Management-mediated predation rate in the caribou-moose-wolf system: spatial configuration of logging activities matters

Landscape complexity can determine the population dynamics of interacting predators and prey. Yet, management plans are commonly developed from aspatial predictive models. This oversight may result in unexpected outcomes or the loss of opportunities to make spatial interventions that would increase a plan's effectiveness. The management of the threatened woodland caribou (Rangifer tarandus caribou), boreal population, provides an example of such shortcomings when using an aspatial approach. Currently, the most influential management recommendation is to maintain at least 65% of undisturbed forests in areas occupied by caribou populations, regardless of the spatial configuration of the forest cover. Using a spatially explicit individual-based model (IBM), we evaluated the effects of the spatial configuration of cuts and roads on the mortality of boreal caribou living in sympatry with wolves (Canis lupus) and moose (Alces alces), an apparent competitor. Starting with a real forest landscape, we created forest management scenarios of the specific spatial distribution of cuts (mosaic, small, or large agglomeration) with increasing disturbance levels. We then ran the IBM with simulated agents, representing individuals of the three species, moving according to movement rules determined from radio-collared individuals. We found that movement responses to land cover types and roads differed among species. For example, caribou and moose generally avoided areas close to roads, contrary to wolves. Those differences influenced the mortality of caribou agents, which not only depended on the levels of disturbance but also depended on the spatial distribution of cuts and roads. After controlling for disturbance level, wolves were more successful when forest management required an extensive road network resulting in relatively high habitat fragmentation. Caribou agents experienced lower mortality in landscapes with low densities of road and disturbance-related edges. The effect remained much stronger, however, for the level than the spatial configuration of human disturbances. Still, our IBM demonstrated how landscape management could be used to manipulate species interactions, with the intent of either increasing or decreasing predation rates on specific populations, depending on management goals.

Automated summary

Landscape complexity plays a key role in determining the population dynamics of interacting predators and prey. The commonly used aspatial predictive models in management plans may lead to unexpected outcomes or missed opportunities for spatial interventions. By utilizing spatially explicit individual-based models, it is possible to manipulate species interactions and achieve specific management goals related to predation rates.