Category: Fish/Shellfish Research
Published: February 2003
Publication number: FPA 02-07
Author(s): Dave Seiler, Greg Volkhardt, and Lori Kishimoto
The Puget Sound Chinook Evolutionary Significant Unit (ESU) was listed under the Endangered Species Act as a threatened species in March 1999 by the National Marine Fisheries Service. The ESU includes 22 populations of chinook salmon, two of which are located in the Lake Washington basin (Ruckelshaus et al. , in press). The North Lake Washington population includes tributaries to the Sammamish River, including Bear Creek and Issaquah Creek. In addition to wild chinook production, an artificial production program (Issaquah Hatchery) exists on Issaquah Creek, and releases approximately 2 million fingerling fall chinook each year. A second population of chinook salmon has been identified in the Cedar River, a tributary to the southern end of Lake Washington. Analysis of genetic data have shown that the Cedar River chinook population is genetically divergent from the North Lake Washington population, and that chinook salmon sampled from Bear Creek and Issaquah Creek are genetically similar (Marshall 2000).
Anticipating this listing, land, waters and fish managers in city, county, state, tribal, and federal government agencies began discussing and planning appropriate responses. In the Lake Washington watershed, it was evident that these planning efforts would be more effective if more were known about the habitat requirements, early life history, freshwater productivity, and survival of chinook salmon. Baseline information was available on the number of spawners, but such counts provide little insight into survival during specific life stages. Estimating the number of juvenile migrants facilitates separating survival into two components, egg-to-migrant (survival during the freshwater stage) and migrant-to-returning adult (primarily survival during the marine stage). This provides a more direct accounting of the role of freshwater habitat in regulating population abundance, and an improved understanding of density independent and density dependent factors affecting the production of migrants (Seiler et al. 1981, Fuerstenberg and Luchetti 1998, Cramer et al. 1999).
In 1999, the Washington Department of Fish and Wildlifeâ€™s Wild Salmon Production Evaluation Unit (WSPE) built upon an existing sockeye fry monitoring program in the Cedar River and Sammamish watershed to assess natural chinook production from the Cedar River and Bear Creek, a tributary of the Sammamish River. Since unmarked samples of juvenile or adult chinook taken in the Lake Washington system would also include naturally produced chinook from Issaquah Creek, we deemed it important to also assess natural juvenile chinook production from this stream. Therefore, we conducted a one-year assessment of natural chinook production from Issaquah Creek in 2000.
To accomplish these evaluations, floating downstream juvenile migrant fish traps were deployed to sample the juvenile salmonids migrating from each stream. Traps were operated from January or February to July on the Cedar River and Bear Creek in order to assess all or nearly all of the chinook migration period. The Issaquah Creek trap was operated from mid-March to July.
This report describes the results of our first and second-year investigations of wild chinook production in the Cedar River, Bear Creek, and Issaquah Creek watersheds within the Lake Washington Basin. It also describes the freshwater production of other anadromous salmonids in these systems, except sockeye fry production from the Cedar River during these two years (Seiler et al. 2001a , Seiler et al. 2001b ).
In each year since in 1992 we have operated an inclined-plane screen (scoop) trap at river mile (RM) 0.7 and 1.0, respectively. Trapping with this gear began each year in late January. In 1999, the scoop trap was used until May and until April in 2000. In March of 1999 and April of 2000, we installed a screw trap at RM 1.0 which was operated until trapping ceased in July. Each trap has advantages and disadvantages. The scoop trap was used early in the season since it was designed to sampled the small chinook and sockeye fry which migrate during that period. Predation of small fry in the trap live box is reduced with this trap since few larger predatory fish are captured using this trap design. The trap also incorporated a removable well that facilitated handling chinook and sockeye fry. The screw trap was used later in the spring because it better captures the larger salmonids that are migrating at that time.
Estimates of juvenile production were developed from trap catches by estimating the proportion of downstream migrants that were captured in the trap (capture rate or trap efficiency). These estimates were made by marking and releasing groups of captured age 0+ chinook, coho smolts, and sockeye fry above the trap. Trap efficiency was estimated by the recaptured proportion of the marked groups. Trap efficiency data was evaluated to determine if flow or other factors influenced trap efficiency.
Age 0+ chinook production from the Cedar River was estimated at 81,000 and 65,000 migrants in 1999 and 2000. The migration timing was bi-modal in both years with a substantial migration of newly emerged fry occurring from January to mid-April and a smaller migration of smolted chinook occurring from mid-April until July. The fry component was estimated at 67,000 in 1999 and 46,000 in 2000, whereas the smolt component ranged from 14,000 in 1999 to 19,000 in 2000. We believe the early â€�"fryâ€ component resulted from the downstream displacement of juveniles due to stream velocity and/or rearing density. Late winter flows in 2000 were lower than the 1999 levels which we believe resulted in the displacement of fewer fry downstream and provided more low velocity habitat for rearing to smolt size. Production of coho, steelhead, and anadromous cutthroat was also estimated in each year (see table below).
Juvenile downstream migrant production estimates for the Cedar River, 1999 and 2000.
In 1999, chinook fry ranged in size from 35-mm to 54-mm during the fry migration period and from 37-mm to 131-mm during the smolt migration period. Similar size ranges were observed in 2000 except that chinook smolts as large as 153-mm were captured that year.
A screw trap was operated at river mile one on Big Bear Creek between February 24 and July 13 in 1999 and from January 24 to July 13 in 2000. Using the approach described for the Cedar River, juvenile production was estimated for wild age 0+ chinook, coho, sockeye, steelhead, and cutthroat.
Natural production of age 0+ chinook was estimated at 15,000 in 1999 and 32,000 in 2000. Chinook migration timing was strongly bimodal in 2000, as was observed in the Cedar River, but was much less so in 1999. In 1999, most chinook migrated as smolts in May and June. Age 0+ chinook fork lengths were similar between years with newly emerged fry averaging between 35 and 40-mm and smolts averaging between 85 and 95-mm by mid-June.
Coho production was estimated at 63,000 in 1999 and 28,000 in 2000. In 1998, 166,000 coho fry were planted in Bear Creek. No releases occurred in 1999. Since these fish were not marked, we were unable to determine the extent, if any, that they may have contributed to the 1999 smolt production. Sockeye production was estimated at 1.5-million in 1999 and 190,000 in 2000. These differences are primarily related to parent-brood escapement levels. The cutthroat smolt migration past the Bear Creek trap was estimated at 3,400 in 1999 and 5,700 in 2000 during the period of trap operation. Steelhead production was fairly consistent between years with an estimated 1,800 smolts produced in 1999 and 2,000 smolts produced in 2000. Adipose marked smolts resulting from steelhead fry releases made up 77% and 82% of the 1999 and 2000 production, respectively.
In 2000, we assessed the stomach contents of cutthroat, steelhead, and coho smolts, and sculpins that were trapped. We found that a substantial number of sockeye fry and few chinook fry were consumed by these species. Sockeye fry were found in 61%, 24%, 42%, and 66% of the cutthroat, coho, steelhead, and sculpins sampled, respectively. When sockeye fry were found in a sample, the average number of sockeye consumed by cutthroat, coho, steelhead, and sculpins were 18, 4, 15, and 7, respectively. More sockeye fry were found to be consumed when fish were left in the trap live box for longer periods. This information was used to adjust the sockeye production estimate in 2000. Using this data, we estimated 3,546 sockeye fry and 92 chinook fry were consumed by cutthroat, steelhead, and coho captured in the Bear Creek screw trap.
A screw trap was installed on March 14, 2000 at river mile two on Issaquah Creek and operated until July 3, 2000. The trap was operated continuously during this period except for periods when large numbers of hatchery fish from the Issaquah Hatchery (RM 3.1) were passing the trap. These periods necessitated suspending trap operation approximately 30% of the time during the March 14 to July 3 trapping interval. Naturally-produced chinook, coho, steelhead, and cutthroat production was estimated from this trapping effort.
Naturally-produced chinook production was estimated at 30,000 during the March 14 to July 3 period. A substantial number of newly emerged chinook fry were migrating when trap operation began and we were unable to estimate, with any certainty, how many of these migrated prior to trap installation. Chinook size at time was similar to that found in the Cedar River and Bear Creek.
During the period of trap operation, we estimated 19,000 naturally-produced coho smolts, 1,100 wild steelhead smolts, and 15,000 wild cutthroat smolts migrated past the screw trap. Based on coho and steelhead migration timing, these estimates represented virtually all of the production of these two species in the basin. By assuming the same timing for Issaquah Creek cutthroat as was found for Bear Creek cutthroat, the total production of Issaquah cutthroat is estimated at 18,000 smolts.
Comparisons Within and Between Streams
We believe the bimodal migration timing exhibited by age 0+ chinook migrants in these watersheds results from a combination of conditions. In the Cedar River, more smolts (19,000 compared to 14,000) and fewer fry (46,000 compared to 67,000) were produced in 2000 relative to 1999. Since the magnitude of late winter stream flows were much less in 2000 compared to 1999, we believe fewer newly emerged fry were washed downstream into the lake. Furthermore, the lower flows provided additional low velocity habitat for rearing to smolt size that year. In addition, fewer cutthroat were present in the river in 2000 which may have also helped to increase survival to smolt size. In Bear Creek, substantially more fry (14,000 compared to 2,000) and smolts (18,000 compared to 13,000) were produced in 2000 relative to 1999. In this believe the higher escapement in 1999 (733 chinook compared to 401 in 1998) was able to more fully seed the habitat with the excess production migrating as fry in 2000. Chinook smolt production was similar between the three streams in 2000. The Cedar River, Bear Creek, and Issaquah Creek produced 18,800, 18,100, and 18,300 chinook smolts, respectively.
Chinook egg-to-migrant survival was estimated at 10.4% in 1999 and 8.0% in 2000 in the Cedar River, and 2.1% in 1999 to 2.4% in 2000 for Bear Creek. It is difficult to compare differences between the two streams since a larger proportion of Cedar River chinook leave the stream as fry. If only the ratio of smolts produced per eggs deposited is examined, smolt/egg productivity ranges from 1.8% in 1999 to 2.3% in 2000 for the Cedar River, 1.9% in 1999 and 1.4% in 2000 for Big Bear Creek, and 0.6% for Issaquah Creek. A strong density dependent effect is evident in these smolt/egg production rates since for each brood year, the highest rates were found where escapements were least and the lowest rates were found where escapements were highest.
Bear Creek sockeye egg-to-migrant survival was estimated at 11% for the 1998 brood and 7.4% for the 1999 brood. In comparing these survival rates with those from two years of trapping on the Sammamish River indicates that egg-to-migrant survival correlates well with the severity of peak flows during the egg incubation period. This relationship has also been observed for Cedar River sockeye (Seiler et al. 2001b ).