Urban ecosystem services assessment along a rural–urban gradient: A cross-analysis of European cities

(This is a sample cover image for this issue. The actual cover is not yet available at this time.)

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright

Author's personal copy Ecological Indicators 29 (2013) 179–190

Contents lists available at SciVerse ScienceDirect

Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind

Urban ecosystem services assessment along a rural–urban gradient: A cross-analysis of European cities Neele Larondelle a , Dagmar Haase b,∗ a

Humboldt University Berlin, Institute of Geography, Lab of Landscape Ecology, Germany Humboldt University Berlin, Institute of Geography, Lab of Landscape Ecology and Helmholtz Centre for Environmental Research – UFZ, Department of Computational 10 Landscape Ecology, Germany b

a r t i c l e

i n f o

Article history: Received 16 September 2012 Received in revised form 13 December 2012 Accepted 26 December 2012 Keywords: Urban ecosystem services Spatial patterns Indicators Trade-offs Cities Europe

a b s t r a c t The main objective of this paper is to present an assessment approach for ecosystem services in an urban context covering the local and the regional scale. It was applied to different European cities. A set of indicators representing important urban ecosystem goods and services – local climate regulation, air cooling and recreation – was tested using spatial data along an urban–rural gradient. The results show that there is neither a typical rural–urban gradient in terms of urban ecosystem service provisioning nor a uniform urban spatial pattern of service provisioning that can serve as a generic model for cities. The results demonstrate that (1) core cities do not necessarily provide fewer ecosystem services compared to their regions and (2) there were no patches found within the four case study cities where all of the indicators report very high performance values. The analysis further shows that a high degree of imperviousness does not necessarily entail low ecosystem service provisioning if an urban structure contains a considerable amount of mature trees which support carbon storage and biodiversity. The results of the present paper provide insights into potentials and trade-offs between different urban ecosystem services that should be considered during urban planning when setting targets and establishing thresholds to protect environmental resources, ecosystem services and biodiversity for residents. © 2013 Elsevier Ltd. All rights reserved.

1. Introduction To date, the role of large settlements and agglomerations in global environmental complexity remains understudied (Decker et al., 2000; Jansson, 2012; Seto et al., 2012). The linkages between land-use dynamics and ecosystem services remain largely unknown, and research on urban ecosystem services is highly demanded (TEEB, 2011; Elmqvist et al., 2012; Carpenter et al., 2009) as “. . .urbanisation is arguably the most dramatic form of land transformation that profoundly influences biological diversity and human life” (Luck and Wu, 2002:327). There is a great need for “[. . .] practical applications, appropriate methods for identification and quantification of individual services, suitable models, indicators and the integration of system components [. . .]” (Burkhard et al., 2010:1), and as city administrations in particular are facing increasingly complex challenges, further development of cross-scale applicable methods is highly required for urban systems. Urban systems differ strongly from rural systems (Haase, 2012) as well as from city to city because the “urban form of a specific city is the result of a variety of influences,

∗ Corresponding author. E-mail address: [email protected] (D. Haase). 1470-160X/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ecolind.2012.12.022

including site and topography, economic and demographic development and planning efforts in the past” (Schwarz et al., 2010). Luck and Wu (2002) mention that physical, ecological, and socioeconomic processes are strongly affected by spatial patterns within their boundaries and beyond. Thus, for example, compact built structures offer less ventilation and air movement than more scattered or open built patterns. Additionally, a lower percentage of sealed land in a given region provides higher rainfall infiltration, higher evapotranspiration and greater moderation of extreme events compared to impermeable surfaces (Haase et al., 2012). Moreover, the configuration of differently built-up and sealed surfaces in a city also leads to different fragmentation and habitat suitability indices (Schwarz, 2010). Assessing the distribution of ecological functions and ecosystem services within cities can increase the success of attempts to reduce the growing urban ecological footprint (Rees and Wackernagel, 1996) and can thus contribute expertise and knowledge to the debate of urban social–ecological justice discussed recently at global and regional scales (Marcuse et al., 2009; Mitchel, 2003). “Cities are dependent on the ecosystems beyond the city limits, but also benefit from internal urban ecosystems (Bolund and Hunhammar, 1999:293)”. The TEEB-Report for Cities (TEEB, 2011) suggests that ecosystem services could be used as a tool; thus, cities have the opportunity to make positive changes, saving on

Author's personal copy 180

N. Larondelle, D. Haase / Ecological Indicators 29 (2013) 179–190

municipal costs, strengthening local (green) economies, enhancing quality of life and securing livelihoods. The main objective of this study is to develop a comparative ecosystem services assessment approach for urban areas. So doing, different proxy-based indicators for regulation and cultural ecosystem services, which are most important for cities as places of population concentration, are developed and valued (i) across different time steps and (ii) across two spatial scales, local and regional. The assessment approach is tested using data from four different European cities. According to Lautenbach et al. (2011), there are only a few studies that assess multiple ecosystem services over more than two time steps and use techniques other than simple benefit-transfer approaches. Hence, the objective of this study to investigate different cities using a multiple ecosystem services approach including synergies and trade-offs within a time series helps to close this gap. The present study is part of the BiodivERsA project URBES (Urban Biodiversity and Ecosystems Services; Elmqvist et al., 2012), which is aimed to address the major knowledge gaps regarding ecosystem services and biodiversity in the urban context and will also create a professional training and communication programme for local authorities together with organisations such as ICLEI (Local Governments for Sustainability) and IUCN (International Union for Conservation of Nature). The latter programme is very important because of “[. . .] a clear lack of information relevant to local scale decision making” (Burkhard et al., 2012:1). Although necessary for implementation at the city and urban regional scale, standardised non-monetary approaches and appropriate indicators for quantification are still missing (Burkhard et al., 2012; Kroll et al., 2011). The study focuses on the case study cities of URBES, Berlin, Salzburg, Stockholm and Helsinki, and develops a straightforward approach to assessing urban ecosystem service provisioning according to the practical needs of planners/stakeholders.

2. Study sites The four cities and urban regions investigated in this study differ in their population size and land-use patterns. These cities are systematically distributed across Europe, including central-eastern (Berlin), western (Salzburg) and northern (Helsinki, Stockholm) cities. All of the cities have at least a medium size of 150,000 inhabitants and have been influenced by different historical urban development and planning systems/schemes. Those characteristics make them exemplary for many medium-to-large size European cities with a good spread from east to west. Because European cities are often dense and to a large extent dependent on ecosystem services from outside their administrative borders (Bolund and Hunhammar, 1999:299), we studied both the core city and the larger urban region, namely the administrative NUTS3 region (Kabisch and Haase, 2011). The NUTS3 region includes the core city (within its administrative boundaries) as well as a larger suburbanised and rural hinterland. When addressing urban ecosystem services and their spatial distribution, it is crucial to gain knowledge about the allocation of these services in the larger urban region beyond the city boundaries. Berlin is Germany’s largest city and the second-most populous city in Europe. Located in north-eastern Germany, Berlin is the centre of the Berlin-Brandenburg Metropolitan region, which is home to 4.6 million residents (Statistisches Landesamt 2010). Located in the European Plains, Berlin is influenced by a temperate seasonal climate. Approximately one-third of the city’s area is composed of forests, parks, gardens, rivers and lakes. The Helsinki Metropolitan Area includes the city of Helsinki and three other cities: Espoo and Vantaa, which immediately border Helsinki to the west and north, respectively, as well as

Kauniainen, which is an enclave within the city of Espoo. The Helsinki Metropolitan Area is the world’s northernmost urban area among those with a population of over one million people, and the city is the northernmost capital of an EU member state. Altogether, 1.1 million people, approximately one in five Finns, live in the Helsinki Metropolitan Area (vrk.fi 2010). Salzburg is the fourth-largest city in Austria and the capital city of the federal state of Salzburg. Salzburg is situated on the banks of the Salzach River, at the northern boundary of the Austrian Alps. As of 2011, this city had approximately 150,000 residents (Statistic Austria 2011). Stockholm is the capital and the largest city of Sweden and constitutes the most populated urban area in Scandinavia. Stockholm has a population of 864,324 in the municipality (2010), 1.4 million in the urban area (2010), and approximately 2.1 million in the Metropolitan Area (2010) (scp.se, 2010). Stockholm is located on Sweden’s south-central east coast, where Lake Mälaren meets the Baltic Sea. The central parts of the city consist of 14 islands that are continuous with the Stockholm archipelago. Over 30% of the city area is made up of waterways, and another 30% is made up of parks and green spaces.

3. Methods 3.1. Data All of the data and references used for the study are listed in Table 1. The spatial data are publicly available, which meets the needs of city and regional planning offices and administrations (Larondelle and Haase, 2012; Lautenbach et al., 2011) that the present study finally aims to support. All of the data were processed and mapped using ArcGIS 10 software. For land-use mapping, 100m resolution Corine Land Cover (CLC) raster data provided by the EEA were chosen.

3.2. Gradient approach Because land-use change has a huge impact on the supply of ecosystem services and the ability that ecosystems have to contribute to human well-being (Lautenbach et al., 2011; Burkhard et al., 2012), land-use change was assessed at two different spatiotemporal scales. Regarding the time-scale, the years 1990, 2000, and 2006 for Berlin and Salzburg and the years 2000 and 2006 for Helsinki and Stockholm were analysed according to CLC availability. To evaluate the provisioning of urban ecosystem services, each of the indicators was chosen with respect to its sensitivity to landuse change. According to Kroll et al. (2011), the target of a study should be a good compromise among precision, broad applicability to a variety of landscapes and adaptability to varying data availability. According to Zipperer et al. (2000:686), to account for human influence on the urban landscape, the basic concept of the ecosystem must incorporate a human component. In our study, this human component is represented by land-use. To identify and assess the impact of the land-use change spatially, a rural–urban gradient was used as reported by Kroll et al. (2011). Around the city centre, which is represented by the centroid of the two most central urban districts, a 30-km concentric buffer was created with 1-km intervals. Gradient analyses are a suitable tool for urban landscape studies, and rural–urban gradients have been commonly used to consider changes of ecological patterns and processes due to urbanisation, following Andersson et al. (2009).

Author's personal copy N. Larondelle, D. Haase / Ecological Indicators 29 (2013) 179–190

181

Table 1 Data used. Data type

Data

Source

Land-use data

Corine Land Cover

http://www.eea.europa.eu/

Census data

Berlin core city Berlin NUTS3 Helsinki core city Helsinki NUTS3 Salzburg core city Salzburg NUTS3 Stockholm core city Stockholm NUTS3

http://www.stadtentwicklung.berlin.de/geoinformation/ http://www.stadtentwicklung.berlin.de/geoinformation/ http://www.aluesarjat.fi http://www.stat.fi/til/vaerak/tau en.html http://www.stadt-salzburg.at http://www.statistik.at/ http://www.statistikomstockholm.se/ http://www.statistikomstockholm.se/

Administrative boundaries

Berlin Helsinki Salzburg Stockholm

http://www.stadtentwicklung.berlin.de/geoinformation/ http://www.hel2.fi http://www.stadt-salzburg.at http://www.stockholm.se

3.3. Urban ecosystem services indicators According to Burkhard et al. (2012), appropriate ecosystem service indicators need to be quantifiable, sensitive to changes in land cover, temporally and spatially explicit and scalable. To assess the multi-functionality and variety that urban landscapes possess a set of different proxy-based indicators were chosen to quantify the urban ecosystem services of (local) climate and air-quality regulation and recreation after Lautenbach et al. (2011). “Since field scale-monitoring schemes for ecosystem services or ecosystem functioning are missing, proxy-based indicators can help to assess the historic development of ecosystem services or ecosystem functioning at the regional scale” (Lautenbach et al., 2011:676). The indicators chosen to assess the mentioned urban ecosystem services are all highly sensitive to land-use and land-use change (Table 2). That sensitivity is an important consideration when dealing with cities because cities are very dynamic environments. Additionally, the services and indicators chosen also belong to the list of key services within the TEEB study that benefit from ecosystem service integration into planning (TEEB, 2011). Each of the indicators can be assessed using proven, scientifically tested empirical methods (cf. Table 2) and also have the potential to assess adaptation to climate change. 3.3.1. Climate regulation and air-quality regulation Climate regulating services play a major role in contributing to human well-being in cities because these services can help to reduce urban heat-island effects (Chen et al., 2006; Kottmeier et al., 2007), mitigate climate change and decrease the strain of air pollution to a minimum. Within this study, the climate regulation and air-quality regulation indicators ‘surface emissivity,’ ‘f-evapotranspiration,’ ‘tree cooling potential’ and ‘carbon sequestration’ were chosen. Urban surface emissivity is strongly connected to temperature patterns but is more easily measured and is thus a good proxy to analyse urban heat stress and temperature contributions (Schwarz et al., 2011). The indicator ‘f-evapotranspiration’ (f-ETP) captures latent heat flows because the relationship is linear and is accordingly a good complement to the valuation of surface emissivity (Schwarz et al., 2011). Evapotranspiration is a proxy that is not only relevant for air temperatures and human comfort but is also highly related to urban water balance (Haase and Nuissl, 2007). The valuation of the tree-cooling potential of urban parks is a good indicator of the temperature balancing service provided by vegetation (Bowler et al., 2010). The indicator ‘carbon storage’ was chosen as an indicator for which supply is not local but rather regional to global. Cities can act as carbon sinks and thus

contribute to global carbon storage in a small but considerable amount (Strohbach et al., 2012). 3.3.2. Recreation and mental and physical health “The recreational aspects of all urban ecosystems, with possibilities to play and rest, are perhaps the highest valued ecosystem service in cities” (Bolund and Hunhammar, 1999:298). However, to date, “[. . .] everyday outdoor recreation in nearby green spaces is often not even mentioned” (Daniel et al., 2012:3). In this study, different data-based methods were used to assess both the maintenance and accessibility of urban green and blue space. The recreational areas selected from the given land-use classes were water classes (CORINE land-use class codes 511, 512, 521, 522, and 523), forest classes (311, 312, and 313) and the two urban-green classes (141 and 142). Particularly for people living in urban environments, the accessibility of green and blue urban areas plays an important role in the quality of life and human well-being because, according to Daniel et al. (2012), recreational activities and nature study offer an opportunity to experience the benefits of ecosystem services directly. 3.4. Assessment The modelling of urban ecosystem service provisioning patterns was conducted using a set of methods applied within a geographical information system (ArcGIS, version 10) and/or different statistical software (MS Excel, version 2010, SigmaPlot, version 12). Empirical values derived from tables found in the literature were used to evaluate the indicators of surface emissivity, f-evapotranspiration and carbon storage. To obtain values for surface emissivity, a lookup table was used that was created by Schwarz et al. (2010). The set used in that study was derived from a regional thermal analysis conducted for the city of Leipzig by creating an index value for different land-use classes considering their thermal emissions. Emission index (i) =





emission (i) ∗ 100 − 100 emission (green space)

(1)

Schwarz et al. (2010) measured f-evapotranspiration using the following equation: f (i) =

max evapotranspiration(i)

ET(0)

(2)

with max evapotranspiration(i) as the maximal evapotranspiration for the considered land-use class and ET(0) as the reference evapotranspiration of 12 mm high grass in a given climate. Carbon storage was valued based on the results of an empirical study undertaken in the city of Leipzig by Strohbach and Haase (2012). The authors used the methodology developed by Yanai et al.

Author's personal copy 182

N. Larondelle, D. Haase / Ecological Indicators 29 (2013) 179–190

Table 2 Indicators for the modelled urban ecosystem services. Urban ecosystem service (UES)

Indicator

Calculation method

Reference

Climate regulation and air quality regulation

Surface emissivity

Emissivity values linked to land-use, look-up Table 3 Evapotranspiration values linked to land-use, look-up Table 3 Cooling potential of tree shadow, empirical data, look-up Table 4 Above-ground C-storage linked to land-use, empirical data, look-up Table 3

Schwarz et al. (2011)

f-Evapotranspiration Tree cooling potential Carbon sequestration Recreation, mental and physical health

Schwarz et al. (2011) Bowler et al. (2010) and own empirical research in the city of Leipzig Strohbach and Haase (2012) and Gibbs (2006)

Recreational area

Area [ha] (urban green, forest, water)

[ha] (Gill et al., 2008)

Recreational area per capita Accessibility of recreational area

Area [ha] (urban green, forest, water)/population size Buffer analyses using ArcGIS 10.0

[ha]

(2010) and adjusted it to the given case study. The most important driver influencing service provisioning is land-use. All of the values of ecosystem service performance with respect to related land-use classes are listed in Table 3. The degree of imperviousness, which serves as an overall proxy of the ability to provide urban ecosystem services, was quantified using empirical data according to Haase and Nuissl (2010). The tree cooling potential was valued by data derived from the literature review based on (i) a study by Bowler et al. (2010) reviewing recent literature on the use of urban green

300 m buffer analysis around settlements [ha] (Barbosa et al., 2007)

spaces to cool towns and cities and on (ii) empirical research in the city of Leipzig (unpublished data) (Table 4).

3.5. Data analysis and standardisation For comparative analysis of the different ecosystem services, their trade-offs and aggregated patterns, the calculated values for all five indicators, ‘surface emissivity,’ ‘f-evapotranspiration,’ ‘tree

Table 3 Indicator values for each land-use class of Corine Land Cover (CLC). CLC-code

CLC-class

f-Valuea

Surface emission (mean)a

MgC/ha

Degree of impervious cover [%]d

111 112 121 122 123 124 131 132 133 141 142 211 222 231 242 243

Continuous urban fabric Discontinuous urban fabric Industrial or commercial units Road and rail networks and associated land Port areas Airports Mineral extraction sites Dump sites Construction sites Green urban areas Sport and leisure facilities Non-irrigated arable land Fruit trees and berry plantations Pastures Complex cultivation patterns Land principally occupied by agriculture, with significant areas of natural vegetation Broad-leaved forest Coniferous forest Mixed forest Natural grasslands Moors and heathland Transitional woodland-shrub Beaches, dunes, sands Bare rocks Sparsely vegetated areas Burnt areas Inland marshes Peat bogs Salt marshes Intertidal flats Water courses Water bodies Coastal lagoons Estuaries Sea and ocean

0.8 0.9 0.8 0.8 0.8 0.8 1.0 1.0 1.0 1.1 1.0 1.1 1.1 1.1 1.1 1.1

143.2 139.4 141.5 145.1 139.9 139.9 137 139 134.8 134.3 138.4 138.9 141.4 135.4 136.6 135.7

4.65b 12.93b 8.52b 0b 8.52b 8.52b 0b 0b 0b 29.38b 5c 5c 16.03b 4.5c 5c 5c

95 60 90 60 85 85 20 20 20 20 40 0 0 0 0 0

1.1 1.3 1.2 1.1 1.1 1.1 1.0 1.0 1.0 1.4 1.4 1.2 1.2 1.2 1.4 1.4 1.4 1.4 1.4

134 137.4 132.8 135 137 136 136 139.1 139.1 139.1 140.4 140.4 140.4 140.4 131.3 131.3 131.3 131.3 131.3

68.31b 72.91b 75.71b 4.5c 4.5c 10.12b 10.12b 0b 10.12b 0b 0b 0b 0b 0b 0b 0b 0b 0b 0b

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

311 312 313 321 322 324 331 332 333 334 411 412 421 423 511 512 521 522 523 a b c d

Evapotranspiration potential (f-value) and emissivity values according to Schwarz et al. (2011). Mg carbon values according to Strohbach and Haase (2012). Mg carbon values according to Gibbs (2006). Imperviousness percentages according to Haase and Nuissl (2010).

Author's personal copy N. Larondelle, D. Haase / Ecological Indicators 29 (2013) 179–190 Table 4 Threshold values applied to assess the tree cooling potential UES. Land-use type

Temperature change (Tcity average − TArea assessed )

Reference

Urban parks Tree cover, forests

−0.945 ◦ C −3.5 ◦ C

100 m buffer around parks (least affected area)

−0.52 ◦ C

Bowler et al. (2010) Own empirical research in the city of Leipzig Mean value of temperature difference day/night according to Bowler et al. (2010)

cooling potential’, ‘carbon sequestration’ and ‘recreation area’, were standardised between 0 and 1 using the following equation:  =

 − min max − min

(3)

183

To analyse the total ecosystem services provisioning potential, the standardised values for the five considered services were recoded to more readily distinguish between areas with high values for service-provisioning and those with lower values: standardised values equalling or greater than 0.7 were set as 1, and lower values were set to 0. The exception is the indicator ‘surface emissivity’, for which values equal to or lower than 0.3 were set to 1, and all of the values above 0.3 were set as 0. This exception was made to ensure comparability between the polarities of the indicators, so that services with high values could be understood as having a high ecosystem service-provisioning potential. In order to guarantee the comparability of the data, some of the values were reversed in direction within the Spidergram figures (Fig. 5) to ensure that all of the high values could be read as providing high ecosystem service-provisioning potential. To assess the change of ecosystem service provisioning potential in Table 5, the maps were converted into raster files with a grid cell size of 100 m × 100 m using the central cell value (=centroid). All other calculations have been done using the shape format.

Fig. 1. Standardised values of urban ecosystem services along an urban–rural gradient (30 km) for the study sites of Berlin, Helsinki, Stockholm and Salzburg for the year 2006.

Author's personal copy 184

N. Larondelle, D. Haase / Ecological Indicators 29 (2013) 179–190

4. Results The spatial patterns of the studied ecosystem services in the four cities show both a considerable heterogeneity among the cities and similarities between coastal and non-coastal cities (Fig. 2). Berlin was characterised by a comparatively high degree of imperviousness across the entire NUTS3 region that does not fall below 0.13 and has its maximum at 2 km (distance from the city centre) with a value of 0.72 (Fig. 1). The values of surface emissivity showed similar patterns up to km 10 and remained overall rather high (above 0.35). After reaching a minimum around the city centre, the indicators of f-evapotranspiration, carbon storage and tree cooling potential rose up to km 13, where the spatial patterns began to change and the degree of imperviousness fell below the (standardised) values calculated for the urban ecosystem services indicators. Based on the analysis of ecosystem service-provisioning potential (Figs. 3–5), the highest provisioning areas were northwest of the core city and in the eastern part, whereas medium potential was found (in the form of spots) in the city centre. Regarding the change of green spaces for recreation in Berlin, all of the suitable land-use classes (urban green space, forest, and water) increased slightly across the entire NUTS3 region from 1990 to 2006. Due to both population decline (Fig. 4) and an increase in total city area, the amount of recreation area per capita that is accessible within a 300 m distance from settlement areas increased from 100 m2 in 1990 to 110 m2 in 2000 and 2006. The supply of urban ecosystem services provided by the core city was smaller than that provided by the NUTS3 region (Fig. 5), particularly regarding tree cooling potential. For surface emissivity, the values for both the core city and the NUTS3 region were fairly similar. The box and whisker plots displayed in Fig. 6.1 show a low level of variance for the indicator of surface emissivity over time, although indicating a slightly higher range for the first 15 km observed. A larger uncertainty over time was plotted for the urban green space distribution, which is shown in Fig. 6.3: the boxes do not show a linear form along the rural–urban gradient of Berlin. The highest values of uncertainty were found in the city centre and at a distance of 20 km. Helsinki was characterised by a comparatively low level of soil imperviousness, which was always below 0.4 but did not develop linearly along the urban–rural gradient: there were highs at km 3, 9, 11 and 20 and lows at km 6, 10, 13 and 19 (Fig. 1). After km 22, the degree of impervious cover remained stable between 0.1 and −0.2. Helsinki showed overall rather high values for f-evapotranspiration, which did not decrease below 0.4 and exhibited a maximum at km 7 (0.67). The ecosystem service-provisioning potential map (Fig. 3.6) for Helsinki indicates areas with the highest potential in the west and small, section-divided patterns along the western boundary of the core city. Regarding the development of recreation areas, the amount of total recreation space increased, especially the green space. Due to population growth, the green space per capita within a 300 m distance from settlement areas decreased slightly from 240 m2 in 2000 to 230 m2 per person in 2006. The mean values of surface emissivity and f-evapotranspiration in the core city were slightly higher than those of the NUTS3 region, contrasting with lower values of carbon storage and tree cooling potential in the core city. Salzburg was characterised by very low values of impervious cover, mainly due to its compact form or historical development. After a 3 km distance from the city centre, the values persisted below 0.2, with a maximum at km 1 of 0.4, which was low for an area close to the city centre compared to the other 3 sites (Fig. 1). The highest values for provisioning services were obtained by the tree cooling potential, which remained at a fairly high level above

0.6 after km 5, with a maximum of 1 in the peri-urban periphery at km 28. Overall, the NUTS3 region of Salzburg revealed rather high potentials for the provision of urban ecosystem services. Regarding the development of recreational areas, Fig. 4 shows an increase in all of the types of blue-green open spaces, whereas the per-capita green space within a 300-m buffer remained constant at 120 m2 due to population growth. The supply of urban ecosystem services within the core city of Salzburg, compared to the NUTS3 region, was higher regarding surface emissivity and f-evapotranspiration but much lower concerning tree cooling potential due to the absence of trees in the core city (Fig. 5). Box and whisker plots display small variances in the indicator values of surface emissivity, with no significant change along the urban–rural gradient (Fig. 6.2). High uncertainties, in contrast, could be detected for the green-space distribution (Fig. 6.4): variances in available green area seemed to increase over time, mainly within the first 5 km distance from the city centre and between km 15 and 20. The total amount of urban green space increased with increasing distance from the city centre. Stockholm can be described using an almost linear decline in the degree of imperviousness along the urban–rural gradient with a maximum at km 2 of 0.5 and a minimum at km 30 of 0.06. The values for f-evapotranspiration were very high on average, almost linearly rising to a maximum at km 30 with a value of 0.7. The values for surface emissivity showed a slight decrease over distance, exhibiting values between 0.3 and 0.5. The map of ecosystem service-provisioning potential (Fig. 3.8) shows very high potentials for urban ecosystem services for the core city of Stockholm, with the highest potentials around the city centre and some medium-valued spots directly within the core city. Concerning recreational areas, Fig. 4 shows an increase of urban green space of more than 2000 ha between 2000 and 2006. Due to considerable population growth in the same time period, the percapita recreational space nevertheless remained stable at a high level of 210 m2 with a 300-m buffer around settlements. The core city of Stockholm provided a lower potential of urban ecosystem services supply compared to the NUTS3 region (Fig. 5); only the values for carbon storage were similar. In summary, the urban–rural gradients for urban ecosystem service provisioning differed strongly among the four study sites. Compared to Stockholm, Salzburg and Helsinki, Berlin was characterised by a high proportion of sealed surface and a low amount of service supply by the core city. Salzburg, in contrast, showed quickly decreasing values for impervious cover after a few kilometres of distance from the city centre and high to very high values for overall ecosystem service provisioning, especially tree cooling potential. Stockholm stands out because of very high values for f-evapotranspiration, which increased almost linearly along the urban–rural gradient. Finally, Helsinki had the most unsteady development of sealed surface, exhibiting high indicator values along the sea-shore in the east and sprawling patterns in the north. Nevertheless, Helsinki had very high values for the overall services supply. Concerning the spatial patterns of urban ecosystem services supply (Fig. 2.1–2.4), Helsinki, Salzburg and Stockholm showed high potential for carbon storage within large, continuous areas of the NUTS3 region, whereas Berlin displayed less potential for carbon storage allocated at smaller and less-uniform patches around the city centre. The indicator of f-evapotranspiration (Fig. 2.5–2.8) showed a similar pattern as that derived for imperviousness (Fig. 3.1–3.4). Both the maps show the central position of Berlin and Stockholm within the NUTS3 region (high centrality) compared to Helsinki, which is limited by the sea to the east, and Salzburg, which is close to the national border at the north. The analysis of ecosystem service-provisioning potential (Fig. 3.5–3.8) displayed high values for Salzburg and Stockholm. The core city of Helsinki was surrounded by high-potential spaces

Author's personal copy N. Larondelle, D. Haase / Ecological Indicators 29 (2013) 179–190

185

Fig. 2. Standardised values of carbon storage potential, time step 2006, for the urban regions of Berlin (1), Helsinki (2), Salzburg (3) and Stockholm (4), and f-evapotranspiration for Berlin (5), Helsinki (6), Salzburg (7) and Stockholm (8).

but had a vast low-potential area along the seaside. Berlin showed the most heterogeneous pattern, with low potentials in the city centre, high potentials north and west of the centre and mediumlevel areas in the south. Both Helsinki and Stockholm had a lot of medium-valued areas in the city centre. Comparing the

ecosystem service provisioning between the core city and the NUTS3 region (Fig. 5), we found that the “urban” and dense-core cities were not necessarily weaker in their ability to provide ecosystem services. Helsinki and Salzburg, for example, reported higher values regarding surface emissivity and f-evapotranspiration

Author's personal copy 186

N. Larondelle, D. Haase / Ecological Indicators 29 (2013) 179–190

Fig. 3. Degree of imperviousness displayed for the urban regions of Berlin (1), Helsinki (2), Salzburg (3) and Stockholm (4) and UES potential for Berlin (5), Helsinki (6), Salzburg (7) and Stockholm (8). The darker the colour the higher the potential provisioning value of UES above average (standardised values ≥ 0.7 = 1 (except: surface emissivity ≤ 0.3 = 1)).

within their centres compared to the overall NUTS3 region. However, the ecosystem services bound to trees and forest cover, such as carbon storage and tree cooling potential, were higher outside the core city boundaries.

Looking at the change of the provisioning potential of urban ecosystem services from 2000 to 2006, the quantification results show that the case study cities differed greatly. For Berlin, a positive development regarding urban ecosystem service provisioning

Author's personal copy N. Larondelle, D. Haase / Ecological Indicators 29 (2013) 179–190

187

Fig. 4. Recreation green space availability (in ha) measured for a 300 m distance from settlement areas (CORINE class 111, 112) for each NUTS3 urban region.

could be detected. A large number of cells developed from nopotential areas to low-potential areas and some high-potential areas (Table 5). In Helsinki, there was a loss of no-potential and high-potential areas in favour of low-potential areas. Salzburg and Stockholm showed negative trends, exhibiting a loss of highpotential areas in favour of no-potential and low-potential areas. 5. Discussion 5.1. Spatial patterns and scales The patterns of high-value areas were highly heterogeneous with respect to the size of the area and the configuration of the Table 5 Changes for urban ecosystem service provisioning potential between 2000 and 2006 (in number of cells for the different assessment classes). Provisioning potential No potential Low potential Medium potential High potential

Berlin −1166 1035 3 140

Helsinki

Salzburg

Stockholm

−376 494 9 −120

562 61 1 −593

4367 −3 −282 −4006

service supply. All four of the cities featured very different urban–rural gradients of ecosystem service provisioning. A primary finding of this study is that there were neither typical spatial patterns of urban ecosystem services supply, such as radio-symmetric and central–peripheral patterns suggested by urbanisation theory or other models (Haase and Nuissl, 2010), nor a typical rural–urban gradient of service supply. More urban studies across Europe could help define such regular patterns or models that would allow for more general assumptions about the spatial configuration of urban ecosystem service provisioning. High ecosystem service-provisioning potential was generally found in areas with low impervious cover. Hot-spot regions within the cities and the NUTS3 regions were often characterised by forests or forested areas because f-evapotranspiration, surface emissivity and carbon storage have a high ecological value in areas with a tree canopy. Interestingly, we did not detect patches where all of the indicators reported high values; thus, optimal sites for ecosystem service provisioning were not found in the four study sites. We found differences in ecosystem service provisioning between the NUTS regions and the core cities. This result suggests that there are services that are better supplied within city boundaries (at a local scale) and others that are more effectively supplied

Author's personal copy 188

N. Larondelle, D. Haase / Ecological Indicators 29 (2013) 179–190

Fig. 5. Spidergrams comparing the urban ecosystem service (UES) potential including five ecosystem services and their trade-offs for two different spatial scales, core city and NUTS3 region, for the urban regions and cities of Berlin, Helsinki, Salzburg and Stockholm (2006).

by forests and large open spaces in the surroundings. “For most general ecosystem services, the share generated by ecosystems within the urban area is expected to be limited compared to the total service” (Bolund and Hunhammar, 1999:300). This statement was found to be only partially true because local climate regulation in the cities of Salzburg and Helsinki was high, resulting from a high share of green urban areas combined with a relatively low share of impervious cover. Fig. 5 shows that the forms of the supply spider diagrams are comparable between the core cities and the NUTS3 regions. This conformity of value distribution between the core city and the NUTS3 region was found even though the core cities of Helsinki and Stockholm only formed approximately 3% of the overall NUTS3 region, Salzburg formed almost 4%, and Berlin formed up to 31%. Thus, we can argue that the NUTS3 region is a reasonable proxy to characterise and estimate the ecosystem service-provisioning potential of a larger city and vice versa, although not the level of the indicator values. In summary, the core cities of the assessed study sites did not provide fewer ecosystem services compared to the entire NUTS3

regions, but overall, no patches were detected where all of the indicators had equally high values at the same time. 5.2. Demand and supply Different proxies for the valuation of urban ecosystem service provisioning have been discussed in this study. However, assessing urban ecosystem service demand requires additional data. A good proxy to address the demand for the regulating services fevapotranspiration, tree cooling potential and surface-emissivity could be the number of extreme heat days per year or heat-related deaths per year. For carbon storage, the demand results from the CO2 emissions per capita. To value the demand for recreational services, per-capita green space targets and the number of tourists as well as an accessibility indicator could be used (Chiesura, 2004; Herzele and Wiedemann, 2003). With further development, the developed ecosystem service quantification model could be used when balancing the supply and demand of urban ecosystem services for cities and urban regions.

Author's personal copy N. Larondelle, D. Haase / Ecological Indicators 29 (2013) 179–190

189

Fig. 6. Boxplots showing the uncertainties over time of two UES indicators, local climate regulation (expressed by the indicator of surface emissivity) and recreation area, along the urban–rural gradient for Berlin (1, 3) and Salzburg (2, 4). The plots were generated using standardised mean values from the years 1990, 2000 and 2006.

5.3. Uncertainties The data and model uncertainties of the present study are mainly related to the input data (land-use) and the respective aggregation and disaggregation procedures to assign model values to these classes (at both the core city and NUTS3 region levels). The most severe level of uncertainties was derived from the use of regional empirical field data collected in other studies and for different sites, which causes uncertainties in terms of the transferability of these indicators/models to the sites investigated and the climate zones assessed, respectively. The assignment of indicator values from other studies to the land-use/cover classes of Corine Land Cover represents another source of uncertainty. When interpreting the results obtained, it is crucial to know the cities and their specific land-use development. Thus, to reduce uncertainties in data assessment and interpretation, we recommend involving local/regional experts and planners in such studies. Within the URBES project, communication with local scientists and practitioners is therefore an essential aspect. 5.4. Application of the results in policy and planning According to Zipperer et al. (2000), land-use decisions must be made along an urban-to-rural continuum because it is essential to incorporate ecological principles into the decision-making process to use and maintain resources sustainably. Because there is a growing need to address ecosystem services in impact assessments and planning processes, it is critical to provide comprehensive studies. The findings of this particular study can be effectively used for urban planning in different ways: (1) to provide an overview

of what ecosystem services are supplied in the core cities and their surrounding regions, (2) to show the spatial patterns of ecosystem services supply, which gives planners the opportunity to compare these data with neighbourhood and infrastructure planning as well as development plans, e.g., concerning ecosystem services interferences with other plans and alternatives, and (3) to initiate communication between urban and regional planning. (4) Additionally, improved communication among urban planners, policy-makers and inhabitants can lead to greater socialenvironmental justice at the city level, and (5) reduced uncertainty and better quantitative knowledge will help to reduce the growing ecological footprint of cities.

6. Conclusions This study contributed to a better understanding of how urban ecosystem service provisioning is distributed across a number of European cities and how this provisioning has changed. Furthermore, the study provides evidence that core cities provide a range of ecosystem services, and thus, it is worth emphasising environmental issues within urban planning and governance in the future. The present study clearly demonstrated the complexity of ecosystem services performance along the urban–rural gradient and thus adds to the understanding of both land-use and ecosystem change in an increasingly human-dominated world. The study further responds to the growing need for data and information on the values of ecosystem(s) services for urban citizens, particularly because the 21st century will be an increasingly urban one. The model presented can also be used to identify synergies and trade-offs between

Author's personal copy 190

N. Larondelle, D. Haase / Ecological Indicators 29 (2013) 179–190

different urban ecosystem services which should be in the focus of further research. Acknowledgements We thank the URBES team for cooperation and the EU/BMBF for funding the project (http://www.biodiversa.org/121). Further thanks go to our colleagues from the New School New York for many fruitful discussions about UES and our team at the Humboldt University for cooperation. Finally, we thank Emily Lorance Rall for her extremely useful comments on an earlier version of this manuscript. References Andersson, E., Ahrné, K., Pyykönen, M., Elmqvist, T., 2009. Patterns and scale relations among urbanization measures in Stockholm. Sweden Landscape Ecol. 83, 1331–1339. Barbosa, O., Tratalos, J., Armsworth, P., Davies, R., Fuller, R., Johnson, P., Gaston, K., 2007. Who benefits from access to green space? A case study from Sheffield, UK. Landscape Urban Plan. 83 (2–3), 187–195, http://dx.doi.org/10.1016/j.landurbplan.2007.04.004. Bolund, P., Hunhammar, S., 1999. Ecosystem services in urban areas. Ecol. Econ. 29, 293–301. Bowler, D.E., Buyung-Ali, L., Knight, T.M., Pullin, A.S., 2010. Urban greening to cool towns and cities: a systematic review of the empirical evidence. Landscape Urban Plan. 97 (3), 147–155, http://dx.doi.org/10. 1016/j.landurbplan.2010.05.006. Burkhard, B., Petrosillo, I., Constanza, R., 2010. Ecosystem services – bridging ecology, economy and social sciences. Ecol. Complex. 7, 257–259. Burkhard, B., Kroll, F., Nedkov, S., Müller, F., 2012. Mapping supply, demand and budgets of ecosystem services. Ecological Indicators 21, 17–29. Carpenter, S.R., Mooney, H.A., Agard, J., Capistrano, D., Defries, R.S., Díaz, S., Dietz, T., Duraiappah, A.K., Oteng-Yeboah, A., Pereira, H.M., Perrings, C., Reid, W.V., Sarukhan, J., Scholes, R.J., Whyte, A., 2009. Science for managing ecosystem services: beyond the millennium ecosystem assessment. Proc. Natl. Acad. Sci. U.S.A. 106 (5), 1305–1312, http://dx.doi.org/10.1073/pnas.0808772106. Chen, X.-L., Zhao, H.-M., Li, P.-X., Yin, Z.-Y., 2006. Remote sensing image-based analysis of the relationship between urban heat island and land use/cover changes. Remote Sens. Environ. 104, 133–146. Chiesura, A., 2004. The role of urban parks for the sustainable city. Landscape Urban Plan. 68, 129–138. Daniel, T.C., Muhar, A., Arnberger, A., Aznar, O., Boyd, J.W., Chan, K.M., Costanza, R., Elmqvist, T., Flint, C.G., Gobster, P.H., Grêt-Regamey, A., Lave, R., Muhar, S., Penker, M., Ribe, R.G., Schauppenlehner, T., Sikor, T., Soloviy, I., Spierenburg, M., Taczanowska, K., Tam, J., von der Dunk, A., 2012. Contributions of cultural services to the ecosystem services agenda. Proc. Natl. Acad. Sci. U.S.A. 1–8, http://dx.doi.org/10.1073/pnas.1114773109. Decker, E.H., Elliott, S., Smith, F.A., Blake, D.R., Rowland, F.S., 2000. Energy and material flow through the urban environment. Annu. Rev. Energy Environ. 25, 685–740. Elmqvist, T. et al., 2012. The URBES Project. Factsheet 1. https://cmsdata. iucn.org/downloads/urbes brochure final print.pdf Gibbs, H.K., 2006. Olson’s Major World Ecosystem Complexes Ranked by Carbon in Live Vegetation: An Updated Database Using the GLC2000 Land Cover Product. Retrieved from http://cdiac.ornl.gov/epubs/ndp/ndp017/ndp017b.html Gill, S.E., Handley, J.F., Ennos, A.R., Pauleit, S., Theuray, N., Lindley, S.J., 2008. Characterising the urban environment of UK cities and towns: a template for landscape planning. Landscape Urban Plan. 87, 210–222. Haase, D., 2012. The nature of urban land use and why it is a special land use case. Unpublished Manuscript for the Ernst Strüngmann Forum: Rethinking Global Land Use in an Urban Era, Frankfurt, September 23–28, 2012.

Haase, D., Nuissl, H., 2007. Does urban sprawl drive changes in the water balance and policy? The case of Leipzig (Germany) 1870–2003. Landscape Urban Plan. 80, 1–13. Haase, D., Nuissl, H., 2010. The urban-to-rural gradient of land use change and impervious cover: a long-term trajectory for the city of Leipzig. J. Land Use Sci. 5, 123–141. Haase, D., Schwarz, N., Strohbach, M., Kroll, F., Seppelt, R., 2012. Synergies, tradeoffs, and losses of ecosystem services in urban regions: an integrated multiscale framework applied to the Leipzig-Halle Region, Germany. Ecol. Soc. 17 (3), http://dx.doi.org/10.5751/ES-04853-170322. Herzele, A., Wiedemann, T., 2003. A monitoring tool for the provision of accessible and attractive urban green spaces. Landscape Urban Plan. 63, 109–126. Jansson, Å., 2012. Reaching for a sustainable, resilient urban future using the lens of ecosystem services. Ecol. Econ., http://dx.doi.org/10.1016/j. ecolecon.2012.06.013. Kabisch, N., Haase, D., 2011. Diversifying European agglomerations: evidence of urban population trends for the 21st century. Popul. Space Place 17, 236–253. Kottmeier, C., Biegert, C., Corsmeier, U., 2007. Effects of urban land use on surface temperature in Berlin: case study. J. Urban Plan. Dev. 133, 128–137. Kroll, F., Müller, F., Haase, D., Fohrer, N., 2011. Rural–urban gradient analysis of ecosystem services supply and demand dynamics. Land Use Policy, http://dx.doi.org/10.1016/j.landusepol.2011.07.008. Larondelle, N., Haase, D., 2012. Valuing post-mining landscapes using an ecosystem services approach—an example from Germany. Ecol. Indic. 18, 567–574, http://dx.doi.org/10.1016/j.ecolind.2012.01.008. Lautenbach, S., Kugel, C., Lausch, A., Seppelt, R., 2011. Analysis of historic changes in regional ecosystem service provisioning using land use data. Ecol. Indic. 11, 676–687. Luck, M., Wu, J., 2002. A gradient analysis of urban landscape pattern: a case study from the Phoenix metropolitan region, Arizona, USA. Landscape Ecol. 17, 327–339. Marcuse, P., Connolly, J., Novy, J., Olivio, I., Potter, J., Steil, J. (Eds.), 2009. Searching for the just city. Routledge, New York. Mitchel, D., 2003. The right to the city – social justice and the fight for public space. Ethics Place Environ. 6 (1), 273–281, http://dx.doi.org/10.1080/ 1366879042000200624. Rees, W., Wackernagel, M., 1996. Urban ecological footprints: why cities cannot be sustainable—and why they are a key to sustainability. Environ. Impact Assess. Rev. 16, 223–248. Schwarz, N., 2010. Urban form revisited – Selecting indicators for characterising European cities. Landscape and Urban Planning 96 (1), 29–47. Schwarz, N., Haase, D., Seppelt, R., 2010. Omnipresent sprawl? A review of urban simulation models with respect to urban shrinkage. Environ. Plan. B: Plan. Des. 37, 265–283. Schwarz, N., Bauer, A., Haase, D., 2011. Assessing climate impacts of planning policies: an estimation for the urban region of Leipzig (Germany). Environ. Impact Assess. Rev. 31, 97–111. Seto, K.S., Reenberg, A., Boone, C.C., Fragkias, M., Haase, D., Langanke, T., Marcotullio, P., Munroe, D.K., Olah, B., Simon, D., 2012. Teleconnections and sustainability: new conceptualizations of global urbanization and land change. Proc. Natl. Acad. Sci. U.S.A., http://dx.doi.org/10.1073/pnas.1117622109. Strohbach, M.W., Haase, D., 2012. Above-ground carbon storage by urban trees in Leipzig, Germany: analysis of patterns in a European city. Landscape Urban Plan. 104 (1), 95–104, http://dx.doi.org/10.1016/j.landurbplan.2011.10.001. Strohbach, M.W., Arnold, E., Haase, D., 2012. The carbon mitigation potential of urban restructuring – a life cycle analysis of green space development. Landscape and Urban Planning 104, 220–229. TEEB – The Economics of Ecosystems and Biodiversity, 2011. TEEB Manual for Cities: Ecosystem Services in Urban Management. Yanai, R., Battles, J., Richardson, A., Blodgett, C., Wood, D., Rastetter, E., 2010. Estimating uncertainty in ecosystem budget calculations. Ecosystems 13 (2), 239–248. Zipperer, W.C., Wu, J., Pouyat, R.V., Pickett, S.T., 2000. The application of ecological principles to urban and urbanizing landscapes. Urban Ecosyst., 685–688.