Aim We investigate the importance of interacting species for current and potential future species distributions, the influence of their ecological characteristics on projected range shifts when considering or ignoring interacting species, and the consistency of observed relationships across different global change scenarios.

Location Europe.

Methods We developed ecological niche models (generalized linear models) for 36 European butterfly species and their larval host plants based on climate and land-use data. We projected future distributional changes using three integrated global change scenarios for 2080. Observed and projected mismatches in potential butterfly niche space and the niche space of their hosts were first used to assess changing range limitations due to interacting species and then to investigate the importance of different ecological characteristics.

Results Most butterfly species were primarily limited by climate. Species dwelling in warm areas of Europe and tolerant to large variations in moisture conditions were projected to suffer less from global change. However, a gradient from climate to host plant control was apparent, reflecting the range size of the hosts. Future projections indicated increased mismatching of already host-plant-limited butterflies and their hosts. Butterflies that utilize plants with restricted ranges were projected to suffer most from global change. The directions of these relationships were consistent across the scenarios but the level of spatial mismatching of butterflies and their host plants increased with the severity of the scenario.

Main conclusions Future changes in the co-occurrence of interacting species will depend on political and socio-economic development, suggesting that the composition of novel communities due to global change will depend on the way we create our future. A better knowledge of ecological species characteristics can be utilized to project the future fate and potential risk of extinction of interacting

species leading to a better understanding of the consequences of changing biotic interactions. This will further enhance our abilities to assess and mitigate potential negative effects on ecosystem functions and services.