Table 4. — Predicted age in days with 70% and 95% confidence 

 limits of larval and juvenile haddock for a given standard length. 



FISHERY BULLETIN: VOL. 86, NO 2 



Table 5. — Predicted age in days with 70% and 95% confidence 

 limits of larval and juvenile Atlantic cod for a given standard length. 



Y = 11.875 + 112.654Xi, r = 0.9129 



(8) 



DISCUSSION 



should be used for the larval and postlarval pe- 

 riod. 



The equation for the estimated ages of larval 

 and juvenile Atlantic cod (Table 7) is as follows: 



Y = 48.202 + 8.628Xi - 121.908^2 + 139.733Z3, 



r = 0.9292. (9) 



When using only the sagittal diameter, the fol- 

 lowing relationship should be applied to larvae 

 and postlarvae: 



Y = 19.364 + 89.560Xi, r = 0.8659. (10) 



Unlike the tedious laboratory methods needed for 

 the enumeration of otolith increments, gross 

 measurements on a limited number of juvenile 

 otoliths could be performed at sea. 



Despite the tedious methodology required for 

 enumerating daily growth increments in larval 

 and juvenile otoliths, the present work suggests 

 that it is feasible to construct age-length keys for 

 Atlantic cod and haddock similar to those com- 

 monly applied in adult population studies (Clark 

 et al. 1982). The value of such growth data is 

 based on several assumptions, however. Since all 

 conclusions drawn depend upon it, reasonable as- 

 surance of the day-increment relationship in the 

 species being investigated is critical (Beamish 

 and McFarlane 1987; Geffen 1987). Confidence in 

 the growth models generated here for Atlantic 

 cod and haddock may be found in the following 

 inferences: 1) the predicted hatch lengths of 3.32 

 mm for haddock and 4.02 mm for Atlantic cod fall 

 within known limits; 2) a high correlation for the 

 length-at-age data with the rearing experiments 

 of Laurence et al. (1981); and 3) the smooth- 



232 



