Vincent de Leijster

146 Chapter 7 We noticed some differences between ecosystem services in the speed of the response. For example, in carbon stock we found that above-ground carbon improved while below- ground carbon remained stable both in coffee and in almond production, and we concluded that below-ground carbon was too slow to be measured. This was in line with several other studies that found that below-ground carbon dynamics are slower than above-ground carbon dynamics (Martin et al., 2013; Yang et al., 2019). Furthermore, we found that butterfly communities responded more rapidly than epiphytes did, as we found lower half-time coefficients. This could be explained by the high mobility of butterflies and the dependency on vegetation structure for the colonization of epiphytes (Cruz-Angón & Greenberg, 2005; Meyer & Sisk, 2001). In Chapter 5, as expected, changes in erosion control were too fast to measure with our chronosequence analysis, as erosion risk is lower in agroforestry than in monoculture. Erosion risk depends strongly on whether the soil is covered by vegetation and litter, and this may change quickly if management is adjusted. For example, we also found large differences in vegetation cover within one year in the almond plantations. Vegetation cover has been shown to strongly reduce erosion rates (Martínez Raya et al., 2006). In this study, I have demonstrated that some ecosystem services respond faster than others, but more information is still needed from other geographical regions and land-use systems to be able to better predict ecosystem service trajectories. RQ 3: How do agroecological interventions influence the economic performance of a farm, and what is the role of farming strategies and incentives? To answer this research question, I turn to the agroecological transition model that I presented in the introduction (Figure 1-2), and I use information from Chapters 4 and 6. Below, I first explain how agroecological interventions directly influenced costs and benefits, then I discuss which farming options and economic incentives were found to be promising in alleviating economic performance, and finally I place the case studies in the agricultural transition model. In the almond case study (Chapter 4), we tested how economic performance was affected by the agroecological interventions related to soil conservation and vegetation diversification, and how the economic incentives (such as price premiums, public subsidies and compensation for ecosystem services) influenced the outcome. We found that vegetation cover in almond orchards negatively affected the yield of the main crop, which became apparent from lower gross and net revenues than in the conventional system. However, compost positively affected yield and quality of the main crop, resulting in higher