As natural areas within agroecosystems are developed or converted to cropland, ecosystem services such as biocontrol and carbon storage may be compromised. We hypothesize that ecosystem services and landowners’ net profits can be increased by optimizing the spatial distribution and management practices of grasslands that are interspersed in agricultural landscapes. We have been testing this hypothesis in a representative Midwestern landscape by developing and parameterizing a spatially explicit model of crop production, soybean pest control practices, carbon storage, and grassland natural enemy abundance and dispersal. Our model will estimate how stressors affect the supply of grassland-derived ecosystem services, focusing on grassland loss from land use change and changes in grassland composition that result from invasive plant species or management practices. This model will inform new decision-making tools that help optimize ecosystem service provisioning.