Evaluation of stream flow effects on smolt survival in the Yakima River basin, Washington

Annual report prepared for: Yakima Basin Joint Board, U.S. Bureau of Reclamation, System Operations Advisory Committee
By: , and 


Study Summary

The influence of stream flow on salmon smolt emigration survival is a topic of widespread management interest. We collected smolt survival data to inform flow management decisions in the Yakima Basin. The Yakima River watershed drains the eastern slopes of the Cascade Mountain Range in central Washington State. The upper basin is comprised of two major tributaries–the Naches River and the upper Yakima River. Headwater storage reservoirs capture runoff during the winter and spring seasons to support downstream irrigation needs. During summer months, water is conveyed through the upper Yakima River and diverted at Roza Dam, a major irrigation diversion that supplies water to the Roza Irrigation District and to a hydroelectric plant located near Yakima, Washington.

To assess smolt survival in the 18 km reach downstream of Roza Dam, a radio telemetry project will be carried out over a three-year timeframe. The first year of study was designed to provide baseline survival estimates at two distinct flow treatments during the spring migration period. The goal was to establish flow treatments that were as divergent as possible in order to maximize the observed effect of environmental conditions on smolt survival. In total, three experimental trials were carried out in 2012–one during low flow conditions (<600 cfs) and two during high flows (>3000 cfs). Data from the first year will be used to determine experimental design requirements to adequately address study objectives in years two and three.

In the spring of 2012, fixed telemetry monitoring stations were established in strategic locations upstream and downstream of Roza Dam. Yearling Chinook salmon Oncorhynchus tshawytscha smolts originating from Cle Elum Hatchery were captured at the Roza Dam fish screen bypass facility, implanted with radio tags, and released upstream of Roza Dam. Each release group of 50 fish was paired with a high or low flow condition. Fish movements were tracked as tagged fish passed each monitoring station during their migration down the upper Yakima River, through Roza Dam, past the Naches River confluence, and eventually through Sunnyside and Prosser Dams. At the conclusion of field data collection, survival rates for each release group were calculated using Cormack-Jolly-Seber mark-recapture models.

Yearling Chinook smolt survival and travel time estimates from 2012 suggest that migration rates and survival rates in the Roza Reach may be associated with stream flow, water temperature, release timing (i.e. migratory disposition), and fish size, but the extent to which each variable influenced survival is yet to be determined. The lowest survival rate (61%) and longest travel time (median 2.26 days) was observed in Release Group 1, which had the smallest size distribution and experienced the lowest flows, lowest temperatures, and earliest release date among the three groups. Release Groups 2 and 3 survived at 96% and 98% respectively and traveled through the Roza Reach in less than eight hours. The primary focus of years two and three of this study will be to collect data that minimizes the effect of confounding explanatory variables, so that flow effects on emigration survival can be quantified independent of these other influential factors.

Study Area

Additional publication details

Publication type Report
Publication Subtype Other Government Series
Title Evaluation of stream flow effects on smolt survival in the Yakima River basin, Washington
Year Published 2012
Language English
Publisher Cramer Fish Sciences
Contributing office(s) Western Fisheries Research Center
Description 31 p.
Country United States
State Washington
Other Geospatial Naches River, Roza Reach, Yakima River
Online Only (Y/N) N
Additional Online Files (Y/N) N
Google Analytic Metrics Metrics page
Additional metadata about this publication, not found in other parts of the page is in this table