salmon floaters were seen from the bridge, but 

 five were found in 1 hour and 50 minutes at sta- 

 tion 5 below the dam. Furthermore, during the 

 period of our study in both 1954 and 1955, 

 greater numbers of floaters were usually found in 

 the 48 km. of river below Bonneville Dam than 

 in any other section of river. No adjustment in the 

 calculated mortality appears necessary for re- 

 cruitment of untagged floaters into the sample 

 area from above the dam. 



The 31 tagged floaters used in the mortality 

 estimates were recovered by our search crews 

 below the dam from July 8 to 17 (table 4). The 

 first tagged floater was found by a fisherman on 

 July 7, the 7th day after the first date carcasses 

 were released (June 30); but it, as well as three 

 other carcasses found by fishermen, was not used 

 in computing the mortality estimate. Our search- 

 ers actually recovered 32 tagged floaters, but one 

 was eliminated from the computations because 

 it was on shore. 



No tagged floaters were recovered by search 

 crews after July 17, although sampling continued 

 through July 22 (table 4). Because no tagged 

 fish were recovered before July 7 or after July 17, 

 we assume that tagged carcasses were available 

 for recovery only during the period July 7 to 17. 

 During this 11-day period, 117 untagged chinook 

 salmon floaters were recovered at search stations. 



The tagged chinook salmon floaters were re- 

 covered from 7 to 16 days after they were released 

 (table 5). Because tagged and untagged carcasses 

 behave in a similar fashion, the smoothed daily 

 percentages of total tags recovered given in table 

 5 can be interpreted to mean that 21.99 percent 

 of the recoverable carcasses of chinook salmon 

 dying on a given day will have a chance of float- 

 ing and becoming available for recovery on the 

 7th day after death (of this 21.99 percent, some 

 will actually float and others will not); 19.87 per- 

 cent will have a chance of becoming available 

 for recovery on the 8th day; 15.71 percent on the 

 9th day; and so forth. Note that an additional 

 assumption is being made here: An individual 

 floating carcass is available for recovery only on 

 the day that it floats because it drifts out of the 

 recovery area in less than 1 day. This assumption 

 is supported by experiments in which chinook 

 salmon floaters were tagged and later recovered 

 (see appendix). These experiments showed that 



476 



floaters rapidly passed through the entire recovery 

 area within a few hours during high riverflows. 



Table 5. — Number and percentage of St tagged chinook 

 salmon floaters recovered below Bonneville Dam on each 

 day after release, 1955 



Tagged floaters Total tagged floaters recovered 

 Days after release recovered on each day after release 



Percent smoothed 

 Number Number Percent by threet 



7 4 12.90 21.99 



8 ' 10 32. 26 19. 87 



9 16 16. 13 15. 71 

 10 0.00 10.46 



n 6 16.13 9 43 



12 4 12.90 10.46 



13 1 3.23 6.28 



14 1 3.23 2.10 



16 .. 0.00 2.10 



16 1 3.23 1.67 



' Two recoveries on July 9 were arbitrarily assigned to releases of June 30 

 and July 1 because the release dates were iinknown. 



Knowledge of the estimated percentage of re- 

 coverable carcasses that have a chance of float- 

 ing and becoming available for recovery any 

 given number of days after death (table 5) 

 enables us to estimate the percentage of recover- 

 able carcasses of fish dying on any given date 

 that have a chance of floating and becoming 

 available for recovery at some time during the 

 July 7 to 17 recovery period. The method for 

 calculation of these estimates is shown in figure 7 

 which, together with table 4, indicates the struc- 

 ture of the 1955 experiment. The figure shows 

 that, of the recoverable carcasses of chinook 

 salmon dying on June 21, an estimated 1.57 

 percent were recoverable 16 days later on July 7, 

 the first day of the recovery period. Similarly, 

 considering recoverable carcasses from mortalities 

 on June 22, 2.10 percent were recoverable 15 

 days later on July 7 and 1.57 percent 16 days 

 later on July 8; consequently, a total of 3.67 

 percent of the carcasses from June 22 mortalities 

 were recoverable during the recovery period. 

 Computations for other days through July 10 

 follow the same pattern. 



We are now ready to estimate Yi, the number 

 of untagged recoverable carcasses in the river 

 during the July 7 to 17 recovery period. As has 

 been noted, of 1,169 carcasses tagged and re- 

 leased on June 30 and July 1, 31 were recovered 

 floating during the recovery period. At the same 

 time, 117 untagged floating chinook salmon car- 

 casses were observed (table 4). Assuming that 

 all tagged carcasses are recoverable, R, = 1,169, 



U.S. FISH AND WILDLIFE SERVICE 



