A comprehensive dataset on cultivated and spontaneously growing vascular plants in urban gardens

This article summarizes the data of a survey of vascular plants in 85 urban gardens of the city of Zurich, Switzerland. Data was acquired by two sampling methods: (i) a floristic inventory of entire garden lots based on repeated garden visits, including all vegetation periods; and (ii) vegetation relevés on two plots of standardized size (10 m2) per garden during the summer. We identified a total of 1081 taxa and report the origin status, i.e., whether a taxon is considered native or alien to Switzerland. Furthermore, the origin of a plant or garden population was estimated for each taxon and garden: each taxon in each garden was classified as being either cultivated or spontaneously growing. For each garden, the number of all native, cultivated, and spontaneously growing plant species is given, along with additional information, including garden area, garden type and the landscape-scale proportion of impermeable surface within a 500-m radius. The dataset is related to the research note entitled “Research Note: Self-reported habitat heterogeneity predicts plant species richness in urban gardens” [1]. It is also linked to a comprehensive dataset on biotic and abiotic soil data and as well as to a dataset on soil-surface dwelling and flying arthropods [2–6].


Data
This article presents data of a survey of vascular plants in 85 urban gardens in the city of Zurich, Switzerland. Two garden types were investigated: allotment (N ¼ 42) and domestic gardens (N ¼ 43). In each garden, we applied two sampling methods: a floristic inventory of entire garden lots (mean area ±SD: 312 ± 155 m 2 ) and sampling on plots of a standardized size (2 Â 10 m 2 ). The two plots were centred within the two main land-use types found in each garden: lawn, meadow, vegetable bed, flower bed or berry patch. We give the origin status for each of the 1081 taxa found, i.e., whether a taxon is considered native or alien to Switzerland (Appendix). Similarly, for each taxon and garden, we estimate the origin of each plant or garden population by classifying each taxon as either cultivated or spontaneously growing (Appendix). Species richness (i.e. number of taxa) of all (S total ), native (S native ), cultivated (S cultivated ) and spontaneous (S spontaneous ) plants was computed for each garden and overall (Table 1). In addition, species richness levels of gardens and land-use types were visualized, and results of the two sampling methods were compared (Figs. 1 and 2). For each garden, additional environmental data is given, such as garden type, area, and urbanization intensity, which was calculated as the landscape proportion of impervious (i.e. built and paved) surface within a 500-m radius ( Table 1). The data is part of an inter-and transdisciplinary investigation of biodiversity, soil quality, ecosystem Specifications

Floristic inventories and vegetation relev es
Data format Raw and aggregated Experimental factors A stratified sampling design with two types of gardens, and two crossed (independent) gradients: an urban intensity and a garden management intensity/habitat spatial heterogeneity gradient.

Experimental features
Gardens were chosen following a stratified sampling design, based on the urban habitat mapping key of the city of Zurich. The (independent) strata included i) garden type (domestic vs. allotment); ii) a garden spatial heterogeneity/management intensity gradient, ranging from extensively managed gardens with a high vertical vegetation structure and a high proportion of native plant species, to intensively managed gardens with a low vertical vegetation structure and a high proportion of alien plant species; and iii) an urban intensity gradient, which ranged from densely to less densely built-up areas of the city. Value of the Data The data is comprehensive as it describes all vascular plants growing on entirely sampled garden lots with a high taxonomic resolution, and plants growing on standardized sampling plots. The data can contribute to comparative studies of community assembly rules of spontaneously versus human assembled urban plant communities. The data can contribute to comparative studies of garden floras to understand mechanisms of plant introductions and invasions. The data can be used to investigate the effects of garden plants on diversity patterns of species of other trophic levels (e.g. herbivors, pollinators), for which data exists from the same gardens The data can be used to investigate above-below ground interactions, as biotic and abiotic soil data exists in the same gardens.
services and social value of urban gardens in Switzerland (www.bettergardens.ch). The data can be linked to biotic and abiotic soil data [2e4], to data of soil surface dwelling and flying arthropods, which were sampled in the same gardens and during the same period, and to arthropod and bird predation data [5,6]. The raw data are available from Mendeley Data https://doi.org/10.17632/ 452pj39jm2.2 [7].

Data source
The data was acquired in the city of Zurich, Switzerland (47 22 0 N, 8 33 0 E). Zurich is located in the temperate climate zone of Europe, with a mean annual temperature of 9.3 C (1981e2010) and mean annual precipitation of 1134 mm [8]. It harbours a population of 0.4 million in an area of approximately 92 km 2 , placing it in the globally most common city class [9].

Garden selection
We collected floristic data in 85 urban gardens (43 domestic and 42 allotment gardens). We defined a domestic garden to be a garden directly adjacent to a single-occupancy or terraced house. Our definition of allotment gardens encompasses spatially clustered garden lots on public land, managed by associations and leased to leisure gardeners as lots of 100m 2 e200m 2 [10].
Gardens were chosen following a stratified sampling design, based on visual criteria defined by the urban habitat mapping key to the city of Zurich [11]. Potentially suitable gardens were identified based on the habitat map of the city of Zurich, aerial images and during field visits. We approached the garden owners initially by letter and thereafter by phone to arrange a visit. If no phone number was available, owners were approached personally. The (independent) strata included i) garden type (domestic vs. allotment), ii) a garden spatial heterogeneity/management intensity gradient, ranging from extensively managed gardens with a high vertical vegetation structure and a high proportion of native plant species, to intensively managed gardens with a low vertical vegetation structure and a high proportion of alien plant species [11], and iii) an urban intensity gradient, which ranged from densely to less densely built-up areas of the city. The urban intensity gradient was quantified as the proportional area of impervious (i.e. built and paved) surface within a 500-m radius around each garden lot (Table 1).
Variance in garden area was kept small and no novel gardens were included. To assure statistical independence among observations, no adjacent garden lots were sampled, and gardens were distributed across the city to include all urban districts. Additionally, with two exceptions, only one garden lot was sampled per allotment garden area. The average pairwise distance between gardens was 4.5 km (SD ± 2.2 km, min.-max. 0.1e11 km).

Floristic data of garden lots
A complete floristic inventory of each garden lot was made during repeated garden visits in 2015, based on the standard determination literature of the Swiss, resp. European (garden) flora [12e17]. Potted plants were included in the inventory. Abundance of each taxon was estimated semiquantitatively on a six-point scale: 1 (1e10 individuals), 2 (11e25 ind.), 3 (26e50 ind.), 4 (51e100 ind.), 5 (101e250 ind.) and 6 (>250 ind.). To account for the different vegetation periods, gardens were visited three times in March/April, May, and July/August. To standardize the sampling effort among gardens, the duration of each visit was restricted to about 1.5 h. Note that the early spring flowering genus Crocus L. was missed. The species richness of each garden is given in Table 1. The list of all taxa and the number of observations per taxon is given in the Appendix. The raw data is available from Mendeley Data https://doi.org/10.17632/452pj39jm2.2 [7]. Table 1 Species richness (i.e. number of taxa) of all (S total ), native (S native ), cultivated (S cultivated ) and spontaneous (S spontaneous ) vascular plants in 85 urban gardens in the City of Zurich, Switzerland. The data are based on a complete floristic inventory of garden lots. For each garden, garden type, area and landscape proportion of sealed surface within a 500-m radius is given. Sealed surface was defined as built and paved land-cover. Note that plants can belong to more than one category (i.e., native, cultivated, spontaneous), since a plant can occur spontaneously in one garden while it is cultivated in another, and both cultivated and spontaneous plants can be native. In each garden, in addition to the floristic inventory, plant species were sampled on two circular plots of 10 m 2 each during the summer. The methodology was based on the survey of angiosperm diversity of the Swiss Biodiversity Monitoring program [18]. Potted plants were not included. On each plot, three vegetation layers (ground vegetation, shrub and tree layer) were roughly distinguished and within each layer, the percentage cover of each taxon was scored on an ordinal scale: 1 (<1% cover), 2 (1e5% cover), 3 (>5e25% cover), 4 (>25e50% cover), 5 (>50e75% cover), 6 (>75e100% cover). Note that due to the heterogenous vegetation structure of gardens, cover sums over 100% within a vegetation layer were allowed. The two plots were centred within the two main garden land-use types found in each garden: lawn, meadow, vegetable bed, flower bed or berry patch. In addition, the levels of soil disturbance were contrasted between the two selected land-use types [2]: in each garden, one of the plots had to be in a low soil disturbance land-use type with mostly perennial vegetation (e.g. lawn), while the other had to be in a high soil disturbance land-use type with mostly annual vegetation (e.g. a vegetable bed). The distribution of species richness in each land use type is given in Fig. 1. A comparison between the two sampling methods is given in Fig. 2. The plot-based data can be linked to urban soil data of land-use types [3]. The raw data is available from Mendeley Data https://doi.org/10.17632/ 452pj39jm2.2 [7].

Taxonomic treatment, origin status and origin of garden populations
Taxonomy largely followed the Checklist of the National Data and Information Centre of the Swiss Flora [19]. In addition, for cultivated ornamental plants, Huxley & Royal Horticultural Society [13], J€ ager [14], and Cullen [15] were consulted. Taxa below the species level, taxa within species complexes, and cultivars were not consequently determined at the lowest possible taxonomic rank. They were mostly grouped into aggregates (e.g. Taraxacum officinale aggr.), Cultivar Groups (e.g. Begonia Semperflorens Cultorum Group), or labeleld CV as cultivars without further distinction (e.g. Rosa CV).
Juillerat et al. [19] were followed to assess the origin status of species, i.e. whether a taxon is considered native or alien to Switzerland. Our definition of native plants encompasses archeophytes, which are taxa introduced to Switzerland before 1500, and neophytes of European origin that have colonized Switzerland spontaneously. A more detailed description of the origin status is given in the raw data [7]. Cultivar groups not derived from native plants were considered to be alien.
The origin of a plant individual or "population" in a garden, i.e. whether a plant or a group of plants was cultivated, or whether it occurred spontaneously, was determined by consulting the Flora of the City of Zurich [20] and/or by asking the garden owners. Intentionally introduced plants, which subsequently formed self-sustaining local populations, were considered to be cultivated. Plants that spontaneously colonized a garden and were subsequently tolerated or even locally favored by the garden owner/tenant were considered to be spontaneous plants. The origin of rare native plants was always verified by asking the garden owners. Meadow and lawn plants originating from seeding were considered to be cultivated plants. In lawns and meadows not deliberately enriched with herbs, only grasses employed in landscaping were considered to be cultivated [13]. Only the herbaceous vegetation layer was considered for meadows, lawn, and vegetables, and the tree layer was excluded altogether. Six plots were excluded due to pseudoreplication (e.g. two plots in lawn of the same garden).
It is important to note that the origin of a garden population can not always be unambiguously retraced and should therefore be interpreted with caution: especially in the case of lawn and grassland plants. Note that a taxon can belong to more than one category (native, cultivated, spontaneous), since a species may occur spontaneously in one garden, while it is cultivated in another, and both cultivated and spontaneous plants can be native.

Acknowledgements
We thank Simon Tresch and Andrea Zanetta for their help with data sampling and field work; the garden owners, the Schweizer Familieng€ artner-Verband and their local sections for providing access to the allotment garden areas, Grün Stadt Zürich for their support, Rainer Nowack for kindly sharing literature and Chris Young, Simon Tresch and Robert Home for the helpful comments. The BetterGardens-project (www.bettergardens.ch) is financed by the Swiss National Science Foundation (Sinergia; grant no. 154416).

Transparency document
Transparency document associated with this article can be found in the online version at https:// doi.org/10.1016/j.dib.2019.103982.