salmon (0. gorhischa), and Hunter (1959), who 

 worked witli pink salmon and chum salmon (0. kein), 

 coni-luded that entry into a river follows increases in 

 stream flow. Shapavalov and Taft (1954) noted a 

 correlation between the general periods of the 

 spawning runs of silver salmon (0. k-isiilch) antl 

 rainfall. They further believed, but were unable to 

 demonstrate ([uantitatively. that fish movement 

 increased with a rise of stream flow. .Mien (195(1), 

 on the other hand, related the movement of silver 

 salmon and chinook salmon (0. Ishawi/I.srha) to 

 nighttime rainfall and low barometric jjressure, 

 respectively. Ellis (1963) believed that the entry 

 of silver salmon and sockeye salmon into rivers was 

 associated with the appearance of atmosplieric 

 warm fronts over the estuary. 



Only Foerster (1929) and Cramer and Ilammack 

 (1952) attributed the sudtk'n movement of salmon 

 to changes in water temperature. Foerster, who 

 reported on sockeye salmon at Cultus T.ake, Rritish 

 Columbia, noted that the numbers of fish arriving 

 at a counting fence synchronized closely with tem- 

 perature change: increases in the daily nui accom- 

 panied declines in temperature. Cramer and Ham- 

 mack, who studied chinook salmon in Deer Creek, 

 a tributary to the Sacramento River, Calif., con- 

 cluded that at the close of a period of clear weather 

 and relatively cool water, sudden increases in water 

 temperature to 75° F. caused an upsurge of fish. 



Andrew and Cleen (19G0), in their ajjpraisal of all 

 available information on the possible effects of dam 

 construction on the Fraser River, British Columbia, 

 devoted considerable attention to the efTects of tem- 

 perature and delay on upstream-migrating salmon. 

 High water temperatures, they concluded, are detri- 

 mental to salmon in their upstream migration be- 

 cause they increase the rate of energy consumption 

 and the incidence of disease and parasites, and may 

 be directly lethal. 



In the matter of delay, Andrew and Geen (1900) 

 cited Thompson (1945) as having shown that a 

 delay of 12 days at Hell's Cate (before construction 

 of the fishways) was sufficient to prexent sockeye 

 salmon from reaching their spawning grounds, and 

 that lesser delays reduced the reproductive cajiacity 

 of the fish. The same authors also referred to an 

 incident in the Fraser Canyon at Vale, British 

 Columbia, in 1955 where the early run to the Stuart 

 River was blocked (J days by higli water. Of an 

 estimated .30,000 to 35,000 sockeye salmon, only 

 2,170 reached the spawning grounds. 



The effect of delay on the productivity of salmon 

 has been illustrated by studies of fish passage at a 

 rock slide on the Habine Rixcr, British Columbia 

 (Godfrey, Hourston, Stokes, and Withler, 1954). 

 Concerning this study, Andrew and Geen (19(i0) 

 stated: 



.... T;iKK''>H of ti'*'' liclow tlio pciiiit of diftu'ult passane and 

 rcrovory of the tagged fisli at h countiiiK fciu-c -!0 miles up- 

 stream showed tliat ."iome of tlie fisli delayed helow the ob- 

 struetion were able to migrate to their spawning grounds but 

 relatively few were able to spawn successfully. Heeause fish 

 were delayed and weakened below the obstruction they were 

 not able to migrate at a normal rate after passing the obstruc- 

 tion. The effective spawning in 1952, when some .sockeye 

 were delayed for extended periods, was estimated as :50 to 42 

 percent of the numbers of female sockeye that rea<'hed the 

 spawning grounds or 7 to 10 percent of the total escapement. 

 From 'M to 40 percent of tlie female sockeye examined on the 

 sp.'iwning ground died unspawned and others die<l after pass- 

 ing the obstruction but before reaching the spawning groimds. 



The previous references indicate that salmon that 

 have been delayed enroute to spawning groiuuls 

 can be affected in two ways. First, some die en- 

 route to the spawning grounds and second, some 

 complete the journey but are unable to spawn 

 stU'cessfully. We conclude that temiierature blocks 

 of the tyiK' outlined in this paper similarly afTect 

 sockeye salmon bound for the Okanogan River 

 spawning groiuuls. 



SUMMARY 



Sockeye .«almon were tagged at Rock Island Dam 

 antl later obscrxed at Zosel Dam on the Okanogan 

 Ri\er in 1953-54 and 19(>2-()3. These exiw-riments 

 were used to detect any changes in tl;e migration 

 time caused by Rocky Reach Dam, which was 

 constructed on the migration route diuing the 

 inter\-ening years. 



Travel time varied greatly, both between and 

 within years. The difTerence between 19()2 and 

 19()3 exc(>ede(l the difference between 1963 and 

 1953-54. The best estimate of the time required 

 for .sockeye salmon to migrate from Rock Island 

 Dam to Zosel Dam is 7 to 9 days, barring major 

 delay due to environment. This travel time has 

 not been increased by Rocky Reach Dam. 



^^'ate^ temperature of the lower Okanogan River, 

 or factors linked with it, is by far the greatest 

 sotnce of delay. Below 70° F.. entry into the 

 Okanogan River is relatively unimpeded. Above 

 70° F., relatively stable or rising temperatures 

 delay entry until a sharp droj) occurs. 



146 



U.S. FISH .\ND WILDLIFE SEKVICE 



