Spatial analyses of threats to ecosystem service hotspots in Greater Durban, South Africa

Study area

Durban is administered by a local government authority known as eThekwini Municipality and is situated in the province of KwaZulu-Natal, South Africa. In 2015 the municipal area of Durban was approximately 229,193 ha in extent (1.4% of the province) with a population 3.55 million (Davids et al., 2016). Durban’s coastline is 98 km long, and is dissected by the rivers of 18 major water catchments and 16 estuaries. Furthermore, Durban is located within the Maputaland-Pondoland-Albany global biodiversity hotspot, rated as such because of its high levels of plant endemism and habitat loss (Mittermeier et al., 2004).

Durban contains urban, peri-urban and rural environments, with approximately two-thirds of the municipal area being rural or semi-rural, where a large proportion of local inhabitants are indigent and directly reliant on ecosystem services for basic needs (Roberts & O’Donoghue, 2013; Sutherland et al., 2014). Social challenges in Durban include high levels of poverty; many densely populated informal settlements; unequal basic service delivery; high rates of urbanisation and dual governance arrangements, whereby eThekwini Municipality jointly administers communal land in the rural northwest and southwest areas of the municipality with the Ingonyama Trust Board (under the jurisdiction of traditional councils) and provincial government (McLean et al., 2016; Sutherland et al., 2014; Davids et al., 2016). These challenges have often meant that socio-economic development priorities have taken preference over environmental and biodiversity concerns (Roberts, 2008).

The Durban Metropolitan Open Space System (D’MOSS) is formally included in eThekwini Municipality’s hierarchy of spatial plans and policies, and serves as a legal instrument for environmental protection through development and use restrictions. D’MOSS comprises natural areas of high biodiversity value within a series of interconnected open spaces, aimed at protecting globally significant biodiversity and the supply of ecosystem services (Roberts & O’Donoghue, 2013; Davids et al., 2016).

Overview of South African spatial planning products considered in this study

In terms of the Constitution of the Republic of South Africa, 1996 (as amended), spatial planning in South Africa is the responsibility of all three spheres of government, local, provincial and national. The national Spatial Planning and Land-use Management Act (SPLUMA), 2013 (Act No. 16 of 2013) provides the framework for all land use management and spatial planning legislation in South Africa. This Act has numerous aims, including the regulation of planning procedures and decision making; addressing spatial imbalances that resulted from the apartheid era and ensuring integration of sustainable development principles in land use planning and regulatory tools.

In terms of SPLUMA, all spheres of government must prepare spatial plans, however, land use management is the responsibility of municipalities, in participation with local traditional councils. In Durban, the traditional councils administer communal lands totalling ±82,266 ha (35.8%) (Davids et al., 2016). SPLUMA and the Municipal Systems Act, 2000 (Act No. 32 of 2000) require that local municipalities develop land use management systems, including the implementation of Integrated Development Plans (IDPs) and municipal spatial development frameworks (SDFs) and associated land use guidelines. EThekwini Municipality’s IDP (2012/2013) considered in this study, was developed with the aim to address key strategic issues identified in the National Spatial Vision and Provincial Growth and Development Strategy, including job creation; reversing the effects of apartheid; access to quality of education, healthcare and social protection; and the transition to a low carbon economy (eThekwini Municipality, 2012b). The IDP states that in order to suitably manage development and minimise impacts on the natural environment and associated ecosystem services, spatial planning must be enhanced and better aligned with the strategic development plans of the Municipality (eThekwini Municipality, 2012b).

Ecosystem service hotspots

The classification, mapping and spatial prioritisation of ecosystem services has received considerable attention (Egoh et al., 2008; Schröter & Remme, 2016; Cimon-Morin, Darveau & Poulin, 2013; Schröter et al., 2017). Among other approaches, the delineation of ‘ecosystem service hotspots’ has been used to spatially prioritise ecosystem services in conservation planning (Cimon-Morin, Darveau & Poulin, 2013). Ecosystem service hotspots either refer to areas of high ecosystem service values of one service, or areas with a combination of multiple services (Schröter & Remme, 2016).

Davids et al. (2016) identified hotspots for 13 ecosystem services in Durban, grouped into five categories (Table 1). The 13 ecosystem services are: carbon storage, water yield (to dams), four sediment retention services (preventing sedimentation of dams, stormwater and sewer pipes and the harbour), three flood attenuation services (relevant to the population, private and public infrastructure) and four nutrient retention services (phosphorus and nitrogen relative to both dams and estuaries) (Table 1; Fig. 1). The original ecosystem service maps for these 13 services were commissioned by the municipality’s Environmental Planning and Climate Protection Department in 2012 and were derived using the InVEST tool, developed by the Natural Capital Project (Tallis & Polasky, 2009). InVEST allows for priority ecosystem service areas to be identified based on a number of factors that either contribute, to or impact on selected ecosystem services (J Glenday, 2012, unpublished data submitted to the eThekwini Municipality). ArcGIS 9.3 was used to estimate the ecosystem service provision based on biophysical properties, land over and location relative to downstream populations, property and infrastructure, along land surface and river channels flow paths (J Glenday, 2012, unpublished data submitted to the eThekwini Municipality). These ecosystem services were linked to a standard 86.9 m resolution and scaled to values ranging from 0 to 100, to indicate relative service provision compared to the maximum found in the study area.

Ecosystem service hotspots were defined by Davids et al. (2016) as the top 50% of ecosystem service provisioning areas (Fig. 1). These ecosystem service hotspots represent the provisioning areas of a particular service that had values greater than the median value (Davids et al., 2016). The median value was used to reduce the importance of outliers present (Press et al., 1992) in the range of ecosystem service values and to avoid the exclusion of large areas providing good ecosystem services (Davids et al., 2016). Approximately 35% of Durban was considered a hotspot for at least one of the 13 ecosystem services assessed (Fig. 2). Majority of the 13 ecosystem service hotspots identified by Davids et al. (2016) were found to be outside of environmental management areas and threatened by habitat transformation, rapid densification, invasive alien plant invasions and pollution, with only half of ecosystem service hotspots located within D’MOSS. The hotspot richness map (Fig. 2) was developed using overlay analysis, which merged the distributions of 13 ecosystem services provisioning maps into a raster grid (with a presence value of 1 for each map) to show the total number of ecosystem services produced in a particular location from a minimum of 0, to a maximum of 13 (Schröter & Remme, 2016; Davids et al., 2016).

Spatial development framework

The SDF forms part of a hierarchy of integrated plans that make up the Land Use Management System currently being established to manage land use and development within Durban (eThekwini Municipality, 2012b). The intention of the SDF is to guide all decisions related to the use, development and planning of land within Durban, providing a strategic framework that spatially indicates how the implementation of the city’s IDP should occur. The SDF is a five-year plan, which is revised annually in line with the IDP. It provides strategic multi-sectoral planning guidance relative to development priorities, transport planning, bulk infrastructure and environmental directives, and acts as a guide to more detailed Local Area Plans, Functional Area Plans, detailed Precinct Plans and Land Use Schemes. This study assesses the proposed changes to land use that could occur through the implementation of the eThekwini Municipality Spatial Development Plans against “the five categories of ecosystem service hotspots and the subset of 13 services.”

The terminologies for land uses in the SDPs were not standardized and categories of land use were grouped as follows: natural areas for environmental protection, nature-based recreation and tourism, amenity, agricultural use and infrastructure.

Spatial planning regions

Durban’s municipal area is divided into four major planning regions, namely, North, South, Central (including Inner West) and Outer West (Fig. 3). Within Durban, only 36% of land is covered by town planning schemes and formally administered by the municipality, approximately 38% is communal land jointly administered by the traditional councils, and the municipality and 26% is peri-urban and jointly administered by local and provincial government (McLean et al., 2016). A process is currently underway to incorporate traditional communal lands and agricultural areas into town planning schemes (McLean et al., 2016).

Under the SDF, four individual Spatial Development Plans (SDP) have been prepared and were adopted in November 2009 and revised in November 2010 and 2011 (eThekwini Municipality, 2012b). Two SDPs (North and Outer West) were selected for use in this study, due to their contrasting general land uses, that is one being more rural in the case of the Outer West SDP and the other more urban in the case of the North SDP.

Historical development applications

Historical development applications assessed in this study include all environmental impact assessments and sand mining applications that were submitted to the Environmental Planning and Climate Protection Department of eThekwini Municipality, between 2009 and 2012. The applications were linked to a point file in GIS, thus the points used in the analyses are merely an indication of where transformation of land may be expected, in the event that applications were successful.

Spatial development plans: outer west and north planning regions in relation to priority areas for the five categories of ecosystem services

The proposed implementation of the SDPs will affect ecosystem service provision in both regions. On average, 60% and 47% of ecosystem service hotspot areas in the North and Outer West respectively, are at risk of being degraded or lost, as they will fall outside of land uses specifically designated to remain natural, such as environment, D’MOSS, public open space, green corridor and amenity (Table 2; Fig. 4; Appendix S3, Fig. 5, Appendix  S4).

On average, approximately 39% for the North and 63% for the Outer West region is proposed to remain within areas designated for environmental purposes, compatible with ecosystem service conservation, namely, D’MOSS, Public Open Space or Amenity. The situation with water yield and nutrient retention to dams is dire with the proposed implementation of the SDP in these two planning regions, leading to about 26% and 21% respectively of the hotspot areas remaining within environmental areas in the North and 36% and 47% respectively in the Outer West. Sediment retention services are also poorly provided for in the North, with only 33% falling within an environmental land-use.

In both regions, the proposed residential land use in the SDPs overlaps with the highest proportions of ecosystem services hotspot areas, with approximately 33% allocated for combined rural and urban residential uses in the Outer West (Appendix S3) and 38% combined rural and urban residential uses in the North (Appendix S4). The other proposed land-uses that are not compatible with ecosystem services, namely industry, commercial, office park and mixed use comprise comparably nominal proportions at between 1.3% and 2.5% in the North and 0.85 and 1% in the Outer West.

The five categories of ecosystem services will be variably affected by the North SDP, with an average of 18% proposed to fall within agricultural areas, of which 15% are carbon storage areas, 24% sediment retention areas, 34% in water yield ecosystem service areas and 15% of flood attenuation.

Development proposals within ecosystem service hotspots

Approximately 36% (n = 658) of all EIA applications within Durban between 2009 and 2012 fell within an ecosystem service hotspot, while 84% (n = 144) of sand mining applications were for locations within ecosystem service hotspots (Tables 3 and 4). The ability to store carbon could be impacted as the highest number of EIA and sand mining applications are noted within carbon storage hotspots (68% and 86% of applications respectively of the total applications made within ecosystem service hotspots, i.e., the 36%).

With respect to areas providing multiple ecosystem services (Fig. 6), the vast majority of EIAs occurred in areas that only produce one ecosystem service (84%), with approximately 9% within areas with two services, 5% within areas providing three services and 2% in areas providing four services. Similarly, for sand-mining applications, 93% of them occurred in areas that only produce one ecosystem service, approximately 4% within areas with two services, 2% within areas providing three services and 1% in areas providing four services.

No EIA or sand mining applications were made in areas with a hotspot richness of five or more services.

Implications of threats to ecosystem services

The findings that the implementation of the SDPs might result in an average transformation or loss of 49% of ecosystem service hotspots, confirm habitat loss as a threat to the continued supply of ecosystem services within urban areas, and the need for effective municipal planning to reduce this threat (Davids et al., 2016; Daily, 2000; Driver, Cowling & Maze, 2003). These threats are emphasised by the finding that approximately 36% of EIA applications and 84% of sand mining applications made between 2009 and 2012, were located within ecosystem service hotspots.

Our study highlights the importance of the use of ecosystem service hotspots in conservation planning (Egoh et al., 2008; Schröter & Remme, 2016; Schröter et al., 2017; Cimon-Morin, Darveau & Poulin, 2013) as a potential tool to both avoid foreseeable threats to ecosystem services and human well-being, and to inform conservation and management to ensure the sustainability of ecosystem services (Davids et al., 2016; Schröter et al., 2017). The implications of the findings and options for management of the ecosystem service hotspots considered in this study, are discussed below.

The potential loss of carbon storage hotspots, as identified in this study, would be counterproductive to local government strategies to combat climate change. Anthropogenic changes have already resulted in the planetary boundaries for climate change and biodiversity loss being exceeded (Steffen et al., 2015; Rockström et al., 2009). The magnitude of change in supply of ecosystem services that were identified in this study, coupled with the inability of remaining natural areas to meet certain biodiversity targets in Durban (McLean et al., 2016), may push Durban into a state where critical environmental thresholds are crossed, and trigger further non-linear system changes (Steffen et al., 2015; Rockström & Karlberg, 2010). Although climate change is recognized as a global phenomenon, responses to climate change are being implemented at the local level, raising the importance of local governance in reducing climate-induced risks on communities (Williams et al., 2018; Surminski et al., 2017).

Over 50% of disaster-related fatalities are attributed to flood events (Wehn et al., 2015) and a third of economic losses are linked to climate induced risks, with the risk of flooding predicted to increase with global warming (Hirabayashi et al., 2013). The proposed increase in developed ‘incompatible’ land uses will result in increased runoff, and coupled with the potential loss of flood attenuation ecosystem services, may exacerbate flooding impacts with consequences for vulnerable indigent communities, particularly those living in informal settlements in flood prone areas (Arunyawat & Shrestha, 2016; Im et al., 2009; Du et al., 2012; Williams et al., 2018). This may also result economic costs related to loss or damage to public and private infrastructure.

The potential impacts on sediment retention services could result in soil erosion and sedimentation of rivers and dams, threatening water quality and aquatic food sources. In addition, the loss of sediment retention services could lead to blockages of stormwater infrastructure, with resultant economic costs related to the maintenance of the same. About a third of sediment retention services are planned to be replaced by agricultural land uses in both regions. In light of potential impacts associated with agricultural practices including increased runoff and soil erosion and resultant effects on water quality (Arunyawat & Shrestha, 2016; Kareiva et al., 2007; Baudron & Giller, 2014), sustainable agricultural practices that limit unwanted impacts on ecosystems and their services to human development are needed (Rockström & Karlberg, 2010).

The substantial amount of nutrient retention hotspots that would be impacted or lost by transformation to rural residential land use may result in increased nutrient loading in estuaries and dams and infiltration of nitrogen into groundwater, impacting on aquatic food sources and drinking water and posing a threat to human health (Cirone & Duncan, 2000). These impacts would also have cost implications for the municipality related to the treatment of water for supply of potable water to residents. In order to mitigate costs and enhance the quality of water supplied, interventions to enhance natural purification of water through wetlands and other habitats and mitigation of water quality degradation due to agriculture and urban development should be considered (Berka, Schreier & Hall, 2001; Houlahan & Findlay, 2004; Verhoeven et al., 2006).

The transformation of water yield service areas to agricultural areas and residential and uses will affect water resources. This impact would depend on a range of factors including the types of vegetation before and after transformation (e.g., affecting evapotranspiration rates and moisture transfer rates to soil), whether the change is permanent or temporary and the proposed land use practices related to, for example, irrigation and fertilization applications (Scanlon et al., 2007). Increases in rain-fed cultivation, increases in built-up areas and decreases in forest cover have been shown in increase water yield and decrease water quality (Arunyawat & Shrestha, 2016; Scanlon et al., 2007). However, increases in surface- or ground-water irrigated agriculture would decrease both water quantity and quality (Scanlon et al., 2007). The condition of natural areas within catchments that supply surface water through runoff must be maintained to ensure yields of high water quality, reduced nutrient loss and reduced soil erosion (Scanlon et al., 2007; Egoh et al., 2008).

Suggestions for improved planning and management of ecosystem services

The inclusion of some ecosystem services into Durban spatial plans such as D’MOSS seems to be partially effective. Some of the reasons include that such planning allocations do not necessarily preclude development, nor do they ensure management of biodiversity in these areas. The challenge remains to efficiently govern ecosystem services in a systematic way. The effective management of ecosystem services would involve the structured incorporation of ecosystem services into decision-making, not only by governments, but also by businesses and individuals and by sectors such as agriculture, forestry, mining and land-use development planning (Pierce et al., 2005; Daily et al., 2009). This would require not only the consideration of conservation and management of natural land to protect and enhance ecosystem service provision, but also the consideration of the impacts of development objectives on the same.

Davids et al. (2016) identified the need for the consideration of ecosystem service areas within the municipal land-use decision-making frameworks through a possible independent conservation strategy for ecosystem services in Durban. This strategy would include the consideration of the large proportion of ecosystem service provisioning areas, lying outside of protected and managed areas, that are under threat of transformation.

With respect to planning for development within Durban, a proactive approach may be required in order to mitigate the impacts of development proposals within important ecosystem service areas. Observations and expert judgement suggest that the fact that no development applications were made in areas with a hotspot richness of five or more, could be attributed to the potential inaccessibility of these areas in terms of slope or the fact that these are located within drainage lines or rivers. These two factors may thus be considered as natural measures of protection for ecosystem services, except for sand mining activities, that generally target rivers and drainage lines.

The potential loss of important ecosystem service areas needs to be carefully considered prior to the implementation of planning and development proposals in Durban. Practical short-term governance solutions would be to include ecosystem service hotspot areas into D’MOSS where they are currently excluded, to develop guidelines for use of ecosystem service hotspots areas and to develop decision-making guidelines for development applications that fall within ecosystem service priority areas.

Potential for future research

This research could be expanded in numerous ways. Firstly, the quantification of ecosystem services for various development scenarios would serve as a far more effective tool for planning future development. This could include quantifying the contribution of cultivated areas to ecosystem services provision, which was not assessed in this study. A spatial analysis to quantify accessibility to areas with high ecosystem service richness (referred to above), could confirm whether natural measures of protection exist for ecosystem service hotspots. The addition of cultural services or non-material NCP, would also provide a more comprehensive understanding of the importance of natural capital for the population of Durban and would provide further motivation for the protection of the same.

More work is required to assess the compatibility of various land uses to ecosystem service provision (Reyers et al., 2009), including research on the ecological processes linked to ecosystem services and how these are affected by land use change and the social and economic factors that drive changes (Fu et al., 2015). This will provide increased understanding of the links between social, economic and ecological factors that could prove more useful for decision-making.