about 1,100 fish per day for several days and flow 

 had increased to between 7,000 and 9,900 c.m.s. 



To understand better the relation of numbers of 

 chinook salmon counted at the dam and numbers 

 floating below the dam, it should be noted that 

 at a water temperature of 14.4° C, salmon car- 

 casses float in about 7 days after death (appendix 

 table 1). Because water temperature averaged 

 13.3° C. during the first 10 days of June 1955, 

 8 or 9 days would elapse between death and the 

 time carcasses first floated. The increase in floaters 

 observed after mid-June was an indicator that 

 after June 3, when total riverflow approximated 

 7,000 c.m.s., the numbers of chinook salmon dying 

 suddenly increased. We conclude on the basis of 

 these observations that mortality may be ex- 

 pected to be high whenever total riverflow rises 

 above 7,000 c.m.s., and low whenever total flow 

 is less than 5,100 c.m.s. 



In some years (1943, 1946, 1950, and 1954) float- 

 ing chinook salmon were recovered below the 

 dam when total flow was less than 5,100 c.m.s. 

 Although data for these periods are limited, the 

 relatively large numbers of floating chinook salm- 

 on in September 1943 and 1950 probably repre- 

 sented a low rate of mortality because very large 

 numbers of chinook salmon were migrating 

 through the area. The search by A. Suomela in 

 1943 (discussed in an earlier section), which re- 

 sulted in the recovery of a large number of car- 

 casses, is not comparable with later searches 

 because he searched in midstream, on shore, in 

 eddies, and any place that fish could be expected 

 to accumulate, whereas in our later searches, 

 floaters were observed only from fixed locations. 

 Fall chinook salmon runs in these 4 years of ap- 

 parently high mortality were large — peak daily 

 counts at Bonneville Dam were 20,000 to 30,000 

 chinook salmon. Water temperatures were about 

 21.1° C, and riverflows were low. Under these 

 conditions, floating chinook salmon were more 

 apparent because they floated in a shorter time, 

 thereby reducing their dispersion downstream 

 from the area of death both before and after they 

 became buoyant. 



The many floaters below the dam in September 

 1954 may have represented a higher rate of loss 

 of chinook salmon than is usual in the fall. Daily 

 counts were relatively low; the peak was only 

 about 8,000. During the 75-percent period of 

 passage of the 1954 fall run, total riverflow was 



the highest in the 10-year period 1946-55, ranging 

 from about 4,700 to 5,500 c.m.s. This higher-than- 

 normal flow may have resulted in a somewhat 

 higher-than-average fall mortality rate. 



Because a total flow of 7,000 c.m.s. or greater 

 is associated with high mortality rates, figure 9 

 may be interpreted as follows: (1) The flow over 

 Bonneville Dam has never reached 7,000 c.m.s. 

 during the fall run of chinook salmon. Observa- 

 tions have substantiated that fall runs generally 

 have experienced relatively low mortality rates. 

 (2) Most of the summer chinook salmon run in 

 every year has been subjected to flows exceeding 

 7,000 c.m.s., which are associated with a high 

 level of mortality. (3) In 4 of the 10 years (1946, 

 1949, 1951, and 1952) the entire 75-percent periods 

 of spring runs were subjected to flows greater 

 .than 7,000 c.m.s., and in 2 of the remaining 6 

 years (1947 and 1948) flows exceeded 7,000 c.m.s. 

 for about two-thirds of the 75-percent period. 

 The remaining four spring runs (1950, 1953, 1954, 

 and 1955) passed the dam when flows were less 

 than 7,000 c.m.s. 



In summary, fall runs of chinook salmon were 

 never subjected to killing flows; spring runs were 

 exposed to killing flows in some years; and sum- 

 mer runs always encountered killing flows. 



WATER TEMPERATURE 



River water temperatures afifect the floating 

 qualities of carcasses and thereby make carcasses 

 more or less evident to obser-vers. Warm water 

 makes carcasses more apparent and cold water 

 makes them less apparent. 



Flotation experiments described in the ap- 

 pendix showed that carcasses require more time 

 to float in cold water than in warm. As a result, 

 in the interval between death and floating, car- 

 casses may be swept farther downstream during 

 periods of low water temperatures than during 

 periods of high temperatures. Furthermore, dur- 

 ing the longer interval between death and float- 

 ing in cold water, scavengers have more oppor- 

 tunity to consume carcasses. 



With this in mind, we examined the relation 

 of water temperatures and numbers of chinook 

 salmon counted at the dam. In 1946-55 the range 

 of temperatures during spring runs (April and 

 May) was 7.8° to 13.9° C; during summer runs 

 (June and July), 10.0° to 19.4° C; and during 

 fall runs (August and September), 17.8° to 21.1° 



484 



U.S. FISH AND WILDLIFE SERVICE 



