Prairie restorations and bees: The potential ability of seed mixes to foster native bee communities

Restoring native habitat is considered critical to conserving native pollinators threatened by habitat loss and degradation, but little is known about whether ecological restorations, most of which do not target pollinators, can predictably support pollinator communities. In this study we compare plant species richness in common commercial prairie seed mixes to remnant prairies, and we take a modeling approach to examine the ability of these seed mixes to attract bee communities relative to native prairie remnants. Using a large data set from native prairies in Iowa consisting of 70 bee species (≈2700 bee specimens) associated with 54 native prairie plants, we constructed species accumulation curves to model the number of bee species potentially attracted to a restoration with the addition of each plant in a seed mix. Our modeling results indicate commercial prairie mixes will accumulate species at rates similar to prairie remnants, but the bee species richness will be lower than remnants because the plant species richness in samples from prairie remnants is twice that of the average commercial seed mix. However, when commercially available seed mixes were modeled to always include four plant species that were exceptionally attractive to native bee species, most mixes attracted significantly more bees than predicted if random species were added. This further suggests that seed mixes and the resulting restorations do not adequately provide for pollinators and could be significantly improved with the addition of a small number of species. Although the particular optimal species additions to seed mixes will vary regionally, adding species functionally equivalent to those we identify may significantly improve restoration of ecological services provided by native bees.