Potential of children’s playgrounds to support pollinators in urban environment: local habitat characteristics matter

The study was conducted in Prague, the capital of the Czech Republic, which is among the greenest metropolitan areas in Europe. Children’s playgrounds represent a widespread but understudied type of urban green space: according to the municipal database, there are approximately 1300 public playgrounds across the city. These playgrounds vary considerably in design and environmental characteristics, ranging from sites with flower beds, spontaneous vegetation, grassy patches, and bare soil to playgrounds dominated by impervious surfaces with minimal natural elements.
We selected 41 playgrounds, and 41 nearby control sites distributed across Prague, spanning the city centre to the outskirts to capture gradients of urbanisation and habitat quality. Sites ranged from semi-natural areas with diverse substrates to structurally simplified sites dominated by impervious surfaces. Control sites consisted of the most common types of urban green spaces, such as parks, public lawns, and grassy patches.
Pollinator surveys were conducted over two consecutive years (2023–2024) during early (June) and late summer (August) periods. Each survey period lasted several consecutive days, with sampling between 09:00 and 17:00 under suitable weather conditions (air temperature > 18 °C, low wind, no precipitation). The study focused on three pollinator groups: bees (without bumblebees), bumblebees, and butterflies. Honeybees (Apis mellifera) were also recorded, but because of their managed status and potentially confounding influence, statistical analyses for bees were conducted both including and excluding them.
Bees (without bumblebees) were sampled during both years at all 82 sites (41 playgrounds and 41 control habitats) using individual sampling using entomological nets. Individuals were collected during standardized 10-minute surveys, covering the focal area to ensure consistent sampling effort. Captured specimens were stored and subsequently identified to species level in the laboratory.
Bumblebees and butterflies were surveyed visually along zig-zag transects to cover whole focal area of given site in a subset of 60 sites (30 playgrounds and 30 control habitats). Bumblebees and butterflies were recorded and identified directly in the field. Bumblebees were classified into four morpho groups based on their colour patterns (gold, white, orange, orange + stripes), whereas butterflies were identified to species level. Each site was surveyed for 10 minutes, with the entire playground or control habitat systematically walked to maximize detection probability.
Immediately before each pollinator survey, we recorded local environmental variables (nectar availability, shading, proportion of impervious surface, bare ground, and vegetated area). We also quantified landscape composition for each pair of sites within 500 m buffers using Sentinel-2 satellite imagery and the Normalized Difference Vegetation Index.