Thomas S. Davis

Forest restoration projects that create canopy gaps can promote floral diversity and may provide reservoirs for bee–flower interactions. However, periods of bee activity must match availability of floral nutrition for bees to benefit from these effects. We model variation in canopy gap size, floral density, pollen protein and bee and floral traits to test the hypothesis that patterns of floral visitation in bee–flower interactions are predicted by patch structure, variation in bee life histories and floral phenotypes. Canopy gap size did not predict overlap of foraging bee abundances and floral protein, but bee life‐history traits did. Foraging activity of large ground‐nesting bees matched accumulation of protein density, whereas activity of small cavity‐nesting bees and kleptoparasitic species preceded peak protein density. Floral visual traits predicted visitation rates, with more interactions for taller species with large floral displays. Both native and invasive species have high nutritional value for bees in canopy gaps and can support frequent and rich interactions. Landscapes with both weedy and native floral species may have higher phenological diversity and provide more foraging opportunities for early‐season bees. We conclude that canopy gap creation is likely to have similar impacts on bee–flower interactions regardless of gap size, but activity periods of larger bees are more aligned with pollen protein density than activity of small bees, and floral visual displays are decoupled from pollen protein content.