Stevens et al.: Radiometric validation of age, growth, and longevity of Sebastes melanostomus 



715 



The most consistent axis in the otolith section for 

 which confident interpretations could be made was 

 along either the sulcus ridge, or along the dorsoventral 

 margin. Final age estimates were resolved for 197 fish, 

 or approximately 76% of the 260 successfully sectioned 

 otoliths. Agreement among readers was relatively low: 

 approximately 24% of age estimates were within ±1 

 year, 61% were within ±5 years, and 87% were within 

 ±10 years. The mean difference in age estimates be- 

 tween readers was 2.9 ±4.0 years. Among the three 

 readers, APE was 10.7%, D was 8.4%, and CV was 

 14.6%. Average percent error, D, and CV estimates were 

 comparable within readers; reader 1 APE was 5.2%, D 

 was 4.1%, and CV was 7.0%. The two oldest fish to be 

 aged were a 90-year-old male (450 mm TL) collected in 

 1999 and an 87-year-old female (546 mm TL) collected 

 in 1985. Both individuals were caught south of Point 

 Conception, California. 



The VBGF fitted to age and length data resulted in 

 distinct growth curves for male and female blackgill 

 rockfish (Fig. 2). This difference is also represented 

 by non-overlapping confidence intervals with respect 

 to the primary VBGF parameters (L M , k; Table 1). The 

 growth coefficient, k, ranged from 0.040 (±0.007, fe- 

 male) to 0.068 (±0.010, male), and asymptotic length 

 was 448 ±14 mm for males to 548 ±28 mm for females. 

 The asymptotic length for females was 32 mm less than 

 the largest female fish sampled (580 mm TL), and for 

 males, was 74 mm less than the largest male sampled 

 (522 mm TL). The fit for all three functions was satis- 

 factory (r 2 =0.81, 0.87; Table 1, Fig. 2). Estimated ages 

 at first, 50%, and 100% maturity, derived from insert- 

 ing published estimates of length-at-maturity (Echever- 

 ria, 1987) into the growth model for each sex, were 15, 

 21, and 22 years for females and 13, 17, and 28 years 

 for males (Table 2). 



Age prediction, age group determination, 

 and core extraction 



A paired sample r-test indicated that there was a sig- 

 nificant difference between male and female average 

 otolith weight (f=4.54, P<0.001), and a student's r-test 

 for slopes indicated a significant difference between male 

 and female average otolith weight-to-age regressions 

 (r m , = 1.967, ?=2.87, P<0.05). Therefore, male and female 

 age estimates and regressions were treated separately. 

 There was no statistical difference between regres- 

 sions involving fish length and average otolith weight 

 (Kruskal-Wallis one-way ANOVA on ranks, #=4.834, 

 P=0.089). A simple linear regression, with average oto- 

 lith weight as the independent variable and estimated 



