Vincent de Leijster

151 Synthesis and conclusion 7 these scenarios as potential hypotheses and to assess the feasibility and likelihood of their development, as well as potential interventions to prevent Scenarios 1-3. Since I investigated coffee over a longer time period than almonds, I can predict its pathway with more certainty (Figure 7-2b). I found that (1) coffee yields were negatively affected by the agroecological intervention, but lower costs and revenues from other products compensated for that, and therefore the coffee agroforestry system had an economic performance that was similar to the conventional system. However, (2) I was able to obtain this finding because farmers had the opportunity to market some of the fruit and timber. If this opportunity had not been available because the supply chains were not present, then agroforestry may have resulted in a lower economic performance than conventional monoculture. (3) In our study case, farmers marketed only 30% of the timber and fruit they produced; therefore, there is still a large untapped potential to further increase economic performance. Moreover, price premiums to compensate for environmentally friendly production or to reward higher quality coffee, and payments for carbon services may improve the economic performance further. However, the institutional context plays an important role in enabling the farmer to benefit from incentives. For example, it depends whether subsidy systems are in place that compensate for environmentally friendly practices or whether traders who buy co-products or pay for ecosystem services are locally present. As a result, the decision regarding which agroecological practice the farmer applies and which farming option is chosen depends not only on the farmer’s capacity in terms of knowledge and investment, but also on the biophysical, institutional and commercial context the farm is embedded in. RQ 4: Which trade-offs or synergies emerge in agroecological transitions? In Chapters 3 and 5 we investigated trade-offs and synergies. First, when we compared agroecological systems to conventional systems we found stronger trade-offs in conventional systems. This was reflected by the fact that most of the investigated regulating and supporting ecosystem services were lower in conventional almond and coffee systems than in their agroecological reference. Nevertheless, the yield of the main crop was higher in both conventional systems, and therefore we can conclude that there are stronger trade- offs between on the one hand provisioning services and on the other hand regulating and supporting services. In other words, agroecological systems provide a more balanced set of ecosystem services and a higher accumulated ecosystem service supply. Looking into the specific relationships between the different ecosystem services, we found a recurring positive relationship between above-ground carbon and habitat provisioning. More specifically, this positive relationship was found between above-ground