Northwest Fisheries Science Center

Display All Information

Document Type: Report
Center: NWFSC
Document ID: 9123
Title: River conditions, fisheries and fish history drive variation in upstream survival and fallback for Upper Columbia River spring and Snake River spring/summer Chinook salmon
Author/Editor: Lisa G. Crozier, Lauren Wiesebron, Elene Dorfmeier, Brian J. Burke
Year: 2017
Institution: National Marine Fisheries Service, Northwest Fisheries Science Center. Seattle, Washington

To identify what factors drove variation in survival during upstream migration, we analyzed a large database of spring/summer Chinook salmon from both the Upper Columbia River spring run and Snake River spring/summer run evolutionarily significant units (ESUs).  We analyzed how individual fish characteristics and environmental conditions were related to adult migration survival from 2004-2015 for 5,062 Upper Columbia and 11,496 Snake River Chinook salmon.

We examined survival over two reaches within the hydrosystem: one in the Columbia River (Bonneville to McNary Dam for both ESUs) and one in the Snake River (Ice Harbor to Lower Granite Dam for Snake River populations only).  Based on a bimodal distribution of migration times, we separated Snake River populations into early- and late-migrating sets, differentiated as "spring run" and "summer run" adults in this report.

To identify the best predictors of survival and fallback over dams, we used a generalized linear modelling approach.  For the Snake River ESU, we initially tested whether run was a significant factor.  If it was, we analyzed spring and summer populations separately.  If not, we analyzed the Snake River ESU populations together.  We used the term stock to differentiate between 1) upper Columbia spring, 2) Snake River spring, and 3) Snake River summer populations.

Temperature had the most consistent influence on survival across all stocks through both reaches.  Temperature generally showed a quadratic relationship with survival in both Columbia and Snake River reaches.  Thus when the two reaches were combined for the Snake River ESU, survival from Bonneville to Lower Granite Dam varied from a low of 20% at temperatures over 20°C to a high of 80% at optimal temperatures (13-16°C).  The year of lowest annual survival for all stocks, 2015, was also the warmest year, with a mean temperature of 17.9°C during the summer run (65% survival from Bonneville to Lower Granite).

Survival from Bonneville to McNary also responded negatively to high spill.  The year of second lowest survival was 2011, when flows were 50% above normal.  These high flows likely reduced adult survival through the Columbia River reach.

Annual and seasonal variation in harvest significantly affected the survival of all stocks.  The year of third lowest survival was 2014, which had normal temperature and flow, but especially high catch during the Snake River summer Chinook migration (21% of the run at large).  An even higher catch rate (25%) exacerbated the impacts from temperature in 2015.  We found a significant interaction between catch and run, such that summer-run populations appeared to suffer more indirect effects of catch as well as higher catch rates in some years.

Fish characteristics important in some of the analyses were hatchery/wild origin, fish age, and a history of juvenile transportation.  However, impacts of these factors were less consistent than those from the primary factors of temperature, spill, and catch.

Survival through the Snake River reach from Ice Harbor to Lower Granite Dam was closely related to temperature and previous travel time in the hydrosystem (Bonneville to Ice Harbor Dam).

Fallback rates were highest and most variable at Lower Granite, followed by McNary and Bonneville Dam. Temperature was important at all dams, although the shape of the relationship varied. Cumulative temperature, which is a combination of travel time and temperature, consistently had a positive correlation with fallback.  Cumulative temperature was a better predictor of fallback rate than travel time alone at three dams.  Flow, spill and prior travel time were also important at several dams.

Managing natural variation in temperature and flow across the enormous Columbia River Basin while accommodating economic and social needs is extremely complex.  Logistical constraints and trade-offs make simplisti

Notes: Prepared for NMFS West Coast Regional Office
Theme: Recovery and rebuilding of marine and coastal species
Foci: Characterize the population biology of species, and develop and improve methods for predicting the status of populations.
Develop methods to use physiological, biological and behavioral information to predict population-level processes.