886 



Fishery Bulletin 92(4), 1994 



the populations in the two systems accounted for 90 

 to 94% of the observed variance. The additional vari- 

 ance explained by assuming a difference between the 

 growth curves for each sex in each system improved 

 the fit to the four data sets by 0.3 to 0.6%. The growth 

 curves estimated for males from lengths at age and 

 from back-calculated lengths in Wilson Inlet differed 

 significantly from those estimated for males in Swan 

 Estuary using the corresponding types of data; the 

 same was true for females (P<0.001 ). 



The percentage of the variance explained by the 

 common curves derived from lengths at age and back- 

 calculated lengths for each sex in each system ranged 

 from 81% for males in the Swan Estuary to 94% for 

 both males and females in Wilson Inlet. The percent- 

 age of the variance explained by assuming that the 

 growth curves determined from lengths at age and 



back-calculated lengths are different was improved 

 by 0.8 and 0.2% respectively for males and females 

 from Wilson Inlet and by 4.2 and 1.6% respectively 

 for males and females from the Swan Estuary. The 

 length at age and back-calculated growth curves for 

 males in Wilson Inlet and Swan Estuary differed sig- 

 nificantly; the same applied for females (P<0.001). 



Discussion 



The present study of the lapilli of C. macrocephalus 

 in Wilson Inlet is the first to demonstrate in a plotosid 

 that each of the otolith's first four translucent zones, 

 and probably all other translucent zones, are formed 

 annually. Furthermore, re-analysis of the data of Nel 

 et al. ( 1985) has shown that this also applies to the 



