296 



Fishery Bulletin 92(2). 1994 



Results 



ANOVA with contrasts showed the average age gen- 

 erated by the three age readers within each age 

 category were all pair-wise significantly different 

 (P<0.001). Reader 3 estimated the youngest ages, 

 followed by Reader 1, and Reader 2 produced the 

 highest ages (Table 1). Age categories were based on 

 ages from Reader 1; therefore, the age ranges in 

 each category for Readers 2 and 3 were greater. 



The reagent blanks plus background for Po-210 

 and Po-209 did not contain significant activity above 

 the background alone (P>0.05). The background 

 activity for Po-209 and Po-210 ranged up to 

 6.413X10- 4 ±1.472xl0" 4 cpm (counts per minute) 

 and 5.600x10^ ±9.75xl0- 5 , respectively (Table 2). 

 Therefore, the unadjusted sample data (including 

 background counts) were reduced by the appropri- 

 ate background only; no adjustment was made for 

 the reagent blank. Specific activity of Pb-210 at the 

 time of separation from Ra-226 ranged from 0.037 

 ±0.007 dpm/g in category 1 to 0.265 ±0.041 dpm/g 

 in category 4 (Table 3). The Pb-210 activity levels 

 were significantly different among the four age cat- 

 egories (P<0.001). 



The average background count (n=66) for the Rn- 

 222 system was 8.500xl0" 3 ±5.27x10^ cpm. The 

 mean reagent blank plus background for Rn-222 

 was not significantly greater than the background 

 alone (P=0.077). Mean Ra-226 activity (n=4) was 

 0.414 ±0.050 dpm/g (Table 3). Ra-226 values from 

 the four age categories were significantly different 



(P<0.001). Therefore, the Ra-226 measurements 

 were specific to each age category. 



The adjusted radiometric ages (from category spe- 

 cific Ra-226 measurements) for age categories 1 to 

 4 respectively were -0.09, 5.21, 17.83, and 22.66 

 years. Although the adjusted radiometric ages were 

 consistently younger than the burnt cross-section 

 ages from Reader 1, there was general agreement 

 between the two methods (Fig. 2). A r>test between 

 the slope of a line fit through the origin vs. the 45° 

 line did not show a significant difference for any of 

 the age readers (P>0.05). 



Discussion 



Principal findings 



The principal result of this study is that the radio- 

 metric ages generally confirmed the burnt otolith 

 ageing criteria that are used to age sablefish by U.S. 

 and Canadian age readers. A factor which facilitated 

 this confirmation is that sablefish otoliths appar- 

 ently accumulate higher levels of radioisotopes than 

 do other fish species that have been previously stud- 

 ied. We found the specific activities of Pb-210 and 

 Ra-226 in sablefish otoliths (Table 3) to be a full 

 order of magnitude greater than values reported in 

 other species (Bennett et al., 1982; Campana et al., 

 1990; Fenton et al., 1991). These large differences 

 in activity levels may be explained by biological and 

 environmental considerations. Radium-226 is incor- 



