Aim Cumulative impact maps are used to identify the spatial distribution of multiple human impacts to species and ecosystems. Impacts can be caused by local stressors which can be managed, such as eutrophication, and by global stressors that cannot be managed, such as climate change. Cumulative impact maps typically assume that there are no interactive effects between stressors on biodiversity. However, the benefits of managing the ecosystem are affected by interactions between stressors. Our aim was to determine whether the assumption of no interactions in impact maps leads to incorrect identification of sites for management.