Indirect effects resulting from species sharing the same enemy can shape spatio-temporal variations in species occurrence. The strength of such effects remains poorly known in natural communities composed of species from different trophic levels interacting in heterogeneous landscapes. Benefiting from a well-known arctic vertebrate community and marked spatio-temporal variations in the density of key prey species, we examined the effects of direct predator-prey and indirect predator-mediated effects on species occurrence in the landscape. We found both positive effects of one prey (lemmings), as well as negative indirect effects of another prey (colonial nesting snow geese) on the occurrence of species (ground-nesting birds) belonging to different guilds and trophic levels but sharing a common predator (arctic fox). However, species using prey refuges available in the landscape were not or less affected by predator-mediated effects. Similarly, the smallest (a passerine) and the largest and most dangerous species (an owl) for the shared predator were not affected by these effects. Our study provides one of the rare empirical evidence of predator-mediated effects ascending the food web (i.e., negative indirect effect of an herbivore on avian predators) and underlines how habitat structure and species traits can modulate the strength of indirect effects in natural communities.

The dataset provided here contains the data used in each of the models presented in the article "Variable strength of predator-mediated effects on species occurrence in an arctic terrestrial vertebrate community". These data are the processed values (as described in the method section and supporting information of the article) used to model the effects of lemming density and distance to a goose colony on the occurrence of various nesting birds. Mapped data is available in supporting information of the article.

The README file contains column descriptions and general information for the analyses.