Category: Fish/Shellfish Research
Published: June 2004
Author(s): Dave Seiler, Greg Volkhardt, and Lindsey Fleischer
Assessment of sockeye fry production began in the Sammamish system in 1997. We placed the trap in the Sammamish River at Bothell where we also operated it during the 1998 season. In 1999, to assess chinook production as well as sockeye, we moved this monitoring program to Bear Creek. Since 1999, as in the Cedar River, this trapping operation has also estimated the populations of coho, steelhead and cutthroat smolts.
The 2001 trapping season was notable for two unusual events. The interval from Fall 2000 through March of 2001 has become known as the â€�"winter without rainâ€. Flows throughout this period were anomalously low as a result. The second anomaly, an earthquake measuring 6.8 on the Richter scale occurred on the morning of February 28. This quake, which was centered in the lower Nisqually Basin, was strong enough to trigger a landslide that temporarily blocked the Cedar River at River Mile 8.
Declining adult sockeye salmon returns in the late 1980's and early 1990's prompted the creation of a multi-agency effort to investigate causes for this decline. To determine which life-stages of sockeye were experiencing poor survival, an evaluation of fry production was undertaken in the Cedar River beginning in 1992. Assessing the sockeye population at this location and life-stage separates freshwater production into river and lake components. This report documents our evaluation during 2001, the tenth year of this project. As in previous years, the primary study goal was to estimate the season total migration of Cedar River wild and hatchery sockeye fry into Lake Washington. These estimates enable calculation of survival rates from egg deposition to lake entry, for hatchery fry from release to the trap, and for both production components from lake entry to subsequent life stages of smolts and adults.
Beginning in January and continuing into June, a floating inclined-plane screen trap located at River Mile (R.M.) 0.7 in the Cedar River was operated to capture a portion of the sockeye fry migrating into Lake Washington (Figure 1). To estimate the capture efficiency of this trap, on 69 nights, dyemarked fry were released upstream of the trap. At base flows, 350 to 400 cfs, capture rates averaged 10%. At the highest flows (800 cfs) the capture rate averaged 4.6%. Stream flows were anomalously low and steady through most of the season, and capture rate varied little compared to previous seasons.
Over the season, 17.2 million hatchery sockeye fry were released into the Cedar River from three locations. All hatchery fry were internally marked by slightly manipulating water temperatures in the hatchery. On most nights of and following hatchery releases, fry caught in the trap were randomly sampled for thermal marks to determine the proportion of hatchery fish present.
Over the 115 nights trapped, 4.0 million sockeye fry were captured. From this catch and the capture efficiency data, we estimated a total of 52.0 million wild and hatchery sockeye fry entered Lake Washington in 2001. Based on otolith analysis and the hatchery release figures, we estimated that this total included 38.5 million wild fry and 13.5 million hatchery produced fry. Average survival to the trap of the 8.4 million hatchery fry released upstream was estimated at 56.5%. Survival was a function of migration distance. Survival of fry released at the Landsburg Hatchery, located 21 miles upstream, averaged 26.3%. Fry released at the Riviera site, located 1.5 miles above the trap survived at an average rate of 75.3%, nearly three times higher. We attribute this difference to the low flows throughout the season, which enabled high predation rates.
Migration timing for wild fry was earlier than in any of the previous nine years. This timing was also 23 days earlier than that predicted by the relationship between timing and February temperature units developed over the previous nine brood years. We attribute this discrepancy to two factors; higher predation rates later in the season as a result of the low flows, and mortality resulting from the 6.8 magnitude earthquake on February 28. This quake triggered a river-blocking landslide at R.M. 8. When flow was restored a short time later, a large quantity of mud was transported down the river, which likely smothered eggs and alevins in the lower river.
Survival from egg deposition to lake entry of wild fry was estimated at 11.3%. This rate is the ratio of 38.5 million wild fry to an estimated deposition of 339 million eggs. Survival of the 2000 brood was the third highest measured thus far, but less than we expected given the low and steady incubation flows. With the peak incubation flow of just 627 cfs, the relationship between peak incubation flow and egg to migrant survival developed over the previous nine broods predicted a survival of 13%. As with migration timing, we attribute the lower survival of eggs and fry to a combination of high predation rates resulting from the anomalously low flows throughout the migration and mortality caused by the earthquake-triggered landslide.
In response to the listing of the Puget Sound Chinook Evolutionary Significant Unit (USE) under the Endangered Species Act as a threatened species, the existing sockeye fry monitoring program in 1999 was expanded to include an assessment of the natural chinook production in the Cedar River. The gear we operate each year starting in January to assess sockeye fry production also captures chinook fry. To capture the larger, later migrating chinook, which we classify as â€�"smoltsâ€, we installed a screw trap at R.M. 1.1, and operated it until July.
Juvenile production was estimated through applying capture rate estimates to catch data. From the start of the season in January through March, we used the capture rate data generated with releases of marked sockeye fry to estimate the migration of chinook fry. Screw trap efficiency was estimated by releasing groups of fin- marked chinook smolts above the trap.
Age 0+ chinook production from the Cedar River was estimated at 32,249 in 2001. Timing was bimodal with smolts emigrating in May and June comprising two thirds (21,400) of the total migration. The fry migration, from January through March, was estimated at 10,800. Egg-to-migrant survival was estimated at 13.5%. We believe that the low flows during the 2001 season allowed a higher than normal proportion of fry to remain in the river longer and grow to smolts before migrating downstream. In comparison, fry have accounted for the majority of the migration in the two previous seasons. Over the season, age 0+ chinook increased in size from less than 40 mm in January to over 100 mm by July.
Over the season, based on actual and projected catches and estimates of capture rates we estimated the migrations of coho, steelhead and cutthroat smolts at 82,462, 1,860 and 2,680, respectively. Figure 1. Site map of the lower Cedar River watershed depicting the fry and screw trap locations, hatchery sockeye release sites, and trap efficiency test release sites for the 2001 trapping season.
We installed a scoop trap on Big Bear Creek 100 yards downstream of the Redmond Way Bridge and operated it from January 27 through April 9. On April 10, we replaced it with a screw trap that fished until July 12. Using the approach described for the Cedar River, downstream migrant production was estimated for wild sockeye fry, age 0+ chinook, coho, steelhead, and cutthroat smolts.
Applying the average scoop and screw trap capture rates of 15% and 21% to respective catches estimated sockeye production at 2.2 million fry.
Production of age 0+ chinook was estimated at 10,588. Migration timing was bi-modal, however most chinook migrated as smolts in May and June.
For the season, we estimated the production of coho, wild steelhead and cutthroat smolts at 21,665, 139 and 2,869, respectively.