How important is green infrastructure in small and medium-sized towns? Lessons from South Africa

G Model

ARTICLE IN PRESS

LAND-3050; No. of Pages 9

Landscape and Urban Planning xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Landscape and Urban Planning journal homepage: www.elsevier.com/locate/landurbplan

How important is green infrastructure in small and medium-sized towns? Lessons from South Africa C.M. Shackleton ∗ , A. Blair, P. De Lacy, H. Kaoma, N. Mugwagwa, M.T. Dalu, W. Walton Department of Environmental Science, Rhodes University, Grahamstown 6140, South Africa

h i g h l i g h t s • • • •

Most greens spaces and GI are under private tenure. There are spatial inequities in the distribution of GI within towns. There are multiple consumptive and non-consumptive uses of GI. Urban residents are willing to participate or contribute to the maintenance of GI.

a r t i c l e

i n f o

Article history: Received 17 August 2015 Received in revised form 29 November 2016 Accepted 14 December 2016 Available online xxx Keywords: Cultural ecosystem services Gardens Green space Parks Private green infrastructure Provisioning ecosystem services Willingness-to-pay

a b s t r a c t The current nature and challenges of urbanisation in sub-Saharan Africa display several unique features only weakly evident on other continents. Key ones include the current high rates of population growth, inadequate planning and governance systems, concentration in small and medium-sized towns, and increasing urban poverty. These shape the extent, nature and use of ecosystem services provided by urban green infrastructure. This paper first examines the location of green infrastructure across nine towns, showing that it is unequal between suburbs and that the bulk is located under private tenure (74%) rather than in public spaces. We then consider the extent and patterns of use of selected provisioning and cultural ecosystem services from green infrastructure in different locations within towns, including private gardens, public parks and street trees. The results show significant use of green infrastructure for a range of provisioning and cultural services as well as its contribution to spiritual and mental wellbeing. Provisioning contributions are both in regular support of livelihood needs as well as increased use after a covariate shock (a flood), both of which help reduce household vulnerability. Lastly, our results show the expressed level of support and willingness-to-pay or work amongst urban residents for green infrastructure and the services it provides. Whilst the composite results indicate marked variation between and within towns, they show that there is widespread use of green infrastructure for both basic needs as well as for more aesthetic and psycho-spiritual appreciation and recreation, in small and medium-sized towns in a developing country such as South Africa. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Urbanisation is a multidimensional process that manifests as increasing human population densities accompanied by changing land cover from one dominated by vegetation or waterways to one dominated by built structures (Elmqvist, Alfsen, & Colding, 2008). As such, it is a profound driver of ecological and socioeconomic change. Sub-Saharan Africa is currently experiencing the highest rates of urbanisation anywhere on the globe, with urban popula-

∗ Corresponding author. E-mail address: [email protected] (C.M. Shackleton).

tions having increased from 14% of the sub-continental population in 1950 to 40% in 2010 and will surpass 50% within the next two decades (United Nations, 2014). Despite the institutionalised suppression of urbanisation in South Africa for almost four decades in the second half of the 20th century, it is already well ahead of this ratio, with an urban population of approximately 64% (World Bank, 2015). The rapidly increasing urban populations and areas in subSaharan Africa demand that urgent attention is given to the environmental and socioeconomic consequences of this demographic and spatial transition. The locus of poverty and vulnerability in sub-Saharan Africa is inexorably adopting an urban visage as many urban centres struggle to accommodate, absorb and

http://dx.doi.org/10.1016/j.landurbplan.2016.12.007 0169-2046/© 2016 Elsevier B.V. All rights reserved.

Please cite this article in press as: Shackleton, C. M., et al. How important is green infrastructure in small and medium-sized towns? Lessons from South Africa. Landscape Urban Plan. (2016), http://dx.doi.org/10.1016/j.landurbplan.2016.12.007

G Model LAND-3050; No. of Pages 9 2

ARTICLE IN PRESS C.M. Shackleton et al. / Landscape and Urban Planning xxx (2016) xxx–xxx

provide for their rapidly burgeoning populations (Angel, Parent, Civco, Blei, & Potere, 2011). Discourses on the transformations required to address the challenges posed by rapid urbanisation are dominated by analyses of large cities which inevitably are national economic hubs. However, this belies the reality that most urban growth in sub-Saharan Africa is occurring in medium- and smallsized cities and towns of less than half a million inhabitants (Cohen, 2006; United Nations, 2014), and that collectively they are already home to more people than those living in cities of one million people or more (Schlesinger, Shackleton, & Drescher, 2015). The models and prescripts of urban development and sustainability derived and examined in large cities may well be found wanting in these smaller ones because of their lower revenue and skills bases (Gwedla & Shackleton, 2015; Pickett et al., 2013; Wisner et al., 2015) as well as their potentially smaller ecological footprint. Until recently, discourses and design and policy proposals to reduce urban ecological footprints and promote sustainability focused on reducing consumption and waste generation, within the paradigm of the so-called sanitary city (Pickett et al., 2013). Within the last decade or two, the potential role of green infrastructure (GI) in contributing cost-effectively to urban sustainability has gained acceptance. This has paralleled the paradigm shift embodied in ecosystem services thinking, which together have revolutionised the foundational discipline of urban ecology. In this context GI is taken to be “the connected network of multifunctional, predominantly unbuilt, space that supports both ecological and social activities and processes” (Kambites & Owen, 2006). Core considerations of the GI concept is that it relates to (1) urban green areas, (2) that are networked and (3) multifunctional (Albert & von Haaren, 2014; Hansen & Pauleit, 2014; Sandström, 2002; Wright, 2011), and as such provide a range of ecosystem goods and services to urban dwellers. Whilst some authors emphasise GI as a planning approach (Pauleit, Liu, Ahern, & Kazmierczak, 2011; Sandström, 2002) many deem it as the physical green spaces and trees on the ground and the corridors between them that provide multiple ecosystem goods and services at various scales (Tzoulas et al., 2007). This view might be more pronounced in developing countries where integrated planning approaches are challenged by the rapid urbanisation rates and high poverty levels. A fourth dimension not explicitly covered in the above definition is the importance of GI in building both social and ecological resilience. Most commonly urban design passively retains some vegetated surfaces to provide some ecosystem services, typically regulating ones such as storm-water attenuation and air purification, or for recreational purposes. Alternatively, designers and planners can proactively incorporate and mould GI into urban designs to simultaneously optimise resilience and sustainability through the ecosystem services provided and reduce the need for and costs of built infrastructure (Ahren, 2007). However, according to Schäffler and Swilling (2013), examples of this are not commonplace. This is perhaps not surprising due to the lack of suitable tools and that the concept, principles and benefits of GI are still relatively poorly disseminated and understood within the planning profession (Albert & von Haaren, 2014; Wright, 2011). The transition to mainstreaming GI into urban design requires both an evolution in urban design principles as well as in urban governance and budgeting processes and structures (Andersson et al., 2014). This is likely to be more challenging in small and medium-sized towns and cities in sub-Saharan Africa because of their generally weaker planning and regulatory institutions and the higher levels of poverty (Wisner et al., 2015). Yet paradoxically, mainstreaming GI could well save infrastructure development and maintenance costs thereby allowing greater funding for social services in such towns. Indeed, smaller urban centres are less path-dependent (i.e. can be more flexible in development trajectories) and therefore, with appropriate vision, have the potential to be more resilient than their larger counterparts (Pickett et al., 2013).

Given the paucity of understanding of GI provision and use in (i) sub-Saharan Africa and (ii) small and medium-sized towns, the objective of this paper was to examine the extent and use of GI in small South African towns using a social-ecological systems lens. Whilst the findings are specific to South Africa, the study has relevance to other sub-Saharan countries because of the focus on poorer and less well-developed towns and cities rather than the large and modern cities. Although the primary quantification of the potential of GI to provide services useful to human wellbeing is via green space abundance and distribution, we also include trees in private and public spaces and streets. To achieve this objective we considered the following key questions: (1) How much green space is there and where is it located? (2) What is the nature and magnitude of GI use by residents for provisioning services? (3) What is the nature and magnitude of GI use by residents for cultural services? and (4) How supportive are urban residents of public green space? Answers to these questions will provide some understanding of GI provision and use for multiple functions in these poorly studied settings and thereby contribute to GI planning and policy in such contexts. Because of the multidisciplinary, mixed-methods approach, the above questions are addressed sequentially by synthesis of results across a number of different studies in a range of small and medium-sized towns in South Africa. Specific methods are presented under each question. Because of the historical legacy of racially segregated planning and development in South Africa, we first briefly describe the spatial segregation that is apparent in most South African urban areas against which any analysis of GI and its benefits must be examined. 1.1. Historical segregation; contemporary separation The racially discriminatory patterns of urban development during the colonial and apartheid periods in South Africa have been well documented (Christopher, 2001). In brief, a plethora of racially biased legislation and ruthless enforcement dictated where black South Africans (which during apartheid included black Africans, Asians and those of mixed race parentage) must live and work and the social and infrastructural services that were provided in different areas. At a macro-scale, millions of black South Africans were required to live in ethnically defined and geographically disparate ‘bantustans’. These areas had little or no economic base and the apartheid government provided relatively little investment in infrastructure and services. Today many towns in the former bantustans are thriving hubs, yet massive backlogs in the provision of housing and services remain and the economic base is limited and so poverty and underdevelopment are a lot higher than in urban areas that were not located in the former bantustans. The racial profile also remains almost exclusively black South African, unlike the more cosmopolitan urban centres elsewhere in the country. The apartheid system also regulated where black South Africans outside of the bantustans could reside. In urbans areas they were required to live in discrete areas zoned as ‘townships’, often on the periphery of the city, or even some distance from it. Like the bantustans, there was no formal economic base in the townships and central government and urban authorities provided only the barest of infrastructure. After the demise of apartheid in the mid1990s, much national government investment has been targeted at improving the living conditions in the townships, but disparities remain in many spheres when compared to other urban residential areas (Donaldson, du Plessis, Spocter & Massey, 2013; Miraftab, 2007). The racial profile of residents in the townships remains almost exclusively black South African. Additionally, as part of the investment in improving living conditions, the post-apartheid government built millions of low-cost housing units in and around towns throughout the country (in and out of former bantustans) of uniform size and appearance, which colloquially are referred to

Please cite this article in press as: Shackleton, C. M., et al. How important is green infrastructure in small and medium-sized towns? Lessons from South Africa. Landscape Urban Plan. (2016), http://dx.doi.org/10.1016/j.landurbplan.2016.12.007

G Model LAND-3050; No. of Pages 9

ARTICLE IN PRESS C.M. Shackleton et al. / Landscape and Urban Planning xxx (2016) xxx–xxx

3

Fig. 1. The location of study towns in South Africa.

as RDP houses or RDP suburbs (named after the Reconstruction and Development Programme). Occupancy of RDP housing is supposed to be reserved for the indigent and is managed by urban municipalities. As a result of the above processes, most South African towns currently display at least three spatially discrete and visually distinctive residential areas (Kaoma & Shackleton, 2014a), namely (1) relatively new, low-cost RDP housing areas occupied by poor black South Africans, (2) townships with a mix of old and new housing occupied mostly by poor, and less poor, black South Africans and (3) suburbs ranging from middle to high income areas, occupied mostly by white South Africans, but with an increasing presence of other racial groups. In some towns, a fourth form is evident in the shape of informal housing areas, typically occupied by new migrants to a town in the hope of being allocated an RDP house (Hunter & Posel, 2012). In the meantime, they settle on vacant land on the edges of towns or apparently unused lands within the town and construct houses from low-cost or scavenged materials. In large and long-established informal areas some local municipalities provide some services, such as piped water, refuse removal, street lights and electricity. 2. Approach and results We present findings against each of the key questions in turn. We briefly outline the broad approaches and sample sizes, along

with supportive reference to published sources where available. The location and distribution of the towns in which we worked are provided in Fig. 1. 2.1. How much green space is there and where is it located? Nine towns in the Eastern Cape province were analysed by means of GIS to map the location of vegetated surfaces in relation to suburb and tenure as evident on 2009–2011 aerial photo images. They were selected to follow on the work of McConnachie, Shackleton, and McGregor (2008) who determined the area of public green space in these towns, but did not consider private green space. Each of the three formal residential forms were delineated and then a 1 km2 grid was superimposed over each (mean town size was 8.0 km2 ). Three grid cells were randomly selected per residential type per town. Within each randomly selected grid we used polygons to digitise all areas that were not under built infrastructure or water (which was negligible; <1% of surface area) as green cover, and differentiated as located on either private space (mostly residential gardens) or public space, using national Surveyor General erven data (equivalent to parcel data) as a guide. Across the nine towns approximately two-fifths (39.9%) of the land surface area was under some built infrastructure or nonvegetated surfaces (Fig. 2). This was highest in the RDP areas because of the higher density of housing units there (McConnachie & Shackleton, 2010). Thus, approximately 60% of the land surface

Please cite this article in press as: Shackleton, C. M., et al. How important is green infrastructure in small and medium-sized towns? Lessons from South Africa. Landscape Urban Plan. (2016), http://dx.doi.org/10.1016/j.landurbplan.2016.12.007

G Model

ARTICLE IN PRESS

LAND-3050; No. of Pages 9

C.M. Shackleton et al. / Landscape and Urban Planning xxx (2016) xxx–xxx

4

Bathurst

50

Bela-Bela Fort Beaufort

40 30

Private green

20

Public green Not green

Town

% o f la n d s u r fa c e a r e a

60

Grahamstown Phalabrowa Queenstown Tzaneen

10

Zeerust

0 Affluent

Township

RDP

Residenal area Fig. 2. The extent and distribution of green infrastructure in different residential areas of medium-sized towns in the Eastern Cape province, South Africa (n = 9).

area was under vegetated surfaces, or GI, supplying some level of ecosystem services. Of the GI, almost three-quarters (73.8%) was on private land and the balance was under municipal/state control. The ratios in the affluent and township areas were relatively similar, but were significantly different in the RDP areas (␹2 = 9.25; p < 0.01), where there was far less public GI, a characteristic already reported for formal recreational green space by McConnachie et al. (2008). The key finding here is that the majority of land in these towns is under some GI, and that most of the GI is under the control of individuals rather than the municipal authorities.

2.2. What is the nature and magnitude of GI for provisioning ecosystem services for household resilience? It is well known that the many urban communities in subSaharan Africa make use of a variety of provisioning goods, most notably firewood, medicinal plants and ‘wild’ foods (edible insects, bushmeat, uncultivated leafy vegetables and fruits, mushrooms) (e.g. Schlesinger et al., 2015). Yet the extent of use, means of procurement and the origin of these provisioning goods has been little studied (which echoes the general dearth of work on urban ecosystem services in sub-Saharan African cities (Haase et al., 2014)). We conducted three separate investigations; one in three small towns in the Eastern Cape province (Davenport, Gambiza, & Shackleton 2011), a second in three small towns in northern South Africa (Kaoma & Shackleton, 2014b), and another in two medium-sized towns of one each in the Eastern Cape and one in the north (Ward & Shackleton 2016). In the first study 200 households from the township suburbs were randomly selected (from houses numbered on the latest aerial photographs) and interviewed regarding their collection of wild resources from the undeveloped municipal lands on the periphery of town. In the second study 150 households were randomly selected (using aerial photographs), fifty each from the township, RDP and informal areas. In the last, 30–40 households were randomly sampled (using aerial photographs) in each of the town suburbs and the township. In each study any randomly selected household unwilling to participate (<3%) was replaced by the next randomly designated household on the list. In total, we had eight towns with data from townships, three from RDP areas, three from informal settlements and two from more affluent suburbs. Considering all the study towns, a mean of 64% township dwellers used at least one wild natural resource, ranging from 27% in Queenstown to 100% in Tzaneen (Fig. 3). Prevalence of use amongst residents in the four northern towns (83 ± 8.7%) was almost double that of the Eastern Cape towns (45 ± 10.2%) (t = 2.8; p < 0.03). These percentages would be somewhat lower if a greater representation of affluent suburbs had been surveyed.

0

20

40

60

80

100

Proporon (%) of township households Fig. 3. The proportion of township households in eight South African towns that use at least one wild natural resource.

For towns where we have data from all three suburbs, it was revealed that the prevalence of use of wild resources was not equal in different areas. Averaged across the three towns it was lowest (29% of households) in RDP areas and highest (82%) in the townships, with informal areas being intermediate between these two (59%). However, because of the greater affluence in townships (relative to RDP areas and informal ones), the proportion buying wild resources, rather than collecting their own, was a lot higher than in the RDP or informal areas. The origin of the purchased resources was uncertain, i.e. were they harvested from urban GI or imported from rural areas. Looking specifically at collected resources and where they were collected provided a more nuanced perspective (Fig. 4). In particular, township dwellers made less use of public urban GI to obtain the provisioning resources that they used, and far greater use of their own private homestead space than did residents of the RDP and informal areas. Overall, the non-cash income provided through the collection of urban sourced provisioning services represented approximately one-fifth of total household cash and non-cash incomes in the three settlements (Tzaneen – 17.7%; Bela–Bela – 16.1% and Zeerust – 26.8%). Not unsurprisingly it was a lot higher amongst households in informal settlements (33.1%) than either the RDP suburbs (16.5%) or townships (13.9%). Such household surveys indicate the prevalence of regular use of urban GI provisioning services. However, this underestimates their significance, especially to the urban poor, during times of shock or hardship. Consequently, we examined use of provisioning services as a coping response to the 2012/2013 floods in three Eastern Cape towns (Sachikonye, 2015). This was done via a questionnaire survey administered to between 70 and 85 households per town in informal settlements who had experienced some degree of physical damage to their dwelling (identified via snowball sampling). The coping responses included one or both of collecting locally sourced wood or plant fibres to patch or rebuild lost structures, or the local collection of any wild natural resources for sale explicitly to raise cash income to purchase building materials to patch or rebuild damaged structures (i.e. over and above any usual collection or trade in wild resources). We quantified the amount of natural resources used or sold and converted that to monetary values by multiplying it by the local price per unit. This was considered against other coping strategies such as selling assets to raise cash, borrowing money from family or friends, seeking additional employment and receipt of emergency aid. The relative monetary value of the contribution of wild resources harvested from GI to the overall suite of coping strategies was 57%, ranging from 46% in Port Alfred to 70% in Port St Johns, with Grahamstown intermediate between these two at 55%.

Please cite this article in press as: Shackleton, C. M., et al. How important is green infrastructure in small and medium-sized towns? Lessons from South Africa. Landscape Urban Plan. (2016), http://dx.doi.org/10.1016/j.landurbplan.2016.12.007

G Model

ARTICLE IN PRESS

LAND-3050; No. of Pages 9

C.M. Shackleton et al. / Landscape and Urban Planning xxx (2016) xxx–xxx

5

P r o p o r t io n o f h o u s e h o ld s ( % )

80 70 60 50 Firewood

40

Fruits

30

Medicines 20

Fencing wood

10 0 Informal

RDP

Township Informal

Collected from home space

RDP

Township

Collected from public urban spaces

Fig. 4. The proportion (%) of households in townships, RDP and informal settlements collecting firewood, fruits, medicines and fencing timber for consumptive use from private and public urban green spaces.

2.3. What is the nature and magnitude of GI use for cultural ecosystem services? We gathered data from several surveys to assess local respondents’ levels of satisfaction with the abundance and distribution of public green spaces within their town. Sample size varied between towns between the separate studies in relation to time and resources. For Fort Beaufort and Port Alfred we interviewed 30 randomly selected households (from aerial photographs) in each suburb (providing 90 per town) (Shackleton & Blair, 2013) via means of a structured questionnaire. The same approach was used in Grahamstown, but with 50 households per suburb (150 in total). In Kingwilliamstown 40 households were randomly selected (using aerial photographs) amongst those closest to each of five public green spaces (200 in total). The results show that even in small towns, where the distances to the rural (and presumably green) surroundings are relatively small, there is significant demand for urban GI for recreational purposes, with just over half of households stating that they use public green spaces to some extent (Table 1). Moreover, in most of the towns we have surveyed, there is a general dissatisfaction with the paucity of public green space available (Table 1). There was a tendency for increasing dissatisfaction about the limited abundance of green space with increasing poverty status of the local suburb (other than in Kingwilliamstown); almost two-thirds of residents in the RDP and township areas felt that there was insufficient public green space available, compared to only 48% of residents from affluent suburbs (␹2 = 5.88; p < 0.02). Overall, just over half the respondents felt that the provision of public urban green space was insufficient. The primary use of public green space was for recreation (>73% of responses), especially family outings and sports/exercise. Across the sample population, respondents enjoyed a mix of different types of public green spaces, including formal public green spaces with sports facilities to more natural, less manicured, green spaces with indigenous vegetation. Residents are also exposed to other forms of green infrastructure, such as in their working environment, school grounds, church yards and trees planted along streets. In these locations, the size of the green space is not a consideration. The benefits that people obtain from these have been little studied, but our results to date indicate some noteworthy relationships. In terms of school grounds, we surveyed by means of a structured questionnaire

(using ranking and Likert scale responses to statements) the attitudes of 16–17 year old school learners to their school grounds in all of the 12 high schools in Grahamstown (24–45 respondents per school). After the survey we counted all the species of trees and shrubs in each school ground. Significant positive relationships were found between woody plant species richness and several perceived attitudinal responses. For example, the rating of the attractiveness of the grounds in general (ten point scale) was positively related to woody plant species richness (r2 = 0.57; p < 0.005), as was the rating of the view from the learners’ class window (r2 = 0.42; p < 0.03). Because these are educational institutions, particularly pertinent was the finding that the learners’ rating of their ability to concentrate at school was also positively related to woody plant species richness in the school grounds (r2 = 0.46; p < 0.02). This suggests that GI is not required solely for immediately tangible needs, but also benefits psychological wellbeing. We undertook similar work amongst the congregants of 30 churches and their use and perceptions of church yard gardens. Eighty-seven percent of the respondents (n = 92) felt that a church garden was necessary. Reasons for requiring a garden included that it (1) added to the sense of peace and tranquillity, (2) helped the respondents pray and “connect” with God, (3) was aesthetically pleasing, (4) made the church look more homely and (5) being in nature helped the respondent relax. The majority of the respondents (60%) felt that if there was no garden surrounding their church, it would diminish the atmosphere in terms of it being a place of spiritual reflection and prayer. Echoing the results from school yards were some positive relationships between the woody plant abundance in church yards and rating of the aesthetic appearance of the garden (r2 = 0.49; p < 0.02) and the rating of the respondents’ spiritual experience (r2 = 0.37; p < 0.05).

2.4. How supportive are urban residents of green infrastructure? The results presented in answering the three previous questions clearly demonstrate that the various benefits gained from GI by individuals contribute to their wellbeing, either materially or psychologically. We could hypothesise therefore, that urban dwellers may take actions within their means to secure or promote such benefits. Such actions could include lobbying or voting for elected officials who advocate for GI, joining environmental non-government organisations (NGOs) or community-based organisations (CBOs) to protect or maintain GI, planting of trees,

Please cite this article in press as: Shackleton, C. M., et al. How important is green infrastructure in small and medium-sized towns? Lessons from South Africa. Landscape Urban Plan. (2016), http://dx.doi.org/10.1016/j.landurbplan.2016.12.007

G Model

ARTICLE IN PRESS

LAND-3050; No. of Pages 9

C.M. Shackleton et al. / Landscape and Urban Planning xxx (2016) xxx–xxx

6

Table 1 The proportion of local residents in seven small and medium-sized South African towns who feel that there is insufficient public urban green space in their town (based on random household surveys).

Fort Beaufort Grahamstown Kingwilliamstown Port Alfred Mean (±SE)

% making use of public green spaces in town

% dissatisfied with the abundance of public green space in town

RDP

Township

Affluent

Mean

RDP

Township

Affluent

Mean

33 45 63 63 51 ± 7

50 42 54 67 53 ± 5

53 70 25 93 60 ± 14

45 52 43 74 54 ± 7

60 76 73 50 65 ± 6

60 62 70 57 62 ± 3

57 46 88 0 48 ± 18

59 61 78 36 59 ± 17

or contributing time or resources to secure GI. Within some of the study towns we collected data pertaining to tree planting (three towns), prevalence of lobbying the local municipality (3 towns) and willingness-to-pay estimates (7 towns). We have previously reported on the extent of tree planting (Shackleton, Chinyimba, Hebinck, Shackleton, & Kaoma, 2015) and so will only repeat the aggregate numbers here in support of the argument being developed. That work showed that 90% of the households surveyed had trees at their homestead (being slightly lower in the RDP areas), with means across sites of 6–10 trees per household comprised of 3–4 species. In terms of raising the issue of insufficient public green space with authorities or elected councillors we undertook a random household survey (30 households per suburb) in both Fort Beaufort and Port Alfred. In both towns there was higher engagement of local officials or elected councillors by more affluent households (37% and 47% in Fort Beaufort and Port Alfred, respectively), even though the affluent areas had more green space. In the township and RDP areas of Fort Beaufort, 20% and 30% of households, respectively, claimed to have discussed the poor provision of public green space with officials or councillors. The corresponding figures for Port Alfred residents were 30% and 17%, respectively. Merging across suburbs, 29% of Fort Beaufort residents claimed to have engaged with officials or councillors, which was similar to the 31% in Port Alfred. Allied to this are several willingness-to-pay (WTP) exercises in several towns. Given the high poverty levels in most South African towns and within specific suburbs within towns we therefore also enquired (in three towns) after respondent’s willingness-to-work (i.e. donate time as part of a community group effort) if they were unwilling or unable to pay anything. There were between 60 and 150 randomly selected respondents per town. In all instances the hypothetical WTP scenario was fully explained and respondents advised that there was no stigma or pressure on having to be willing to pay anything and that a zero response was perfectly acceptable. Thereafter bid estimates were open (for Fort Beaufort, Port Alfred and Kingwilliamstown) or in increasing increments between zero and R500 per month for the other sites. In each scenario, the hypothetical payment would be made to a local NGO (to be established) rather than the local municipal authority because of the high levels of dissatisfaction with local authorities in much of South Africa. Overall, approximately half of the respondents were willing to pay some amount towards establishment or maintenance of urban GI (green spaces or street trees) in their neighbourhood (Table 2), ranging from a low of 18% in Fort Beaufort to 86% in Tzaneen. For the three towns where the data were collected in specific suburbs, there was a slight trend of an increasing proportion of respondents willing to pay with increasing wealth of the suburb, from approximately one-fifth of respondents in the poorest area to one-third in the more affluent. An opposite trend was apparent when considering respondents’ willingness to give time rather than cash, ranging from 81% willing in the RDP areas to 29% willing in the more affluent areas.

Proportion (%) of residents stating there is enough green space

Town

70 60

y = 5.76x- 26.87 r² = 0.471

50 40 30 20 10 0 8

10 12 14 16 Proporon (%) of town that is public green space

Fig. 5. The degree of satisfaction with the amount of public green space relative to proportional area of green space in four towns in the Eastern Cape.

In terms of provision of public green space, municipalities need to consider how much space is sufficient to meet the perceived needs of local residents. Whilst various guidelines are available for European countries and some of the larger metropolitan centres in South Africa, there are none for small or medium-sized towns in South Africa. As a first approach we used the data in Table 1 regarding the proportion of residents who were satisfied (inverse of the numbers in Table 1) with the abundance of public green space in their town and investigated if there was any relationship with the actual proportion of public green space as measured by McConnachie et al. (2008). Whilst recognising the inadequate sample size of this first attempt, Fig. 5 indicates that levels of satisfaction may be positively related to the abundance of public green space and suggests that a proportional area of approximately 18–20% might be expected to satisfy 80% or more of residents at the town scale, which is approximately double what most towns currently have. 3. Discussion The results provide some clear and new insights into the abundance, distribution, use and importance of green space as the basis of GI in small and medium-sized towns in South Africa. Such settings in a developing country have rarely been included in debates around the role of GI in urban planning and sustainability. This is particularly germane as our results indicate that approximately 60% of the town surface area is under green cover thereby providing some level of ecosystem services, which is almost double the average of 32.6% for 100 cities worldwide reported by Dobbs, Nitschke, and Kendal (2014). Thus, whilst developing country towns, as represented by our South African sample, are economically poor, they appear to be GI rich. All ten South African towns surveyed by McConnachie et al. (2008) had greater than 9 m2 per capita of public green space as recommend by the World health Organization. In comparison, Dobbs et al. (2014) found that only one quarter of

Please cite this article in press as: Shackleton, C. M., et al. How important is green infrastructure in small and medium-sized towns? Lessons from South Africa. Landscape Urban Plan. (2016), http://dx.doi.org/10.1016/j.landurbplan.2016.12.007

G Model

ARTICLE IN PRESS

LAND-3050; No. of Pages 9

C.M. Shackleton et al. / Landscape and Urban Planning xxx (2016) xxx–xxx

7

Table 2 The proportion of respondents in seven South African ‘towns’ who were willing to pay some amount on a monthly basis or work for the establishment or maintenance of green spaces or street trees in their suburb town (based on random household surveys other than towns marked (*) where the surveys were administered to manager/owners of every second commercial outlet along the main street of each town). Town

Bela Bela Fort Beaufort Kingwilliamstown Port Alfred Tzaneen Aggregate of 4 “poor” towns (*) Aggregate of 4 “rich” towns (*) Mean (±SE)

% respondent’s willing to work for urban greening in their suburb or street

% of respondent’s willing to pay for urban greening in their suburb or street RDP

Township

Affluent

Mean

RDP

Township

Affluent

Mean

– 3 47 13 – – – 23 ± 13

– 20 53 13 – – – 29 ± 12

– 30 – 37 – – – 34 ± 4

53 18 50 21 86 45 68 49 ± 9

– 100 43 100 – – – 81 ± 19

– 71 33 83 – – – 62 ± 15

– 58 – 0 – – – 29 ± 29

– 76 38 61

towns and cities in their worldwide sample met this recommendation. Echoing findings from developed countries and large cities, the results show that urban GI contributes to urban livelihoods and wellbeing in multiple ways. Much previous literature, largely from developed regions, has focussed on the provision and use of green spaces for recreational purposes (e.g. Shanahan, Lin, Gaston, Bush, & Fuller, 2015), but with a growing body of work linking this to human physical and mental health (e.g. Keniger, Gaston, Irvine, & Fuller, 2013). Our results add to this body of knowledge, showing both direct use for recreational (including sports) as well as psychological or mental health benefits, are displayed through the positive relationships between woody plant species richness and spiritual appreciation of churchgoers and of high school learners. Less evident in the broader knowledge and international debates is the extent of use of GI for provisioning services. Our results show that this was extensive, as was also reported by Gopal and Nagendra (2014) in the slum areas of Bangalore, India. Such high provisioning use is possibly a result of either many people being relatively recent migrants to the urban setting and/or the high levels of urban poverty. The latter is supported to some extent by the greater use among households in the poorer suburbs, and echoes many rural studies reporting higher reliance on collected provisioning services amongst poorer households than the more well-off (Angelsen et al., 2014; Shackleton & Shackleton, 2006). However, those in the more affluent areas still make extensive use of provisioning services but many do so through purchase from vendors or by planting suitable species in their homesteads. Nonetheless, a proportion of more affluent households continue to use GI for provisioning services, even in developed countries such as the USA (McLain, Poe, Hurley, Lecompte-Mastenbrook, & Emery, 2012; Poe, McLain, Emery, & Hurley, 2013) and Switzerland (Kilchling, Hansmann, & Seeland, 2009). The former explanation is less certain with the results showing no clear relationship between prevalence of use and length of residence in the town. This parallels the findings of Schlesinger et al. (2015) who showed no reduction in use of provisioning resources with increasing residence within six urban areas in sub-Saharan countries. However, the recent broaderscale survey of 32 rural communities in the Ecuadorian Amazon (Gray, Bozigar, & Bilsborrow, 2015) shows a decline in provisioning use of natural resources over 11 years, which the authors argued was a result of multiple influences, including urbanisation. Although widespread, the provision and use of GI was spatially and socially mediated. By socially mediated we mean perceptions and use differed between different social groups, of which we emphasised economic differences, which in the South African context also mirror racial classes to a large extent. By spatially mediated we mean that the distribution of GI was not even, nor was its use. The unequal distribution of GI benefits is common in many cities of the world (Dobbs et al., 2014) and has been roundly criticised

– – 58 ± 11

within a social justice or political ecology paradigm (McConnachie and Shackleton, 2010; Pedlowski, Da Silva, Adell, & Heynen, 2003; Wen, Zhang, Harris, Holt, & Croft, 2013). This requires that planners consider not only the amount of GI (either per capita or as a percentage of the urban area), but also its distribution throughout the city (Hashem, 2015) which will improve accessibility for residents in suburbs currently some distance from GI. However, as recently argued by Lin, Fuller, Bush, Gaston, and Shanahan (2014), proximity is not the only factor that drives use or none use of urban parks, but the amenities available and predisposition to nature/outdoors are also important. Clearly this relates to recreational use, whereas for provisioning use proximity is likely to be particularly important. Provisioning use is likely to be significant in supporting livelihood resilience (such as coping with shocks), especially amongst more vulnerable populations. Thus, it is quite paradoxical that the newest and centrally planned parts of the towns, i.e. the RDP suburbs designated for poorer households, lack planned urban green space. The higher private GI in affluent areas is a reflection of the larger plot size per homestead than is found in the poorer areas of town. The larger plot size encourages maintenance of some GI and is also perceived to add value of the property, which acts as an incentive. In terms of amount of GI our results indicate that the majority of residents felt that there was insufficient public green space. The relationship between satisfaction levels and proportion of public green space indicates that approximately double the current area is required. Interestingly, this matches the figure reported by Shackleton and Blair (2013) who asked those residents who expressed dissatisfaction, how much more public green space they required, with the modal answer being double the current amount. The desire for the benefits of GI is also indicated in the high proportion of households that plant trees in their home gardens, as has been reported in many other countries (e.g. Conway & Bang, 2014), as well as the reasonable proportion that were willing to either pay or work to maintain public urban green space. Whilst the amount of public GI might be insufficient, it is important to consider not only public green spaces in the provision of ecosystem services and benefits. Our results show that the bulk of GI in the small and medium-sized towns was under private tenure, paralleling the results of studies in developed countries (GonzálezGarcía & Gomez-Sal, 2008; Iverson & Cook, 2000). However, other than carbon sequestration, the contribution of residential gardens to urban GI and the services they provide has not received equal attention as public GI (Cameron et al., 2012). Consequently, many reports of the benefits of urban GI will be marked underestimates. Urban residents maintained gardens and trees for their aesthetic beauty as well as to provide shade, shelter and provisioning goods such as firewood and fruits (Kaoma & Shackleton, 2014b; Lubbe, Siebert, & Cilliers, 2010). Clearly the small size of many private gardens means that the provision of fruits and firewood does not meet the regular needs of most households. However, these may

Please cite this article in press as: Shackleton, C. M., et al. How important is green infrastructure in small and medium-sized towns? Lessons from South Africa. Landscape Urban Plan. (2016), http://dx.doi.org/10.1016/j.landurbplan.2016.12.007

G Model LAND-3050; No. of Pages 9

ARTICLE IN PRESS C.M. Shackleton et al. / Landscape and Urban Planning xxx (2016) xxx–xxx

8

be important safety-nets during times when cash or labour to procure the goods elsewhere is limited. Indeed, the notion of urban GI (private or public) as safety-nets has hardly been explored, and we hypothesise that it is likely to be a lot higher in developing countries than developed ones or poorer urban areas than richer ones, and so deserves attention. Our results show that provisioning resources were extremely important as a coping strategy for poor households in informal settlements to recover from floods. 4. Conclusion The concept of green infrastructure (GI) is gaining acceptance as a key element in the evolution of sustainable cities. This is because of the variety of ecosystem services that it provides for human benefit as well as the potentially greater cost-effectiveness of GI above built infrastructure engineering solutions. Greater advocacy of GI needs to be supported by robust data, information and insights across a range of benefits, scales and settings. In this paper we have considered different benefits in a little explored setting, namely small and medium-sized towns in a developing country. Concentrating at the household and individual scale, the results show that multiple benefits are obtained from GI, although the distribution of formal public green spaces (a major component of publically available GI) is inequitable across different areas of the towns and thus populations. In general, the poorer areas have lower proportions of public green spaces and trees. Similarly, they have lower areas of private green spaces because of the higher density of housing structures. The lower availability of GI in poorer areas is likely to be particularly telling because of the greater reliance on GI for provisioning services than in the more affluent areas or towns. Provisioning benefits such as firewood, fruits and medicines are not only important contributors to daily consumptive needs, but they also provided more than half of the ‘value’ needed to recover from shocks such as floods. Consequently, any constraints on the amount of or access to GI could severely impact livelihood resilience and vulnerability. It was not unsurprising, therefore, that respondents took, or were willing to take, a number of measures to secure or improve the availability or maintenance of GI such as discussions with local officials or elected councillors, planting of trees, or willingness to pay, or work, for GI. Our results also showed that GI was important and appreciated for cultural services such as recreation and aesthetic benefits. Additionally, we demonstrated clear connections between some attributes of GI, such as tree species abundance or richness and selfreported psychological or mental wellbeing of learners and church congregants. Acknowledgements We are grateful for the funding of different facets of this work by either SANPAD (project 10/58), or the South African Research Chairs Initiative of the Department of Science and Technology and the National Research Foundation of South Africa (any opinion, finding, conclusion or recommendation expressed in this material is that of the authors and the NRF does not accept any liability in this regard). We are grateful to Bronwyn Mclean for making Fig. 1. References Ahren, J. (2007). Green infrastructure for cities: The spatial dimension. In V. Novotny, & P. Brown (Eds.), Cities of the future: Towards integrated sustainable water and landscape management (pp. 267–283). London: IWA. Albert, C., & von Haaren, C. (2014). Implications of applying the Green Infrastructure concept in landscape planning for ecosystem services in peri-urban areas: An expert survey and case study. Planning Practice and Research, http://dx.doi.org/10.1080/02697459.2014.973683

Andersson, E., Barthel, S., Borgström, S., Colding, J., Elmqvist, T., Folke, C., et al. (2014). Reconnecting Cities to the Biosphere: Stewardship of Green Infrastructure and Urban Ecosystem Services. Ambio, 43, 445–453. Angel, S., Parent, J., Civco, D. L., Blei, A., & Potere, D. (2011). The dimensions of global urban expansion: Estimates and projections for all countries, 2000–2050. Progress in Planning, 75, 53–107. Angelsen, A., Jagger, P., Babigumira, R., Belcher, B., Hogarth, N. J., Bauch, S., et al. (2014). Environmental income and rural livelihoods: A global comparative analysis. World Development, 64, S12–S28. http://dx.doi.org/10.1016/j. worlddev.2014.03.006 Cameron, R. W., Blanuˇsa, T., Taylor, J. E., Salisbury, A., Halstead, A. J., Henricot, B., et al. (2012). The domestic garden—Its contribution to urban green infrastructure. Urban Forestry and Urban Greening, 11, 129–137. Christopher, A. J. (2001). Urban segregation in post-apartheid South Africa. Urban Studies, 38, 449–466. Cohen, B. (2006). Urbanization in developing countries: Current trends, future projections, and key challenges for sustainability. Technology in Society, 28(1e2), 63e80. Conway, T. M., & Bang, E. (2014). Willing partners? Residential support for municipal urban forestry policies. Urban Forestry and Urban Greening, 13, 234–243. Davenport, N., Gambiza, J., & Shackleton, C. M. (2011). Use and users of municipal commonage around three small towns in the Eastern Cape: South Africa. Journal of Environmental Management, 92, 1149–1460. Dobbs, C., Nitschke, C. R., & Kendal, D. (2014). Global drivers and tradeoffs of three urban vegetation ecosystem services. PLoS One, 9(11), e113000. Donaldson, R., du Plessis, D., Spocter, M., & Massey, R. (2013). The South African area-based urban renewal programme: Experiences from Cape Town. Journal of Housing and the Built Environment, 28, 629–638. Elmqvist, T., Alfsen, C., & Colding, J. (2008). Urban systems. In S. E. Jorgensen (Ed.), Encyclopaedia of Ecology (pp. 3665–3672). Oxford, UK: Elsevier. González-García, A., & Gomez-Sal, A. (2008). Private urban spaces or patios as a key element of urban ecology in tropical America. Human Ecology, 36, 291–300. Gopal, D., & Nagendra, H. (2014). Vegetation in Bangalore’s slums: Boosting livelihoods, well-being and social capital. Sustainability, 6, 2459–2473. Gray, C. L., Bozigar, M., & Bilsborrow, R. E. (2015). Declining use of wild resources by indigenous peoples of the Ecuadorian Amazon. Biological Conservation, 182, 270–277. Gwedla, N., & Shackleton, C. M. (2015). The development visions and attitudes towards urban forestry of officials responsible for greening in South African towns. Land Use Policy, 42, 17–26. Haase, D., Larondelle, N., McPhearson, T., Schwarz, N., Hamstead, Z., Kremer, P., et al. (2014). A quantitative review of urban ecosystem services assessment: Concepts models and implementation. Ambio, 43, 413–433. Hansen, R., & Pauleit, S. (2014). From multifunctionality to multiple ecosystem services? A conceptual framework for multifunctionality in green infrastructure planning for urban areas. Ambio, 43, 516–529. Hashem, N. (2015). Assessing spatial equality of urban green spaces provision: A case study of Greater Doha in Qatar. Local Environment, 20, 386–399. Hunter, M., & Posel, D. (2012). Here to work: The socioeconomic characteristics of informal dwellers in post-apartheid South Africa. Environment & Urbanisation, 24, 285–304. Iverson, L. R., & Cook, E. A. (2000). Urban forest cover of the Chicago region and its relation to household density and income. Urban Ecosystems, 4, 105–124. Kambites, C., & Owen, S. (2006). Renewed prospects for green infrastructure planning in the UK. Planning Practice and Research, 21, 483–496. Kaoma, H., & Shackleton, C. M. (2014a). Homestead greening is widespread amongst the urban poor in three medium-sized South African towns. Urban Ecosystems, 17, 1191–1207. Kaoma, H., & Shackleton, C. M. (2014b). Collection and use of urban tree products by households in poorer residential areas of three South African towns. Urban Forestry and Urban Greening, 13, 244–252. Keniger, L. E., Gaston, K. J., Irvine, K. N., & Fuller, R. A. (2013). What are the benefits of interacting with nature? International Journal of Environmental Research and Public Health, 10, 913–935. Kilchling, P., Hansmann, R., & Seeland, K. (2009). Demand for non-timber forest products: Surveys of urban consumers and sellers in Switzerland. Forest Policy and Economics, 11, 294–300. Lin, B. B., Fuller, R. A., Bush, R., Gaston, K. J., & Shanahan, D. F. (2014). Opportunity or orientation? Who uses urban parks and why. PLoS One, 9(1), e87422. Lubbe, C. S., Siebert, S. J., & Cilliers, S. S. (2010). Political legacy of South Africa affects the plant diversity patterns of urban domestic gardens along a socio-economic gradient. Scientific Research and Essays, 5, 2900–2910. McConnachie, M. M., & Shackleton, C. M. (2010). Public green space inequality in small towns in South Africa. Habitat International, 34, 244–248. McConnachie, M. M., Shackleton, C. M., & McGregor, G. (2008). Extent of public green space and alien species in ten small towns in the thicket biome, South Africa. Urban Forestry and Urban Greening, 7, 1–13. McLain, R., Poe, M., Hurley, P. T., Lecompte-Mastenbrook, J., & Emery, M. R. (2012). Producing edible landscapes in Seattle’s urban forest. Urban Forestry and Urban Greening, 11, 187–194. Miraftab, F. (2007). Governing post-apartheid spatiality: Implementing city improvement districts in Cape Town. Antipode, 39, 602–626. Pauleit, S., Liu, L., Ahern, J., & Kazmierczak, A. (2011). Multifunctional green infrastructure planning to promote ecological services in the city. In J. Niemelä

Please cite this article in press as: Shackleton, C. M., et al. How important is green infrastructure in small and medium-sized towns? Lessons from South Africa. Landscape Urban Plan. (2016), http://dx.doi.org/10.1016/j.landurbplan.2016.12.007

G Model LAND-3050; No. of Pages 9

ARTICLE IN PRESS C.M. Shackleton et al. / Landscape and Urban Planning xxx (2016) xxx–xxx

(Ed.), Handbook of urban ecology (pp. 272–285). Oxford: Oxford University Press. Pedlowski, M. A., Da Silva, V. A., Adell, J. J., & Heynen, N. C. (2003). Urban forest and environmental inequality in Campos dos Goytacazes, Rio de Janeiro, Brazil. Urban Ecosystems, 6, 9–20. Pickett, S. T. A., Boone, C. G., McGrath, B. P., Cadenasso, M. L., Childers, D. L., Ogden, L. A., et al. (2013). Ecological science and transformation to the sustainable city. Cities, 32, 510–520. Poe, M. R., McLain, R. J., Emery, M., & Hurley, P. T. (2013). Urban forest justice and the rights to wild foods, medicines, and materials in the city. Human Ecology, 41, 409–422. Sachikonye, M. (2015). Natural resource use as a coping and adaptation strategy to floods of vulnerable populations in the Eastern Cape. pp. 98. Grahamstown: Rhodes University. Sandström, U. F. (2002). Green infrastructure planning in urban Sweden. Planning Practice and Research, 17, 373–385. Schäffler, A., & Swilling, M. (2013). Valuing green infrastructure in an urban environment under pressure—The Johannesburg case. Ecological Economics, 86, 246–257. Schlesinger, J., Drescher, A., & Shackleton, C. M. (2015). Socio-spatial dynamics in the use of wild natural resources: Evidence from six rapidly growing medium-sized cities in Africa. Applied Geography, 56, 107–115. Shackleton, C. M., & Blair, A. (2013). The abundance of urban public green space influences residents’ perceptions, uses and willingness to get involved in two small towns in South Africa. Landscape & Urban Planning, 113, 104–112. Shackleton, C. M., & Shackleton, S. E. (2006). Household wealth status and natural resource use in the Kat River Valley, South Africa. Ecological Economics, 57, 306–317. Shackleton, S. E., Chinyimba, A., Hebinck, P., Shackleton, C. M., & Kaoma, H. (2015). Multiple benefits and values of trees in urban landscapes in two towns in northern South Africa. Landscape and Urban Planning, 136, 76–86. Shanahan, D. F., Lin, B. B., Gaston, K. J., Bush, R., & Fuller, R. A. (2015). What is the role of trees and remnant vegetation in attracting people to urban parks? Landscape Ecology, 30, 153–165. Tzoulas, K., Korpela, K., Venn, S., Yli-Pelkonen, V., Kazmierczak, A., Niemela, J., et al. (2007). Promoting ecosystem and human health in urban areas using Green Infrastructure: A literature review. Landscape and Urban Planning, 81, 167–178. United Nations. (2014). Department of Economic and Social Affairs, Population Division. In World urbanization prospects: The 2014 revision, highlights (ST/ESA/SER.A/352). pp. 32. New York: UN.

9

Ward, C. D., & Shackleton, C. M. (2016). Natural resource use: Incomes and poverty along the rural-urban continuum of two medium-sized South African towns. World Development, 78, 80–93. Wen, M., Zhang, X., Harris, C. D., Holt, J. B., & Croft, J. B. (2013). Spatial disparities in the distribution of parks and green spaces in the USA. Annals of Behavioral Medicine, 45(Suppl. 1), S18–S27. Wisner, B., Pelling, M., Mascarenhas, A., Holloway, A., Ndong, B., Faye, P., et al. (2015). Small cities and towns in Africa: insights into adaptation challenges and potentials. In S. Pauleit, A. Coly, S. Fohlmeister, P. Gasparini, G. Jørgensen, S. Kabisch, W. J. Kombe, S. Lindley, I. Simonis, & K. Yeshitela (Eds.), Urban vulnerability and climate change in Africa: A multidisciplinary approach (pp. 153–196). Dordrecht: Springer. World Bank. (2015). Urban population. http://data.worldbank.org/indicator/SP. URB.TOTL.IN.ZS. Accessed April 2016 Wright, H. (2011). Understanding green infrastructure: the development of a contested concept in England. Local Environment, 16, 1003–1019. Charlie Shackleton is professor and research chair in the Department of Environmental Science, Rhodes University, South Africa. His research interests include natural resource use by rural and urban communities and urban greening and forestry for human wellbeing and sustainable towns. Andrew Blair is former postgraduates in Department of Environmental Science, Rhodes University, South Africa. Peter De Lacy is former postgraduates in Department of Environmental Science, Rhodes University, South Africa. Humphrey Kaoma is former postgraduates in Department of Environmental Science, Rhodes University, South Africa. Noster Mugwagwa is former postgraduates in Department of Environmental Science, Rhodes University, South Africa. Mwazvita Dalu is current postgraduates in Department of Environmental Science, Rhodes University, South Africa. Wesley Walton is former postgraduates in Department of Environmental Science, Rhodes University, South Africa.

Please cite this article in press as: Shackleton, C. M., et al. How important is green infrastructure in small and medium-sized towns? Lessons from South Africa. Landscape Urban Plan. (2016), http://dx.doi.org/10.1016/j.landurbplan.2016.12.007