Vincent de Leijster

117 Chronosequence analysis of economic performance of agroforestry coffee farms in Colombia 6 related to lower costs of pest and weed control. However, use of agrochemicals is often not related to canopy management (Cerda et al., 2017; Jezeer et al., 2018). The benefits of agroforestry may vary along its life cycle. As economic indicators (e.g. yield, price, costs, income from other sources) may be respond to the development of the canopy layer (Siles et al., 2010). A Costa Rican study on coffee and a Ghanaian and Bolivian study on cocoa, experimentally demonstrated variation in yield development between shaded and unshaded conditions. These studies solely included yields, but not costs, marketed prices and additional benefits from co-products, and therefore did not provide insights into farm economic performance. The approach of space-for-time substitution, widely used in landscape ecology, has been used to monitor the development of ecological and economic performance indicators in agroforestry (Jain and Singh, 2000; Kattwinkel et al., 2011; Nijmeijer and Harmand, 2019). In this approach, sites that differ in age-since-intervention are compared as each represent different steps in time and in this way provide long-term information in a short period. For example, space-for-time substitution was used to assess the development of costs, benefits and net revenues of Indian poplar agroforestry systems during the 8 year rotation period (Jain and Singh, 2000). We used this approach also in a recent ecological study in the same region and found that the supply of ecosystem services in Colombian coffee plantations improved over time since implementation of agroforestry (De Leijster et al., 2021). Here we aim to understand how economic indicators develop in coffee plantations since trees were planted. We used a space-for-time substitution approach to reconstruct the development of economic indicators from newly established agroforests to mature agroforests (1 to 40 years after transition from monoculture) and compare this to the net revenue of unshaded coffee plantations. First, we analyzed how net revenue developed, and which specific costs and benefits contribute to these developments. Secondly, we examined the development of non-marketed products and services (tree fruits, timber and carbon) since implementation of agroforestry. Thirdly, we examined which factors might explain economic performance of coffee plantations, including characteristics of the supply chain (certificates and intermediary), location (altitude), management intensity (pest control, herb control, fertilization), farm in general (farm size, time since coffee pruning), and canopy (tree density, species richness and tree spatial arrangement). We assess the development of economic performance in three planting regimes that differ in tree spatial arrangement (trees dispersed, in alley formation or in living fences). The findings of this study will contribute to a better understanding of whether ecosystem services in tropical agroecosystems may be restored while providing profit. This will allow assessing whether external investments are needed to compensate for potential income gaps.