Landscapes provide ecosystem services to agricultural systems by supporting pollinators and predators of crop pests, services valued at $US 3 and $US 4.5 billion/yr respectively.  Habitat management is the practice of providing nectar, pollen and shelter to beneficial insects in cropping systems, often in the form of flowering strips. However, the potential for flowering strips to increase biocontrol and pollination depends on the existing abundances of beneficial insects in the landscape, and highly simplified landscapes may support fewer beneficial insects than more diverse landscapes.  To test the effect of landscape characteristics on the success of habitat management, we sampled beneficial insects at soybean fields with and without flowering strips across a range of landscape diversity. Soybean plants infested with aphids were used to compare the level of pest reduction and sunflowers were used to measure the level of pollination. Habitat management in the form of flowering strips has the potential to support increased levels of both biocontrol and pollination, leading to more sustainable crop production.
Results/Conclusions

Across landscapes, both seed weight and seed set of sunflowers were significantly higher in fields with flowering strips.  Furthermore, landscape characteristics influenced the effect of flowering strips on seed set.  Sunflower seed set was negatively related to the abundance of perennial grasses and forbs in the landscape in control fields, indicating that floral resources in these landscapes attracted pollinators away from our sites and diluted pollination effects.  In contrast, sunflower seed set in fields with flowering strips was positively correlated with abundance of perennial grasses and forbs in the landscape.  This suggests that there is a greater total abundance of pollinators in landscapes with more perennial habitat, and that the flowering strips were able to successfully attract pollinators to the sunflowers in these landscapes.  In the biocontrol study, natural enemies dramatically suppressed soybean aphids in all sites.  Across treatments and sites, the number of aphids on soybean plants from which predators were excluded was significantly higher than on soybean plants that were exposed to predators.  There were no differences in strength of aphid suppression between landscapes or treatments in 2008, probably due to low overall aphid pressure.