than the stated age. Conversely, error may have been introduced using the calculated lengths-at-age 

 because of variable timing of annulus formation. Also, if mostly faster-growing older fish were selected 

 for measurements of annuli on scales, as suggested previously, a larger K value may have resulted because 

 these specimens had a faster rate of growth. 



An independent assessment of growth was obtained from winter flounder marked with Petersen disc 

 tags, released in the study area, and later recaptured after various periods at large. The fish used in this 

 analysis included females from 201 to 406 mm and males from 205 to 398 mm at time of tagging. The 

 recaptured fish were caught after 90 to 1,065 d at large and growth ranged from 1 to 149 mm in females 

 and 1 to 85 ram in males. The data were fit to a two-parameter (K and Loo) von Bertalanify model 

 described by Fabens (1965) for fish of unknown age, but whose increase in length is known for varying 

 time periods. The estimates of Loo (423 mm for females; 375 for males) were similar to those described 

 by the 1983 calculated length-at-age data, but the K values (0.35; 0.31) were closer to the 1977-83 aging 

 data model. However, less confidence can be placed on the parameters determined from the tagged fish 

 due to variability in the data and smaller sample size. Growth of individuals also varied because of the 

 particular season of release or recapture and effects of tagging may have influenced these results. 



Mortality and survival 



Mortality and survival were calculated using age and length data from 1978-79 and 1981-83. Data 

 from 1977 were not used because relatively few fish were aged and from 1980 because of previously 

 mentioned inconsistencies of data from that survey. Both methods of estimation used were time-specific 

 (Ricker 1975). Apparent large variability in estimated abundance of individuals of specific year-classes, 

 most likely due to changes in survey methodology, sample selection, and sampling error, made cohort-specific 

 methods of estimating survival less reliable or simply not possible (NUSCo 1984). The time-specific 

 method of Robson and Chapman (1961) has cin advantage in that the age determinations of older fish do 

 not have to be known with certainty, although the representativeness of the youngest age used is very 

 important (Ricker 1975). As noted previously, relative abundance of small winter flounder (ages 1 and 2) 

 in the population age structure were not accurately measured. Therefore, only ages 3 and older were used 

 in the calculations. 



49 



