Project Title

Salmonid Watershed Analysis Model (SWAM)

Description

Habitat loss and alteration is the leading cause of species' declines worldwide, therefore, habitat restoration and protection is a prominent conservation strategy. In order to explore the relationship between habitat and populations, we have developed a modeling method for examining the relationship between GIS based habitat data and Pacific salmon (Oncorhynchus spp.) abundance at two very different spatial scales: stream reach and watershed.

SWAM is comprised of a series of spatial and statistical analyses that relate salmonid population counts (e.g., redd counts, adult counts, juvenile counts) at streams in a particular basin to habitat characteristics. SWAM identifies large-scale habitat features (anthropogenic and natural) correlated with fish abundance in a given subbasin. It provides a map of where the highest densities of fish in a particular basin are likely to occur, a series of ecological hypotheses about factors driving salmon abundances in a particular basin, and a list of important factors to consider when setting up monitoring projects or management experiments.

SWAM characterizes the relationship between habitat and salmon (Oncorhynchus spp.) populations in any given subbasin. The response variable is a time series of spawner counts collected at numerous reaches in a given subbasin. Predictive variables consist of habitat data characterized from geospatial data layers of land use type (i.e., grazing, water diversions, logging, mining, urbanization), landscape characteristics (i.e., geology, topography, vegetation), and climatic conditions (i.e., air temperature, precipitation). Consistent relationships between habitat and salmon abundance over time are used to predict relative salmonid densities in areas of the basin that lack abundance data.

As of October 2003, SWAM has been completed in the Salmon River basin in ID, the Snohomish, Yakima, and Wenatchee basins in WA, and the John Day and Willamette River basins in OR. Two summary documents have been put together to describe the SWAM approach and combined results from multiple basins; these are available below. The first is a magazine article that appeared in UPDATE magazine. The second is a report synthesizing several projects related to salmon recovery planning. The final results for the Snohomish and Salmon Rivers have been published; links to the manuscripts are provided below. Willamette River results are currently in review at the Canadian Journal of Fisheries and Aquatic Sciences. As soon as the manuscript is published, it will be posted on this site. Statistical results for the Wenatchee, John Day, and Yakima Rivers are described in an informal report also available below.

New analyses using ideas, results, and analysis components from the original SWAM papers are underway.

• Exploration and quantification of the effects of scale on the correlation between landscape condition and redd distribution across multiple basins.
• A synthesis of SWAM results across multiple basins to compare the impacts of climate, land cover, land-use, land form, and geology.
• An analysis of redd distribution across multiple basins using salmon returns at Bonneville as an index of population size.
• Use of metrics of landscape structure (e.g., connectivity, patch size, adjacency) as predictors of salmon redd distribution.
• Impacts of landscape condition on water quality in lowland agricultural streams.

River basins in which SWAM analyses are complete - Salmon River, Snohomish River and Wenatchee, John Day, and Yakima Rivers.

Summary Documents:

Steel, E.A., L. Johnston, B.E. Feist, G. Pess, R.E. Bilby, D. Jensen, T. Beechie, and J. Myers. 2003. Pacific salmon recovery planning and the Salmonid Watershed Analysis Model (SWAM): a broad-scale tool for assisting in the development of habitat recovery plans. Endangered Species Update. 20(1):1-32.

Summaries for the Salmon Habitat in Recovery Planning (SHRP) Document, The Chinook Salmon Life Cycle Model, and the Salmonid Watershed Analysis Model (SWAM)

Basin-specific Results:

Salmon River
Feist, B.E., E.A. Steel, G.R. Pess, and R.E. Bilby. 2003. The influence of scale on salmon habitat restoration priorities. Animal Conservation 6:271-282. Reprinted with permission from Animal Conservation, The Zoological Society of London, and Cambridge University Press

Snohomish River
Pess, G.R., D.R. Montgomery, R.E. Bilby, E.A. Steel, B.E. Feist, and H.M. Greenberg. 2002. Landscape characteristics and land use on coho salmon (Oncorhynchus kisutch) abundance, Snohomish River, Washington State, USA. Canadian Journal of Aquatic and Fisheries Science. 59:613-623.

Wenatchee, John Day, and Yakima Rivers
BPA Status Report, January 28, 2003

Willamette
Steel, E.A., B.E. Feist, D. Jensen, G.R. Pess, M. Sheer, J. Brauner, and R.E. Bilby. 2004. Landscape models to understand steelhead (Oncorhynchus mykiss) distribution and help prioritize barrier removals in the Willamette basin, OR, U.S.A. Can. J. Fish. Aquat. Sci. 61:999-1011.

Investigators

Blake Feist, Ashley Steel, George Pess, and Mindi Sheer (Conservation Biology Division)

Collaborators

Bob Bilby (Weyerhaeuser Company); Dave Montgomery and Harvey Greenberg (University of Washington), and Dave Jensen (Frank Orth Statistical Consultant), and Marta Danielsdottir (University of Washington)

Support

Bonneville Power Administration

Watershed Home  |  Landscape Ecology and Recovery Science Home