Influence of Natural and Anthropogenic Disturbances on Spawning Sockeye Salmon Distributions in the Cedar River, Washington

Abstract

Human management of river ecosystems dampens natural disturbance regimes that drive critical habitat-forming processes. However, the quantification of processes can be difficult and is typically accomplished through surrogate analyses. In 2001, a channel-damming landslide in the Cedar River, Washington, changed habitat conditions and spawning distributions of Sockeye Salmon Oncorhynchus nerka. We evaluated annual changes in a number of key indicators of river processes during the pre- and post-disturbance periods. We used an exhaustive chi-square automatic interaction detection (CHAID) model to evaluate the effect of changes in habitat complexity, discharge characteristics, and parental generation spawning distributions on 54 years (1952–2006) of weekly spawning fish surveys. Before the landslide, the spatial and temporal patterns of Sockeye Salmon distributions were heavily influenced by the spawning locations of their parents. Densities of spawning fish were highest in the uppermost river reaches. After the landslide, habitat complexity and fish densities increased in the lower, landslide-affected river reaches. The CHAID model analyses indicated that fluvial habitat complexity, river discharge characteristics, and the density of fish in respective parental generations were significant factors explaining the spawning distributions by reach. Our analyses of the salmon population response to annual changes in habitat complexity affirm that our metrics were useful surrogates for measuring the habitat-forming processes restored when the landslide reconnected the river channel with its floodplain. Furthermore, this study provides an “order of magnitude” insight into the amount of restoration that will elicit a population-scale response in salmon.

Received August 27, 2013; accepted January 21, 2014

ACKNOWLEDGMENTS

This work was funded by the U.S. Environmental Protection Agency's Science to Achieve Results program (Grant Number R827149-01-0), the H. Mason Keeler Endowment for Excellence, and the Galen and Helen Maxfield Fisheries Scholarship. We are grateful to Steve Foley (deceased) of WDFW and to Hans Berge and Terry Butler of the King County Department Natural Resources and Parks for providing data and critical insights. We also thank Dave Beauchamp and Dave Montgomery for their critical reviews. In addition, Mason Bryant, Jonathan Frodge, Michele Koehler, Laura Rattner, and Kurt Fausch suggested improvements to earlier versions of this manuscript. Finally, we thank three anonymous reviewers for their helpful suggestions.