Wild bee visitors and their association with sown and unsown floral resources in reconstructed pollinator habitats within an agriculture landscape.

In the Midwestern United States where the landscape has been largely converted from tallgrass prairie to row crops, thousands of private land parcels have been enrolled in federal pollinator habitat reconstruction programmes.To examine the outcome of pollinator habitat reconstruction, we randomly selected 19 sites that were in the third growing season post restoration and surveyed plant, floral and wild bee richness and abundance. Floral sources were divided into three categories: species from the seed mix (sown), species from the soil seed bank or surrounding landscape (unsown), and species that were likely unintentionally sown (contamination).Seventy‐two percent of captured bees were collected from sown flowers. The majority of oligolectic bees were Asteraceae specialists. Using linear regressions with bee abundance and richness as response variables and floral density and diversity as predictor variables, we showed that sown floral density was positively correlated with the total abundance and richness of bees collected from sown flowers, indicating that the selected pollinator friendly flowers provide support for wild bee communities. Conversely, total plant stem density or richness or the total floral density or richness was not significantly correlated with bee abundance or richness.The plant–pollinator network nestedness and modularity were positively correlated with the ratio of the unsown floral richness at each site, which suggests that the association between unsown flowers and wild bees can increase the long‐term stability of the plant‐pollinator network, even though the unsown species do not directly increase the bee abundance or richness. [ABSTRACT FROM AUTHOR]