AIM: Although there is a vast body of literature on the causes of variation in species composition in ecological communities, less effort has been invested in understanding how interactions between these species vary. Since interactions are crucial to the structure and functioning of ecological communities, we need to develop a better understanding of their spatial dynamics. Here, we apply novel numerical tools to data on species interactions, and reveal that they vary more, and in response to different climate variables, than species do. LOCATION: Eurasia. METHODS: We used a measure of Locality Contribution to Beta-Diversity to evaluate the compositional uniqueness of 51 host--parasite communities across Eurasia, using publicly available data. We measured uniqueness based on the species composition, and based on potential and realized biotic interactions. RESULTS: We show that interactions vary more, over space, than species do. In particular, we show that interactions respond to some climatic variables that have no effect on species distribution or dissimilarity. MAIN CONCLUSIONS: Species interactions provide far more resolution than species occurrences alone, while still retaining all information about species occurrences. We suggest that they be put front and center in analyses of communities, especially in a biogeographic context.