The 1983 data were examined for potential sources of bias or error 

 in the abundance estimates by comparing proportions of marked and 

 recaptured fish by sex, length, and station. More females were marked 

 (n=3,127; 60%) and recaptured (n=247; 68%) than males (2,069, 40%; 116, 

 32%) . The increase in proportion of females recaptured in comparison to 

 those tagged was significantly different, although no plausible 

 explanation can be offered for this phenomenon. One possible reason was 

 that males left the river at a greater rate than females. However, 

 other data seemed to indicate the opposite effect as 54% of the returns 

 of disc-tagged males were from the river while only 39% of the females 

 were. Another possibility was that after spawning, when emission of 

 eggs and milt ceased, more males than females were misclassified by sex. 

 For instance, a change of 10 females to males (3% of the recaptures) 

 would have made the difference non-significant. Errors in 

 classification by sex did not affect the abundance estimate since the 



15 



