CRONE and BOND: LIFE HISTORY OF COHO SALMON 



males near each female and the number of males 

 on the riffles but not with females also were 

 recorded. 



The estimates of the total number of spawners 

 based on the periodic counts on the spawning 

 riffles were obtained in the following manner. The 

 counts of both sexes were plotted against time 

 (Figure 3). In the figure each point for 1965 and 

 1967 represents the average of three successive 

 daily counts of spawners, and each point for 1963 

 and 1964 represents a daily count. A curve was 

 drawn by eye through each set of points, and the 

 resulting area under the curve represents the 



5 10 



NOVEMBER 



SURVEY DATE 



Figure 3.-Stream survey estimates of numbers of spawning 

 echo salmon in Sashin Creek, based on periodic counts on 

 spawning riffles, 1963-65 and 1967. Area under each curve is 

 spawning effort, expressed in fish-days. 



spawning effort in fish-days (see Table 3 and 

 section on Redd Life). The estimates of the total 

 number of spawners were then derived by divid- 

 ing the total number of fish-days by the average 

 redd life (the number of days a female spends on 

 the spawning site or redd). The method was 

 modified from McNeil (1966). The average of the 

 mean redd life computed for coho salmon females 

 in 1965 and 1967 was used to calculate total 

 number of spawners in 1963 and 1964. 



As indicated in the following tabulation, es- 

 timates of the total number of spawners derived 

 from the stream survey data were much higher 

 than the counts at the weir, except for 1965. 



We also estimated the size of the escapement in 

 1965 and 1967 by conducting a mark-recapture 

 experiment using the Bailey modification of the 

 Petersen formula as given by Ricker (1958). In 

 1965, 46 adult coho salmon (32 females and 14 

 males) were tagged before spawning; in 1967, 73 

 unspawned coho salmon (28 females and 45 males) 

 were tagged. The tags used were plastic Petersen 

 disks. Marked-to-unmarked ratios were obtained 

 from observations made during the visual cen- 

 suses and these were used to estimate the popula- 

 tions. Based on the marked-to-unmarked ratios, 

 the estimated number of coho salmoh spawners 

 (both sexes) in Sashin Creek was 221 in 1965 and 

 370 in 1967 (Table 4). 



The estimates of escapement size in 1965 and 

 1967 based on marked-to-unmarked ratios were 

 much higher than either the counts at the weir or 

 the estimates based on spawning effort and redd 

 life (Table 4). Several possible sources of error 

 existed in estimating numbers of spawners from 

 spawning effort and redd life: 1) The levels of 

 spawning activity were lower at low streamflows, 

 when visibility was good, and higher at high 

 streamflows (Figure 4), when visibility was re- 

 stricted. (In other words, the least accurate counts 

 of spawners occurred when the greatest numbers 

 were spawning.) 2) Some redds were occupied only 

 at night (indicated from our limited observations). 

 3) The assumption that the mean spawning life of 

 females was equal to that of males could be invalid. 



901 



