Hong Kong

a r t i c l e i n f o a b s t r a c t

Article history:Received 28 June 2014Received in revised form 20 August 2014Accepted 28 September 2014

Keywords:Urban treesStone-retaining wallsCompact cityPublic perceptionUrban nature conservationPlace attachment

Mature urban trees dwelling on old masonry walls are a rare urban ecological and landscape treasure.These ‘stonewall’ trees denoting synergy between nature (trees) and culture (masonry) are worthpreserving. This study aims to identify factors influencing public attitude toward this important butthreatened natural-cum-cultural asset in a compact city milieu. Evidence is solicited from a face-to-facequestionnaire survey involving 800 citizens of Hong Kong. Results indicate that Hong Kong peoplegenerally appreciate the existence of stonewall trees. Younger adults harbored doubts about their value,whereas the older ones were more sympathetic. Local residents of the suburbs that host the majority ofstonewall trees registered stronger commitment than non-local residents. Community attachment hadpositive effects across both resident groups, but length of tenure was not a key factor. This suggests thataffinity for stonewall trees is a function of the affective bonding between people and the communitywhere they live. Deep engagement with the local community elevated the desire for preserving the wallsand associated trees. The findings are important for understanding public perception and enlistingsupport from the larger society for preserving the urban asset against mounting pressures for urbandevelopment and renewal.

� 2014 Elsevier Ltd. All rights reserved.

Introduction As a typical compact city, Hong Kong exemplifies the chronic

Trees and cities have co-existed for millennia under spontane-ous or cultivated regimes. Urban trees contribute to the well-beingof people, environmental quality, and the long-term sustainabilityof cities (Beer, Delshammar, & Schildwacht, 2003; Chiesura, 2004;Jim, 1998a, 2003; Konijnendijk, 2008; Miller, 1997). Yet the term‘urban tree’ carry ambivalent or incongruous undertones. While‘urban’ implies artificiality and difficult conditions for non-humanlife, ‘tree’ conjures up an image of nature and serenity. Various citygovernments have attempted to resolve the conflicts between thenatural and artificial elements with divergent levels of achieve-ment. Compact cities, in particular, are characterized by a tighturban fabric with limited opportunities for planting. Many treesfail to compete for space, while some are relegated to cramped cor-ridors and tiny parcels occluded in various urban infrastructures.Urban development and renewal have threatened the viability ofurban trees, some of which are hundred years old and have beengranted heritage status (Jim, 1998b, 2010).

spatial conflict between green-space conservation and urban devel-opment. Over half of the land is steep natural terrain, while onlyabout 15% is suitable for urban development (EnvironmentalProtection Department, 2006). Urban population and infrastructureare highly concentrated. Hong Kong, with a land area of 1108 km2, ishome to 7.18 million people. The average population density hasreached 6330 persons/km2 and is among the highest in the world.The keen competition for developable land and the phenomenalcost of land and real estates have resulted in grave shortage ofplantable space in urban areas.

Nonetheless, nature has enlisted unexpected sites for spontane-ous vegetation growth in the cramped interstices of the compactcity. Urban Hong Kong is endowed with an unmatched ecologicalhabitat in the form of stone-retaining walls. Most of these wallsare apparently inhospitable for plants, but the traditional methodand style of wall construction permitted spontaneous colonizationby a host of plants which includes some tree species. The oldeststone-retaining walls were constructed using traditional Chinesetechniques over 100 years ago. They are often blanketed by densevegetation, including large trees up to 21 m tall (Jim, 1998b, 2010;Jim & Chen, 2010). The largest specimen trees are as old as the cen-tury-old walls. Not many cities in the world have such a relatively

A.Y. Lo, C.Y. Jim / Cities 42 (2015) 130–141 131

large number of old stone walls with spontaneous vegetation tobestow a distinct urban natural asset.

In Hong Kong, the majority of mature ‘stonewall trees’ exist inthe Central and Western (CW) District. Together with the old stonewalls to which they firmly attach, these trees have become a culturallegacy for the community and an indispensable part of the collectivememories of local residents. People may develop a sentimentalattachment to such natural companions and respectable doyens.However, urban renewal and construction of new infrastructurehave brought demolition, degradation and modification of oldmasonry walls that host these trees, resulting in demise and massiveinjury of this rare ‘natural-cum-cultural’ asset of the city. The dis-turbed habitats become unsuitable for tree growth, thus triggeringgradual decline of the remnant trees. A series of events triggeredpublic outrage and attracted extensive media coverage. Localresidents and environmentalists managed to secure major adjust-ments to these construction projects to protect the endangerednatural–cultural hybrids from massive unrecoverable damage.

Statutory protection is needed to preserve the valuable stone-wall trees. While local councils lack power and resources to act,the city government remains incapable of effectively managingurban heritage trees (Jim, 2002, 2003). Conservation efforts havebeen dominated by a handful of NGOs and concerned local residentsand scientists. In Hong Kong, the community as a whole is notactively involved in urban governance (Ng, 2002). Wider commu-nity support has to be enlisted to reinforce the bottom-up effortsand drive changes at the institutional level. This will be a challeng-ing task, because the oldest masonry walls are restricted to a fewrelatively small communities and viewed by some citizens as anoutdated and anachronistic infrastructure that could be replacedby new ones. Understanding society’s attitude toward stonewalltrees will be instrumental to maximize public support for drivingchanges in official planning and management practice. Objectiveand systematic evidence on the public’s views is lacking and moreis needed to inform conservation and publicity campaigns.

This study explores Hong Kong people’s attitude toward stone-wall trees representing an important component of nature-in-cityworthy of conservation. Past events showed that residents of theCW District (hereafter called ‘local residents’) reacted most stronglyto proposed removal of these trees. The study seeks evidence ontheir greater concern and, more importantly, identifies possible rea-sons. Understanding the latter could help develop strategies forrecruiting support from non-local residents. Our pilot research indi-cated that attitude toward urban green spaces is closely associatedwith the individual’s concern about the wider community (Lo & Jim,2010). Attempts to solicit support from the larger society could ben-efit from articulating the advocacy of nature conservation in termsof broader social and community imperatives. Studies of placeattachment offer useful insights for identifying key socio-culturalfactors influencing people’s attitude (Lin & Lockwood, 2014;McCunn & Gifford, 2014; Ryan, 2005). In this paper, we introducethe concept of place attachment to the study of the relationshipbetween heritage trees and people in a compact city milieu.

The remainder of this paper begins with an elaboration on thestonewall trees in Hong Kong. This is followed by a brief reviewof relevant studies of place attachment. Our research methodsare then described and results are presented. At the end of thepaper we discuss the implications of the research findings forunderstanding public perception and motivating the bulk of thepopulation.

Stonewall trees in Hong Kong

Hong Kong was established in 1842 as a British colony at thesubtropical coast of China. The city consists of 18 Council Districtsdistributed among three centrally governed regions, namely, New

Territories, Kowloon, and Hong Kong Island. The first colonialurban settlements landed on the northern coastal strip of HongKong Island, beginning with the CW District which remains theadministrative and commercial core of the city. The District is aculturally significant site steeped in historical icons and connota-tions, as symbolized by a variety of Victorian-style architecturesand old-time ambience reminiscent of genius loci.

Rapid urbanization since the Second World War fuelled produc-tion and consumption of developable land, which is increasinglyexpensive (Chiu, 2006; Jim, 1998a). Some of the urban Districtshave been filled to over 45,000 persons/km2 (i.e. Yau Tsim Mong,Wong Tai Sin, and Kwun Tong Districts). High-rise and high-den-sity development mode is spatially prevalent. Built-up areas arerestricted to merely 23% of the land (Planning Department,2013). The land is subject to substantial topographical constraints,with about 80% at above 100 m elevation associated with steepslopes that are unsuitable for development. The overall urbanmatrix has little plantable areas by design or by default for inser-tion of greenery. Besides physical constraints due to the lack ofgrowth spaces, the microclimatic and soil conditions of thecramped planting sites are unfavorable for tree growth.

Some species of trees managed to establish and survive in harshurban habitats, such as old masonry walls. Most of these imposingstructures have existed for some decades to more than a century.They include the walls of historical buildings, geotechnical retain-ing structures, and ruins of monuments. In Hong Kong, oldmasonry walls are concentrated in the urban core developed inthe nineteenth and early twentieth century, notably the northernstrip of Hong Kong Island and the Kowloon Peninsula (Jim,1998a, 1998b). Some are situated at roadsides and adjacent tomain streets, while others are found in obscure lanes and behindbuildings. Elaborately handcrafted in line with traditional Chinesemasonry design, they represent the confluence of time-honoredscience and technology in offering a pragmatic solution tomaximize developable land in a hilly terrain. The artistic shapes,engineering services and endurance of old masonry walls havewon appreciation from local residents.

Walls appear to be harsh and difficult to live in and thereforeunlikely sites for trees. Nevertheless, some 30 trees species arefound on old masonry walls in Hong Kong (Jim, 1998b, 2013). Witha rough surface with lots of gaps between adjacent stone blocksand soil in the crevices for spontaneous plant growth withouthuman intervention, they provide fine landscape and urban eco-logical elements to the otherwise nondescript streetscape (Fig. 1)(Jim, 2014; Jim & Chen, 2010). Due to existence at elevated posi-tions on the artificial cliffs, they could escape from the otherwiserampant root and stem damages that inflict many urban trees inthe city. A recent study has found that in Hong Kong, old stonewalls support some 1200 trees in built-up areas that are occupiedby artificial surfaces and have few undeveloped sites (Jim, 2010).Most of the walls have one to three trees, with an exceptionalone having fifty. These trees are mostly young or small at 1–3 mhigh, the tallest one being 21 m (Fig. 2). Considering the extraordi-narily harsh environment of stone walls, especially verticality,exposure, moisture deficiency, and lack of suitable substrate, thenumber of individual trees and tree species and their aggregatebiomass are remarkable (Jim, 1998a, 1998b, 2010; Jim & Chen,2010). The old CW District has the greatest concentration of oldmasonry walls and companion trees (Fig. 3).

Stonewall trees in the urban core of Hong Kong are a precious‘natural-cum-cultural’ asset. The old masonry walls denote a syn-ergy between nature (trees) and culture (masonry), exemplifyingthe harmonious co-existence of culture and nature in a compacturban environment. However, urban renewal and developmentthreaten to displace the century-aged walls from the city. The spa-tial conflict was heightened in 2006, when one of the proposed

Fig. 1. One of the largest stonewall trees in Hong Kong, a 21 m tall Chinese Banyan (Ficus microcarpa), perches on a masonry retaining wall that is over a century old andembedded in the densely developed residential neighborhood.

132 A.Y. Lo, C.Y. Jim / Cities 42 (2015) 130–141

alignments of a major underground railway extension projectthreatened the cream of the masonry-wall and associated stone-wall-tree stocks at Forbes Street in CW District (Fig. 2). Anotherredevelopment project might also endanger a cluster of meritori-ous stonewall trees at a former Police Married Quarters site atHollywood Road in the same District. Moreover, the branches ofmany large stonewall trees hanging above carriageways are com-monly perceived as hazardous to road users. Three recent fatal treecollapse incidents, all involving ground-growing trees, have per-haps unnecessarily triggered knee-jerk reactions in the administra-tion to remove or drastically prune many large trees, includingstonewall ones. Since new retaining walls are made of reinforcedconcrete that do not permit tree colonization, the old masonrywalls have become a threatened and relict heritage desperatelyin need of an effective conservation shield.

Wider public endorsement is instrumental to legitimizeconservation initiatives and mobilize political energy essential

for attracting legislative attention and resource allocation. Appeal-ing to community attachment is an effective strategy for recruitingsupport from the larger society to the extent in which stonewalltrees are widely seen as the community’s heritage. Further insightscan be drawn from studies of place attachment.

Place attachment

Landscape perception and preference are place-specific(Swanwick, 2009). Place means more than just space (Tuan,1974, 1977); it refers to space that has been given meaningthrough personal, social, or cultural processes (Low & Altman,1992). Place attachment is a useful concept for understanding atti-tude toward urban green spaces. It can be defined as the affectivebonding between an individual or community and the surroundingenvironmental settings, encompassing a range of psychological,cognitive, social and cultural elements (Casakin et al., in press;

Fig. 2. An imposing linear urban forest dwells on a 120-year old stone retaining wall in the Central & Western District of Hong Kong.

A.Y. Lo, C.Y. Jim / Cities 42 (2015) 130–141 133

Kaltenborn & Bjerke, 2002; Low & Altman, 1992; Ryan, 2005;Stedman, 2006; Walker & Ryan, 2008). The concept implies a grad-ual transformation of the experience of a space or a piece of landinto a culturally or socially meaningful and shared symbol, thatis, place (Low, 1992). As Brehm, Eisenhauer, and Krannich (2006,p. 144) suggest, ‘positive cognition related to a specific settingallows people to acquire a sense of belonging to places that givemeaning to their lives’. Through the vehicle of place, individualsand communities apprehend and embed themselves into a varietyof life experiences with the environs.

Research has found a strong and complex relationship betweenplace attachment and landscape preference (Tapsuwan, Leviston, &Tucker, 2011; Walker & Ryan, 2008). A common observation is thatattachment to a place could strengthen concern about degradinglandscapes and motivates environmental action. Vaske andKobrin (2001), for example, conclude that environmentallyresponsible behavior is a function of place attachment and identity.

Similar findings pertaining to green space and landscape percep-tion have been reported by Walker and Ryan (2008) and Lokocz,Ryan, and Sadler (2011) on rural New England in the U.S.,Budruk, Thomas, and Tyrrell (2009) on an Indian city, Arnbergerand Eder (2012) on Vienna’s urban and suburban sites, and Linand Lockwood (2014) on Australia’s protected areas. In Norway,Kaltenborn and Bjerke (2002) have shown that affective bondingto the place had a positive effect on perceived attractiveness oflandscapes with historically significant elements. These studieslend empirical support to our suggestion that place attachmentcan help understand people’s preference for urban trees, particu-larly those of cultural and historical significance.

Apart from attachment to the physical environment, Brehmet al. (2006) argue that localized social attachments involvingfriends, relatives, and others known in the community also playan important role. The emphasis on social dimensions echoesRyan’s (2005) binary approach, which distinguishes ‘place-specific

Fig. 3. Map showing the location and alignment of old stone retaining walls in the Central & Western District of Hong Kong. They are situated mainly in the hilly part of theold city core developed since the 1840s.

134 A.Y. Lo, C.Y. Jim / Cities 42 (2015) 130–141

attachment’ from ‘conceptual attachment’. According to Ryan(2005, p. 38), place-specific attachment is held by neighbors andrecreational park users tied to a particular site and are unwillingto change location, whereas conceptual attachment is held byvolunteers and experts who are knowledgeable about the urbannatural areas and prefer a particular type of landscape rather thana particular place. A similar analytical scheme is adopted by Linand Lockwood (2014), who argue that spatially localized attach-ment accords importance with a particular place, whereas general-ized attachment applies to multiple places that share a commonsource for place formation. People affectively attached to a placein a generic way could cognitively engage with alternative sitesthat deliver similar services that contribute to the transformationof the experience of a landscape into a place.

Place substitutability explains the key difference between thetwo forms of attachment advanced by Ryan (2005) and Lin andLockwood (2014). Individuals who are deeply engaged with thepast and present of where they live and the people around themtend to appreciate and identify themselves with the unique placeswithin their communities that can hardly be found elsewhere.Older communities in the urban core have indicated a differentview about urban green spaces: veteran working-class residentsare attracted by a particular social space offered by neighborhoodparks, and not primarily the physical attributes. Youngermiddle-class residents, on the other hand, are more impressed bythe aesthetic and ecological values of parks, which are notrestricted to locality (Lo & Jim, 2010).

Mature stonewall trees in Hong Kong are site-specific. They arefound in a few old communities and large clusters with higherconservation values are located at the western portion of the CW

District. The notion of specific and generic attachment could offerimportant insights into Hong Kong people’s attitude toward thispublic asset. Drawing on these previous studies, we hypothesizedthat stonewall trees are more appreciated by local residents, espe-cially those having lived in these communities for decades, andthat the appreciation is a function of the attachment to the placeand people around. These questions were addressed by an empir-ical study involving a social survey conducted in Hong Kong, asdescribed in the next section.

Data collection

The city of Hong Kong is divided into 18 administrative districts,each being managed by a District Council. Due to resource con-straints, the social survey was only administered in 13 selectedDistricts in Hong Kong, which are home to nearly 80% of HongKong population (Table 1). These include the CW District, wherethe majority of mature stonewall trees are situated, and 12 otherDistricts. The sampling strategy was designed to collect an ade-quate number of observations from the CW District to allow for acomparison between local and non-local residents. A number ofstonewall trees also exist in the Wan Chai District, neighboringthe CW District. As we did not collect sample from Wan Chai, the‘non-local’ sample consists of those residents living at a distancefrom clusters of stonewall trees (i.e. outside CW and Wan Chai Dis-tricts). About one-fourth of sampling quota was assigned to thewestern portion of the CW District, including Kennedy Town, SaiYing Pun, and Pok Fu Lam. The rest was distributed among theother twelve sampling sites according to the size of local popula-tion. As a result, a larger proportion (23.8%) of respondents than

Table 1Distribution of actual population and respondents by sampling sites. Source: Censusand Statistics Department (2013).

District Census Sample Difference% of totalpopulation

% of respondents

Hong Kong IslandCentral & Western (CW) 3.5 23.8 �20.3Eastern 8.3 8.8 �0.5Southern 3.9 5.0 �1.1

New TerritoriesKwai Tsing 7.1 7.5 �0.4Tuen Mun 6.8 10.0 �3.2North 4.3 5.0 �0.7Sha Tin 9.0 10.0 �1.0Sai Kung 6.2 7.5 �1.3

KowloonYau Tsim Mong 4.4 3.1 1.3Shum Shui Po 5.4 4.4 1.0Kowloon City 5.3 3.8 1.5Wong Tai Sin 6.0 5.0 1.0Kwun Tong 8.9 6.3 2.6

Total 79.1 100

A.Y. Lo, C.Y. Jim / Cities 42 (2015) 130–141 135

the actual share of population (3.5%) came from the CW District. Inother sampling sites, discrepancies between the sample and actualpopulation were controlled below or around 3% (Table 1). Samplingquota was specified for each District according to the distributionof residents in four age groups, i.e. 15–24, 25–44, 45–64, and 65or above (Census and Statistics Department, 2013). This strategyachieved a locally representative sample with no statisticallysignificant difference from individual Districts in terms of agedistribution (see Appendix for details).

Face-to-face interviews were conducted in public areas. Onemajor public park or multiple neighborhood parks were selectedfrom each of the selected districts as study sites, based on proxim-ity to housing clusters, pedestrian traffic, and suitability for execut-ing intercept survey. Park visitors and pedestrians passing-by theparks were invited to participate in the research. Systematicsampling procedures were adopted: one out of every three adultvisitors or passers-by was invited to participate in the survey. Sam-pling quota was determined by age distribution in the district con-cerned, as mentioned above. Trained interviewers were instructedto approach a particular age group when the sampling quota forthe remaining age groups had been exhausted.

A university-based research service provider administrated thequestionnaire survey. Several undergraduate students with priorinterview experience were recruited as interviewers and trainedin the survey procedure. Survey activities commenced in October2012 and finished in February 2013.

Questionnaire design

The questionnaire solicited respondents’ attitude toward stone-wall trees, perceived importance of tree functions, and the level ofcommunity attachment. Respondents’ personal characteristicswere also recorded in order to statistically control socio-economicfactors in regression analysis.

Dependent variable

Prior to responding to a series of questions, participants of thesurvey were presented a factual description of stonewall trees thatprovided background information, including how these trees growfrom pre-war stone-retaining walls, where they are located (i.e.mostly the CW District), and the fact that modern concrete walls

no longer allow trees to colonize. Pictures showing two local exam-ples were used for illustration. Attitude was gauged by sevenclosed-ended questions that described the intrinsic contributionsand importance of these trees and the need of protecting them,such as ‘Stonewall trees are part of the community history thatmoney cannot buy’, ‘It is our moral responsibility to protect stone-wall trees’, and ‘Damage of stonewall trees is not recoverable’. Fiveresponse options were available to respondents, ranging from‘Strongly disagree’ to ‘Strongly agree’. The seven attitude itemswere combined into a composite scale and used as a dependentvariable in regression analysis.

Independent variables

Another set of questions probed respondents’ general percep-tion of urban trees. These questions specified the functional bene-fits of urban trees, such as cooling, dampening wind, interceptingrainwater, etc., and were conceptually distinct from those compris-ing the dependent variable, which emphasized aesthetic and ethi-cal dimensions. Including these items as independent variablescould control for their effects on attitude toward stonewall treesin regression modelling. A five-point Likert scale was employedto gauge respondents’ opinions about these functions, with optionsranging from ‘Very important’ to ‘Not at all important’.

The third part of the questionnaire included eight questionsthat assessed respondents’ attachment to the community theycurrently lived in. These questions described various elements ofcommunity attachment, including emotions, interest in localhistory, engagement in community affairs, familiarity, and socialrelationship. Survey statements were adapted from varioussources, including Brehm et al. (2006), Budruk et al. (2009), andTapsuwan et al. (2011), etc. and modified to match local circum-stances. Responses to these questions were elicited based on afive-point Likert scale, with options ranging from ‘Strongly agree’to ‘Strongly disagree’.

Respondents were also asked if they were aware of stonewalltrees prior to participating in this survey; available response optionsincluded ‘yes’, ‘little’, or ‘not at all’. The final section of the question-naire recorded personal characteristics of respondents, includingage, personal income, highest educational qualification, sex, the Dis-trict where they currently live in and tenure in this District.

Responses to these questions form the basis of a regressionanalysis. Attitude toward stonewall trees was regressed on theperceived importance of tree functions, the level of communityattachment and socio-economic factors to examine the statisticalrelationship between them. In addition, the Principal ComponentsAnalysis, with varimax rotation, was conducted to ascertain thedimensions of the attitudinal variables. The ANOVA techniquewas used to compare CW residents’ attitude toward stonewalltrees with the rest of the sample. Results of the quantitative anal-ysis were reported in the next section.

Results

Demographics

A total of 800 face-to-face interviews were completed in theselected Districts. A comparable number (787) of individuals con-tacted on-site refused to participate or terminated the interviewhalfway through. This yielded a completion rate of 50.4%. Responserate is 60.7% if those who agreed to participate but were not qual-ified or their age group had exceeded the quota were taken intoaccount (417 individuals). One hundred and ninety interviews, or23.8%, were conducted in the CW District (Table 1).

In the full sample, the number of male respondents (53.2%)exceeded females (46.8%). Most of them identified themselves in

136 A.Y. Lo, C.Y. Jim / Cities 42 (2015) 130–141

the 25–44 age group (38.8%), followed by the 45–64 group (33.8%).Senior respondents in their sixties or older (14%) were as many asyounger ones aged 24 or below (13.5%). About 41.1% of respon-dents held a tertiary degree, 31.2% possessed a high school qualifi-cation, and 27.6% are at a lower education level (junior high schoolor below). For personal income, 17.5% of respondents earned lessthan HK$8,000 (HKD/USD: 7.8/1.0) per month, and 22.0% hadHK$8,000–$14,999. There were 16.1% of respondents in the highestincome group, i.e. $25,000 or above, and 15.1% in the $15,000–$24,999 group. Non-income earners, including retirees,housewives, job seekers and students, accounted for 27.1%. Nearlytwo-third (62.5%) had heard of stonewall trees prior to participat-ing in the survey, whereas 37.5% was unaware. The average lengthof residence was 18.3 years.

Attitude toward stonewall trees

Descriptive statistics for each survey statement about publicattitude toward stonewall trees are reported in Table 2. The major-ity of respondents acknowledged the aesthetic contributions ofstonewall trees (64.5%) and their historical significance (78.9%).Many of them were concerned about the unrecoverable damageon these trees (69.1%). The rather high average scores range from3.55 to 3.87. Results indicate a generally high level of interestand concern among Hong Kong people about these natural ele-ments embedded within the city.

Support for preserving these trees was moderately strong.Around half of the respondents accepted the responsibility for pro-tection (56%) and recognized the uniqueness of this public asset(48.9%). Fewer believed that people they knew of would supportpreservation (41.9%) and assigned higher priority to this endeavor

Table 2Descriptive statistics and factor loading for attitude toward stonewall trees.

Statement

Stonewall trees can improve the outlook of the communityStonewall trees are part of the community history that money cannot buyDamage on stonewall trees is not recoverableIt is our moral responsibility to protect stonewall treesStonewall trees are a unique urban treasureI think people I know would support preserving stonewall treesPreserving stonewall trees is more important than urban redevelopment

Eigenvalues% of variance explained

a Percentage of respondents indicating ‘agree’ or ‘strongly agree’.b Measured on a five-point scale: 1 = ‘strongly disagree’, 5 = ‘strongly agree’.

Table 3Descriptive statistics and factor loading for perceived importance of tree functions.

Statement %Im

Trees can provide shading on hot days 9Trees can sequester carbon dioxide and mitigate greenhouse effect 9Greening rooftops can cool down building and reduce indoor temperature 7Trees can reduce urban temperature and avoid overheating 9Urban green spaces can trap rainwater and avoid street flooding 5Trees can dampen winds and reduce damage on urban infrastructure and building 5Trees can provide shelter on rainy days 3

Eigenvalues% of variance explained

a Percentage of respondents indicating ‘Important’ or ‘Very important’.b Measured on a five-point scale: 1 = ‘Not at all important’, 5 = ‘Very important’.

than urban redevelopment (40.0%). A possible reason for themoderate support is the lack of prior knowledge about stonewalltrees among some 37.5% of the sample, as reported in the previoussection. Consequently, the call for indicating support to preserva-tion was met with hesitation, uncertainty or apathy, which is evi-denced by some 30–40% of ‘neutral’ responses across the last fouritems in Table 2. The potential costs of preservation and mainte-nance, which would be shared among all taxpayers in Hong Kong,might also contribute to the moderate level of public support.

Factor analysis extracted only one factor, suggested that theseattitude questions can be reduced to one dimension. A compositescale was created by combining all seven items; it has a Cronbach’salpha of 0.84 and therefore considered to be reliable. Higher scoresdenote more positive view.

Perceived importance of tree functions

Table 3 suggests that respondents expressed general recogni-tion of main ecosystem services provided by urban trees, whichhas two dimensions. More than 95% of respondents affirmed therole of urban trees in providing shading on hot days and mitigatinggreenhouse effect. The vast majority (93.6%) were impressed bythe cooling effects of trees, although fewer (76.1%) consideredthe role of rooftop greening in cooling down buildings to be impor-tant. All of these items recorded a mean score of 4 or close to 4.Factor analysis confirmed that these four items, concerning themoderation of indoor, ambient or global temperature, convergedtoward one dimension (Factor 1). The four items formed a reliableadditive scale, with a Cronbach’s alpha of 0.69. Higher scoresdenote greater perceived importance.

% Agreea Averageb S.D. Factor loading

64.5 3.55 0.86 .73378.9 3.87 0.82 .74969.1 3.68 0.85 .71456.0 3.47 0.88 .75348.9 3.33 0.90 .79841.9 3.26 0.83 .72440.0 3.22 0.91 .555

3.64652.091

portantaAverageb S.D. Factor loading

Factor 1 Factor 2Temperaturerelated

Non-temperaturerelated

5.2 4.32 0.62 0.7955.8 4.41 0.62 0.7336.1 3.92 0.80 0.5083.6 4.28 0.63 0.7745.1 3.51 0.89 0.7823.9 3.46 0.97 0.7437.2 3.12 0.98 0.622

2.096 1.69529.947 24.215

A.Y. Lo, C.Y. Jim / Cities 42 (2015) 130–141 137

On the other hand, non-cooling functions attracted lower ranks,as indicated by the lower mean scores ranging from 3.12 to 3.51.More than half of the respondents (55.1%) acknowledged thecapacity of urban trees for rainwater interception and therebyavoiding street flooding, but nearly one-third of them were unde-cided. A comparable number of individuals (53.9%) held a positiveview about the wind-dampening function of urban trees whichcould minimize damage on infrastructure and buildings. Shelteragainst heavy rains found support from only 37.2% of respondents.These three items loaded on the same factor (Factor 2), althoughscale reliability was modest (Cronbach’s alpha = 0.57).

Community attachment

As shown in Table 4, the eight questions that elicited commu-nity attachment were divided into two factors. The first factor,place attachment, portrays connections between people and place.Most of respondents were familiar with the community theycurrently lived in (83.1%) and emotionally engaged with it(82.1%). Over half of them expressed interest in learning about

Table 4Descriptive statistics and factor loading for community attachment.

Statement % Agreea

I am familiar with the community I live in 83.1I have deep emotional attachment to my community 82.1I am interested in learning about the history of my community 59.8I am very concerned about the social affairs of my community 56.5I hope that old buildings in my community will be preserved 52.6I have good relationship with people in my community 60.6I have many friends in my community 58.1My community is full of human touch 57.7

Eigenvalues% of variance explained

a Percentage of respondents indicating ‘agree’ or ‘strongly agree’.b Measured on a five-point scale: 1 = ‘strongly disagree’, 5 = ‘strongly agree’.

Table 5Attitude toward stonewall trees by sampling area.

Sampling area N Average

Central & Western District 187 25.47Other Districts 583 24.10

** p < .01.

Table 6Descriptive statistics for model variables.

Description Ra

Demographic variablesPlace of residence (1 = CW District, 0 = other districts) 0–Sex (1 = male, 0 = otherwise) 0–Age (1 = below 45, 0 = otherwise) 0–Education (1 = tertiary degree, 0 = otherwise) 0–Personal Income (1 = > HK$15,000, 0 = otherwise) 0–Prior knowledge about stonewall trees (1 = yes/little, 0 = otherwise) 0–Length of residence in current district (number of years) 0.8

Attitudinal variablesAttitude toward stonewall trees 11Importance of tree functions – temperature related 6–Importance of tree functions – non-temperature related 4–Community attachment – place attachment 5–Community attachment – social ties 3–

the history of their community (59.8%) and concern about localaffairs (56.5%). A comparable proportion of respondents (52.6%)agreed that old buildings in the community were worth preserving.The composite scale consisting of these five items yielded asatisfactory alpha value of 0.72. Higher scores denote higher levelof attachment.

The remaining three questions loaded on the ‘social ties’ factor.Over half of the respondents believed that they had a goodrelationship with neighbors (60.6%) and had many friends in thecommunity (58.1%). The statement that ‘My community is full ofhuman touch’ received agreement from 57.7% of respondents.These three items formed a reliable scale with an alpha value of0.78.

Linear regression

Initial analysis compared the average scores of attitude towardstonewall trees between the CW District and other Districts.Results suggested that CW residents indicated stronger affinityfor these trees than the rest of the sample (Table 5). A regression

Averageb S.D. Factor loading

Factor 1 Factor 2Place attachment Social ties

3.61 0.78 0.5573.95 0.66 0.7233.56 0.80 0.7993.54 0.77 0.7453.47 0.87 0.4723.61 0.77 0.8633.55 0.84 0.8523.56 0.79 0.663

2.426 2.16230.324 27.028

S.D. 95% Confidence interval F statistics

4.127 24.88–26.07 14.530**

4.321 23.75–24.45

nge Average S.D. Cronbach’s alpha

1 0.24 0.43 –1 0.53 0.50 –1 0.52 0.50 –1 0.41 0.49 –1 0.31 0.46 –1 0.63 0.48 ––68 18.30 12.71 –

–35 24.43 4.31 0.8420 16.93 1.93 0.6915 10.09 2.08 0.5725 18.14 2.68 0.7215 10.72 2.00 0.78

Table 7Regression analysis for attitude toward stonewall trees.

Unstandardized coefficient (standard error in parentheses)

Full sample Residents of the CW District only Residents of other District

Demographic variablesPlace of residence (CW District) 0.749*

(0.340)Education (tertiary) 0.187 0.233 0.092

(0.330) (0.616) (0.398)Age (below 45) �0.669* 0.242 �0.875*

(0.333) (0.646) (0.392)Sex (male) 0.125 �0.115 0.145

(0.293) (0.555) (0.349)Income (>HK$15,000) 0.204 �0.044 0.385

(0.326) (0.590) (0.393)Length of residence 0.000 �0.018 0.009

(0.012) (0.019) (0.016)Prior knowledge �0.510 �0.043 �0.663

(0.302) (0.599) (0.352)Attitudinal variablesImportance of tree functions (temperature related) 0.192* �0.077 0.261**

(0.084) (0.161) (0.099)Importance of tree functions (non-temperature related) 0.387** 0.385** 0.410**

(0.074) (0.129) (0.090)Place attachment 0.531** 0.804** 0.448**

(0.068) (0.134) (0.080)Social ties �0.112 �0.041 �0.128

(0.086) (0.183) (0.099)(Constant) 8.716 8.593 8.876

(1.596) (2.938) (1.900)

Adj. R2 0.20 0.27 0.16F 17.586 7.830 11.761Sig. 0.000 0.000 0.000N 750 184 566

* p < .05.** p < .01.

138 A.Y. Lo, C.Y. Jim / Cities 42 (2015) 130–141

model was constructed to identify explanatory factors. Dummyvariables were created for demographic variables, as shown inTable 6 which displays the descriptive statistics of model variables.

Regression analysis of the full sample confirmed the explana-tory power of the place of residence variable, controlling for otherfactors (Table 7). Another significant factor was respondent’s age:younger individuals were less likely than the older ones to ratestonewall trees as important. Other demographic variables, includ-ing length of residence, demonstrated limited impacts on thedependent variable. Multi-collinearity was not observed as all pre-dictors yielded a Tolerance value exceeding .10 and a VarianceInflation Factor (VIF) below 5.

Perception of the functions of urban tree, whether related totemperature or not, was linked to attitude toward stonewall trees.Both of the perceived importance variables produced positiveeffects, suggesting that those who appreciated the environmentalcontributions of urban trees articulated a stronger preference forstonewall trees specifically. Apart from perceived benefits, placeattachment also demonstrated linkage. Affective bonding of peopleto the place created significant positive impacts. Bonding to otherpeople, i.e. social ties, did not attain statistical significance. Theregression model explained about 20% variations in the dependentvariable.

In a further analysis, the full sample was split into two groups,namely, residents of the CW Districts and other Districts (Table 7).There are no major differences between the two alternativemodels, except for age and perceived importance of temperature-related functions losing explanatory power in the CW-only sub-sample. It is important to note that effects of place attachmentoperated across the two sub-samples, despite the fact that mostof stonewall trees are found in the CW District. This means that

those individuals who were emotionally and culturally engagedwith the community where they lived nurtured a more positiveview about stonewall trees, regardless of location.

Discussion and conclusion

Mature urban trees dwelling on old masonry walls are a rareurban treasure. This study is the first to document public attitudetoward this unique form of urban green space. It has shown thatmembers of the Hong Kong public generally appreciated the exis-tence and intrinsic worth of stonewall trees. Many of themapplauded efforts on preserving this valuable natural-cum-cultural asset, although some hesitated to accept the moralresponsibility and costs of protection at the expense of urbanredevelopment. Younger adults aged below 45 harbored doubtsabout value of the stonewall trees, whereas the older ones weremore sympathetic. This corroborates Jim and Shan’s (2013)findings from another Chinese city that elderly embraced morepositive views toward urban green space, but stands at odds withSwanwick’s (2009) observation from the U.K. that those aged 25–44 requested greater access to local green spaces than other agegroups.

Nature-in-city elements can be characterized by their positionin a gradient based on the degree of naturalness (Kowarik, 1990).Formal urban parks tend to have a low score on this gradient,whereas natural woodlands tend to have high score. Stone wallscolonized by trees in urban areas denote a hanging urban forestthat can secure a rather high score. Older residents tend to havebetter understanding and appreciation of this spontaneous natu-ral endowment than younger members. Nature readily availableat close quarters within built-up areas is more relished by older

A.Y. Lo, C.Y. Jim / Cities 42 (2015) 130–141 139

people, as they are likely to have greater emotional attachmentto the community, which can be translated into affinity for nat-ure. The higher level of exposure to environmental educationamongst the younger generations through formal school curricu-lum in their formative years (Suggate et al., 1998) does not seemto have imbued a stronger sense of nature admiration andenjoyment.

Despite expression of reasonably strong support to protectstonewall trees, when respondents are asked to choose betweenprotection and redevelopment, the same support has waned by anotable margin. The pragmatic urge to satisfy housing needs in acity that is beset chronically by grave deficiency of readily develop-able land and affordable housing has subliminally weakened thehigh-order conservation goal. The routine obliteration of naturein local urban development and the excessively manicured urbanparks as poor substitute of nature may have muffled the innatenature empathy. With respect to support of stonewall tree protec-tion, respondents rate themselves higher than people they know.The results indicate an interesting discrepancy, according moreconfidence to oneself than others with reference to the specificnature conservation issue.

The findings lend support to our research hypothesis. Residentsof the CW District registered stronger commitment than the rest ofthe city. Place attachment is related to preference for or access togreen elements in cities (Arnberger & Eder, 2012; Budruk et al.,2009; McCunn & Gifford, 2014). Broadly consistent with thesestudies, we found a positive relationship between perception ofurban trees and place attachment. Affinity for stonewall trees is afunction of the affective bonding between people and the commu-nity they live. Deep engagement and substantial cultural experi-ence with local community elevated the desire for preservingstonewall trees. In a similar vein, Jorgensen and Anthopoulou(2007) have noted in the U.K., the elderly particularly valued urbanwoodland for their links with the past. Our findings also resonatewith Lo and Jim’s (2010) study that people’s concern about thewider community influences their preference for urban greenspaces. The image of old masonry walls and associated trees isetched on the perception of a community. Affinity for this uniqueform of urban green space can be explained in terms of the rela-tionship between people and communal space (Burgess,Harrison, & Limb, 1988; Konijnendijk, 2008).

Apart from the people–place relationship, perception of theecosystem-service functions of urban trees also plays a role. Thissuggests that stonewall trees are recognized not only as a cul-tural endowment, but also as a source of material benefits. Whilethe cultural dimension tends to be place-specific, the utilitarianone is ‘conceptual’ (Ryan, 2005) in the sense that these benefitsare not limited to this form of tree assemblages. Good relation-ships with people in the same community, however, did notstrengthen affinity. A possible reason is that most of the stone-wall trees stand along roadsides or in tiny pocket parks withlow-quality physical space that is not conducive to prolongedand intimate social interactions. Unlike urban neighborhoodparks, therefore, these green spaces are unable to serve a sub-stantive social role (Coles & Bussey, 2000; Lo & Jim, 2010;Ward Thompson, 2002).

The influence of residential place attachment did not distin-guish between local and non-local residents. What makes a differ-ence is the stronger attachment to the community where theylived, whether being a member of the CW District or not. Placespecificity might contribute to the more positive view of local res-idents. Yet the disposition of non-local residents was probably aconceptual or generic one (Lin & Lockwood, 2014; Ryan, 2005),

as it was linked to attachment to their own community, outsidethe CW District, where old masonry walls favorable for plantgrowth are sporadic if not completely absent. Community-engagednon-local residents appreciated the existence of a type of land-scape that is regarded as a valuable asset of a community, eventhough they were not a residential member of this community.Community attachment, regardless of the location of concern,could motivate support for the conservation of stonewall trees.Appeal to the bonding between people and place can reach bothlocal and non-local residents.

The findings have implications for soliciting public supportfrom the larger society. In Hong Kong, conflicts between preserv-ing heritage trees and redeveloping the dilapidated city core con-tinue to linger. The public is aware of the chronic shortage ofdeveloping land and frustrated at the extremely expensive realestate. As a result, the general attitude is to utilize every piece ofurban land, whereas amenity tree planting and nature conserva-tion in the city are given a low priority (Jim, 1998a, 2003; Lo &Jim, 2012). Mounting pressures for urban development andrenewal in recent years threaten to displace some unrecoverableheritage landscape. While some local residents are motivated toact and protest against adverse changes in the living environmentforced upon their community, the rest of the society who have lit-tle immediate material interest might remain somewhat passive.However, as we have shown, non-local residents are likely toextend their support to the extent in which they are deeplyengaged in their own community, regardless of the location andlength of residence. Attachment to their own place predisposesnon-local residents to a positive perception of a landscape thatsymbolizes the historical and cultural uniqueness of a community,such as the western part of the CW District, which was the earliestcolonial settlement in Hong Kong where a relatively large numberof old masonry walls carrying mature trees stand. Non-local resi-dents with a strong sense of community tend to embrace conser-vation of this natural-cum-cultural asset. The altruistic extensionof appreciation and support across district boundaries could bemobilized to promote other conservation causes that are loca-tion-specific.

Community elites and environmentalists need to augmentsupport from both local and non-local residents in defending nat-ure conservation initiatives that may potentially incur significantcosts to taxpayers. Hong Kong people habor a rather utilitarian ifnot indifferent attitude toward urban trees conservation (Lo &Jim, 2012; Lo, 2012) as well as wider environmental issues (Lee,2003). Citywide conservation campaigns could benefit from artic-ulating the endeavor in terms of broader social and communityimperatives, and move well beyond the narrow focus on ecologi-cal or historical value. Preserving a place offers an additionalargument for citizens to participate in an essentially environmen-tal campaign to wean the deeply-entrenched mode of urbandevelopment and renewal from conservation lethargy and igno-rance. Such place-based attempts can help address the pressingneed for raising public awareness and recognition that heritagetrees are worth preserving as an inheritance of the community(Jim, 2005, 2014).

Some aspects of this research warrant further investigation. Forexample, the survey was only administered in selected CouncilDistricts in Hong Kong. A geographically more comprehensivesampling strategy could increase the demographic representativesof the sample. Furthermore, many participants of the surveyproved to be unfamiliar with stonewall trees. Survey designerswould need to provide potential respondents more backgroundinformation to facilitate informed response.

Appendix: Age distribution in census and sample by sampling area

District Age group Chi-Square^

Census (% of population) Sample (% of respondents)

15–24 25–44 45–64 >65 18–24 25–44 45–64 >65

Hong Kong IslandCentral & Western 10.7 34.3 29.7 13.9 14.2 42.6 32.1 11.1 4.168Eastern 10.8 30.1 33.1 15.7 10.0 40.0 38.6 11.4 2.637Southern 11.9 30.4 33.3 13.3 17.5 30.0 37.5 15.0 0.712

New TerritoriesKwai Tsing 12.8 30.6 31.1 14.4 11.7 40.0 33.3 15.0 0.953Tuen Mun 13.2 32.3 33.9 9.5 13.8 36.3 36.3 13.8 0.842North 14.5 31.0 32.3 10.2 22.5 30.0 35.0 12.5 1.251Sha Tin 12.4 31.9 33.6 11.3 11.3 36.3 38.8 13.8 0.509Sai Kung 13.0 34.5 31.0 8.9 10.0 50.0 30.0 10.0 3.150

KowloonYau Tsim Mong 10.1 34.7 28.6 14.2 8.0 40.0 36.0 16.0 0.358Shum Shui Po 11.8 29.6 30.6 16.5 8.6 40.0 31.4 20.0 1.180Kowloon City 11.2 31.7 30.2 15.4 20.0 26.7 26.7 26.7 3.949Wong Tai Sin 12.8 27.8 32.3 16.7 17.5 37.5 32.5 12.5 1.778Kwun Tong 12.3 29.4 31.0 15.9 14.0 36.0 28.0 22.0 1.276

^ All values not significant at 5% level.

140 A.Y. Lo, C.Y. Jim / Cities 42 (2015) 130–141

References

Arnberger, A., & Eder, R. (2012). The influence of green space on communityattachment of urban and suburban residents. Urban Forestry & Urban Greening,11(1), 41–49.

Beer, A. R., Delshammar, T., & Schildwacht, P. (2003). A changing understanding ofthe role of greenspace in high-density housing. Built Environment, 29(2),132–143.

Brehm, J. M., Eisenhauer, B. W., & Krannich, R. S. (2006). Community attachments aspredictors of local environmental concern: The case for multiple dimensions ofattachment. American Behavioral Scientist, 50(2), 142–165.

Budruk, M., Thomas, H., & Tyrrell, T. (2009). Urban green spaces: A study of placeattachment and environmental attitudes in India. Society & Natural Resources,22(9), 824–839.

Burgess, J., Harrison, C. M., & Limb, M. (1988). People, parks and the urban green: Astudy of popular meanings and values for open spaces in the city. Urban Studies,25(6), 455–473.

Casakin, H., Hernández, B., & Ruiz, C. (in press). Place attachment and place identity inIsraeli cities: The influence of city size. Cities. http://www.doi.org/10.1016/j.cities.2014.07.007.

Census and Statistics Department (2013). Hong Kong Monthly Digest of Statistics: Theprofile of Hong Kong Population analysed by District Council District, 2012. HongKong Special Administrative Region Hong Kong.

Chiesura, A. (2004). The role of urban parks for the sustainable city. Landscape andUrban Planning, 68, 129–138.

Chiu, R. L. H. (2006). Planning, land and affordable housing in Hong Kong. HousingStudies, 22(1), 63–81.

Coles, R. W., & Bussey, S. C. (2000). Urban forest landscape in the UK – Progressingthe social agenda. Landscape and Urban Planning, 52, 181–188.

Environmental Protection Department (2006). Sustainable development for the 21stcentury: Environmental baseline study reports. Hong Kong Government, HongKong.

Jim, C. Y. (1998a). Impacts of intensive urbanization on trees in Hong Kong.Environmental Conservation, 25(2), 146–159.

Jim, C. Y. (1998b). Old stone walls as an ecological habitat for urban trees in HongKong. Landscape and Urban Planning, 42, 29–43.

Jim, C. Y. (2002). Planning strategies to overcome constraints on greenspaceprovision in urban Hong Kong. Town Planning Review, 73(2), 127–152.

Jim, C. Y. (2003). Protection of urban trees from trenching damage in compact cityenvironments. Cities, 20(2), 87–94.

Jim, C. Y. (2005). Monitoring the performance and decline of heritage trees in urbanHong Kong. Journal of Environmental Management, 74, 161–172.

Jim, C. Y. (2013). Drivers for colonization and sustainable management of tree-dominated stonewall ecosystem. Ecological Engineering, 57, 324–335.

Jim, C. Y. (2014). Ecology and conservation of strangler figs in urban wall habitats.Urban Ecosystems, 17, 405–426.

Jim, C. Y., & Chen, W. Y. (2010). Habitat effect on vegetation ecology and occurrenceon urban masonry walls. Urban Forestry & Urban Greening, 9(3), 169–178.

Jim, C. Y. (2010). Old masonry walls as ruderal habitats for biodiversity conservationand enhancement in urban Hong Kong. In N. Müller, P. Werner, & J. G. Kelcey(Eds.), Urban biodiversity and design. Chichester, UK: Wiley-Blackwell.

Jim, C. Y., & Shan, X. (2013). Socioeconomic effect on perception of urban greenspaces in Guangzhou, China. Cities, 31, 123–131.

Jorgensen, A., & Anthopoulou, A. (2007). Enjoyment and fear in urban woodlands –Does age make a difference? Urban Forestry and Urban Greening, 6, 267–278.

Kaltenborn, B. P., & Bjerke, T. (2002). Associations between landscape preferencesand place attachment: A study in Røros, Southern Norway. Landscape Research,27(4), 381–396.

Konijnendijk, C. C. (2008). The forest and the city: The cultural landscape of urbanwoodland. Dordrecht: Springer.

Kowarik, I. (1990). Some responses of flora and vegetation to urbanization inCentral Europe. In H. Sukopp, S. Hejny, & Kowarik (Eds.), Urban ecology: Plantsand plant communities in urban environments (pp. 45–74). The Hague, TheNetherlands: SPB Academic.

Lee, Y. S. F. (2003). Environmental consciousness in Hong Kong. Southeast AsianStudies, 41(1), 15–35.

Lin, C.-C., & Lockwood, M. (2014). Forms and sources of place attachment: Evidencefrom two protected areas. Geoforum, 53, 74–81.

Lo, Alex Y. (2012). The encroachment of value pragmatism on pluralism: Thepractice of the valuation of urban green space using stated-preferenceapproaches. International Journal of Urban and Regional Research, 36(1),121–135.

Lo, A. Y., & Jim, C. Y. (2010). Differential community effects on perception and use ofurban greenspaces. Cities, 27(6), 430–442.

Lo, A. Y., & Jim, C. Y. (2012). Citizen attitude and expectation towards greenspaceprovision in compact urban milieu. Land Use Policy, 29(3), 577–586.

Lokocz, E., Ryan, R. L., & Sadler, A. J. (2011). Motivations for land protection andstewardship: Exploring place attachment and rural landscape character inMassachusetts. Landscape and Urban Planning, 99(2), 65–76.

Low, S. M. (1992). Symbolic ties that bind: Place attachment in the plaza. In I.Altman & S. M. Low (Eds.), Place attachment (pp. 165–185). New York.

Low, S. M., & Altman, I. (1992). Place attachment: A conceptual inquiry. In I. Altman& S. M. Low (Eds.), Place attachment (pp. 1–12). New York: Plenum Press.

McCunn, L. J., & Gifford, R. (2014). Interrelations between sense of place,organizational commitment, and green neighborhoods. Cities Part A, 41(0),20–29.

Miller, R. W. (1997). Urban forestry: Planning and managing urban greenspaces (2nded.). Englewood Cliffs, NJ: Prentice Hall.

Ng, M. K. (2002). Property-led urban renewal in Hong Kong: Any place for thecommunity? Sustainable Development, 10, 140–146.

Planning Department (2013). Planning statistics: Land utilization in Hong Kong. HongKong Government, Hong Kong. <http://www.pland.gov.hk/pland_en/info_serv/statistic/landu.html> Accessed 09.06.14.

Ryan, R. L. (2005). Exploring the effects of environmental experience on attachmentto urban natural areas. Environment and Behavior, 37(1), 3–42.

Stedman, R. C. (2006). Understanding place attachment among second homeowners. American Behavioral Scientist, 50(2), 187–205.

Suggate, B. B., Paul Hart, K. P., Ho, R. K. P., Ofwono-Orecho, J. K. W., Peries, M.,Robottom, I., Tsalikii, E., & Van Staden, C. (1998). An overview of significantinfluences and formative experiences on the development of adults’environmental awareness in nine countries. Environmental Education Research,4, 445–464.

A.Y. Lo, C.Y. Jim / Cities 42 (2015) 130–141 141

Swanwick, C. (2009). Society’s attitudes to and preferences for land and landscape.Land Use Policy, 26. Land Use Policy, 26(Suppl. 1), S62–S75.

Tapsuwan, S., Leviston, Z., & Tucker, D. (2011). Community values and attitudestowards land use on the Gnangara groundwater system: A sense of place studyin Perth, Western Australia. Landscape and Urban Planning, 100(1–2), 24–34.

Tuan, Y. F. (1974). Topophilia: A study of environmental perception, attitudes, andvalues. Englewood Cliffs, NJ: Prentice-Hall.

Tuan, Y. F. (1977). Space and place: The perspective of experience. Minneapolis:University of Minnesota Press.

Vaske, J. J., & Kobrin, K. C. (2001). Place attachment and environmentallyresponsible behavior. The Journal of Environmental Education, 32(4), 16–21.

Walker, A. J., & Ryan, R. L. (2008). Place attachment and landscape preservation inrural New England: A Maine case study. Landscape and Urban Planning, 86(2),141–152.

Ward Thompson, C. (2002). Urban open spaces in the 21st century. Landscape andUrban Planning, 60, 59–72.