716 



Fishery Bulletin 102(4) 



age as the dependent variable, was sufficient to pre- 

 dict age. Log normalizing the regressions to stabilize 

 the variance in older age estimates was unsuccessful 

 (Cochran's test: «=0.05, 36 df, C=0.4748, P=0.486). The 

 final regressions are given in Figure 3. 



Fourteen age groups based on the predicted ages of 

 unsectioned otoliths were chosen. These groups consist- 



700 



600 



500 



400 



■2> 300 



200 



100 



o males 

 X females 



female 



Female age = 548(1 - e -°<*°('*««») t ^=181 (/=0.87) 

 Male age = 448(1 - e" 0068 !'* 237 !), n=167 (/=0.87) 



10 20 30 40 50 60 



Otolith section age (yr) 



70 



80 



90 



100 



Figure 2 



Blackgill rockfish (Sebastes melanostomus) von Bertalanffy growth functions 

 plotted for males and females. Observed and expected values, as well as the 

 parameters of the equations, are given. Note that the same juvenile samples 

 (/; = 16) were included in both male and female equations. 



ed of four juvenile groups, and five male and five female 

 adult groups I Table 3). Fish lengths ranged from 82 mm 

 to 580 mm TL, and predicted age ranged from 1 to 69 

 years. The number of otolith cores per age group ranged 

 from 11 to 59, representing 7 to 32 fish per group. Total 

 sample weight for each age group ranged from 0.4649 g 

 to 1.6424 g. Whole otolith weight ranged from 0.041 

 to 0.842 g, and average individ- 

 ual core weight for the adult age 

 groups ranged from 0.025 g to 

 0.028 g. The process of extracting 

 the core inadvertently destroyed 

 some otoliths in the grinding pro- 

 cess, leading to smaller samples 

 for some groups. 



Radiometric analysis 



Radiometric analysis of all age 

 groups (n = 14) resulted in the 

 successful determination of 210 Pb 

 activity for all samples, and lim- 

 ited success for 226 Ra (Table 4). 

 Activities of 210 Pb increased, as 

 expected, fivefold from juvenile to 

 adult age groups, and ranged from 

 near 0.01 dpm/g for the juvenile 

 samples to over 0.05 dpm/g for the 

 oldest age groups. Error associ- 

 ated with these measurements 

 ranged from 3.7 to 9.2 % (Is). The 

 detection of 226 Ra activity was 

 met with some technical difficul- 

 ties. Because of poor radium 

 recovery, radium measure- 

 ments were unreliable in 

 three samples and radium 

 was lost in four samples. 

 Therefore, an average of the 

 reliable 226 Ra measurements 

 was used because of the rel- 

 ative consistency of levels 

 measured in these samples 

 (0.0643 ±0.0035 dpm/g, 

 n=l). The use of a single 

 estimate for 226 Ra activ- 

 ity was acceptable prior to 

 refinement of the technique 

 (Andrews et al., 1999b). Cal- 

 culated 210 Pb: 226 Ra ratios 

 increased as expected from 

 0.172 to 0.845 and 0.912 for 

 the oldest groups (Table 4). 



Age estimate accuracy 



Radiometric ages were in 

 agreement with predicted 

 ages, as evidenced by concor- 

 dance of 210 Pb: 226 Ra activity 



