Beyond the average: understanding the heterogeneity of environmental impacts of green coffee in Brazil and the Democratic Republic of the Congo

Coffee production provides income for millions of families, but it also leads to many social-ecological challenges. Life Cycle Assessment (LCA) studies of coffee show that most greenhouse gas emissions occur at the preproduction and production stages. In general, different coffee LCA studies lead to different results, which highly depend on the production practices, set system boundaries, selected methods, and impact categories. There is a need for more primary, transparent, regional and site-specific data covering different environmental impact categories to feed into coffee LCA studies. In this study, we assessed the environmental impacts of producing green coffee based on primary data of five case studies in Brazil and four case studies in the Democratic Republic of the Congo (DRC), covering a total of 60 farms. We considered the impact categories global warming, acidification, eutrophication, land use, water use, ecotoxicity, and carcinogenic and non-carcinogenic human toxicity. Our findings revealed that the environmental impacts of producing green coffee vary widely, even more so than reported in the literature. In this study, we have found that this variability is not only the result of differences in production and management practices but also depends on varying yields. In the DRC in particular, climate-induced yield variability accounted for more variance in our LCA results than differences between farms, countries, or case studies. Further research should focus on incorporating climate change and yield scenarios into LCA studies more effectively to improve our understanding of the different sources of variability in LCA results. • LCAs based on primary data from a very large sample of 60 coffee farms. • One of the first empirical LCA studies of coffee production in Sub-Saharan Africa. • Current coffee LCAs probably underestimate environmental impacts. • Climate change increases the uncertainty of LCA results for agriculture.