Understanding the spatial and temporal foraging patterns of pollinators is essential to conserving these organisms in human‐modified landscapes, such as agroecosystems (Cranmer et al. 2011). Particularly, understanding foraging fidelity (i.e., returns to a particular plant species or area) is crucial for providing supportive habitats. Numerous approaches have been used to discern the movement and foraging of individual bees (reviewed by Mola and Williams 2019), ranging from highly technological to more traditional methods like mark–release–recapture experiments. Although mark–recapture experiments are often hampered by low recapture rates, density dependence, and logistical constraints (Mola and Williams 2019), they can help us understand short‐term local movement patterns (Dorchin et al. 2013) and fidelity (Ogilvie and Thomson 2016) of highly mobile organisms like bees.

We conducted eight mark–release–recapture experiments (Appendix S1) at five large-scale commercial plant nurseries in southern California, USA, surrounded by both suburban and natural landscapes. Experiments were conducted in summer and autumn 2018, between 11:00 and 14:00 on consecutive clear, sunny days. Three nurseries were sampled in both seasons, one was sampled in summer only, and one in autumn only. A single observer hand-netted male and female Halictus ligatus on patches of blooming plants over areas up to 2 ha. If H. ligatus was not sufficiently abundant, we then focused on the few most abundant species of non-Apis bees, including solitary species. Captured bees were iden- tified, marked on the thorax (Fig. 1) using nontoxic paint, then released. Paint color was used to indicate the plant species on which the bee was captured. The following day, we returned and netted marked and unmarked bees in the same and nearby patches where we marked bees the previous day. All individuals were retained in vials to ensure no individual was recaptured twice.

Both data files have a "metadata" tab describing the variables.

"recapture.xlsx" contains data for each individual bee that was captured on the first day of an experiment, e.g. the bee's genus, the nursery, the season, the plant on which it was captured, whether it was recaptured on the second day of the experiment, whether it switched plants, etc.

"electivity.xlsx" contains data used to calculate an electivity index for each species of plant in the experiment. For each species of plant at each nursery in each season, provided are data on the patch area as well as how many bees were captured on that plant species. For reference, see Jenkins, S. H. (1979). Seasonal and year-to-year differences in food selection by beavers. Oecologia, 44(1), 112-116.