RALSTON and WILLIAMS: AGEING OF TROPICAL FISHES 



incremental zone with its adjacent discontinuous 

 zone. When counting daily increments, we enumer- 

 ated the dark discontinuous zones. 



From the 94 sectioned otoliths examined for the 

 presence of g^rowth increments (Fig. 2), a total of 

 852 determinations of otolith growth rate (i.e., in- 

 crement width) were completed. Note that no incre- 

 ment width data were collected at otolith lengths 

 in excess of 7,500 ^im, although otoliths as long as 

 9,594 pim were measured and used in the regression 

 analysis of log(FL) on log(OL) (see Figure 1). Beyond 

 7,500 i^m (corresponding to 329 mm FL), the pat- 

 tern of otolith growth became increasingly irregular, 



and clearly distinguishable daily increments, com- 

 posed of well-defined incremental and discontinuous 

 zones, were difficult to resolve. 



The data show that as otolith length increased the 

 growth rate of the otolith declined (Fig. 3). No de- 

 tectable difference in the relationship between 

 otolith growth rate and otolith length could be at- 

 tributed to sex. A partitioned analysis of covariance 

 of the log-transformed data (Table 1) failed to reveal 

 differences in either the slopes or adjusted means 

 of males and females. Thus, data for the two sexes 

 were combined. 



The mean growth rate of the otolith d(OL)/d<, 



E 

 a. 





o 



k- 

 o 



"o 

 O 



1000 2000 3000 4000 5000 6000 7000 8000 



Otolith Length (jim) 



Figure 3.— The relationship between otolith growth rate and otolith length plotted for males, females, 

 and specimens of unknown sex. 



Table 1 .— Partloned analysis of covariance of log-transformed otolith growtti rate 

 d(OL)ldt. Otolith length (OL) was used as the covariate and sex was the treat- 

 ment variable. The data were divided into two separate linear partitions: data for 

 which OL <3,700 ^im and OL >3.700 ^im. 



