FISHERY BULLETIN: VOL. 87, NO. 3, 1989 



o 



z 



MAY 



Figure 1. — Mean lengths adjusted for age of northern anchovy spawned in 

 1980-84. For each year, fish are grouped by back-calculated date of hatching, 

 and length at capture is adjusted to mean age of 208 days. 



1984 were intermediate between those of 1980- 

 82 and 1983. 



Fish hatched during 1983 did not gi'ow more 

 slowly throughout the entire larval and juvenile 

 stage than during other years. Otolith increment 

 widths from 104 juveniles hatched in March 1980 

 were compared with otolith increment widths 

 from 99 juveniles hatched in March 1983 (Fig. 2). 

 In fish smaller than 40 mm in length, increment 

 width was not different. Above 40 mm, the oto- 



lith increment widths were smaller in the 1983 

 sample than in the 1980 sample (P < 0.01); the 

 95% confidence limits for the two cohorts did not 

 overlap. These results show that the difference 

 in lengths of the two cohorts indicated by the 

 analysis of covariance is largely due to difference 

 in growth rate after a size of 40 mm, rather than 

 to a reduction of growth throughout life. 



Back-calculated lengths and gi'owth rates of 

 1980 northern anchovy juveniles were greater 



(A 



C 



o 

 u 



E 



a 



i 



I- 



z 

 111 



s 



LU 

 CC 

 O 



z 



10 



30 40 50 



LENGTH (mm) 



Figure 2. — Average otolith increment width versus length of 104 northern 

 anchovy spawned in March 1980 and of 99 anchovy spawned in March 1983. 

 Shaded areas are 95% confidence intervals. 



648 



