Urban areas as hotspots for bees and pollination but not a panacea for all insects

Urbanisation is an important global driver of biodiversity change, negatively impacting some species groups whilst providing opportunities for others. Yet its impact on ecosystem services is poorly investigated. Here, using a replicated experimental design, we test how Central European cities impact flying insects and the ecosystem service of pollination. City sites have lower insect species richness, particularly of Diptera and Lepidoptera, than neighbouring rural sites. In contrast, Hymenoptera, especially bees, show higher species richness and flower visitation rates in cities, where our experimentally derived measure of pollination is correspondingly higher. As well as revealing facets of biodiversity (e.g. phylogenetic diversity) that correlate well with pollination, we also find that ecotones in insect-friendly green cover surrounding both urban and rural sites boost pollination. Appropriately managed cities could enhance the conservation of Hymenoptera and thereby act as hotspots for pollination services that bees provide to wild flowers and crops grown in urban settings.


Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.

n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly The statistical test(s) used AND whether they are one-or two-sided Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals) For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted Give P values as exact values whenever suitable.

For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated Our web collection on statistics for biologists contains articles on many of the points above.

Software and code
Policy information about availability of computer code Data collection

Data analysis
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors/reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.

Data
Policy information about availability of data All manuscripts must include a data availability statement. This statement should provide the following information, where applicable: -Accession codes, unique identifiers, or web links for publicly available datasets -A list of figures that have associated raw data -A description of any restrictions on data availability Theodorou, Panagiotis Dec 24, 2019 Data on landscape heterogeneity were obtained using Quantum GIS v2.18 (https://qgis.org) and land cover data obtained from Geofabrik GmbH (https://www.geofabrik.de/).
All statistical analyses were performed in R statistical software v 3.5.2 using the packages picante v. Field-specific reporting Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection. Field work, collection and transport

Field conditions
We used a paired study design at flower-rich sites in nine independent German cities and nine nearby, equivalent, flower-rich rural sites to test the impact of urbanization on pollinator biodiversity and the ecosystem service of pollination. We used metabarcoding to evaluate flying insect diversity. To quantify pollination, we potted greenhouse-raised, insect-pollinator dependent red clover plants.
We employed pan-traps to sample insects and to compare the diversity of flying insects at urban with those at rural sites, we measured species richness of the four main orders of flying insect pollinators: Diptera, Lepidoptera, Coleoptera, and Hymenoptera. In addition, we monitored all insects visiting the flowers of the experimental red clover plants (10 plants in each site) for five hours at each site in order to estimate flower visitation rates. Each urban-rural site pair was visited at the same time and for a total of five consecutive warm, non-windy days between June and August 2014. To determine the main ecological correlates of insect biodiversity and pollination in both rural and urban flower-rich sites, we gathered a series of local (patch) and landscape-scale variables potentially related to insect pollinators and pollination. These were (1) local flowering plant richness and abundance using 10 randomly placed 1 square meter quadrats at each of our sampling sites, (2) the proportion of semi-natural cover (grassland, meadows and scrub vegetation), (3) the proportion of forest, (4) the extent of arable (=agricultural) cover, (5) the proportion of residential and (6) commercial/industrial land cover, (7) the extent of botanical gardens, public parks and allotments, (8) landscape diversity and (9) edge density, as total length of 'green cover' (semi-natural and forest cover, botanical gardens, public parks, and allotments) patch edges divided by the total area, and which represents a quantification of landscape configuration. Given our paired 'urban-rural' experimental design, the rationale in our statistical analyses was to use site pair as a random factor and to compare between ecosystem type (urban versus rural). We controlled for potentially confounding local and landscape factors, unless we specifically aimed to model their relationship to predictor variables: dimensions of biodiversity and pollination.
Our research sample is the insect community (Diptera, Lepidoptera, Coleoptera, and Hymenoptera) samples in each of our sites. It is characterized by overall insect biomass, species number and phylogenetic diversity.
We determined adequate sample size in our preliminary study, which has already been published ( Insects were sampled using three blue, three yellow and three white pan traps (diameter: 42 cm, height: 2.8 cm) mounted on a stick at vegetation height at each site. Each pan trap was 2/3 filled with unscented soapy water and emptied every day for a total of five consecutive warm, non-windy days between June and August 2014. Insects from traps were killed on-site using 95% ethanol and stored in a -20°C freezer. Insect samples from each site were washed, dried and weighed using a balance. For assessment of species richness, we used next generation sequencing (NGS)-based metabarcoding. All people involved are listed in acknowledgements.
Each urban-rural site pair was visited at the same time and for a total of five consecutive warm, non-windy days at one point between 12/6/2014 and 10/8/2014. We sampled insects in cities and rural locations in central and eastern Germany.
No data were excluded from the analyses.
We used a highly replicated and statistically robust experimental design across multiple, paired sites.
All samples collected for our analyses were from well defined, pre-selected locations, and samples were collected from all locations. Thus our sampling design was 'fully crossed'.
All analyses were performed blind. This is especially relevant for our experimental pollination data, in which we collected seed from open versus closed flowers, and for our next-generation-sequencing data, in which we meta-barcoded pan-trapped insects. For these two datasets, bags/tubes containing material were given a unique number code that did not contain details of treatment, then data were collected/generated, and only afterwards were treatments allocated to processed data.
Temperatures exceeded 16°C, wind speed was less than 2 m/s at 1 m above ground level, and skies were sunny (<50% cloud cover) on all sampling days. These conditions are optimal to sample insects in our region.