Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

5.1 Introduction

More than half of the global human population is living in urban areas, and the trend towards further urbanization is strongly increasing (MEA 2005; United Nations 2008). Hence, the majority of people globally will experience “nature” and related ecosystem services primarily within the urban fabric (Gilbert 1989; McKinney 2002; Miller and Hobbs 2002; Miller 2005; Goddard et al. 2010). There is increasing evidence that urban land uses affect profound changes in all environmental components and that humans are the main drivers of change (Sukopp et al. 1979; Pickett et al. 2001; Alberti et al. 2003; Grimm et al. 2008). Urban growth has been identified as a major threat to biodiversity (e.g. Hansen et al. 2005), but at the same time, urban regions can harbour an array of species (Sukopp and Werner 1983; Gilbert 1989; Pyšek 1993; McKinney 2002) and contribute to the conservation of biodiversity. However, distinct urban ecosystems cannot replace totally the habitat function of (near-)natural systems (Kowarik 2011).

As plants provide an array of provisioning, regulating, and cultural ecosystem services in urban regions (Tzoulas et al. 2007), biodiversity is also of paramount importance from a social perspective, due to the simple fact that nature is here at the nearest proximity to people (Miller and Hobbs 2002). Exposure to natural systems has been found to positively affect human well-being and health (Kaplan and Kaplan 1989; Mitchell and Popham 2008). The biodiversity of urban green matters beyond the simple function of biomass. In green spaces of Sheffield, the degree of psychological benefit for people was positively correlated with species richness of plants and to a lesser extent of birds (Fuller et al. 2007).

A better understanding of the ways in which biodiversity is affected by urbanization and how plant species may thrive in urban settings is needed to optimize strategies aiming at conserving and enhancing urban nature. Urban ecology has a long history, mainly in Europe (Sukopp 2002) and has developed rapidly over the last few decades, as illustrated by Marzluff et al. (2008). The Danish plant geographer Joakim Schouw (1823) described “Plantae urbanae” in the early nineteenth century and was among the first who found that distinct plant species have an affinity for urban conditions. A wealth of urban studies has revealed thus far an array of biotic responses to urbanization, including changes in phytodiversity patterns at different scales.

At the regional scale, cities appear to be hot spots of plant species richness with usually higher numbers of plant species compared with rural surroundings (Haeupler 1975; McKinney 2002; Hope et al. 2003; Kühn et al. 2004; Knapp et al. 2009). This holds surprisingly for both native and introduced species (Kühn et al. 2004). A common feature of urban floras is a high proportion of introduced species (Kowarik 1995; Pyšek 1998). Urban floras of 54 European cities have an average proportion of 40% alien species, ranging from 20 to 60% (Pyšek 1998). For some North American cities, a similar range (19–46%) and average (35%) has been reported (Clemants and Moore 2003). The richness of introduced species in urban regions has been explained by cities functioning as important points of entry and as foci for the secondary release of introduced species, with trade, traffic, and horticulture as most prominent dispersal pathways (Hodkinson and Thompson 1997; Dehnen-Schmutz et al. 2007; Kowarik and von der Lippe 2007).

Generally, the number of plant species increases with the size of the city and the human population (Klotz 1990; Pyšek 1993). Comparisons between the floristic assemblages of urbanized and rural areas in Germany revealed that floras of urbanized areas have a reduced phylogenetic diversity with a few species-rich lineages (Knapp et al. 2008a), and many species traits of urban plant species differ significantly from species assemblages of rural areas (Knapp et al. 2008b, 2009).

Changes in urban climate, hydrology, and soils due to urban land use have resulted in a heterogeneous mosaic of highly fragmented sites that add a range of dry, warm, and nutrient-rich habitats to the previously existing set of sites (Sukopp et al. 1979; Oke 1982; Gilbert 1989; Sukopp and Wittig 1998; Pickett et al. 2001). Correspondingly, grid comparisons show that urban regions harbour fewer species with hygromorphic leaves than rural areas, but more species with scleromorphic or succulent leaves and species that are indicators of nutrient-rich, warm, and dry conditions (Knapp et al. 2009). However, as the data from grid cells usually represent highly heterogeneous areas, it is difficult to scale down interpretations to habitat scale.

At the city level, variation in species richness and composition has been abundantly analyzed along urban-rural gradients (McDonnell and Hahs 2008). In one of the first gradient studies, Kunick (1982) found total plant species richness to peak in the transition zone between densely built-up inner-city areas of Berlin and the adjacent areas, while most non-native species have been found in the urban core. A few analyses of changes in urban flora over large periods of time revealed a significant turn-over in species composition, an increase in total species richness, and a decline in rare native species and archaeophytes (pre-1500 aliens) while the number of neophytes (post-1500 aliens) markedly increased. However, neophytes are generally less frequent compared to native species (cf. Kowarik 1990 for Berlin; Chocholoušková and Pyšek 2003 for Plzẽ; Knapp et al. 2010 for Halle).

Urbanization increases the fragmentation of habitats (Alberti 2005; Robinson et al. 2005) and integrating habitat features into analyses of large floristic data sets helps identifying the role of individual land use types for species richness (e.g. Celesti-Grapow et al. 2006 for Rome). Analyses at the habitat scale in Berlin showed that about two-thirds of native plant species were able to colonize sites subject to high levels of human-mediated disturbances, whereas the remaining species were confined to more natural sites. Numbers of introduced species peaked at profoundly changed sites, those of native species on sites subject to a medium level of human-induced disturbance (Kowarik 1990). At the community scale, introduced species were most prevalent in vegetation types that can be assigned to early succession stages or to naturally disturbed ecosystems such as pioneer sites in floodplains (Kowarik 1995). For an array of land use types in Berlin, ranging from (near-)natural to profoundly transformed by human agency, a wealth of studies, summarized in Sukopp (1990), illustrate the environmental conditions of various urban habitats and the corresponding species assemblages–revealing Berlin as one of the best-studied urban systems in Europe.

Characteristics of total urban floras and spatial patterns of plant species has been broadly analyzed in many cities, and a couple of studies suggest that biodiversity patterns in urban settings are deeply affected by human-driven mechanisms that lead to habitat changes but also drive the species composition by setting filters for plant selection and management of urban green spaces (Alberti et al. 2003; Grimm et al. 2008; Kowarik 2011). Analyses of large data sets at the city scale have revealed general trends in urban biodiversity patterns (e.g. Knapp et al. 2009), but it is yet challenging to disentangle the underlying mechanisms as different kinds of factors may interfere with each other at the habitat level.

The first three studies completed in the frame of the graduate program “Urban Ecology Berlin” aimed at a better understanding of patterns of urban phytodiversity by disclosing the functioning of possibly interfacing key mechanisms that are related to urban climate, human-mediated dispersal, and the effects of environmental parameters at different scales. Further ongoing studies address different approaches to enhance ecosystem functions on urban land.

5.2 Towards a Better Understanding of Mechanisms Underlying Biodiversity Patterns

Profound changes in urban climate with increased temperatures, less cold winters, and a longer vegetation period have often been hypothesized to affect the distribution of plant species (e.g. Sukopp and Wurzel 2003). As other than habitat-related factors could also drive the spatial configuration of species assemblages in urban settings, the first study aims at identifying consequences of the urban heat island on the performance of woody species that have become abundant in urban regions (Säumel 2006).

The occurrence of most plant species is strictly dispersal-limited, and the high fragmentation of urban habitats is supposed to profoundly affect the distribution of plant species (Bierwagen 2007). At the same time, the frequency of human-mediated dispersal processes is expected to peak in urban regions and could possibly counteract dispersal limitations of plants in urban settings. The second study thus addresses the question of to which extent traffic on urban motorways functions as dispersal vector of plant species (von der Lippe 2006).

Urbanization is multidimensional and key mechanisms cannot be combined into aggregated variables such as population density or total paved areas (Alberti et al. 2003). While the first two studies aim at identifying the functioning of distinct key mechanisms in urban settings, the third study addresses the complex interrelationships of different environmental and landscape parameters. Taking urban wastelands as model ecosystems, this study aims at disentangling the relative importance of environmental and landscape factors at the habitat and city scales that drive species composition and establishment (Westermann 2009).

5.2.1 Response of Urban Tree Species to Increased Temperatures

Global warming is expected to enhance invasions by non-native species (Dukes and Mooney 1999; Walther 2004) and effects of urban heat islands may anticipate impacts of global warming. Although urban heat islands have often been hypothesized to affect the survival, spread and distribution of non-native species in urban settings (e.g. Sukopp and Wurzel 2003), studies on the response of species to increased temperatures in terms of morphological and allocational traits are rare. Aiming to disentangle species’ responses to changing temperature, we tested by experimental approaches the hypothesis that urban heat islands favour the growth of non-native species.

As model species, we used the native Norway maple (Acer platanoides) and two non-native tree species, tree of heaven (Ailanthus altissima) and box elder (Acer negundo). All are successful colonizers of urban habitats in Central European cities. However, only Ailanthus altissima is largely confined to metropolitan agglomerations, which has been hypothesised as being an effect of urban heat islands (Sachse et al. 1990; Kowarik and Säumel 2007).

In a field experiment, we exposed potted one-year-old saplings of the model species along an urban-rural gradient from the centre to the hinterlands of Berlin. Saplings of the same species have also been exposed for 2 years in climate chambers to assess their response to different temperature regimes: low temperature (10/05°C), elevated temperature (20/15°C), and control (15/10°C). We determined whether a moderate temperature increase or decrease alters growth and leaf display of the exposed saplings and analyzed more than 40 foliar, stem, and root traits to evaluate how temperature influences the plasticity of plant morphology. We here summarize the results by using a qualitative and quantitative classification approach (Table 5.1).

Table 5.1 Summary of interspecific variation in overall response of three tree species (Ailanthus altissima, Acer negundo, A. platanoides) due to warming (elevated temperature, ET) or chilling (low temperature, LT) compared to the control
Full size table

5.2.1.1 Intra- and Interspecific Variations

Our results illustrate the importance of temperature for growth and survival of tree species that successfully colonize urban habitats. The field experiment revealed the duration of frost periods and the amount of frost intensities as crucial factors for survival of the early developmental stages. Both Acer species performed higher tolerance to frost injury than Ailanthus, resulting in differing survival rates after field exposure during a strong winter. The survival rates in the Acer species decreased with declining influence of the urban climate, while all individuals of Ailanthus died due to early deep frost (von der Lippe et al. 2005).

In the climate chamber experiments, moderate increase or decrease in temperature altered saplings architecture and growth, morphology, and biomass allocation (Table 5.1). In general, the overall species’ growth performance benefits from warming. In Ailanthus, chilling led to fragile saplings with a likely low competitive strength. Species differed significantly in temperature sensitivity of stem, branch, leaf, and root traits. Generally, Ailanthus showed the greatest response, while Acer platanoides showed a low and Acer negundo an intermediate level of response to temperature.

In Ailanthus, most traits related to size and number of plant modules tend to increase with increasing temperature and to decrease with decreasing temperature. Stem, branch, and leaf traits of both Acer species mainly decreased with decreasing temperature but did not change due to warming (Table 5.1). Exceptions of these general patterns were found in branch and leaf traits. In Acer negundo, branch traits and leaf number increased with decreasing temperature. Independently of the temperature regime, Acer negundo reached the greatest main stem height of all species, which suggests a high competitiveness.

In Ailanthus, the relative growth rate was markedly reduced due to chilling, while warming strongly enhanced related parameters. Chilling sensitivity was high in Ailanthus and low in both Acer species. Consistently, Ailanthus had a narrow tolerance range of temperature, while the Acer species acclimated to a wide temperature range (Table 5.1). Air temperature sum on the day of the first bud break was lowest in Acer negundo, intermediate in Acer platanoides, and highest in Ailanthus. Time to the first bud break was strongly shortened by warming in all species.

5.2.1.2 Phenotypic Plasticity

Phenotypic plasticity is the ability of an organism to express different phenotypes depending on the biotic or abiotic environment (Bradshaw 1965). A phenotypically plastic plant is able to optimize growth in an altered environment, while a phenotypically stable or fixed plant may suffer from a reduced growth rate. Many authors argue that a high level of plasticity enhances the invasion potential of introduced species (Davis et al. 2000; Yamashita et al. 2002; Durand and Goldstein 2001; Huxman and Smith 2001; Daehler 2003). Our study reveals a divergent plasticity of morphological and allocational traits in the focal species. All are invasive, but only Ailanthus performed a high level of plasticity, while Acer negundo showed an intermediate and Acer platanoides a low level of plasticity to temperature. We thus conclude that phenotypic plasticity may enhance the invasion of urban habitats by introduced species (cf. Ailanthus), but is no obligatory prerequisite for such a colonisation success (cf. Acer platanoides).

As a second important result, we found that the amount of plasticity depended on the direction of temperature change (i.e. chilling or warming). The maple species, but not Ailanthus, exhibited in most traits a significant higher plasticity to chilling than to warming. In allocational and growth traits, all species showed higher plasticity to warming than to chilling. The growth performance of Ailanthus altissima in the cold climate chamber highlighted that high plasticity not necessarily leads to higher fitness and invasiveness (see Scheiner 1993a, b; Via et al. 1995) and cannot be considered as adaptive (Sultan 1992, 1995).

5.2.1.3 Linking Experimental Results with Distribution Patterns

Our experimental results support the hypothesis that urban heat island effects favour the survival of juvenile trees in winter. That holds surprisingly also for the native Norway maple. In both maple species, an enhanced allocation of biomass to roots in the first growing period may support saplings to withstand frost injury and to take advantage of increased access to water and nutrients in the subsequent resprouting phase. Maple saplings exposed in the urban hinterland showed a decreased survival, but the whole cohort did not die-off as did Ailanthus saplings. This result corresponds well with the overall distribution pattern of both maple species, which are not restricted to urban habitats in Central Europe.

As Ailanthus is more sensitive to chilling than both maple species, mild winters seem to be a crucial prerequisite for its establishment. Hence, reduced frost periods and intensities due to the urban heat island are supposed to enhance population growth in this species and contribute to explain its confinement to urban centres in Central Europe.

The climate chamber experiments revealed significant differences in overall temperature sensitivity of the exposed species that can be well related to their distribution patterns. Ailanthus’ growth is favoured by warming and strongly limited by chilling, while Acer negundo and Acer platanoides have a wide temperature tolerance (Table 5.1). Compared to both maple species, Ailanthus showed a substantial growth reduction due to chilling and a conspicuous increase in growth due to warming. In addition, a prolonged growing season as a typical outcome of urban climate will also favour the spread of A. altissima.

Water limitations and increased hydraulic stress as consequences of warming are supposed to hamper plant establishment. Our results provide strong evidence that Ailanthus meets this challenge much better than the Acer species. In the warm climate chamber, we found a reduced allocation to leaves as transpiring tissue, contrasting to an increased allocation to roots, and within the root system to fine and coarse roots. This will enhance the resource exploitation (water, nutrients) and reduce water loss by transpiration. The increased stem volume provides a larger cross-sectional area for vascular movement of water, and the positive correlation of cumulative foliar biomass and stem basal area indicates a better water supply for the whole plant. These findings amplify previous results on water-saving mechanisms (Trifilo et al. 2004) and deep below ground exploration (Pan and Bassuk 1986) in Ailanthus.

Due to a higher leaf turnover in the warm chamber, Ailanthus produced much more leaf litter than the other species. This might enhance its competitive ability because increased temperatures enhance allelopathic effects of the leaflets (Lawrence et al. 1991). While stem growth patterns in Ailanthus and A. platanoides can be attributed to the ontogenetic drift due to warming, the branching patterns of A. negundo might reflect a functional adaptation to chilling. Our results thus suggest a high adaptive capacity of Acer negundo across a broad range of temperatures, which contributes to explain its invasion success under both cold and warm climates.

The low responsiveness of A. platanoides to temperature clearly demonstrates the limits in generalizing positive effects of warmer temperatures as drivers of plant invasions. In summary, A. platanoides showed a low plasticity of morphological, allocational, and growth traits due to warming. Correspondingly, other factors than recent warming should have driven the successful spread of this species in North America and Europe. In Central Europe, increased nitrogen inputs are supposed to favour the colonisation of urban habitats and forests (Sachse et al. 1990). In North America, A. platanoides might benefit from more efficient light, water, or nutrient use compared to native broad-leaved species (Kloeppel and Abrams 1995) and of biotic interactions. The latter include a lower seed predation compared to native congeners (Meiners and Handel 2000) and reduced inhibitory effects of soil biota compared to its native range (Reinhart and Callaway 2004).

5.2.1.4 Anticipated Effects of Global Warming

To predict future effects of climate change on the abundance and distribution of species, a “space-for-time” approach has been used by linking current distribution patterns with actual climatic conditions (e.g. Iverson and Prasad 1998; Thomas et al. 2004). As cities actually show distinct urban heat island effects in different climatic regions (Arnfield 2003), they can be used as laboratories to analyze anticipated effects of globally increased temperatures on plant performance and resulting distribution patterns. Our results suggest that global warming will enhance invasions by some, but not by all, species that currently successfully spread in urban habitats. As a response to global warming, Ailanthus is expected to spread far beyond urban areas in Central Europe and to expand its range northwards. In contrast, the spread of both maple species will be less triggered (Acer negundo) – or not at all (A. platanoides) – by increased temperatures.

5.2.2 Vehicles as Dispersal Vectors of Plant Species

A dense road network is one of the most distinctive features that differentiate urban regions from their rural surroundings. The role of this omnipresent linear structure both for the fragmentation and connectivity of urban plant populations is still little understood. Distribution patterns of plants along roadsides suggest an important role of traffic as a dispersal vector. In particular, large gaps between newly established populations and the next known seed sources can frequently be observed during roadside expansion of plant species (Scott and Davison 1985; Griese 1996; Lavoie et al. 2007). This phenomenon points to a recurrent impact of human-mediated long-distance dispersal on population dynamics in this habitat.

Urban roadsides usually harbour a large number of non-native species. The flora of urban roadsides in Berlin housing areas contains 36% non-native species (Langer 1994), and roadsides on an urban to rural gradient in East Berlin contained 22% neophytes (Schmitz 2000). A similar percentage of 26% of introduced species was reported for the roadside flora in an urban area in South Africa (Cilliers and Bredenkamp 2000). Verges of high-use roads were shown to comprise more non-native species and a higher non-native plant cover than abandoned roads or unpaved roads of the same region (Parendes and Jones 2000; Gelbard and Belnap 2003), suggesting an impact of traffic density on roadside plant composition. Records of roadside invasions over time often revealed a very rapid spread along roadside corridors that by far exceeds the primary dispersal ability of the plant species involved (Ernst 1998; Pyšek et al. 2002; Heger and Böhmer 2005).

Although roadsides can be acknowledged as migration corridors this way, the vector, i.e. the underlying processes that cause these distributional patterns, can only be retraced indirectly. The possible processes leading to plant migration along roadsides can be divided into two main causes: (1) altered site conditions at roadsides that provide a suitable migration corridor and (2) seed dispersal by vehicles. The relative contribution of each of these mechanisms is hard to measure, as the outcome, that is, distributional patterns, integrates both of them. While the habitat characteristics of the corridor that facilitate plant migrations along roads are widely studied experimentally (Tyser and Worley 1992; Pauchard et al. 2003; Watkins et al. 2003; Johnston and Johnston 2004; Truscott et al. 2005), the key features of the main vector that possibly drives dispersal along roads, that is, dispersal by vehicles, are barely understood. The aim of this study was therefore to identify the mechanisms and properties of plant dispersal by vehicles and its role in plant migrations along roads. A focus was on the facilitation of plant invasions by this vector.

Adhesive dispersal by vehicles has long been suggested to play a prominent role in roadside expansions of plants (Ridley 1930). By taking samples of mud from the surface of vehicles, a handful of studies have demonstrated that adhesive seed transport by vehicles is actually possible and obviously correlates to the amount of unpaved surfaces a vehicle is driven through (Clifford 1959; Schmidt 1989; Lonsdale and Lane 1994; Hodkinson and Thompson 1997; Zwaenepoel et al. 2006). A different method for seed sampling was used by Wace (1977) who sampled the cumulative seed deposition inside a carwash by monthly samples of mud from the settling tanks. Although these studies revealed a high potential for human-mediated dispersal by vehicles, it is difficult to project the quantitative impacts on roadside plant populations from these data as both the rate of seed deposition and the transport distances remain unknown.

The limited knowledge about the effect of traffic as a dispersal vector is obviously due to the methodological limitation that seed deposition along roads cannot easily be separated in the shares that originate either from vehicle dispersal or from other adjacent seed sources. Simply exposing seed traps to roadsides would integrate both sources. A crucial prerequisite for this study was therefore to develop an experimental approach that could reliably estimate the magnitude of seed deposition by vehicles at roadsides as well as the spatial effectiveness and direction of this vector. Looking for sampling sites which are insulated from dispersal vectors other than traffic, we finally found a series of three long motorway tunnels in the north-western part of Berlin. To standardise the sampling procedure, a special kind of seed trap had to be designed, which could be mounted on the pavement of the motorway verges inside the tunnel (Fig. 5.1).

Fig. 5.1
figure 1_5

Flat seed trap with a removable funnel construction for installation at the road verge of motorway tunnels

Full size image

5.2.2.1 Seed Deposition at the Road Verge and Transport Distances

The seed deposition trapped at the road verges varies between 635 and 1,579 seeds per square metre and year, which is within the range of the seed rain in early successional stages, alpine turf, or acidic grassland (von der Lippe and Kowarik 2007a). Thus, within roadside habitats with sparse vegetation, the seed deposition of vehicles has the potential to add a similar proportion to the seed rain as the resident vegetation.

Within a transect of seed traps in the longest tunnel, seed deposition did not decline markedly with distance from the tunnel entrance (von der Lippe and Kowarik 2007a, Fig. 5.2), suggesting recurrent long-distance dispersal by traffic over the range of 550 m that was covered by the transect.

Fig. 5.2
figure 2_5

Annual seed deposition along 400-m seed-trap transects in both lanes of an urban motorway tunnel. Solid line: outbound lane; dotted line: inbound lane

Full size image

A comparison of the tunnel samples with the composition of the flora around the tunnel entrances revealed that long distance-dispersal over more than 250 m occurred significantly more frequent in non-native than in native species (von der Lippe and Kowarik 2007a). In individual cases, dispersal over several kilometres could be revealed, such as for the Australian annual herb Chenopodium pumilio, where the closest known populations are in a distance of more than 5 km from the tunnels.

There was a notable high proportion of rape and cereal seeds in the samples (24% of the total seeds), which can be attributed to transport losses (von der Lippe and Kowarik 2007b). The densities of each of these species in the seed samples were strongly related to one particular direction of traffic, probably tracing the major transport routes of the harvest.

5.2.2.2 Species Composition of the Tunnel Flora and Directed Dispersal

The species transported by vehicles represent 12.5% of the present flora of Berlin. Non-native species constituted 54.5% of all viable seeds in the samples and 50.0% of the species. Compared to the city zones adjacent to the sampling sites, non-native species were over-represented in the samples (von der Lippe et al. 2005). In addition, 39 species (19.1%) of the tunnel flora proved to be problematic invasive weeds in some parts of the world, including Acer negundo, Buddleja davidii, Lupinus polyphyllus, Solidago canadensis, and Robinia pseudoacacia (von der Lippe and Kowarik 2007a).

The species that were transported by traffic come from a broad variety of different habitat types, including both elements of urban-industrial vegetation and semi-natural communities (von der Lippe and Kowarik 2008a). Among the habitat types, most species of the tunnel flora can be assigned to ruderal ecotones and grassland habitats. Also, species of woodlands, weed communities, and urban vegetation are overrepresented in the samples.

The spectrum of species found in the tunnels had a higher similarity with the Berlin roadside flora than with the species composition of the vegetation around the tunnel entrances. The frequency of the species in the tunnel traps is significantly correlated to their frequency in the roadside vegetation (von der Lippe and Kowarik 2008b, Fig. 5.3). This points to a strong feedback between roadside habitats as donors and recipients of seeds.

Fig. 5.3
figure 3_5

Relationship between the frequency of species in the Berlin roadside flora in seven independent road sections (Langer 1994) and their mean frequency in the seed samples from 25 seed traps in five separate tunnel lanes of an urban motorway. Numbers within bars are sample sizes (species numbers within each group). The Spearman rank correlation is significant (rS = 0.547, p < 0.001)

Full size image

Overall, significantly more seeds were transported from the city towards the surrounding areas than vice versa, with a significantly higher proportion of seeds of non-native species in samples from carriageways leading out of the city. Indicator species analysis revealed that only few species can be confined to samples from lanes leading into the city, while mostly species of urban habitats were significantly related to samples from the outbound lanes (von der Lippe and Kowarik 2008a). The findings demonstrate that dispersal by traffic reflects different seed sources that are associated with different traffic directions and may thus exchange propagules along the urban-rural gradient.

5.2.2.3 Implications for Nature Conservation and Roadside Management

The observed dimension of seed deposition at roadsides suggests a strong impact of urban mobility on biodiversity patterns. Given its spatial effectiveness, human-mediated dispersal by vehicles may counteract urban fragmentation and the related habitat isolation. While this probably supports urban plant diversity in the short run, it also poses new risks of invasion and homogenization of urban floras.

Two outcomes of the study are of particular concern for nature conservation goals: first, the high share of non-native species deposited by vehicles and the large distances that can be bridged by this vector potentially support unwanted plant invasions along road corridors. Second, the directed transport of propagules in the direction out of the city may help to overcome dispersal limitation of isolated urban populations of invasive plant species and assist their migration to the countryside. The function of cities as centres of new introductions of plant species is thus combined with an effective vector that may lead to a quick “suburbanization” of these species. It is an open question, however, to which extent roadside populations of invasive species may function as foci for the invasion of habitats in the surrounding landscape. So far, effective invasions from roadsides to interior habitats were revealed for dry grasslands (Gelbard and Belnap 2003).

The management of roadsides can serve various purposes. Traditionally, erosion control and traffic safety were the primary goals but with an increasing road network, additional demands arose, like maintenance and development of roadside biodiversity and avoidance of biological invasions (Berger 2005).

The invasion of roadside verges by fast spreading non-native plants is not easy to control. Besides the difficulties of sustainable removing established populations of invasive plant species (Schepker and Kowarik 2002), the results of this study demonstrate the important role of recurrent introductions of a species to the same site with the help of traffic-mediated dispersal. Many approaches for the management of roadside invasions focus on local eradication of invasive species, often using chemical control (Berger 2005). Prior to this, however, an important implication of the findings about traffic as an important vector in plant invasions is to avoid invasion foci by roadside plantings.

If an invasion process is once in progress, efforts of eradication can be warrantable, if a species causes detrimental impacts and if success of the control measures can be expected. The results of this study suggest that management approaches on a local scale are unlikely to succeed in the control of invasive plants at roadsides. Even small isolated populations of invasive species at roadsides could act as invasion foci, which can rapidly negate control efforts. Therefore, roadside management of invasive plants demands for a perception of roads as a coherent dispersal network that affects distribution of propagules on a regional scale.

5.2.3 Environmental and Landscape Factors Shaping Species Composition on Urban Wasteland

Urbanization promotes the fragmentation of habitats (Alberti 2005; Robinson et al. 2005) and leads to a highly heterogeneous matrix, which is a distinctive characteristic of urban environments (Cadenasso et al. 2007). Thus far, only few studies addressed the question of how the spatial configuration of the urban matrix affects the floristic composition of urban habitat patches (e.g. Wania et al. 2006). As long-distance dispersal may counteract the effects of habitat fragmentation, the species-specific dispersal potential may strongly govern the colonization of fragmented habitat patches. In addition, environmental filters at the local scale are supposed to drive the composition of species assemblages of urban habitats, which are often different from natural ecosystems.

To better understand the mechanisms that shape urban biodiversity patterns, it is thus necessary to disentangle the relative importance of the spatial habitat configuration within the urban matrix, the local environmental conditions, and the dispersal abilities of individual species. A fuller understanding of the interplay between local factors and landscape parameters that affect urban biodiversity is also important for the management of urban habitats (Angold et al. 2006).

When urban wastelands remain unused for longer periods of time, initial successional stages are followed by herb- and shrub-dominated stages that finally develop into forests. Due to the frequency of disturbances in cities, early- and mid-successional stages are more common than late successional stages (Rebele 1994). Mainly in shrinking cities, however, larger woodlands can emerge on urban wastelands (Kowarik 2005). Up to now, soil, microclimatic, and land use predictors were used to explain the species assemblages of urban wastelands (Godefroid and Koedam 2007; Godefroid et al. 2007; Muratet et al. 2007). The question how these predictors govern successional mechanisms in urban habitats was however not considered explicitly.

We here aim at analyzing the relative importance of environmental and landscape predictors during the course of urban wasteland succession. Then, the predictability of the occurrences of species was examined with regard to their dispersal ability. To better distinguish successional effects from those of habitat heterogeneity due to strongly divergent starting conditions, one urban land-use type, i.e. abandoned railway areas, was studied. We analyzed species composition and measured directly microclimatic and soil parameters in stands that had been assigned to four successional stages, based on the dominance of annual, perennial herbs, shrubs, or trees (see Fig. 5.4 as an example). Landscape predictors (e.g. proportions of different habitats and sealed areas) within a 500-m buffer were analyzed using the Berlin Digital Biotope Mapping (SenStadt 2008). Data were analyzed by canonical correspondence analysis (CCA). To add the perspective of individual species dispersal ability, we compared the relative importance of landscape predictors to that of local environmental predictors. We then analyzed how strong landscape predictors improve the predictability of species occurrence (CCA) and how the change in predictability correlates to dispersal-related plant traits (regression tree). We hypothesized that the increase in predictability due to the inclusion of landscape predictors mainly occurs in species with a low dispersal capacity (see Westermann et al. 2011).

Fig. 5.4
figure 4_5

Wasteland succession on urban railway areas (Berlin, Südgelände). The woodland stage is prevailed by a clonal population of the native poplar (Populus tremula), but also non-indigenous species can play an important role, e.g. the North American black locust (Robinia pseudoacacia), here on the right side

Full size image

5.2.3.1 Relative Importance of Landscape and Environmental Variables

In a CCA including environmental and landscape variables, the most important predictors were PAR, C/N-ratio, and the proportion of ruderal and other habitat types (Fig. 5.5). PAR and temperature maxima exerted a significant and strong influence on species composition. This is likely due to differences in vegetation height and density during succession. The incoming radiation at the soil surface determines the surface temperature and thereby the air temperature near the surface (Stoutjesdijk and Barkman 1992).

Fig. 5.5
figure 5_5

Important environmental and landscape predictors driving plant succession in abandoned railway areas. Biplot of studied plots and predictors based on a CCA; eigenvalues of the first and second axes: 0.631 and 0.408. Significance of all canonical axes: 0.001

Full size image

The C/N ratio was the only significant soil predictor. In particular, plots in annual-dominated stages had high C/N ratios. The C/N ratios decreased during succession because the nitrogen content increased from annual- to tree-dominated stages. Correspondingly, in post-mining sites, the total nitrogen content increased in the course of succession (Frouz et al. 2008), thus showing a similar pattern. Hence, well-known mechanisms from non-urban successional series appear to govern also the succession in abandoned urban railway areas.

The proportion of ruderal and woodland habitats in the local vicinity of a plot exerted a significant and strong influence on the species composition. Muratet et al. (2007) argue that migration of species among large wastelands might have a significant impact on local species composition. Hence, the area covered with plants in the surroundings of a plot may act as propagule source and may thus affect the species composition of abandoned railway areas.

In a variation partitioning analysis, environmental predictors accounted for a higher portion of the variation in species data compared to landscape predictors. However, including landscape variables in a joint model together with environmental variables considerably increased the explained variation in the CCA (Westermann et al. 2011). These findings expand upon results from a study in Brussels, which found that density and function of built-up areas predominantly determine the urban plant assemblages (Godefroid and Koedam 2007). The present study also includes predictors that represent overgrown areas. In this case, the importance of built-up areas as predictor for the species composition is clearly lower but nevertheless significant. Built-up areas may function as dispersal barriers, preventing propagule exchange between fragmented habitats. Early successional stages on urban railway areas were often located in a close proximity to densely built-up neighbourhoods. In particular, propagules of short-lived species are able to reach these sites. This may be due to wind dispersal or due to human-mediated dispersal that favours small-seeded and lightweight species, as shown for traffic (von der Lippe and Kowarik 2008b).

5.2.3.2 Linking Landscape Variables with Species Traits

With a regression tree, we analyzed how the change in predictability due to the inclusion of landscape predictors in a CCA correlates to dispersal-related plant traits. The most important predictor in the regression tree model was the seed longevity index (SLI), followed by seed size, terminal velocity, and seed production. Species with a long-term persistent soil seed bank showed the largest increase in predictability due to the inclusion of landscape predictors in the CCA model. In contrast, species with higher wind-dispersal ability and a more transient soil seed bank had a significantly smaller improvement in predictability. Thus, the increase in predictability mainly occurs in species with low dispersal ability as hypothesized (Westermann et al. 2011).

The most important predictors, SLI and seed size, are more related to long-term persistence of a species on a given site than to dispersal, thus stressing the importance of a persistent soil seed bank in fragmented habitats (Bekker et al. 1998). With increasing isolation, the exchange of seeds between populations is hindered. In such cases, species with life-history characteristics that ensure a high persistence of the seed bank are promoted (Maurer et al. 2003). Such species traits foster stable populations once habitat patches have been colonized.

5.3 Enhancing Ecosystem Services in Urban Settings

Plants are known to mitigate negative features of urban environments, e.g. by reducing heat stress and pollution loads, and to offer multiple chances for urban residents to interact with natural components, with several feedback loops on human health and well-being (Tzoulas et al. 2007; Bowler et al. 2010). Contact with nature makes citizens also aware of the importance of biodiversity conservation and inspires human dwellers to interact with nature (Miller 2005). As the human population is markedly increasing, such ecosystem services become increasingly important. To strengthen them is thus a major challenge for urban planning. Hereby, a key question is to which extent ecosystem services rely simply on biomass or whether biodiversity matters as suggested for some psychological benefits in urban green spaces (Fuller et al. 2007).

Dust in the air is a prominent problem for urban health, and the function of vegetation to reduce this threat has been mainly analyzed for trees along roads or for distinct parks (e.g. Freer-Smith et al. 1997; Beckett et al. 2000; Endlicher et al. 2007). It is promising, however, to also consider the potential functions of the ground vegetation along urban roads, as this type of vegetation is directly exposed to both major pollution sources and the space where people move. Depending of the way of maintenance, vegetation along roads can also harbour an array of species (F. Weber).

While in strongly increasing cities green spaces are often limited and intensively used and managed, contrasting tendencies emerge in cities which are subject to shrinking (Kowarik and Körner 2005; Giseke 2007; Langner and Endlicher 2007). Here, new potentials for green spaces arise due to the reduction of built infrastructure, but often, the financial resources for designing new parks or later maintaining them are limited (Langner 2009). To date, ideas for developing aesthetically attractive green spaces in residential areas are of high practical relevance, in particular in areas where the density of housing estates has been reduced. With a field experiment situated directly in demolition sites, we aim at combining cultural ecosystem services with conservation interests by establishing attractive semi-natural meadows relying on regional provenances of native species (L. Fischer).

5.3.1 Regulating Services of Roadside Vegetation

Roads as key elements of the human society are omnipresent components of most landscapes throughout the world. Nowadays, the public awareness for the societal dependency on road traffic and for related environmental damages is growing and ecological impacts came into the scientific focus. The vast majority of studies on ecological impacts of roads have been developed in rural sites (Foreman et al. 2003).

Roadsides within urban landscapes differ widely in terms of shape, density, traffic, or adjoining usage compared to roads in rural landscapes (Van der Ree 2009). In particular, biotic and abiotic conditions on urban roadsides vary due to the location within the urban mosaic (Wittig 2002). Characteristic for habitats at road verges is a prevailing imprinting due to plant stressors like mechanical disturbances, de-icing agents, increased warming, dusts, heavy metals, and maintenance measures. Traffic also leads to the deposition of plant propagules at road verges (von der Lippe and Kowarik 2007a, b; see above). Generally, the growing conditions for plants next to roads stand for a well nutrient supply, at most slightly acidic to alkaline soils and rather droughty conditions (Langer 1994; Stottele 1995; Foreman et al. 2003). We use the term roadside vegetation in a broad sense relating to the entirety of plants growing on roadsides in urban landscapes. It includes all types of spontaneous and cultivated vegetation occurring in paving joints, at road verges, and tree planting sites (Wittig 2002).

In this chapter, we focus on the benefits people obtain from roadside vegetation, applying the concept of ecosystem services (MEA 2005). Roadside vegetation can give relief from environmental pollution in urban landscapes, first of all because it is situated very next to motor vehicle traffic. However, ecosystem services of roadside vegetation are an uncharted territory and nearly not considered in urban ecology research so far. Analogue to vegetation of other unsealed spaces, roadside vegetation mainly helps to sustain human health and quality of life due to services, which are due to the role of ecosystems in regulating climate and air quality. We here give a first overview on regulating services offered by roadside vegetation which are yet to be thoroughly valuated and made utilisable (Table 5.2).

In urban street spaces, people’s health is seriously affected by traffic-related emissions (Samet et al. 2000; WHO 2006; UNEP 2007). Urban structure plays an important role as buildings form vertical barriers hindering the dispersion of particulates, leading to a higher particulate pollution within residential sites by local traffic compared to areas with high traffic surrounded by open space (Capannesi et al. 1993). Currently, the political pressure to act increases as established emission limits have been exceeded seriously in urban areas (UNEP 2007). By the need to improve air quality, the capacity of plants for particulate deposition, and as a result the filtration of dust-laden air, recently gained more scientific attention (Jim and Chen 2008; Litschke and Kuttler 2008). Particulate immobilisation and air filtration is a promising regulating service of urban roadside vegetation, although the effectiveness is still not proved. Proximity to pollution sources could lift removal efficiency of vegetation (Jim and Chen 2008). Freer-Smith et al. (1997) observed that the number of particles counted on leaf surfaces decreased as distance from the motorway increased. In order to maximize the efficiency of filtration, vegetation is required as near as possible to the emission source (Litschke and Kuttler 2008). A planting concept, or the existing vegetation, should provide as great a plant surface as possible near to the emission source without significantly reducing air exchange. Conceivable solutions could include loose ground-level vegetation with adequate spacing between the plants to minimize the effects on airflow, combined with façade greenery (Litschke and Kuttler 2008).

Additionally to urban climate effects, the heat release from traffic increases air temperature in streets canyons (Swaid and Hoffman 1990; Hong et al. 2009). Micro climate regulation and passive air quality melioration have been reported also for small urban green sites with trees (Shashua-Bar and Hoffman 2000).

Furthermore, roadside vegetation is beneficial to city dwellers concerning noise pollution. Vegetation does attenuate noise depending on the density, height, length, and width of vegetation barriers, which influence noise reduction by diffusion, whereas leaf size and branching characteristics have resonant absorption properties (Aylor 1971; Cook and VanHaverbeke 1977; Fang and Ling 2003). However, the actual acoustic effects are small if the vegetation belt is rather narrow (Kragh 1980). Vegetation in cities in general and roadside vegetation in particular could be beneficial to people in terms of coping with noise: Szeremeta and Zannin (2009) found that visual features such as vegetation or fauna contrasting the “grey” urban surroundings and acoustic conditions being different from traffic noise proved to be important factors influencing the perception of traffic noise within an urban park.

Roadside vegetation also offers biological monitoring with plants as a low-cost and effective method to estimate levels of environmental pollution and their impact on biological receptors (Mulgrew and Williams 2000; Oliva and Fernández Espinosa 2007). Biomonitoring by plants can be used as a standardized method (Bargagli 1998; Moreno et al. 2003).

It is widely acknowledged that green spaces and natural elements generally benefit people (e.g. Bowler et al. 2010). However, space for nature development in urban landscapes is often scarce. As roadside vegetation is in the immediate vicinity of people’s everyday life, its potentials and services should be explored thoroughly and considered in urban planning decisions. Up to now, the filtration of airborne particles by plant surfaces was studied mostly for trees and some shrubby species (Beckett et al. 1998; Freer-Smith et al. 2004). First evidences suggest, however, that spontaneous vegetation may also contribute to the immobilisation of airborne particles (Jim and Chen 2008; Litschke and Kuttler 2008; Weber et al. unpubl. data). Further research should thus also consider ecosystem services of herbaceous roadside vegetation.

Table 5.2 Overview of regulating ecosystem services which can be related to roadside vegetation, and selected literature references
Full size table

5.3.2 Urban Meadows: Linking Conservation Goals with Cultural Ecosystem Services

Being widely established in the city, urban meadows originate and sustain due to a wide array of uses and objectives. As grassland covers about 5% of the city’s surface (SenStadt 2008), these areas are of high relevance for Berlin’s residents in terms of recreation, sports, and experience of nature. Also for plant species, urban meadows such as in old parks can be a “refuge” (Maurer et al. 2000; Peschel 2000). Some reasons are that gaining a good yield is unimportant for landowners here, and with the continuance of the sites such as in historic parks, they are often maintained over long periods of time in a traditional way (Wilhelm and Andres 1998).

The potential of urban grasslands becomes especially interesting for biodiversity conservation when considering the remarkable decline of extensive grasslands in agriculture (Wesche et al. 2009). Here, grassland extent and quality change due to intensification of agriculture (Reidsma et al. 2006) but also with the abandoning of unproductive sites (Henle et al. 2008). Additionally, urbanization leads to the fragmentation of many grassland areas in the cultural landscape (Ricketts and Imhoff 2003; Antrop 2004; Wittig et al. 2010). However, it is uncertain to which extent grasslands occur in urban settings today, and if they are of similar quality to agricultural – i.e. traditional – grasslands. In addition, aesthetically attractive grasslands are integral parts of designed and semi-natural green spaces, where they may contribute to human well-being as a part of aesthetic experience. Aesthetic experiences often lead to direct actions or generate feelings like that of identity as found for European agricultural landscapes (Gobster et al. 2007).

It is thus challenging to consider both conservation value and cultural significance of urban grasslands. Combining both functions could strengthen the potential to maintain or establish species-rich grasslands in urban settings. Taking Berlin as an example, we here first present an overview of the function of different urban land use types in harbouring legally protected grasslands and then illustrate an experimental approach in establishing species-rich grasslands in residential areas.

5.3.2.1 Protected Grasslands in Urban Settings

With analyzing the Berlin Biotope Mapping (Senstadt 2008), we performed a first area-wide GIS-based analysis of quantity and quality of urban grassland biotope types (Fischer et al. unpubl. data). We hereby compared legally protected and non-protected grasslands in agricultural areas to that of typical urban land use types (airports, historic parks, other urban areas). We determined that of all grassland cover in Berlin, 43% are legally protected grassland types. Only one-third of these grasslands are located in agricultural areas but the majority of more than 70% lies in other land use categories. Covering just 2% of Berlin’s surface, airports and historic parks contain one-third of all protected dry grasslands.

Largest grassland patches with a high connectivity on different scales were determined for airports and partially for agricultural areas. Grassland patches in urban green spaces outside of historic parks and airports are least connected with other grassland habitats. In airports, grasslands of special conservation interest are highly connected as the areas themselves are large and other biotope types are less common due to the land use type itself. With that, airports represent core areas for grassland habitats in the inner city. Also, the smaller grassland patches in historic parks may contribute to the inner city habitat network, as patches are highly connected with other non-sealed surroundings, whereas their small patch size might originate from the landscape garden design.

These results demonstrate the prominent role of urban land use types to harbour grassland habitats of special conservation interest outside of agricultural areas. In parts, urban grassland may compensate for the high decline of traditional grasslands in the cultural landscape. As our analyses also showed that most of the protected grasslands are located outside of conservation areas, the results suggest that private and public landowners need to be involved in strategies to conserve urban grassland. Especially within the next years, when both inner city airports in Berlin are being changed to green spaces and housing areas, aims of biodiversity conservation and options for recreation of urban dwellers need to be balanced. This offers an excellent opportunity to “conserve biodiversity where people live and work” (Miller and Hobbs 2002).

5.3.2.2 Establishing Species-Rich Grasslands in Residential Areas

In dense residential areas, usually more intensely managed grassland types such as lawns prevail. In shrinking urban regions where residential houses and associated infrastructure are demolished, the evolving green space offers the chance to establish meadows oriented on extensive grassland types. Here, establishing new meadows might match objectives of nature conservation, planning, and aesthetics. The initiation of grasslands on demolition sites requires careful consideration of various background factors, though. For example, an important issue for planners and landowners is the question about financial advantages of meadows compared to traditional landscape design with frequently cut lawns. Not only the initiation but also the later maintenance needs to be a cheap option for large-scale greenings (Giseke 2007; Langner 2009). It is equally important that the created meadows have an appealing character to residents whilst being appropriate for their daily use. These social and psychological benefits associated with biodiversity (Fuller et al. 2007) are especially important for residents who rarely have the possibility to experience nature (Miller 2005; Samways 2007).

For nature conservation, we can determine two distinct advantages of creating extensive meadows in shrinking residential areas. First, species that are on the decline in the cultural landscape may be at least partially supported in these areas – considering that the importance of urban green spaces for biodiversity generally grows (Goddard et al. 2010). Second, a positive interaction is to be expected for residents interested in nature and its conservation (Miller 2005; Dunn et al. 2006).

The establishment of extensive grasslands on urban demolition sites has to meet several requirements: (a) technically, as dispersal limitations of native grassland species need to be overcome, and target species have to be established over time. Limitations linked to abiotic, e.g. soil characteristics, need to be incorporated. (b) The new meadow type has to withstand the typical pressure of use in residential areas. (c) The meadows have to be perceived as an attractive alternative to common landscape design but also to non-maintained wasteland.

Our study area Marzahn-Hellersdorf in the north-eastern part of Berlin is a large-scale residential housing area of the 1980s, when about 100,000 apartments were built. Especially young families moved in back then, resulting now in an oversupply of kindergartens and schools due to severe demographic changes – e.g. the proportion of elderly people is predicted to increase another 190% within the next 20 years (Bezirksamt Marzahn-Hellersdorf von Berlin, n.d.). The area is a part of the program ‘Stadtumbau Ost’ (Senstadt 2010), meaning that apartments were torn down within the last years (Bezirksamt Marzahn-Hellersdorf von Berlin 2007). Nevertheless, shrinkage of the quarter is still apparent with various demolition and wasteland areas, and evolving free spaces of no future or interim use.

In these surroundings, we test different treatment types which involve strategies of reintroduction of species and assisted migration (Ricciardi and Simberloff 2009; Vitt et al. 2010): (a) seed mixtures of regional provenances, (b) the same seed mixtures combined with a mycorrhiza inoculation, and (c) seed transfer (Heudrusch ®) from species-rich grassland communities of nearby conservation areas. For the first treatment, we composed seed mixtures oriented on comparable grasslands in the cultural landscape outside of Berlin. The mycorrhiza inoculation should give a start up aid for the vegetation, especially in dry times, and increase soil stability (Rillig and Mummey 2006; Chaudhary et al. 2009). With seed transfer, we apply a traditional method which transfers not only target species but also species of other functional groups such as mosses. These methods are known as robust treatments for extreme sites (Kirmer and Tischew 2006; Tischew et al. 2010; Kiehl 2010; Kiehl et al. 2010), but have not been tested in urban settings yet.

We randomly applied the three treatment types on tilled soil in 4 × 4 m² plots in fall 2008 on 11 sites in Berlin-Hellersdorf. Each site also has a control plot. Plots were neither marked nor fenced, as we wanted residents to use these areas just as normal. In the following 2 years, we studied the success of our target species and the spontaneous vegetation (mapping of all species, their cover, and reproductive potential), and analyzed environmental parameters (e.g. soil stone content, potassium, pH, frequency of dogs and people).

First results show a significant increase in species richness in the treated meadow plots already after the first 2 years, e.g. species number increased from 28 species on average in control plots to over 40 species in both seeding treatments. The highest species numbers were found in plots where regional seeds were sown and mycorrhiza was inoculated. Many target species are already abundant. In those variants where seeds were applied, of the 26 sown species, in average, we found 11 (seed mixture only) and 14 (seeds combined with mycorrhiza) target species. Nine of the target species were among the most frequent 20 species found in the seed mixture treatments. This shows that spontaneous vegetation and target species are nearly equally present. Seed transfer had a lower rate of target species (on average 4 of 16 target species). In all treatments, target species richness was increasing in the second year of establishment. Blooming aspects were changing throughout the course of the year (Fig. 5.6).

Fig. 5.6
figure 6_5

New types of urban grasslands are initiated in large-scale housing areas in Berlin. Blooming aspect of marguerite Leucanthemum ircutianum in summer of the second year of establishment

Full size image

Statistical modelling which incorporates environmental variables showed that the abundance of target species is strongly depending on abiotic parameters such as the stone content in the soil. In contrast, human-mediated disturbances by recreational activities or walking dogs had less influence on the occurrence of target species in the non-fenced plots. This latter surprising result suggests promising chances of establishing species-rich grassland vegetation in residential areas.

In the long run, we anticipate the emergence of novel grassland types in residential areas, which are characterized by a mixture of typical grassland species (target species) and ruderal species that colonize the plots from adjacent seed sources or the propagule bank. Further monitoring should test the successful establishment of target species over longer periods of time.

A future part of the study will aim at the perception and preferences of the experimentally established grasslands by the neighbourhood. It will be of high relevance for application which meadow type they prefer, and if simple methods such as seed mixtures on tilled soil are sufficient to create meadows which appeal to residents. If this could be proved statistically, our project could be a new impulse to combine cultural ecosystem services with nature conservation interests.