• emergence success;
  • evolution;
  • groundwater flow;
  • habitat selection;
  • hatching success;
  • spawning ground;
  • substrate


1. Despite the great interest in characterising fish reproductive habitat, the relationship between the selection of a given spawning site and individual fitness has not been experimentally tested.

2. In this study, we used an in situ experimental approach to determine (i) the relative contribution of substrate characteristics as well as hydrological and physicochemical variables to small-scale redd site selection by brook charr, Salvelinus fontinalis (Mitchill) and (ii) if hatching and emergence success, used as a surrogate of fitness, are improved in selected compared to non-selected sites in both lake and stream habitats.

3. Our results show that upwelling groundwater flow was always significantly higher in selected than in non-selected sites in both lake and stream habitats. We found no significant difference in the mean geometric substrate diameter and no consistent differences in substrate composition between selected and non-selected sites. Oxygen concentration was higher (significantly so in three of four comparisons) and conductivity tended to be lower in selected than in non-selected sites, while temperature showed no significant or consistent variations. We found a significant positive relationship between the selection of a given spawning habitat and hatching and emergence success in these systems.

4. These results show that the main cue that brook charr use to select their spawning sites is upwelling groundwater in lake and stream habitats, and that active selection of these sites increases individual fitness. This suggests that natural selection acted on the same cues in lentic and lotic environments; this could have been highly adaptive in a species that used both habitats as colonisation routes after the last glaciation.