Vincent de Leijster

214 set of ecosystem services and that farm design (spatial arrangement of vegetation) and farm management may interfere with the ecological performance. Therefore, I conclude that multiple agroecological practices may best be combined on farm scale to optimize a wide range of ecosystem services. Second, the work described in this dissertation provides more evidence that ecosystem services generally follow nonlinear trajectories. The pathways may either be asymptotic (fast and then slow) or sigmoid (slow, fast, slow) in response to changes in land use or management. The pathways of many ecosystem services related closely to the development of vegetation structural complexity. Examples are carbon stock that follows tree growth development, erosion control, which follows ground cover vegetation, and biodiversity increase after more plant species establish over time. I propose that relationships between vegetation development and ecosystem services result in rather consistent shapes of ecosystem service trajectories in various regions. In contrast to the shape, can we not compare the speed of change between the Mediterranean and tropical regions, since the primary productivity rates differ greatly between these regions. Furthermore, do some ecosystem services have faster responses than others. For example, above-ground carbon stock increases more rapidly than below-ground carbon stock, and butterfly communities build up more rapidly than epiphyte communities because of the differences in mobility. In answer to the third research question, I found that agroecological interventions affect the farm economic performance, and that farming strategies and incentives may alter this effect. Both for almonds and coffee we found that the yield of the main crop was in most cases reduced by the agroecological interventions, except for compost application. However, agroecological management often also affects the quality of the produced product, management costs, and income from other products produced on the farm and these factors also positively influence economic performance. Crop diversification may be a promising strategy to generate income as well as to achieve more variation in the vegetation on the plantation, with all the ecological benefits that this entails. Further, certain external incentives can compensate for yield gaps. Payment for carbon sequestration is an option, but I showed that its potential benefits for farmers are very low since carbon prices in the voluntary market are low. I found that price premiums and greening subsidies provide better potential to compensate for yield losses. For the final question, I concluded that trade-offs among provisioning ecosystem services on the one hand and regulating and supporting services on the other were most frequently occurring. Moreover, we found that regulating and supporting ecosystem services often bundled, for example, carbon storage, erosion control and habitat provision for biodiversity. I suggest that crop diversification may be key to reduce trade-offs and optimize