Abstract
This thesis presents insights into various issues connected to the practice and adoption of agroforestry in the village cluster of Tiby in rural Mali. In the context of the sustainable livelihoods framework, it is looked at if and how farmers diversify with agroforestry and how this can contribute to reduced livelihood risk and more sustainable land-use. A main focus lies on perceptions of risks and benefits associated to agroforestry practices and multifunctional trees, which are analysed with the help of Bayesian belief netwoks. Moreover, implications to the methodological strategy used for field research and data analysis with Bayesian belief networks are proposed with the aim to contribute to similar future research.