In Sub-Saharan Africa, agroforestry has been identified as the most sustainable remedy to counter declining farm productivity. Over the last decades, researchers and other actors have promoted several agroforestry technologies to improve farm productivity. Sometimes, the promotion message provided through extension assumes a homogenous smallholder farmers’ context. However, smallholder farmers’ social and farm contexts are heterogeneous. Smallholder farmers make different choices of which technologies fit their contexts. A range of factor categories influence and (re)shape choice decisions of smallholder farmers. In this paper, the authors seek to articulate the importance of socio-technological factors shaping smallholder farmers’ choices of specific agroforestry technologies on their farms. Knowledge of these factors provides insights that inform the design of refined farmer context-based extension messages, consequently enhancing the scaling-up of agroforestry technologies. The Decomposed Theory of Planned Behaviour was used as the main framework to understand smallholder farmers’ choice decisions among agroforestry technologies. We used a mixed methods approach. Quantitative data were collected from 277 randomly selected farming households in the eastern highlands of Uganda. Qualitative data that complemented the quantitative were collected using focus group discussions. An alternative-specific conditional logit model was used to model smallholder farmers’ agroforestry choices. Results indicated that the number of tree species desired by the farmer and the perceived value of the technology were the most critical factors that commonly influence smallholder farmers’ choice of agroforestry technologies. The influence of other factors such as gender, the number of training sessions attended, total land owned, peer influence and perceived behavioural control were technology-specific, suggesting the need to tailor agroforestry interventions to specific farmer categories.

在撒哈拉以南非洲，农林业被认为是应对农业生产力下降的最可持续的补救措施。在过去的几十年里，研究人员和其他参与者推广了几种农林业技术，以提高农场生产力。有时，通过推广提供的宣传信息假设了同质的小农背景。然而，小农的社会和农场环境各不相同。小农对适合其环境的技术做出不同选择。一系列因素类别影响和（重新）塑造了小农户的选择决定。在本文中，作者试图阐明影响小农选择其农场特定农林业技术的社会技术因素的重要性。对这些因素的了解提供了洞察力，为改进农民基于上下文的扩展信息的设计提供信息，从而促进农林业技术的推广。计划行为分解理论被用作理解小农户在农林业技术中选择决策的主要框架。我们使用了混合方法。从乌干达东部高地随机选择的 277 个农户收集了定量数据。使用焦点小组讨论收集补充定量的定性数据。一个替代特定的条件 logit 模型被用来模拟小农户的农林业选择。结果表明，农民期望的树种数量和对技术的感知价值是通常影响小农选择农林业技术的最关键因素。其他因素的影响，如性别、参加培训课程的次数、拥有的土地总量、同伴影响和感知的行为控制是特定于技术的，这表明需要针对特定​​农民类别调整农林业干预措施。