734 



Fishery Bulletin 93(4), 1995 



The methods used by NMFS researchers to esti- 

 mate migration rates, timing, and survival were de- 

 tailed by Raymond (1979). In brief, unique batch 

 marks were applied by some combination of freeze 

 brands and fin clips to yearling (stream-type mi- 

 grants which were offspring of spring and summer 

 stocks) wild or hatchery chinook salmon collected at 

 hatcheries, from scoop traps and purse-seines, and/ 

 or at hydroelectric dams. The marked chinook salmon 

 were then released from the collection sites and re- 

 captured at downstream scoop-trap or purse-seine 

 sites, or from gatewells or collection facilities at dams. 



The estimated population of chinook salmon pass- 

 ing a capture site was derived from the formula N 

 = n/CE, where N - the estimate of the total num- 

 ber of chinook salmon passing (either for the un- 

 marked population as a whole or for specific mark 

 groups); n = the number of chinook salmon collected 

 (unmarked or marked); and CE = the collection effi- 

 ciency. Collection efficiency was determined from 

 separate groups of chinook salmon that were collected 

 semiweekly at each capture site from the unmarked 

 population offish that was passing each capture site 

 and subsequently marked uniquely for semiweekly 

 upstream releases at each capture site. Estimates of 

 collection efficiency for each site were derived from 

 the formula CE = (r/(R-10%R) x 100%), where CE = 

 the collection efficiency; R = the number of chinook 

 salmon marked and released upstream specifically 

 for collection efficiency estimates; and r = the num- 

 ber of chinook salmon recaptured from collection ef- 

 ficiency (R) releases. The number of chinook salmon 

 released was decreased by 10% to account for sus- 

 pected mortalities due to handling, marking, release 

 procedures, and migration between the upstream 

 release site back to the capture site. These methods 

 also assumed that any adverse effects of handling, 

 marking, and/or release procedures were equal for 

 all release groups (Raymond, 1979). 



Collection efficiency generally decreased as river 

 flow increased. At Ice Harbor Dam, collection effi- 

 ciency curves were fitted by regression techniques 

 to paired data sets of individual collection efficiency 

 estimates with the corresponding mean river flow 

 during the period the estimates were made (Ray- 

 mond, 1979). These curves were then used in future 

 years to predict collection efficiency under various 

 flows. At other capture sites, the data were consid- 

 ered too variable to develop reliable collection effi- 

 ciency curves. In these cases, real-time estimates of 

 collection efficiency were continually obtained dur- 

 ing the period when fish were captured at the site. 



The population estimate (AD was made in the fol- 

 lowing manner: if 15 marked chinook salmon of a 

 particular group or 2,000 unmarked chinook salmon 



were captured at a collection site during a 24-hour 

 period when the CE = 2%, then the estimated num- 

 ber of marked chinook salmon or the total number of 

 unmarked chinook salmon which passed the collec- 

 tion site during the period would have been 750 (15/ 

 0.02) or 100,000 (2,000/0.02). Total population esti- 

 mates were the sum of daily population estimates 

 over the period of time when a specific, marked group 

 of fish passed the site or over the period when the 

 unmarked population passed. 



Travel time of marked fish between two sites was 

 determined by subtracting the date of release from a 

 collection site (or the median passage date offish at 

 one capture site) from the median date of passage at 

 a downstream capture site. Migration rates of fish 

 were determined by dividing the distance between 

 two sites by the travel-time estimate between the 

 two sites. Survival estimates were made by dividing 

 the population estimate at a downstream capture site 

 by either the number offish released at an upstream 

 collection site or the population estimate at a cap- 

 ture site. 



Nearly all fish used for marking in the first study 

 years were products of natural spawning. The per- 

 centage of hatchery chinook salmon varied with 

 hatchery output each year, but 100% of the Snake 

 River stock was wild before 1966. Raymond (1988) 

 estimated that from 1966 to 1969 hatchery fish rep- 

 resented about 15% of the chinook salmon migration 

 that reached the upper dam (Ice Harbor Dam, 1966- 

 68; Lower Monumental Dam, 1969) on the lower 

 Snake River. According to his estimates, this percent- 

 age increased to 45-55% from 1970 to 1976 (the up- 

 per dam was Little Goose Dam, 1970-74; Lower 

 Granite Dam, 1975-present) and averaged greater 

 than 80% from 1981 to 1984. 



The 1973-79 NMFS yearly point estimates of sur- 

 vival (Fig. 2) (1973-75 in Raymond [1979]; 1976-79 

 [see Footnote 2]) were used by NMFS researchers 3 

 to indicate the effects of the recently completed Snake 

 River hydropower dams on juvenile fish survival. 

 Particularly low juvenile fish survivals were observed 

 under the low-flow conditions in 1973 and 1977, 

 whereas survival estimates did not vary much un- 

 der a broad range of higher flows. In the early 1990's, 

 computer models were developed by the Northwest 

 Power Planning Council, state fishery agencies, and 

 tribes in the Pacific Northwest to predict survivals 

 of juvenile fish migrating from Lower Granite Dam 



3 Sims, C. W., and F. J. Ossiander. 1981. Migrations of juve- 

 nile chinook salmon and steelhead trout in the Snake River 

 from 1973 to 1979, a research summary. Final Report to U.S. 

 Army Corps of Engineering, Portland, OR, 31 p. Northwest 

 Fish. Sci. Cent., NMFS. 



