Developing agricultural systems that produce food and income to sustain smallholder livelihoods in the tropics, without further compromising ecosystem functioning and biodiversity conservation, is an important societal challenge. Agroforestry systems pose a promising approach to deal with the twin challenges of local development and conservation of ecosystem services like biodiversity. But while the ecological benefits of these systems are clear, they are still believed to underperform economically, compared to intensified conventional systems. More insight in the relations between crop productivity, biodiversity and smallholder livelihoods is needed to identify systems that can minimise trade-offs between economic and environmental performance or even provide double dividends. This thesis therefore aims to i) assess the economic and environmental outcomes of smallholder management systems to identify trade-offs and seek opportunities for double dividends, and ii) identify opportunities and constraints for smallholder farmers to adopt associated management strategies. Due to their economic and ecological importance worldwide, this thesis focuses on coffee and cocoa smallholder systems; with in-depth empirical case study on smallholder coffee systems in San Martín, Peru.
This thesis shows that contrary to popular belief, agroforestry systems can offer competitive business opportunities in comparison to the expanding, conventionally intensified systems and can reconcile farmer livelihoods and conservation of biodiversity and other ecosystem services. First of all, no trade-offs were found between coffee yields, forest butterfly species richness and above-ground carbon storage and shaded coffee systems supported biodiversity and carbon storage. Importantly, no evidence for a negative relation between coffee yields and shade was found across a shade cover range of 0-80%. Secondly, this thesis provides evidence that systems with high shade levels can perform equally well or better compared to plantations with lower shade levels and/or with higher input levels. There were no differences between net income and benefit-cost ratio (BCR) for plantations with different shade management practices. Rather, net income and BCR were lower for plantations with higher input practices. This was explained to some extent by higher costs of intensified systems, both for flexible (inputs and labour) and fixed costs (land and equipment), while costs were lower for plantations with higher shade levels. Third, there were no trade-offs between butterfly diversity and above-ground carbon storage with net income for this case study of Peruvian coffee farmers. Lastly, the results suggest that improving livelihood assets is important for decision making, and the actionable assets differ for shade and input management; whilst human, social and natural assets may limit or enhance adoption of environmentally-friendly management systems, financial and physical assets may affect adoption of input management strategies. Importantly, it became clear that the benefits of agroforestry systems are very diverse and location-specific and that there is no blueprint for management systems that will provide double dividends under all circumstances. In order to reconcile economic and ecological goals in coffee and cocoa systems, comprehensive multidisciplinary analyses are needed, including for other regions, to be able to draw generalizable conclusions and deepen our insight in trade-offs between economic and environmental performance.