The productivity of agroforestry systems (AFSs) and the provisioning of associated ecosystem services (ESs) are threatened by increasing cropping intensification and climate change. Compared to full-sun coffee, maintaining shade cover might protect against climate variability, forest degradation, and pests/diseases attack. However, there may be trade-offs between yields and ESs. While the impacts of shade have been reviewed, a global synthesis to understand how shade trees and a range of practices and biophysical factors contribute to yield gaps and ESs across management types is lacking. This research aims at integrating quantitative data on management practices (shade cover percentage, tree/coffee plant densities, and nutrient inputs), coffee/tree characteristics, soil properties, topographic attributes and climate, in order to compare the responses of coffee yields, biodiversity and ES supply across management types. We conducted a meta-analysis with a total of 142 papers that fulfilled the inclusion criteria and analyzed the individual and combined effects of these factors. The results show that factors such as shade cover and tree/plant densities affected more the yields than climatic and biotic factors. AFSs under moderate shade cover levels (35–50%), with tree densities between 100 and 250 trees ha⁻¹, at low altitudes, and with steeper slopes provided as much yields as full-sun systems (4.1 ± 2.88 tons ha⁻¹ yr⁻¹). While increased shade cover (>51%) decreased yields, the results also show that more diversified AFSs can support biodiversity conservation and provide ESs such as stored carbon, infiltration, pollination, and water runoff. The generalized linear models (GLMs) analyses showed that coffee yield variations are highly context-dependent and factors such as the incidence of pest/leaf rust, altitude, soil quality, available water, and biodiversity may play an important role. The novelty of this research is that the effect of shade, coffee plants, and input management practices is taken into account, along with the effects of a range of site-specific biophysical factors; along with the application of a case study to test robustness of the meta-analysis. This study provides an understanding of the effects of management type gradients on coffee yield and whether the low yields in AFSs compared to conventional coffee monocultures, are compensated by the provisioning of ESs.

农林业系统 (AFS) 的生产力和相关生态系统服务 (ES) 的提供受到日益增加的种植集约化和气候变化的威胁。与全日照咖啡相比，保持遮荫可以防止气候变化、森林退化和病虫害侵袭。但是，收益率和 ES 之间可能存在权衡取舍。虽然已经审查了遮荫的影响，但缺乏了解遮荫树以及一系列实践和生物物理因素如何导致不同管理类型的产量差距和 ES 的全球综合。本研究旨在整合管理实践（遮荫百分比、树木/咖啡植物密度和养分投入）、咖啡/树木特征、土壤特性、地形属性和气候的定量数据，为了比较不同管理类型的咖啡产量、生物多样性和 ES 供应的反应。我们对总共 142 篇符合纳入标准的论文进行了荟萃分析，并分析了这些因素的个体和综合影响。结果表明，树荫覆盖和树木/植物密度等因素对产量的影响大于气候和生物因素。中等树荫覆盖水平 (35–50%) 下的 AFS，树木密度在 100 到 250 棵树公顷之间 结果表明，树荫覆盖和树木/植物密度等因素对产量的影响大于气候和生物因素。中等树荫覆盖水平 (35–50%) 下的 AFS，树木密度在 100 到 250 棵树公顷之间 结果表明，树荫覆盖和树木/植物密度等因素对产量的影响大于气候和生物因素。中等树荫覆盖水平（35-50%）下的 AFS，树木密度在 100 到 250 棵树公顷之间^-1，在低海拔和陡峭的斜坡上提供与全日照系统一样多的产量（4.1 ± 2.88 吨公顷^-1年^-1）。虽然增加遮荫覆盖 (>51%) 会降低产量，但结果还表明，更多样化的 AFS 可以支持生物多样性保护并提供 ES，例如储存的碳、渗透、授粉和径流。广义线性模型 (GLM) 分析表明，咖啡产量变化高度依赖于环境，害虫/叶锈病的发生率、海拔、土壤质量、可用水和生物多样性等因素可能发挥重要作用。这项研究的新颖之处在于考虑了遮荫、咖啡植物和投入管理实践的影响，以及一系列特定地点的生物物理因素的影响；以及应用案例研究来测试荟萃分析的稳健性。