FISHERY BULLETIN: VOL. 75, NO. 4 



TABLE 2. — Mean back-calculated total length ± 1 standard deviation and annual 

 percent increase in mean total length for male summer flounder captured in Dela- 

 ware Bay during 1966-68. Included for comparison are mean back-calculated 

 lengths from other studies. 



'Lengths given for Eldridge at the end of year 1 and 2 are estimates of the average observed 

 length frequency. 



TABLE 3. — Mean back-calculated total length ± 1 standard deviation and annual percent increase in 

 mean total length for female summer flounder captured in Delaware Bay during 1966-68. Included for 

 comparison are mean back-calculated lengths from other studies. 



'Lengths given for Eldridge at the end of year 1 and 2 are estimates of the average observed length frequency. 



formed at the end of the first year. Eldridge de- 

 cided that Poole's calculated length at 1 yr seemed 

 too high when compared with observed length fre- 

 quencies, so he considered this first well-defined 

 annulus to be formed at first spawning, or at the 

 end of the flounder's third year. We considered the 

 first well-defined annulus to be formed at age 2. 

 Therefore, Poole's age 1 fish = our age 2 fish = 

 Eldridge's age 3 fish. Work by Richards (1970) 

 supported our age interpretation. He found sum- 

 mer flounder growth curves generated by analog 

 simulation only fit Poole's length data when 

 Poole's age-groups were shifted 1 yr forward, i.e., 

 his age 1 fish were made age 2. Richards did not 

 examine Eldridge's age data. 



Comparing Poole's (1961) lengths to ours after 

 adjustment for age interpretation, we find them 

 similar except for age 5 females. With age in- 

 terpretation adjustment, Eldridge's (1962) 

 lengths for males are smaller than ours except at 

 ages 2 and 3 when they are larger, and his lengths 

 for females are noticeable larger until age 5 when 

 they begin to agree quite well. 



The length-frequency distribution of the 1966 

 commercial catch and the 1966-71 research catch 



826 



revealed that both were primarily composed of 

 age-groups 2 through 5. Figure 3, using the 1966 

 and 1968 research catch because lengths were by 

 sex, is representative of this distribution. This age 

 composition is similar to the age composition re- 

 ported by Poole (1961) for the sport fishery catch of 

 Great South Bay, N.Y., after adjustment is made 

 for age interpretation differences. 



Equations representing growth rates from Wal- 

 ford's growth transformation (Rounsefell and 

 Everhart 1953) are: 



for males L t+1 = 141.91 + 0.767 (L t ) 



Correlation coefficient = 0.996 

 Standard error of estimate = 7.39 



for females L t+1 = 136.72 + 0.843(L,) 



Correlation coefficient = 0.998 

 Standard error of estimate = 6.20 



where L t +1 = fish length (millimeters) at time t 

 plus 1 yr 

 L t = fish length (millimeters) at time t. 



