Empirical evidence is mounting that the composition of landscape matters for the suppression of pests in agricultural crops. Non-crop habitats like forest, for instance, have been associated with higher levels of predation and parasitism of several agricultural pests. Yet, the identification of priority areas where implementation of conservation biological control is most effective has received only limited attention. This can be explained in part by the fact that landholders base management decisions at convenient land management scales such as the field, but there is also a lack of practical guidelines on spatial planning for ecosystem services. For this purpose, mechanistic, process-based models can be advantageous. In this presentation I will show using spatially-explicit simulation models how functional traits of predators and the spatial arrangement of their source habitats in the landscape can influence pest populations in crops. I will further highlight how biocontrol mediated by parasitoids can be disrupted by broad-spectrum pesticide use and demonstrate that there are strong thresholds in the proportion of the landscape subject to such pesticide applications for acquiring meaningful levels of biocontrol. This information can inform management strategies that recognize the importance of larger scale issues and may be better suited to capitalize on the pest control services provided by natural enemies.