Authors: Udy, Kristy; Reininghaus, Hannah; Scherber, Christoph; Tschartnke, Teja Date of study: July 2015 Project name: Plant-pollinator interactions along an urbanization gradient from cities and villages to farmland landscapes Type of data: pollinator observations on plant plots. Observations were done by by counting numbers of pollinator visits in a defined time frame. Analysis: we summed abundances for every observation day. The response variables are: number of pollinator visits, plant richness and pollinator morphspecies richness. The influencing variables are: landscape type (a factor with three levels) and plant species richness (numeric). Plant species richness was influenced by landscape type and was always tested in separate models. Generalized least squares models and linear mixed-effects models were used to test the data. The proportional abundance of the seven different pollinator groups was tested using multinomial models against the explanatory variables. Bipartite networks (N=12) were created from the plant-pollinator interactions for each site and their structure analyzed with network level metrics using the bipartite package. The network level metrics used were: robustness, interaction evenness and Shannon diversity of interactions. R (version 3.5.1; R Core Team 2018) was used to run analyses. Figures: Figure 3 shows the relationship between plant species richness and landscape - measured using a linear mixed effects model. Figure 4 shows the relationships between numbers of visits by each pollinator group and landscape type - measured using a generalized least squares model. Figure 5 shows the proportional abundance of each pollinator group in each landscape and with increasing plant species richness - measured using multinomial models. Figure 6 shows pollinator morphospecies richness in each landscape type and with increasing plant species richness - measured using a generalized least squares model. Figure 7 the networks are split by landscape type and were created using the 'plotweb' function in the Bipartite package. Figure 8 shows the different network metrics measured against landscape type, the network metrics were calculated using the Bipartite package. Dataset metadata Dataset 1 Morphospecies level number of visits to each site and plant species code: identifies each garden, based on site name (first two letters), site type (see landuse) and garden number (e.g. GA1). group2: morphospecies groups, bees and bumblebees are grouped into 'wild_bee' pollinator group. See 'group' below. site: type of site. city: code for place name. landuse: AG = agriculture, VI = village, CI = city. species_fam: latin family name for each pollinator species. groups: morphospecies groups are: bee, bumble bee, butterfly, diptera, honey bee, syrphid flies and wasp. visits: number of visits by each pollinator at each site. individual: number of individuals of each morphospecies seen at each site. plant: plant species that pollinator was observed on. plant_richness: flowering plant richness in surrounding landscape (within 20 m of plant plot). flower_cover: prportion flower cover in surrounding landscape (within 20 m of plant plot). prop_garden_500: proportion of gardens within 500 m radius of site. prop_greenSpace_500: proportion of other types of greenspace within 500 m radius of site (e.g. public park). prop_greenSpaceU_500: prop_urban_500: proportion of built-up urban area within 500 m radius of site (e.g. buildings). code_time: letter on end of code identifies time of day pollinators were observed. M = morning, MI = midday. Dataset 2 Number of pollinator visits grouped by pollinator group type for each landuse type. Same variables as dataset 1. Dataset 3 Network metrics per site.