Vincent de Leijster

140 Chapter 7 7.1 Synthesis The UN have declared the coming decade (2021-2030) the decade of ecosystem restoration, and this demonstrates the urgency of reversing land degradation. Agriculture has a key role to play, not only because it is currently the strongest driver of land degradation but also because it covers more than a third of the Earth’s land. In this thesis I have examined how agroecological management transitions in woody croplands may contribute to rehabilitating ecological performance, and how economic benefits can be maintained during this transition. More specifically, I have examined agroecological transitions in Spanish almond orchards (Chapters 3 and 4) and Colombian coffee plantations (Chapters 5 and 6). I have tested how multiple ecosystem services respond to agroecological interventions, and how they interact (Chapters 3 and 5). In these same systems, I have investigated how agroecological interventions affect the economic performance of a farm , and how the economic outcome can be influenced by farming options and incentives (Chapter 4 and 6). I have studied short- term developments in the almond case study and long-term developments in the coffee case study. In this final chapter, I first give a summary of the findings per chapter and then I answer and discuss the research questions (Section 1.6). In Chapter 3 , three agroecological interventions were implemented in five almond orchards, namely ‘no tillage with spontaneous vegetation’, ‘green manure’, and ‘compost’. These interventions relate to the agroecological principles of soil conservation and vegetation diversification. Together with my co-authors, I evaluated whether after one year the ecosystem services on these sites differed from those of conventional tillage management, which was the mainstream practice in the region. We found that the accumulated ecosystem service supply was 17-28% higher on the agroecological sites than on the conventional tillage sites. However, each practice supported a different set of ecosystem services (Figure 3-2). Vegetation cover under ‘no tillage’ and ‘green manure’ improved understory carbon content, plant species richness and soil microbial activity, which are indicators of the ecosystem services ‘carbon storage’, ‘habitat provisioning’ and ‘nutrient cycling’. Compost application supported soil-related ecosystem service indicators such as ‘soil microbial activity’, ‘soil and tree nutrient content’ and ‘soil carbon storage’. Moreover, compost also improved the provisioning services as indicated by heavier kernels. Insect-mediated ecosystem services, ‘pest control’ and ‘pollination’, were not affected by the agroecological treatments, but we concluded that changes may develop over longer timescales or larger spatial scales. Although we found no significant effect of treatments on almond production, we did find a negative relationship between vegetation cover and almond production (Figure 3-3). In combination with findings of other studies, this suggests that trade-offs may develop between regulating