Résumé
Trophic interactions in multiprey systems can be largely determined by
prey distributions. Yet, classic predator–prey models assume spatially
homogeneous interactions between predators and prey. We developed a
spatially informed theory that predicts how habitat heterogeneity alters
the landscape-scale distribution of mortality risk of prey from predation,
and hence the nature of predator interactions in multiprey systems. The
theoretical model is a spatially explicit, multiprey functional response
in which species-specific advection–diffusion models account for the
response of individual prey to habitat edges. The model demonstrates that
distinct responses of alternative prey species can alter the consequences
of conspecific aggregation, from increasing safety to increasing predation
risk. Observations of threatened boreal caribou, moose and grey wolf
interacting over 378 181 km2 of human-managed boreal forest support this
principle. This empirically supported theory demonstrates how distinct
responses of apparent competitors to landscape heterogeneity, including to
human disturbances, can reverse density dependence in fitness correlates.