Complementing urban agriculture and green spaces is important for ecosystem functions and biodiversity in cities: a systematic review and meta-analysis

•1. As cities expand, they encroach on agricultural land, impacting food production and natural habitats. While urban agriculture could help address these issues, the impact of increased food production on ecosystem functions—particularly with regards to soil conservation, climate regulation and biodiversity—remains poorly understood. •2. To fill this knowledge gap, we conducted a meta-analysis to quantify the effects of urban agriculture on ecosystem functions (i.e. soil quality and climate regulation) and biodiversity (i.e. plants, birds, mammals, arthropods and insect pollinators). We estimated ecosystem functions and biodiversity in urban agriculture by comparing them with other urban green spaces (e.g. parks, residential gardens, green roofs) and conventional rural farms. •3. Our results overall showed that urban agriculture is in an intermediate state between the other urban green spaces and conventional rural farms in terms of environmental impacts. Urban agriculture had a positive effect on the studied ecosystem functions relative to conventional rural farms (+25%, CI95: +12% to +39%), but was equivalent in its provision of these functions to other green spaces. Urban agriculture also had a positive effect on biodiversity compared to conventional rural farms (+38%, CI95: −10 to +111%), but a negative effect compared to green spaces (−12%, CI95: −29 to +8%). Specifically, urban agriculture had 39% (CI95: −59 to −10%) lower plant diversity and abundance than that seen in other green spaces. •4. Practical implications. Our results suggest that urban agriculture's contribution to urban sustainability requires complementarity with other green spaces to effectively support ecosystem functions and biodiversity in cities. We also detected substantial issues with the classification and characterisation of urban agriculture plots. Improving conceptual and methodological consistency will continue to be a crucial challenge for urban agriculture research.