Pollinator and plant datasets

Floral and pollinating insect surveys were conducted from June 7th to July 7th, 2022. All sites were surveyed once, and enhanced and control pairs were surveyed on the same day to minimise possible confounding temporal and climatic effects. A 100 x 2 m linear transect divided into 10 m sections was set up on each site, or for sites smaller than 100 m long, two parallel transects were set up 20 m apart to create a total length of 100 m. Floral abundance was recorded along transects in 1 m2 quadrats placed randomly in each section (10 x 1 m2 quadrats x site). The total number of floral units per quadrat for each species was recorded. One floral unit was defined as a head (e.g., Trifolium repens), an umbel (e.g., Heracleum sphondylium), or a capitulum (e.g., Centaurea nigra) (outlined by Tarrant et al., 2013 and Carvell et al., 2006). Only open and fresh flowers were included in floral counts. Native wildflower species were identified using Rose (1981) and horticultural plants identified using iNaturalist (2022), Pl@ntNet (2022), and Seek (2022). Pollinating insect surveys were conducted along the same transect as the floral survey on warm, sunny days, with little wind (<5 Beaufort) (O'Connor et al., 2019) between 09:30 and 19:00. Each transect was walked either by two surveyors each observing 1 m to the left or right of the transect, or by one surveyor observing 1 m to the left of the transect line then repeating the transect and observing 1 m to the right, resulting in equal areas surveyed. Surveyors walked at a slow pace and recorded a plant-pollinator visit whenever an insect visitor touched the sexual organs of a receptive flower (Tarrant et al., 2013).