Introduction

Urban flora comprises plant species established through deliberate cultivation and natural colonisation, the latter including both species originally native to the area and those that were introduced by human activity but have become naturalised. This last category is often further subdivided into those that were naturalised before 1500AD (archaeophytes) and those after 1500AD (neophytes).

Between the theoretical extremes of totally natural and totally artificial plant communities there is a spectrum, any point along which is determined by the relative proportions of human and natural processes acting. The term ‘semi-natural’ recognises the inevitability of human influences on those plant communities near to the theoretical ‘natural’ end of the spectrum (Peterken 1993). Urbanisation can not only encapsulate patches of originally rural semi-natural plant communities but also provide opportunities for new plant communities to arise on derelict and neglected land. Kendle and Forbes (1997) have used the term ‘spontaneous vegetation’ to describe such plant communities which are effectively a sub-set of semi-natural vegetation since natural processes of colonisation still play a very significant part in their establishment. This paper is concerned with the plant species present in urban semi-natural vegetation.

Urban areas are traditionally considered to support a higher diversity of plant species than the surrounding countryside (Gilbert 1989), the total number of species increasing as the city population size increases (Sukopp 2004). Two primary reasons suggested for this are the relatively high number of alien (non-native) species available and the heterogeneity of the urban landscape (Kowarik 1995). Studies of temporal changes to urban flora suggest that increases in species richness reflect an increase in neophytes which masks a decline in native species. Both Tait et al. (2005), studying Adelaide, Australia, and Chocholousková and Pysek (2003), studying Plzeň in the Czech Republic, found this trend over a period of 100 years or more. Godefroid (2001) studied the flora of Brussels from 1940 to the beginning of the 21st Century and found no change in overall species numbers but an increase in the proportion of alien species. DeCandido (2004), looking at Pelham Park, New York, over a similar time period, found an increase in vascular plant species overall but a decline in the proportion of natives and an increase in the proportion of aliens. Pysek (1998) in a study of 54 central European cities, found that with increasing city size the proportion of neophytes also increases.

Roy et al. (1999), however, found no evidence that urban landcover increases overall plant species richness. Their data did suggest that complete urbanisation doubles the proportion of alien species and that the availability of urban habitats, together with high levels of disturbance, are important factors in maintaining the urban flora. Kühn and Klotz (2006), studying the effects of urbanisation on plant communities, found evidence for homogenisation in natives and archaeophytes but a tendency towards the opposite in neophytes.

Studies of spatial variations in semi-natural vegetation along a rural–urban gradient, reveal a more complex situation. Moffatt et al. (2004) found that urban riparian forest had significantly lower understorey native and overall species diversity and a higher proportion of exotics, than suburban, rural and high quality forest areas. Burton et al. (2005) also found that species diversity in the regeneration layer of woody riparian vegetation decreased linearly towards the urban centre as the proportion of non-native species increased. Porter et al. (2001) found no clear pattern in the species richness of woody vegetation along a forest–urban gradient and concluded that landscaping preference largely determined diversity patterns.

The results from Porter et al. (2001) point towards the significance of Kowarik’s (1995) second primary reason for higher urban species richness, the heterogeneity of the urban landscape. Wania et al. (2006) also concluded that habitat variability might be decisive for the species richness of cities.

Ecologists characterise habitat patches by their species composition, structural diversity, ecological dynamics and value to biodiversity. However, planners are concerned with resolving competing demands for land-use and developing spatial strategies for future urban development that respond to the broader economic, social and environmental needs of the human community (DCLG 2007). Part of this process involves surveying, categorising and evaluating urban green space, the results of which have a direct bearing on the future development and management of such spaces. It is therefore important to understand how the semi-natural plant communities associated with different planning designations of urban green space might contribute to urban plant biodiversity.

This paper utilises existing datasets to investigate the relationship between plant species (categorised as natives, archaeophytes, neophytes, casuals and those designated because of concern over their conservation status) existing in semi-natural urban plant communities and urban green space designated by the municipal government for Leeds, a large city in the north of England. Designations are principally according to land-use but those based on value to nature conservation are also included.

The concept of the city-region, comprising a central urban core together with its commuter hinterland encompassing smaller satellite settlements and intervening rural areas, has grown in significance over the past half-century (ODPM 2006). City planning departments may therefore be responsible not only for large urban areas but also smaller settlements and rural areas, together with their respective designated green space, which will vary in terms of the ecological processes impacting upon it. Consequently, this study examines data from both the central urban area of Leeds and the satellite settlements overseen by the same municipal authority.

Urban green space in Leeds

Leeds is situated approximately 320 km north of London in the county of West Yorkshire, on the eastern fringes of the Pennine hills (53°48′ N, 1°32′ W) and has a population of about 720,000 (Office for National Statistics 2006). It has a long history of manufacturing although, with the decline of the UK textile and heavy manufacturing industries, its current economic success has been built upon recent rapid growth in the financial and legal sectors. Leeds Metropolitan Borough Council (MBC), the municipal authority, has a boundary encompassing an area of 563 km2, of which about 175 km2 was categorised as built-up in 1994 (Leeds City Council 1994), the remainder comprising mainly farmland but with significant areas of woodland, moorland and river valley, some of which has been despoiled by industry, now largely gone. English Nature (1998) has divided England into Natural Areas (NA) based on geology, geomorphology, soil and semi-natural vegetation types. The bulk of Leeds MBC lies within the Coal Measures NA although the northern edge overlaps the Southern Pennines and Pennine Dales Fringe NAs. The soils of these areas tend to be acid to neutral. To the east, Leeds falls within the Southern Magnesian Limestone NA with its more basic soil and corresponding differences in semi-natural plant communities (see Fig. 1).

Fig. 1
figure 1

Leeds Metropolitan Borough showing distribution of built-up areas and sampled tetrads; boundaries of relevant Vice-Counties and Natural Areas are also given

Full size image

Public and semi-public urban green space in Leeds is designated in the current Unitary Development Plan (Leeds City Council 2001) according to both descriptions of land-use and recognition of patches with significant nature conservation or landscape value. The principal categories of the former comprise non-overlapping but often contiguous areas and are summarised in Table 1. Designations for nature conservation value are summarised in Table 2. Some patches have been designated both as urban green space and for their nature conservation values.

Table 1 Categories of urban green space defined in the Leeds Unitary Development Plan (adopted from Leeds City Council 2001)
Full size table
Table 2 Categories of site designated for conservation value (ecological or geological) and defined in the Leeds Unitary Development Plan (adapted from Leeds City Council 2001)
Full size table

Methods

Records of plant species’ presence were obtained from a comprehensive survey of semi-natural plant communities in the Watsonian Vice-County of Mid-west Yorkshire (VC 64). Vice-Counties are sub-divisions of counties, developed originally in 1859 and still used for recording botany (Dandy 1969). All of Leeds north east of the River Aire occurs within VC 64. The survey, which covered all of VC 64 (3,544 km2), took place between 1984 and 2004 (Abbott 2005) and involved recording the presence of all non-cultivated plant species growing in each tetrad (2 km × 2 km sampling unit) within the Vice-County.

For this study data were obtained from all complete tetrads that occurred within the VC 64 sector of Leeds Metropolitan Borough and which had a minimum of 10% built-up cover, as shown on the Leeds Landscape Assessment (Leeds City Council 1994). These comprised 27 from a contiguous area of urban and peri-urban landscape in the city of Leeds (central) and 20 from surrounding towns and villages (satellite)—see Fig. 1.

Data from the Vice-County Census Catalogue of the Botanical Society of the British Isles (Stace et al. 2003) were used to categorise species into native, archaeophyte, neophyte and casual. This last category comprised mainly garden escapes and hybrids. Species designated because of concern over their conservation status were identified from the latest available spreadsheet of conservation designations, produced by the Joint Nature Conservation Committee, the statutory advisor to Government on UK and international nature conservation (JNCC 2007). The four levels of designation applicable to the species encountered were vulnerable and near threatened [both red listings based on 2001 IUCN (World Conservation Union) guidelines]; nationally rare (in 15 or fewer hectads in Great Britain) and nationally scarce (in 16–100 hectads in Great Britain). Numbers of species in each conservation category ranged between 0 and 2 per tetrad, except for one case each of 3 and 4. Consequently numbers for the four categories were summed to give a total number of conservation-designated species per tetrad.

The type and extent of urban green space within a tetrad was determined from GIS vector datasets, provided by Leeds MBC. Protected playing pitches and allotments (Table 1) each have a distinct physical structure, the former an open expanse of mown grass with or without some woody vegetation around the periphery, the latter normally intensively cultivated for vegetables and/or flowers. Consequently these two green space types were treated separately for analysis. In contrast, the other three categories, ‘green space’, ‘proposed green space’ and ‘other protected open land’ are not distinguishable from each other in terms of their physical structure (which can include a wide variety of habitat types), merely their status in the development planning process. Consequently, their areas were summed for each tetrad to provide a ‘general green space’ category.

Data were analysed to investigate:

  • If there were differences in the areas of built-up and total designated green space between central and satellite tetrads.

  • If there were differences in the numbers of native, neophyte, archaeophyte, casual and conservation-designated species between central and satellite tetrads.

  • If there were correlations between the numbers of native, neophyte, archaeophyte, casual and conservation-designated species and the areas of each of: general green space; allotments; pitches and the three nature conservation designations (Site of Special Scientific Interest, SSSI; Site of Ecological or Geological Interest, SEGI; Leeds Nature Area, LNA), both individually and collectively.

The majority of datasets in all categories (central, satellite and total) differed significantly from normal distributions, tested for by the Shapiro–Wilk W-test (Shapiro and Wilk 1965) and so non-parametric statistical tests were used for analysis, the Mann–Whitney U-test for comparing central tetrads with satellite ones and Spearman’s rank correlation coefficient when testing for correlations between species’ categories and between numbers within a species category and area of designated green space.

Results

Differences between central and peripheral tetrads

Areas for both built-up land and total designated green space per tetrad were significantly greater in the central tetrads than the satellite ones (Table 3). It might be expected that these differences and the closer proximity of rural landscape to the satellite tetrads compared with the central ones might affect their species composition. However, comparisons of median numbers between central and satellite tetrads showed no significant differences for any of the five categories of plant species, although, for neophytes there were more in the central tetrads and the difference was close to being statistically significant (P = 0.0541) at the 5% level (Table 4).

Table 3 Comparisons between central and satellite tetrads for areas of built-up (>10%) and total designated green space per tetrad in Leeds Metropolitan Borough (Vice-County 64 sector)
Full size table
Table 4 Comparison of median values for number of each category of plant species per tetrad between central and satellite tetrads in Leeds Metropolitan Borough (Vice-County 64 sector)
Full size table

The distribution of total species showed a cluster of tetrads with relatively high numbers (300 or more) towards the north west edge of the city itself. Five of the eight satellite tetrads in the Southern Magnesian Limestone Natural Area also contained 300 or more species in total (Fig. 2a). Inspection of both native and neophyte distributions (Fig. 2b, c respectively) show similar patterns, while the distribution pattern for archaeophytes (Fig. 2d) shows less obvious clustering within the city. The two tetrads each with > 40 archaeophytes include most of two adjacent satellite settlements in the north east corner of the Borough, within the Southern Magnesian Limestone NA. Casual species occurred in considerably smaller numbers than the other categories but the tetrad with the highest number also occurred in the north west of the city.

Fig. 2
figure 2

Numbers of each plant species’ category (a) total species, (b) natives, (c) neophytes, (d) archaeophytes in sampled tetrads from Vice-County 64 sector of Leeds Metropolitan Borough. Shading indicates built-up area. Numbers indicate Natural Areas (1 Coal Measures, 2 Pennine Dales fringe, 3 Southern Magnesian Limestone, 4 Southern Pennines)

Full size image

The majority of the small number of species designated because of conservation concern were native (Table 5).

Table 5 Numbers of conservation-designated plant species in central and satellite tetrads in Leeds Metropolitan Borough (Vice-County 64 sector), displayed according to species category and conservation designation
Full size table

Patterns of distribution between the four categories of species were similar in central and peripheral tetrads, the only notable difference being six conservation-designated neophyte species in central tetrads compared with three in peripheral ones. This overall similarity in the distribution pattern of the four categories was also evident from the close and significant correlations between each of the pairs of species’ categories (Table 6).

Table 6 Correlation analysis of the numbers of plant species per tetrad in each of the four species categories (native, archaeophyte, neophyte and casual) against each other, using Spearman’s rank correlation coefficient (r s)
Full size table

Correlations between tetrad species composition and green space category

The three tiers of nature conservation designations were not evenly distributed across the study area. Of the top tier SSSI sites, there are only seven, widely scattered and occupying a total area of 153.29 ha. One of these is a reservoir (108.61 ha), designated for its birds, and another a small quarry, designated for its geological features. There are only ten 2nd tier SEGIs occupying 130.70 ha, although the largest is subdivided into four component patches. In contrast, LNAs, which are principally designated to provide urban communities with easily accessible, local patches of semi-natural landscape, irrespective of the particular species or ecological communities present, are represented by 50 sites distributed relatively evenly across the study area and with a total area of 532.65 ha. Initially, the three tiers of nature conservation-designated sites were combined. ‘Total green space’ comprises all general green space, allotments, playing pitches and non-overlapping nature conservation-designated sites. To investigate possible correlations between numbers of species from the different categories and the categories of green space described above, Spearman’s rank correlation coefficients (r s) were determined. The results are given in Table 7.

Table 7 Correlation analysis of the numbers of plant species per species category against areas of built-up and green space designations (ha) within tetrads in Leeds Metropolitan Borough (Vice-County 64 sector), using Spearman’s rank correlation coefficient (r s)
Full size table

There was a significant positive correlation between total numbers of species and both total areas of green space and the sum of areas with nature conservation designations. Numbers of natives and archaeophytes correlated positively only with areas of nature conservation designations. Both neophytes and casuals correlated positively with areas of built-up landscape, nature conservation designations and total green space. The former also correlated positively with general green space and playing pitches.

Significant positive correlations were found between native species richness and the areas of SSSIs and SEGIs. Numbers of both neophytes and casuals showed significant positive correlations with areas of SEGIs and LNAs. In both cases the strength of the correlation was greatest with the area of LNAs, the 3rd tier sites. No correlations were found between numbers of archaeophytes and areas of any tier of nature conservation-designation and neither was evidence found of any correlation between the small numbers of species designated because of concern over their conservation status and areas of any category of green space.

Discussion

This study, utilising existing datasets based on tetrad sampling units, has examined the relationship between categories of plant species growing in semi-natural communities and green space designated by a city municipal authority, according to land-use and nature conservation value. The authority’s boundary is located as much as 14.5 km outside the central urban area, encompassing satellite settlements and tracts of rural landscape. It might be expected that satellite settlements, with their smaller individual built-up areas surrounded by a rural matrix, would experience a proportionally greater impact by rural processes than a large central urban area, and hence be subjected to the equivalent of an edge effect (Dawson 1994), thereby developing different plant communities. Proportionally, the greatest difference between central and satellite tetrads was in numbers of neophytes, with a greater median value in central tetrads, although the difference was not statistically significant. Nonetheless, this tendency is in keeping with the findings of Tait et al. (2005), Chocholousková and Pysek (2003), Godefroid (2001), DeCandido (2004) and Pysek (1998).

Relatively large numbers of total species, natives and neophytes were found towards the north west edge of the central urban area. This corresponds to the findings of McKinney (2006), who recognised the importance of suburban and urban fringe habitats in supporting an overall species richness often greater than the surrounding countryside and urban core, and Zerbe et al. (2003) who found the highest number of species per square kilometre in the transition zone between city centre and the outskirts, an urban area often characterised by open spaces like large parks, urban forests and larger patches of derelict land. Such features contribute to urban landscape heterogeneity, potentially a key factor in the diversity of plant species found (Wania et al. 2006).

The significantly positive correlations between numbers of all four categories of plant species and total areas of nature conservation-designated sites emphasise the importance in urban landscapes of semi-natural areas in supporting plant species diversity. Criteria for the designation of all such categories of site are based on those for SSSI (Nature Conservancy Council 1989) and do not inevitably include species richness, since some sites may be designated for a particular rare species or as a good example of a particular habitat type. All LNAs (3rd tier sites) contain semi-natural vegetation but some are designated primarily for their amenity and educational value and may not have been the subject of a detailed survey. Nonetheless, the fact that LNAs contribute most in terms of area and number within the three categories of nature conservation-designated sites, suggests they may play a significant role in supporting plant species richness. However, discriminating between the three tiers of nature conservation designation emphasises the importance of the upper tier sites for native species richness and the lower tier sites for neophyte and casual species richness.

The results of this study are less clear cut in identifying the importance of other categories of designated green space for plant species richness. Positive correlations between numbers of neophytes per tetrad and areas of both pitches and ‘general green space’ suggest these categories may be important in contributing to urban plant species richness. However, while all pitches have a degree of consistency in their habitat structure, comprising an area of open intensively managed grassland (but with variations in the type of boundary and surrounding land) the category ‘general green space’ (excluding nature conservation-designated areas), in contrast, can cover a much wider variety of habitat structures (e.g. open grassland, clumps of trees, patches of woodland, green corridors along roads and various wetland habitats).

Moreover, the different categories of green space considered above do not account for all ecologically important space in the built-up area. The Leeds MBC Unitary Development Plan identifies significant areas of green corridor within the city (3,452 ha) which include some designated sites from all categories considered above and significant areas of non-designated green space, most of which is in private ownership and not accessible to the public (e.g. areas of farmland and private golf courses). Private gardens also make a very significant contribution to the overall urban green infrastructure, 21.8–26.8% of the urban area in a survey of five major UK cities (Loram et al. 2007) compared with 29.6% of the built-up area identified as designated public green space in the current study. The ecological contribution of public green space needs to be seen within this context.

The grain of the designated green space varied quite markedly and, in the case of areas designated for nature conservation, was always below 25% of a tetrad, ranging from 0.24 to 85.25 ha per tetrad. This level of mismatch in grain size between the type of green space and the sampling unit might be expected to increase the chances of correlations between species numbers and the area of a particular type of green space being masked through the effects of factors elsewhere in the tetrads. The fact that such correlations have been detected indicates that either they are real, and might even prove stronger were a smaller area of sampling unit to be used, or that factors elsewhere in the tetrads follow a similar pattern in influencing species’ numbers. Examples of the latter are most likely to be categories of green space not considered in this analysis, particularly non-designated and in private ownership.

This study provides evidence for the importance of designated green space to urban plant species richness but it is limited in the extent to which it considers the type and ecological value of such space. Nonetheless, the fact that a municipal authority designates any green space, irrespective of its ecological quality, is recognition of its value to the urban community, a recognition that has been growing in recent years (CABE Space 2004). Initially this focused largely on formal parks, recreational areas and sports pitches and their health and amenity value to people. However, recently the ecological value of such space has become more of an issue as the value of contact with nature has become a particular focus of interest. Evidence has been accumulating of benefits, not only for physical health (Bird 2004), but also for mental well-being (Rohde and Kendle 1994; Korpela and Hartig 1996; Kowarik and Körner 2005; Bird 2007; Korpela and Ylen 2007; Tzoulas et al. 2007). Providing accessible, semi-natural green space in urban areas is therefore seen as increasingly important and is reflected in the publication of guidelines for accessibility (Handley et al. 2003). There is also evidence of public preference for more varied vegetation in urban parks, play areas and green spaces (Dunnett et al. 2003), which could be interpreted as incipient support for more semi-natural vegetation and the greater structural variety within it. Incorporating more semi-natural vegetation in to public urban green space requires appropriate management, which may well necessitate significant changes from what was done previously. In the UK specific guidance on this has recently become more available, dealing not only with the plant community establishment and management but also political, social and economic aspects (TCPA 2004; CABE Space 2006).

In Leeds, the designation of LNAs, introduced in the city’s nature conservation strategy (Leeds City Council Countryside Service 1991) for public access as much as for biodiversity conservation, anticipated the importance of accessible semi-natural green space. The results of this study suggest that such designations may also contribute to the overall plant species richness of the urban area but, again, principally through increasing numbers of neophytes and casuals.

Conclusion

It is well documented that in the urban areas species richness is often greater than in the surrounding rural areas and that higher proportions of neophytes and casuals (including non-natives) contribute significantly towards this. There is evidence that urban landscape heterogeneity plays an important role in this but less is known of how specific types of urban green space contribute to semi-natural plant species composition. This study emphasises the importance of more semi-natural patches for native species but also how neophytes and casuals may be exploiting a wide range of urban green spaces.

Biodiversity is seen as an important indicator of sustainable development (DEFRA 2007) and so these findings have implications for planners with respect to both the urban green space designation process and subsequent management of the green space so designated. Pressures for development threaten many urban green spaces, particularly those that have the equivalent of a 3rd tier designation or none at all. This study suggests that such sites may be of more value for neophytes and casuals, typically effective colonisers of new habitats. Loss of such sites, on an individual basis, may not have serious implications for urban plant biodiversity, provided that similar new sites become available for colonisation at an appropriate rate. Further studies incorporating field surveys of designated sites, ideally as part of a monitoring programme, would enable higher resolution data to be obtained and, over time, provide information on the longer term effects of site designation on plant community diversity.