demonstrated the validity of the use of otoliths for 

 the age determination of yellowtail flounder. 



Eggs were separated from the ovarian tissue by 

 washing with a gentle stream of water through a 

 series of four fine mesh screens (mesh sizes 1.52, 

 0.98, 0.51, 0.14 mm). After separation the eggs 

 were placed in a gallon jar and diluted with water 

 to 3,000 ml. Large samples were first divided using 

 a plankton splitter and only half of the sample 

 diluted. The lid of the gallon jar was modified to 

 hold a 1-ml Hensen-Stemple pipette which ex- 

 tended approximately 15 cm into the jar. The jar 

 was then inverted 10 times and the sample taken 

 before any settling of the eggs occurred. The sub- 

 sample was placed onto a gridded Petri dish and 

 the eggs counted with a dissecting microscope. A 

 minimum of three subsamples were counted for 

 each fish. The coefficient of variation was com- 

 puted and ranged from <1 to 18% (mean = 7.5%). 

 Fecundity \/as estimated by multiplying the mean 

 number of eggs from the subsamples by 3,000, or 

 6,000 if the sample had been split. 



Results and Discussion 



Linear regressions, correlation coefficients (r), 

 and coefficients of determination (r 2 ) were com- 

 puted from data transformed to common 

 logarithms. These were: 



F = 0.986L 3858 (Figure 1! 



r = 0.885, r 2 = 0.784 



(1) 



and fecundity vs. age (t = 4.84, df = 47, PO.001). 

 Gonad weight, therefore, contributed most to the 

 variation in fecundity and would be the best 

 parameter to measure in estimating fecundity. 

 However, since the relationship between ovary 

 weight and fecundity varies seasonally, depend- 

 ing on the stage of development, this conclusion 

 may be valid only for prespawning fish. 



In addition to the 50 pairs of ovaries collected by 

 us, we estimated the fecundity of 14 fish (lengths 

 29-46 cm, ages 2-6 yr) from the southern New 

 England stock collected in 1976 by the Northeast 

 Fisheries Center, National Marine Fisheries Ser- 

 vice, NOAA, Woods Hole, Mass. The regression 

 lines for fecundity vs. length and fecundity vs. age 

 for these fish were not significantly different 

 (P>0.25) from our regressions when compared 



5n 



CO 



o 



O 4 



LU 



o 

 to 



3- 



F = . 9857 L 3 95S HOWEL L B KESL ER 

 r - . 885 



—i — i— i — | — i — i — i—i — | i i — i—i — | — i — i — i — i— | — i— i — i i | — i— i — n — pi 

 25 30 35 40 45 50 55 



TOTAL LENGTH (cm) 



F = 240,700A ! 294 (Figure 2) (2) 



r = 0.812, r 2 = 0.659 



FIGURE 1. — Yellowtail fecundity plotted against length. Solid 

 line is the fitted curve for the southern New England population, 

 and the dashed line that of the Grand Bank population. 



F = 62,150G 0678 (Figure 3) (3) 



r = 0.941, r 2 = 0.885 



were F, L, A, and G are fecundity (10 6 eggs/ 

 female), length (centimeters), age (years), and 

 gonad weight (grams), respectively. In all equa- 

 tions the slopes were significantly different from 

 zero (PO.001). 



The coefficient of determination for Equation (3) 

 shows that 88.5% of the variation in fecundity was 

 related to gonad weight independent of both 

 length and age. This was more than the variation 

 related to length alone (78.4%, Equation (D) or 

 age alone (65.9% , Equation (2)). Furthermore, the 

 correlation coefficient for fecundity vs. gonad 

 weight was significantly higher than that for 

 fecundity vs. length (t = 3.85, df = 47, P <0.001), 



5-i 

 CO 



O 4- 



o 



3- 



r =812 



F » 2O550 A ' 



PITT (1971) 



10 



"I 

 12 



AGE (YR) 



FIGURE 2. — Yellowtail fecundity plotted against age. Solid line 

 is the fitted curve for the southern New England population, and 

 the dashed line that of the Grand Bank population. 



878 



