PARENT SESSION
Contributed Oral Session 62: Fish Predation and Morphology
Tuesday, August 9, 1:30 PM - 5:00 PM, Meeting Room 518 C, Level 5, Palais des congrès de Montréal

Adding boulders to increase juvenile salmonid density: Effects of visual isolation and predation risk.

Dolinsek, Ivan^*,1, Venter, Oscar², Grant, James¹, Biron, Pascale², ¹ Department of Biology, Montreal, Quebec, Canada² Department of Geography, Planning and Environment, Montreal, Quebec, Canada

ABSTRACT- Habitat heterogeneity has long been implicated as an important component of habitat quality for fishes. We examined a relatively untested hypothesis that increasing habitat heterogeneity increases visual isolation, thereby shrinking territory size and increasing the density of territorial animals. Specifically, we tested Kalleberg's (1958) hypothesis that the density of territorial salmonid fishes can be increased by adding boulders to a stream, while having no effect on non-territorial fishes. In each of eight reaches in Catamaran Brook and the Little Southwest Miramichi River, in New Brunswick, Canada, we added 36 boulders (median diameter = 0.21 m) to one 3x2 m site, removed all boulders in another site, and left one site as a control in each of two years. We then tested three alternative hypotheses to explain this increase in density: the velocity refuge effect, the predator refuge effect, and the territory size effect. As predicted by the visual isolation hypothesis, adding boulders increased the density of salmonid fishes, primarily juvenile Atlantic salmon, by 2.6-fold, but had no significant effect on non-territorial fishes. A GIS viewshed analysis showed that the visible area and the distance-to-nearest neighbour - and hence the territory size - were also significantly smaller in the boulder-added treatment than in either the control or boulder-removed treatments. Also, fish in the boulder-added treatments selected microhabitat to maximize their field of view, further supporting the visual isolation and territory size hypotheses. We found no support for the velocity refuge hypothesis; the snout velocity used by salmon did not differ between the treatments. Our results showed partial support for the predator refuge hypothesis; in the boulder sites, Atlantic salmon were closer to cover and showed reduced anti-predator behaviour compared to the other treatments, but did not select microhabitats closer to cover. Our results suggest that juvenile salmon are attracted to complex environments, partly for anti-predator reasons and mainly for the decreased visibility of these sites causing a reduction in territory size, allowing a higher density.

Key words: habitat heterogeneity, Kalleberg, Atlantic salmon, territory size