Restoration of native plants on highly altered lands depends on the restoration of ecosystem processes and services. Pollination may be inadequate for sites located distant from habitats supporting source populations of pollinators. We investigated the spatial scale at which high quality forest and prairie habitat influences movement of pollinators into degraded, formerly mined land prior to restoration. We predicted that richness and abundance of pollinators would decline with distance from high quality habitat, possibly resulting in lower plant reproductive output. The scale of this effect should differ for pollinators depending on their ecology and dispersal distances. Local densities of floral resources could also affect pollinator communities. We sampled bees and butterflies at 24 sentinel patches of native plants planted within formerly mined, reclaimed grassland in south central Ohio, USA. The patches were located 10-400 m from forest edge habitat and 10-700 m from the edge of a restored prairie habitat offering abundant floral resources in June-August, with a minimum of 50 m between patches. We present pollinator abundance and diversity from water bowl traps taken in April-July. We also present data on flower visitation rates and seed production in the patches of native plants in each plot.
Generic richness of bees declined with distance from restored prairie habitat and forest edge. Distance from restored habitat explained more of the variance in bee generic diversity (partial r2 = 0.30, p = 0.03) than distance from forest (partial r2 = 0.25, p = 0.09). The richness of butterflies also declined with distance from the restored habitat (R2=0.35, p = 0.03), but not from the forest edge. Neither butterfly nor bee abundance was affected by the distance from forest or the restored habitat, although butterfly abundance declined significantly with distance from restored habitat in the July sample. Local floral density did not affect the abundance or diversity of bees or butterflies. Forest influenced the composition of bees in plots. The genus Andrena, which nests in forest soils, and its cleptoparasite, Nomada, only occurred in plots nearest forest. Visitation rates to Penstemon digitalis (June) and Liatris pychnostachia (July) did not vary with landscape position. In contrast, visitation to Coreopsis tinctoria (August) flowers declined significantly with distance from the restored habitat, but not forest. Here, restored prairie maintains local pollinator diversity and forest contributes unique species over small spatial scales, which translate into improved reproduction for some native plant species.