With the intensification of human activities, preserving animal populations is a contemporary challenge of critical importance. In this context, the umbrella species concept is appealing because preserving a single species should result in the protection of multiple co-occurring species. Practitioners, though, face the task of having to find suitable umbrellas to develop single-species management guidelines. In North America, boreal forests must be managed to facilitate the recovery of the threatened boreal caribou (Rangifer tarandus). Yet, the effect of caribou conservation on co-occurring animal species remains poorly documented. We tested if boreal caribou can constitute an effective umbrella for boreal fauna. Birds, small mammals, and insects were sampled along gradients of post-harvest and post-fire forest succession. Predictive models of occupancy were developed from the responses of 95 species to characteristics of forest stands and their surroundings. We then assessed the similarity of species occupancy expected between simulated harvested landscapes and a 90 000-km2 uncut landscape. Managed landscapes were simulated based on three levels of disturbance, two timber-harvest rotation cycles, and dispersed or aggregated cut-blocks. We found that management guidelines that were more likely to maintain caribou populations should also better preserve animal assemblages. Relative to fragmentation or harvest cycle, we detected a stronger effect of habitat loss on species assemblages. Disturbing 22%, 35%, and 45% of the landscape should result, respectively, in 80%, 60%, and 40% probability for caribou populations to be sustainable; in turn, this should result in regional species assemblages with Jaccard similarity indices of 0.86, 0.79, and 0.74, respectively, relative to the uncut landscape. Our study thus demonstrates the value of single-species management for animal conservation. Our quantitative approach allows for the evaluation of management guidelines prior to implementation, thereby providing a tool for establishing suitable compromises between economic and environmental sustainability of human activities.

The northern redbelly dace, Phoxinus eos (Cope, 1862), is subject to predation by brook trout, Salvelinus fontinalis (Mitchill, 1814), in Canadian Shield lakes, particularly when individuals migrate to the pelagic zone at sunset to feed on zooplankton and fish shoals break up into single individuals. The objectives of the present study were to (i) determine whether northern redbelly dace react to skin extracts from conspecifics and thus potentially use chemical alarms to detect predators in nature, (ii) characterize the fright reaction in northern redbelly dace under different concentrations of alarm substance, and (iii) estimate the active space of the alarm substance in this species. Northern redbelly dace responded to skin extracts of conspecifics with a series of antipredator behaviours. The dace moved closer to the substrate and away from the area where the alarm substance was injected, increased both the cohesion and polarization of their schools, and performed more dashing and freezing behaviours. The observed responses were closely correlated with the concentration of the skin extract solution, suggesting that individuals could use this sensitivity to different concentrations of alarm substance to estimate the risk of predation in nature. The dilution experiment allowed us to estimate that 1 cm2 of northern redbelly dace skin contains enough alarm substance to generate a response of individuals in 110 558 L of water, which corresponds to a cube with sides of approximately 4.8 m. Indirect evidence suggests that the northern redbelly dace could use chemical alarm cues to assess the risk of predation by brook trout in nature.

No abstract