Copyright © 2023 Dutta, De Barba, Selva, Fedorca, Maiorano, Thuiller, Zedrosser, Signer, Pflüger, Frank, Lucas and Balkenhol.Introduction: Connected landscapes can increase the effectiveness of protected areas by facilitating individual movement and gene flow between populations, thereby increasing the persistence of species even in fragmented habitats. Connectivity planning is often based on modeling connectivity for a limited number of species, i.e., “connectivity umbrellas”, which serve as surrogates for co-occurring species. Connectivity umbrellas are usually selected a priori, based on a few life history traits and often without evaluating other species. Methods: We developed a quantitative method to identify connectivity umbrellas at multiple scales. We demonstrate the approach on the terrestrial large mammal community (24 species) in continental Europe at two scales: 13 geographic biomes and 36 ecoregions, and evaluate the interaction of landscape characteristics on the selection of connectivity umbrellas. Results: We show that the number, identity, and attributes of connectivity umbrellas are sensitive to spatial scale and human influence on the landscape. Multiple species were selected as connectivity umbrellas in 92% of the geographic biomes (average of 4.15 species) and 83% of the ecoregions (average of 3.16 species). None of the 24 species evaluated is by itself an effective connectivity umbrella across its entire range. We identified significant interactions between species and landscape attributes. Species selected as connectivity umbrellas in regions with low human influence have higher mean body mass, larger home ranges, longer dispersal distances, smaller geographic ranges, occur at lower population densities, and are of higher conservation concern than connectivity umbrellas in more human-influenced regions. More species are required to meet connectivity targets in regions with high human influence (average of three species) in comparison to re...