FISHERY BULLETIN: VOL. 85, NO. 3 



80 



70 



60- 



50- 



40- 



30- 



20 



70 



60 



50 



40 



o 



z 



30 



20 



10 



70 



60 



50 



40- 



30 



20 - 



10 



EARLIER PERIOD 



O MALE 



• FEMALE 



DIV 2J (1973 -78) 



LATER PERIOD 



(S 



DIV 2J (1979-83) 



DIV 3K (1970-78) 



DIV 3K (1979-83) 



• ® 



• ® 



DIV 3L (1968-78) 



DIV 3L (1979-83) 



0L__i ' I   i__i I I I I I I I I I I I I I I I I I I I I 1 I I — 1 — I — I — I — I — I — I — I — I — I — I — I — L 



2 4 6 8 10 12 14 16 18 20 22 24 2 4 6 8 10 12 14 16 18 



AGE (YRS) 



Figure 11. — A comparison between two time periods for growth curves of male and female witch flounder in NAFO Div. 2J, 3K, 



and 3L (circled points indicate only one observation). 



(using slopes of the curves as an expression of rate 

 of change of length with age) between the sexes 

 are similar up to about ages 10-12 after which 

 they diverge, particularly if the mean length at 

 age is used as a criterion. While the curves do not 

 appear to fit the mean data points very well in the 

 older ages (and to a much lesser degree the very 

 young ages), the correlation coefficients were all 

 high (Table 1) and all highly significant (P < 



0.001). The main reason for this is that the num- 

 ber of observations over age 15 is usually <1%. 

 Taking the 1968-78 males in Div. 3L as an exam- 

 ple, even adding a second term to the regression 

 (i.e., c (Log,. Age)'-^) increases the amount of ex- 

 plained variation by <3%. In nearly all cases the 

 mean data points in older ages are below the fit- 

 ted lines (Fig. 11) suggesting that if observations 

 were more numerous in the older ages, the actual 



622 



