Abstract

For effective management of stream salmonids, it is essential to (i) assess the productive potential of a stream in relation to species-specific habitat requirements, and to (ii) identify the key factors underlying any bottleneck periods during the life cycle of a fish. For this purpose, this PhD-thesis focuses on the mechanisms of habitat selection by juvenile salmonids in boreal streams.

Habitat preference curves for depth, water velocity, substrate and instream cover for brown trout (Salmo trutta L.) in river Kuusinkijoki, northeastern Finland, indicated that larger trout preferred deeper stream areas than age-0 trout did. In summer, all size-classes of trout preferred small substrates, whereas in winter, areas with cobble-boulder substrates were preferred. Winter presents a bottleneck period for trout in boreal streams; therefore winter habitat curves should be incorporated into habitat-hydraulic models when estimating habitat suitable for riverine trout in areas with severe winter conditions. The preference curves of age-0 trout were validated by correlating age-0 trout density with habitat availability at multiple sites in two boreal rivers where trout densities were monitored in 1988–1995. Substrate preference curve was effective in predicting trout densities among sites, whereas among-year variation in trout densities was best predicted by depth-related preference curves.

The responses of age-0 brown trout and grayling (Thymallus thymallus (L.)) to enhancement structures were investigated in artificial stream flumes. For both species a crucial habitat factor was the availability of flow refuges, especially in winter. In another experiment, age-1 trout dominated over age-0 trout when competing for velocity and overhead cover they both found suitable, emphasizing the role of intraspecific interactions in habitat selection by trout. These results suggest that the provision of a broad diversity of microhabitats should be a major goal in rehabilitation programs for fishery purposes.

A new method, combining GIS-assisted (Geographical Information System) approach with geostatistical tools, facilitated the detection of fish distribution patterns in a spatially heterogenous stream habitat. The method will likely prove valuable when determining appropriate sampling scale(s) for future studies of fish habitat selection in relation to benthic prey. Unlike Arctic bullhead (Cottus poecilopus, Heckel), trout did not show any aggregation with their benthic prey according to spatially-referenced data on the distribution of lotic fishes and benthic macroinvertebrates within a stream reach.