The difference between fall chinook salmon 

 counts'' at Bonneville and The Dalles Dams is an 

 estimate of the number of fish removed by the 

 fishery and spawninfi' (aside from mortalities anil 

 passage through navigation locks where they are 

 not counted) between the two dams." 



By subtracting the number of returns to the four 

 selected hatcheries and the numl)er removed l)y the 

 fishery from this estimate, we obtain a maximum 

 estimate of the fish that spawned between the two 

 dams at sites other than the four hatcheries — these 

 are the fish that did not stray into the four selected 

 hatcheries. The estimate is maximum because it 

 includes losses between the two dams and fish that 

 strayed from the four hatcheries. Finally, by ap- 

 plying age composition data and estimates of stray- 

 ing (from observed straying of fish marked at 

 Spring Creek Hatchery, table 15), we obtain esti- 

 mates of the munl)er of 1961-brood nonhatchery 

 fish spawned at the four selected hatcheries (table 

 16). 



An example of the information in table 16 fol- 

 lows. In 1963 an estimated 11,759 fall chinook 

 salmon spawned between the two dams but not at 

 the four subject hatcheries. An estimated 14 per- 

 cent, or 1,6-16, were 1961-brood fish. We assumed 

 that these fish are the (100-8.7) =91.3 percent of 

 the nonhatchery fish that did not stray into the 

 four hatcheries. The 8.7 was the observed percent- 

 age of the :2-year-()ld fish straying from Spring- 

 Creek Hatchery (table 15). It follows that (1,646/ 

 0.913) — 1,646=157 nonhatchery fish strayed into 

 the hatcheries. Similar calculations for 1964, 1965, 

 and 1966 give an estimated total of '2,126 non- 

 hatchery fish of the 1961 brood that were spawned 

 at the four selected hatcheries during the 4 years. 



Total return of marked ( Ad-RM and Ad-only) 

 and unmarked fish to the four iiatcheries was 1,898 

 and 26,538 fish, respectively (appendix table 5). 



» U.S. Arm.v. Corps of EngineiTS. 1963-66. Annual fish iiassage 

 report ; Xi.rtli Pacific Division, B(^nnevilIp. TIip Dalles, MeNar.v 

 and Ice Harbor Dams. Columhia and Snake Rivers. Oregon and 

 Wasliinston, ino.'i-66. U.S. Arm.v Eng. Dist.. Corps Ens.'.. Port- 

 land (Oreg.) and Walla Walla (Wash.) . Various pagination. 



" In an analysis of Hslnva.v counts (1957-65) at Bonneville ana 

 The Dalles Dams (among others). Fredd " found that the dif- 

 ferences in counts of fall chinook salmon generally exceeded 

 estimates of the numbers removed by the fishery and the numbers 

 spawned in intermediate areas. He concluded that counting errors 

 could not be a major contributor to the discrepancy, thus indicat- 

 ing a "loss" of tish between the two dams. 



" Fredd, Lcniis C. 1966. Analysis of differences in fish counts at 

 Columbia River dams. 1957-6.'), Fish Comm. Oreg. (U.S. .\rmy 

 Corps Eng, Portland Dist,. contract DA-.3.")-026-CIVENG-65-44), 

 47 pp. [Processed.] 



CONTRIBUTION OF COLUMBIA RIVER HATCHERIES 



Table 16. — Estimated number of nonhatchery chinook 

 salmon of 1961 brood spawned at four hatcheries ' between 

 Bonneville and The Dalles Dams, 1963-66 



Year of run 



Item Total 



1963 1964 1965 1966 



Count at Bonneville Dam 2 121,184 154,534 134,469 135,095 545,282 



Count at The Dalles Dam 2 66,473 56,160 87,096 69,018 278,737 



Hatchery returns ' ..-. -- 24,377 28,409 12,025 31,023 95,834 



CatcllS.- ---- -- 18.575 23,338 25.051 7,008 73,972 



Bonneville count minus The 



Dalles count minus hatchery 



return minus catch 11,759 46,637 10,297 28,046 96,739 



Percent 1961 brood'..- 14 78 66 9 



Number of 1961 brood year not 



straying --. .-- -- 1,646 36,377 6,796 2,524 47,343 



Percent straying s. . 8.7 4.3 4,7 



Estimated number of 1961- 



broo<l fish straying into four 



iiatcheries I 157 1,634 335 2,126 



' Spring Creek, Little White Salmon, Oxljow, and Cascade Hatcheries. 



- Counts for period August 26 to September 30. . 



3 Cliinook catcli between Bonneville and The Dalles Dams lor period 

 August 26 to September 30. Sport catch not included. 



• Estimated age composition from fishery samples. 



5 Estimates from Spring Creek mark (Ad-LV-RM), table 11. Assumed 

 tor age 5 because of the small number of returns. 



Subtracting the estimated number of nonhatchery 

 strays, 2,126, from the unmarked return, we found 

 the marked to unmarked ratio becomes 1,898/ 

 24,412 = .078. Estimated survival of marked rel- 

 ative to the unmarked tish is then (.078/.1193) 

 100=65.4 percent. 



This procedure for estimating the number of 

 nonhatchery fish entering the hatcheries is sub- 

 ject to error. If, for example, the percentage stray- 

 in<>- was as large as 10 percent (for each age 

 a-roup), then the estimated total of 1961-brood 

 nonhatchery Hsh that entered the four hatcheries 

 would lie 5,260: the estimate of the survival of 

 marked to unmarked fish would then be 74.6 i)er- 

 ceiit. In spite of the inexactness, however, it .seems 

 clear that the small value of the marked to un- 

 marked ratio observed in the hatchery returns rela- 

 tive to that in the hatchery releases cannot be 

 solely attributed to straying of nonhatchery fish. 

 Furtlierni(U-e, it appears that the total survival of 

 marked (Ad-R^I) relati\e to unmarked 1961- 

 brood-year fish could not ha\e exceeded 70 percent. 



In summary, it appears from the present data 

 that assumption 6 (eciuality of proportions marked 

 at each hatchery) is satisfied, but for the present 

 we must assume that marked and unmarked hatch- 

 ery fish have the same ocean distribution and ma- 

 turity schedules. In addition, we will assume that 

 total survival of marked fish was 70 percent of the 

 survival of unmarked fish— and further, that all 

 additional mortality of marked fish occurred dur- 

 ing their first year of life. 



TO FALL CHINOOK SALMON HARVEST 383 



