In first-order headwaters, fish population metrics may be regulated similarly to those in larger order systems downstream, resulting in temporal covariance within watersheds.  The small size of headwater streams relative to drainage area, their close linkage with terrestrial processes, and partial isolation of fish populations may decouple this linkage.  Biogeoclimatic (ecoregion) parameters may influence this decoupling by constraining or augmenting local population dynamics.  We examined abundance, growth, movement, and habitat associations of salmonid fishes in headwater and mainstem tributaries within a major sub-basin of the Columbia River, Washington, USA, for emergence of patterns in those traits at the landscape, tributary and local scales.   This sub-basin consisted of a relatively dry and a relatively wet ecoregion, with some differences in temperature regimes.  We hypothesized that biogeoclimatic association results in broad patterns of density in both the headwaters and the sub-basin tributaries of our study system, and that variation in demographic parameters (e.g., growth and emigration) would follow observed trends in density at either the ecoregion or drainage scale.  Alternatively, these responses might be more strongly influenced by local population effects such as density (intraspecific competition) or local demography/size structure.  We compared overall fish population density and temporal variation in headwaters with that in main tributaries to evaluate whether fish population density varies most distinctly among ecoregions or among individual drainages within the entire sub-basin. 

Results/Conclusions

The dry ecoregion showed higher fish density than the wet ecoregion, but fish density and temporal variation in density were strongly correlated between the headwaters and the main tributaries, suggesting that population size was influenced more at the drainage scale than at the ecoregion scale.  Ecoregion associations did, however influence the response of fish density to habitat features such as stream pool frequency.  Estimated emigration rate, based on mark-recapture assays, was higher in the wet ecoregion than in the dry ecoregion, and was density-dependent in both.  Individual growth was influenced by the size structure of the local fish population and did not significantly differ between ecoregions.  Our results imply that ecoregional contexts may be important to consider when streams are managed for fish production because they appear to be associated with overall productivity and habitat affinity; however some population metrics appear to be under local control.